ROLE OF INTERCELLULAR ADHESION MOLECULE 2 (ICAM-2) IN THE MURINE IMMUNE SYSTEM by CARMINE CARPENITO B . S c , The University of British Columbia, 1987 M . S c , The University of British Columbia, 1990 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE D E G R E E OF DOCTOR OF PHILOSOPHY in THE FACULTY OF G R A D U A T E STUDIES (Department of Medical Genetics, Genetics Program) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June 1997 © Carmine Carpenito, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of /%P?fpL- &faJGTtCi (^^J,r% f^OG/^ The University of British Columbia Vancouver, Canada Date TYj<0£ 2^)h 71 DE-6 (2/88) ABSTRACT Intercellular adhesion molecule-2 (ICAM-2; CD102) is one of three ligands for the p2 leukocyte integrin LFA-1 (CD11a/CD18). Although ICAM-2 expression is limited to lymphocytes, monocytes, granulocytes, and endothelium, the biological role of ICAM-2 has remained unknown. In this thesis, the murine ICAM-2 cDNA was cloned in order to assist the functional investigation. Sequence analysis of both the cDNA and the genomic clone revealed that ICAM-2 is a member of the immunoglobulin superfamily. The cDNA and antibody were used to examine the role of ICAM-2 in T cell activation and leukocyte transendothelial migration. In order to examine the role of ICAM-2 in antigen presentation to T cells, murine fibroblastic L cells expressing allogeneic class II MHC (l-E d) were transfected with the ICAM-2 cDNA and tested for the ability to stimulate splenic T cells. The expression of ICAM-2 significantly increased the stimulation of T cells in an LFA-1-dependent manner. The increased T cell response was also observed when the class II MHC and ICAM-2 were expressed in separate cells combined together. This indicated that ICAM-2 is actually transmitting a costimulatory signal rather than merely enhancing T cell adhesion to the antigen presenting cell. T cells stimulated with ICAM-2-transfected L cells expressing the class II MHC were able to respond to an allogeneic secondary stimulation. In contrast, T cells stimulated with L cells expressing only the class II MHC were not able to respond to allogeneic stimulation in the secondary response. These results indicated that ICAM-2 may provide a necessary costimulatory signal to the T cell that is required for the aversion of an anergic state. Endothelial cells transfected with the ICAM-2 cDNA were examined for the ability to assist leukocyte migration. A system was set up in which transendothelial migration could be easily quantitated. It was found that a lymphocytic cell line and bone marrow neutrophils were able to utilize ICAM-2 for the migratory process without destroying the endothelial monolayer. These results demonstrate that ICAM-2 is able to play a role in two physiological processes that are of central importance to the normal function of the immune system. iii TABLE OF CONTENTS Abstract ii List of tables v List of figures vi List of abbreviations viii Acknowledgments x Chapter 1: Introduction 1:1 Cellular adhesion 1 1:1.1 Discovery of adhesion molecules 4 1:1.2 Adhesion molecules in the immune system 4 1:2 Leukocyte integrins 5 1:2.1 Nomenclature 6 1:2.2 Tissue distribution 7 1:2.3 Structure and biosynthesis 9 1:2.4 Chromosomal location of leukocyte integrin genes 11 1:3 Leukocyte adhesion deficiency (LAD) 11 1:4 Identification of LFA-1 counter-receptors 12 1:4.1 Intercellular adhesion molecule 1 (ICAM-1) 13 1:4.2 Intercellular adhesion molecule 2 (ICAM-2) 23 1:4.3 Intercellular adhesion molecule 3 (ICAM-3) 27 1:5 Thesis objectives 29 1:6 References 30 Chapter 2: Cloning and characterization of murine ICAM-2 2:1 Introduction 46 2:2 Materials and Methods 49 2:3 Results 67 2:4 Discussion 86 2:5 References 96 Chapter 3: Costimulatory role of ICAM-2 in T cell response to allogeneic class II M H C 3:1 Introduction 104 3:2 Materials and Methods 109 3:3 Results 112 3:4 Discussion 128 3:5 References 134 a Chapter 4: Role of ICAM-2 in leukocyte transendothelial migration 4:1 Introduction 140 4:2 Materials and Methods 143 4:3 Results 149 4:4 Discussion 163 4:5 References 170 Chapter 5: Summary and Discussion 5:1 Conclusion and future directions 176 5:2 References 184 iv LIST OF TABLES T A B L E 1 Various members of the immunoglobulin, integrin, and selectin families involved in leukocyte function 3 T A B L E 2 Red bead fluorescence does not interfere with yellow cell fluorescence 162 v LIST OF FIGURES Figure 1 Comparison of human and murine ICAM sequences 68 Figure 2 Alignment of protein sequences from the P C R clone with human ICAM-2 70 Figure 3 Nucleic acid and protein sequence of the murine ICAM-2 cDNA (G3-1.1) 72 Figure 4 Comparison of amino acid sequence between cDNA-encoded (G3-1.1) protein and human ICAM-2 73 Figure 5 Expression of murine ICAM-2 by Northern blot analysis 75 Figure 6 Genomic Southern blot analysis of murine ICAM-2 76 Figure 7 Nucleotide sequence of the murine ICAM-2 genomic clone 78 Figure 8 Southern blot analysis of murine ICAM-2 genomic clone (BM 1-1.1) 79 Figure 9 Restriction map of the murine ICAM-2 genomic ICAM-2 clone 80 Figure 10 Immunopurification of ICAM-2 81 Figure 11 Adhesion of murine splenic T cells to purified ICAM-1 and ICAM-2 83 Figure 12 Adhesion of splenic T cells to L cells transfected with ICAM-2 84 Figure 13 Flow cytometric analysis of RT10.3 cells 113 Figure 14 Adhesion of splenic T cells to RT10.3 cells expressing ICAM-1 and ICAM-2 116 Figure 15 Schematic representation of T cell stimulation by allogeneic MHC class II and ICAM-2 117 Figure 16 Dose response of splenic T cells to RT10.3 cells expressing ICAM-1 or ICAM-2 119 Figure 17 Kinetics of the allogeneic T cell response to ICAM-1- or ICAM-2-transfected RT10.3 cells 120 Figure 18 Effects of antibodies on allogeneic T cell response to RT10.3 cells 121 Figure 19 Schematic representation of T cell response to mixed stimulators 123 Figure 20 Stimulation of T cells with RT10.3 cells and ICAM-2-transfected L cells 124 Figure 21 Schematic representation of secondary stimulation of T cells 126 vi Figure 22 Secondary responses of T cells 127 Figure 23 Flow cytometric analysis of SVEC4.10 cells 150 Figure 24 Immunoprecipitation of ICAM-1 and ICAM-2 from transfected SVEC4.10 cells 151 Figure 25 Kinetics of bead diffusion across S V E C monolayer 153 Figure 26 Ability of TIL1 cells to utilize ICAM-1 and ICAM-2 for migration across the SVEC4.10 monolayer 154 Figure 27 SVEC4.10 cells expressing ICAM-1 or ICAM-2 are able to mediate TIL1 transendothelial migration in an LFA-1-dependent manner 156 Figure 28 Expression of adhesion molecules on mouse bone marrow neutrophils 157 Figure 29 Neutrophil binding to SVEC4.10 monolayers 158 Figure 30 Neutrophil migration across SVEC4.10 monolayers 160 Figure 31 Antibody blocking of neutrophil migration across SVEC4.10 monolayers 161 Figure 32 Simultaneous assessment of neutrophil migration and SVEC4.10 monolayer permeability 164 vii LIST OF ABBREVIATIONS Ag anitgen A P C antigen-presenting cell bp base pair B S A bovine serum albumin C A M cell adhesion molecule C D cluster of differentiation cDNA complementary DNA C M V cytomegalovirus cpm counts per minute CTL cytolytic T lymphocyte D M E M Dulbecco's modified minimum essential medium DNA deoxyribonucleic acid dNTP deoxynucleotidetriphosphate DTT dithiothreitol E C M extracellular matrix EDTA ethylenediamine tetraacetic acid Fab antigen-binding fragment EtBr ethidium bromide F A C S fluorescence-activated cell sorter F C S fetal calf serum FITC fluorescein isothiocyanate fMLP f-Met-Leu-Phe H B S S Hank's balanced salt solution HEV high endothelial venule hr hour HSA heat stable antigen H U V E C human umbilical vein endothelial cell ICAM intercellular adhesion molecule IFN interferon ig immunoglobulin IL interleukin IPTG isopropylthio-p-D-galactoside kb kilobase kD kilodalton LAD leukocyte adhesion deficiency LFA-1 leukocyte function-associated antigen-1 L P S lipopolysaccharide mAb monoclonal antibody Mac-1 macrophage antigen-1 Mad mucosal addressin MALA-2 murine activation lymphocyte antigen-2 2-ME 2-mercaptoethanol M H C major histocompatibility complex min minute MLR mixed lymphocyte reaction M O P S 3-[N-Morpholino]-propane-sulfonic acid viii Mr molecular mass mRNA messenger RNA N C A M neural cell adhesion molecule NK natural killer cell NP-40 nonidet p-40 O R F open reading frame O V A ovalbumin P A G E polyacrylamide gel electrophoresis PBL peripheral blood lymphocytes P B S phosphate buffered saline P C R polymerase chain reaction P E G polyethylene glycol PHA phytohaemagglutinin P M A phorbol 12-myristate 13-acetate P M S F phenylmethylsulfonyl fluoride PSGL-1 P-selectin glycoprotein ligand-1 R B C red blood cell R G D Arg-Gly-Asp RNA ribonucleic acid S D S sodium dodecyl sulphate S E M standard error of the mean S S C saline sodium citrate buffer S S P E saline sodium phosphate EDTA buffer TAE Tris acetate EDTA buffer T c cytolytic T lymphocyte TcR T cell receptor [3H]-TdR [methyl-3H]-thymidine deoxyribose T h T helper lymphocyte TE Tris EDTA buffer TNE Tris sodium chloride EDTA buffer Tris tris(hydroxymethyl)aminomethane TWEEN-20 polyoxyethylene-sorbitan monolaurate UTR untranslated region UV ultraviolet V C A M vascular adhesion molecule V L A very late antigen X-gal 5-bromo-4-chloro-3-indolyl-p-D-galactoside ix ACKNOWLEDGMENTS I would like to acknowledge the input of numerous people who have contributed to this thesis. Starting with my supervisor, Dr. Fumio Takei, for his guidance and support in leading this sometimes difficult project through the dark moments. I also thank my advisory committee, Dr. Graeme Dougherty, Dr. Keith Humphries, Dr. Dixie Mager, and Dr. Ann Rose, for critical comments and discussion. In addition, I would like to express my gratitude to Andrew Pyszniak, Mike Ohh, and Dean Thacker for their support and friendship, to Blythe Miyagawa for the work in cloning the genomic ICAM-2, and every member of the TFL who has ever answered any of my endless and sometimes thoughtless questions. x C H A P T E R 1 INTRODUCTION 1:1 C e l l u l a r A d h e s i o n The ability of cel ls to adhere both to one another and to extracel lular matrix components is an essent ia l requirement for maintaining organ structure and integrity as well as function. Ce l l surface molecules mediate these adhes ive interactions and as such play critical roles in development, t issue organizat ion, cell migration and function in multicellular organisms (Hynes, 1987; Ede lman , 1989; Turner, 1992). The immune sys tem is compr ised of a multiplicity of cel ls spec ia l ized in the recognit ion and presentat ion of non-self ant igens. In contrast to the rather permanent adhes ion observed in sol id organs, the cel ls of the immune system (leukocytes) only temporari ly adhere to their targets (Dustin and Springer, 1989; van Kooyk et al., 1989). This transient adhes ion is regulated by the activation state of the leukocytes. The express ion profile of cell adhes ion molecules ( C A M s ) on leukocytes can a lso determine which counter-receptors on opposing cel ls are uti l ized, thus affecting the extent to which the appropriate immune response is del ivered. Molecular character izat ion of C A M s expressed by leukocytes as well as non-hematopoiet ic cel ls has greatly advanced the understanding of the immune response. A s the list of C A M s participating in var ious aspects of the immune response expands , it is apparent that these molecu les share structural similarity. T h e s e structurally and functionally related adhes ion receptors of the immune system are grouped into three famil ies ( T A B L E 1). 1 Collect ively, these molecules al low cells of the immune sys tem to function in leukocyte recirculation, inf lammation, and ant igen-specif ic removal of pathogens by mediating adhes ion between leukocytes and the appropriate targets (Dustin and Spr inger, 1991; Bev i lacqua , 1993; Car los and Har lan, 1994). This thesis is concerned with intercellular adhes ion molecule-2 ( ICAM-2), a member of one of these protein famil ies, and its potential role in antigen presentation as well as leukocyte transendothel ial migration. 2 T A B L E 1 Var ious members of the immunoglobul in, intergih, and selectin families involved in leukocyte function Immunoglobulin (Ig) Superfamily: common feature is varying numbers of an Ig-like domain, which consists of 100-110 amino acids arranged in two anti-parallel p-sheets stabilized together by an intradomain disulphide bond. Mediates leukocyte adhesion to hematopoietic and non-hematopoietic cells. Important in antigen recognition and leukocyte recirculation. ICAM subfamily ICAM-1 ICAM-2 ICAM-3 Expression lymphoid cells, myeloid cells, fibroblasts, endothelium, keratinocytes endothelium, lymphoid cells, myeloid cells, platelets lymphoid cells, myeloid cells Counter-receptors and ligands LFA-1, Mac-1, CD43, fibrinogen LFA-1 LFA-1, a d p 2 Leukocyte Integrin (p 2 ) Family: common feature is a heterodimeric structure with one of four a-chains non-covalently associated with pychain. Expression is restricted to leukocytes. Binding capacity controlled by activation state of leukocyte playing critical role in inflammation and antigen recognition. Expression Counter-receptors and ligands LFA-1 lymphoid cells, myeloid cells ICAM-1,-2,-3 Mac-1 lymphoid cells, myeloid cells ICAM-1, fibrinogen p150/95 lymphoid cells, myeloid cells fibrinogen a d p 2 lymphoid cells, myeloid cells ICAM-3 Selectin Family: common features include N-terminal lectin-like domain, an epidermal growth factor repeat, followed by a variable number of modules similar to complement binding proteins. Function primarily in leukocyte trafficking and inflammation. Expression E-selectin endothelium L-selectin lymphoid cells, myeloid cells P-selectin endothelium, platelets Counter-receptors and ligands sialyl Lewis*3, E-selectin sialyl Lewis*'3, GlyCAM-1, CD34, MadCAM-1 sialyl Lewis"3, PSGL-1 1:1.1 Discovery of adhesion molecules The ex is tence of C A M s was postulated before their actual d iscovery based on the reasoning that a mechan ism must exist to fix cel ls in posit ion giving rise to organ structure and thus spec ia l ized function (Edelman, 1976; Ede lman , 1977). The initial approach employed to identify cell adhes ion molecules involved attempting to disrupt cel lular adhes ion to adjacent cel ls with antibodies (Muller and Ger i sch , 1978). It was observed that s l ime mold and neural cell aggregates could be disrupted with Fab fragments of IgG purified from polyclonal ant isera raised against the respect ive cell type. Th is disruption could then be reversed with the addit ion of var ious fractions of solubi l ized membrane proteins that al lowed cellular adhes ion to recur. Th is provided key ev idence that cell adhes ion was mediated by speci f ic membrane-bound molecu les. The next step was to examine what other cell types also exp ressed C A M s and to identify the individual C A M s responsible for the adhes ion as wel l as their poss ib le function. In the case of the immune sys tem, C A M s have been identified by monoc lona l ant ibodies (mAb) recognizing the speci f ic molecules (Kohler and Milstein, 1975). 1:1.2 Adhesion molecules in the immune system The ability of leukocytes to recognize and eliminate foreign invaders is crucial to the function of the immune system. W h e n a T lymphocyte encounters an antigen present ing cell ( A P C ) or a target cel l , there is an initial weak adhes ion between the two cel ls (Shortman and Goldste in , 1979). If the T cell receptor (TcR) recogn izes the 4 appropriate ant igen, then the T cell will elicit its helper or cytolytic function. If the T cell does not recognize the appropriate antigen, it will detach and continue circulating throughout the body. In the immune sys tem, severa l l ines of ev idence supported the notion that C A M s were involved in immune function. S ince T c R recognit ion of an t i gen :MHC was thought to provide very weak adhes ion , other molecu les were postulated to mediate the actual adhes ion (Martz, 1975; Shor tman and Golds te in , 1979; Martz, 1980). Ant igen- independent adhes ion of cytolytic T lymphocytes (CTL) to target cel ls provided the first experimental ev idence that suggested immune cell adhes ion involved accessory molecules (Martz, 1975; Zagury et al., 1975). This non-specif icity of conjugate formation between C T L s and target cel ls lacking the cognate ant igen lead researchers to attempt to identify the molecules involved in the adhes ion . 1:2 Leukocyte integrins The first adhes ion molecules d iscovered to play a role in the immune sys tem were members of the integrin superfamily. These molecules are heterodimeric cell sur face molecu les, each mediating specif ic interactions with other cel ls or extracel lular matrix components (Hynes, 1987). The integrin superfamily is divided into subfami l ies based on one of three commonly shared p subunits (p^ p2, p3); al though the p subunits are distinct, they are nonetheless structurally similar. B e c a u s e of the leukocyte-restricted express ion of the p2 leukocyte integrins (Larson and Springer, 1990), this subfamily is involved solely in leukocyte function. Members of the p2 integrins share a common p chain which assoc ia tes noncovalently with unique a cha ins (Hogg, 1989). 5 The p 2 chain (CD18) can assoc ia te with either the oc L chain (CD11a) , a M chain (CD11b) , a x chain (CD11c) , or a d forming L F A - 1 , M a c - 1 , p150/95, o r a d p 2 , respect ively (Corbi etal., 1987; Corbi etal., 1988a; Larson etal., 1989; Dani lenko e r a / . , 1995; V a n der V ie ren et al., 1995). The p 2 leukocyte integrins are the most extensively studied family of adhes ion molecu les functioning in the immune sys tem. Al though Mac-1 was the first p 2 integrin to be recognized (Springer et al., 1979), it w a s used primarily as a marker for myeloid cel ls. The first p 2 leukocyte integrin to be character ized functionally was the lymphocyte funct ion-associated antigen-1 ( L F A - 1 , C D 1 1 a / C D 1 8 ) . It was first def ined by a mAb screened for the ability to interfere with C T L killing of target cel ls and T cell proliferation (Davignon et al., 1981; Kurz inger et al., 1981; Pierres et al., 1982). More intense examinat ion indicated that the A b blockage occurred at the stage of conjugate formation (between the T cell and the target cell) rather than the actual function elicited (Krensky et al., 1984; Spi ts et al., 1986). In contrast, ant ibodies against the T c R are able to block the function of cytolytic T cel ls but not the initial adherence to target cel ls (Kaufman and Berke, 1983; Martz, 1987). This provided further ev idence that there is an initial adhes ion event which p recedes the functional activity of the leukocyte and b lockage of this preliminary step abrogates leukocyte function. 1:2.1 Nomenclature The leukocyte integrins share both structural and functional similarit ies (Hogg, 1989; Larson and Springer, 1990). The commonly shared p 2 subunit has been 6 ass igned to the cluster of differentiation (CD)18 and the a cha ins of L F A - 1 , M a c - 1 , and p150/95 have been ass igned the designat ions C D 1 1 a , C D 1 1 b , and C D 1 1 c , respectively. The recently d iscovered a chain of ad(32 (Dani lenko et al., 1995) has yet to be ass igned a cluster of differentiation. 1:2.2 Tissue Distribution The distribution of LFA-1 is restricted to cel ls of hematopoiet ic origin. LFA-1 is exp ressed by most leukocytes (Kurzinger et al., 1981; Krensky et al., 1983), with the except ion of s o m e murine t issue macrophages (St rassman et al., 1985). LFA-1 is exp ressed on ~ 5 0 % of bone marrow cel ls. It is a lso present on the majority of thymocytes, neutrophils, macrophages, monocytes, and peripheral lymphocytes. L F A -1 express ion is absent or low on myeloid and erythroid precursor cel ls (Miller et al., 1985; C a m p a n a et al., 1986). In B cell and myeloid l ineages, LFA-1 first appears in the pre-B cell and late myeloblast s tages, respectively. However, LFA-1 may be lost from B cel ls during terminal differentiation to p lasma cel ls (Miedema et al., 1985). S i nce most leukocytes express L F A - 1 , it is not surprising that ant ibodies against LFA-1 have been shown to inhibit a multitude of immune functions by blocking adhes ion to counter-receptors on opposing cel ls (Springer et al., 1987; Larson and Springer, 1990; Spr inger, 1990). Essent ia l ly every leukocyte function requiring adhes ive interactions can in s o m e way be abrogated with LFA-1 A b s . The distribution of Mac-1 is more restricted than that of L F A - 1 . Express ion of Mac-1 is conf ined to monocytes, macrophages, granulocytes, large granular 7 lymphocytes, and immature B cel ls (Flotte et al., 1983; Spr inger and Unk less , 1984; de la Hera et al., 1988). Mac-1 is a lso expressed on C D 5 + B cel ls and lymphokine act ivated killer cell precursors. It is a multifunctional receptor that can bind multiple l igands. Initially, an anti-Mac-1 A b was shown to block monocyte and granulocyte binding to erythrocytes coated with C3b i , a complement component (Bel ler et al., 1982). Eventual ly, Mac-1 was demonstrated to bind any cell type coated with C3b i (Rothlein and Springer, 1985; R a m o s et al., 1988). Under certain condit ions, Mac-1 can a lso bind an LFA-1 ligand (Diamond et al., 1991). Mac-1 has a lso been reported to bind the soluble l igand fibrinogen (Wright et al., 1988). T h e s e adhes ive interactions are important in myeloid cell function. Ant i -Mac-1 A b s can block homotypic aggregat ion of neutrophils, granulocyte transendothel ial migration, as well as monocyte and neutrophil chemotaxis and adherence (Anderson et al., 1986; Dana et al., 1986; Wal l is etal., 1986). The p150/95 protein is similar to Mac-1 in distribution. It is a lso found on some activated lymphocytes. Like M a c - 1 , p150/95 can also act as a C3b i receptor (Micklem and S i m , 1985; Malhorta et al. , 1986). It a lso probably functions as a cell adhes ion molecule. Ant ibodies against p150/95 are able to inhibit neutrophil and peripheral blood monocyte binding to endothel ium, phagocytosis, and chemotax is (Anderson et al., 1986; Ke izer et al., 1987a). In some cases , the p150/95 protein can also contribute adhes ive strength in C T L conjugate formation with target cel ls depending on its express ion levels on the activated T cells and lymphoid cell l ines (Keizer et al., 1987b). 8 The a d p 2 molecule has been detected on peripheral blood leukocytes. Human lymphocytes express low levels of this molecule (Van der V ie ren et al., 1995), whereas only a smal l subset of canine C D 8 + T cel ls express a d p 2 (Dani lenko et al., 1995). Human granulocytes a lso express a d p 2 , which can be upregulated from an intracellular pool by f - M L P treatment of the granulocytes. Immunohistological staining of human sp leen has revealed that a d p 2 express ion is local ized to the red pulp cords and s inuses on smal l and large mononuclear cel ls and granulocytes. Express ion of a d p 2 in the white pulp is limited to scattered dendritic cel ls. Funct ional inhibition by antibody blocking has yet to be done. However, the predominant express ion of a d p 2 on spec ia l i zed cel ls in t issues suggests that the major functions of a d p 2 may be restricted to speci f ic microenvironments. 1:2.3 Structure and biosynthesis The two complementary D N A s (cDNA) encoding the p 2 and the a L subunits have been isolated and character ized in both the human (Kishimoto et al., 1987a; Law et al., 1987; Larson etal., 1989) and murine sys tems (Wilson et al., 1989; Kau fmann et al., 1991). T h e s e two c D N A s encode separate subunit precursors of 72 kD (p 2 ) and 130 kD ( a L ) . High mannose N-glycoside carbohydrate groups are then added to these two bare polypeptide backbones (Sastre et al., 1986a). Heterogeneity in the glycosylat ion pattern between cel ls has been observed; N-linked carbohydrates are sulfated only on T cel ls (Dahms and Hart, 1985). In addit ion, sialylation patterns differ between B and T cel ls (Takeda, 1987). In order to further process the glycosylated 9 precursors to the complex type N-linked carbohydrate structures, the p 2 a r | d the a L subuni ts must first assoc ia te intracel lular^ (Ho and Springer, 1983; Spr inger et al., 1984). Th is o l igosacchar ide modification occurs in the Golg i apparatus from which the mature LFA-1 molecu les are transported to the cell surface. The deduced amino acid sequences based on the c D N A s e q u e n c e s indicates that the a L and the p 2 subunits have the c lass ica l features of integral type I membrane proteins (Kishimoto et al., 1987a; Law et al., 1987; Larson et al., 1989). Both have large extracel lular domains with multiple potential N-glycosylat ion si tes, short t ransmembrane and cytoplasmic domains. The p 2 subunit has an unusual ly high cyste ine content (7.4%) concentrated over a four-fold repeat region (20%) spanning 186 amino acid residues near the C-terminus (Kishimoto et al., 1990). It is bel ieved that this g ives the p 2 subunit a rigid tertiary globular structure. The ocL subunit shares striking homology with the ocM and a x subunits (44% in human). A significant feature of a subunits is the three tandem repeats which contain putative divalent cation binding si tes (Kishimoto et al., 1990). T h e s e domains are similar to the C a + + binding " E F - h a n d loop" sites in calmodul in, troponin C , and parvalbumin. T h e s e may be the putative metal-binding sites responsible for the Mg + + - dependen t adhes ion d isp layed by the leukocyte integrins (Rothlein and Springer, 1986). Another key feature of the ot L subunit is the p resence of an I domain which has been specula ted to be directly involved in binding to counter structures on opposing cel ls (Randi and Hogg , 1994; Huang and Spr inger, 1995). 10 1:2.4 Chromosomal location of leukocyte integrin genes The chromosomal localization of the genes encoding the human leukocyte integrins has provided some interesting results. The a subunits a L , a M , and a x were local ized to the s a m e band on chromosome 16 (16p11.1-p13) (Corbi et al., 1988b); the location of the a d subunit has not been mapped. This defined a cluster of genes belonging to the s a m e family of proteins involved in leukocyte adhes ion . Isolation of the genomic c lones of the a subunits has revealed that these genes have ana logous intron/exon organizat ion (Sastre et al., 1986b; Corbi et al., 1990; W o n g et al., 1996). The protein homology, exon/intron boundary similarity, and proximal ch romosomal location sugges ts that the leukocyte integrins arose by gene duplication of an ancestra l a subunit. The human p 2 gene is located on chromosome 21 and not found near other p g e n e s (Corbi etal., 1988b). 1:3 Leukocyte Adhesion Deficiency (LAD) Pat ients with an inherited leukocyte integrin-dependent immunodef ic iency have been identified (Anderson and Springer, 1987). T h e s e patients consistent ly exhibit a defect in adherence-dependent leukocyte functions and suffer from recurrent life-threatening bacterial and fungal infections (Fischer et al., 1988). Symptoms assoc ia ted with this d i sease include recurrent gingivitis, defect ive neutrophil mobility and phagocytos is , lack of pus formation, and absence of lymphocytes and granulocytes in infected lesions despite chronic leukocytosis. The defect in leukocyte function w a s found to be secondary to an abnormality in adhes ion (Crowley et al., 11 1980). Leukocytes from these patients were found to lack express ion of three leukocyte integrins ( L F A - 1 , M a c - 1 , p150/95) (Springer et al., 1984; Arnaout et al., 1984; Beatty et al., 1984). However, levels of cell surface express ion vary between patients and correlates with the severity of the d isease . Moderately deficient patients may survive to adulthood if their recurrent infections are adequately treated while the severe ly deficient patients require a bone marrow transplant to extend life expectancy past early chi ldhood (Fischer et al., 1990). The def ic iency in L F A - 1 , M a c - 1 , and p150/95 express ion (ad(32 not examined) was traced to abnormal express ion of the p 2 subunit (Marlin et al., 1986; Kishimoto e ra / . , 1987b). Normal amounts of these three a chain precursors are detected in patient cel ls (L isowska-Grospier re et al., 1986). However , a b s e n c e of a functional p 2 subunit prevents proper intracellular associat ion of the heterodimers, which is required for complete process ing and transport to the cell sur face. Express ion of these three integrins can be rescued by introduction of the human or murine p 2 c D N A encoding a protein which can assoc ia te with any of the human a chain precursors (Hibbs et al., 1990). The rescued molecu les function in exact ly the s a m e manner that they do on normal leukocytes. 1:4 Identification of LFA-1 counter-receptors Stimulated lymphoid cel ls form homotypic aggregates that are LFA-1 -dependen t (Rothlein and Springer, 1986). Lymphoid cel ls from L A D patients do not form these aggregates s ince these cel ls do not express L F A - 1 . However, when L A D lymphocytes (LFA-1") are combined with normal L F A - 1 + lymphoid cel ls, coaggregates form in an 12 LFA-1 -dependen t manner. This lead investigators to search for LFA-1 counter-receptors on L A D cel ls which eventually lead to the d iscovery of a family of LFA-1 receptors on var ious cell types. 1:4.1 Intercellular Adhesion Molecule-1 (ICAM-1) The observat ion that stimulated lymphocytes were able to adhere to lymphocytes from L A D patients in an LFA-1-spec i f i c manner sugges ted that there exists a structure on L A D cel ls which could interact with LFA-1 (Rothlein and Spr inger, 1986). In order to identify the LFA-1 counter-receptor, mAbs raised against L A D cel ls were sc reened for their ability to block the LFA-1-dependent aggregat ion of st imulated lymphoid cel ls (Rothlein et al., 1986). The first mAb obtained from this procedure recognized a heavily glycosylated protein of ~86-114 kD that was subsequent ly termed intercellular adhes ion molecule 1 ( ICAM-1, CD54) . Murine ICAM-1 was similarly identified using a mAb that inhibited an in vitro mixed lymphocyte reaction (MLR) (Takei, 1985). Unlike L F A - 1 , ICAM-1 express ion is more widely distributed. It is detected on cel ls of both hematopoiet ic and nonhematopoiet ic origin (Dustin et al., 1986). A l though express ion may be very low on nonhematopoiet ic cel ls and resting lymphoid cel ls, ICAM-1 can be greatly upregulated on these cel ls by var ious cytokines (Dustin et al., 1986; Dustin et al., 1988a). This increased ICAM-1 express ion correlates with increased LFA-1-dependent adhes ion of lymphocytes to induced cel ls. ICAM-1 express ion is a lso c losely coordinated with the progression of the immune 13 response indicating that the LFA-1 : ICAM-1 interaction plays very important roles in all aspec ts of the immune sys tem. 1:4.1 a) ICAM-1 cDNA cloning The genes coding for both the human (S immons et al., 1988; Staunton et al., 1988) and murine ICAM-1 (Horley et al., 1989; S iu et al., 1989) have been c loned and s e q u e n c e d . ICAM-1 is a t ransmembrane protein with a polypeptide backbone of 55 kD. Both the human and mouse spec ies have multiple potential g lycosylat ion sites which are extensively util ized. The degree of glycosylat ion can vary depending on the cell type as ev idenced by the s ize variation (85-115 kD) seen in the immunoprecipi tated product (Dustin et al., 1986; Makgoba et al., 1988). The amino acid s e q u e n c e of ICAM-1 shares sufficient homology with other protein s e q u e n c e s and harbors a common repeated structural motif to p lace it in the immunoglobul in supergene family (Wil l iams, 1987; Wi l l iams and Barclay, 1988). Membersh ip within this superfamily is based on the presence of a conserved tertiary protein structure, an Ig-like structural motif. The Ig-like domain consists of approximately 110 amino ac ids which fold in two anti-parallel p-sheets. Five Ig-like domains are present in the extracel lular portion of ICAM-1 fol lowed by short t ransmembrane and cytoplasmic regions (S immons et al., 1988; Staunton et al., 1988; Horley et al., 1989; S iu et al., 1989) . In addit ion to these domains, ICAM-1 also displays overal l structural homology with two unique C A M s , myelin assoc ia ted glycoprotein and neural cell adhes ion molecule (Dustin et al., 1988b), bearing multiple Ig-like domains. The homology 14 between human and murine ICAM-1 is somewhat limited (65% nucleotide homology, 5 0 % amino acid homology). Genom ic cloning has revealed that the five Ig-like domains of ICAM-1 are encoded by separate exons (Voraberger et al., 1991; Degitz et al., 1991). The human ICAM-1 gene has been mapped to ch romosome 19 (Katz et al., 1985) and the murine ICAM-1 has been mapped to ch romosome 9 (Bal lantyne et al., 1991). Identification of ICAM-1 as a ligand for LFA-1 was the first example of an interaction occur ing between superfamil ies functioning in the immune sys tem (Marlin and Spr inger, 1987). 1:4.1 b) T i s s u e distr ibution a n d cy tokine induct ion In contrast to its counter-receptor, L F A - 1 , ICAM-1 express ion is not restricted to cell of the hematopoiet ic l ineage. ICAM-1 has been detected at low levels on cel ls including resting leukocytes, follicular dendritic cel ls, f ibroblasts, vascu lar endothel ium, kerat inocytes, synovial cel ls, and certain epithelial cel ls (Dustin et al., 1986; Pobe r et al., 1986; te V e l d e et al., 1987; Dustin et al., 1988a; Mentzer et al., 1988; Rothlein et al., 1988; Dustin and Springer, 1988a). However, ICAM-1 levels can be dramatical ly increased by var ious cytokines (Dustin et al., 1986; Pober et al., 1986; Rothlein et al., 1988). Th is is most prominent on endothelial cel ls, f ibroblasts, mesenchyma l cel ls, and epithelial cel ls (Dustin etal., 1986; Wantz in etal., 1989; Munro etal., 1989; Dustin and Spr inger, 1991). Cytok ines such as 1 ( 3 , T N F - a , IFN-y, and L P S are pro-inflammatory and cause the upregulation of ICAM-1 on cel ls able to receive the signal mediated by these molecules (Dustin et al., 1986; Rothlein etal., 1988). It is of interest 15 to note that the cytokine receptors which cause ICAM-1 upregulation a lso cause the express ion of other genes . O n e example of this is IFN-y induction of both ICAM-1 and M H C c lass II (O 'Connel l and Edidin, 1990; Farrar and Schreiber , 1993). Th is al lows appl icat ion of T helper (T h) function to appropriate target cel ls s ince both adhes ion between the two cel ls and efficient antigen presentation are required for T h function (Springer, 1990; Dustin and Springer, 1991). Express ion of ICAM-1 a lso plays a signif icant role in inflammation and migration of leukocytes out of the b loodstream and into the si tes of infection (d iscussed in chapter 4) (Bev i lacqua, 1993; Car los and Har lan, 1994). This upregulation usually requires de novo synthesis (Rothlein et al., 1988; Johnson et al., 1989) but in certain instances rapid mobil ization of ICAM-1 from intracellular stores has been observed (Dougherty et al., 1988). ICAM-1 induction has a lso been observed on lymphocytes. Rest ing B and T cel ls lack detectable ICAM-1 express ion (Dustin et al., 1986; Clark et al., 1986; Wawryk et al., 1989). However , express ion is increased during activation of the lymphocytes thus al lowing more efficient interaction with L F A - 1 + cel ls and coordination of the immune response s ince lymphocytes must communicate with each other to ensure a response. 1:4.1 c) Physiological significance of LFA-1 MCAM-1 interaction The formal proof that ICAM-1 was a counter-receptor for LFA-1 c a m e from studies using artificial membranes bearing purified human ICAM-1 protein (Marlin and Spr inger, 1987). ICAM-1 protein incorporated into planar lipid bi layers was able to bind L F A - 1 + cel ls. The adhes ion was specif ical ly inhibited by pretreating the cel ls with 16 anti-LFA-1 mAb or anti-ICAM-1 mAb treatment of the membranes. Such adhesion was found to require divalent cations such as Mg + + or C a + + since the chelating agent EDTA would block the adhesion. This is in agreement with the sequence data of LFA-1 which indicates the presence of a divalent cation binding domain within the extracellular region of the a L subunit (Larson et al., 1989). Energy and functional microfilaments are also required for binding as demonstrated by sodium azide together with 2-deoxy-D-glucose which block energy production and T lymphoblastoid binding to ICAM-1-bearing membranes (Marlin and Springer, 1987). Cell binding is temperature dependent, being maximal at 37°C, reduced at 14°C, and completely inhibited at 4°C. The adhesion is also inhibited by cytochalasin B, a microfilament inhibitor. It has been found that LFA-1 is indeed attached to several cytoskeletal proteins (Burn et al., 1988; Kupfer et al., 1990; Pavalko and LaRoche, 1993). Reciprocal studies using purified LFA-1 incorporated into membranes do not mimic the same binding requirements (Dustin and Springer, 1989; Kishimoto et al., 1990). This indicates that the active, dynamic process is occurring on the LFA-1-bearing cell side of the interaction. Several integrins which act as receptors for extracellular matrix components bind their ligands through Arg-Gly-Asp (RGD) sequences (Ruoslahti and Pierschbacher, 1986; Hynes, 1987). Oligopeptides containing the RGD sequence can block the interaction between these integrins and their ligands. The human ICAM-1 does not contain an RGD sequence, but does contain several RGD-like sequences (Staunton et al., 1988). However, the RGD and RGD-like oligopeptides could not 17 inhibit binding of L F A - 1 + cel ls to ICAM-1 membranes (Marlin and Springer, 1987). This sugges ts that leukocyte integrin binding specificity has diverged from that of other integrins. Leukocytes bearing LFA-1 circulate throughout the body as single unattached cel ls even though they continuously encounter I C A M - 1 + cel ls. It w a s suspec ted and later observed that in order for L F A - 1 + cel ls to adhere to I C A M - 1 + cel ls , the L F A - 1 + cel ls need to be in an activated state (Dustin and Springer, 1989; van Kooyk et al., 1989). Lymphocyte activation by P M A , or antibody triggering through C D 3 or T c R crossl inking results in a strongly increased LFA-1-dependent adhes iveness . The increased adhes ion following C D 3 or T c R antibody crossl inking appeared almost immediately after activation and returned to normal after 30 min. It has been specu la ted that this relatively quick increased adhes ion may be due to a conformational change in LFA-1 exposing a binding site. Ev idence for this hypothesis c o m e s from a m A b directed against an LFA-1 epitope which is only detected when the lymphocyte is activated (Dransfield and Hogg, 1989). Express ion of this epitope d isappears when M g + + is removed, temperature is lowered, or metabol ic inhibitors are present. R e l e a s e of a low molecular weight lipid cal led integrin modulat ing factor (IMF) has been reported to a lso induce increased adhes ion (Hermanosk i -Vosatka et al., 1992) and possib le conformational change. The increased adhes ion may a lso be control led by the redistribution of LFA-1 or ICAM-1 on the cell sur face. LFA-1 local izat ion has been detected at regions of focal contact with ICAM-1 (Kupfer and Singer , 1989a; Kupfer and Singer, 1989b). In another study, LFA-1 express ion was 18 found to be uniformly distributed on the cell surface, however, avid LFA-1 (able to bind ICAM-1-coa ted beads) was only detected at local ized regions (Pyszn iak et al., 1994). A s wel l , ICAM-1 express ion local ized to uropods has also been observed (Dougherty etal., 1988). The LFA-1 : ICAM-1 attachment is a highly regulated interaction and plays a major role in adhes ion-dependent immunological responses. Ant ibodies against LFA-1 or ICAM-1 have been shown to interfere with a multitude of immune responses (Martz, 1987; Spr inger et al., 1987; Dustin and Springer, 1991). In addit ion to blocking C T L -mediated killing of target cel ls (Davignon et al., 1981; Pierres et al., 1982; Dia lynas et al., 1982), ant ibodies against LFA-1 and ICAM-1 are also able to block natural killer (NK) cel l -mediated cytotoxicity and ant ibody-dependent cytotoxicity mediated by granulocytes and peripheral blood mononuclear cells (Kohl et al., 1984; M i e d e m a et al., 1984; Schmidt et al., 1985). These two antibodies together have been shown to inhibit virtually every adhes ion-dependent immune response. Lymphokine-act ivated killer (LAK) cell mediated cytolysis (Nishimura et al., 1985) and ingestion of Staphylococcus by neutrophils (Ross et al., 1985) are also b locked by these ant ibodies. The LFA-1 : ICAM-1 interaction is a lso involved in T lymphocyte functions. T h e s e ant ibodies block T cell proliferation in response to v i ruses, al loant igens, xenoant igens, and mitogens (Davignon et al., 1981; Pierres et al., 1982; Krensky et al., 1983; Hildreth and August , 1985; Dougherty and Hogg, 1987). T cel l -dependent ant ibody responses by B cel ls are a lso inhibited by ant i -LFA-1 mAbs as is interleukin-2 (IL-2) production by T cell hybrids stimulated by al logeneic A P C s (Davignon et al., 19 1981; Kau fman and Berke, 1983; Go lde et al., 1985; Howard et al., 1986; F ischer et al., 1986). The inhibition exper iments are consistent with the notion that LFA-1 is involved in the strengthening of cellular adhes ion which facilitates the execut ion of the immune response. In the case of lymphocyte proliferation or cytolysis, the ant i -LFA-1 m A b s can cause dissociat ion of preformed conjugates as long as the ant ibodies are added no later than two hours after the initial contacts are made (Krensky et al., 1984; Spi ts et al., 1986). After such a time, the antibodies cannot inhibit these functions, further support ing the idea that LFA-1 : ICAM-1 interactions are involved in the induction of proliferation or cytotoxicity rather than the process itself. In addit ion to the immune responses mentioned above, LFA-1 and ICAM-1 play signif icant roles in other functions. Transendothel ia l migration of leukocytes has been shown to have an L F A - 1 - and ICAM-1-dependent component (Oppenhe imer -Marks et al., 1991; Kavanaugh et al., 1991; Furie et al., 1991; Bev i lacqua, 1993; Car los and Har lan, 1994). Ant ibodies blocking these interactions inhibit leukocyte binding to and subsequent migration across endothelial cel ls. Administrat ion of blocking ant ibodies to rabbits has a lso interfered with leukocyte migration (Barton et al., 1989; Doerschuk et al., 1990; Lo etal., 1992). The LFA-1 and ICAM-1 antibody blocking studies are in agreement with the results obtained from genetical ly deficient mice. ICAM-1 deficient mice display abnormal inflammatory responses and increased numbers of circulating neutrophils and lymphocytes (Sligh et al., 1993; X u et al., 1994). T h e s e mice exhibit impaired neutrophil migration in response to chemical peritonitis as well as de layed type 20 hypersensit ivity. The mutant cel ls from these animals failed to stimulate a l logeneic lymphocytes in a M L R , although they proliferated normally as responder cel ls. ICAM-1 deficient mice were also resistant to septic shock, by reducing either T cell activation or neutrophil infiltration. M ice deficient for the (32 subunit display similar impairment in the inflammatory response to chemical peritonitis (Wilson et al., 1993; Bul lard et al., 1996). T h e s e mice a lso display delays in transplantation rejection. LFA-1-def ic ient mice also exhibited defect ive leukocyte function (Schmits et al., 1996). Lymphocytes d isp layed a dec reased response in an M L R . The inflammatory response to chemica l peritonitis w a s a lso reduced. T h e s e mice were also not able to reject grafted immunogenic tumors. Perturbation of normal adhes ion between LFA-1 and ICAM-1 has been demonstrated to coincide with the metastatic capacity of certain cel ls. This adhes ive interaction al lows adjacent cel ls to remain attached. However, if the LFA-1 : ICAM-1 binding is not present, the total adhes ion is lessened and the l ikelihood that the cel ls can d issoc iate is increased. In B-lymphoid tumors, ICAM-1 express ion correlates with metastas is (Boyd et al., 1989; Wawryk et al., 1989). Large B cell tumors which form bulky, solitary m a s s e s consistently exhibit intermediate to strong ICAM-1 express ion . Converse ly , l ymphoma cel ls which show a diffuse, infiltrative, non-adhes ive pattern of metastas is exhibit little or no ICAM-1 express ion. In contrast to B cell tumors, ICAM-1 express ion s e e m s to be an indicator of metastatic progression of me lanoma cel ls (Johnson et al., 1988; Johnson et al., 1989). ICAM-1 is not normally found on quiescent melanocytes or benign melanocyt ic lesions. However, an increase in tumor 21 s ize is accompan ied by an upregulation in ICAM-1 and results in an increased probability of metastasis. This increased ICAM-1 express ion may be a s ide effect of cytokines re leased from lesion-infiltrating lymphocytes. A s well , the I C A M - 1 + melanoma cel ls may interact with L F A - 1 + recruited leukocytes caus ing a reduction in homotypic aggregat ion between melanoma cel ls. The d issociated me lanoma cel ls can attach to the LFA-1 molecules on the mobile leukocytes present in the lesion. Express ion of ICAM-1 on lymphomas may not only affect d issociat ion from the primary tumor, but a lso e s c a p e from the immune survei l lance of circulating leukocytes. Burkitt's l ymphoma cel ls downregulate LFA-1 and ICAM-1 (Clayberger et al., 1987; Gregory et al., 1988). It is thought that lack of these adhes ion molecu les on the tumor cel ls may contribute to their inability to initiate an effective immune response thus leading to e s c a p e from immunosurvei l lance. This is supported by the observat ion that LFA-1 and ICAM-1 reappear on these tumor cel ls after severa l p a s s a g e s in culture and become sensi t ive to lysis by Epste in-Barr virus (EBV)-speci f ic C T L s . In addit ion to its role in mediating cellular contact in the immune sys tem, ICAM-1 has a lso been shown to function in other capaci t ies. It has been shown, for example , to serve as a receptor for the major serotype of rhinoviruses (Greve et al., 1989; Staunton et al., 1989a; Tomass in i et al., 1989). Rhinov i ruses are the primary causat ive agent of the common cold (Sperber and Hayden , 1988). Viroid particles infect cel ls express ing I C A M - 1 . A soluble version of ICAM-1 lacking the t ransmembrane and cytoplasmic domains is able to specif ical ly block rhinovirus infections v ia ICAM-1 (Marlin et al., 1990). In addition to the common cold, ICAM-1 22 also acts as a receptor for Plasmodium falciparum infected erythrocytes (Berendt et al., 1989). The initial event in the pathogenesis of malaria is the adherence of infected erythrocytes to endothel ium in the liver. In malaria, e levated levels of cytokines have been detected, and these cytokines are also responsible for the induction of ICAM-1 express ion on resting endothel ium. Similar to rhinovirus infection, an immunoadhes in vers ion of ICAM-1 is able to inhibit P. falciparum-\nfected erythrocyte adhes ion to ICAM-1 sur faces (Staunton etal., 1992). 1:4.2 Intercellular Adhesion Molecule-2 (ICAM-2) 1:4.2 a) ICAM-2 cDNA cloning Al though ICAM-1 interaction with LFA-1 plays a major role in the overal l function of the immune sys tem, it is not the only LFA-1-dependent interaction. Initially, it was observed that the LFA-1-dependent binding of T cel ls to endothel ial cel ls contains both an ICAM-1-dependent and an ICAM-1- independent component (Dustin and Springer, 1988a). The ICAM-1-dependent pathway is inducible whereas the I C A M - 1 -independent pathway is constitutive. In addit ion, purified LFA-1 in planar membranes is ab le to strongly bind an ICAM-1" T cell line (Dustin and Springer, 1989). This predicted the ex is tence of at least one alternate LFA-1 counter-receptor. A human c D N A encod ing this molecule was c loned by screening C O S cel ls transfected with a p lasmid-based express ion library (Staunton et al., 1989b). The C O S cel ls were then se lected for their ability to adhere to purified human LFA-1 protein in the presence of ant i - ICAM-1 antibody. This new LFA-1 ligand was designated ICAM-2 (CD 102). 23 I C A M - 2 is an integral membrane protein with two extracellular Ig-like domains . T h e s e two domains share highest sequence identity with the two N-terminus Ig-like domains of human ICAM-1 (34%) suggest ing that the crucial interactions involved in adhes ion to LFA-1 are mediated by these two domains. The LFA-1 binding region of ICAM-1 has been mapped to domain 1 and extends partly into domain 2 indicating that the similarity is both structural as well as functional (Staunton et al., 1990). S ince Mac-1 is a lso able to bind ICAM-1 through the third Ig-like domain (Diamond et al., 1990; D iamond et al., 1991), it suggests that ICAM-2 is not able to act as a counter-receptor for M a c - 1 . The inability of ICAM-2 to mediate adhes ion to Mac-1 has been demonstrated (de Fougerol les et al., 1995). Despi te the structural and functional similarity between ICAM-1 and ICAM-2 , there is an apparent difference in affinity for L F A - 1 . ICAM-1 is able to mediate LFA-1 adhes ion more effectively than ICAM-2 . Initial ev idence to support this differential binding capabil i ty of LFA-1 came from severa l adhes ion and detachment studies (Kishimoto et al., 1990). ICAM-1-transfected C O S cel ls adhere more avidly to L F A - 1 -coated plastic petri plates than do ICAM-2-transfected cel ls ( ICAM-2 C O S cel ls wash off more readily). Ant ibodies against ICAM-1 are able to effectively inhibit cel lular binding to LFA-1 -coa ted plastic when the LFA-1 is coated at a low density. However, when the LFA-1 is coated at a high density, the ant i- ICAM-1 m A b s are not able to completely block the binding. This implies that when a limited number of LFA-1 molecu les are avai lable, the higher affinity l igand, ICAM-1 , will outcompete the lower affinity l igand, ICAM-2 . This differential binding capacity is speci f ic to the tertiary 24 structure of the binding sites. Extension of the ICAM-2 binding site by the addit ion of five extra Ig-like domains does not enhance LFA-1 adhes ion (Damle et al . , 1992a). A s wel l , removal of Ig-like domains 3-5 from ICAM-1 does not diminish binding. T h e s e results are consistent with the idea that the strength of the LFA-1 : ICAM binding is determined by the structure of the binding site and not by the d is tance that the binding domain extends from the cell surface. The gene encoding ICAM-2 is a single copy gene located on ch romosome 17 in humans (Hogg et al., 1991) and chromosome 11 in the murine sys tem (Kuramoto et al., 1994). C lon ing the human c D N A has revealed that it consis ts of a core protein of 29 kD with six potential N-glycosylation sites yielding a mature glycoprotein of 50-60 kD (Staunton etal., 1989b; de Fougerol les etal., 1991). Al though express ion patterns of ICAM-1 and ICAM-2 are different, there is some overlap. ICAM-1 is exp ressed at low levels on a subpopulat ion of lymphocytes, monocytes, and endothel ial cel ls (Dustin et al., 1986; Pober et al., 1986; te Ve lde et al., 1987; Dustin et al., 1988a; Mentzer et al., 1988; Rothlein et al., 1988; Dustin and Springer, 1988a), but is strongly induced on these cel ls and f ibroblasts and epithelial cel ls by cytokines and inflammatory agents (Dustin et al., 1986; Pober etal., 1986; Rothlein et al., 1988). In compar ison , ICAM-2 express ion is restricted to lymphocytes, monocytes, and vascu lar endothel ium (Staunton etal., 1989b; de Fougerol les etal., 1991; Nortamo etal., 1991). In addit ion, I C A M - 2 express ion is not affected by cytokines. ICAM-2 has also been found on platelets and granulocytes (Diacovo et al., 1994). 25 1:4.2 b) P h y s i o l o g i c a l s i g n i f i c a n c e of L F A - 1 : I C A M - 2 interaction Relat ively little investigation has been done on the physiological function of the L F A - 1 : I C A M - 2 interaction. B a s e d on its express ion on endothel ium and its ability to mediate LFA-1-dependen t adhes ion of lymphocytes, it is suspected that ICAM-2 is involved in lymphocyte recirculation (Dustin and Springer, 1991; Car los and Har lan, 1994). In one study, a 22 residue peptide corresponding to a region in domain 1 of human ICAM-2 was shown to inhibit a B cell line from binding to endothel ial cel ls (Li et al., 1993a). This s a m e peptide is a lso able to activate N K cel ls and T cel ls (Li et al., 1993b; Somersa lo et al., 1995; Li et al., 1995). The peptide-act ivated N K cel ls have increased cytotoxicity and are more readily able to migrate through a Boyden chamber . It is suspec ted that this peptide may transmit a signal to the cell through LFA-1 s ince binding of the peptide coinc ides with increased phosphorylat ion of two proteins (Somersa lo et al., 1995). The activated state can dictate the increased function of the cel l . The ability of the ICAM-2 protein to transmit a s ignal is a lso observed in lymphocyte activation (d iscussed in chapter 3). Al though tested only in an artificial sys tem, the ICAM-2 protein is able to augment the primary activation s ignal del ivered through the T c R / C D 3 complex (Damle et al., 1992a; Damle et al., 1992b). Other studies have implicated ICAM-2 express ion in mal ignancy. Endothel ia l ICAM-2 express ion is significantly higher in malignant lymph nodes than in nonmal ignant nodes (Renkonen et al., 1992). High express ion of ICAM-2 was a lso detected on Hodgkin 's d isease-der ived cell l ines (Ellis et al., 1992). A s well , Burkitt's l ymphoma (BL) cel ls are lysed by EBV-spec i f i c C T L s when ICAM-2 is present on the B L cel ls 26 even if other C A M s such as ICAM-1 are absent (Khanna et al., 1993). However, susceptibi l i ty to lysis is increased if ICAM-1 is a lso present suggest ing that ICAM-2 can play a key role in the immune system when ICAM-1 is absent. Natural killer (NK) cel ls are a lso able to utilize the LFA-1 : ICAM-2 pathway for the lysis of target cel ls (Jackson etal., 1992; Ka tsanas etal., 1994; Helander et al., 1996). 1:4.3 In terce l lu la r A d h e s i o n Mo lecu le - 3 (ICAM-3) In similar fashion to the way that ICAM-2 was identified, a third LFA-1 counter-receptor w a s a lso identified. The LFA-1-dependent adhes ion of one human lymphoid cell line to purified LFA-1 was greater than the total I C A M - 1 - and ICAM-2-dependent adhes ion (de Fougero l les et al., 1991; de Fougerol les and Springer, 1992). A third molecule, I C A M - 3 (CD50) , was identified by several groups. O n e group character ized I C A M - 3 based on avai lable sequence information from severa l C A M s (Vazeux et al., 1992). Another group generated antibodies against I C A M - 3 and subsequent ly express ion c loned the c D N A (Fawcett et al., 1992). A third group utilized amino acid s e q u e n c e data from the purified protein to construct ol igonucleot ides which were used to sc reen a c D N A library (de Fougerol les et al., 1993) ICAM-3 is a lso a typical t ransmembrane glycoprotein with a core polypeptide of 56 kD containing 15 potential N-glycosylat ion si tes. The degree of glycosylat ion is evident in the s ize of the mature protein (124 kD). ICAM-3 contains five Ig-like domains, each with varying similarit ies to the corresponding domains of the other members of the I C A M subfamily. B a s e d on s e q u e n c e compar ison and the predicted number of domains, I C A M - 3 is most similar to 27 ICAM-1 with 5 2 % overall amino acid identity and 7 7 % identity in the second domain . The first two domains of ICAM-3 are 3 7 % identical to the two Ig-like domains of I C A M -2. E v e n though the members of the I C A M subfamily share structural and functional similarit ies, their distribution is unique possibly reflecting distinct physiological funct ions. I C A M - 3 is detected on resting lymphocytes, monocytes, neutrophils, and epidermal Langerhans cel ls (de Fougerol les and Springer, 1992; A c e v e d o et al., 1993; Staquet et al., 1995). It is not inducible by cytokines or other inf lammatory agents. Rest ing lymphocytes bind LFA-1 primarily through ICAM-3 ; this fact coupled with the observed higher express ion of ICAM-3 on resting lymphocytes and monocytes compared to ICAM-1 or ICAM-2 suggests that LFA-1 : ICAM-3 may be critical in initiating an immune response (de Fougerol les and Springer, 1992; de Fougero l les et al., 1994). Simi lar to ICAM-1 and ICAM-2 , purified ICAM-3 is a lso able to provide a cost imulatory s ignal which enhances the primary signal del ivered through the T c R / C D 3 complex (de Fougero l les et al., 1994). An t i - ICAM-3 mAb can inhibit peripheral blood lymphocyte proliferation in response to phytohemagglut inin, as well as a l logeneic and anti-specif ic proliferation (de Fougerol les ef al., 1994). The observat ion that a cocktai l of an t i - ICAM-1, -2 , and -3 mAbs can effectively inhibit homotypic aggregat ion to the s a m e extent that ant i -LFA-1 ant ibodies can suggests that there may not be any more members in the I C A M subfamily. In addition, one group has demonstrated that I C A M -3 can a lso adhere to a d ( 3 2 (Van der V ieren ef al., 1996). 28 1:5 T h e s i s ob jec t i ve The interaction and functional s igni f icance of LFA-1 : ICAM-1 adhes ion has been extensively studied. It is involved in virtually every leukocyte function tested. However , the physiological s igni f icance of the LFA-1 : ICAM-2 interaction has not been examined as thoroughly. Al though preliminary studies are indicative of possib le funct ions, they are still not definitive. Based on distribution studies and preliminary T cell activation data, it is speculated that ICAM-2 may play a role in delivering a necessary costimulatory signal to lymphocytes in order for them to proliferate. It is a lso suspec ted that ICAM-2 may play a role in leukocyte recirculation. The overal l objective of this thesis was to examine the possib le roles of murine ICAM-2 in the immune and inflammatory responses. Understanding of ICAM-2 funct ions in the murine sys tem will provide a better basis for the construction of an animal model in the future. W h e n this project was initiated, murine ICAM-2 was not yet character ized. Therefore, the speci f ic a ims of this research were a) to character ize murine ICAM-2 through gene cloning b) to determine the role of ICAM-2 in lymphocyte activation c) to examine the involvement of ICAM-2 in transendothel ial migration of leukocytes. 29 1:6 R e f e r e n c e s A c e v e d o A , del P o z o M A , Arroyo A G , Sanchez -Ma teos P, G o n z a l e z - A m a r o R, and Sanchez -Mad r i d F (1993) Distribution of ICAM-3-bear ing cel ls in normal human t issues. Express ion of a novel counter-receptor for LFA-1 in epidermal Langerhans cel ls. Am. J. Path. 143:774. Ande rson D C , Miller L J , Schmals te ig F C , Rothlein R, and Spr inger T A (1986) Contr ibut ions of the Mac-1 glycoprotein family to adherence-dependent granulocyte funct ions: structure-function assessmen ts employing subunit-specif ic monoc lonal ant ibodies. J. Immunol. 137:15. A n d e r s o n D C and Spr inger T A (1987) Leukocyte adhes ion def ic iency: an inherited defect in the M a c - 1 , L F A - 1 , and p150/95 glycoproteins. Annu. Rev. Med. 38:175. Arnaout M A , Spi ts H, Terhoust C , Pitt J , and Todd RFI (1984) Def ic iency of a leukocyte glycoprotein (LFA-1) in two patients with Mo1 deficiency. J. Clin. Invest. 74:1291. Bal lantyne C M , Kozak C A , O'Br ien W E , and Beaudet A L (1991) Ass ignment of the gene for intercellular adhes ion molecule-1 (/cam-1) to proximal mouse ch romosome 9. Genomics 9:547. Barton R W , Rothlein R, Ks iazek J , and Kennedy C (1989) The effect of anti-intercellular adhes ion molecule-1 on phorbol-ester- induced rabbit lung inf lammation. J. Immunol. 143:1278. Beatty P G , Harlan J M , R o s e n H, Hansen J A , O c h s HD, Pr ice T, Taylor D, and Klebanoff S J (1984) A b s e n c e of monoclonal-ant ibody-def ined protein complex in a boy with abnormal leukocyte function. Lancet / :535. Bel ler D l , Spr inger TA , and Schre iber R D (1982) Ant i -Mac-1 select ively inhibits the mouse and human type three complement receptor. J. Exp. Med. 156:1000. Berendt A R , S immons DL, Tansey J , Newbold C l , and Marsh K (1989) Intercellular adhes ion molecule-1 is an endothelial cell adhes ion receptor for P lasmod ium falc iparum. Nature 341:57. Bev i lacqua M P (1993) Endothel ial- leukocyte adhes ion molecules. Annu. Rev. Immunol. 11:767. Boyd A W , Dunn S M , Fecondo J V , Culvenor J G , Duhrsen U, Burns G F , and Wawryk S O (1989) Regulat ion of express ion of a human intercellular adhes ion molecule ( ICAM-1) during lymphohematopoiet ic differentiation. Blood 73:1896. 30 Burn P, Kupfer A , and S inger S J (1988) Dynamic membrane-cytoskeleta l interactions: speci f ic assoc iat ion of integrin and talin ar ises in vivo after phorbol ester treatment of peripheral blood lymphocytes. Proc. Natl. Acad. Sci. USA 85:479. Bul lard D C , Scharf fet ter-Kochanek K, McArthur M J , C h o s a y J G , McBr ide M E , Montgomery C A , and Beaudet A L (1996) A polygenic mouse model of psoriasi form skin d i sease in CD18-def ic ient mice. Proc. Natl. Acad. Sci. USA 93:2116. C a m p a n a D, Sher idan B, T idman N, Hoffbrand A V , and J a n o s s y G (1986) Human leukocyte funct ion-associated ant igens on lymphohemopoiet ic precursor cel ls. Eur. J. Immunol. 16:537. Car los T M and Harlan J M (1994) Leukocyte-endothel ia l adhes ion molecu les. Blood 84:2068. Clark E A , Ledbetter J A , Holly R C , Dinndorf P A , and S h u G (1986) Polypept ide on human B lymphocytes assoc ia ted with cell activation. Hum. Immunol. 16:100. C layberger C , Wright A , Medei ros L J , Kol ler TD , Link M P , Smith S D , Warnke R A , and Krensky A M (1987) A b s e n c e of cell surface LFA-1 as an e s c a p e from immunosurvei l lance. Lancet 2(8558):533. Corb i A L , Garc ia-Agui la r J , and Spr inger T A (1990) G e n o m i c structure of an integrin a lpha subunit, the leukocyte p150/95 molecule. J. Biol. Chem. 265:2782. Corb i A L , Kishimoto TK , Miller L J , and Spr inger T A (1988a) The human leukocyte adhes ion glycoprotein Mac-1 (complement receptor type 3, CD11b ) a lpha subunit: c loning, primary structure, and relation to the integrins, von Wi l lebrand factor and factor B. J. Biol. Chem. 263:12403. Corb i A L , Larson R S , Kishimoto TK , Spr inger TA , and Morton C C (1988b) Ch romosoma l location of the genes encoding the leukocyte adhes ion receptors L F A - 1 , M a c - 1 , and p150/95. Identification of a gene cluster involved in cell adhes ion . J. Exp. Med. 167:1597. Corb i A L , Mil ler L J , O 'Connor K, Larson R S , and Spr inger T A (1987) c D N A cloning and complete primary structure of the alpha subunit of a leukocyte adhes ion glycoprotein, p150/95. EMBOJ. 6 :4023. Crowley C A , Curnutte JT , Ros in R E , Andre-Schwar tz J , Gal l in J l , K lempner M, Snyde rman R, Southwick F S , S tosse l T P , and Baboir B M (1980) A n inherited abnormali ty of neutrophil adhes ion: its genetic t ransmission and its assoc iat ion with a miss ing protein. N. Eng. J. Med. 302:1163. 31 Dahms N M and Hart G (1985) Lymphocyte funct ion-associated antigen 1 (LFA-1) contains N-l inked o l igosacchar ides. J. Immunol. 134:3978. Damle NK , K lussman K, and Aruffo A (1992a) Intercellular adhes ion molecule-2, a second counter-receptor for C D 1 1 a / C D 1 8 (leukocyte funct ion-associated antigen-1), provides a costimulatory signal for T-cel l receptor-initiated activation of human T cel ls. J. Immunol. 148:665. Damle NK, K lussman K, Linsley P S , Aruffo A , and Ledbetter J A (1992b) Differential regulatory effects of intercellular adhes ion molecule-1 on cost imulat ion by the C D 2 8 counter-receptor B7 . J. Immunol. 149:2541. D a n a N, Styrt B, Griffin J , Todd R F , K lemper M, and Arnaout M A (1986) Two functional domains in the phagocyte membrane glycoprotein Mo1 identified with monoc lona l ant ibodies. J. Immunol. 137:3259. Dani lenko D M , Rossi t to P V , V a n d e r V i e r e n M, Trong HL, McDonough S P , Affolter V K , and Moore P F (1995) A novel canine leukointegrin, a d p 2 , is expressed by speci f ic macrophage subpopulat ions in t issue and a minor C D 8 + lymphocyte subpopulat ion in peripheral b lood. J. Immunol. 155:36. Dav ignon D, Martz E, Reyno lds T, Kurzinger K, and Spr inger T A (1981) Monoc lona l ant ibody to a novel lymphocyte funct ion-associated antigen (LFA-1) : mechan ism of blocking of T lymphocyte-mediated killing and effects on other T and B lymphocyte funct ions. J. Immunol. 127:590. de Fougero l les A R , Diamond M S , and Spr inger T A (1995) Heterogeneous glycosylat ion of I C A M - 3 and lack of interaction with Mac-1 and p150/95. Eur. J. Immunol. 25 :1008. de Fougero l les A R , Kl ickstein L B , and Spr inger T A (1993) Cloning and express ion of intercellular adhes ion molecule 3 reveals strong homology to other immunoglobul in family counter-receptors for lymphocyte funct ion-associated antigen 1. J. Exp. Med. 177:1187. de Fougero l les A R , Qin X , and Spr inger T A (1994) Character izat ion of the function of intercellular adhes ion molecule ( ICAM)-3 and compar ison with ICAM-1 and ICAM-2 in immune responses . J. Exp. Med. 179:619. de Fougero l les A R and Spr inger T A (1992) Intercellular adhes ion molecule 3, third adhes ion counter-receptor for lymphocyte funct ion-associated molecule 1 on resting lymphocytes. J. Exp. Med. 175:185. 32 de Fougero l les A R , Stacker S A , Schwart ing R, and Spr inger T A (1991) Character izat ion of ICAM-2 and ev idence for a third counter-receptor for L F A - 1 . J. Exp. Med. 174:253. Degitz K, L ian-J ie L, and C a u g h m a n S W (1991) Cloning and character izat ion of the 5'-transcriptional regulatory region of the human intercellular adhes ion molecule 1 gene. J. Biol. Chem. 266:14024. de la Hara A , A l ve rez -Mon M, Sanchez -Madr id F, Mart inez A C , and Durantez A (1988) Co-exp ress ion of Mac-1 and p150/95 on C D 5 + B cel ls. Structural and functional character izat ion in a human chronic lymphocytic leukemia. Eur. J. Immunol. 18:1134. D iacovo T G , de Fougero l les A R , Bainton D F , and Spr inger T A (1994) A functional integrin l igand on the surface of platelets: intercellular adhes ion molecule-2. J. Clin. Invest. 94 :1243. Dia lynas D, Loken M, Sarmiento M, and Fitch F W (1982) Identification of lysis-relevant molecu les on the surface of C T L : primary screening of monoclonal ant ibodies for the capaci ty to block cytolysis by cloned T cell l ines. Adv. Exp. Med. Biol. 146:547. D iamond M S , Staunton D E , de Fougerol les A R , Stacker S A , Garc ia-Agu i la r J , Hibbs M L , and Spr inger T A (1990) ICAM-1 (CD54) : a counter-receptor for Mac-1 ( C D 1 1 b / C D 1 8 ) . J. Cell Biol. 111:3129. D iamond M S , Staunton D E , Marlin S M , and Spr inger T A (1991) Binding of the integrin Mac-1 ( C D 1 1 b / C D 1 8 ) to the third immunoglobulin-l ike domain of ICAM-1 (CD54) and its regulation by glycosylat ion. Cell 65 :961. Doerschuk C M , Winn R K , C o x s o n H O , and Harlan J M (1990) CD18-dependen t and -independent mechan isms of neutrophil emigration in the pulmonary and systemic microcirculat ion of rabbits. J. Immunol. 144:2327. Dougherty G J and Hogg N (1987) The role of monocyte lymphocyte function-assoc ia ted antigen-1 (LFA-1) in accessory cell function. Eur. J. Immunol. 17:943. Dougherty G J , Murdoch S , and Hogg N (1988) The function of human intercellular adhes ion molecule-1 ( ICAM-1) in the generat ion of an immune response. Eur. J. Immunol. 18:35. Dransf ield I and Hogg N (1989) Regulated express ion of M g 2 + binding epi tope on leukocyte integrin a subunits. EMBO J. 8:3759. 33 Dustin M L , Rothlein R, Bahn A K , Dinarello C A , and Spr inger T A (1986) Induction by IL-1 and interferon, t issue distribution, biochemistry, and function of a natural adherence molecule ( ICAM-1). J. Immunol. 137:245. Dustin M L , S inger K H , Tuck DT, and Spr inger T A (1988a) Adhes ion of T lymphoblasts to ep idermal kerat inocytes is regulated by gamma interferon and is mediated by intercellular adhes ion molecule-1 ( ICAM-1). J. Exp. Med. 167:1323. Dust in M L and Spr inger T A (1988a) Lymphocyte function assoc ia ted antigen-1 ( LFA-1) interaction with intercellular adhes ion molecule-1 ( ICAM-1) is one of at least three mechan i sms for lymphocyte adhes ion to cultured endothel ial cel ls. J. Cell Biol. 107:321. Dustin M L and Spr inger T A (1989) T-cel l receptor cross-l inking transiently st imulates adhes i veness through L F A - 1 . Nature 341:619. Dustin M L and Spr inger T A (1991) Ro le of lymphocyte adhes ion receptors in transient interactions and cell locomotion. Annu. Rev. Immunol. 9:27. Dustin M L , Staunton D E , and Spr inger T A (1988b) Supergene famil ies meet in the immune sys tem. Immunol. Today 9:213. Ede lman G M (1976) Science 192:218. Ede lman G M (1977) Summary : understanding select ive molecular recognit ion. Cold Spring Harbor Symposium on Quantitative Biology 41(2): 891. Ede lman G M (1989) Topobiology. Sci. Amer. 260:76. Ell is P A , Hart D N , Col ls B M , Nimmo J C , MacDona ld J E , and A n g u s H B (1992) Hodgkin 's cel ls express a novel pattern of adhes ion molecules. Clin. Exp. Immunol. 90:117 Farrar M A and Schre iber R D (1993) The molecular cell biology of interferon-y and its receptor. Annu. Rev. Immunol. 11:571. Fawcett J , Ho lness C L L , Needham LA, Turley H, Gatter K C , M a s o n D Y , and S immons DL (1992) Molecular cloning of ICAM-3 , a third l igand for L F A - 1 , constitutively exp ressed on resting leukocytes. Nature 360:481. F ischer A , Durandy A , Sterkers G , and Griscel l i C (1986) Ro le of the LFA-1 molecule in cel lular interactions required for antibody production in humans. J. Immunol. 136:3198. 34 Fischer A , L isowska-Grosp ier re B, Anderson D C , and Spr inger T A (1988) The leukocyte adhes ion def iciency: molecular basis and functional consequences . Immunodef. Rev. 1:39. F ischer A , L isowska-Grosp ier re B, Mazero l les F, Le Deist F, Pe rez N, D imanche-Boitrel MT, and Griscel l i C (1990) Leukocyte adhes ion def ic iency as a model for the study of the functional role of L F A - 1 . Spr inger et al. (eds): Leukocyte adhes ion molecu les . New York, Spr inger-Ver lag, Chapter 6. Flotte T, Spr inger TA , and Thorbecke G J (1983) Dendrit ic cell and macrophage staining by monoclonal ant ibodies in t issue sect ions and epidermal sheets . Am. J. Path. 111:112. Fur ie M B , Tanc inco M C A , and Smith C W (1991) Monoc lona l ant ibodies to leukocyte integrins C D 1 1 a / C D 1 8 and C D 1 1 b / C D 1 8 or intercellular adhes ion molecule-1 inhibit chemoattractant-st imulated neutrophil transendothel ial migration in vitro. Blood 78:2089. Go lde W T , Kappler J W , Greenste in J , Mal issen B, Hood L, and Marrack P (1985) Major histocompatibil i ty complex-restr icted antigen receptor on T cel ls. VIII Ro le of the LFA-1 molecule. J. Exp. Med. 161:635. Gregory C D , Murray R J , Edwards C F , and Rick inson A B (1988) Downregulat ion of cell adhes ion molecules L F A - 3 and ICAM-1 in Epste in-Barr virus-posit ive Burkitt's l ymphoma underl ies tumor escape from virus-specif ic T cell survei l lance. J. Exp. Med. 167:1811. G reve J M , Davis G , Meyer A M , Forte C P , Yos t S C , Marlor C W , Kamarck M E , and McCle l l and A (1989) The major human rhinovirus receptor is I C A M - 1 . Cell 56:839. He lander T S , Ca rpen O, Turunen O, Kovanen P E , Vaher i A , and T imonen T (1996) ICAM-2 redistributed by ezrin as a target for killer cel ls. Nature 382:265. Hermanowsk i -Vosa tka A , V a n Strijp J A G , Swiggard W J , and Wright S D (1992) Integrin modulat ing factor-1: a lipid that alters the function of leukocyte integrins. Cell 68 :341 . Hibbs M L , Ward law A J , Stacker S A , Anderson D C , Lee A , Roberts T M , and Spr inger T A (1990) Transfect ion of cel ls from patients with leukocyte adhes ion def ic iency with an integrin p subunit (CD18) restores lymphocyte funct ion-associated antigen-1 express ion and function. J. Clin. Invest. 85:674. 35 Hildreth J E K and August J T (1985) The human lymphocyte funct ion-associated (HLFA) antigen and a related macrophage differentiation antigen (HMac-1) : functional effects of subunit-specif ic monoclonal ant ibodies. J. Immunol. 134:3272. Ho M K and Spr inger T A (1983) Biosynthesis and assembly of the a lpha and beta subunits of M a c - 1 , a macrophage glycoprotein assoc ia ted with complement receptor function. J. Biol. Chem. 258:2766. Hogg N (1989) The leukocyte integrins. Immunol. Today 10 :111 . Hogg N, Bates P A , and Harvey J (1991) Structure and function of intercellular adhes ion molecu le-1. Chem. Immunol 50:98. Horley K J , Carpeni to C , Baker B, and Take i F (1989) Molecular cloning of murine intercellular adhes ion molecule ( ICAM-1). EMBO J. 8:2889. Howard D R , E a v e s A C , and Take i F (1986) Lymphocyte funct ion-associated antigen (LFA-1) is involved in B cell activation. J. Immunol. 136:4013. Huang C and Spr inger T A (1995) A binding interface on the I domain of lymphocyte funct ion-associated antigen-1 (LFA-1) required for speci f ic interaction with intercellular adhes ion molecule 1 ( ICAM-1). J. Biol. Chem. 270:19008. Hynes R O (1987) Integrins: a family of cell surface receptors. Cell 48 :549. J a c k s o n A M , A lexandrov A B , Prescott S , J a m e s K, and Ch isho lm G D (1992) Ro le of adhes ion molecu les in lymphokine-act ivated killer cell killing of bladder cancer cel ls: further ev idence for a third ligand for leukocyte funct ion-associated ant igen-1. Immunol. 76:286. Johnson J P , S tade B G , Ho lzmann B, Schwab le W, and Riethmuller G (1989) De novo express ion of intercellular adhes ion molecule 1 in me lanoma correlates with increased risk of metastas is . Proc. Natl. Acad. Sci. USA 86:641. J o h n s o n J P , S tade B G , Hupke U, Ho lzmann B, and Riethmuller G (1988) The me lanoma progress ion-associated antigen p3.58 is identical to the intercellular adhes ion molecule, I C A M - 1 . Immunobiol. 178:275. Katsan is E, Bausero M A , X u H, Orchard P J , X u Z, Mclvor R S , Brian A A , and B lazer B R (1994) Transfect ion of the mouse ICAM-1 gene into murine neuroblastoma enhances susceptibi l i ty to lysis, reduces in vivo tumorigenicity and dec reases ICAM-2 -dependent killing. Cane. Immunol. Immunother. 38:135. 36 Katz F E , Parker M, Stanley K, Murray L J , Clark E A , and G r e a v e s M F (1985) C h r o m o s o m e mapping of cell membrane antigens expressed on activated B cel ls. Eur. J. Immunol. 15:103. Kau fmann Y and Berke G (1983) Monoc lona l cytotoxic T lymphocyte hybr idomas capab le speci f ic killing activity, antigen respons iveness, and inducible interleukin secret ion. J. Immunol. 131:50. Kau fmann Y , Tseng E, and Spr inger T A (1991) Cloning of the murine lymphocyte funct ion-associated molecule-1 a-subunit and its express ion in C O S cel ls. J. Immunol. 147:369. Kavanaugh A F , Lightfoot E, Lipsky P E , and Oppenhe imer -Marks N (1991) Ro le of C D 1 1 / C D 1 8 in adhes ion and transendothel ial migration of T cel ls: analys is using CD18-def ic ient T cell c lones. J. Immunol. 146:4149. Ke ize r G D , Borst J , V i sse r W , Schwart ing R, de Vr ies J E , and Figdor C G (1987b) Membrane glycoprotein p150/95 of human cytotoxic T cell c lones is involved in conjugate formation with target cel ls. J. Immunol. 138:3130. Ke ize r G D , te V e l d e A A , Schwart ing R, Figdor C G , and de Vr ies J E (1987a) Ro le of p150/95 in adhes ion , migration, chemotaxis, and phagocytosis of human monocytes . Eur. J. Immunol. 17:1317. K h a n n a R, Burrows S R , Suhrbier A , Jacob C A , Griffin H, Misko IS, Scu l ley T B , R o w e M, R ick inson A B , and M o s s DJ (1993) E B V peptide epitope sensit izat ion restores human cytotoxic T cell recognition of Burkitt's lymphoma cel ls. Ev idence for a critical role for I C A M - 2 . J. Immunol. 150:5154. Kishimoto T K , Hol lander N, Roberts T M , Anderson D C , and Spr inger T A (1987b) Heterogeneous mutations in the beta subunit common to the L F A - 1 , M a c - 1 , and p150/95 glycoproteins cause leukocyte adhes ion def iciency. Cell 50:193. Kishimoto T K , Larson R S , Corbi A L , Dustin ML , Staunton D E , and Spr inger T A (1990) Leukocyte integrins. Spr inger et al. (eds): Leukocyte adhes ion molecules. N e w York, Spr inger -Ver lag, Chapter 1. Kishimoto T K , O 'Connor K, Lee A , Roberts T M , and Spr inger T A (1987a) Clon ing of the beta subunit of the leukocyte adhes ion proteins: homology to an extracel lular matrix receptor def ines a novel supergene family. Cell 48 :681 . Kohl S , Spr inger TA , Schmals t ieg F C , Loo L S , and Anderson D C (1984) Defect ive natural killer cytotoxicity and polymorphonuclear leukocyte ant ibody-dependent cel lular cytotoxicity in patients with L F A - 1 / O K M - 1 deficiency. J. Immunol. 133:2972. 37 Kohler G and Milstein C (1975) Cont inuous cultures of fused cel ls secret ing antibody of predef ined specificity. Nature 256:495. Krensky A M , Robb ins E, Spr inger TA , and Burakoff S J (1984) L F A - 1 , L F A - 2 , and L F A - 3 ant igens are involved in CTL-target conjugation. J. Immunol. 132:2180. Krensky, A M , Sanchez -Madr id F, Robbins E, Nagy J , Spr inger TA , and Burakoff S J (1983) The functional s igni f icance, distribution, and structure of L F A - 1 , L F A - 2 , and L F A - 3 : cell sur face ant igens assoc ia ted with CTL-target interactions. J. Immunol. 131:611. Kupfer A , Burn P, and Singer S J (1990) The PMA- induced speci f ic assoc iat ion of LFA-1 and talin in intact c loned T helper cel ls. J. Mol. Cell. Immunol. 4 :317. Kupfer A and S inger S J (1989a) The specif ic interaction of helper T cel ls and ant igen-present ing B cel ls. IV. Membrane and cytoskeletal reorganization in the bound T cel ls as a function of antigen dose . J. Exp. Med. 170:1697. Kupfer A and Singer S J (1989b) Cel l biology of cytotoxic helper T cell functions: immunof luorescence microscopic studies of single cel ls and cell coup les . Annu. Rev. Immunol. 7:309. Kuramoto T, Mash imo T, Koike R, Miyawaki S , Y a m a d a J , M iyasaka M, and Ser ikawa T (1994) The a lymphoplas ia (a/y) mutation co-segregates with the intercellular adhes ion molecule-2 ( ICAM-2) on mouse chromosome 11. Inter. Immunol. 6 :991. Kurz inger K, Reyno lds T, Germain R N , Davignon D, Martz E, and Spr inger T A (1981) A novel lymphocyte funct ion-associated antigen (LFA-1) : cel lular distribution, quantitative express ion , and structure. J. Immunol. 127:596. Larson R S , Corb i A L , Berman L, and Spr inger T A (1989) Pr imary structure of the leukocyte funct ion-associated molecule-1 a lpha subunit: an integrin with an embedded domain defining a protein superfamily. J. Cell Biol. 108:703. Larson R S and Spr inger T A (1990) Structure and function of leukocyte integrins. Immunol. Rev. 114:181. Law S K A , G a g n o n J , Hildreth J E K , Wel ls C E , Will is A C , and W o n g A J (1987) The primary structure of the beta subunit of the cell surface adhes ion glycoproteins L F A - 1 , C R 3 , and p150/95 and its relationship to the fibronectin receptor. EMBO J. 6 :915. Li R, Nortamo P, Kantor C , Kovanen P, T imonen T, and Gahmberg C G (1993b) A leukocyte integrin binding peptide from intercellular adhes ion molecule-2 st imulates T cell adhes ion and natural killer cell activity. J. Biol. Chem. 268:21474. 38 Li R, Nortamo P, Va lmu L, To lvanen M, Huuskonen J , Kantor C , and Gahmberg C G (1993a) A peptide from ICAM-2 binds to the leukocyte integrin C D 1 1 a / C D 1 8 and inhibits endothel ial cell adhes ion. J. Biol. Chem. 268:17513. Li R, X i e J , Kantor C , Koist inen V , Altieri D, Nortamo P, and Gahmberg C G (1995) A peptide der ived from the intercellular adhes ion molecule-2 regulates the avidity of the leukocyte integrins C D 1 1 b / C D 1 8 and C D 1 1 c / C D 1 8 . J. Cell Biol. 129:1143. L isowska-Grosp ie r re B, Bohler M, F ischer A , M a w a s C , Springer, TA , and Griscel l i C (1986) Defect ive membrane express ion of the LFA-1 complex may be secondary to the a b s e n c e of the beta chain in a child with recurrent bacterial infection. Eur. J. Immunol. 16:205. Lo S K , Everitt J , G u J , and Malik A B (1992) Tumor necrosis factor mediates exper imental pulmonary e d e m a by ICAM-1 and CD18-dependen t mechan isms . J. Clin. Invest. 89 :981. M a k g o b a M W , Sanders M E , Luce G E G , Dustin ML , Spr inger TA , Clark E A , Mannon i P, and S h a w S (1988) ICAM-1 is a ligand for LFA-1-dependent adhes ion of B, T and myeloid cel ls. Nature 331:86. Malhorta V , Hogg N, and S im R B (1986) Ligand binding by the p150/95 antigen of U937 monocyt ic cel ls: properties in common with complement receptor type 3 (CR3) . Eur. J. Immunol. 16:1117. Marl in S D , Morton C C , Anderson D C , and Spr inger T A (1986) LFA-1 immunodef ic iency d isease : definition of the genetic defect and chromosomal mapping of a lpha and beta subunits by complementat ion in hybrid cel ls. J. Exp. Med. 164:855. Marl in S D and Spr inger T A (1987) Purif ied intercellular adhes ion molecule-1 ( ICAM-1) is a l igand for lymphocyte funct ion-associated antigen 1 (LFA-1) . Cell 51 :813. Marl in S D , Staunton D E , Spr inger TA , Stratowa C , Sommergruber W, and Mer luzz i V J (1990) A soluble form of intercellular adhes ion molecule-1 inhibits rhinovirus infection. Nature 344:70. Martz E (1975) Early steps in speci f ic tumor cell lysis by sensi t ized mouse T lymphocytes. I. Resolut ion and characterizat ion. J. Immunol. 115:261. Martz E (1980) Immune T lymphocyte to tumor cell adhes ion . Magnes ium sufficient, ca lc ium insufficient. J. Cell Biol. 84:584. Martz E (1987) LFA-1 and other accessory molecules functioning in adhes ion of T and B lymphocytes. Hum. Immunol. 10:3. 39 Mentzer S J , Rothlein R, Spr inger TA , and Fal ler D V (1988) Intercellular adhes ion molecule-1 ( ICAM-1) is involved in the cytolytic T lymphocyte interaction with human synovia l cel ls. J. Cell. Physiol. 137:173. Mick lem K J and S im R B (1985) Isolation of complement- f ragment- iC3b-binding proteins by affinity chromatography. Biochem. J. 231:233. M i e d e m a F, Tetteroo P A T , Hessel ink W G , Werner G , Spits H, and Mel ief C J M (1984) Both F c receptors and LFA-1 on human Ty-lymphocytes are required for antibody-dependent cel lular cytotoxicity (killer cell activity). Eur. J. Immunol. 14:518. M i e d e m a F, Tromp J F , van't V e e r M B , P o p p e m a S , and Mel ief C J M (1985) Lymphocyte funct ion-associated antigen 1 (LFA-1) is a marker of mature ( immunocompetent) lymphoid cel ls. A survey of lymphoproliferative d i s e a s e s in man. Leuk. Res. 9:1099. Mil ler B A , Antognett i G , and Spr inger T A (1985) Identification of cell sur face ant igens present on murine hematopoiet ic stem cel ls. J. Immunol. 134:3286. Mul ler K and Ger i sch G (1978) A specif ic glycoprotein as a target site of adhes ion blocking Fab in aggregat ing dictyostelium cel ls. Nature 274:445. Munro J M , Pobe r J S , and Cotran R S (1989) Tumor necrosis factor and interferon-g a m m a induce distinct patterns of endothelial activation and leukocyte accumulat ion in skin of Papio anubis. Am. J. Pathol. 135 :121 . Nish imura T, Y a g i H, Yag i ta H, Uch iyama Y , and Hashimoto Y (1985) Lymphok ine-act ivated cel l -associated antigen involved in broad-react ive killer cel l -mediated cytotoxicity. Cell. Immunol. 94:122. Nor tamo P, Sa l cedo R, T imonen T, Patarroyo M, and Gahmberg C G (1991) A monoc lona l antibody to the human leukocyte adhes ion molecule intercellular adhes ion molecule-2: cel lular distribution and molecular distribution and molecular character izat ion of the antigen. J. Immunol. 146:2530. O 'Conne l l K A and Edidin M (1990) A mouse lymphoid endothel ial cell line immortal ized by simian virus 40 binds lymphocytes and retains functional character ist ics of normal endothel ial cel ls. J. Immunol. 144:521. Oppenhe imer -Marks N, Davis L S , Bogue DT, Ramberg J , and Lipsky P E (1991) Differential utilization of ICAM-1 and V C A M - 1 during the adhes ion and transendothel ial migration of human T lymphocytes. J. Immunol. 147:2913. 40 Pava lko F M and L a R o c h e S M (1993) Activation of human neutrophils induces an interaction between the integrin p2 subunit (CD18) and the actin binding protein a-actinin. J. Immunol. 151:3795. Pierres M, Gor id is C , and Golstein P (1982) Inhibition of murine T cel l -mediated cytolysis and T cell proliferation by a rat monoclonal antibody immunoprecipitat ing two lymphoid cell sur face polypeptides of 94,000 and 180,000 molecular weight. Eur. J. Immunol. 12:60. P o b e r J S , G imbrone M A , Lapierre LA, Mendrick DL, Fiers W , Rothlein R, and Spr inger T A (1986) Over lapping patterns of activation of human endothel ial cel ls by interleukin-1, tumor necros is factor, and immune interferon. J. Immunol. 137:1893. Pyszn iak A M , We lde r C A , and Take i F (1994) Cel l surface distribution of high-avidity LFA-1 detected by soluble ICAM-1-coated microspheres. J. Immunol. 152:5241. R a m o s O F , Ka i C , Yefenof E, and Klein E (1988) The elevated natural killer sensitivity of targets carrying surface-at tached C 3 fragments require the availability of the iC3b receptor (CR3) on the effectors. J. Immunol. 140:1239. Rand i A M and Hogg N (1994) I domain of p2 integrin lymphocyte funct ion-associated antigen-1 contains a binding site for l igand intercellular adhes ion molecu le-1 . J. Biol. Chem. 269 :12395. R e n k o n e n R, P a a v o n e n T, Nortamo P, and Gahmberg C G (1992) Express ion of endothel ia l adhes ion molecules in vivo. Increased endothel ial ICAM-2 express ion in lymphoid mal ignancies. Am. J. Path. 140:763. R o s s G D , Thompson R A , Walport M J , Spr inger TA , Watson J V , Ward R H R , L ida J , N e w m a n S L , Harr ison R A , and Lachman P J (1985) Character izat ion of patients with an increased susceptibi l i ty to bacterial infections and a genet ic def ic iency of leukocyte membrane complement receptor type 3 and the related membrane antigen L F A - 1 . Blood 66 :882. Rothlein R, Cza jkowsk i M, O'Nei l l M M , Marlin S D , Mainolfi E, and Mer luzz i V J (1988) Induction of intercellular adhes ion molecule-1 on primary and cont inuous cell l ines by pro-inf lammatory cytokines. J. Immunol. 141:1665. Rothlein R, Dustin M L , Marlin S D , and Spr inger T A (1986) A human intercellular adhes ion molecule ( ICAM-1) distinct from L F A - 1 . J. Immunol. 137:1270. Rothlein R and Spr inger T A (1985) Complement receptor type three-dependent degradat ion of opson ized erythrocytes by mouse macrophages . J. Immunol. 135:2668. 41 Rothlein R and Spr inger T A (1986) The requirement for lymphocyte function-assoc ia ted antigen 1 in homotypic leukocyte adhes ion stimulated by phorbol esters. J . Exp. Med. 163:1132. Sas t re L, Kishimoto T K , G e e C , Roberts T, and Spr inger T A (1986a) The mouse leukocyte adhes ion proteins Mac-1 and L F A - 1 : studies on m R N A translation and protein glycosylat ion with emphas is on Mac -1 . J. Immunol. 137:1060. Sas t re L, R o m a n J M , Tep low D B , Drewer W J , G e e C E , Larson R S , Rober ts T M , and Spr inger T A (1986b) A partial genomic D N A clone for the a lpha subunit of the mouse complement receptor type 3 and cellular adhes ion molecule M a c - 1 . Proc. Natl. Acad. Sci. USA 83:5644. Schmidt R E , Bartley G , Levine H, Sch lossman S F , and Ritz J (1985) Funct ional character izat ion of LFA-1 ant igens in the interaction of human N K c lones and target cel ls. J. Immunol. 135:1020. Schmi ts R, Kundig T M , Baker D M , Shumaker G , S imard J J L , Duncan G , W a k e h a m A , Shah in ian A , van der Heiden A , Bachmann M F , Ohash i P S , Mak T W , and Hickstein D D (1996) LFA-1-def ic ient mice show normal C T L responses to virus but fail to reject immunogenic tumor. J. Exp. Med. 183:1415. Shor tman K and Goldste in P (1979) Target cell recognit ion by cytolytic T cel ls: different requirements for the formation of strong conjugates or for proceeding to lysis. J. Immunol. 123:833. S i m m o n s D, M a k g o b a M W , and S e e d B (1988) ICAM-1 , an adhes ion l igand for L F A - 1 , is homologous to the neural cell adhes ion molecule N C A M . Nature 331:624. S iu G , Hedr ick S M , and Brian A A (1989) Isolation of the murine intercellular adhes ion molecule-1 ( ICAM-1) gene: ICAM-1 enhances ant igen-specif ic T cell activation. J. Immunol. 143:3813. Sl igh J E , Bal lantyne C M , Rich S R , Hawkins HK, Smith C W , Bradley A , and Beaudet A L (1993) Inflammatory and immune responses are impaired in mice deficient in intercellular adhes ion molecule 1. Proc. Natl. Acad. Sci. USA 90:8529. Somersa lo K, Carpen O, S a k s e l a E, Gahmberg C G , Nortamo P, and T imonen T (1995) Act ivat ion of natural killer cell migration by leukocyte integrin-binding peptide from intercellular adhes ion molecule-2 ( ICAM-2). J. Biol. Chem. 270:8629. Sperber S J and Hayden F G (1988) Chemotherapy in rhinovirus co lds. Antimic. Agents. Chemo. 32:409. 42 Spits H, van Schoo ten W, Keizer H, van Seventer G , van de Rijn M, Terhorst C , and de Vr ies J E (1986) Al loant igen recognition is preceded by nonspeci f ic adhes ion of cytotoxic T cel ls and target cel ls. Science 232:403. Spr inger T A (1990) Adhes ion receptors of the immune sys tem. Nature 346:425. Spr inger T A , Dustin ML , Kishimoto TK , and Marlin S D (1987) The lymphocyte funct ion-associated L F A - 1 , C D 2 , and L F A - 3 molecules: cell adhes ion receptors of the immune sys tem. Annu. Rev. Immunol. 5:223. Spr inger TA , Gal f re G , S e c h e r D S , and Milstein C (1979) M a c - 1 : a differentiation ant igen identified by monoclonal antibody. Eur. J. Immunol. 9:301. Spr inger T A , Thompson W S , Miller L J , Schmals t ieg F C , and Anderson D C (1984) Inherited def ic iency of the M a c - 1 , L F A - 1 , p150/95 glycoprotein family and its molecular bas is . J. Exp. Med. 160:1901. Spr inger T A and Unk less J C (1984) Ana lys is of macrophage differentiation and function with monoclonal ant ibodies. A d a m s D O and Hanna M G (eds): Contemporary topics in immunobiology, volume 14. New York, P lenum Press , chapter 1. Staquet M J , Peguet J , Jacquet C , Dezutter-Dambuyant C , and Schmitt D (1995) Express ion of I C A M - 3 on human epidermal dendritic cel ls. Immunobio. 192:249. Staunton D E , Dustin M L , Er ickson H P , and Spr inger T A (1990) The arrangement of the immunoglobul in- l ike domains of ICAM-1 and the binding sites for LFA-1 and rhinovirus. Cell 61:243. Staunton D E , Dustin M L , and Spr inger T A (1989b) Funct ional cloning of ICAM-2 , a cel l adhes ion ligand for LFA-1 homologous to ICAM-1 . Nature 339:61. Staunton D E , Marl in S D , Stratowa C , Dustin ML , and Spr inger T A (1988) Pr imary structure of intercellular adhes ion molecule 1 ( ICAM-1) demonstrates interaction between members of the immunoglobul in and integrin supergene famil ies. Cell 52:925. Staunton D E , Mer luzz i V J , Rothlein R, Barton R, Marlin S D , and Spr inger T A (1989a) A cell adhes ion molecule, I C A M - 1 , is the major surface receptor for rhinoviruses. Cell 56:849. Staunton D E , Ockenhouse C K , and Spr inger T A (1992) So lub le intercellular adhes ion molecule 1-immunoglobulin G1 immunoadhesin mediates phagocytos is of malar ia-infected erythrocytes. J. Exp. Med. 176:1471. 43 St rassman G , Spr inger TA , Haski l l S J , Miragl ia C C , Lanier LL , and A d a m s D O (1985) Ant igens assoc ia ted with the activation of murine mononuc lear phagocytes in vivo: differential express ion of lymphocyte funct ion-associated antigen in the severa l s tages of development . Cell. Immunol. 94:265. S t rassman G , Spr inger TA , Somers S D , and A d a m s D O (1986) Mechan i sm of tumor cell capture by activated macrophages: ev idence for involvement of lymphocyte funct ion-associated (LFA)-1 antigen. J. Immunol. 136:4328. T a k e d a A (1987) Sialylat ion patterns of lymphocyte funct ion-associated antigen 1 (LFA-1) differ between T and B lymphocytes. Eur. J. Immunol. 17:281. Take i F (1985) Inhibition of mixed lymphocyte response by a rat monoc lona l antibody to a novel murine lymphocyte activation antigen (MALA-2) . J. Immunol. 134:1403. te V e l d e A A , Ke izer G D , and Figdor C G (1987) Differential function of LFA-1 family molecu les ( C D 1 1 a and CD18) in adhes ion of human monocytes to me lanoma and endothel ial cel ls. Immunology 61 :261. Tomass in i J E , G r a h a m D, DeWitt C M , Lineberger DW, Rodkey J A , and Co lonno R J (1989) c D N A cloning reveals that the major rhinovirus receptor on H e L a cel ls is intercellular adhes ion molecule 1. Proc. Natl. Acad. Sci. USA 86:4907. Turner M L (1992) Cel l adhes ion molecules: a unifying approach to topographic biology. Biol. Rev. 67:359. V a n der V ie ren M, Trong HL, W o o d C L , Moore P F , St. John T, Staunton D E , and Gal lat in W M (1995) A novel leukointegrin, a d p 2 , binds preferentially to ICAM-3 . Immunity 3:683. van Kooyk Y , van de Wei l -van K e m e n a d e P, Wede r P, Kuijpers T W , and Figdor C G (1989) Enhancement of LFA-1-media ted cell adhes ion by triggering through C D 2 or C D 3 on T lymphocytes. Nature 342:811. V a z e u x R, Hoffman P A , Tomita J K , Dickinson E S , J a s m a n R L , St. John T, and Gal lat in W M (1992) Clon ing and characterization of a new intercellular adhes ion molecule I C A M - R . Nature 360:485. Voraberger G , Scha fe r R, and Stratowa C (1991) Cloning of the human gene for intercellular adhes ion molecule 1 and analys is of its 5'-regulatory region: induction by cytokines and phorbol esters. J. Immunol. 147:2777. Wal l is W J , Hickstein DD, Schwar tz B R , June C C , O c h s HD, Beatty P G , Klebanoff S J , and Har lan J M (1986) Monoclonal antibody-defined functional epi topes on the 44 adhesion-promot ing glycoprotein complex (Cdw18) of human neutrophils. Blood 6Z:1007. Wantz in G L , Ralfk iaer E, Avnstorp C , Cza jkowsk i M, Marlin S D , and Rothlein R (1989) Kinet ics and character izat ion of intercellular adhes ion molecule-1 ( ICAM-1) express ion on kerat inocytes in var ious inflammatory skin lesions and malignant cu taneous lymphomas. J. Am. Acad. Dermatol. 20:782. Wawryk S O , Novotny J R , Wicks IP, Wi lk inson D, Maher D, Salvar is E, W e l c h K, F e c o n d o J , and Boyd A W (1989) The role of LFA-1 / ICAM-1 interaction in human leukocyte homing and adhes ion. Immunol. Rev. 108:135. Wi l l iams A F (1987) A year in the life of the immunoglobul in superfamily. Immunol. Today 8:298. Wi l l iams A F and Barc lay A N (1988) The immunoglobul in superfamily - domains for cell sur face recognit ion. Annu. Rev. Immunol. 6:381. Wi lson R W , Bal lantyne C M , Smith C W , Montgomery C , Bradley A , O'Br ien W E , and Beaude t A L (1993) G e n e targeting yields a CD18-mutant mouse for study of inf lammation. J. Immunol. 151:1571. Wi lson R W , O'Br ien W E , and Beaudet A L (1989) Nucleot ide s e q u e n c e of the c D N A from the mouse leukocyte adhes ion protein C D 1 8 . Nucleic Acids Res. 17:5397. Wright S D , Wei tz J l , Huang A J , Levin S M , Si lverstein S C , and Loike J D (1988) Comp lemen t receptor type three ( C R 3 , C D 1 1 b / C D 1 8 ) of human polymorphonuclear leukocytes recognizes f ibrinogen. Proc. Natl. Acad. Sci. USA 85:7734. W o n g DA, Dav is E M , L e B e a u M, and Spr inger T A (1996) Clon ing and chromosomal local izat ion of a novel gene encoding a human (32-integrin a subunit. Gene 171:291. X u H, G o n z a l o J A , St. Pierre Y , Wi l l iams IR, Kupper T S , Cotran R S , Spr inger TA , and Gut ie r rez -Ramos J C (1994) Leukocytosis and resistance to sept ic shock in intercellular adhes ion molecule 1-deficient mice. J. Exp. Med. 180:95. Zagoury D, Bernard J , Th ierness N, Fe ldman M, and Berke G (1975) Isolation and character izat ion of individual functionally reactive cytotoxic T lymphocytes: conjugat ion, killing and recycling at the single cell level. Eur. J. Immunol. 5:818. 45 Chapter 2 Cloning and characterization of murine ICAM - 2 T h e work presented in this chapter appears in the following publ icat ions: i) O h h M, Smith C A , Carpeni to C , and Take i F (1994) Regulat ion of intercellular adhes ion molecule-1 gene express ion involves multiple m R N A stabil ization mechan i sms : effects of interferon-y and phorbol myristate acetate. Blood 84 :2632 ii) Carpeni to C , Ohh M, and Take i F (1995) Cloning and character izat ion of murine intercellular adhes ion molecule-2 ( ICAM-2): a functional and molecular analys is . Allerg. Immunol. (Life Sci. Adv.) 14:43 iii) X u H, Bickford J K , Luther E, Carpeni to C , Take i F, and Spr inger T A (1996) Character izat ion of murine intercellular adhes ion molecule-2. J. Immunol. 156:4909 2:1 Introduction The immune sys tem consists of a network of cel ls that maintain bas ic de fenses against microorganisms, parasi tes, and cancer cel ls. Leukocytes travel throughout the multicellular organism surveying for the presence of these d isease-caus ing pathogens. A s d i scussed in the previous chapter, many different cell surface proteins exp ressed by leukocytes have been shown to play a role in determining their functional activity. A m o n g the most important are members of the immunoglobul in (Ig) supergene family (Wil l iams and Barclay, 1988; Springer, 1990). T h e s e molecules share sequence and structural similarit ies with the variable and constant domains of Igs and contain at least one region of a conserved Ig-like tertiary protein structure, cal led a homology unit (Ig-like domain). The homology unit consists of approximately 100-110 amino ac ids which 46 fold into a sandwich of two anti-parallel (3-sheets (Amit et al., 1986; Wi l l iams, 1987; A lzar i et al., 1988; Wi l l iams and Barclay, 1988). This structure is stabi l ized by a conserved disulphide link between the two sheets. The domain may a lso harbor severa l potential glycosylat ion sites. Molecu les with these Ig-like domains may be soluble proteins, t ransmembrane proteins, or g lycophosphol ip id (GPI) at tached membrane proteins. The genomic structure of most of these molecu les shows that e a c h Ig-like domain is often encoded by one exon (Hunkapil ler and Hood , 1989). Introns separat ing the exons are all in phase. This supports the notion that a primordial gene, which encoded an Ig-like structure, dupl icated, t ransposed throughout the genome, and diverged from its ancesteral gene giving rise to the many different Ig-like structures encoded in the genome (Will iams and Barclay, 1988; Hunkapi l ler and Hood , 1989). A l though the members of the Ig superfamily perform a variety of funct ions, the central focus of their role in the immune system is cell sur face recognit ion. Var ious members are known to play a role specif ical ly in leukocyte function. A smal l subset of these molecu les display not only structural but a lso functional similarity. The members of the I C A M subfamily are t ransmembrane glycoproteins with varying numbers of Ig-like domains (S immons et al., 1988; Staunton et al., 1988; Horley et al., 1989; S iu et al., 1989; Staunton et al., 1989; X u et al., 1992; Fawcett et al., 1992; V a z e u x et al., 1992; de Fougero l les et al., 1993). A s descr ibed in the previous chapter, three I C A M s ( ICAM-1 , -2 , -3) have been identified. The key functional aspect which they share is the ability to bind a common counter-receptor expressed on leukocytes, L F A - 1 . 47 The three known LFA-1 counter-receptors ( ICAM-1, -2 , and -3), are shown to be typical type I membrane glycoproteins. Three I C A M s have been identified and character ized in the human sys tem. ICAM-1 and ICAM-2 have been character ized in the murine sys tem. The murine vers ions of the I C A M s are ass igned their designat ion based on structural homology, s ize similarity, sequence identity, and similar distribution patterns. A l though all three I C A M s have been shown to bind L F A - 1 , their relative contributions to LFA-1-dependent immune responses, particularly ICAM-2 and -3 , are still unclear. The physiological s igni f icance of the LFA-1 : ICAM-1 interaction has been examined thoroughly (Springer, 1990; Dustin and Springer, 1991). A s ment ioned previoulsy in chapter 1, studies on the L F A - 1 : I C A M - 2 interaction have not been definitive. Funct ional examinat ion of the LFA-1 : ICAM-2 interaction has involved purified ICAM-2 and ant i -TcR mAb crossl inking or antibody b lockade of an immune response (Damle et al., 1992; de Fougerol les et al., 1994) or ICAM-2 peptide studies on leukocyte activation and transendothel ial migration (Li et al., 1993a; Li et al., 1993b; Somersa lo et al., 1995; Li et al., 1995). This project takes a different approach in examin ing the signi f icance of the LFA-1 : ICAM-2 interaction. Sys tems to examine the role of ICAM-2 in T cell stimulation and transendothel ial migration were estab l ished. The effect on each of these aspects of the immune system were examined when the I C A M - 2 c D N A was introduced into these sys tems. W h e n this project w a s intiated, ICAM-2 in the murine system had not yet been identified. The first step towards examining the role of ICAM-2 in the immune sys tem 48 w a s to c lone the c D N A encoding the murine ICAM-2 . In this chapter, cloning and functional character izat ion of murine ICAM-2 is descr ibed. The results in this chapter demonstrate the high degree of similarity between human and murine ICAM-2 in their s e q u e n c e s and function. The binding properties of the murine ICAM-2 (purified protein and c D N A - e n c o d e d ) to murine LFA-1 are a lso examined. 2 . 2 Ma te r i a l s a n d M e t h o d s 2:2.1 Animals B A L B / c mice were purchased from Char les River C a n a d a , Q u e b e c , C a n a d a and maintained in the Joint An imal Facility of the B .C . C a n c e r Resea rch Centre . 2:2.2 Cell lines and antibodies The mouse fibroblast L cell line (Sanford et al., 1948) w a s maintained in Du lbecco 's modif ied minimum essent ia l media ( D M E M ) containing 1 0 % fetal calf serum ( F C S ) and antibiotics (50 U/ml penicill in and 50 u.g/ml streptomycin). NS-1 (Kohler et al., 1976), a B A L B / c p lasmacytoma cell line, and B W 5 1 4 7 (Ralph and Nako inz , 1973), an A K R thymic leukemia cell line, were maintained in D M E M containing 5 % F C S and antibiotics. The T lymphoma line E L - 4 (Old et al., 1965) w a s maintained in R P M I 1640 medium containing 10% F C S . Al l monoc lonal ant ibodies (mAb) were used as purified immunoglobul ins. The rat m A b YN1/1 .7 .4 ( l gG 2 a ) recognizes the murine ICAM-1 and has been descr ibed (Takei , 1985; Horley et al., 1989). The rat ant i -mouse ICAM-2 (3C4, l g G 2 a ) antibody 49 used for functional and inhibition studies was purchased from Pharmingen (San Diego, C A ) and the antibody used for the purification of ICAM-2 w a s a generous gift from Dr. T. Spr inger (Centre for Blood Resea rch , Harvard Medica l School ) , and has been character ized (Xu et al . , 1996). The rat hybr idoma cell l ines FD441 .8 (Dialynas et al., 1982) which produces ant i -mouse LFA-1 (CD11a , A T C C TIB 213, l g G 2 b ) and R1-2 (Ho lzmann and W e i s s m a n , 1989; Ho lzmann et al., 1989) which produces rat anti-mouse V L A - 4 (CD49d , A T C C HB227 , l g G 2 b ) were obtained from Amer i can Type Culture Col lect ion (Rockvi l le, MD) . 2:2.3 PCR cloning a murine specific probe for ICAM-2 Total R N A from E L - 4 cel ls was isolated using the ac id-phenol extraction method (Chomczynsk i and S a c c h i , 1987). Ce l l s (5 x 10 6) were w a s h e d twice with phosphate buffered sal ine ( P B S ) and dissolved in 0.5 ml of solution D (4 M guanidinium thiocyanate, 25 m M sodium citrate pH 7.0, 0 .5% sarkosy l , 0.1 M 2-mercaptoethanol) . Fol lowing this, 50 uJ of 2 M sodium acetate pH 4.0, 0.5 ml water-saturated-phenol, and 0.1 ml chloroform:isoamyl alcohol (49:1) were added. This solution w a s centri fuged at 14,000 g for 15 min at 4°C. The aqueous layer, which w a s free of genomic D N A , was mixed with 0.5 ml isopropanol and put at -70°C for 1 hr. The solution was centri fuged at 14,000 g for 20 min at 4°C and the pellet was then suspended in 0.3 ml solution D fol lowed by 0.3 ml of ethanol. The solution was kept at -70°C for another hour and then centri fuged for 20 min at 4°C. The pellet was then washed twice with 1 ml 7 0 % ethanol and finally suspended in 20 uJ of diethylpyrocarbonate (DEPC)- t rea ted water. 50 Total R N A (5 |ug) from E L - 4 cel ls was combined with 1.5 u.g ol igo-dT primers (12 -18 m e r ) in a final vo lume of 30 ul containing 50 m M Tris pH 8.3, 60 m M KCI, 3 m M M g C I 2 , 10 m M DTT, 500 u M d N T P s (Pharmacia ; Ba ie d'Urfe, P Q ) , 4 ug acetylated B S A (Promega; Mad ison , Wl) , 40 units of R N a s i n (Promega) and 600 units of Moloney murine leukemia virus reverse transcriptase (Canadian Life Techno log ies ; Burl ington, ON) . This mixture was incubated at 42°C for 1 hr to generate s ingle-st randed c D N A (Sambrook et al., 1989). In a final vo lume of 50 ul, 1/5 of the above c D N A reaction w a s combined with the degenerate ol igonucleot ides 5 ' - G T C A A C T G ( C / T ) A G ( C / T ) ( A / T ) C C ( A / T ) C ( A / C ) T G - 3 ' and 5 ' - C A C A G ( C / T ) ( C / G ) ( A / C ) G ( A / G ) C A G G A G A A ( A / G ) T T - 3 ' (1 u.M concentrat ion each) in a mixture containing 10 m M Tris pH 8.3, 50 m M KCI, 1.5 m M M g C I 2 , 0 . 1 % gelatin, and 1.5 units of Taq po lymerase (Canad ian Life Techno log ies) . This cocktai l mix was subjected to 30 rounds of po lymerase chain amplif ication ( P C R ) (94°C, 30 sec ; 50°C, 60 sec ; 72°C, 60 sec) . The amplif ied D N A fragment (450 bp) w a s then purified by agarose gel e lectrophoresis. It w a s then blunt ended by incubating the fragment in 50 m M Tris pH 7.5, 10 m M MgCI 2 , 5 m M DTT, 500 U.M d N T P s , and 5 units of T 4 D N A polymerase (Canadian Life Technolog ies) at 37°C for 5 min. The ends of the fragment were then phosphorylated with 5 units T 4 polynucleot ide k inase (Canadian Life Technologies) in 60 m M Tris pH 7.6, 10 m M M g C I 2 , 2.5 m M DTT, 1 m M A T P at 37°C for 1 hr. Finally, the P C R fragment w a s ligated into a Smal cut Bluescipt vector (pBST) (Pharmacia) with 2 units of T 4 D N A l igase (Canad ian Life Technolog ies) in 50 m M Tris pH 7.6, 10 m M M g C I 2 , 1 m M A T P , 1 m M DTT, 5 % P E G - 8 0 0 0 at 4°C for 8-12 hrs. The ligation mixture w a s then 51 t ransformed into competent E. coli D H 5 a cells (Canadian Life Technolog ies) and plated on L B agar containing ampicil l in (50 u.g/ml) and colour select ion with 0 .02% I P T G and 0.04% X-ga l . Individual colonies were subsequent ly p icked for plasmid mini-preps and sequenced by the Sanger dideoxynucleot ide chain termination method (Sanger et al., 1977) using the T 7 sequenc ing kit (Pharmacia) , [a- 3 2 P]-dCTP (3000 Ci /mmol) , T 3 and T 7 primers. 2:2.4 DNA isolation and Southern blot analysis High molecular weight D N A was isolated from B A L B / c sp leen cel ls (Gross -Bel lard et al., 1977). The cel ls (1 x 10 8) were washed 3 X with P B S and suspended in 2 ml of T N E buffer (10 m M Tris pH 8.0, 0.15 M N a C l , 10 m M E D T A ) and gently mixed. To this suspens ion , 20 pi of 2 0 % S D S and 100 u.g of proteinase K (S igma Chemica l C o m p a n y ; St. Louis, MO) were added and incubated at 37°C for 12 hrs. The lysate w a s then extracted 3 X TNE-satura ted-phenol , 3 X with TNE-sa tu ra ted -(phenolxhloroform) (1:1), and finally 3X with chloroform:isoamyl a lcohol (24:1). The aqueous phase was dia lyzed against two changes of 4 I of T E buffer (10 m M Tris pH 8.0, 1 m M E D T A ) each for 16 hrs at 4°C. The D N A w a s quantitated spectrophotometr ical ly ( A 2 6 0 = 1.00 for 50 ucj/ml) and purity was determined (A 2 6 o/A 2 8o ~ 2.0) on an L K B U L T R O S P E C 4050 (Cambridge, England). P lasmid mini-preps were prepared by alkaline lysis (Brinboim and Dolby, 1979). Bacter ia l cultures were grown in 2 ml of L B broth overnight at 37°C and pel leted. The pellet was suspended in 0.2 ml G T E buffer (50 m M glucose, 25 m M Tris pH8.0 , 10 m M 52 E D T A ) containing 4 mg/ml of lysozyme and incubated for 5 min at room temperature. To this mixture, 0.4 ml of 1% S D S , 0.2 M N a O H was added and incubated on ice for 5 min. Finally, 0.3 ml of 3 M potassium acetate, 2 M acet ic acid was added and the mixture was centrifuged (14,000 g) for 20 min at 4°C. The clear supernatant was recovered and extracted with 0.6 ml a phenol:chloroform:isoamyl a lcohol solution (25:24:1), and mixed with 1 ml of isopropanol. Nucle ic ac ids were al lowed to precipitate by incubating on ice for 10 min fol lowed by spinning in a microfuge for 15 min at 4°C. The pellet was then dissolved in 0.1 ml of T E buffer containing 200 Lig/ml of R N a s e A (S igma Chemica l Co. ) , 20 U/ml R N a s e Tj (Boehr inger Mannhe im; Lava l , P Q ) and incubating first at 37°C for 30 min and then at 50°C for 5 min. To the aqueous phase , 60 jal of 7.5 M ammonium acetate pH 7.0 was added and incubated on ice for 5 min. The solution was centrifuged for 5 min and the supernatant recovered was mixed with 160 uJ of ethanol. This was then centrifuged for 5 min at room temperature and the pellet was washed with 1 ml of 7 0 % ethanol. The D N A was then d isso lved in 50 uJ of T E buffer and ready for either sequenc ing or restriction enzyme analys is . Large sca le plasmid purification was done in a slightly different fashion. Overnight bacterial cultures grown in 500 ml of L B broth were pelleted and suspended in similar fashion as the minipreps. They were lysed and neutral ized as before. However , the isopropanol-precipitated nucleic ac ids were then purified by the Wizard Maxiprep kit (Promega). The isolated D N A was then ethanol precipitated and washed with 7 0 % ethanol before being d issolved in sterile T E buffer. 53 The restriction enzymes used for restriction digest analys is were purchased from either Canad ian Life Technolog ies , Pharmac ia , or Boehr inger Mannhe im. The condit ions and temperature for the digest ions were those recommended by the manufacturer. Approximately 1 u.g of plasmid D N A was digested with 5 units of e n z y m e for 2-3 hrs. High molecular weight genomic D N A (12 u.g) was digested with a large e x c e s s of enzyme(s) (10 units/ucj of genomic DNA) for 6-8 hrs. Both types of samp les were then precipitated by adding 1/10 volume of 2.5 M sod ium acetate pH 4.5 and 2 vo lumes of cold ethanol, and incubated on dry ice for 20 min. The solution was microfuged (14,000 g) for 20 min at 4°C and the pellet washed with 1 ml of 7 0 % ethanol . The air dried pellet was then redissolved in 10-30 uJ of T A E buffer (40 m M Tris acetate pH 7.2, 20 m M sodium acetate, 1 m M E D T A ) . To this, 1/10 vo lume D N A loading buffer (0.25% xylene cyanol , 0 .25% bromophenol blue, 5 0 % glycerol in T A E ) was added and loaded onto an 0.8% T A E agarose gel. P lasmid digests were e lect rophoresed at 50 volts for 3-4 hrs. Digested genomic D N A was e lect rophoresed at 25 volts for 16-18 hrs. Ethidium bromide (Canadian Life Technolog ies) w a s present in both the agarose gel and the T A E running buffer at a concentrat ion of 500 ng/ml. Molecu lar weights were determined from molecular weight s tandards such as Hindlll digested X D N A and X (Hindlll + EcoRI) fragments. The gels were photographed under U V light with a Polaroid camera . The Southern blot hybridization w a s done as descr ibed (Southern, 1975). The gel was first treated in 0.1 M HCI for 15 min fol lowed by treatment in 0.5 M N a O H , 1.5 M N a C l for 30 min, and finally neutral ized in 1.0 Tris pH 7.0, 2.5 M N a C l for an addit ional 30 min (Sambrook et al., 1989)). The gel was 54 then layered on top of a strip of 3 M M Whatman filter paper (Schle icher and Schue l l ; K e e n e , NH) soaked in 2 0 X S S C (3 M N a C l , 0.15 M sodium citrate pH 7.0). The ends of the filter paper are soaked in a pool of 2 0 X S S C buffer. A p iece of Ze taProbe G T membrane (B ioRad Laborator ies; M iss i ssauga , ON) cut to the s ize of the gel w a s first soaked in double-dist i l led water for 10 min and then layered on top of the gel . Four wetted p ieces of Whatman paper were then layered on top of the Ze taProbe membrane, fol lowed by a stack of paper towels (~10 cm) cut to the s a m e s ize as the gel and a g lass plate. The D N A was covalently fixed to the membrane by U V c ross-linking in a U V Stratal inker 1800 (Stratagene; La Jol la , C A ) . The membrane was prehybridized in a solution of 6 X S S C , 1 0 % deion ized formamide (Canad ian Life Technologies) , 1% S D S , 1% Blotto (Carnat ion instant sk im milk), 2 m M E D T A , 0.5 mg/ml heat denatured sa lmon sperm D N A (Pharmacia) at 60°C for 4 hrs. D N A probes were rad io labe led by the procedure deve loped by Feinberg and Voge ls te in (Feinberg and Vogels te in , 1983) using the T 7 ol igolabell ing kit (Pharmacia) . D N A (20 ng) was mixed with 100 ng of randomly generated hexanucleot ides and boiled for 3 min and immediately cooled on ice. In addit ion to the cocktai l buffer, a nucleotide mix (dATP, d G T P , d T T P , [a- 3 2P]-dCTP) w a s a lso added with T 7 po lymerase (final vo lume of 10 ul) and al lowed to proceed accord ing to manufacturer 's protocol. The reaction was stopped by the addition of 2 u.l 10 M N a O H . The ol igolabel led-probe was then added to the hybridization solution containing 6 X S S C , 1 0 % deion ized formamide, 1% S D S , 1% Blotto, 2 m M E D T A , 0.5 mg/ml denatured sa lmon sperm D N A , and 10% dextran sulfate (Turhan et al., 1988). The 55 filter w a s incubated in this solution at 60°C for 16 hrs and then washed 3 X for 30 min e a c h at 65°C in 0.3X S S C , 0 .1% S D S , 0 .1% sodium pyrophosphate. The filter was then exposed to Kodak X A R film at -70°C for 16-36 hrs. 2:2.5 Isolation of a cDNA clone and sequence analysis The P C R fragment (described in 2:2.3) was used to screen a (B6 x CBA)F. , mouse lung c D N A library in the XZAP II vector (Stratagene; Short et al., 1988). The library w a s titrated according to manufacturer 's protocol. Briefly, X L 1 - B l u e bacterial cel ls were st reaked on L B agar to obtain a single colony which w a s used to inoculate 40 ml of L B med ia supplemented with 0 .2% maltose and 10 m M M g S 0 4 . T h e s e cel ls were grown overnight and resuspended in 20 ml of 10 m M M g S 0 4 . A dilution from the phage library (~10 5 phage particles) was incubated with 5 ml of the host cel ls at 37°C for 15 min, mixed with 50 ml of soft agar ( N Z Y C M + 0.8% agar) and plated on N Z Y C M + 1.5% agar. T h e s e plates were incubated overnight at 37°C. The plates were chil led at 4°C for 2 hrs before being lifted onto duplicate nitrocellulose sheets . The sheets were first wetted in water and then in 1 M N a C l before being laid on the plates for 4 min. E a c h plate was lifted in duplicate and the orientation w a s marked with india ink. The filters were then p laced, D N A side up, on Whatman paper saturated with 1.5 M N a C l , 0.5 M N a O H for 4 min then transferred to another Wha tman paper saturated with 1 M Tris pH 7.0, 3 M N a C l for 4 min (Sambrook et al., 1989). The filters were then baked at 80°C for 2 hrs and incubated in a prewash step (50 m M Tris pH 8.0, 1 M N a C l , 1 m M E D T A , 0 .1% S D S for 1 hr at 42°C) to remove bacterial debris. The filters 56 were then incubated in a prehydrization solution of 6 X S S C , 0 .5% S D S , 1% Blotto, 0.5 mg/ml denatured sa lmon sperm D N A for 2 hrs at 50°C. The filters were then transferred into a hybridization solution (same composi t ion as above plus the rad io labe led probe) for 14 hrs at 50°C. Three w a s h e s of 20 min each at 50°C were then done: a) 5 X S S C , 0 .1%, b) 2 X S S C , 0 .1% S D S , c) 0.5X S S C , 0 . 1 % S D S . The filters were then exposed to Kodak X A R film. The autoradiographs were al igned with their respect ive dupl icates and only spots which appeared in both lifts of the s a m e plate were chosen for the next round of screening which was cont inued until all p laques on the plates hybridized with the probe. The X Z A P II vector has the unique ability to al low in vivo exc is ion and recircularization of any c loned insert within the X vector to form a phagemid containing the insert in p B S T ( S K ) (Short et al., 1988). Inside E. coli, proteins provided from a helper phage recognize D N A synthesis initiation and termination s e q u e n c e s in the X ZAP II. The end result is to d isplace the single-stranded D N A between the two sites recognized by the phage proteins. The displaced single-stranded D N A is then circular ized and packaged as a phagemid which can infect bacterial host cel ls. The phagemid can propagate as double-stranded p B S T with the c loned insert inside the cel l and will appear as a bacterial colony. The plasmid D N A can be isolated as descr ibed above and used for sequenc ing and restriction enzyme digest ion. S e q u e n c e data was generated from sequenc ing both strands of the insert. 57 2:2.6 RNA isolation and Northern blot analysis Total R N A was isolated from the various lymphoid and non-lymphoid organs (Davis et al., 1986). Sp leen , thymus, and bone marrow were teased with tweezers and p a s s e d through a syr inge to generate single cell suspens ions . Heart, lung, kidney, and liver were cut into smal l p ieces and immersed into liquid nitrogen. The frozen p ieces were then minced with a mortar and pestle until a fine powder was generated and al lowed to thaw before being lysed in a guanidinium isothiocyanate solut ion (4 M guanid in ium isothiocyanate, 25 m M sodium acetate pH 6.0, 0.12 M 2-mercaptoethanol) . The single cell suspens ions of sp leen, thymus, and bone marrow were a lso lysed in this s a m e solution. The lysate was then layered on a caes ium chlor ide cushion (5.7 M CsCI , 25 m M sodium acetate pH 6.0) and centri fuged in a B e c k m a n L8 -60M ultracentrifuge (174,000 g in an SW41 rotor for 20 hr at 20°C). The denser R N A is separated from the D N A , proteins, and lipids and isolated at the bottom of the cush ion. This pellet was d issolved in 20 u.l of DEPC- t rea ted doubly-disti l led water and an aliquot (10 u.g R N A ) was combined with R N A loading buffer such that the final solution contained 5 0 % deionized formamide, 36 m M 3-[N-Morphol ino]-propane-sulfonic acid (1X M O P S ) pH 7, 7% formaldehyde, 6% glycerol, 0 .5% bromophenol blue. The samp les were heated to 95°C for 2 min and then loaded into the wel ls of a 1% agarose gel in 1X M O P S buffer with 2 % formaldehyde and 1X M O P S as the running buffer. The gel was run at 120 volts for 3-4 hrs. Molecular weights were determined from the R N A ladder (Canadian Life Technologies) . After e lectrophoresis, the lane with the R N A ladder was cut from the rest of the gel and washed twice in 500 58 ml of 10X S S C for 15 min each . The gel s lab was then stained in 200 ml 1X S S C containing 1 L i g / m l Ethidium bromide and destained in 2 changes of 500 ml of 1X S S C for 1 hr e a c h . The slab was then photographed under U V light. The gel with the isolated R N A samples was rinsed for 20 min each in two changes of 500 ml of 10X S S C in order to remove the formaldehyde. R N A in the gel was then transferred onto a Ze taP robe membrane by capil lary action for 16 hrs using 10X S S C . The membrane was then rinsed in 2 0 X S S C and U V fixed with a U V Stratal inker 1800. After this, the membrane was incubated in a prehybridization solution containing 1.5X S S P E (20X S S P E : 3.6 M N a C l , 0.2 M N a H 2 P 0 4 pH 7.4, 20 m M E D T A ) , 1% S D S , 0 .5% Blotto, 0.5 mg/ml heat denatured sa lmon sperm D N A for 2 hrs at 60°C. The membrane was then transferred to a hybridization solution containing 1.5X S S P E , 1% S D S , 0 .5% Blotto, 0.5 mg/ml heat denatured sa lmon sperm D N A , and ol igolabel led probe for 16 hrs at 60°C. The filter w a s then washed in a) 2 X S S C , 0 . 1% S D S at 20°C for 15 min, b) 0.5X S S C , 0 . 1 % S D S at 20°C for 15 min, c) 0.3X S S C , 0 . 1 % S D S 65°C for 30 min. The filter w a s then exposed to Kodak X A R film. The filter w a s then str ipped of the rad io labe led probe by boiling in three changes of 500 ml each of 0.1 X S S C , 1% S D S for 20 min. The filter was then subjected to another prehybridization and probing with another ol igolabel led probe. 2:2.7 Genomic cloning of the murine ICAM-2 High molecular weight D N A from B A L B / c sp leen cel ls was digested with EcoRI. The D N A w a s electrophoresed on a 0.8% agarose T A E gel and three s ize-se lec ted 59 (4.0-5.5 kb, 5.5-7.0 kb, 7.0-9.0 kb) segments were isolated by electro-elution. Al iquots of e a c h fraction were electrophoresed on a 0.8% T A E agarose gel at 35 volts for 5 hrs and Southern blotted as descr ibed above. The filter was then probed with the isolated I C A M - 2 c D N A G 3 - 1 . 1 . The fraction (5.5-7.0 kb) with the most intense band (~ 6.5 kb) w a s ligated into EcoRI cut A.gt10 arms (750 ng genomic D N A with 5 u.g X arms) and packaged as phage particles using the X Packag ing Extract (Invitrogen; S a n Diego, C A ) . The library was screened as descr ibed above for the c D N A cloning, except that the host bacterial strain was E. coli C600Hf l . The probe was generated from the c loned c D N A , G 3 - 1 . 1 . Severa l genomic c lones were isolated (three rounds of screening) and subc loned into EcoRI cut p B S T . The c lones were found to have the s a m e restriction enzyme pattern and thus only one, B M 1 - 1 . 1 , w a s ana lyzed further by partial sequenc ing and restriction enzyme mapping. 2:2.8 Purification of cell surface ICAM-2 The purification of murine ICAM-2 from B W 5 1 4 7 cel ls required two preliminary s teps, biotinylation of smal l amount of B W 5 1 4 7 and the coupl ing of the ant i -mouse ICAM-2 (3C4) antibody to Affi-gel 10 beads. In the coupl ing process, 5 mg of purified 3 C 4 ant ibody in 0.1 M sodium bicarbonate buffer pH 8.0, 0 .85% N a C l w a s incubated with 2.5 ml of Affi-gel 10 beads (BioRad) prewashed with 10 ml with ice cold isopropanol and then 20 ml of ice cold double-disti l led water. The beads were then incubated with the antibody overnight with gentle agitation at 4°C. Res idua l succ in imide esters were inactivated with the addition of 0.1 M Tris pH 7.5. In the other 60 step, B W 5 1 4 7 cel ls were washed twice with biotin labeling buffer ( H B S S minus phosphates , plus 10 m M N a H C 0 3 pH 7.4) at 4°C. Immediately before use, a stock solution of sul fo-N-hydroxy succin imide biotin (sNHS-biot in) (Pierce Chemica l C o . ; Rockford , IL) was prepared at 10 mg/ml in labelling buffer. B W 5 1 4 7 cel ls (1 X 10 7) were suspended in 1 ml of labelling buffer and 40 u.1 of biotin stock solution w a s added and gently vortexed (von Boxberg et al., 1990). The cel ls were then incubated on ice for 20 min with occas iona l shaking fol lowed by five w a s h e s of 30 ml of D M E M + 2 % F C S . After label ing, the cel ls were d issolved in lysis buffer containing 1.0% nonidet P-40 (NP-40) , 120 m M N a C l , 4 m M MgCI 2 , 20 m M Tris pH 7.5, 4 ug/ml phenylmethylsul fonyl fluoride ( P M S F , S igma Chemica l Co. ) , and 50 m M L-lys and incubated on ice for 1 hr with occass iona l vortexing. The solution was then microfuged at 14,000 g for 10 min to pellet nuclei and other debris. The resulting supernatant (1 ml) containing biotinyalted proteins was ready to be used as a tracer in the subsequent purification process . The large sca le purification of murine ICAM-2 is similar to that descr ibed for the murine ICAM-1 (Horley et al., 1989). B W 5 1 4 7 cel ls were grown in spinner bottles containing D M E M + 5 % F C S equil ibrated with 5 % C 0 2 . The cel ls were harvested when the cell density reached 2-3 X 10 6 per ml. Harvested B W 5 1 4 7 cel ls (2 X 10 9) were washed twice with 50 ml of P B S and suspended in 40 ml 10 m M Tris pH 7.5. The cel ls were then passed through a 26 gauge needle three t imes. Nucle i were removed by low speed centrifugation (1,100 R P M in a B e c k m a n T J - 6 centrifuge) for 15 min at 4°C. After spinning at 19,000 R P M (oakridge tubes in a J A - 2 0 rotor) for 50 min 61 at 4°C, the pellet was resuspended in 50 ml of ice cold 10 m M Tris pH 7.5, 1 m M E D T A , 0 .85% N a C l and homogenized on ice for 5 min with a homogenizer . A n equal vo lume of 10 m M Tris pH 7.5, 1 m M E D T A , 0 .85% N a C l , 2 % Triton X - 1 0 0 containing 2 m M P M S F , 10 Lig /ml leupeptin, and 2 u.g/ml aprotinin was added and combined with lysate from cell surface biotinylated B W 5 1 4 7 cells (tracer). The solubi l ized membrane fraction w a s then mixed with a magnet ic stir bar for 30 min on ice and centr i fuged at 10,000 R P M for 30 min. The remaining supernatant was then combined with the 3 C 4 coupled Affi-gel 10 beads for 8 hrs under constant agitation at 4°C. The beads were then w a s h e d with 400 ml of 1% Triton X -100 , 10 m M Tris pH 7.5, 1 m M E D T A , 0 .85% N a C l fol lowed by 100 ml of 0 . 1% Triton X -100 , 10 m M Tris pH 7.5, 1 m M E D T A , 0 .85% N a C l , both at 4°C. Bound protein was eluted with 0.1 M glycine pH 2.9, 0 . 1 % Triton X -100, 0 .85% N a C l and each 1 ml fraction col lected was neutral ized with 0.5 M N a 2 C 0 3 . A 10 L I I aliquot from each fraction was subjected to SDS-po lyac ry lamide gel e lectrophoresis ( S D S - P A G E ) as outlined previously (Sambrook et al., 1989). The fractions were ana lyzed on a 10% polyacrylamide gel and electrophoretical ly transferred onto an Immobilon-P membrane (Millipore, Bedford, MA) by electroblotting at 350 m A for 2 hrs using a Tr is-glycine-methanol transfer sys tem (0.192 M glycine, 25 m M Tris pH 8.3, 2 0 % methanol) (Liu et al., 1994). The membrane was then placed in blocking buffer (2% B S A , 0 .05% Tween-20 in P B S ) for 2 hrs at room temperature and w a s h e d three t imes in wash buffer (0.1% B S A , 0 .05% Tween-20 in P B S ) . The filter w a s then incubated in wash buffer containing streptavidin conjugated to horse radish perox idase (1:10,000 dilution) (Cedar lane; Hornby, ON) for 30 min at room 62 temperature. After three w a s h e s of five min each in wash buffer, the membrane was w a s h e d in P B S for five min and place in an enhanced chemi luminescence solution (Amersham; Oakvi l le, ON) for 1 min. The membrane was exposed to X-ray film (Kodak X A R ) for 2-15 min and all fractions which d isplayed a 50-55 kD band were pooled and concentrated in a Centr icon C-30 (Amicon; Beverly, MA) . The purity and yield w a s a s s e s s e d by S D S - P A G E and silver staining of the gel (Ohsawa and Ebata , 1983). The purified protein was quantitated by compar ison of band intensity with that of known amounts of B S A . 2:2.9 Isolation and activation of murine splenic T lymphocytes Murine sp lenic T cel ls were isolated from B A L B / c mice, 8-12 weeks old, using a nylon wool (Po lysc iences , Warr ington, PA) column as descr ibed (Julius et al., 1973). Briefly, 2 gm of nylon wool was steril ized by autoclave and incubated overnight in a packed co lumn containing RPMI 1640 + 5 % F C S at 37°C in a 5 % C 0 2 humidified a tmosphere. S ing le cell suspens ion of sp leen cel ls were then incubated in the nylon wool co lumn for 1 hr. Non-adherent T cel ls were then eluted with R P M I 1640 + 5 % F C S at a flow rate of ~ 1 ml/min and the contaminating red blood cel ls were lysed with a Tr is -ammonium chloride solution (17 m M Tris, 140 m M N H 4 C I , pH 7.2; Hunt, 1979). The cel ls were then incubated in RPMI 1640 + 5 % F C S containing 50 ng/ml phorbol 12-myristate 13-acetate (PMA) for 20 min at 37°C at a concentrat ion of 1 x 1 0 6 cel ls/ml (Pyszn iak et al., 1994). Fol lowing this, the T cel ls were washed three t imes with Hank 's Ba lanced Salt Solut ion ( H B S S ) and labeled with the f luorescent dye Ca l ce in -63 A M (Molecular Probes , Eugene . O R ) according to the manufacturer 's protocol. T h e s e labeled T cel ls were then ready to be used in subsequent adhes ion assays . 2:2.10 Binding of splenic T cells to purified ICAM-2 Purif ied murine ICAM-2 was covalently coupled to microwell plates (Falcon 3072, Bec ton Dick inson, Lincoln Park, NJ) as descr ibed previously (Horley et al., 1989). Purif ied recombinant soluble ICAM-1 , provided by Andrew Pyszn iak (Terry Fox Laboratory, B C C a n c e r Resea rch Centre), (Welder et al., 1993) was a lso coupled to microwel ls and used as a positive control. Wel ls of 96-well microplates were treated with 100 u.l 0 .2% glutaraldehyde, 0.1 M sodium carbonate-HCI pH 9.0 for 1 hr at room temperature. The wel ls were then washed three t imes with 0.1 M sod ium carbonate-HCI pH 9.0 buffer and 50 u.l of poly-L- lysine (50 u.g/ml) in 0.05 M sod ium carbonate pH 9.0 buffer. The plates were then incubated for 2 hr at room temperature and washed three t imes with 0.05 M sodium carbonate pH 9.0 buffer. The wel ls then received 50 ul of 0 .2% glutaraldehyde in 0.05 M sodium carbonate pH 9.0 for 1 hr at room temperature and washed three t imes with 0.05 M sodium carbonate pH 9.0 buffer. E a c h well then received 25 ul of either ICAM-1 (100 ng), ICAM-2 (100 ng), or 1% ovalbumin in 0.1 M sodium carbonate buffer pH 9.0. After a 2 hr incubation at room temperature, the wel ls were washed and received 150 ul of 1% B S A in 0.1 M sod ium carbonate buffer in order to neutralize the free glutaraldehyde groups. The P M A -activated Ca lce in- labe led splenic T cel ls suspended in H B S S / 5 % F C S and d ispensed into the wel ls (10 5 cells/well) in a final volume of 100 u l Al l condit ions were tested in 64 triplicate. Un less otherwise stated, anti- ICAM-1 A b or ant i - ICAM-2 A b were added to the wel ls 15 min prior to the addition of the cel ls at a concentrat ion of 2 Lig /ml or 4 Lig /m l , respectively. An t i -CD11a antibody (2 Lig/ml) was incubated with sp lenic T cel ls for 15 min at 37°C prior to addition to the wel ls. The plates were centri fuged at 300 g for 1 min, incubated for 8 min at 37°C, and washed five t imes with H B S S + 5 % F C S (37°C). F luo rescence of the remaining bound cel ls was measured by a CytoF luor 2300 (Millipore) and compared with a standard curve. 2:2.11 Cloning of ICAM-2 cDNA into expression vector The murine ICAM-1 (Horley et al., 1989) and ICAM-2 c D N A s were c loned into the mammal ian express ion vector p B C M G S (Karasuyama et al., 1990). The c D N A s were oriented in p B S T ( S K ) such that the 5' end of the c D N A w a s adjacent to the Xhol site and the 3' end was adjacent to the Notl site. The insert w a s liberated from p B S T by digest ion with Notl and Xhol and was c loned into p B C M G S digested with Notl and Xhol, with the Xhol site being c losest to the promoter. The selectable markers for p B C M G S are ampicil l in resistance in bacteria (plated on L B agar with 50 Lig /ml of ampicil l in) and neomycin resistance in eukaryotic cel ls (selected in D M E M + 5 % F C S + 0.5 mg/ml G418) . p B C M G S is an ep isomal express ion vector where replication is based on a bovine papi l loma virus replication sys tem. It maintains a relatively high copy number (20-100 copies/cel l ) and the C M V - b a s e d promoter /enhancer sys tem aid in high express ion of the insert. 65 < 2:2.12 Binding of splenic T cells to L cell transfectants expressing I CAMs L cel ls were transfected with either murine ICAM-1 (Horley et al . , 1989) or murine ICAM-2 c D N A in the express ion vector p B C M G S by the poly-L-ornithine (S igma Chemica l Co.) method (Dong et al., 1993). Transfectants were then selected in D M E M + 5 % F C S containing 0.5 mg/ml G 4 1 8 for 5-8 days. G418-res is tant cel ls were then ana lyzed for express ion of ICAM-1 and ICAM-2 by flow cytometry. The L cell t ransfectants were grown for 2 days in D M E M + 5 % F C S containing 0.5 mg/ml G 4 1 8 . The cel ls were harvested with P B S + 2.5 m M E D T A and incubated with anti-ICAM-1 or ant i - ICAM-2 ant ibodies (3 X 1 0 5 cel ls and 4 |^g/ml antibody in 100 uJ H B S S + 2 % F C S + 0 . 1 % N a N 3 ) on ice for 20 min. The secondary stain was goat (Fab) 2 -ant i -rat IgG conjugated to f luorescein isothiocyanate ( G a R l g G - F I T C ) (Cooper B iomedica l , W e s t Chester , P A ) . Dead cel ls were stained with 2 u.g/ml propidium iodide. The staining w a s ana lyzed on a F A C S t a r (Becton-Dickson). The L cel ls express ing ICAM-1 or ICAM-2 were plated in 96-well flat-bottom microwell plates (Falcon 3072, Becton Dickinson) in 200 ul of D M E M + 5 % F C S and grown for 2 days to a subconfluent monolayer. Adhes ion of PMA-ac t iva ted splenic T cel ls to these monolayers was examined in the s a m e way as descr ibed above for adhes ion to purified ICAM-2 . 66 2:3 R e s u l t s 2:3.1 PCR sequence analysis The nucleotide identity between the human and murine ICAM-1 is low (50%) (Horley et al., 1989; S iu et al., 1989) and using the human c D N A as a probe to clone the murine c D N A presented some difficulties. In order to avert such problems in cloning the murine ICAM-2 c D N A , a murine specif ic probe was preferred. Th is was ache ived by exploiting the conserved sequences among the known members of the I C A M subfamily. Am ino acid sequences of human ICAM-1 , ICAM-2 , and murine ICAM-1 (only ones avai lable when project began) d isplayed a high degree of sequence identity around the cysteine residues that form disulphide l inkages between the two p-sheets of their Ig-like domains. These covalent bonds are thought to stabi l ize the Ig-like structure and are an excel lent point to begin analysis of conservat ion of sequences among members of the I C A M subfamily. A s seen in Figure 1, there is a strong degree of conservat ion around the cysteine residues in domains 1 and 2 at both the amino acid and nucleot ide levels. W h e n the nucleotide sequence in these two regions is examined closely, it is apparent that there is a consensus sequence with very little degeneracy in these two regions conserved among all three I C A M members . In order to c lone a murine-speci f ic probe for ICAM-2 , these degenerate ol igonucleot ides were used to P C R amplify ICAM-l ike D N A fragments. E L - 4 cel ls were a good source for this amplif ication s ince they do not express detectable levels of ICAM-1 and thus any ICAM- l ike D N A fragments detected would be either ICAM-2 , ICAM-3 , or s o m e novel 67 a ) Regionl mICAM-1 DAQVSIHPREAFLPQGGSVQ|VNCSSSC|KEDLSLGLETQWLKDE- LESGPNWKLF 53 h I C A M - 1 QTSVSPSKVILPRGGSVLVTCSTSC 3QPKLLGIETPLPKKELLLPGNNRKVY 52 h I C A M - 2 SDEKVFEVHVRPKKLAVEPKGSLEVNCSTTC ^IQPEVGGLETSL - NKI LLDEQAQWKHY 57 P GS V CS C G ET L K mICAM-1 ELSEIGEDSSPLCFENCGTVQSSASATITVYSFPESVELRPLPAWQQVGKDLTLRCHV 111 h I C A M - 1 ELSNVQEDSQPMCYSNCPDGQSTAKTFLTVYWTPERVELAPLPSWQPVGKNLTLRCQV 110 h I C A M - 2 LVSNISHDTVLQCHFTCSGKQESMNSNVSVYQPPRQVILTLQPTLVAVGKSFTIECRV 115 S D C C Q V Y P V L P V G K T C V mICAM-1 h I C A M - 1 h I C A M - 2 mICAM-1 h I C A M - 1 h I C A M - 2 DGGAPRTQLSAVLLRGEEILSRQPVGGHPKDPKEITFTVLASRGDH--G-ANFSCRTE|166 EGGAPRANLTVVLLRGEKELKREPAVG — EPAEVTTTVLV-RRDHH-G-ANFSCRTE PTVEPLDSLTLFLFRGNETLHYETFGKAAPAPQEATATFNS-TADREDGHRJNFSCLAV P L L R G L P E T T D G LDLRPQGLALFSNVSEARSLRTFDLPATIP LDLRPQGLELFENTSAPXQLQTFVLPATPP LDLMSRGGNIFHKHSAPKMLEIYEPVSDSQ LDL G F S L Region 2 196 192 202 b) Regionl mICAM-1 GTG AAC TGT TCT TCC TCA TGC V a l A s n Cys S e r S e r S e r Cys Region 2 mICAM-1 AAT TTC TCA TGC CGC ACA GAA A s n Phe S e r C y s A r g T h r G l u h I C A M - 1 GTG ACA TGC AGC ACC TCC TGT V a l T h r Cys S e r T h r S e r Cys hICAM-1 AAT TTC TCG TGC CGC ACT GAA A s n Phe S e r Cys A r g T h r G l u h I C A M - 2 GTC AAC TGC AGC ACC ACC TGT V a l A s n Cys S e r T h r T h r Cys h I C A M - 2 AAC TTC TCC TGC CTG GCT GTG A s n Phe S e r Cys Leu A l a V a l s e q u e n c e GTC AAA TGC ACC ACC ACA TGC G CC T TGT T T C T s e q u e n c e AAC TTC TCC TGC CGC ACA GAA T G TG G T TG o l i g o GTC AAC TGC AGC ACC ACA TG T T T T C comp. CAC AGC CAG GCA GGA GAA ATT o l i g o T GC G Figure 1 Comparison of human and murine ICAM sequences, a) The amino acid s e q u e n c e of mouse ICAM-1 is al igned with the human ICAM-1 and human ICAM-2 sequences . The fourth line shows the amino acid res idues common to all three sequences . The boxed regions indicate the areas of conserved homology from which the degenerate ol igonucleot ides were constructed, b) S e q u e n c e analys is from these two regions including nucleic acid sequence , a consensus sequence , and the ol igonucleot ide sequence used to P C R amplify a c D N A contained within these two regions. Degeneracy at each position of the ol igonucleotide is indicated by the bases al igning above each other. 68 I C A M . The P C R products obtained were character ized by partial sequenc ing (20 c lones analyzed) and only two were similar to the human ICAM-2 gene. The other 18 either had no long open reading frame or no similarity to any publ ished sequences . Compar i son of amino acid sequence encoded by the cloned P C R fragment with the publ ished human ICAM-2 sequence revealed 6 3 % amino acid identity in a region in domains 1 and 2 spanning 134 residues (Figure 2). The identity w a s evenly distributed throughout the sequence with several stretches of cont inuous identity. B a s e d on this, it w a s conc luded that the P C R fragment was the partial murine equivalent of the human ICAM-2 gene. Interestingly, none of the P C R - d e r i v e d c lones were I C A M - 1 . The 6 3 % amino acid identity between the 2 P C R c lones and the human ICAM-2 gene strongly indicates that the c lones obtained were derived from the murine homologue of I C A M - 2 . O n e of these c lones was used as a probe to sc reen a lung c D N A library because the express ion of ICAM-2 in lung is high (shown in sect ion 2:3.3). 2:3.2 Analysis of mouse ICAM-2 cDNA sequence Of the approximately 2 X 1 0 5 phage from a lung c D N A library that were sc reened with the c loned P C R fragment, eight phage were isolated which hybridized to the probe. T h e s e c lones were converted to insert-containing p B S T plasmid and digested with EcoRI and Xhol which liberated the inserts of three s i zes (1 of ~0.6 kb, 4 of - 0 . 9 kb, 3 of ~1.2 kb). S ince c D N A s of ~ 0.9 kb were found to be incomplete, the 1.2 kb c D N A s were ana lyzed further. These three larger c D N A s gave similar restriction enzyme digest patterns, thus only one (G3-1.1) was used for more extensive 69 P C R c l o n e h I C A M - 2 P D M G G L E T P T N K I M L E E H P Q G K W K Q F L V S N V S K D T V F F C H F T C S G K Q H S E P E V G G L E T S L N K I L L D E Q A Q - - W K H Y L V S N I S H D T V L Q C H F T C S G K Q E S M P G G L E T N K I L E Q WK L V S N S DTV C H F T C S G K Q S P C R c l o n e S L N I R V Y Q P P A Q V T L K L Q P P R V F V G E D F T I E C T V S P V Q P L E R L T L S L L R G h I C A M - 2 N S N V S V Y Q P P R Q V I L T L Q P T L V A V G K S F T I E C R V P T V E P L D S L T L F L F R G N V Y Q P P QV L L Q P V VG F T I E C V V P L L T L L RG P C R c l o n e R E T L K N Q T F G G A E T V P Q E A T A T F N S T A L K K D G L h I C A M - 2 N E T L H Y E T F G K A A P A P Q E A T A T F N S T A D R E D G H E T L T F G A P Q E A T A T F N S T A DG F i g u r e 2 Alignment of protein sequence from the PCR clone with human ICAM-2. E L -4 c D N A w a s used as a template for P C R amplif ication with the degenerate ol igonucleot ides from Reg ion 1 and Reg ion 2. The c loned fragment w a s sequenced and the only open reading frame generated the amino acid sequence shown in the top line. It w a s al igned with human ICAM -2 sequence and residues common to both were shown on the third line. 70 s e q u e n c e analys is . Smal ler fragments of the G3-1.1 insert generated by cutting the insert with Stul, Pstl, Asp700l, Aval, Styl, and Nhel were sequenced on both strands with the T 3 and T 7 primers. The G3-1.1 insert is 1124 nucleot ides (nt) with the largest open reading frame (ORF) beginning at the A T G initiation codon at posit ion 123 and ending with the stop codon T G A at position 954. The 5' untranslated region (5' U T R ) is 122 nt in length (Figure 3). The 3' U T R is 170 nt in length with an 18 nt poly(A) tail 14 nt after the polyadenylat ion signal ( A A T A C A ) . The deduced amino acid sequence of the c D N A - e n c o d e d protein indicates an O R F of 277 residues in length with sequence characterist ic of a t ransmembrane protein. Of the 277 residues, the first 19 residues make up the hydrophobic leader s e q u e n c e which is important in directing the translated protein to the cell sur face. The 203 extracel lular res idues are fol lowed by a hydrophobic 26 amino acid t ransmembrane domain and a hydrophil ic 29 amino acid cytoplasmic domain . The extracel lular region encodes two Ig-like domains containing five potential N-glycosylat ion sites. The first domain has four cysteine res idues which probably form two disulphide bonds to further stabil ize the tertiary structure of the domain . In contrast, the second domain has only two cysteine residues to form one disulphide bond. The sequence identity between murine and human ICAM-2 is 6 0 % at the amino acid level and 7 0 % at the nucleotide level. A s shown in Figure 4, the similarity is spread throughout the coding region with a stretch of near identity in the t ransmembrane and cytoplasmic regions. This indicates that the G3-1.1 c D N A is the murine homolgue of ICAM-2 . 71 CGGGGGAGCGCCAGGCTTCACTCCCCGACCTGTAGCAGACATCTCTC 47 CCTAACCCTCCAGGCAGCCGTCAGCTGTGCCCCTGAAGCCCATAGACTCCACAGACCCCACAGACCCCACCTGAG 122 ATGTCTTCTTTTGCTTGCTGGAGCCTGTCTCTTCTTATCCTGTTCTACAGCCCAGGGTCTGGTGAGAAGGCCTTT 197 M e t S e r S e r P h e A l a C y s T r p S e r L e u S e r L e u L e u I l e L e u P h e T y r S e r P r o G l y S e r G l y G l u L y s A l a P h e 6 - 1 9 +1 GAGGTCTACATATGGTCCGAGAAGCAGATAGTAGAAGCCACAGAGTCTTGGAAAATCAACTGCAGCACCAACTGC 272 G l u V a l T y r l l e T r p S e r G l u L y s G l n l l e V a l G l u A l a T h r G l u S e r T r p L y s I l e A s n C y s S e r T h r A s n C y s 31 - - - C H O - - -GCAGCCCCAGACATGGGCGGCCTGGAGACGCCCACGAATAAAATAATGTTGGAAGAGCATCCTCAAGGGAAGTGG 347 A l a A l a P r o A s p M e t G l y G l y L e u G l u T h r P r o T h r A s n L y s I l e M e t L e u G l u G l u H i s P r o G l n G l y L y s T r p 56 AAACAGTTCTTAGTCTCAAACGTCTCCAAAGACACGGTCTTCTTTTGCCATTTCACGTGTTCGGGAAAGCAGCAC 422 L y s G l n P h e L e u V a l S e r A s n V a l S e r L y s A s p T h r V a l P h e P h e C y s H i s P h e T h r C y s S e r G l y L y s G I n H i s 81 - - - C H O - - -TCGGAGAGTCTCAACATCAGGGTGTACCAGCCTCCAGCTCAAGTCACACTGAAGCTGCAGCCGCCTCGGGTGTTT 497 S e r G l u S e r L e u A s n l l e A r g V a l T y r G l n P r o P r o A l a G l n V a l T h r L e u L y s L e u G l n P r o P r o A r g V a l P h e 106 GTGGGTGAAGACTTCACCATTGAGTGCACGGTGTCCCCTGTGCAGCCCCTTGAGAGGCTCACCCTCTCTCTGCTC 572 V a l G l y G l u A s p P h e T h r l l e G l u C y s T h r V a l S e r P r o V a l G l n P r o L e u G l u A r g L e u T h r L e u S e r L e u L e u 131 CGTGGCAGAGAGACCCTGAAGAATCAGACCTTTGGGGGAGCAGAAACTGTCCCCCAAGAGGCCACAGCCACGTTC 647 A r g G l y A r g G l u T h r L e u L y s A s n G l n T h r P h e G l y G l y A l a G l u T h r V a l P r o G l n G l u A l a T h r A l a T h r P h e 156 - - - C H O - - -AACAGCACAGCTCTGAAAAAGGACGGTCTCAACTTTTCCTGCCAGGCTGAGCTGGATCTACGGCCCCATGGTGGG 722 A s n S e r T h r A l a L e u L y s L y s A s p G l y L e u A s n P h e S e r C y s G l n A l a G l u L e u A s p L e u A r g P r o H i s G l y G l y 181 - - - C H O - - - - - - C H 0 - - -TATATCATCCGCAGCATCTCGGAGTACCAGATCCTTGAAGTCTATGAGCCGATGCAGGACAACCAAATGGTCATC 797 T y r l l e l l e A r g S e r l l e S e r G I u T y r G l n i l e L e u G l u V a l T y r G l u P r o M e t G l n A s p A s n G l n M e t V a l H e 206 ATCATCGTGGTGGTGTCAATACTGCTGTTCTTATTTGTGACATCTGTCCTGCTATGCTTTATCTTTGGCCAGCAC 872 II e l l e V a l V a l V a l S e r l l e L e u L e u P h e L e u P h e V a l T h r S e r V a l L e u L e u C y s P h e l l e P h e G l y G l n H i s 231 TGGCACAGAAGACGGACAGGCACCTACGGGGTGCTAGCTGCCTGGAGGAGGCTGCCCCGAGCCTTTCGGGCACGT 947 T r p H i s A r g A r g A r g T h r G l y T h r T y r G l y V a l L e u A l a A l a T r p A r g A r g L e u P r o A r g A l a P h e A r g A l a A r g 256 CCCGTGTGAGCCCACGTTGCCAGGCCCCTGGTGGTTACCAGAACTCAACATGGCACCTTCAAGGTGTGGTTCGGC 1022 P r o V a l * * * 258 ACTGGCTGAAGGACTGTGGCGGCAGCAGCAGATGCGGGGGACATTTCCTCTCCTTTTTAGCCTCAATACAAATAT 1096 CTGGATTTCAAAAAAAAAAAAAAAAAA 1124 Figure 3 Nucleic acid and protein sequence of the murine ICAM-2 cDNA (G3-1.1). The complete nucleotide sequence of the G3-1.1 c lone is shown with the predicted amino acid residue beneath its corresponding codon. The predicted N-terminal s ignal pept ide is underl ined ( ), potential glycosylat ion sites are marked by — C H O — , cyste ine res idues are in bold, the t ransmembrane region is underl ined with a thick line {^^^^m), the translation termination codon is marked with *** beneath it, and the polyadenylat ion signal sequence A A T A C A is boxed. The amino acid sequence is numbered from the predicted c leavage site of the signal peptide. 72 G 3 - 1 . 1 SGEKAFEVYIWSEKQIVEATESWKINCSTNCAAPDMGGLE 40 hICAM-2 SDEKVFEVHVRPKKLAVEPKGSLEVNCSTTCNQPEVGGLE 40 S EK FEV K VE S NCST C P GGLE G 3 - 1 . 1 TPTNKIMLEEHPQGKWKQFLVSNVSKDTVFFCHFTCSGKQ 80 hICAM-2 TSLNKILLDEQAQ--WKHYLVSNISHDTVLQCHFTCSGKQ 78 T NKI L E Q WK LVSN S DTV CHFTCSGKQ G 3 - 1 . 1 HSESLNIRVYQPPAQVTLKLQPPRVFVGEDFTIECTVSPV 120 hICAM-2 ESMNSNVSVYQPPRQVILTLQPTLVAVGKS-FTI ECRVPTV 118 S N VYQPP QV L LQP V VG FTIEC V V G 3 - 1 . 1 QPLERLTLSLLRGRETLKNQTFGGAETVPQEATATFNSTA 160 hICAM-2 EPLDSLTLFLFRGNETLHYETFGKAAPAPQEATATFNSTA 158 PL LTL L RG ETL TFG A PQEATATFNSTA G 3 - 1 . 1 LKKDGL-NFSCQAELDLRPHGGYIIRSISEYQILEVYEPM 199 hICAM-2 DREDGHRNFSCLAVLDLMSRGGNIFHKHSAPKMLEIYEPV 198 DG NFSC A LDL GG I S LE YEP G 3 - 1 . 1 QDNQMVIIIVVVSILLFLFVTSVLLCFIFGQHWHRRRTGT 239 hICAM-2 SDSQMVIIVTVVSVLLSLFVTSVLLCFIFGQHLRQQRMGT 238 D QMVII VVS LL LFVTSVLLCFIFGQH R GT G 3 - 1 . 1 YGV LAAWRRLPRAFRARPV hICAM-2 YGVRAAWRRLPQAFRP YGV AAWRRLP AFR 258 254 F i g u r e 4 Comparison of amino acid sequence between cDNA-encoded (G3-1.1) protein and human ICAM-2. The amino acid sequence derived from the c loned G3-1.1 c D N A is al igned with the human ICAM-2 sequence . The third line shows the residues shared by the two sequences . 73 2:3.3 Northern blot analysis Express ion of murine ICAM-2 m R N A in lymphoid and non-lymphoid t issues was examined by Northern blot analys is (Figure 5). A single spec ies of ~ 1.2 kb was detected in the var ious t issues which expressed ICAM-2 . Low levels were detected in bone marrow cel ls, while sp leen and thymus displayed moderate level of ICAM-2 express ion . ICAM-2 was not detected in liver t issue. However, low levels of ICAM-2 were s e e n in heart and kidney t issues, while very high levels were detected in lung t issue. This express ion pattern of murine ICAM-2 is similar to that of human ICAM-2 . The high level of express ion in lung is expected because lung t issue is abundant in endothel ium, which constitutively exp resses high levels of ICAM-2 . 2:3.4 Genomic cloning The mouse ICAM-2 c D N A was then used as a probe in genomic Southern blot analys is of B A L B / c sp leen D N A digested with var ious restriction e n z y m e s (which do not cut the G3-1.1 c D N A ) alone and in combinat ion (Figure 6). In four c a s e s , a single band w a s detected (BamHI, EcoRI, Hindlll, Xhol). In the c a s e of Dral, the blot d isp layed two bands (possibly due to the presence of a Dral site within an intron). Doub le digests a lso revealed a single banding pattern. It appears that murine ICAM-2 , like its human homologue, is present as a single copy gene in the mouse genome. The genomic organizat ion of ICAM-2 could provide information about its evolution as well as its regulation in certain t issues. In order to isolate a mouse I C A M -2 genomic c lone, B A L B / c sp leen D N A was digested with EcoRI and s i ze se lected in 74 1 2 3 4 5 6 7 A c t i n F i g u r e 5 Expression of murine ICAM-2 by Northern blot analysis. Total R N A (10 p,g) isolated from var ious murine t issues (BALB/c ) was run on a formaldehyde gel and blotted onto a Zeta-probe filter. Source of R N A in each lane are as follows: 1) heart, 2) lung, 3) liver, 4) kidney, 5) bone marrow, 6) sp leen, 7) thymus. The filter was probed with the G3-1.1 c D N A as well as an actin probe. 75 1 2 3 4 5 6 7 8 9 23.1 F i g u r e 6 Genomic Southern blot analysis of murine ICAM-2. Sp leen D N A (10 ug) from B A L B / c mice was digested with the indicated restriction enzymes , e lect rophoresed on a 0 .8% agarose gel, and blotted. E n z y m e s in each lane are as fol lows: 1) BamHI, 2) Dral, 3) EcoRI, 4) Hindlll, 5) Xhol, 6) BamHI/EcoRI, 7) EcoRI/Hindlll, 8) EcoRI/Xhol, 9) Hindlll/BamHI. The filters were then probed with the G3-1.1 c D N A . 76 the range of 5.5-7.0 kb. A Xgt10 library was constructed with the D N A and sc reened with the G3-1.1 c D N A . O n e clone isolated, BM1-1 .1 , was partially sequenced (Figure 7) and mapped by restriction enzyme analys is in conjunction with Southern blotting (Figure 8). It was found to be 6.5 kb in length. The coding sequence res ides in a 5 kb region and contains an exon coding for the 5' U T R and signal peptide, two exons encod ing the two extracellular Ig-like domains, and an exon encoding the t ransmembrane and cytoplasmic domains, and the 3' U T R . The exons are separated by phase I introns (split exons after the first nucleotide of a codon) and the map of the genomic c lone is shown in Figure 9. Sequenc ing a lso generated 233 bp of sequence 5' to the A T G translation initiation start site. Al though this is a rather short stretch of sequence , there is no apparent T A T A box or CAAT- l i ke sequences . There is however a transcription initiation consensus sequence , A T T C T T , at nucleot ide at -229 (with respect to the translation start codon) suggest ing that there are promoter-l ike s e q u e n c e s farther upstream. 2:3.5 Purification of ICAM-2 Prel iminary binding studies of PMA-act iva ted splenic T cel ls to purified ICAM-2 could be used for examining parameters of the LFA-1 : ICAM-2 adhes ion . The murine ICAM-2 w a s isolated from B W 5 1 4 7 cel ls, an A K R thymic leukemia cell line, because of high express ion . The protein purified by mAb affinity chromatography w a s ana lyzed by S D S - P A G E . The si lver stained gel (Figure 10) revealed that the isolated protein was a single band with a molecular mass of 50-55 kD. The protein was homogenous s ince 77 a t t c t t ^ a g g c c c t a a a g g c t t g g g a g c t g g t c t g t g c a t a t t g t t t c c t g a t c t c a g a t a a t t a g a g g a a a t g a g c t c a c t g g c a c a g a g g a g a t t g t g g a t t t c a g t t g g g a g c g c c a g g c t t c a c t c c c c g a c c t g t a g c a g a c a t c t c t c c c t a a c c c t c c a g g c a g c c g t c a g c t g t g c c c c t g a a g c c c a t a g a c t c c a c a g a c c c c a c a g a c c c c a c c t g a g A T G T C T T C T T T T . . . . T T C T A M e t S e r S e r P h e . . . . P h e T y CAGCCCAGgtaagccagctcccaggggtttcag.. intronl..cagtggttgattttccagGGJCJGGJGAGAAG. . . . A T C A rSerProG l y S e r G l y G l u L y s I l e A GGGJGTACCgtgagtggctctgctgccgt.. intron2..cctccttaactccgctgcagAGCCTCCAGCJCAA CTTGAA r g V a l T y r G 1nProProAlaGln LeuGlu GJCJAJGgtgaggggaggatccgtaga.. intron.3..ccacgtcctttgcctcccagAGGCGAJGCAGGAC. . . .GCACGTCC ValTyrG 1uProMetGlnAsp A l a A r g P r C G T G T G A g c c c a c g t t g c c a g g c c c c t g g t g g t t a c c a g a a c t c a a c a t g g c a c c t t c a a g g t g t g g t t c g g c a c t g g c t g a a g o V a l * * * g a c t g t g g c g g c a g c a g c a g a t g c g g g g g a c a t t t c c t c t c c t t t t t a g c c t c a a t a c a a a t a t c t g g a t t t c . . . . Figure 7 Partial nucleotide sequence of the murine ICAM-2 genomic clone. The partial nucleotide sequence of the ICAM-2 gene determined from the 6.6 kb EcoRI fragment (BM1-1.1) is listed above. Untranslated sequences are in lower case letters, protein coding sequences are in upper case, and intron sequences in bold italics (lower case). Amino acid sequence from the coding region is shown below the corresponding codons of the exons. The consensus transcription start site (ATTCTT) is boxed. Figure 8 Southern blot analysis of murine ICAM-2 genomic clone (BM1-1.1). The 6.6 kb EcoRI insert (750 ng) containing the murine ICAM-2 genomic c lone was digested with var ious restriction enzymes (1- Xmnl, 2- Aval, 3- BamHI, 4- Kpnl, 5- Ncol, 6-Smal), e lect rophoresed, and blotted onto a filter. The filter was then probed with the 192 bp fragment generated by Stul digestion of the G3-1.1 c D N A which represented the 5' U T R and the portion encoding the signal peptide. The filter was a lso probed with the 411 bp fragment generated by Styl digestion of the G3-1.1 c D N A which represents the t ransmembrane and cytoplasmic portions as well as the 3' U T R . 79 Restriction enzymes Aval A BamHI B Kpnl K Ncol N Smal S XmnI X B K r—W-f A N lkb F i g u r e 9 Restriction map of the murine ICAM-2 genomic clone. By combining data from the restriction digestion and Southern blots of Figure 8 with sequence information from the partial sequencing of the genomic c lone as well as from the c D N A clone, a map of the restriction e n z y m e sites in the BM1-1.1 c lone is shown above. The locations of the four exons are indicated by filled boxes and the relative posit ions of the restriction enzyme sites are also shown. Mr X 10 F i g u r e 10 Immunopurification of ICAM-2. Murine ICAM-2 immune-affinity purified from B W 5 1 4 7 p lasma membranes was subjected to nonreducing S D S - 1 0 % P A G E and silver staining. 81 no other bands were detected in the silver stained gel . Approximately 5 ug of murine I C A M - 2 w a s obtained per litre of cultured B W 5 1 4 7 cel ls. 2:3.6 Binding to ICAM-2 protein The purified murine ICAM-2 was used to examine the binding capabil i ty of I C A M - 2 . PMA-ac t i va ted splenic T cel ls adhered to purified ICAM-1 and ICAM-2 coup led to plastic (Figure 11). This was LFA-1-spec i f i c as an t i -CD11a antibody inhibited the binding to background levels. It was also dependent on divalent cat ions such as magnes ium, s ince E D T A also inhibited the binding. Approximately 5 4 % and 3 6 % of T cel ls bound ICAM-1 and ICAM-2 , respectively. Ce l l adhes ion to ICAM-2 was inhibited by the ICAM-2 antibody (3C4) (11% of cel ls bound) but not by the ICAM-1 antibody. Ce l l adhes ion to ICAM-1 was inhibited by the ICAM-1 antibody, but was unaffected by the ICAM-2 antibody. The binding to ICAM-2 w a s further dec reased to 7% when the ICAM-2 antibody concentrat ion was increased from 4 u.g/ml to 15 u.g/ml. A s observed previously, resting T cel ls did not bind to the purified proteins. 2:3.7 Binding to cell surface ICAM-2 In order to examine whether the G3-1.1 c D N A encodes a protein capab le of binding L F A - 1 , L cel ls were transfected with the c D N A . F low cytometry revealed that I C A M - 2 w a s detected on the surface of the transfected cel ls (Figure 12a). P M A -activated sp len ic T cel ls readily adhered to monolayers of the transfected L cel ls express ing ICAM-1 or ICAM-2 (Figure 12b). Al though the level of ICAM-2 on the 82 L 0 ICAM-2 + 5 mM EDTAl B ICAM-1 .'• • OVA + CD11a Ab binding 0 10 20 30 40 50 60 cells bound (%) Figure 11 Adhesion of murine splenic T cells to purified ICAM-1 and ICAM-2. I C A M - 2 , I C A M - 1 , and ovalbumin were immobil ized on microculture wel ls. To each wel l , 1 0 5 ca lce in A M labeled splenic T cel ls (PMA-act ivated or resting) were added and al lowed to proceed for 8 min at 37°C. Blocking ant ibodies were added to P M A -act ivated T cel ls 15 min prior to the addition of cel ls to the wel ls. The ant ibodies were present at a concentrat ion of 2 u.g/ml, except for ICAM-2 antibody which w a s present either at 4 u.g/ml or 15 u.g/ml (indicated by *). Unbound T cel ls were removed by wash ing five t imes with H B S S containing 5 % F C S . Resul ts are expressed as a mean of triplicate wel ls ± S E M . 83 A) ICAM-1 Ab ICAM-2 Ab L cells L:ICAM-1 L:ICAM-2 Figure 12 Adhesion of splenic T cells to L cells transfected with ICAM-2. A) L cel ls, untransfected or transfected with the ICAM-1 (L: ICAM-1) or ICAM-2 (L : ICAM-2) c D N A were subjected to flow cytometry using the appropriate antibody and F(ab ' ) 2 goat anti-rat IgG conjugated to F ITC. ICAM-1 and ICAM-2 express ion are shown by the sol id histogram and secondary antibody alone is shown by the outl ined histogram. B) (next page) L cel ls express ing ICAM-1 or ICAM-2 c D N A were grown in 96 well plates. PMA-ac t i va ted splenic T cel ls labeled with calcein A M were added to e a c h well containing a subconf luent monolayer of L cel ls, and binding was measured as descr ibed in Figure 11. 84 B) 0 25 50 75 100 cells bound (%) (figure legend on previous page) 85 t ransfected L cel ls was higher than that of ICAM-1 , T cel ls adhered to ICAM-1 transfectants (86% of input cells) more efficiently than to ICAM-2 transfectants (62%). T h e adhes ion of T cel ls to transfected L cel ls was effectively inhibited by the anti-C D 1 1 a m A b and the control antibody against C D 4 9 d had no effect on adhes ion . The ant i - ICAM-2 mAb also inhibited the T cell adhes ion to ICAM-2- t ransfected L cel ls, whereas the T cell adhes ion to ICAM-1-transfected L cel ls was effectively inhibited by the m A b to I C A M - 1 . 2:4 D i s c u s s i o n In the human sys tem, three counter-receptors for LFA-1 ( ICAM-1 , ICAM-2 , and ICAM-3) have been identified and character ized (S immons et al., 1988; Staunton et al., 1988; Staunton et al., 1989; Fawcett et al., 1992; V a z e u x et al., 1992; de Fougero l les et al., 1993). W h e n this project began, only ICAM-1 had been identified in the murine sys tem (Horley et al., 1989; S iu et al., 1989). In an attempt to isolate addit ional I C A M -like c D N A s in the murine sys tem, a P C R - b a s e d strategy w a s explored using degenerate ol igonucleot ides based on avai lable sequence information from the known I C A M s . The primer sequences were determined on the basis of conserved cysteine res idues and adjacent amino acid sequences from the first two Ig-like domains in the human I C A M - 1 , ICAM-2 , and the murine ICAM-1 (the only known I C A M s e q u e n c e s at the time). T h e s e ol igoprimers were used to P C R amplify murine ICAM- l ike D N A fragments which were then used as a probe to screen for complete c D N A s . With this strategy, full length c D N A c lones encoding var ious molecules can be isolated. O n e 86 group had previously c loned the murine ICAM-1 using the human ICAM-1 c D N A as a probe (Siu et al., 1989). S ince the nucleotide homology between the two D N A fragments w a s relatively low (50%), the str ingency of the screen was a lso low making the screen ing process more difficult. EL -4 cel ls were chosen as a source for P C R amplif ication because they exhibit an LFA-1-dependent adhes ion pathway and ICAM-1 express ion is low (Wuthridge, 1992). This favoured the amplif ication of murine I C A M -like molecu les not yet identified. O n e of the P C R c lones generated by the above approach was found to partially encode a protein with significant similarity to the human ICAM-2 . This sugges ted that the P C R c lone w a s derived from the murine equivalent of the human ICAM-2 . It a lso demonstrated that this strategy has the potential to isolate novel c D N A c lones encod ing related molecules. In the cloning of the human I C A M - R ( ICAM-3) c D N A , a similar strategy was used (Vazeux et al., 1992). The ol igonucleot ides used as P C R primers were derived from sequences of the second domain of I C A M - 1 , ICAM-2 , N C A M , M A G , P E C A M - 1 , and V C A M - 1 , whereas the primers used in this study span domains one and two of ICAM-1 and ICAM-2 . A s well , the template in the I C A M - R P C R cloning w a s genomic D N A , whereas in this study it was s ingle-stranded c D N A der ived from m R N A of E L - 4 cel ls. A limitation of using c D N A as a template for P C R amplif ication is that the overall ef fect iveness is determined by express ion of the m R N A as wel l as relative levels of other similar m R N A s . In the case of the I C A M - R cloning, these limitations were circumvented by using genomic D N A as a template. A drawback in using genomic D N A is the presence of introns which may be lengthy and 87 thus interfere with efficient synthesis of the P C R fragment. Wh ichever template is used , the degenerate ol igonucleot ides make it possib le to c lone novel ICAM-l ike s e q u e n c e s without ant ibodies or specif ic protein sequence information. The murine ICAM-2 c D N A encodes a type I t ransmembrane protein with two extracel lular Ig-like domains extending from the cell surface fol lowed by a hydrophobic t ransmembrane domain and a hydrophil ic cytoplasmic tail. The posit ions of the cyste ine res idues are conserved. The first domain contains four cyste ine residues forming two disulphide l inkages, while the second domain contains only one disulphide bond formed by the two cysteine residues. The presence of four cyste ines in the first domain of murine ICAM-2 is also seen in human (S immons et al., 1988; Staunton et al., 1988) and murine ICAM-1 (Horley et al., 1989; S iu et al., 1989), human ICAM-2 (Staunton et al., 1989), human ICAM-3 (Fawcett et al., 1992; V a z e u x et al., 1992; de Fougero l les et al., 1993), and human (Osborn et al., 1989) and murine V C A M - 1 (Araki et al., 1993). Interestingly, these are all members of the Ig superfamily which adhere to members of the integrin superfamily, LFA-1 and V L A - 4 (Marlin and Spr inger, 1987; E l i ces et al., 1990). Within the ICAM-2 extracellular region are five potential N-glycosylat ion sites (Asn-X-Ser /Thr) . The molecular mass of the ICAM-2 apoprotein deduced from the c D N A sequence is 28 kD but the mature protein is 55 kD. This relatively high increase in molecular mass is probably due to extensive glycosylat ion. ICAM-1 has a 55 kD polypeptide backbone with the mature protein being ~ 90-100 kD and the I C A M - 3 polypeptide backbone has a molecular mass of 57 kD with the mature protein migrating as a band of 124 kD in S D S - P A G E . ICAM-1 and I C A M - 3 have eight 88 and 15 potential N-linked glycosylat ion sites, respectively (S immons et al., 1988; Staunton et al., 1988; Fawcett et al., 1992; V a z e u x et al., 1992; de Fougero l les et al., 1993). Th is extensive glycosylat ion in the I C A M s yields an average increase in mass of ~ 4.5 kD per glycosylat ion site. C o m m o n va lues of N-glycosidic o l igosacchar ides are ~ 2-2.5 kD per site (Asada et al., 1991; Gahmberg et al., 1991; Nor tamo et al., 1991). Th is higher than average ol igosacchar ide content in I C A M s is conserved not only among the members of the subfamily but a lso between the murine and human spec ies and may play a role fine tuning the adhes ive property of the L F A - 1 : I C A M interactions. W h e n the human ICAM-1 sequence is altered such that a glycosylat ion in the third Ig-like domain is destroyed, it's ability to bind Mac-1 is enhanced (Diamond et al., 1991). However, it has a lso been demonstrated that unglycosylated ICAM-2 purified from a bacterial express ion system is able to bind LFA-1 (Gahmberg et al., 1991). The role of the carbohydrate moiety in ICAM-2 remains unclear. Ana lys i s of the c D N A sequence demonstrates the overall structural identity with the human ICAM-2 . The murine ICAM-2 shares sequence similarity with the human ICAM-2 (60% at the amino acid level and 7 0 % at the nucleic acid level) with a stretch of near identity spanning the t ransmembrane and cytoplasmic regions (TM and Cyto). In the c a s e of ICAM-2 , there is an 8 5 % and a 7 0 % amino acid identity between the murine and human homologue over the T M and Cyto regions, respectively. Interestingly, the sequence identity between I C A M s of the s a m e spec ies is higher in the LFA-1-b ind ing extracellular domains than in the T M and Cyto regions. Nonethe less , the conservat ion in the T M and Cyto regions is high between human and 89 their murine equivalents. These regions may be important in the function of their speci f ic I C A M s and thus are maintained by sequence conservat ion ac ross spec ies . T h e s e regions may function in localization on the cell sur face, interaction with the cytoskeleton, and signal l ing. One study has indicated that both human ICAM-1 and I C A M - 2 cytoplasmic tails are able to interact with the a-act inin, a cytoskeletal protein which can anchor actin f i laments to the cell membrane (Carpen et al., 1992; He iska et al., 1996). A l so , C O S cel ls transfected with ICAM-1 exhibited punctated ICAM-1 staining on the cell surface (Kishimoto et al., 1990; Ca rpen et al., 1992). However , when the c D N A is altered such as to generate a GPI- l inked version of I C A M - 1 , the staining became diffuse. In addit ion, ICAM-2 is able to interact with ezr in (Helander et al., 1996), a membrane-organiz ing protein which is thought to act as a cytoskeletal linker for membrane-bound proteins. NK-resistant murine target cel ls express ICAM-2 in a diffuse pattern. However, transfection of ezrin into these cel ls c a u s e s the murine I C A M - 2 to redistribute to uropods and the target cel ls become sensi t ized to NK- lys is . The express ion of murine ICAM-2 , as determined by Northern blot analys is , is similar to the express ion pattern of human ICAM-2 . ICAM-2 is constitutively exp ressed on lymphocytes, granulocytes, platelets, and endothel ium (de Fougero l les et al., 1991; D iacovo et al., 1994). Detection of ICAM-2 in lymphoid organs such as sp leen , bone marrow, and thymus may be due to the leukocytes present (Xu et al., 1992). ICAM-2 is a lso detected in heart, lung, and kidney, possibly due to leukocytes and endothel ial cel ls present in these organs. Lung t issue is particularly rich in endothel ium and this is reflected in the high level of express ion of ICAM-2 . T h e s e express ion patterns have 90 been conf i rmed by another group looking at murine ICAM-2 (Xu et al., 1992) and dupl icates the human pattern. In addition, ICAM-2 has been shown to be non-inducible (de Fougero l les et al., 1991; Nortamo et al., 1991; Ohh et al., 1994). The only time ICAM-2 has been found on cel ls other than the ones ment ioned above or at higher levels than normally found in vivo is on neop lasms (Roos , 1991; El l is et al., 1992; Renkonen etal., 1992). The isolation of a murine genomic ICAM-2 D N A fragment al lowed the examinat ion of its structure. This is important in understanding its regulation as well as the evolut ionary origin of members of the I C A M subfamily. The genomic organizat ion of the murine ICAM-2 gene confirms its ass ignment as a member of the Ig superfamily (Wil l iams and Barclay, 1988; Hunkapi l ler and Hood, 1989). The exon/intron boundar ies of the gene are reflected in the structural domains of the protein. The s ignal peptide is encoded by the first exon. The next two Ig-like domains are encoded by separate exons and finally the fourth exon codes for the t ransmembrane and cytop lasmic domains. T h e s e exons are separated by phase I introns in which the intron appears after the first nucleotide of a codon. P h a s e I introns have been found between numerous Ig domain-l ike exons of other members of the Ig superfamily (Hunkapi l ler and Hood , 1989). By maintaining uniform intron phase in these molecu les , it al lows the possibil ity of alternate exon usage without altering the reading f rame. Uniform intron phase is a lso important for Ig superfamily evolut ion, because it ensures that molecu les with multiple Ig-like domains can be constructed by exon dupl icat ion and shuffling meanwhi le maintaining the correct reading frame. B a s e d on 91 s e q u e n c e homology and intron/exon boundar ies, gene duplication and subsequent d ivergence from a primordial domain are the accepted origin of Ig-like molecu les (Owens et al., 1987; Wi l l iams and Barclay, 1988; Hunkapi l ler and Hood , 1989). The formation of a distinct subfamily is somewhat less clear. Members of the C D 1 1 family have been found to map to a region of one chromosome (Corbi et al., 1988). The evolut ion of this family can be explained by unequal cross ing over between ch romosomes and local ized to one region. However, ICAM-1 and ICAM-2 in both the human and murine sys tems are found to be on separate ch romosomes (Katz et al., 1985; Bal lantyne et al., 1991; Hogg et al., 1991; Kuramoto et al., 1994). Thus , unequal cross ing over cannot explain the generat ion of the I C A M subfamily in the human and murine sys tems. It may be that the ancestral ICAM- l ike gene dupl icated and inserted in var ious locations in the genome. Subsequen t d ivergence may have occurred at each of the insertion sites. Ana lys i s of the region 5' of the coding region failed to reveal any obvious regulatory sequences . The genomic c lone isolated did however p o s s e s s a consensus transcription initiation sequence . A larger portion of the 5' region of the murine ICAM-2 gene has been c loned by another group (Xu et al., 1992). They found that further upstream of the transcription initiation site is a TATA- l i ke sequence and an inverted C A A T box. No known transcription factor binding sites were identified in the promoter region. In contrast, the 5' upstream region of ICAM-1 has NF-KB and A P - 1 binding sites (Voraberger et al., 1991). ICAM-2 express ion is not inducible by cytokines and thus it is not surprising that ICAM-2 does not p o s s e s s NF-KB and A P - 1 binding sites 92 which have been shown to be important in transcription of inducible genes such as ICAM-1 (Angel etal., 1987; Edbrooke etal., 1989). S o m e resting leukocytes express ICAM-2 in addition to LFA-1 (de Fougero l les et al., 1991; Nortamo et al., 1991). However, they do not spontaneous ly aggregate b e c a u s e the avidity of LFA-1 does not appear to be high enough (Rothlein and Spr inger, 1986; Dustin and Springer, 1989; van Kooyk et al., 1989). In our exper iments, sp len ic T cel ls did not adhere to purified murine ICAM-1 or ICAM-2 unless previously activated with P M A . The adhes ion was speci f ic as determined by inhibition with C D 1 1 a , ICAM-1 , and ICAM-2 mAbs . PMA-ac t iva ted sp len ic T cel ls a lso demonstrated the s a m e specificity towards cell surface ICAM-1 and ICAM-2 . T cel ls did not adhere to L cel ls unless they were transfected with ICAM-1 or ICAM-2 c D N A . Binding of T cel ls to ICAM-2-transfected L cel ls was consistently lower than to I C A M - 1 -transfected L cel ls, despite the higher level of ICAM-2 express ion . This suggests that cell adhes ion mediated by LFA-1 : ICAM-2 is not as efficient as that mediated by L F A -1: ICAM-1. Simi lar results have been obtained by other groups. W h e n LFA-1 express ion is limiting, ICAM-1 binding is preferred over ICAM-2 binding (Dustin et al., 1989; Kishimoto etal., 1990). A s well , ICAM-1 adhes ion is a lso stronger than ICAM-2 . The compar ison between ICAM-2 and ICAM-3 binding to LFA-1 has not been examined . Never the less, leukocytes expressing activated LFA-1 are able to adhere to I C A M - 2 + cel ls in the absence of ICAM-1 express ion. Certain cel ls , including unstimulated endothel ial cel ls, express relatively high levels of ICAM-2 but only low levels of ICAM-1 (Dustin and Springer, 1988). Leukocytes express ing activated LFA-1 93 may be able to adhere to these cells by binding to ICAM-2 and thus facilitate subsequent functions of the immune response. The binding sites for LFA-1 adhes ion in ICAM-1 has been mapped by site directed mutagenes is (Staunton et al., 1990). The first domain of ICAM-1 is critical for binding to L F A - 1 . B a s e d on the sequence conservat ion of the I C A M s and their functional similarity in binding L F A - 1 , it may be speculated that the l igand recognit ion si tes lie in homologous posit ions and contain key conserved res idues. Two key res idues involved in LFA-1 binding are glutamic acid at posit ion 34 and glutamine at posit ion 73. T h e s e are conserved in all five known I C A M sequences . Interestingly, a peptide spanning residues 21-42 of the human ICAM-2 is able to block binding to L F A -1 indicating the importance of these residues (Li et al., 1993b). The binding of ICAM-2 to LFA-1 has been well estab l ished in vitro (de Fougero l les et al., 1991; X u et al., 1992; X u et al., 1996), however the functional s igni f icance of this cell adhes ion pathway in vivo remains unknown. It is of interest that inf lammatory responses of ICAM-1-def ic ient mice are only partially defect ive (Sligh et al., 1993; X u et al., 1994). Whether ICAM-2 is compensat ing for the ICAM-1 def ic iency remains to be seen . It also remains to be determined whether the high level of ICAM-2 express ion in lung contributes to neutrophil migration in pneumonia (Burns et al., 1994). B a s e d on the limited express ion pattern of ICAM-2 , lymphocytes and endothel ium, it may be possib le that ICAM-2 functions in T cell activation and leukocyte transendothel ial migration. With the c D N A encoding murine ICAM-2 and the 94 3 C 4 antibody, the potential functional roles of ICAM-2 in an al logeneic T cell response and in leukocyte transendothel ial migration are examined further in this thesis. 95 2:5 R e f e r e n c e s Alzar i P M , L a s c o m b e M, and Poljak R J (1988) Three-d imensional structure of ant ibodies. Annu. Rev. Immunol. 6:555. Amit A G , Mar iuzz i R A , Phil l ips S E V , and Poljak R J (1986) Three-d imens iona l structure of an antigen-antibody complex at 2.8 A resolution. Science 233:747. A n g e l P, Imagawa M, Ch iu R, Stein B, Imbra R J , Rahmsdor f H J , Jonat C , Herrl ich P, and Karin M (1987) Phorbol ester- inducible genes contain a common c/s element recognized by a TPA-modu la ted frans-acting factor. Cell 49:729. Arak i M, Arak i K, and Vassa l l i P (1993) Cloning and sequenc ing of mouse V C A M - 1 c D N A . Gene 126:261. A s a d a M, Furukawa K, Kantor C , Gahmberg C G , and Kobata A (1991) Structural study of the sugar cha ins of human leukocyte cell adhes ion molecu les C D 1 1 / C D 1 8 . Biochem. 30 :1561. Bal lantyne C M , Kozak C A , O'Br ien W E , and Beaudet A L (1991) Ass ignment of the gene for intercellular adhes ion molecule-1 (/cam-1) to proximal mouse ch romosome 9. Genomics 9:547. C a r p e n O, Pal la i P, Staunton D E , and Spr inger T A (1992) Assoc ia t ion of intercellular adhes ion molecule-1 ( ICAM-1) with actin-containing cytoskeleton anda-act in in. J. Cell Biol. 118:1223. Carpeni to C , Ohh M, and Take i F (1995) Cloning and character izat ion of murine intercellular adhes ion molecule-2 ( ICAM-2): a functional and molecular analys is . Allerg. Immunol. (Life Sci. Adv.) 14:43. C h o m c z y n s k i P and Sacch i N (1987) Single-step method of R N A isolation by acid guanidin ium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162:156. Corb i A L , Larson R S , Kishimoto TK , Spr inger TA , and Morton C C (1988) Ch romosoma l location of the genes encoding the leukocyte adhes ion receptors L F A - 1 , M a c - 1 , and p150/95. Identification of a gene cluster involved in cell adhes ion . J. Exp. Med. 167:1597. Damle NK , K lussman K, and Aruffo A (1992) Intercellular adhes ion molecule-2, a second counter-receptor for C D 1 1 a / C D 1 8 (leukocyte funct ion-associated antigen-1), 96 provides a costimulatory signal for T-cel l receptor-initiated activation of human T cel ls. J. Immunol. 148:665. Dav ignon D, Martz E, Reyno lds T, Kurzinger K, and Spr inger T A (1981) Monoc lona l ant ibody to a novel lymphocyte funct ion-associated antigen (LFA-1) : mechan ism of blocking of T lymphocyte-mediated killing and effects on other T and B lymphocyte functions. J. Immunol. 127:590. Dav is L G , Dibner M D , and Battey J F (1986) Bas ic methods in molecular biology. E lsev ier S c i e n c e Publ ishing C o . , Inc., New York. de Fougero l les A R , Kl ickstein LB , and Spr inger T A (1993) Clon ing and express ion of intercellular adhes ion molecule 3 reveals strong homology to other immunoglobul in family counter-receptors for lymphocyte funct ion-associated antigen 1. J. Exp. Med. 177:1187. de Fougero l les A R , Qin X , and Spr inger T A (1994) Character izat ion of the function of intercellular adhes ion molecule ( ICAM)-3 and compar ison with ICAM-1 and ICAM-2 in immune responses . J. Exp. Med. 179:619. de Fougero l les A R and Spr inger T A (1992) Intercellular adhes ion molecule 3, third adhes ion counter-receptor for lymphocyte funct ion-associated molecule 1 on resting lymphocytes. J. Exp. Med. 175:185. de Fougero l les A R , Stacker S A , Schwart ing R, and Spr inger T A (1991) Character izat ion of ICAM-2 and ev idence for a third counter-receptor for L F A - 1 . J. Exp. Med. 174:253. D iacovo T G , de Fougero l les A R , Bainton D F , and Spr inger T A (1994) A functional integrin l igand on the surface of platelets: intercellular adhes ion molecule-2. J. Clin. Invest. 94 :1243. Dia lynas D, Loken M, Sarmiento M, and Fitch F W (1982) Identification of lysis-relevant molecu les on the surface of C T L : primary screening of monoclonal ant ibodies for the capaci ty to block cytolysis by cloned T cell l ines. Adv. Exp. Med. Biol. 146:547. D iamond M S , Staunton D E , Marlin S M , and Spr inger T A (1991) Binding of the integrin Mac-1 (CD11b /CD18 ) to the third immunoglobulin-l ike domain of ICAM-1 (CD54) and its regulation by glycosylat ion. Cell 65 :961. Dong Y , Skoul tchi A l , and Pol lard J W (1993) Efficient D N A transfection of quiescent mammal ian cel ls using poly-L-ornithine. Nucleic Acids Res. 21 :771. Dustin M L , Garc ia-Agui la r J , Hibbs M L , Larson R S , Stacker S A , Staunton D E , Ward law A J , and Spr inger T A (1989) Structure and regulation of the leukocyte adhes ion 97 receptors for LFA-1 and its counter-receptors, ICAM-1 and ICAM-2 . Cold Spring Harbor Symposium on Quantitative Biology 54:753. Dustin M L and Spr inger T A (1988) Lymphocyte function assoc ia ted antigen-1 (LFA-1) interaction with intercellular adhes ion molecule-1 ( ICAM-1) is one of at least three mechan isms for lymphocyte adhes ion to cultured endothel ial cel ls. J. Cell Biol. 107:321. Dustin M L and Spr inger T A (1989) T-cel l receptor cross-l inking transiently st imulates adhes i veness through L F A - 1 . Nature 341:619. Dustin M L and Spr inger T A (1991) Ro le of lymphocyte adhes ion receptors in transient interactions and cell locomotion. Annu. Rev. Immunol. 9:27. Edbrooke M R , Burt DW, Chesh i re J K , and W o o P (1989) Identification of c/s-acting s e q u e n c e s responsib le for phorbol ester induction of serum amylid A gene express ion via a nuclear factor icB-like transcription factor. Mol. Cell. Biol. 9:1908. E l i ces M J , Osborn L, Takada Y , Crouse C , Luhowskyj S , Hemler M E , and Lobb R R (1990) V C A M - 1 on activated endothel ium interacts with the leukocyte integrin V L A - 4 at a site distinct from the VLA-4/ f ibronect in binding site. Cell 60:577. El l is P A , Hart D N , Col ls B M , Nimmo J C , McDona ld J E , and A n g u s H B (1992) Hodgkin 's cel ls express a novel pattern of adhes ion molecu les. Clin. Exp. Immunol. 90:117. Fawcett J , Ho lness C L L , Needham LA, Turley H, Gatter K C , M a s o n DY , and S i m m o n s DL (1992) Molecular cloning of ICAM-3 , a third l igand for L F A - 1 , constitutively exp ressed on resting leukocytes. Nature 360:481. Fe inberg A P and Vogels te in B (1983) A technique for radiolabell ing D N A restriction endonuc lease fragments to high speci f ic activity. Anal. Biochem. 132:6. Gahmbe rg C G , Nortamo P, Z immermann D, and Ruoslaht i E (1991) The human leukocyte-adhesion l igand, intercel lular-adhesion molecule 2: express ion and character izat ion of the protein. Eur. J. Biochem. 195:177. Gross -Be l la rd M, Oudet P, and C h a m b o n P (1977) Isolation of high molecular weight D N A from mammal ian cel ls. Eur. J. Biochem. 36:32. He iska L, Kantor C , P a r r T , Critchley D R , Vil ja P, Gahmberg C G , and C a r p e n O (1996) Binding of the cytoplasmic domain of intercellular adhes ion molecule-2 ( ICAM-2) to a -actinin. J. Biol. Chem. 271:26214. 98 Helander T S , Ca rpen O, Turunen O, Kovanen P E , Vaher i A , and T imonen T (1996) ICAM-2 redistributed by ezrin as a target for killer cel ls. Nature 382:265. Hogg N, Bates P A , and Harvey J (1991) Structure and function of intercellular adhes ion molecu le-1. Chem. Immunol 50:98. Ho lzmann B, Mclntyre B W , and W e i s s m a n IL (1989) Identification of a murine Peyer ' s patch-speci f ic lymphocyte homing receptor as an integrin molecule with an a lpha chain homologous to human V L A - 4 a lpha. Cell 56:37. Ho lzmann B and W e i s s m a n IL (1989) Peyer 's patch-specif ic lymphocyte homing receptors consist of a VLA-4- l i ke a lpha chain assoc ia ted either of two integrin beta cha ins , one of which is novel. EMBO J. 8:1735. Horley K J , Carpeni to C , Baker B, and Take i F (1989) Molecular cloning of murine intercellular adhes ion molecule ( ICAM-1). EMBO J. 8:2889. Hunkapi l ler T and Hood L (1989) Diversity of the immunoglobul in gene superfamily. Adv. Immunol. 44:1. Hunt S V (1979) Separat ion of lymphocyte populat ions. Handbook of exper imental immunology. D M Wei r (Ed). Blackwel l Scientif ic, Oxford. Chapter 24. Jul ius M H , S impson E, and Herzenberg L A (1973) A rapid method for the isolation of functional thymus-der ived murine lymphocytes. Eur. J. Immunol. 3:645. K a r a s u y a m a H, Kudo A , and Melchers F (1990) The proteins encoded by the V p r e . B and y 5 pre-B cel l -specif ic genes can assoc ia te with each other and with a. heavy chain. J. Exp. Med. 172:969. Katz F E , Parker M, Stanley K, Murray L J , Clark E A , and G r e a v e s M F (1985) C h r o m o s o m e mapping of cell membrane ant igens expressed on activated B cel ls. Eur. J. Immunol. 15:103. Kishimoto T K , Larson R S , Corbi A L , Dustin ML , Staunton D E , and Spr inger T A (1990) Leukocyte integrins. Spr inger et al. (eds): Leukocyte adhes ion molecules. N e w York, Spr inger-Ver lag, Chapter 1. Kohler G , Howe S C , and Milstein C (1976) Fus ion between immunoglobul in-secret ing and nonsecret ing mye loma cell l ines. Eur. J. Immunol. 6:292. Kuramoto T, Mash imo T, Koike R, Miyawaki S , Y a m a d a J , M iyasaka M, and Ser i kawa T (1994) The alymphoplas ia (a/y) mutation co-segregates with the intercellular adhes ion molecule-2 ( ICAM-2) on mouse chromosome 11. Inter. Immunol. 6 :991. 99 Li R, Nortamo P, Kantor C , Kovanen P, T imonen T, and Gahmberg C G (1993a) A leukocyte integrin binding peptide from intercellular adhes ion molecule-2 st imulates T cell adhes ion and natural killer cell activity. J. Biol. Chem. 268:21474. Li R, Nortamo P, V a l m u L, To lvanen M, Huuskonen J , Kantor C , and Gahmberg C G (1993b) A peptide from ICAM-2 binds to the leukocyte integrin C D 1 1 a / C D 1 8 and inhibits endothel ial cell adhes ion. J. Biol. Chem. 268:17513. Li R, X i e J , Kantor C , Koist inen V , Altieri D, Nortamo P, and Gahmberg C G (1995) A peptide der ived from the intercellular adhes ion molecule-2 regulates the avidity of the leukocyte integrins C D 1 1 b / C D 1 8 and C D 1 1 c / C D 1 8 . J. Cell Biol. 129:1143. Liu L, D a m e n J E , Cutler R L , and Krystal G (1994) Multiple cytokines stimulate the binding of a common 145-kilodalton protein to She at the Grb2 recognit ion site of She . Mol. Cell. Biol. 14:6926. Marl in S D and Spr inger T A (1987) Purif ied intercellular adhes ion molecule-1 ( ICAM-1) is a l igand for lymphocyte funct ion-associated antigen 1 (LFA-1) . Cell 51:813. Nortamo P, S a l c e d o R, T imonen T, Patarroyo M, and Gahmberg C G (1991) A monoc lona l antibody to the human leukocyte adhes ion molecule intercellular adhes ion molecule-2: cellular distribution and molecular distribution and molecular character izat ion of the antigen. J. Immunol. 146:2530. Ohh M, Smith C A , Carpeni to C , and Take i F (1994) Regulat ion of intercellular adhes ion molecule-1 gene express ion involves multiple m R N A stabil ization mechan isms : effects of interferon-y and phorbol myristate acetate. Blood 84:2632 O h s a w a K and Ebata N (1983) Si lver stain for detecting 10 femtogram quantit ies of protein after polyacrylamide gel electrophoresis. Anal. Biochem. 135:409. Old L, B o y s e E, and Stockert E (1965) The G leukemia ant igen. Cancer Res. 25 :813. Osborn L, Hess ion C , Tizard R, Vassa l l o C , Luhowskyj S , C h i - R o s s o G , and Lobb R (1989) Direct express ion cloning of vascu lar adhes ion molecule 1, a cytokine- induced endothel ial protein that binds to lymphocytes. Cell 59:1203. O w e n s G C , Ede lman G M , and Cunn ingham B A (1987) Organizat ion of the neural cell adhes ion molecule ( N - C A M ) gene: alternative exon usage as the bas is for different membrane-assoc ia ted domains. Proc. Natl. Acad. Sci. USA 84:294. P ier res M, Gor id is C , and Golste in P (1982) Inhibition of murine T cel l -mediated cytolysis and T cell proliferation by a rat monoclonal antibody immunoprecipitat ing two 100 lymphoid cell sur face polypeptides of 94,000 and 180,000 molecular weight. Eur. J. Immunol. 12:60. Pyszn iak A M , We lde r C A , and Take i F (1994) Ce l l surface distribution of high-avidity LFA-1 detected by soluble ICAM-1-coated microspheres. J. Immunol. 152:5241. Ra lph P and Nako inz I (1973) Inhibitory effects of lectins and lymphocyte mitogens on murine lymphomas and myelomas. J. Nat. Cancer Inst. 51:883. R e n k o n e n R, P a a v o n e n T, Nortamo P, and Gahmberg C G (1992) Express ion of endothel ia l adhes ion molecules in vivo. Increased endothel ial ICAM-2 express ion in lymphoid mal ignancies. Am. J. Path. 140:763. R o o s E (1991) Adhes ion molecules in lymphoma metastasis. Cancer Metastasis Rev. 10:33. Rothlein R and Spr inger T A (1986) The requirement for lymphocyte funct ion-assoc ia ted antigen 1 in homotypic leukocyte adhes ion stimulated by phorbol esters. J. Exp. Med. 163:1132. Sambrook J , Fri tsch E F , and Maniatas T (1989) Molecular c loning: a laboratory manua l , 2nd E d . Co ld Spr ing Harbor Laboratory, Co ld Spr ing Harbor, N Y . Sanford K K , Ear le W R , and Likely G D (1948) The growth in vitro of single isolated t issue cel ls. J. Natl. Cancer Inst. 9:229. S a n g e r F, Nicklen S , and Cou lson A R (1977) D N A sequenc ing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA Z4 :5463. Short J M , Fe rnandez J M , Sorge J A , and Huse W D (1988) X Z A P : a bacter iophage X express ion vector with in vivo excis ion properties. Nucleic Acids Res. 16:7583. S i m m o n s D, M a k g o b a M W , and S e e d B (1988) ICAM-1 , an adhes ion l igand for L F A - 1 , is homologous to the neural cell adhes ion molecule N C A M . Nature 331:624. S iu G , Hedr ick S M , and Brian A A (1989) Isolation of the murine intercellular adhes ion molecule-1 ( ICAM-1) gene: ICAM-1 enhances ant igen-specif ic T cell activation. J. Immunol. 143:3813. Sl igh J E , Bal lantyne C M , Rich S R , Hawkins HK, Smith C W , Bradley A , and Beaudet A L (1993) Inflammatory and immune responses are impaired in mice deficient in intercellular adhes ion molecule 1. Proc. Natl. Acad. Sci. USA 90:8529. 101 Somersa lo K, Ca rpen O, S a k s e l a E, Gahmberg C G , Nortamo P, and T imonen T (1995) Act ivat ion of natural killer cell migration by leukocyte integrin-binding peptide from intercellular adhes ion molecule-2 ( ICAM-2). J. Biol. Chem. 270:8629. Southern E M (1975) Detection of speci f ic sequences among D N A fragments separated by gel electrophoresis. J. Mol. Biol. 98:503. Spr inger T A (1990) Adhes ion receptors of the immune sys tem. Nature 346:425. Staunton D E , Dustin ML , Er ickson H P , and Spr inger T A (1990) The arrangement of the immunoglobul in- l ike domains of ICAM-1 and the binding sites for LFA-1 and rhinovirus. Cell 61 :243. Staunton D E , Dustin M L , and Spr inger T A (1989b) Funct ional cloning of ICAM-2 , a cell adhes ion l igand for LFA-1 homologous to ICAM-1 . Nature 339:61. Staunton D E , Marl in S D , Stratowa C , Dustin ML , and Spr inger T A (1988) Pr imary structure of intercellular adhes ion molecule 1 ( ICAM-1) demonstrates interaction between members of the immunoglobul in and integrin supergene famil ies. Cell 52:925. Take i F (1985) Inhibition of mixed lymphocyte response by a rat monoclonal antibody to a novel murine lymphocyte activation antigen (MALA-2) . J. Immunol. 134:1403. Turhan A G , Humphr ies R K , C a s h m a n J D , Cuthbert DA, E a v e s C J , and E a v e s A C (1988) Transient suppress ion of clonal hemopoies is assoc ia ted with pregnancy in a patient with a myeloproliferative disorder. J. Clin. Invest. 81:1999. van Kooyk Y , van de Wei l -van K e m e n a d e P, W e d e r P, Kuijpers T W , and Figdor C G (1989) Enhancement of LFA-1-media ted cell adhes ion by triggering through C D 2 or C D 3 on T lymphocytes. Nature 342:811. V a z e u x R, Hoffman P A , Tomita J K , Dickinson E S , J a s m a n R L , St. John T, and Gal lat in W M (1992) Cloning and characterizat ion of a new intercellular adhes ion molecule I C A M - R . Nature 360:485. von Boxberg Y , Wutz R, and Schwarz U (1990) U s e of the biotin-avidin sys tem for label l ing, isolation and characterizat ion of neural cel l -surface proteins. Eur. J. Biochem. 190:249. Voraberger G , Scha fe r R, and Stratowa C (1991) Cloning of the human gene for intercellular adhes ion molecule 1 and analysis of its 5'-regulatory region: induction by cytokines and phorbol esters. J. Immunol. 147:2777. 102 Wil l iams A F (1987) A year in the life of the immunoglobul in superfamily. Immunol. Today 8:298. Wi l l iams A F and Barc lay A N (1988) The immunoglobul in superfamily - domains for cell sur face recognit ion. Annu. Rev. Immunol. 6:381. Wuthr idge R P (1992) Monoc lona l antibodies targeting murine LFA-1 induce L F A -1/ ICAM-1- independent homotypic lymphocyte aggregat ion. Cell. Immunol. 144:22. X u H, Bickford J K , Luther E, Carpeni to C , Takei F, and Spr inger T A (1996) Character izat ion of murine intercellular adhes ion molecule-2. J. Immunol. 156:4909. X u H, G o n z a l o J A , St. Pierre Y , Wi l l iams IR, Kupper T S , Cotran R S , Spr inger TA , and Gut ie r rez -Ramos J C (1994) Leukocytosis and resistance to sept ic shock in intercellular adhes ion molecule 1-deficient mice. J. Exp. Med. 180:95. X u H, Tong IL, de Fougerol les A R , and Spr inger T A (1992) Isolation, character izat ion, and express ion of mouse ICAM-2 complementary and genomic D N A . J. Immunol. 149:2650. 103 Chapter 3: Costimulatory role of ICAM-2 in T cell response to allogeneic class II MHC T h e work presented in this chapter appears in the following publication: Carpeni to C , Pyszn iak A M , and Takei F (1997) ICAM-2 provides a costimulatory s ignal for T cell stimulation by al logeneic c lass II M H C . Scand. J. Immunol. 45 :248 3:1 In t roduc t ion T lymphocytes bearing the a|3 T cell receptor (TcR) recognize p rocessed ant igenic pept ides presented by the major histocompatibil ity complex (MHC) of proteins on the surface of antigen presenting cel ls ( A P C s ) . The interactions between T cel ls and A P C s are compr ised of both ant igen-specif ic and ant igen-nonspeci f ic components (Rothenberg, 1992; Gu inan et al., 1994). The first s tage of these interactions involves the random, low level adhes ion between A P C s and T cel ls occurr ing most prominently in lymphoid t issues (Hemler, 1990; Springer, 1990; Kishimoto et al., 1991). During this weak interaction between T cel ls and A P C s , the an t i gen :MHC complex, if present in sufficient quantity, is recognized by the T c R and a primary activation signal is del ivered within the T cell (Yague et al., 1985; Dembick et al., 1986; Sai to et al., 1987; Berzofsky et al., 1988). After ligation of the T c R , the T cell is competent to respond to var ious secondary or costimulatory s ignals (Geppert et al., 104 1990; W e a v e r and Unanue, 1990). In contrast to the primary signal del ivered by the T c R which is both ant igen-specif ic and MHC-rest r ic ted, the costimulatory s ignal is neither ant igen-speci f ic nor MHC-rest r ic ted. The second signal provides the necessa ry enhancement to the primary signal such that cytokine secret ion, cel lular proliferation, and effector function are possib le (Cerdan et al., 1992a; Ce rdan et al., 1992b; Boussiot is et al., 1993). The two signal model proposed that a second s ignal is required to amplify the T c R signal (Bretscher and C o h n , 1970; Jenk ins and Schwar tz , 1987; Schwar tz , 1990; Jenk ins , 1992). If the costimulatory s ignal is not del ivered, the T cel ls become unresponsive in an ant igen-specif ic manner. The T cel ls are anerg ic to subsequent stimulation, however they are still v iable s ince they are able to respond to exogenous ly added IL-2. Adhes ion molecules at the T cell sur face are excel lent cand idates for mediating the necessary costimulatory s ignals b e c a u s e the cel ls are ideally situated to provide a pathway for regulatory information to be transmitted. T h e s e molecu les make ideal candidates for the t ransmission of this information. The most extensively studied pair of adhes ion molecules which have been shown to deliver the required secondary signal is the C D 2 8 : B 7 (Linsley and Ledbetter, 1993; Boussiot is et al., 1996). C D 2 8 is a T cel l-specif ic molecule that can interact with B7 molecu les on A P C s . B7 can also bind a CD28- l i ke l igand, C T L A 4 . However, C T L A 4 actually transmits a negat ive costimulatory signal (Janeway and Bottomly, 1994; W a l u n a s et al., 1994; Wate rhouse etal., 1995; Tivol etal., 1996; Bluestone, 1997) Inhibition of the immune response is possib le by b lockage of the adhes ive interactions between A P C s and T cells. Ant ibodies directed against the cell adhes ion 105 molecu les mediating the initial interactions or against the T c R inhibits the delivery of the primary s ignal and the immune response is inhibited (Martz, 1987; Dustin and Spr inger, 1991; Gu inan et al., 1994). If these T cells are removed from the inhibitory condit ions, they are able to respond when rechal lenged with the initial ant igen. However , if the b lockage occurs at the level of the C D 2 8 : B 7 interaction either with ant ibodies or with C T L A 4 - l g fusion protein, a soluble high affinity counter receptor for B7 , the T cel ls become unresponsive in an ant igen-specif ic manner (Jenkins et al., 1991; Harding et al., 1992; Tan et al., 1992). They will not be able to mount any kind of an immune response when subsequent ly chal lenged with the ant igen. In summary, the functional outcome of a b lockade at the level of adhes ion or T c R signal ing prevents antigen recognition resulting in immunosuppress ion. B lockade of the C D 2 8 : B 7 interaction results in the induction of T cell anergy. The LFA-1 : ICAM-1 , -2, -3 pathways have also been examined for their potential roles in the t ransmission of costimulatory s ignals to T cel ls (van Seventer et al., 1990; van Seventer et al., 1991a; van Seventer et al., 1991b; Damle et al., 1992a; Damle et al., 1992b; de Fougero l les et al., 1994). The signi f icance of these interactions has been observed in the mixed lymphocyte reaction (MLR) . S ince m A b s against LFA-1 and the three I C A M s can either completely or partially inhibit the response, it was plausible to suspect that they were playing a role in T cell costimulat ion (de Fougero l les et al., 1994). However, A P C s in an M L R express a large number of cell sur face molecu les which may function not only to enhance T c e l l : A P C adhes ion , but a lso in cost imulat ion. A culture system was developed in order to examine these 106 potentially relevant molecules on an individual basis. The purified molecu les were coupled to microwells along with a mAb against either the T c R or the C D 3 complex. The primary s ignal can be del ivered to the T c R by the m A b and the potential cost imulatory s ignal may be transmitted by the purified protein. The proliferative T cell response to the proteins is measured by [ 3H]-thymidine uptake. Al l three LFA-1 : ICAM pathways have been shown to play a role in T cell proliferation. The extent of the T cell st imulation by the three I C A M s is identical to the relative LFA-1 affinities for each I C A M , with ICAM-1 inducing the strongest proliferation. In addit ion, the T cel ls when pr imed with ICAM-1 and ICAM-2 costimulation and subsequent ly re-stimulated through var ious receptors, display varying proliferative responses (Damle et al., 1992a; Damle et al., 1992b). It appears that ICAM-1 primed T cel ls are more respons ive to B7 costimulat ion in a secondary response than they are to B7 in a primary response. However , ICAM-2 primed T cel ls do not respond to B7 costimulat ion in a secondary response to any greater extent than in a primary response. The maturation state of the T cel ls a lso appears to be important in responding to subsequent re-stimulation. Ant igen primed or memory T cel ls do not respond well to re-stimulation with ICAM-1 or I C A M - 2 . Th is has been demonstrated in both a proliferative response as well as re lease of cytokines (Semnani et al., 1994). The preferred cho ice of costimulat ion for naive T cel ls is the LFA-1 : ICAM-1 pathway whereas memory T cel ls appear to be most respons ive to the C D 2 8 : B 7 pathway (Damle et al., 1992b). A s wel l , previously act ivated T cel ls are more B7 responsive and ICAM-1 and ICAM-2 unrespons ive than naive T cel ls. 107 Although purified proteins and ant i -TcR or an t i -CD3 m A b activation has uncovered much information about costimulation, the system used is highly artificial and has severa l potential shortcomings. The mAb affinity for the T c R / C D 3 complex is much higher than that of the an t igen /MHC (Sagerstrom et al., 1993) and therefore may not be representat ive of the actual physiological interactions that occur between T cel ls and A P C s . Furthermore, although cell surface ICAM-2 has been shown to mediate adhes ion to L F A - 1 + cel ls (Xu et al., 1992; X u et al., 1996), it may not be as effective in delivering a costimulatory s ignal as the purified protein due to g lycocalyx hindrance. The functional role of cell surface ICAM-2 in T cell activation is yet to be understood. The objective of the work presented in this chapter was to examine the role of ICAM-2 in T cell activation under more physiological condit ions. The LFA-1 : ICAM-2 pathway w a s functionally isolated by creating antigen presenting cel ls. Mur ine fibroblast L cel ls (H-2 k) express ing c lass II l - E d molecules were transfected with the murine ICAM-2 c D N A and examined for the ability to stimulate a l logeneic splenic T cel ls (H-2 k ) . The proliferative response was compared to that with untransfected and ICAM-1- t ransfected l - E d L cel ls. The induction of an anergic state was also examined by rechal lenging the stimulated T cells in a secondary response. A s wel l , the T cell stimulation w a s repeated with two separate cel ls express ing the l - E d and the ICAM-2 molecu les. Al l the exper iments were des igned to address the issue of whether cell sur face ICAM-1 and ICAM-2 merely enhance T c R recognit ion of a l logeneic M H C or whether they actually provide a costimulatory signal essent ia l for the avo idance of anergy. 108 3:2 M a t e r i a l s a n d M e t h o d s 3:2.1 Animals C 3 H / H e , B A L B / c , and C 5 7 B L / 6 mice used in this study were bred at the Joint An ima l Facil ity of the B .C . C a n c e r Resea rch Centre from the founders purchased from J a c k s o n Laborator ies (Bar Harbor, ME) . 3:2.3 Cell lines and antibodies The murine fibroblast L cell line (H-2 k ; Sanford et al., 1948) was maintained in D M E M containing 10% F C S . The murine fibroblast line RT10.3 .BCI (referred to as R T 1 0 . 3 here after) is a transfected L cell line which expresses the murine c lass II M H C l - E d molecule (Germain and Quil l , 1986; Ruberti et al., 1992) and was a generous gift from Dr. W . Jeffer ies (Biotechnology laboratory, University of British Columbia) . It was a lso maintained D M E M + 10% F C S . Al l ant ibodies were used as purified Ig. The ant ibodies which recognize the murine cell sur face ICAM-1 ( l gG 2 a ) , ICAM-2 ( l gG 2 a ) , C D 1 1 a ( l gG 2 b ) and C D 4 9 d ( l gG 2 b ) have been descr ibed in chapter 2. The rat ant i -mouse V C A M - 1 (A429, l g G 2 a ) was purchased from Pharmingen (San Diego, C A ) . The biotinylated m A b s to l -A d ( A M S -32.1 , l g G 2 a ) , l - E d (AMS-16 , l g G 2 a ) , B7-1 (1G10, l g G 2 a ) , and B7-2 ( G L 1 , l g G 2 a ) were a lso purchased from Pharmingen. Al l commercia l ant ibodies were d ia lyzed against 1X P B S to remove sodium az ide. The hybridoma cell l ines that produce rat ant i -mouse C D 2 4 (M1/69.16.11, A T C C TIB 125, l g G 2 b ) (Springer et al., 1978) and murine anti-rat 109 IgK ( R G 7 / 9 . 1 , A T C C TIB 169, l g G 2 b ) were obtained from Amer ican Type Culture Col lect ion (Rockvi l le, MD) . 3:2.4 Transfection of l-E* L cells with murine ICAM-1 and ICAM-2 cDNAs The murine ICAM-1 (Horley et al., 1989) and ICAM-2 c D N A s in the express ion vector p B C M G S N e o (Karasuyama et al., 1990) were transfected into R T 1 0 . 3 cel ls by the poly-L-ornithine method (Dong et al., 1993). Transfectants were se lected in D M E M containing 1 0 % F C S and G 4 1 8 (500 |ag/ml, Canad ian Life Technolog ies) . Bulk transfectants that expressed high levels of ICAM-1 and ICAM-2 were isolated by panning directly with purified anti- ICAM-1 or ant i - ICAM-2 mAb immobi l ized on petri d ishes (Falcon 1001, Becton Dickinson). The isolated cel ls were expanded and express ion levels of var ious cell surface molecules were tested by flow cytometry as descr ibed in the previous chapter. Ce l ls (3 X 10 5) were stained with I C A M - 1 , ICAM-2 , V C A M - 1 , or C D 2 4 antibodies (4 (ag/ml). The secondary stain w a s T IB169-F ITC (descr ibed in 3:2.3) and dead cel ls were stained with propidium iodide (2 u.g/ml). The cel ls were also stained with biotinylated l-A d , l -E d , B7 -1 , and B7-2 ant ibodies and counter-stained with streptavidin-FITC. The stained cel ls were ana lyzed on a F A C S t a r (Becton Dickinson). 3:2.5 Cell adhesion assay Adhes ion of PMA-act iva ted splenic T cel ls to the ICAM-1 and ICAM-2 transfected R T 1 0 . 3 cel ls was a s s e s s e d as descr ibed in the previous chapter. 110 3:2.6 Proliferation assay of primary allogeneic response Splen ic T cel ls were isolated from C 3 H / H e (H-2 k) mice (2-10 months old) using nylon wool as descr ibed in the previous chapter (Julius et al., 1973). L cel ls (H-2 k ) and R T 1 0 . 3 (H-2 k express ing l -E d ) cel ls were harvested with P B S + 2.5 m M E D T A and irradiated with 12,000 rads using a Phil ips R T 250 X-ray machine. Sp len ic T cel ls (2.5 X 10 5 ) were combined with varying numbers of irradiated stimulator cel ls in 200 uJ of R P M I 1 6 4 0 + 5 % F C S + 5 X 10~5 M p -mercaptoethanol (P-ME) in 96-well flat bottom plates (Fa lcon 3072, Becton Dickinson). Ant ibodies for inhibition studies were used at 4 u.g/ml. The plates were cultured for 4, 5, or 6 days at 37°C in a 5 % C 0 2 humidif ied a tmosphere. The cultures were then pulsed with 1 u.Ci/well of [methyl- 3H]-thymidine ( 3 H-TdR) (DuPont, Boston, MA) for a further 8 hrs as a measure of newly synthes ized D N A . The radiolabeled cel ls were then harvested onto filters and the radioactivity on the filters w a s measured in a p-plate liquid scintillation counter ( L K B Wa l lac 1205, Turku, Finland). The results are expressed as mean of triplicate exper iments ± standard error mean (cpm + S E M ) . In a separate set of exper iments, 2.5 X 1 0 5 sp len ic T cel ls were combined with 1.25 X 10 4 irradiated untransfected RT10 .3 cel ls and 5 X 1 0 3 L cel ls express ing either ICAM-1 (L: ICAM-1) or ICAM-2 (L: ICAM-2) from the previous chapter. The proliferative response to the mixed cell stimulation was measured as descr ibed above. 111 3:2.7 Secondary stimulation assay Splen ic T cel ls (1.0 X 10 7) from C 3 H / H e mice were cultured with irradiated R T 1 0 . 3 transfectant cel ls (8 X 10 5) in 5 ml of RPM11640 + 5 % F C S + 5 X 10" 5 M p -ME in 6-well flat bottom plates (Falcon 3046, Becton Dickinson) for 6 days. The T cel ls were then harvested and washed three t imes with 10 ml RPM11640 + 5 % F C S . The T cel ls were incubated without any stimulus for overnight in RPM11640 + 5 % F C S + 5 X 10" 5 M p -ME. 2 X 1 0 5 of these T cel ls from each primary stimulation were then combined with 3 X 1 0 5 irradiated (7500 rads) B A L B / c (H-2 d ) , C 3 H / H e (H-2 k ) , or C 5 7 B L / 6 (H-2 b ) sp leen cel ls in 96 well round bottom plates ( Fa lcon 3077, Becton Dickinson) in 200 ul of RPM11640 + 5 % F C S + 5 X 10" 5 M p-ME and incubated for 4 days . T cell proliferation to the secondary st imulus was measured as descr ibed for the primary a l logeneic response. Inhibition studies used purified C D 1 1 a antibody at a concentrat ion of 4 jag/ml. 3:3 R e s u l t s 3:3.1 Expression of ICAM-1 and ICAM-2 on RT10.3 cells R T 1 0 . 3 cel ls ( l -E d - t ransfected L cells) express ing murine ICAM-1 or ICAM-2 were generated by the transfection of the appropriate c D N A s , fol lowed by panning with one of the ant i - ICAM mAbs . F low cytometric analysis of the transfected cel ls showed that the express ion levels of ICAM-1 and ICAM-2 were equivalent on the respect ive cel ls (Figure 13a). There was a low level of endogenous ICAM-2 express ion on the untransfected and the ICAM-1-transfected cel ls. Four other potential costimulatory 112 A) no 1 ° A b ICAM -1 A b ICAM -2 A b V C A M - 1 A b H S A A b C O F i g u r e 13 Flow cytometric analysis of RT10.3 cells. RT10 .3 cel ls transfected with the murine ICAM-1 and ICAM-2 c D N A were ana lyzed for var ious cel l sur face molecules. A) The top panel of histograms represents I C A M - 1 , ICAM-2 , V C A M - 1 , and M1/69 express ion on the transfected RT10 .3 cel ls as determined by indirect immunof luorescence. B) The bottom panel of histograms (next page) represents expression of M H C c lass II l-A and l-E molecules, as well as the cotimulatory molecu les B7-1 and B7-2 . (figure l e g e n d o n p r e v i o u s page) molecu les , V C A M - 1 , B7 -1 , B7-2 , and heat stable antigen (HSA, CD24 ) , were not detected on these cel ls (Figure 13a and 13b). The cel ls a lso expressed equivalent levels of l - E d (Figure 13b). I-Ad was not detected on the surface of the R T 1 0 . 3 cel ls. 3:3.2 Adhesion of PMA-activated splenic T cells to RT10.3 cells In order to determine the relative binding capaci t ies of the I C A M s on the transfected R T 1 0 . 3 cel ls, the adhes ion of PMA-act iva ted splenic T cel ls to the R T 1 0 . 3 cel ls w a s determined. Approximately 18% of T cel ls adhered to untransfected cel ls whereas express ion of ICAM-1 and ICAM-2 on RT10 .3 cel ls increased adhes ion to 8 4 % and 6 7 % respect ively (Figure 14). The I C A M - 1 - and ICAM-2-media ted adhes ion w a s inhibited by the ant i -LFA-1 mAb whereas the isotype control ant ibody (ant i -CD49d mAb) had no effect on adhes ion. The ant i- ICAM-1 mAb inhibited T cell adhes ion to the ICAM-1- t ransfected RT10 .3 cel ls and the ant i - ICAM-2 m A b inhibited the adhes ion to ICAM-2- t ransfected RT10 .3 cel ls. T h e s e results indicate that ICAM-1 and ICAM-2 on the transfected RT10 .3 cel ls are functional and readily mediated T cell adhes ion . The low level of endogenous ICAM-2 appears to be too low to contribute significantly to the overal l adhes ion . 3:3.3 Primary allogeneic response of splenic T cells against RT10.3 cells The role of ICAM-2 in a T cell response against a l logeneic c lass II M H C w a s examined (Figure 15). Murine splenic T cel ls from C 3 H / H e mice (H-2 k ) were st imulated with I C A M - 1 - and ICAM-2-transfected RT10 .3 cel ls (H-2 k express ing l -E d ) . 115 0 25 50 75 100 cells bound (%) F i g u r e 14 Adhesion of splenic T cells to RT10.3 cells expressing ICAM-1 and ICAM-2. R T 1 0 . 3 cel ls were grown in 96-well flat bottom wel ls. To each wel l , 10 5 PMA-ac t i va ted splenic T cel ls labeled with calcein A M were added with the appropriate blocking antibody (4 |ig/ml). The binding assay and w a s h e s were performed as descr ibed previously. Resul ts are expressed as a mean of triplicate wel ls ± S E M . 116 Irradiated RT10.3 cells H-2k:l-Ed F i g u r e 15 Schematic representation of T cell stimulation by allogeneic MHC class II and ICAM-2. Murine splenic T cel ls from C 3 H / H e mice (H-2 k) were purif ied by a nylon wool co lumn and combined with irradiated stimulator cel ls (RT10.3) express ing ICAM-1 or ICAM-2 . T h e RT10 .3 cel ls (H-2 k) express the al logeneic M H C c lass II molecule l -E d . After 5 days , the cel ls were pu lsed with [ 3H]-TdR and the proliferative responses were examined as descr ibed in Mater ia ls and Methods. Since the R T 1 0 . 3 cel ls are of C 3 H / H e origin, the C 3 H / H e T cel ls will respond to the a l logeneic l - E d M H C expressed on the surface of the RT10 .3 cel ls. The optimal a l logeneic response, determined by [ 3H]-TdR incorporation, w a s ach ieved by incubating 2.5 X 1 0 5 sp lenic T cel ls with 1.5 X 1 0 4 RT10 .3 cel ls (Figure 16) for 5 days (Figure 17). The al logeneic response pattern to the RT10 .3 cel ls is similar regardless of whether the cel ls are ICAM-transfected or untransfected. Under these condit ions, the T cell response to untransfected RT10 .3 cells (nearly 20000 cpm) was substantial ly higher than the control ( less than 200 cpm) in which no RT10 .3 cel ls were added (Figure 18). The express ion of ICAM-1 and ICAM-2 on RT10 .3 cel ls significantly enhanced the T cell stimulation by 2.5-3 fold. Ant i -LFA-1 mAb inhibited the stimulation of T cel ls by I C A M - 1 - and ICAM-2-transfected RT10 .3 cel ls, whereas it had little effect on the stimulation with the untransfected RT10 .3 cel ls. The T cell stimulation with untransfected RT10 .3 cel ls was LFA-1- independent . A s wel l , the low level of endogenous ICAM-2 express ion did not appear to have any stimulatory effect on the T cel ls s ince the an t i -CD11a mAb did not have any affect on the background response. Ant i - ICAM-1 and ant i - ICAM-2 mAbs also specif ical ly inhibited T cell stimulation with the transfected RT10 .3 cel ls express ing the relevant I C A M s , although the levels of inhibition by these m A b s were lower than that of the ant i -LFA-1 mAb. Control mAb (ant i -CD49d) did not have an effect on the stimulation. In order to el iminate the possibil i ty that the RT10 .3 cells were behaving as feeder cel ls and supplying proliferative cytokines to the T cel ls, 4 day supernatants from irradiated and non-118 60 100 1000 10000 100000 Number of stimulator cells F i g u r e 16 Dose response of splenic T cells to RT10.3 expressing ICAM-1 or ICAM-2. Sp len ic T cel ls (2.5 x 10 5) were cultured with increasing numbers of each irradiated R T 1 0 . 3 transfectant. Proliferative responses were measured on day 5 and expressed as a mean of triplicate wel ls ± S E M . 119 60 - 40-E Q. c o a mmm • > CO o Q. o o £ 20 OC "O H CO o..-. RT10.3 RT10.3:ICAM-1 RT10.3-.ICAM-2 I 4 5 6 Days in culture 7 8 F i g u r e 17 Kinetics of the allogeneic T cell response to ICAM-1- or ICAM-2-transfected RT10.3 cells. Sp len ic T cel ls (2.5 x 10 5) were cultured with each type of irradiated R T 1 0 . 3 transfectant (1.5 x 10") for the indicated periods. The proliferative T cel l r esponses were determined as descr ibed previously. 120 80 • RT10.3 responses + CD11a Ab + ICAM-1 Ab + ICAM-2 Ab + CD49d Ab F igu re 18 Effects of antibodies on allogeneic T cell response to RT10.3 cells. Sp len ic T cel ls (2.5 x 10 s) were cultured with irradiated I C A M - 1 - or ICAM-2-t ransfected RT10 .3 cel ls (1.5 x 10 4) in the presence of the indicated ant ibodies (4 | ig/ml). Proliferative responses were measured on day 5 as descr ibed previously and expressed as a mean of triplicate cultures + S E M . irradiated R T 1 0 . 3 cel ls (transfected and untransfected) were incubated with the T cel ls. The T cel ls did not proliferate to any notable level (all were under 1000 cpm). 3:3.4 Allogeneic response to mixed stimulator cells There are two possib le explanat ions for the increased T cell response to I C A M -1 and ICAM-2 express ion on RT10 .3 cel ls. One explanat ion is that express ion of ICAM-1 and ICAM-2 increases the LFA-1-dependent physical interaction of T cel ls to the R T 1 0 . 3 cel ls and therefore results in enhanced T c R recognit ion of pep t i de :MHC complex. The other possib le explanat ion is that the I C A M s present on the RT10 .3 cel ls provide costimulatory s ignals which augment the primary s ignal transmitted through the T c R . In order to distinguish between these two poss ib le mechan isms, T cel ls were combined with untransfected RT10 .3 cel ls and L cel ls transfected with ICAM-1 or ICAM-2 (Figure 19). If the ICAM-1 and ICAM-2 molecu les were actually transmitting costimulatory s ignals, then the two signals may be spatially separated and T cell proliferation should still be observed. Untransfected RT10 .3 cel ls express ing I-E d were combined with L cel ls express ing ICAM-1 (L: ICAM-1) or ICAM-2 (L: ICAM-2) but not l -E d , and used to stimulate splenic T cel ls from C 3 H / H e mice (H-2 k ) . The mixture of these cel ls were able to stimulate T cells above the level of control cultures in which untransfected L cel ls plus RT10 .3 cel ls were used to stimulate T cel ls (Figure 20). The presence of untransfected L cel ls did not increase the T cell response to untransfected RT10 .3 cel ls. However, the presence of L: ICAM-1 and L : ICAM-2 with untransfected RT10 .3 cel ls greatly increased the T cell response above the control 122 F i g u r e 19 Schematic representation of T cell response to mixed stimulators. The stimulatory role of ICAM-2 in al logeneic T cel l response was examined using a mixture of two populat ions of stimulator cel ls separate ly express ing ICAM-2 and l -E d . Untransfected R T 1 0 . 3 (H-2 k : l -E d ) cel ls which do not express ICAM-1 or ICAM-2 are combined with L cel ls (H-2 k) transfected with ICAM-2 and used to stimulate T cel ls (H-2 k). media only! UCAM-2 UCAM-1 L cells RT10.3 + L:ICAM-2 RT10.3 + L:ICAM-1 RT10.3 + L cells RT10.3:ICAM-2 RT10.3:ICAM-1 RT10.3 6 20 40 60 80 [3H]-TdR incorporation (cpm/1000) Figure 20 Stimulation of T cells with RT10.3 and ICAM-2-transfected L cells. Irradiated untransfected RT10 .3 cel ls (1.25 x 10") were combined with irradiated L cel ls (5 x 10 3) transfected with ICAM-1 ( L I C A M - 1 ) or ICAM-2 ( L I C A M - 2 ) and incubated with 2.5 x 10 5 splenic T cel ls. A s negative controls, T cel ls were incubated without stimulator cel ls (media alone), or with L cel ls, L I C A M - 1 , or L I C A M - 2 without R T 1 0 . 3 cel ls. ICAM-1-t ransfected RT10 .3 (RT10.3: ICAM-1) and ICAM-2-t ransfected R T 1 0 . 3 (RT10.3 : ICAM-2) were also used as stimulators in posit ive controls. The proliferative responses of T cel ls were measured on day 5 and expressed as a mean of triplicate cultures + S E M . 124 culture. Al though the T cell response to L I C A M - 1 and L I C A M - 2 was slightly greater than that against untransfected L cel ls, the responses to RT10 .3 + L: ICAM-1 and R T 1 0 . 3 + L I C A M - 2 were still greater than the additive effect of responses to RT10 .3 cel ls and to L I C A M - 1 or L I C A M - 2 . These results indicate that ICAM-1 and ICAM-2 on R T 1 0 . 3 cel ls not only can enhance the physical interaction between T cel ls and R T 1 0 . 3 cel ls, but that they can also deliver an essent ia l costimulatory molecule through LFA-1 on T cel ls. 3:3.5 Secondary stimulation A b s e n c e of a costimulatory signal able to augment the primary s ignal from the T c R leads to an anergic state in which the T cel ls are functionally para lyzed upon subsequent restimulation with the appropriate antigen. In order to examine whether ICAM-1 and ICAM-2 mediated s ignals are sufficient to avert an anergic state, T cel ls st imulated with RT10 .3 transfected with ICAM-1 or ICAM-2 were later combined with a l logeneic A P C s and the T cell response was measured (Figure 21). Sp len ic T cel ls from C 3 H / H e mice (H-2 k) stimulated with RT10 .3 cel ls (H-2 k express ing l -E d ) for 6 days were harvested and subsequent ly rechal lenged with irradiated sp leen cel ls from B A L B / c (H-2 d ) , C 5 7 B L / 6 (H-2 b ) , or C 3 H / H 3 (H-2 k) mice. T cel ls st imulated with I C A M -transfected RT10 .3 cel ls in primary culture displayed a vigorous response to B A L B / c sp leen cel ls in the secondary stimulation (Figure 22). In contrast, T cel ls st imulated with untransfected RT10 .3 cel ls in primary culture did not respond significantly to B A L B / c sp leen cel ls. T cells recovered from primary cultures, regardless of I C A M 125 C3H/He (H-2k) C3H/He (H-2K) BALB/c (H-2d) C57BL/6 (H-2b) primary allogeneic response RT10.3 (H-2k:I-Ed) irradiated spleen cells • 6 days overnight culture in media only 4 days l3H]-TdR uptake Figure 21 Schematic representation of secondary stimulation of T cells. Sp len ic T cel ls (H-2 k) are cultured with the irradiated RT10 .3 transfectants in a primary a l logeneic response for 6 days. The T cel ls are then transfered into fresh med ia without stimulators for overnight. T h e T cel ls are then combined with irradiated sp leen cells from either a C 3 H / H e mouse (H-2 k), a B A L B / c mouse (H-2 d), or a C 5 7 B L / 6 (H-2 b) and cultured for 4 days before examining the proliferative response. S e c o n d a r y s t i m u l u s : media BALB/c BALB /c (+Ab) C57BL /6 C 5 7 B L / 6 (+Ab) C 3 H / H e medial B A L B / c 1 B A L B / c (+Ab) C 5 7 B L / 6 C 5 7 B L / 6 (+Ab) C 3 H / H e med ia B A L B / c B A L B / c (+Ab) C 5 7 B L / 6 C 5 7 B L / 6 (+Ab) C 3 H / H e ZZZ3I T77?f\ Primary stimulus- RT10.3:ICAM-2 y///////y///77777. 50 100 [ 3 H ] - T d R i n c o r p o r a t i o n (cpm/1000) 150 F i g u r e 22 Secondary responses of T cells. Sp len ic T cel ls were st imulated with R T 1 0 . 3 , R T 1 0 . 3 : I C A M - 1 , or RT10 .3 : ICAM-2 cel ls for 6 days. The T cel ls from each response were then harvested and cultured ovdernight in media before being comb ined with irradiated sp leen cel ls from the indicated strains of mice in round-bottom wel ls (2 x 10 5 T cel ls + 3 x 10 5 sp leen cel ls). An t i -CD11a antibody w a s added where indicated by (+Ab) at a concentration of 4 u.g/ml. Proliferation w a s examined on day 4 and expressed as a mean of triplicate cultures + S E M . 127 express ion on the stimulator cel ls, showed equal and substant ial responses to C 5 7 B L / 6 sp leen cel ls but not to syngene ic C 3 H / H e sp leen cel ls. This indicated that the T cel ls were functional and had remained viable in culture. The responses in the secondary stimulation were inhibited by the ant i -LFA-1 mAb. T h e s e results indicate that T cel ls st imulated by untransfected RT10 .3 cel ls in primary cultures became specif ical ly non-responsive to H-2 d cel ls. T h e s e anergic cel ls were still v iable s ince they were able to respond to ant i -CD3 cross-l inking (3 u.g/ml) with P M A (5 ng/ml) [RT10.3: (3.52 ± .23) x 1 0 5 cpm; RT10 .3 ICAM-1 : (3.15 ± .37) x 1 0 5 cpm; R T 1 0 . 3 I C A M - 2 : (2.95 + .14) x 1 0 5 cpm]. In contrast, primary stimulation with I C A M -transfected R T 1 0 . 3 cel ls greatly enhanced the T cell responses to a l logeneic B A L B / c sp leen cel ls in secondary stimulation. 3:4 Discussion The functional role of cell surface ICAM-2 in the stimulation of T cel ls with a l logeneic c lass II M H C was examined in this chapter. In contrast to I C A M - 1 , whose role in T cel l activation has been demonstrated in var ious sys tems (van Seventer et al., 1990; van Seventer etal., 1991a; Damle etal., 1992b; Damle etal., 1992c), the role of I C A M - 2 has not been thoroughly investigated. The functional role of ICAM-2 in T cell activation has been previously investigated using ant i - ICAM-2 mAb to inhibit the T cell response or testing the effects of recombinant I C A M - 2 / F c fusion proteins on T cel ls st imulated by ant i -TcR cross-l inking (Damle et al., 1992a; Damle et al., 1992b). In this study we have used murine fibroblast L cel ls transfected with l - E d and ICAM-1 or 128 I C A M - 2 . T h e s e cel ls are able to act as A P C s and can stimulate a l logeneic T cel ls from C 3 H / H e (H-2 k ) mice. This system has a number of advantages. First, the function of cell sur face ICAM-2 , not purified recombinant proteins, can be determined. S e c o n d , the stimulation of primary T cel ls and not c loned T cell l ines can be examined , s ince the T cell response to a l logeneic c lass II M H C in primary stimulation is strong enough for detect ion. Third, stimulation of T cel ls with A P C s , not by a n t i - T c R / C D 3 c ross-linking, is being studied which is more physiological ly relevant. The affinity of mAb for T c R / C D 3 is est imated to be 1 0 3 - 1 0 4 fold higher than that of T c R for A g / M H C (Matsui et al., 1991; W e b e r et al., 1992). The difference in affinity may be important when T c e l h A P C interactions are examined. Finally, l -E d - t ransfected L cel ls (RT10.3) do not express B7 or H S A , two cell surface molecules demonstrated to del iver costimulatory s ignals to T cel ls (Linsley et al., 1991; G immi et al., 1991; Liu et al., 1992). Therefore, the functional role of murine ICAM-2 on surrogate A P C s in T cell stimulation can be examined in the absence of these important costimulatory molecules. This study has demonstrated that ICAM-2 expressed on RT10 .3 cel ls is not only able to mediate LFA-1-dependent T cell adhes ion but a lso plays a signif icant role in the activation of resting cel ls. Sp len ic T cel ls (H-2 k) mount a relatively low but signif icant proliferative response to untransfected RT10 .3 cel ls (H-2 k) express ing l -E d . However , those stimulated T cells s e e m to become unresponsive when cha l lenged with H-2 d sp leen cel ls from B A L B / c mice in a secondary stimulation. Th is does not s e e m to be due to cell death in the primary culture because they are able to respond to a n t i - C D 3 / P M A stimulation and to third party a l logeneic sp leen cel ls from C 5 7 B L / 6 mice 129 (H-2 b ) . In fact, the ant i -H-2 b response of these cel ls was much higher than the secondary response to H-2 d sp leen cel ls. Therefore, the failure of these T cel ls to mount a secondary ant i -H-2 d response cannot be expla ined by the lack of clonal expans ion of the ant i -H-2 d reactive T cel ls in the primary culture and implies an active mechan i sm to induce an anergic state. In contrast, the express ion of ICAM-1 or I C A M -2 on R T 1 0 . 3 cel ls greatly enhances the stimulation of T cel ls in an LFA-1 -dependen t manner in primary cultures. This stimulation a lso prevents T cell un respons iveness to subsequent a l logeneic stimulation. Thus , T cells stimulated with ICAM-transfected R T 1 0 . 3 cel ls in primary cultures display a vigorous response to H-2 d sp leen cel ls in secondary stimulation. The T cell response to RT10 .3 cel ls is reflected in the adhes ion profile of P M A -act ivated sp lenic T cel ls to RT10 .3 cel ls. Therefore, the enhanced response of T cel ls to ICAM-transfected RT10 .3 cel ls may be a result of improved T cell adhes ion to A P C s which would facilitate more efficient T c R recognition of A g / M H C . However, this study has demonstrated that ICAM-1 or ICAM-2 and l - E d expressed on separate L cel ls are still able to stimulate T cel ls. This is further indicative of the costimulatory potential of ICAM-2 as well as ICAM-1 on A P C s . Similar exper iments have a lso shown that I C A M -1 can provide a costimulatory s ignal (van Seventer et al., 1991b). T cel ls are able to proliferate when the primary ant igen-specif ic signal and the costimulatory s ignal are spatial ly separate (Dubey et al., 1995). These findings are consistent with the results from this study indicating that cell surface ICAM-1 and ICAM-2 provide a necessary costimulatory s ignal which augments the signal provided by engagement of the T c R 130 with A g bound to M H C . Other groups have shown that purified ICAM-1 and ICAM-2 can provide potent costimulatory s ignals to ant i -CD3-act ivated T cel ls. However , this study demonstrates that ICAM-2 on A P C s can also act as a potential costimulatory molecule. T h e s e results demonstrate that ICAM-2 and al logeneic c lass II M H C express ion is sufficient for T cell activation in addition to rescuing the T cell from a state of a l lo -unrespons iveness. The quest ion as to whether TcR-der ived and costimulatory s ignals may be present on separate cel ls has been addressed by severa l groups. Al though the results from this study and by others (van Seventer et al., 1990; van Seventer et al., 1991) indicate that the two signals may be present on separate cel ls, other groups have argued and demonstrated the opposite (Galvin et al., 1992; Liu and Janeway , 1992; J a n e w a y and Bottomly, 1994). T h e s e investigators argue that the two s ignals must be on the s a m e cell for an optimal T cell response. If the s ignals could be present on separa te cel ls, then autoreactive T cel ls could become activated on a regular bas is . This is not observed and the quest ion as to whether the two s ignals can be del ivered by two cel ls requires further investigation to determine if the two cell stimulation is a reflection of the t issue culture condit ions. The costimulatory s ignal generated by the interaction between B7 on A P C s and C D 2 8 on T cel ls has been extensively studied (Linsley and Ledbetter, 1993; Boussiot is et al., 1996). This study suggests that ICAM-2 as well as ICAM-1 may provide an alternative costimulatory pathway for T cel ls when A P C s do not express B7 . However , the exact nature of the ICAM-media ted signal appears to be unclear. Boussiot is et al. 131 (1993) have previously demonstrated that NIH3T3 cel ls transfected with human I C A M -1 and D R 7 were able to induce T cell proliferation in primary cultures. However , the st imulated T cel ls were not able to respond to NIH3T3 cel ls transfected with D R 7 and ICAM-1 or B7 in secondary stimulations. This apparently contradicts the results from our study which indicates that express ion of ICAM-1 or ICAM-2 and a l logeneic c lass II M H C provides adequate stimulation for T cel ls in the avers ion of an anerg ic state. However , the fact that we used sp leen cel ls, instead of transfected NIH3T3 cel ls, as stimulators in the secondary culture may help to explain the anomaly. Pro fess iona l A P C s present in a sp leen cell population express ing B7 , I C A M s , as well as c lass II M H C may be responsib le for the secondary stimulation in our sys tem. This secondary stimulation is LFA-1-dependent as demonstrated by the inhibition using an t i -CD11a mAb . Whether the b lockade is occurring at the level of adhes ion or cost imulat ion is uncertain. What is c lear is that p resence of ICAM-1 or ICAM-2 in primary culture is ab le to rescue T cel ls from an state of unrespons iveness to a l logeneic sp leen cel ls in a secondary stimulation. It has been previously shown by one group that ICAM-1 costimulat ion in a primary culture enhances B7 respons iveness in a secondary response (Damle etal., 1992b). A l though the profile of cytokines re leased in these exper iments was not examined , it has been investigated by others (Boussiot is et al., 1993; Boussiot is et al., 1994; S e m n a n i et al., 1994). W h e n T cells receive a costimulatory s ignal v ia I C A M - 1 , IL-2 is re leased. This has also been observed when B7-1 provides the secondary s ignal . If ICAM-1 and B7-1 are together providing the cost imulat ion, the level of IL-2 132 re leased increases drastically if either are providing the signal a lone (Dubey et al., 1995). However , the cytokine profile for ICAM-2 has not been examined . If IL-2 is detected with ICAM-2 stimulation, it will provide further support that ICAM-2 can in fact act as a c lass ic costimulatory molecule. It may also display other cytokines reflecting the unique immune response initiated by the distribution pattern of ICAM-2 . Cost imulat ion by LFA-1 : ICAM-1 interaction has shown that G M - C S F is a lso re leased (Semnan i et al., 1994). Whereas , costimulation by L F A - 3 : C D 2 results in IL-5 re lease in addit ion to IL-2. The finding that murine ICAM-2 on A P C s may function as a costimulatory molecule for T cel ls in the absence of B7 or H S A has important implications to antigen presentat ion by cell types not normally considered to be 'professional ' A P C s . Inflammatory cytokines such as IFN-y are known to induce c lass II M H C express ion on s o m e non- lymphoid cel ls that do not express B7 or H S A (O'Connel l and Edid in, 1990; Farrar and Schreiber , 1993). S o m e of these cel ls, including endothel ial cel ls, constitutively express ICAM-2 . It is possib le that those cel ls may function in antigen presentat ion under certain condit ions. The capacity for an endothel ial cell line to function as an A P C has been reported (St. Louis et al., 1993). In addit ion, demonstrat ion that the two required s ignals for T cell activation may be del ivered by two separa te cel ls a lso raises the possibility that C A M s may not be an absolute requirement on an A P C . A cell with antigenic peptides and M H C on the sur face but no costimulatory adhes ion proteins might still be able to elicit a T cell response if an adjacent cell exp resses the appropriate costimulatory molecules. Furthermore, our 133 results ra ises the possibil ity that non-professional A P C s express ing c lass II M H C and I C A M s but not B7 or H S A may play a role in graft rejection and auto immune d i seases . M ice treated with ant i -LFA-1 and ant i- ICAM-1 mAbs while undergoing a card iac allograft were found to be tolerant to subsequent skin allograft (Isobe et al., 1992; Isobe and lhara, 1993). Further studies are required to a s s e s s the functional role of I C A M - 2 in vivo as to whether ant i - ICAM-2 mAb may be able to block a host versus graft response to t issues such as the endothel ium. 134 3:5 R e f e r e n c e s Berg N N and Ostergaard HL (1995) Character izat ion of intercellular adhes ion molecule-1 ( ICAM-1 )-augmented degranulat ion by cytotoxic T cel ls. ICAM-1 and anti-C D 3 must be co- local ized for optimal adhes ion and stimulation. J. Immunol. 155:1694. Berzofsky J , Brett S , Streicher HZ, and Takahash i H Z (1988) Ant igen process ing for presentat ion to T lymphocytes: functional mechan isms and implications for the T-cel l repetoire. Immunol. Rev. 106:5. B luestone J A (1997) Is C T L A - 4 a master switch for peripheral T cell to lerance? J. Immunol.! 758:1989. Bouss io t is V A , F reeman G J , Gray G S , Gr ibben J G , and Nadler L M (1993) B7 but not intercellular adhes ion molecule-1 costimulation prevents the induction of human al loant igen-speci f ic tolerance. J. Exp. Med. 178:1753. Boussiot is V A , F reeman G J , Gr ibben J G , and Nadler L M (1996) The role of B 7 - 1 / B 7 -1 : C D 2 8 / C T L A - 4 pathways in the prevention of anergy, induction of productive immunity and down-regulat ion of the immune response. Immunol. Rev. 153:5. Bouss io t is V A , F reeman G J , Griffin J D , Gray G S , Gr ibben J G , and Nadler L M (1994) C D 2 is involved in maintenance and reversal of human al loant igen-specif ic c lonal anergy. J. Exp. Med. 180:1665. Bretscher P and C o h n M (1970) A theory of sel f-nonself discrimination. Science 169:1042. Brunet J F , Denizot F, Luciani M F , Roux-Dosse to M, S u z a n M, Mattei M G , and Goldste in P (1987) A new member of the immunoglobul in superfamily - C T L A - 4 . -Nature 328:267. C e r d a n C , Martin Y , Courcou l M, Brailly H, M a w a s C , Birg F, and Ol ive D (1992a) Pro longed IL-2 receptor a lpha /CD25 express ion after T cell activation via the adhes ion molecu les C D 2 and C D 2 8 . Demonstrat ion of combined transcriptional and post-transcript ional regulation. J. Immunol. 149:2255. Ce rdan C , Razana jaona D, Martin Y , Courcou l M, Pavon C , M a w a s C , Ol ive D, and Birg F, (1992b) Contr ibut ions of the C D 2 and C D 2 8 T lymphocyte activation pathways to the regulation of the express ion of the colony-stimulating factor ( C S F - 1 ) gene. J. Immunol. 149:373. Damle NK , K lussman K, and Aruffo A (1992a) Intercellular adhes ion molecule-2, a second counter-receptor for C D 1 1 a / C D 1 8 (leukocyte funct ion-associated antigen-1), 134 a provides a costimulatory s ignal for T-cel l receptor-initiated activation of human T cel ls. J. Immunol. 148:665. Damle NK, K lussman K, Linsley P S , Aruffo A , and Ledbetter J A (1992b) Differential regulatory effects of intercellular adhes ion molecule-1 on costimulat ion by the C D 2 8 counter-receptor B7 . J. Immunol. 149:2541. Damle NK, K lussman K, Linsley P S , and Aruffo A (1992c) Differential costimulatory effects of adhes ion molecules B7 , ICAM-1 , L F A - 3 , and V C A M - 1 on resting and ant igen-pr imed C D 4 + T lymphocytes. J. Immunol. 148:1985. de Fougero l les A R , Qin X , and Spr inger T A (1994) Character izat ion of the function of intercellular adhes ion molecule ( ICAM)-3 and compar ison with ICAM-1 and ICAM-2 in immune responses . J. Exp. Med. 179:619. Demb ic Z , H a a s W, W e i s s S , McCubrey J , Kiefer H, von Boehmer H, and Ste inmetz M (1986) Transfer of specificity by murine a and p T-cel l receptor genes . Nature 320:232. Dia lynas D, Loken M, Sarmiento M, and Fitch F W (1982) Identification of lysis-relevant molecu les on the surface of C T L : primary screening of monoclonal ant ibodies for the capaci ty to block cytolysis by cloned T cell l ines. Adv. Exp. Med. Biol. 146:547. Dong Y , Skoul tchi A l , and Pol lard J W (1993) Efficient D N A transfection of quiescent mammal ian cel ls using poly-L-ornithine. Nucleic Acids Res. 21 :771. Dubey C , Croft M, and Swa in S L (1995) Cost imulatory requirements of naive C D 4 + T cel ls. ICAM-1 or B7-1 can costimulate naive C D 4 T cell activation but both are required for opt imum response. J. Immunol. 155:45. Dustin M L and Spr inger T A (1991) Ro le of lymphocyte adhes ion receptors in transient interactions and cell locomotion. Annu. Rev. Immunol. 9:27. Farrar M A and Schre iber R D (1993) The molecular biology of interferon-y and its receptor. Annu. Rev. Immunol. 11:571 Germa in R N and Quil l H (1986) Unexpected express ion of a unique mixed- isotype c lass II M H C molecule by transfected L-cells. Nature 320:72. Gepper t T, Dav is L, Gur H, Walchol tz M, and Lipsky (1990) A c c e s s o r y cell s ignals involved in T-cel l activation. Immunol. Rev. 117:5. G immi C D , F reeman G J , Gr ibben J G , Sugi ta K, F reedman A S , Morimoto C , and Nadler L M (1991) B cell surface antigen B7 provides a costimulatory s ignal that 135 induces T cel ls to proliferate and secrete interleukin-2. Proc. Natl. Acad. Sci. USA 88:6575. Gu inan E C , Gr ibben J G , Boussiot is V A , F reeman G J , and Nadler L M (1994) Pivotal role of the B 7 : C D 2 8 pathways in transplantation tolerance and tumor immunity. Blood 84:3261. Harding F A , McArthur J G , G r o s s J A , Raulet D H , Al l ison J P (1992) CD28-med ia ted signal l ing co-st imulates murine T cel ls and prevents induction of anergy in T cell c lones. Nature 356:607. Hemler M (1990) V L A proteins in the integrin family: Structures, functions, and their roles on leukocytes. Annu. Rev. Immunol. 8:365. Ho lzmann B, Mclntyre B W , and W e i s s m a n IL (1989) Identification of a murine Peyer ' s patch-speci f ic lymphocyte homing receptor as an integrin molecule with an a lpha chain homologous to human V L A - 4 a lpha. Cell 56:37. Ho lzmann B and W e i s s m a n IL (1989) Peyer 's patch-speci f ic lymphocyte homing receptors consist of a VLA-4- l i ke a lpha chain assoc ia ted either of two integrin beta cha ins, one of which is novel. EMBO J. 8:1735. Horley K J , Carpeni to C , Baker B, and Takei F (1989) Molecular cloning of murine intercellular adhes ion molecule ( ICAM-1). EMBO J. 8:2889. Isobe M, Yag i ta H, Okumura K, and lhara A (1992) Spec i f ic accep tance of card iac allograft after treatment with antibodies to ICAM-1 and L F A - 1 . Science 255:1125. Isobe M and lhara A (1993) To lerance induction against card iac allograft by anti-ICAM-1 and LFA-1 treatment: T cel ls respond to in vitro al lostimulation. Transplant Proc. 25 :1079. J a n e w a y C A and Bottomly K (1994) S igna ls and s igns for lymphocyte responses . Cell 7_6:275. Jenk ins M K (1992) The role of cell division in the induction of clonal anergy. Immunol. Today 13:69. Jenk ins M K and Schwar tz R H (1987) Ant igen presentation by chemical ly modified sp lenocytes induces ant igen-specif ic T cell unrespons iveness in vitro and in vivo. J. Exp. Med. 165:302. Jenk ins M K , Taylor P S , Norton S D , and Urdahl K D (1991) C D 2 8 del ivers a costimulatory s ignal involved in ant igen-specif ic IL-2 production by human T cel ls. J. Immunol. 147:2461. 136 Jul ius M H , S impson E, and Herzenberg L A (1973) A rapid method for the isolation of functional thymus-der ived murine lymphocytes. Eur. J. Immunol. 3:645. K a r a s u y a m a H, Kudo A , and Melchers F (1990) The proteins encoded by the V p r e . B and y 5 pre-B cel l -specif ic genes can assoc ia te with each other and with u, heavy chain. J. Exp. Med. 172:969. Kishimoto T, Larson R, Corbi A , Dustin M, Staunton D, and Spr inger T (1991) The leukocyte integrins. Immunol. Today 46:149. L insley P S , Brady W , Grossmai re L, Aruffo A , Damle NK, and Ledbetter J A (1991) Binding of the B cell activation antigen B7 to C D 2 8 cost imulates T-cel l proliferation and IL-2 m R N A accumulat ion. J. Exp. Med. 173:721. L insley P S and Ledbetter J A (1993) The role of the C D 2 8 receptor during T cell responses to ant igen. Annu. Rev. Immunol. 11:191. Liu Y , J o n e s B, Aruffo A , Sul l ivan K M , Linsley P S , Janeway C A (1992) Heat-stable ant igen is a costimulatory molecule for C D 4 + T cell growth. J. Exp. Med. 175:437. Mar tz E (1987) LFA-1 and other accessory molecules functioning in adhes ion of T and B lymphocytes. Hum. Immunol. 10:3. Matsui K, Boni face J J , R e a y P A , Schi ld H, F a z e k a de St. Growth B, and Davis M M (1991) Low affinity interaction of pep t ide-MHC complexes with T cell receptors. Science 254:1788. O 'Conne l l K A and Edidin M (1990) A mouse lymphoid endothel ial cell line immortal ized by simian virus 40 binds lymphocytes and retains functional character ist ics of normal endothelial cel ls. J. Immunol. 144:521. Rothenberg E (1992) The development of functionally responsive T cel ls. Adv. Immunol. 51:85. Rubert i G , Sel l ins K S , Hill C M , Germain R N , Fathman C G , and Livingstone A (1992) Presentat ion of antigen by mixed isotype c lass II molecules in normal H -2 d mice. J. Exp. Med. 175:157. Sagers t rom C G , Kerr E M , Al l ison J P , and Davis M M (1993) Act ivat ion and differentiation requirements of primary T cel ls in vitro. Proc. Natl. Acad. Sci. USA 90:8987. 137 Sai to T, W e i s s A , Mil ler J , Norcross M A , and Germain R N (1987) Spec i f ic ant igen- la activation of transfected human cells express ing murine Ti ap -human T 3 receptor comp lexes . Nature 325:125. Sanford K K , Ear le W R , and Likely G D (1948) The growth in vitro of s ingle isolated t issue cel ls. J. Natl. Cancer Inst. 9:229. Schwar tz R H (1990) A cell culture model for T lymphocyte clonal anergy. Science 248:1349. S e m n a n i RT , Nutman T B , Hochman P, S h a w S , and van Seventer G A (1994) Cost imulat ion by purified intercellular adhes ion molecule 1 and lymphocyte function-assoc ia ted antigen 3 induces distinct proliferation, cytokine and cell sur face antigen profi les in human "naive" and "memory" C D 4 + T cel ls. J. Exp. Med. 180:2125. Spr inger T A (1990) Adhes ion receptors of the immune sys tem. Nature 346:425. Spr inger T, Gal f re G , Seche r D S , and Milstein C (1978) Monoc lona l xenogene ic ant ibodies to murine cell surface antigens: identification of novel leukocyte differentiation ant igens. Eur. J. Immunol. 8:539. St. Louis J D , Lederer J A , and Lichtman A H (1993) Cost imulatory deficient antigen presentat ion by an endothel ial cell line induces a nonproliferative T cell activation response without anergy. J. Exp. Med. 178:1597. Take i F (1985) Inhibition of mixed lymphocyte response by a rat monoclonal antibody to a novel murine lymphocyte activation antigen (MALA-2) . J. Immunol. 134:1403. Tan P, Anaset t i C , Hansen J A , Ledbetter J A , Linsley J A (1992) Induction of al loant igen hypo- respons iveness in human T lymphocytes by blocking interaction of C D 2 8 with its natural l igand B 7 / B B - 1 . J. Exp. Med. 177:165. Tivol E A , Schwei tzer A N , and Sharpe A H (1996) Cost imulat ion and autoimmunity. Curr. Opin. Immunol. 8:822. van Seven te r G A , Newman W, Sh imizu Y , Nutman T B , Tanaka Y , Horgan K J , Gopa l T V , Enn is E, O'Sul l ivan D, Grey H, and S h a w S (1991a) Ana lys is of T cell stimulation by superant igen plus major histocompatibil ity complex c lass II molecu les or by C D 3 monoc lona l antibody: costimulation by purified adhes ion l igands V C A M - 1 , I C A M - 1 , but not E L A M - 1 . J. Exp. Med. 174:901. van Seventer G A , Sh imizu Y , Horgan K J , Luce G E G , W e b b D, and S h a w S (1991b) Remote T cell co-stimulation via LFA-1 / ICAM-1 and C D 2 / L F A - 3 : demonstrat ion with 138 immobi l ized l igand/mAb and implication in monocyte-mediated co-st imulat ion. Eur. J. Immunol. 21 :1711. van Seven te r G A , Sh imizu Y , Horgan K J , and S h a w S (1990) The LFA-1 l igand I C A M -1 provides an important costimulatory signal for T cell receptor-mediated activation of resting T cel ls. J. Immunol. 144:4579. W a l u n a s TL , Lenschow D J , Bakker C Y , Linsley P S , F reeman G J , G r e e n J M , T h o m p s o n C B , and Bluestone J A (1994) C T L A - 4 can function as a negat ive regulator of T cell act ivation. Immunity 1:405. Wate rhouse P, Penn inger J M , T imms E, W a k e h a m A , Shah in ian A , Lee K P , T h o m p s o n C B , Gr iesser H, and Mak T W (1995) Lymphoprol i ferat ive d isorders with early lethality in mice deficient in Ctla-4. Science 270:985. W e a v e r C T and Unanue E R (1990) The costimulatory function of ant igen-present ing cel ls. Immunol. Today 11:49. W e b e r S , Traunecker A , Oliveri F, Gerhard W, and Karjalainen K (1992) Spec i f ic low-affinity recognit ion of major histocompatibil ity complex plus peptide by soluble T-cel l receptor. Nature 356:793. X u H, Bickford J K , Luther E, Carpeni to C , Take i F, and Spr inger T A (1996) Character izat ion of murine intercellular adhes ion molecule-2. J. Immunol. 156:4909. X u H, Tong IL, de Fougerol les A R , and Spr inger T A (1992) Isolation, character izat ion, and express ion of mouse ICAM-2 complementary and genomic D N A . J. Immunol. 149:2650. Y a g u e J , Whi te J , Co lec lough C , Kappler J , Pa lmer E, and Marrack P (1985) The T cell receptor: the a and p1 chains define idiotype, and antigen and M H C specif icity. Cell 42 :81 . 139 Chapter 4 Role of ICAM - 2 in leukocyte transendothelial migration 4:1 Introduction The immune system is compr ised of an integrated complex of functionally distinct cel ls and organs. Leukocytes move continuously throughout the body using the b loodstream and lymphatic vesse ls as pathways. They circulate throughout the bloodstream and cross capil lary networks into various lymphoid and non-lymphoid t issues (Picker and Butcher, 1992; Bev i lacqua, 1993; Car los and Har lan, 1994). Lymphocy tes re-enter the vasculature via the efferent lymphatic channe ls such as the thoracic and mesenter ic ducts. This ensures that the entire functional repertoire of the immune sys tem can effectively survey the host and maximize the immune response to the pathogen. Very few, if any, immune responses are initiated in the bloodstream. Therefore, the entry of leukocytes into lymph nodes is a key regulatory step in the normal function of the immune system. The molecular characterizat ion of leukocyte adhes ion to the endothel ium has made it c lear that the vascular lining plays an active role in p rocesses such as inf lammation and recirculation in the immune response. Leukocytes can be recruited to si tes of inflammation by chemoattractant gradients (Furie et al., 1991; Kavanaugh et al., 1991; Oppenhe imer -Marks et al., 1991). A n upregulation of adhes ion molecu les at the site of inflammation coinc ides with the enhanced transendothel ial migration of leukocytes (Kishimoto et al., 1990; Lusc inkas et al., 1991; Dobr ina et al., 1991). 140 Circulat ing lymphocytes are imported into secondary lymphoid organs such as lymph nodes and Peyer ' s patches through physiological ly distinct post-capi l lary venules, termed high endothel ial venules (HEV) . T h e s e venules are cal led H E V because they are lined with tall, cuboidal , and metabolical ly active endothel ial cel ls. Th is is in contrast to normal endothel ial cel ls which line the rest of the vasculature. The H E V in the secondary lymphoid t issues provide a port of exit for normal circulating lymphocytes to gain entry into the lymphatic sys tem. A sequence of adhes ive events which al lows leukocytes to migrate out of the b loodstream into the extravascular spaces has been identified (Harlan et al., 1992). V e n o u s flow transports the leukocytes throughout the vasculature resulting in random contact with the vesse l wal l . After the initial contact, some of the leukocytes appear to roll a long the endothel ial surface (Zimmerman et al., 1992). Th is s lows the passage of the leukocyte al lowing for subsequent adhes ive interactions. Fol lowing the rolling, the leukocytes adhere more efficiently to the endothel ium taking on a flattened and spread out morphology (Smyth et al., 1993). After a firm attachment, the leukocytes can then crawl over the endothel ial surface searching for an opening and s q u e e z e between endothel ia l cel ls (diapedesis) (Smith, 1992; S tosse l , 1993). O n c e in the subendothel ia l t issue, the leukocyte can migrate to an area of inflammation or to secondary or tertiary lymphoid t issues. Three famil ies of adhes ion receptors mediate the var ious s tages of the extravasat ion process (Bevi lacqua, 1993; Car los and Har lan, 1994). The select in family of proteins are involved in the initial adhes ion which results in rolling under 141 condit ions of venous flow (von Andr ian et al., 1991; Lawrence and Spr inger, 1991; A b a s s i et al., 1993; Ley et al., 1993; Lawrence and Springer, 1993). Adhes ion between the (32 integrins and Ig-like counter-receptors mediates the firm adhes ion (von Andr ian et al., 1991; Muller and Weig l , 1992; Butcher, 1991; Spr inger, 1994). The s a m e pair of adhes ion molecules and P E C A M - 1 also mediate the transendothel ial migration (Muller et al., 1993; Bird et al., 1993; Vaporc iyan et al., 1993). Migration through the subendothel ia l t issue is mediated by the integrins adhes ion with extracel lular matrix components (Hakkert et al., 1991; Kuijpers et al., 1993). The speci f ic molecu les involved in each step depends on many factors including activation state of both the leukocyte and the endothel ium, leukocyte type, and p resence of activators (Butcher, 1991; Springer, 1994). The role of var ious C A M s in transendothel ial migration has been examined both with ant ibody blocking studies as well as knockout mice. The LFA-1 : ICAM-1 interaction has been examined and found to play a critical role in migration (Furie et al., 1991; Kavanaugh et al., 1991; Oppenhe imer -Marks et al., 1991). Not only does it mediate neutrophil migration across an endothelial barrier, it a lso ass is ts lymphocytes in the p rocess . M ice deficient for ICAM-1 (Xu et al., 1994) or LFA-1 (Schmits et al., 1996) a lso display altered migratory patterns. Lymphocytes and neutrophils play important roles in normal recirculation as well as inf lammation. Both of these p rocesses utilize common molecules. However, the antibody against LFA-1 is able to block transendothel ial migration more effectively than the antibody against I C A M - 1 , suggest ing that ICAM-1 is not the only LFA-1 ligand utilized in the transendothel ial 142 migration p rocess (Furie et al., 1991; Kavanaugh et al., 1991; Oppenhe imer -Marks et al., 1991). A l though ant ibodies to ICAM-1 block lymphocyte binding to and migration ac ross non-lymphoid endothel ial cel ls, they have no effect on l ymphocy te :HEV interactions (May and Ager , 1992; Ager , 1994). Based on the constitutive endothel ial express ion of ICAM-2 (de Fougerol les et al., 1991), it is a possib le candidate as an alternate LFA-1 ligand in leukocyte migration across endothel ial barriers. However, the role of ICAM-2 in transendothel ial migration across lymphoid and non- lymphoid endothel ial cel ls has not been examined. The objective of the work presented in this chapter was to examine the potential role of ICAM-2 in transendothel ial migration of lymphocytes and neutrophils, two leukocyt ic cell types that play key roles in recirculation and inf lammation. The first step in examining the role was to set up a assay system that quantitates the migrated leukocytes. A n endothel ial cell line was grown on a porous membrane which provided an effective barrier such that the leukocytes above the endothel ial cel ls could not pass ive ly c ross the endothel ial cel ls. I C A M - 1 - and ICAM-2-t ransfected cel ls were also examined for their ability to mediate migration in this sys tem. 4:2 Materials a n d m e t h o d s 4:2.1 Animals C 3 H / H e mice used in this study were bred at the Joint An ima l Facil ity of the B . C . C a n c e r R e s e a r c h Centre from the founders purchased from J a c k s o n Laborator ies (Bar Harbor, M E ) . 143 4:2.2 Cell lines and antibodies The murine endothel ial cell line S V E C 4 . 1 0 , derived from S V 4 0 infection of peripheral lymph node stroma, was maintained in D M E M containing 5 % F C S and has been descr ibed previously (O'Connel l and Edidin, 1990). The S V E C 4 . 1 0 line has been shown to retain the morphology and functional characterist ics of normal endothel ial cel ls. The murine lymphoma cell line TIL1 was derived from a tumor initiated in C 3 H / H e mice by inoculation with IL-7-producing F s a - R f ibrosarcoma cel ls (McBr ide et al., 1992). The non-adherent tumor infiltrating lymphocytes were expanded and separated from the adherent fibroblastoid cel ls. TIL1 and S V E C 4 . 1 0 cel ls were a generous gift from Dr. G Dougherty (Terry Fox Laboratory, B .C . C a n c e r Resea rch Center) . TIL1 cel ls were maintained in D M E M containing 1 0 % F C S . Al l ant ibodies were used as purified Ig. The ant ibodies which recognize the murine cell sur face ant igens ICAM-1 ( l gG 2 a ) , ICAM-2 ( l gG 2 a ) , LFA-1 ( l gG 2 b ) , V L A - 4 ( l g G 2 b ) , and V C A M - 1 ( l gG 2 a ) have been descr ibed in the previous two chapters. The rat ant ibody R B 6 - 8 C 5 ( l gG 2 b ) which recognizes the murine granulocyte antigen Gr-1 has been descr ibed previously (Hestdal et al., 1991) and was purchased from Pharmingen . The rat ant ibodies which recognize mouse P E C A M - 1 (CD31) ( M E C 1 3 . 3 , l g G 2 a ) and mouse C D 2 5 (3C7, l g G 2 b ) were a lso purchased from Pharmingen . The hybr idoma cell l ines that produce ant i -mouse C D 4 4 (KM81 , A T C C TIB 241 , l g G 2 a ) (Miyake et al., 1990) and ant i -mouse Mac-1 (M1/70.15.11.5, A T C C TIB 128, l g G 2 b ) 144 (Springer et al., 1979) were obtained from American Type Culture Collection (Rockville, MD). 4:2.3 Expression of ICAM-1 and ICAM-2 on SVEC4.10 cells The murine ICAM-1 (Horley et al., 1989) and ICAM-2 cDNAs in the expression vector pBCMGSNeo (Karasuyama et al., 1990) were transfected into SVEC4.10 cells by the poly-L-ornithine method (Dong et al., 1993). Transfectants were selected in DMEM containing 5% FCS and G418 (500 ug/ml). Bulk transfectants that expressed high levels of ICAM-1 and ICAM-2 were isolated by panning directly with purified anti-ICAM-1 or anti-ICAM-2 mAb immobilized on petri dishes (Falcon 1001) as described in Chapter 3. The isolated cells were expanded and expression levels of various surface molecules were tested by flow cytometry as described in previous chapters. The secondary stain was GaRlgG-FITC (Cooper Biomedical) and dead cells were stained with 2 u,g/ml propidium iodide. 4:2.4 Immunoprecipitation The SVEC4.10 cells and their transfectants were grown for 2 days and the cells were harvested as described for flow cytometric analysis. The cells were then washed and labeled with biotin as described previously (von Boxberg et al., 1990). The plasma membranes were solubilized and incubated with either anti-ICAM-1 mAb or anti-ICAM-2 mAb coupled to Affi-Gel 10 beads (from chapter 2) for 4 hours at 4°C. The beads were then washed and the bound fraction washed eluted as described. The fraction 145 w a s then subjected to 1 0 % S D S - P A G E and transferred to an Immobi lon-P membrane (Millipore) which was probed with streptavidin-horseradish perox idase. The filter was then incubated in an enhanced chemi luminescence solution (Amersham) and exposed to X- ray film for 2-8 minutes. 4:2.5 Isolation of murine bone marrow neutrophils Bone marrow cel ls were obtained by flushing out mice tibia and femurs with H B S S ( C a + + / M g + + free) containing 0 .1% B S A (Hart et al., 1986). The cel ls were p a s s e d through a 27 gauge needle to obtain a single cell suspens ion before being combined with a hypotonic Tr is-ammonium chloride solution to lyse erythrocytes. The pelleted cel ls were washed with C a + + / M g + + free H B S S + 0 . 1 % B S A and then suspended in 3 ml of 4 5 % Percol l (density = 1.0575 g/ml; Pharmac ia) in C a + + / M g + + free H B S S . This was loaded onto a Percol l density gradient in a 15 ml polycarbonate tube. The gradient consis ted of 2 ml of 6 2 % (density = 1.0776 g/ml), 5 5 % (density = 1.0693 g/ml), and 5 0 % (density = 1.0634 g/ml) Percol l in C a + + / M g + + free H B S S layered success ive ly onto 3 ml of an 8 1 % Percol l cushion (density = 1.1002 g/ml). The gradient was then centrifuged at 1600 g for 30 min at 10°C. The cel ls banding between the 6 2 % and 8 1 % layers were harvested with a Pasteur pipette and washed twice with C a + + / M g + + free H B S S . Ce l l purity was a s s e s s e d by flow cytometry using the R B 6 - 8 C 5 antibody. 146 4:2.6 Adhesion and transmigration assays Adhes ion of leukocytes to S V E C 4 . 1 0 monolayers was quantitated as descr ibed before in the previous two chapters. TIL1 cel ls and neutrophils were labeled with Ca lce in A M and tested for their ability to adhere to S V E C 4 . 1 0 subconf luent monolayers express ing ICAM-1 or ICAM-2 . Blocking ant ibodies were added 15 min prior to the addit ion of the labeled leukocytes to the monolayer-containing wel ls. Transmigrat ion studies were performed using D M E M without phenol red ( D M E M p ) which can interfere with Ca lce in A M f luorescence. T h e s e a s s a y s were performed in 24 well plates (Falcon 3047, Becton Dickinson). E a c h well contained a transmigration insert (Falcon 3097, Becton Dickinson) with D M E M p " + 5 % F C S in both the upper and lower chambers . The transmigration insert consis ts of a plast ic circular shel l with an inert porous polycarbonate filter (8 u.m pores) at the bottom. S V E C 4 . 1 0 and their transfectants were trypsinized and 10 5 cel ls were added inside the insert. The filter membranes were precoated with fibronectin (100 u.l of 20 ucj/ml in P B S ; S i g m a Chemica l Co. ) for 1 hr before the endothelial cel ls were added . The S V E C 4 . 1 0 cel ls were able to grow as a monolayer on the porous membrane without falling into the lower chamber . After a 36-40 hr incubation at 37°C, fresh D M E M p " + 5 % F C S was added to both the top and lower chambers and Ca lce in A M labeled leukocytes (2 x 10 5 ) were gently added to the top chamber. The plates were then p lace in an incubator for var ious t imes. The inserts were then removed and the bottom s ide of the membrane w a s gently washed with D M E M p " + 5 % F C S . The extent of transmigration w a s determined by the level of yel low f luorescence in the bottom of the 24-wel l plate. 147 4:2.7 Examination of endothelial monolayer permeability The endothel ial monolayer grown on the transmigration inserts was examined for permeabil i ty. The S V E C 4 . 1 0 monolayer was grown on the inserts as descr ibed for transmigrat ion assays . The media was also replaced before the assay . In the top chamber , yel low f luorescently labeled beads (0.98 um F luorospheres, Molecular P robes , Eugene , O R ) were added in a vo lume of 300 uJ (0 .1% sol ids by volume). The a s s a y s were al lowed to proceed like transmigration a s s a y s and the degree of permeabil i ty was determined by the yel low f luorescence at the bottom of the wel l . 4:2.8 Simultaneous quantitation of yellow and red fluorescence In order to determine whether varying amounts of red and yel low f luorescent dye could be quantitated independently, different combinat ion of filters were employed. In a 96-well plate, Ca lce in AM- labe led bone marrow neutrophils (yellow) were added at varying concentrat ions (2.5 x 10 4 , 5 x 10 4 , 7.5 x 10 4 , 1 x 10 5) in 100 ul of D M E M p " + 5 % F C S . For each cell concentrat ion, varying amounts of red- labeled beads (1 u.m Fluoresbri te P C Red microspheres; Po lysc iences Inc., Warr ington, P A ) were also present (0.025%, 0.05%, 0.075%, 0 .10% sol ids/volume). The yel low f luorescence in the wel ls w a s determined using the B filter for both excitation (k= 485 ± 1 0 nm) and emiss ion (X= 530 ± 12.5 nm) readings. The red f luorescence w a s examined using the D filter for excitation (k= 590 ± 10 nm) and the E filter for emiss ion (k= 645 ± 20 nm). Al l f luorescent readings were taken with a CytoFluor 2300 F luorescent reader 148 (Millipore) and the software program Cy toCa lc (version 01.00.04). Neutrophil transmigrat ion was a s s e s s e d in the presence of red-labeled beads . Inserts containing the S V E C 4 . 1 0 monolayers was establ ished as before. However, 2 x 1 0 5 Ca lce in A M -labeled neutrophils were then combined with 0 .10% red-labeled beads and added to the migration inserts. At var ious time points, the inserts were removed and yel low f luorescence (transmigration) and red f luorescence (permeability) were a s s e s s e d for e a c h wel l . 4:3 R e s u l t s 4:3.1 Expression of ICAM-1 and ICAM-2 on SVEC4.10 cells The S V E C 4 . 1 0 endothel ial cell line does not express endogenous ICAM-1 or I C A M - 2 . Th is murine cell line was transfected with the ICAM-1 or ICAM-2 c D N A s and panned with the appropriate antibody. F low cytometric analys is revealed that the levels of express ion of the two murine I C A M s were equivalent (Figure 23). The adhes ion molecule V C A M - 1 was also not detected on the surface of S V E C 4 . 1 0 cel ls. In Figure 24, the immunoprecipitat ion reveals that the ICAM-1 and ICAM-2 molecu les are of the appropriate s ize , 90-95 kD and 50-55 kD respectively, indicating that there w a s no abnormal glycosylat ion patterns in the S V E C 4 . 1 0 cell line. 4:3.2 Time course of diffusion The endothel ial cel ls were grown to conf luency on the membrane inserts. A l though the presence of medium ( D M E M p " + 5 % F C S ) in the insert made it difficult to 149 n o l ° A b ICAM-1 A b ICAM-2 A b VCAM-1 A b PECAM-1 A b C D 4 4 A b SVEC4.10 SVEC4.10: ICAM-1 SVEC4.10: ICAM-2 IW UB iW lino ""W Iffi W "to ' in Itt' TO IW IW *T Figure 23 F /cw cytometric analysis of SVEC4.10 cells. The endothelial cell line S V E C 4 . 1 0 was transfected with the ICAM-1 ( S V E C : I C A M - 1 ) or I C A M - 2 ( S V E C : I C A M - 2 ) and subjected to flow cytometry using the indicated ant ibodies and F(ab') 2 goat anti-rat IgG-FITC as the secondary stain. I C A M - 1 , ICAM-2 , V C A M - 1 , P E C A M - 1 , and C D 4 4 express ion are shown. I C A M - 1 A b 0 1 2 I C A M - 2 A b 0 1 2 -214 -111 74 -46 •30 Figure 24 Immunoprecipitation of ICAM-1 and ICAM-2 from transfected SVEC4.10 cells. Untransfected and ICAM-transfected S V E C 4 . 1 0 cel ls were sur face biotinylated and ICAM-1 and ICAM-2 were immunoprecipitated using the appropriate antibody conjugated to beads . The samp les were subjected to S D S - 1 0 % P A G E ( S V E C 4 . 1 0 in lane 0, S V E C 4 . 1 0 : I C A M - 1 in lane 1, SVEC4.10 : ICAM - 2 in lane 2) and detected by Western blotting using peroxidase-conjugated streptavidin. 151 v isual ize the conf luency of the monolayer, the permeabil i ty could still be examined using f luorescent ly- labeled beads. The S V E C 4 . 1 0 monolayers appear to be intact as demonstrated by the time course in Figure 25. The beads diffuse efficiently ac ross the membrane in the insert when S V E C 4 . 1 0 cells are not present. However , when the endothel ia l cel ls are present, the beads do not effectively c ross the monolayer. The S V E C 4 . 1 0 cel ls form tight junctions with adjacent cel ls that do not al low the pass ive diffusion of 0.98 u,m beads. 4:3.3 Transmigration of TIL1 cells across an endothelial monolayer A preliminary test to examine whether the set up would actually support the transmigration of TIL1 cel ls was carried out. Ca lce in A M labeled TIL1 cel ls were added into the inserts containing S V E C 4 . 1 0 monolayers. After var ious time intervals, the inserts were removed from the wel ls and the migrated cel ls were quantitated by the measurement of yel low f luorescence. TIL1 cel ls diffuse across the membrane filter readily (Figure 26). However, untransfected S V E C 4 . 1 0 cel ls ( ICAM-1", ICAM-2") were able to form an impermeable barrier across the porous membrane in which few TIL1 cel ls (11% of total cells) were detected in the lower chamber after 16 hours of transmigrat ion. I C A M - 1 - and ICAM-2-transfected S V E C 4 . 1 0 cel ls were able to promote migration of TIL1 cel ls across the endothel ial monolayer. In both c a s e s , approximately 3 8 % of total TIL1 cel ls had migrated after 16 hours. There is an apparent dif ference in the kinetics of ICAM-1-media ted and ICAM-2-med ia ted migration. The migration of TIL1 cel ls appears faster through S V E C 4 . 1 0 cel ls 152 Time (hr) F i g u r e 25 Kinetics of bead diffusion across SVEC monolayer. F luorescent ly labeled beads (0.98 urn diameter) were added in the top chamber of migration inserts with and without the var ious S V E C 4 . 1 0 transfectant monolayers grown on the porous membrane . The inserts were removed at var ious t imes to a s s e s s the diffusion of beads in the bottom chamber. The results from each time point are exp ressed as a mean of dupl icate wel ls + S E M . 153 0 $ , , , 0 5 10 15 20 Time (hr) F i g u r e 26 Ability of TIL 1 cells to utilize ICAM-1 and ICAM-2 for migration across the SVEC4.10 monolayer. TIL1 cel ls labeled with calcein A M were added to the top chamber of the migration insert. The migration insert w a s in p lace until the indicated t imes afterwhich the insert was removed and migration w a s a s s e s s e d by the amount of f luoresecence in the bottom chamber. Al l points are exp ressed as a mean of dupl icate wel ls ± S E M . 154 express ing ICAM-1 than ICAM-2 in the initial 8 hours. However, over longer periods of t ime (16 hours), the total number of migrated TIL1 cel ls is equal . This sys tem not only demonst ra tes that it is adequate in quantitating transendothel ial migration, it a lso demonstrates that ICAM-1 and ICAM-2 are able to mediate the migration p rocess in an LFA-1 -dependen t manner (Figure 27). 4:3.4 Isolation of neutrophils B o n e marrow neutrophils isolated by density centrifugation were a s s e s s e d for purity using the R B 6 - 8 C 5 antibody. F low cytometry revealed that the isolated populat ion w a s ~ 9 0 % pure (Figure 28). The population a lso d isp layed the identical posit ive staining pattern for ICAM-2 , L F A - 1 , and Mac -1 . Isotype controls for l g G 2 a (CD31)and l g G 2 b (CD25) were negative. 4:3.5 Neutrophil binding to SVEC4.10 cells Isolated neutrophils were labeled with Ca lce in A M and tested for their ability to adhere to the S V E C 4 . 1 0 cel ls and their transfectants. A s seen in Figure 29, the neutrophils adhered to the untransfected S V E C 4 . 1 0 cel ls (47% of total input cells). However , ICAM-1 and ICAM-2 express ion on the endothel ial cell line increased binding to 8 4 % and 7 6 % respectively. This increased adhes ion was LFA-1-spec i f i c as demonstrated by the an t i -CD11a antibody which inhibited binding to background levels. The control antibody against C D 4 9 d has no effect on adhes ion . 155 40 • SVEC4.10 migration + CD11 a Ab + CD49d Ab F i g u r e 27 SVEC4.10 cells expressing ICAM-1 or ICAM-2 are able to mediate TIL1 transendothelial migration in an LFA-1-dependent manner. TIL1 cel ls labeled with ca lce in A M were examined for their ability to migrate across I C A M - 1 - and I C A M - -transfected S V E C 4 . 1 0 monolayers. Ant ibodies were added to TIL1 cel ls 20 min prior to addit ion of cel ls to the monolayer. The migration was al lowed to p roceed for 6 hrs. T h e results are expressed as a mean of duplicate wells ± S E M . 156 LOG FLUORESCENCE INTENSITY F i g u r e 28 Expression of adhesion molecules on mouse bone marrow neutrophils. M o u s e neutrophils isolated by differential centrifugation from bone marrow were ana lyzed by flow cytometry. Neutrophils (2.5 x 105) were stained with the indicated primary antibody (750 ng in 100 (il final volume). The samp les were then counter-sta ined with F(ab') 2 goat anti-rat IgG-FITC. The ordinate represents the relative cell number and the a b s c i s s a represents the log f luorescence intensity. 157 100 • SVEC4.10 • SVEC4.10:ICAM-1 binding +CD11aAb + CD49d Ab F i g u r e 29 Neutrophil binding to SVEC4.10 monolayers. Ca lce in A M labeled neutrophi ls (1 x 10 5) were added to wells with untransfected S V E C 4 . 1 0 monolayers and I C A M - 1 - and ICAM-2-transfected monolayers. The neutrophils were a l lowed to bind for 8 min and washed as descr ibed previously. Blocking ani tbodies were added (4 |Lig/ml) 15 min prior to addition of the neutrophils to the wel ls. Resu l ts are presented a s a mean of triplicate wel ls + S E M . 158 4:3.6 Neutrophil transendothelial migration across SVEC4.10 monolayers The express ion of ICAM-1 and ICAM-2 greatly ass is ted the transendothel ial migration of neutrophils ac ross S V E C 4 . 1 0 cells (Figure 30). Express ion of ICAM-1 and I C A M - 2 al lows an increase in neutrophil migration (both 4 6 % of total input cel ls after 8 hours) compared with untransfected S V E C 4 . 1 0 cel ls (15%). Transmigrat ion occurs through untransfected S V E C 4 . 1 0 cel ls at a slower rate than the transfected cel ls. The migration profile of ICAM-1-media ted migration is very similar to that of I C A M - 2 in both rate and absolute migration at each time point. In Figure 31 , the migration p rocess is shown to be partially LFA-1-dependent with the an t i -CD11a antibody inhibition bringing the level of migration to almost background levels (18%). The control ant ibody against C D 4 9 d had no effect. 4:3.7 Simultaneous quantitation of migration and permeability It is possib le that some of the migrated neutrophils are leaking through holes opened up in the endothel ial monolayer by previously migrated neutrophils. To examine this possibil ity, Ca lce in AM- labe led neutrophils combined with red f luorescent beads were added onto the endothel ial monolayer bearing membrane. The first step in such an examinat ion is to ensure that the filter setup will not al low the yel low f luorescence of the neutrophils to interfere with the red f luorescence of the beads . Us ing filter D for excitation (X= 590 ± 10 nm) and filter E for emiss ion (X= 645 ± 20 nm), it w a s possib le to measure red f luorescence without any interference from the yel low cel ls ( T A B L E 2). Using filter B for both excitation (X= 485 ± 10 nm) and 159 100 ^ 75-no monolayer SVEC4.10 SVEC4.10:ICAM-1 SVEC4.10.-ICAM-2 co E 50 25 T " 4 X / Time (hr) 8 10 12 F i g u r e 30 Neutrophil migration across SVEC4.10 monolayers. Migration of calcein A M labeled neutrophils ac ross S V E C 4 . 1 0 and ICAM-transfected monolayers was a s s e s s e d as descr ibed for TIL1 cel ls. The migration at each time point is expressed as a mean of dupl icate wel ls + S E M . 160 c o • M B +-* C O rj) E 3 C l ) c migration + CD11 a Ab + CD49d Ab F i g u r e 31 Antibody blocking of neutrophil migration across SVEC4.10 monolayers. Ca lce in A M labeled neutrophils were incubated with ant ibodies (4 ug/ml) in H B S S + 5 % F C S for 20 min prior to addition to the migration inserts. The a s s a y w a s al lowed to p roceed for 6 hrs and the results are expressed as a mean of dupl icate wel ls + S E M . 161 TABLE 2 Red bead fluorescence does not interfere with yellow cell fluorescence. Ca lce in A M labeled neutrophils were combined with varying concentrat ions of red f luorescently labeled beads (0.98 pirn diameter) in a vo lume of 100 pJ of H B S S + 5 % F C S . T h e s e combinat ions were p laced in 96 well plates and the f luorescence readings are expressed as a mean of triplicate wel ls ± S E M . For each combinat ion, the yellow f luorescence reading (determined with Filter B for both excitation and emiss ion) is the top number in the table below and the red f luorescence reading (determined with Filter D for excitation and Filter E for emission) is the bottom italicized number in parentheses. no beads 0.025% beads 0.05% beads 0.075% beads 0.10% beads no cells 30 + 2 (45 ± 1) 38 + 2 (109 ± 1) 41 + 1 (185 ±2) 52 + 4 (249 ±4) 60 + 2 (332 + 14) 2.5 x 104 cells 854 + 63 (45 ± 1) 876 + 50 (109 + 2) 893 + 42 (181 + 4) 887 + 84 (248 ± 8) 881 + 15 (345 + 8) 5.0 x 104 cells 1552 + 72 (45 ± 1) 1523 + 58 (109 ±2) 1509 + 27 (176 ±5) 1478 + 65 (262 ±3) 1509 + 11 (328 + 5) 7.5 x 104 cells 2409 + 39 (45 ± 1) 2367 + 34 (108 ±3) 2287 + 93 (182 + 3) 2374 + 41 (264 ±5) 2307 + 85 (339 ±8) 1.0 x 105 cells 2911+16 (45 ± 1) 2816 + 59 (107 ±2) 2743 + 62 (179 + 3) 2792 + 113 (258 ±6) 2744 + 96 (336 ± 4) emiss ion (A= 530 + 12.5 nm), the yel low f luorescence could be measured without interference from the red beads. Using this combinat ion of filters, it is poss ib le to examine migration and permeabil i ty simultaneously. The time course of neutrophil migration in the p resence of red beads (Figure 32a) is identical to that without beads (Figure 30). The beads , in the presence of migrating neutrophils, do not appear to move through the membrane (Figure 32b). This suggests that previously migrated neutrophils do not leave gaping holes in the endothel ial monolayer and subsequent migration is not due to pass ive diffusion through the exposed membrane pores. 4 : 4 D i s c u s s i o n The importance of the LFA-1 glycoprotein in the adhes ion and extravasat ion of leukocytes across endothelial cel ls has been documented in many sys tems (Bev i lacqua, 1993; Car los and Har lan, 1994). Var ious groups have demonstrated that LFA-1 on leukocytes interacts with ICAM-1 on endothel ial cel ls in the process of t ransendothel ia l migration (Furie etal., 1991; Kavanaugh et al., 1991; Oppenhe imer -Marks et al., 1991). In certain studies, the role of the LFA-1 : ICAM-1 interaction has been attributed to the actual migration across the endothel ial barrier. However, the ICAM-1 component , as determined by antibody inhibition, can be smal ler than the LFA-1 component . This has been attributed to the utilization of an alternate LFA-1 counter-receptor present on endothelial cel ls. ICAM-2 , present on endothel ial cel ls, may mediate the LFA-1 adhes ion. However, the precise role of ICAM-2 in leukocyte 163 OA 1 1 1 1 1 1 2 3 4 5 6 7 8 9 B) 80-T . Time (hr) F i g u r e 32 Simultaneous assessment of neutrophil migration and SVEC4.10 monolayer permeability. Ca lce in A M labeled neutrophils were combined with red f luorescently labeled beads (0.98 |im) and added to S V E C 4 . 1 0 monolayers . Both migration of neutrophils and diffusion of beads across the monolayers w a s a s s e s s e d as descr ibed in the Materials and Methods. A) The time course of neutrophil migration ac ross the monolayer is shown. B) The time course of bead diffusion ac ross the monolayer is shown. 164 transmigration is still unclear. The purpose of this work was to examine the possibil ity of whether ICAM-2 on endothel ial cel ls could assist leukocyte transmigrat ion. The work done in this chapter revolved around establ ishing an in vitro sys tem to determine if ICAM-2 express ion on endothelial cel ls would al low leukocyte transmigrat ion. Initial concerns about the permeabil i ty of the endothel ial monolayer were d ismissed when it was shown that the f luorescently- labeled beads could not to diffuse ac ross the monolayer. Endothel ial cel ls transfected with the ICAM-2 c D N A were able to permit migration of both a T cell line (TIL1) and bone marrow neutrophils. The increased migration in the ICAM-transfected endothel ial cel ls w a s L F A - 1 -dependent as demonstrated by the an t i -CD11a antibody. T h e s e exper iments were prel iminary and the complete specificity was not demonstrated with the ant i - ICAM-1 and ant i - ICAM-2 antibody inhibition. A s well , the inhibition by the an t i -CD11b and anti-C D 1 8 ant ibodies w a s not examined. The migration of leukocytes across the endothel ial barrier is a multistep p rocess which p roceeds by the success ive formation of adhes ive bonds between receptors on leukocytes and counter-receptors on endothelial cel ls. Interference with any of the adhes ive interactions would inhibit all subsequent interactions. The migration a s s a y in this chapter is a static model sys tem in which no flow was appl ied to mimic veneous flow. Therefore, rolling of leukocytes, mediated by select ins, is unlikely to play a role. It is difficult to determine from these results whether ICAM-2 on endothel ial cel ls al lows more efficient leukocyte adhes ion or is directly involved in their transmigrat ion. However , previous studies have shown that LFA-1 : ICAM-1 interaction play a critical 165 role in the actual migration process (Kavanaugh et al., 1991; Oppenhe imer -Marks et al., 1991). T cel ls are able to bind endothelial cel ls in an ICAM-1- independent manner. T h e s e cel ls remain bound with the addition of blocking ant ibodies but are not able to migrate as they would if the antibodies were not added. The exper iments in this project do not examine whether the increased transmigration by ICAM-2 express ion is a direct result of the migration step or simply due to increased adhes ion . Other molecu les have also been examined for the exact role they play in the migration process . P E C A M - 1 (CD31) has been shown to be involved in monocyte migration (Muller et al., 1993). Addit ion of ant i -CD31 antibody does not alter monocyte binding to the endothel ial cel ls but does arrest the monocyte motility at the junction between adjacent endothel ial cel ls. It is very likely that ICAM-2 express ion on the endothel ial cel ls increases the binding capacity. If ICAM-2 , like I C A M - 1 , can mediate the migration process , addition of the ant i - ICAM-2 antibody after the initial binding has occurred should inhibit further migration. Leukocy tes adherent on endothelial cel ls appear to crawl over the luminal sur face search ing for intercellular junctions to squeeze between (Bev i lacqua, 1993; Car los and Har lan, 1994). The leukocytes move across the endothel ial cel ls to the abluminal sur face and on to the extravascular t issue. The d iapedes is involves intercellular junction d isassembly and may leave gaps in the endothel ium. T h e s e gaps may facilitate subsequent d iapedes is unless quickly repaired. A s demonstrated by the prel iminary exper iments in the chapter, neutrophil migration ac ross the S V E C 4 . 1 0 monolayer is not accompan ied by pass ive bead (red f luorescent) diffusion. Al though 166 this does not el iminate the possibility that holes are still present in the endothel ial cell monolayer, they are nonetheless not big enough to allow 1 u.m beads to pass through. S ince neutrophils are 10-20 u.m in diameter, they are much bigger than any gap possib ly left in the monolayer and their migration must be an active p rocess involving neutrophil shape distortion. The LFA-1 : ICAM interactions are not the only molecular pair ings involved in leukocyte migration. In examining the leukocyte-endothel ial cell adhes ions , it has become apparent that other adhes ion pathways play prominent roles in leukocyte migration. In fact, these pathways are all super imposed upon each other and many factors dictate which may be more significant than the others in var ious situations. The activation state of the endothel ium is one such factor. Under inflammatory condit ions, ICAM-1 is upregulated on endothel ial cel ls and leukocytes have a stronger preference of ICAM-1 over ICAM-2 under such condit ions (Dustin et al., 1986; Dustin et al., 1988; Dustin and Spr inger, 1991). Other molecules such as V C A M - 1 are a lso upregulated in inf lammation and may alter the adhes ive capacity of the endothel ium (Rice and Bev i lacqua , 1989; R ice et al., 1990; Car los et al., 1990). The combinat ion of different molecular pair ings between leukocytes and endothel ium can determine whether leukocytes migrate out of the bloodstream in one t issue or another. Under inflammatory condit ions, neutrophils are the first leukocytes present, whereas lymphocytes and monocytes make up the majority of migrating leukocytes at later time points (Munro et al., 1989; Leung et al., 1991; Norris et al., 1991; Bev i lacqua , 1993; Ager , 1994). Mo lecu les involved in this process, as determined by ant ibody inhibition, 167 include L F A - 1 , M a c - 1 , and ICAM-1 . Under normal recirculation, neutrophils do not migrate out of the bloodstream, but lymphocytes and monocytes exhibit spon taneous migration ac ross endothel ium. Molecu les involved in this as determined by genetical ly deficient mice include the select in family members (Mayadas et al., 1993; Labow et al., 1994; Tedder et al., 1995). ICAM-2 has not been examined in normal recirculation. However , this study indicates that presence of ICAM-2 on endothel ial monolayers does promote lymphocyte (TIL1) migration in an LFA-1-dependent manner. A l though the ant i - ICAM-2 antibody inhibition has not yet been done, it does appear that ICAM-2 may play a role in normal recirculation. The filter setup used to dist inguish between red and yel low f luorescence can a lso be used to determine which cell type may preferentially utilize which molecule for migration. In a heterogeneous cell populat ion, similar to the b loodst ream, different cel ls may be labeled with red and yel low f luorescent dyes and tested for s imul taneous migration. The ability of ant ibodies against adhes ion molecules on both leukocytes and endothel ia l cel ls to block migration provides an attractive opportunity to treat var ious d i sease states. Ant ibodies against L F A - 1 , M a c - 1 , and ICAM-1 have been shown to inhibit a ser ies of inflammatory responses (Carlos and Har lan, 1994). W h e n neutrophils are the major cause of t issue damage, these ant ibodies in combinat ion or somet imes a lone can decrease the extent of damage . In the c a s e of ischemia/reperfus ion, these antibodies have been shown to at least partially prevent the d a m a g e after a card iac or renal transplant (Cosimi et al., 1990; Har lan et al., 1992; Isobe et al., 1992). Vascu la r occ lus ion initiates the damage to the endothel ium which 168 is further magnif ied after reperfusion by activation of the inflammatory sys tem and adhes ion of neutrophils. The ant ibodies are able to block the neutrophil adhes ion and thus inhibit the subsequent damage. The ant ibodies are a lso able to block the immune response generated by an organ allograft. One site of attack by the immune sys tem is the endothel ium in the allograft. S ince endothelial cel ls express ICAM-2 , they may be a target for neutrophil attachment. A s wel l , endothelial cel ls have been shown to generate a T cell response (St. Louis et al., 1993) which may be ICAM-2-dependent . Th is provides a goal for immune system modulation in which the ant i - ICAM-2 antibody may inhibit both neutrophil adhes ion and the T cell response to endothel ium. 169 4:5 References A b a s s i O , Kishimoto T K , Mclntire LV, Anderson D C , and Smith C W (1993) E-select in supports neutrophil rolling under condit ions of flow. J. Clin. Invest. 92:2719. A g e r A (1994) Lymphocyte recirculation and homing: roles of adhes ion molecu les and chemoattractants. Trends Cell Biol. 4 :326. Bev i lacqua M P (1993) Endothel ial- leukocyte adhes ion molecules. Annu. Rev. Immunol. 11:767. Bird IN, Spragg J H , A g e r A , and Mathews N (1993) Studies of lymphocyte transendothel ia l migration: analys is of migrated cell phenotypes with regard to C D 3 1 ( P E C A M - 1 ) , C D 4 5 R A and C D 4 5 R O . Immunol. 80:553. Butcher E C (1991) Leukocyte-endothel ia l cell recognit ion: three (or more) s teps to specif icity and diversity. Cell 67 :1033. Car los T M and Harlan J M (1994) Leukocyte-endothel ia l adhes ion molecu les. Blood 84:2068. Car los T M , Schwar tz B R , Kovach NL, Y e e E, R o s s o M, Osborn L, C h i - R o s s o G , N e w m a n B, Lobb R, and Harlan J M (1990) Vascu la r cell adhes ion molecule-1 lymphocyte adherence to cytokine-activated cultured human endothel ial cel ls. Blood 76:965. Cos im i A B , Cont i D, Delmonico FL , Preffer F l , W e e S L , Rothlein R, F a a n e s R, and Colv in R B (1990) In vivo effects of monoclonal antibody to ICAM-1 (CD54) in nonhuman primates with renal allografts. J. Immunol. 144:4604. de Fougero l les A R , Stacker S A , Schwart ing R, and Spr inger T A (1991) Character izat ion of ICAM-2 and ev idence for a third counter-receptor for L F A - 1 . J. Exp. Med. 174:253. Dobr ina A , M e n e g a z z i R, Car los T N , Nardon E, Cramer R, Zacch i T, Har lan J M , and Patr iarca P (1991) Mechan ism of eosinophi l adherence to cultured vascu lar endothel ia l cel ls: eosinophi ls bind to the cytokine- induced endothel ial l igand vascu lar cell adhes ion molecule-1 via the very late antigen-4 integrin receptor. J. Clin. Invest. 88:20. Dong Y , Skoul tchi A l , and Pol lard J W (1993) Efficient D N A transfection of quiescent mammal ian cel ls using poly-L-ornithine. Nucleic Acids Res. 21 :771. 170 Dustin M L , Rothlein R, Bahn A K , Dinarello C A , and Spr inger T A (1986) Induction by IL-1 and interferon, t issue distribution, biochemistry, and function of a natural adherence molecule ( ICAM-1). J. Immunol. 137:245. Dustin M L and Spr inger T A (1988) Lymphocyte function assoc ia ted antigen-1 (LFA-1) interaction with intercellular adhes ion molecule-1 ( ICAM-1) is one of at least three mechan i sms for lymphocyte adhes ion to cultured endothel ial cel ls. J. Cell Biol. 107:321. Dustin M L and Spr inger T A (1991) Ro le of lymphocyte adhes ion receptors in transient interactions and cell locomotion. Annu. Rev. Immunol. 9:27. Fur ie M B , Tanc inco M C A , and Smith C W (1991) Monoc lona l ant ibodies to leukocyte integrins C D 1 1 a / C D 1 8 and C D 1 1 b / C D 1 8 or intercellular adhes ion molecule-1 inhibit chemoattractant-st imulated neutrophil transendothel ial migration in vitro. Blood Z8:2089 . Hakkert B C , Kuijpers T W , Leeuwenberg J F M , van Mourik J A , and R o o s D (1991) Neutrophi l and monocyte adherence and migration across monolayers of cytokine-activated endothel ial cel ls: the contribution of C D 1 8 , E L A M - 1 , and V L A - 4 . Blood 78:2721. Har lan J M , Winn R K , Vedde r N B , Doerschuk C M , and R ice C L (1992) In vivo models of leukocyte adherence to endothel ium, in Harlan J M and Liu D Y (eds), Adhes ion : its role in inflammatory d isease , New York, NY , F reeman. Hart P H , S p e n c e r LK, Nulsen M F , McDona ld P J , and F in lay -Jones J J (1986) Neutrophil activity in abscess-bear ing mice: comparat ive studies with neutrophils isolated from peripheral b lood, elicited peritoneal exudates, and a b s c e s s e s . Infect. Immun. 51:936. Hestda l K, Ruscett i F W , Ihle J N , J a c o b s e n S E W , Dubois C M , Kopp W C , Longo DL, and Kel ler J R (1991) Character izat ion and regulation of R B 6 - 8 C 5 antigen express ion on murine bone marrow cel ls. J. Immunol. 147:22. Ho lzmann B, Mclntyre B W , and W e i s s m a n IL (1989) Identification of a murine Peyer ' s patch-speci f ic lymphocyte homing receptor as an integrin molecule with an a lpha chain homologous to human V L A - 4 a lpha. Cell 56:37. Ho lzmann B and W e i s s m a n IL (1989) Peyer 's patch-speci f ic lymphocyte homing receptors consist of a VLA-4- l i ke a lpha chain assoc ia ted either of two integrin beta cha ins, one of which is novel. EMBO J. 8:1735. Horley K J , Carpeni to C , Baker B, and Take i F (1989) Molecular cloning of murine intercellular adhes ion molecule ( ICAM-1). EMBO J. 8:2889. 171 Isobe M, Yag i ta H, Okumura K, and lhara A (1992) Speci f ic accep tance of card iac allograft after treatment with ant ibodies to ICAM-1 and L F A - 1 . Science 255 :1125. K a r a s u y a m a H, Kudo A , and Melchers F (1990) The proteins encoded by the V p r e . B and y 5 pre-B cel l -speci f ic genes can assoc ia te with each other and with u. heavy chain. J. Exp. Med. 172:969. Kavanaugh A F , Lightfoot E, Lipsky P E , and Oppenhe imer -Marks N (1991) Ro le of C D 1 1 / C D 1 8 in adhes ion and transendothel ial migration of T cel ls: analys is using CD18-def ic ient T cell c lones. J. Immunol. 146:4149. Kishimoto T K , Larson R S , Corbi A L , Dustin ML , Staunton D E , and Spr inger T A (1990) Leukocyte integrins. Spr inger et al. (eds): Leukocyte adhes ion molecu les. New York, Spr inger -Ver lag, Chapter 1. Kui jpers T W , Mul E P , B lom M, Kovach NL, G a e t a F C A , Tol lefson V , E l ices M J , and Har lan J M (1993) Freez ing adhes ion molecules in a state of high-avidity binding b locks eosinophi l migration. J. Exp. Med. 178:279. Labow M A , Norton C R , Rumberger J M , Lombard-Gi l looly K M , Shuster D J , Hubbard J , Bertko R, Knaack P A , Terry R W , Har ison ML , et al. (1994) Character izat ion of E-select in-def icient mice: demonstrat ion of overlapping function of the endothel ial se lect ins. Immunity 1:709. Lawrence M B and Spr inger T A (1991) Leukocytes flow on a select in at physiological f low rates: distinction from and prerequisite for adhes ion through integrins. Cell 65:859. Lawrence M B and Spr inger T A (1993) Neutrophils roll on E-select in. J. Immunol. 151:6338. Leung D Y , Pobe r J S , and Cotran R (1991) Express ion of endothel ial- leukocyte adhes ion molecule-1 in elicited late phase allergic reactions. J. Clin. Invest. 87:1805. Ley K, Tedder T F , and K a n s a s G S (1993) L-selectin can mediate leukocyte rolling in untreated mesenter ic venules in vivo independent of E- and P-select in. Blood 82:1632. Lusc inskas F W , Cybu lsky M l , Kiely J M , Peck ins C S , Davis V M , and Gimbrone M A (1991) Cytokine-act ivated human endothelial monolayers support enhanced neutrophil transmigration via a mechan ism involving both endothel ia l- leukocyte adhes ion molecule-1 and intercellular adhes ion molecule-1. J. Immunol. 146:1617 172 M a y M J and A g e r A (1992) ICAM-1- independent lymphocyte transmigration across high endothel ium: differential up-regulation by interferon gamma. Eur. J. Immunol. 22:219. M a y a d a s T N , Johnson R C , Rayburn H, Hynes R O , and W a g n e r DD (1993) Leukocyte rolling and extravasat ion are severely compromised in P selectin-deficient mice. Cell 74 :541 . McBr ide W h , Thacker J D , C o m o r a S , Economu J S , Kel ley D, Hogge D, Dubinett S M , and Dougherty G J (1992) Genet ic modification of a murine f ibrosarcoma to produce interleukin 7 st imulates host cell infiltration and tumor immunity. Cancer Res. 52 :3931. Miyake K, Med ina KL , Hayash i S . , O n o S , Hamaoka T, and K incade P W (1990) Monoc lona l ant ibodies to P g p - 1 / C D 4 4 block lympho-hemopoies is in long-term bone marrow cultures. J. Exp. Med. 171:477. Mul ler W A and Weig l S A (1992) Monocyte-select ive transendothel ial migration: d issect ion of the binding and transmigration phases by an in vitro assay . J. Exp. Med. 176:819. Mul ler W A , Weig l S A , Deng X , and Phil l ips D M (1993) P E C A M - 1 is required for transendothel ia l migration of leukocytes. J. Exp. Med. 178:449. Munro J M , Pober J S , and Cotran R S (1989) Tumor necrosis factor and interferon-g a m m a induce distinct patterns of endothel ial activation and leukocyte accumulat ion in skin of Papio anubis. Am. J. Pathol. 135:121. Norris P, Pos ton R N , Thomas D S , Thornhil l M, Hawk J , and Haskard D O (1991) The express ion of endothel ial leukocyte adhes ion molecule-1 (ELAM-1 ) , intercellular adhes ion molecule-1 ( ICAM-1), vascular cell adhes ion molecule-1 ( V C A M - 1 ) in exper imental cutaneous inflammation: a compar ison of ultraviolet B erythema and de layed hypersensit ivity. J. Invest. Dermatol. 96:763. O 'Conne l l K A and Edidin M (1990) A mouse lymphoid endothel ial cell line immortal ized by simian virus 40 binds lymphocytes and retains functional character ist ics of normal endothel ial cel ls. J. Immunol. 144:521. Oppenhe imer -Marks N, Davis L S , Bogue DT, Ramberg J , and Lipsky P E (1991) Differential utilization of ICAM-1 and V C A M - 1 during the adhes ion and transendothel ial migration of human T lymphocytes. J. Immunol. 147:2913. P icker L J and Butcher E C (1992) Physio logical and molecular mechan isms of lymphocyte homing. Annu. Rev. Immunol. 10:561. 173 Rice G E and Bev i lacqua M P (1989) A n inducible endothel ial cell sur face glycoprotein mediates me lanoma adhes ion . Science 246:1303. R i ce G E , Munro J M , and Bev i lacqua M P (1990) Inducible cell adhes ion molecule 110 ( INCAM-110) is an endothel ial receptor for lymphocytes. J. Exp. Med. 171:1369. Schmi ts R, Kundig T M , Baker D M , Shumaker G , S imard J J L , Duncan G , W a k e h a m A , Shah in ian A , van der Heiden A , Bachmann M F , Ohash i P S , Mak T W , and Hickstein DD (1996) LFA-1-def ic ient mice show normal C T L responses to virus but fail to reject immunogenic tumor. J. Exp. Med. 183:1415. Smith C W (1992) Transendothel ia l migration, in Harlan J M and Liu D Y (eds), A d h e s i o n : its role in inflammation, New York, N Y , F reeman. Smyth S S , Joneck is C C , and Par ise L V (1993) Regulat ion of vascu lar integrins. Blood 81:2827. Spr inger T A (1994) Traffic s ignals for lymphocyte recirculation and leukocyte emigrat ion: the multistep paradigm. Cell 76 :301. Spr inger TA , Gal f re G , S e c h e r D S , and Milstein C (1979) M a c - 1 : a differentiation ant igen identified by monoclonal antibody. Eur. J. Immunol. 9 :301. St. Louis J D , Lederer J A , and Lichtman A H (1993) Cost imulatory deficient ant igen presentat ion by an endothel ial cell line induces a nonproliferative T cell activation response without anergy. J. Exp. Med. 178:1597. S tosse l T P (1993) O n crawling of animal cel ls. Science 260:1086. Tedder T F , S teeber DA, and P izcueta P (1995) L-selectin-deficient mice have impaired leukocyte recruitment into inflammatory sites. J. Exp. Med. 181:2259. Vaporc i yan A A , DeL isse r H M , Y a n H C , Mendiguren II, Thorn S R , J o n e s M L , Ward P A , and A lbe lda S M (1993) Involvement of platelet-endothelial cell adhes ion molecule-1 in neutrophil recruitment in vivo. Science 262:1580. von Andr ian U H , C h a m b e r s J D , M c E v o y L M , Bargatze R F , Arfors K E , and Butcher E C (1991) Two step model of leukocyte-endothel ial interaction in inf lammation: distinct roles for L E C A M - 1 and the leukocyte (32 integrins in vivo. Proc. Natl. Acad. Sci. USA 88:7538. von Boxberg Y , Wutz R, and Schwarz U (1990) U s e of the biotin-avidin sys tem for label l ing, isolation and characterizat ion of neural cel l -surface proteins. Eur. J. Biochem. 190:249. 174 X u H, G o n z a l o J A , St. Pierre Y , Wi l l iams IR, Kupper T S , Cotran R S , Spr inger TA , and Gut ie r rez -Ramos J C (1994) Leukocytosis and resistance to sept ic shock in intercellular adhes ion molecule 1-deficient mice. J. Exp. Med. 180:95. Z immerman G A , Prescott S M , and Mclntyre T M (1992) Endothel ia l cell interaction with granulocytes: Tethering and signall ing molecules. Immunol. Today 13:93. 175 Chapter 5 5:1 D i s c u s s i o n The physiological function of ICAM-2 is the least understood of the three LFA-1 counter-receptors. The role of the LFA-1 : ICAM-1 interaction has been examined extensively and found to participate in virtually every adhes ion-dependent function a s s a y e d (Martz, 1987; Springer, 1990; Dustin and Springer, 1991). The L F A - 1 : I C A M - 3 interaction has also been studied and found to play a role in T cell activation (de Fougero l les et al., 1994). However, the role of ICAM-2 in the immune response has not been clarif ied. The goal of this thesis was to character ize further the functional role of the LFA-1 : ICAM-2 interaction. The first objective was to isolate the murine ICAM-2 c D N A and use it as a tool to examine express ion, t issue distribution, and binding capaci ty. The next objective was to investigate possib le functional roles of ICAM-2 . The c D N A w a s used to test the possib le role of ICAM-2 in T cell activation and leukocyte transendothel ial migration. The cloning of the murine ICAM-2 c D N A involved a unique P C R - b a s e d strategy combining sequence information of similar molecules. The unifying feature in the clustering of sequence data was the Ig domain structure (Wil l iams, 1987; Wi l l iams and Barc lay, 1988). Th is structural feature is present in molecules which function within the immune sys tem. Ant igen recognition is the fundamental focus of most members in the immunoglobul in superfamily. However, Ig-like molecules have been identified in certain invertebrate spec ies indicating that early ancestors of these domains were 176 present prior to the special izat ion of the immune sys tem. Ig superfamily members are also present in insects; these molecules are involved in the nervous sys tem and function in axon gu idance and fasciculat ion (Harrelson and G o o d m a n , 1988). The Ig domain may have branched out and was widely adopted in evolution because of its stability conferred by the disulphide-bonded p-strand structure (Wil l iams and Barc lay, 1988; Hunkapi l ler and Hood , 1989). It is commonly accepted that Ig-like s e q u e n c e s were der ived by gene duplication and divergence from a primordial domain . The d ivergence occurred to a level such that only the bas ic structure remained. In terms of function, these Ig-related structures probably played a primordial role in cell :cel l interaction. From an early role in cell adhes ion, the cell surface molecu les became likely candidates to control the behaviour of cel ls in the development of an immune sys tem. In the invertebrate Caenorhabditis elegans, certain developing neural cel ls undergo a programmed cell death fol lowed by phagocytos is (Horvitz et al., 1982; Hedgecock et al., 1983). In some cases , the neural cell is destroyed by recognit ion of a neighboring cel l . If this primitive natural killer cell specificity could be modif ied to include a requirement of a pathogen infection of cel ls, then an immune sys tem may have evo lved. The structural conservat ion of Ig-like domains through evolution suggests that they are stable configurations which can assist in many cellular funct ions. The main tenance of I C A M molecules across spec ies indicates that they play critical roles in the immune sys tem. Amino acid sequence analys is has revealed that the amino acid identity between human and murine ICAM-2 (60%) is higher than that for ICAM-1 177 (53%). In examining c ross-spec ies binding, it was found that human LFA-1 is able to a lso bind murine ICAM-1 and ICAM-2 in addition to the human counter-parts (Johnston et al., 1990; X u et al., 1992). Human:mur ine hybrid cel ls have shown that a L and p 2 of human and mouse are able to promiscuously co-assoc ia te and form interspecies a p comp lexes detected at the cell surface (Marlin et al., 1986; Larson and Springer, 1990). T h e s e hybrid complexes were used in c ross-spec ies binding a s s a y s and found that the human a L :mur ine p 2 complex is capab le of binding human I C A M - 1 . Var ious res idues in both LFA-1 and the I C A M counter-receptors have been conserved in both the mouse and human genome. Alteration of glutamic acid at residue 34 and glutamine at residue 73 in human ICAM-1 disrupts the ability to bind LFA-1 (Staunton et al., 1990). Interestingly, both of these residues are a lso present in the other four character ized human and murine I C A M s . In addit ion, the crystal structure of human I C A M - 2 has revealed that the glutamic acid residue lies in a p-strand which may interact with the l-domain of LFA-1 (Casasnovas et al., 1997). The degree of s e q u e n c e conservat ion is therefore high enough to use it as a bas is for c ross -spec ies cloning of ICAM- l ike molecules. In addition to the cloning of murine ICAM-2 of this thesis, can ine ICAM-1 has also been P C R amplif ied based on human and murine ICAM-1 s e q u e n c e s (Smith et al., 1991). This may provide a s imple method for identifying yet uncharacter ized ICAM-l ike molecules. The genomic organizat ion of the murine ICAM-2 gene confirms its ass ignment as a member of the Ig superfamily. A s with other members of this family, there is good correlation between the intron/exon organizat ion and the structural domains of 178 the protein (Wil l iams and Barclay, 1988; Hunkapil ler and Hood , 1989). The leader peptide, the two Ig-like domains, and the t ransmembrane/cytoplasmic domains are encoded by four exons separated by three type I introns. T h e s e type I introns are commonly found between Ig-like exons of other members of the Ig superfamily. A uniform intron phase is important for the evolution of the Ig superfamily because it ensures that molecu les with multiple Ig-like domains can be restructured by exon dupl icat ion and shuffling without altering the reading frame. G e n e dupl icat ion and d ivergence from a primordial domain are the likely origin of Ig-related molecu les. The cor respondence between protein domains and exon structure, as well as the uniform phase of the introns supports this theory of evolution. Further support c o m e s from x-ray crystal lography studies. Ev idence has shown that spl ice sites in genes usually map to amino acid residues located at the protein surface (Craik et al., 1982). This has been conf irmed with the crystal lographic structure of the Ig domains of immunoglobul ins and H L A molecules (Bjorkman et al., 1987; A lzar i et al., 1988). The genomic ICAM-2 clone isolated from this project did not contain enough s e q u e n c e upstream of transcription to a s s e s s the regulatory region adequately. However , another group has analyzed this region more thoroughly (Xu et al., 1992). They have found that this region contains a TATA- l i ke sequence , an inverted C A A T box, and a consensus transcription initiation sequence . However, no binding sites for the transcription factors NF-KB and A P - 1 were identified as they were for the ICAM-1 5' upst ream region (Voraberger et al., 1991). This is not surprising s ince ICAM-2 express ion is not inducible and the transcription factors NF-KB and A P - 1 are involved 179 in gene induction (Angel et al., 1987; Edbrooke et al., 1989). The only c a s e s in which I C A M - 2 express ion is elevated is on neoplast ic cel ls (Roos, 1991; Ell is et al., 1992; Renkonen et al., 1992). ICAM-2 appears to be a constitutively exp ressed gene. C p G is lands have been shown to be assoc ia ted with the 5' region of constitutively exp ressed genes and methylation can play an important role in gene express ion (Bird, 1986; Gard iner -Garden and Frommer, 1987; Cedar , 1988). Constitut ively exp ressed g e n e s contain C p G is lands which are unmethylated through all s tages of the cel l cyc le. Whether methylation plays a role in ICAM-2 express ion remains to be s e e n . T i ssue-speci f ic factors regulating the somewhat limited distribution pattern of I C A M - 2 a lso remain to be identified. One group has shown that the human ICAM-2 promoter is able to function in the murine system in a similar t issue-speci f ic manner observed in the human sys tem (Cowan etal., 1996). The ex is tence of a subfamily of Ig-related molecules ( ICAM-1 , -2 , and -3) with the ability to bind LFA-1 attests to the importance of these adhes ion pathways. A n inability to make these interactions in vivo results in the cl inical symptoms assoc ia ted with L A D (Anderson and Springer, 1987). The functional s igni f icance of each of these interactions has been examined to varying degrees. However, the LFA-1 : ICAM-2 contact is the least understood. Based on the structural and functional homology with ICAM-1 and ICAM-3 , ICAM-2 was tested for its ability to assist in T cell activation. T cell activation using purified ICAM-2 protein has shown that it can stimulate proliferation (Damle et al., 1992a; Damle et al., 1992b). Every LFA-1 -dependen t response appears to be greater for ICAM-1 than for ICAM-2 . It was initially thought 180 that the preference of ICAM-1 over ICAM-2 by LFA-1 was due to the smal ler s ize of the ICAM-2 (only two domains) protein not being able to extend its binding site as c lose to L F A - 1 . However, the difference may actually be qualitative. A modif ied vers ion of ICAM-2 with an extra five domains from CD31 has not shown any difference in proliferation from the native ICAM-2 molecule. It may be that LFA-1 can exist in multiple act ivated conformational states (Binnerts et al., 1994). Ev idence for this has been presented in which LFA-1 can recognize ICAM-1 but cannot recognize ICAM-3 . However , LFA-1 recognizing ICAM-3 is able to still bind to I C A M - 1 . The role of LFA-1 : ICAM-2 interaction in T cell proliferation to al loantigen has been examined . The response can be inhibited by ant ibodies against either molecule and thus providing a potential target for immune sys tem modulat ion in allograft accep tance . Ant ibodies have been administered previously for increased graft survival . The quest ion of whether the ant ibodies are merely non-speci f ic immunosuppress ive agents or whether they are assist ing in the induction of a to lerance state depends on each specif ic situation. The use of LFA-1 and ICAM-1 m A b s has been documented to increase the survival of var ious allografts. In one c a s e involving card iac allograft in mice, these two antibodies together were able to prolong the survival of the graft (Isobe et al., 1992). In addit ion, these mice were able to accept an al lo-skin graft indicating that the immune system had become tolerant to the al lotype of the graft. The next step would be to examine if the ICAM-2 ant ibody could a lso increase the survival of an allograft. If this is possib le, then whether the nature of the engraftment is immunosuppress ive or immunotolerant could be examined . 181 Endothel ium in solid organs express ICAM-2 and can be a site of an immune response. Ant ibodies may be able to block the immune response and prolong engraftment. The role of ICAM-2 in transendothel ial migration was examined . My work sugges ts that ICAM-2 present on endothel ial cel ls ass is ts neutrophil migration in an LFA-1 -dependen t manner. S ince neutrophils are the first leukocyt ic cel l type to migrate the endothel ium at areas of inflammation (Bevi lacqua, 1993; Car los and Har lan, 1994; Ager , 1994), ICAM-2 may also play a role in inflammation as well as recirculation. Ant i - ICAM-2 treatment may inhibit the migration and thus the inf lammatory response. The express ion of ICAM-2 on platelets (Diacovo et al., 1994) may a lso be tied in with migration. Al though post-neutrophil migration did not indicate that there were any large holes in the endothelial monolayer, it did not exc lude the possibil i ty that the monolayer had nonetheless been damaged . The migration p rocess involves junction d isassembly and it would be difficult to imagine that the junction would immediately reassemble . The holes would have to be repaired in order to prevent vascu lar leakage. O n e possibil ity is that the ICAM-2 on platelets may adhere to the trailing end of a migrating leukocyte and stop at the monolayer initiating the c a s c a d e of events required to mediate wound repair and acting like a "vascular band-aid". I C A M - 2 deficient mice may provide the answer to this possib le role of ICAM-2 on platelets. 182 The functional role of ICAM-2 has remained elusive. The exper iments descr ibed in this thesis have attempted to shed some light on this issue. I have shown that ICAM-2 can assist in T cell activation by providing a costimulatory s ignal . It can a lso mediate neutrophil migration across an endothelial monolayer. Therefore, I C A M -2 may play a significant role in immune and inflammatory responses. Ultimately these funct ions should a lso be tested in vivo. The c D N A and mAb speci f ic for murine I C A M -2 descr ibed in this thesis will be an important tool for such future studies. 183 5:2 R e f e r e n c e s A g e r A (1994) Lymphocyte recirculation and homing: roles of adhes ion molecu les and chemoattractants. Trends Cell Biol. 4:326. A lzar i P M , L a s c o m b e M, and Poljak R J (1988) Three-d imensional structure of ant ibodies. Annu. Rev. Immunol. 6:555. A n g e l P, Imagawa M, Ch iu R, Stein B, Imbra R J , Rahmsdor f H J , Jonat C , Herrl ich P, and Kar in M (1987) Phorbol ester- inducible genes contain a common cis e lement recognized by a TPA-modu la ted frans-acting factor. Cell 49:729. Bev i lacqua M P (1993) Endothel ial- leukocyte adhes ion molecules. Annu. Rev. Immunol. 11:767. Binnerts M E , van Kooyk Y , S immons DL, and Figdor C G (1994) Distinct binding of T lymphocytes to I C A M - 1 , -2, or -3 upon activation of L F A - 1 . Eur. J. Immunol. 24 :2155. Bird A P (1986) CpG- r i ch is lands and the function of D N A methylat ion. Nature 321:209. Bjorkman P J , S ape r M A , Samraou i B, Bennett W S , Strominger J L , and Wi ley D C (1987) Structure of the human c lass I histocompatibil ity ant igen, H L A - A 2 . Nature 329:506. Car los T M and Harlan J M (1994) Leukocyte-endothel ia l adhes ion molecu les. Blood 84:2068. C a s a s n o v a s J M , Spr inger TA , Liu J , Harr ison S C , and W a n g J (1997) Crystal structure of ICAM-2 reveals a distinctive integrin recognition surface. Nature 387:312. C e d a r H (1988) D N A methylation and gene activity. Cell 53:3. C o w a n P J , Shinke l TA , Witort E J , Bar low H, Pea rse M J , and d 'Apice A J (1996) Target ing gene express ion to endothelial cel ls in t ransgenic mice using the human intercellular adhes ion molecule 2 promoter. Transplant. 62 :155. Craik C S , Sprang S , Fletterick R, and Rutter W J (1982) Intron-exon spl ice junctions map at protein sur faces. Nature 299:180. Damle NK , K lussman K, and Aruffo A (1992a) Intercellular adhes ion molecule-2, a second counter-receptor for C D 1 1 a / C D 1 8 (leukocyte funct ion-associated antigen-1), provides a costimulatory signal for T-cel l receptor-initiated activation of human T cel ls. J. Immunol. 148:665. 184 Damle NK , K lussman K, Linsley P S , Aruffo A , and Ledbetter J A (1992b) Differential regulatory effects of intercellular adhes ion molecule-1 on costimulat ion by the C D 2 8 counter-receptor B7 . J. Immunol. 149:2541. de Fougero l les A R , Qin X , and Spr inger T A (1994) Character izat ion of the function of intercellular adhes ion molecule ( ICAM)-3 and compar ison with ICAM-1 and ICAM-2 in immune responses . J. Exp. Med. 179:619. D iacovo T G , de Fougero l les A R , Bainton DF , and Spr inger T A (1994) A functional integrin l igand on the surface of platelets: intercellular adhes ion molecule-2. J. Clin. Invest. 94 :1243. Dustin M L and Spr inger T A (1991) Ro le of lymphocyte adhes ion receptors in transient interactions and cell locomotion. Annu. Rev. Immunol. 9:27. Edbrooke M R , Burt DW, Chesh i re J K , and W o o P (1989) Identification of c/'s-acting s e q u e n c e s responsib le for phorbol ester induction of serum amylid A gene express ion via a nuclear factor KB-like transcription factor. Mol. Cell. Biol. 9:1908. Ell is P A , Hart D N , Col ls B M , Nimmo J C , McDona ld J E , and A n g u s H B (1992) Hodgkin 's cel ls express a novel pattern of adhes ion molecules. Clin. Exp. Immunol. 90:117. Gard iner -Garden M and Frommer M (1987) C p G islands in vertebrate genomes . J. Mol. Biol. 196:261. Harre lson A L and G o o d m a n C S (1988) Growth cone gu idance in insects: fascicl in II is a member of the immunoglobul in superfamily. Science 242:700. Hedgecock E M , Suls ton J E , and Thompson J N (1983) Mutation affecting programmed cell deaths in the nematode Caenohabditis elegans. Science 220:1277. Horvitz H R , Ell is H M , and Steinberg P W (1982) Programmed cell death nematode development . Neurosci. Commentaries 1:56. Hunkapi l ler T and Hood L (1989) Diversity of the immunoglobul in gene superfamily. Adv. Immunol. 44:1 • Isobe M, Yag i ta H, Okumura K, and lhara A (1992) Speci f ic accep tance of card iac allograft after treatment with ant ibodies to ICAM-1 and L F A - 1 . Science 255 :1125. Johnston S C , Dustin M L , Hibbs ML , and Spr inger T A (1990) On the spec ies specif icity of the interaction of LFA-1 with intercellular adhes ion molecules. J. Immunol. 145:1181. 185 Larson R S and Spr inger T A (1990) Structure and function of leukocyte integrins. Immunol. Rev. 114:181. Marl in S D , Morton C C , Anderson D C , and Spr inger T A (1986) LFA-1 immunodef ic iency d isease : definition of the genet ic defect and chromosomal mapping of a lpha and beta subunits by complementat ion in hybrid cel ls. J. Exp. Med. 164:855. Mar tz E (1987) LFA-1 and other accessory molecules functioning in adhes ion of T and B lymphocytes. Hum. Immunol. 10:3. R e n k o n e n R, P a a v o n e n T, Nortamo P, and Gahmberg C G (1992) Express ion of endothel ia l adhes ion molecules in vivo. Increased endothel ial ICAM-2 express ion in lymphoid mal ignancies. Am. J. Path. 140:763. R o o s E (1991) Adhes ion molecules in lymphoma metastasis. Cancer Metastasis Rev. 10:33. Smith C W , Entman ML , Lane C L , Beaudet A L , Ty TI, Youke r K, Hawkins HK, and A n d e r s o n D C (1991) Adherence of neutrophils to canine card iac myocytes in vitro is dependent on intercellular adhes ion molecule-1. J. Clin. Invest. 88:1216. Spr inger T A (1990) Adhes ion receptors of the immune sys tem. Nature 346:425. Staunton D E , Dustin ML , Er ickson H P , and Spr inger T A (1990) The arrangement of the immunoglobul in- l ike domains of ICAM-1 and the binding sites for LFA-1 and rhinovirus. Cell 61 :243. Voraberger G , Schafer R, and Stratowa C (1991) Cloning of the human gene for intercellular adhes ion molecule 1 and analysis of its 5'-regulatory region: induction by cytokines and phorbol esters. J. Immunol. 147:2777. Wi l l iams A F (1987) A year in the life of the immunoglobul in superfamily. Immunol. Today 8:298. Wi l l iams A F and Barc lay A N (1988) The immunoglobul in superfamily - domains for cell sur face recognit ion. Annu. Rev. Immunol. 6 :381. X u H, Tong IL, de Fougerol les A R , and Spr inger T A (1992) Isolation, character izat ion, and express ion of mouse ICAM-2 complementary and genomic D N A . J. Immunol. 149:2650. 186