You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Functional analysis of murine CD43 shedding : a role for the CD43 cytoplasmic tail in nuclear signalling Seo, Wooseok 2008

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2008_spring_seo_wooseok.pdf [ 3.23MB ]
JSON: 24-1.0066325.json
JSON-LD: 24-1.0066325-ld.json
RDF/XML (Pretty): 24-1.0066325-rdf.xml
RDF/JSON: 24-1.0066325-rdf.json
Turtle: 24-1.0066325-turtle.txt
N-Triples: 24-1.0066325-rdf-ntriples.txt
Original Record: 24-1.0066325-source.json
Full Text

Full Text

FUNCTIONAL ANALYSIS OF MURINE CD43 SHEDDING: A ROLE FOR THE CD43 CYTOPLASMIC TAIL IN NUCLEAR SIGNALLING  by WOOSEOK SEO M.Sc., The University of Calgary, 2002 B.Sc., Chong-Ju University, 1998  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Experimental Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  March 2008 © Wooseok Seo, 2008  ABSTRACT CD43, a representative of the leukocyte mucin family proteins, is a transmembrane protein highly expressed on most lymphohemopoietic cells and is believed to play a role in the regulation of leukocyte activation and/or migration. CD43 was shown to be proteolytically shed from human cells and high concentrations of soluble CD43 have been found in human plasma. The biological significance of CD43 shedding however remains enigmatic. To study the functional significance of CD43 shedding, we initiated our study by investigating whether CD43 shedding also occurs in the murine system and confirmed using flow cytometry, Western blot and ELISA techniques that murine CD43 is cleaved from the cell surface as is observed in the human system. To examine the biological significance of CD43 shedding, we designed and constructed non-sheddable forms of murine CD43. Ectopic expression of non-sheddable CD43 molecules in primary CD43 deficient bone marrow cells showed that these CD43 mutants have serious negative impacts on cell viability, revealing CD43 shedding as an essential process and implying that the CD43 mutants interfered with intracellular signaling processes. Our data support the hypothesis that CD43 ectodomain shedding is a requirement for release of the cytoplasmic domain and its translocation to the nucleus. In support of our hypothesis, we confirmed that the CD43 cytoplasmic domain is localized in the nucleus and is modified by SUMO (small ubiquitin-like modifier) peptides. In an attempt to determine the functional significance of CD43 nuclear translocation and SUMO modification, we examined nuclei from hemopoietic cells  more closely and observed that the CD43  cytoplasmic tail is localized in a subnuclear structure called promyelocytic nuclear bodies, which control many nuclear functions including apoptosis. Consistent with this observation we find that leukocytes from CD43 deficient mice have an increased apoptotic response upon growth factor  ii  withdrawal. We conclude that nuclear translocation of the CD43 cytoplasmic tail serves to control the apoptotic response in leukocytes and that CD43 functions as an anti-apoptotic molecule.  iii  TABLE OF CONTENTS Abstract……………………………………………………………………..……......…………....ii Table of Contents……………………………………………….…………….………..…………iv List of Tables…………………………………………………………………………........…… List of Figures…………………………………………………………..…………..………...….vii List of Abbreviations……………………………………………..…………………..……...…...xi Acknowledgments………………………………………………………...………………..…...xiv CHAPTER 1 INTRODUCTION……………………………………………………………….....1 1. Mucins…………………………………………………………………………………..1 1.1 Mucin family………………………………………………………………..…1 1.2 Sialomucins……………………………………………………………………2 1.2.1 Muc-1……………………………………………………………......3 1.2.2 CD34…………………………………………………………..…….4 1.2.3 PSGL-1………………………………………………………..…….5 1.2.4 CD45……………………………………………………………..….6 1.2.5 GlyCAM-1 and MAdCAM-1…………………………………….....7 2. CD43……………………………………………………………………………………8 2.1 Structure and expression of CD43………………………………………….....8 2.2 Glycosylation of CD43………………………………………………………10 3. History of CD43…………………………………………………………………….…13 3.1 Discovery of CD43………………………………………………………..…13 3.2 Clues about the function of CD43…………………………………………...14 4. Function of CD43……………………………………………………………………..15 4.1 B cells and CD43………………………………………………………….…15 4.2 CD43 knockout mouse…………………………………………………….…16 iv  4.3 CD43 as a physical barrier………………………...………………………....18 4.3.1 CD43 as a physical barrier………...……...………………………..18 4.3.2 More than a physical barrier……………………………………….19 4.3.3 CD43 and leukocyte migration…………………………………….20 4.3.4 Anti-adhesive properties of cells from CD43 null mice…………...22 4.3.5 CD43 and ERM proteins………………………………………...…23 4.4 CD43 as a signaling molecule………………………………………………..24 4.4.1 CD43 and leukocyte actvation……………………………………..24 4.4.2 Mechanism of CD43 signaling…………………………………….25 4.4.3 CD43 and apoptosis………………………………………………..25 4.5 Possible ligands for CD43…………………………………………………...27 4.5.1 ICAM-1…………………………………………………………….27 4.5.2 Influenza A virus…………………………………………………...28 4.5.3 MHC-I……………………………………………………………...28 4.5.4 Sialoadhesin………………………………………………………..28 4.5.5 Galectin-1…………………………………………………………..29 4.5.6 E-selectin…………………………………………………………...29 4.6 Diseases and CD43…………………………………………………………..30 4.6.1 AIDS and CD43……………………………………………………30 4.6.2 Cancer and CD43…………………………………………………..30 4.6.2 Tuberculosis and CD43…………………………………………….31 5. Shedding of CD43…………………………………………………………………….32 5.1 Shedding of CD43……………………………………………………………32 5.2 Shedding……………………………………………………………………..33 5.2.1 Shedding………………………………………………...................33 v  5.2.2 ADAM……………………………………………………………..34 5.2.3 APP………………………………………………………………...35 5.2.4 Notch……………………………………………………………….37 5.2.5 CD44……………………………………………………………….39 6. Nuclear localization…………………………………………………………………...39 6.1 Nuclear localization of CD43………………………………………………..39 6.2 Nuclear localization………….………………………………………………41 7. Sumoylation…………………………………………………………………………...44 7.1 SUMO………………………………………………………………………..44 7.2 Mechanisms of sumoylation…………………………………………………45 7.3 Function of SUMO…………………………………………………………..47 7.4 PML nuclear body……………………………………………………………51 8. Thesis objectives………………………………………………………………………56 CHAPTER 2 MATERIALS AND METHODS………………………………………………….57 CHAPTER 3 RESULTS…………………………………………………………………………64 1. Shedding of CD43……………………………………………………………………64 1.1 Down-regulation of CD43…………………………………………………...64 1.1.1 CD43 is down-regulated from cell surface……………………...…64 1.1.2 Inducers of CD43 down-regulation………………………………..66 1.1.3 CD43 down-regulation from different cell types…………………..69 1.2 Secretion of CD43 from cell surface………………………………………...71 1.2.1 CD43 shedding examined by ELISA………………………………71 1.2.2 Disappearance of endogenous CD43………………………………75 1.2.3 CD43 shedding examined by Western blotting…………………….76 2. Non-sheddable CD43…………………………………………………………….…...80 vi  2.1 Construction of non-sheddable CD43 chimeras……………………………..80 2.2 Toxicity of non-sheddable CD43 chimeras…………………………………..81 2.3 Physiological relevance of toxicity of non-sheddable CD43 chimeras……...83 2.4 Rescue of the toxicity caused by non-sheddable CD43 chimeras……….......88 3. Functional significance of the cytoplasmic tail of CD43……………………………..93 3.1 Nuclear localization of CD43……………………………………………..…93 3.2 Negative impacts on cell growth of CD43-GFP fusion……………………...97 3.3 Nuclear localization of CD43 studied by confocal microscopy…………..…99 3.4 Sumoylation of CD43……………………………………………………..…99 3.5 Colocalization of the CD43 cytoplasmic tail at PML nuclear bodies………103 3.6 CD43 and apoptosis regulation……………………………………………..106 4. Conclusion…………………………………………………………………..……….114 CHAPTER 4 SUMMARY AND DISCUSSION……………………………………………….115 References…………………………………………………………………..…………………..123 Appendix I: Map of pMPSF…………………………………………………………………....138 Appendix II: Constructs made to dissect the mechanisms of shedding of CD43…………… ...139 Appendix III: List of publications……………………..……………..……..………………..…140  vii  LIST OF TABLES Table 1. Proposed functions of CD43…………………………………………………………...21 Table 2. Proteins that are known to be localized in PML nuclear bodies……………………….55  viii  LIST OF FIGURES Figure 1.1: Structure of murine CD43.……………………………………………………………9 Figure 1.2: CD43 glycosylation………………………………………………………………….11 Figure 1.3: Processing of amyloid precursor protein…………………………………………….36 Figure 1.4: Notch synthesis and processing by proteolysis……………………………………...38 Figure 1.5: CD44 processing by proteolysis……………………………………………………..40 Figure 1.6: Mechanism of nuclear transport……………………………………………………..43 Figure 1.7: Mechanism of sumoylation………………………………………………………….46 Figure 1.8: Targets for SUMO…………………………………………………………………...49 Figure 1.9: Immunofluorescence staining of PML nuclear bodies……..………………..………52 Figure 3.1.1: Down-regulation of CD43 on mouse cell lines……………………………………65 Figure 3.1.2: Down-regulation of CD43 on mouse primary cells……………………………….67 Figure 3.1.3: Down-regulation of CD43 on mouse primary T cells……………………………..68 Figure 3.1.4: Down-regulation of CD43 of Gr-1+ cells………………………………………….70 Figure 3.1.5: Down-regulation of CD43 on mouse mast cells and macrophages………………..72 Figure 3.1.6: CD43 shedding examined by ELISA……………………………………………...74 Figure 3.1.7: Down-regulation of endogenous CD43……………………………………………77 Figure 3.1.8: CD43 shedding examination by Western blotting…………………………………79 Figure 3.2.1: Construction of non-sheddable CD43……………………………………………..82 Figure 3.2.2: Toxicity of non-sheddable CD43 chimeras (GFP+)...…………………………..…84 Figure 3.2.3: Toxicity of non-sheddable CD43 chimeras (cell numbers)……………………......85 Figure 3.2.4: Non-sheddable CD43 chimera on non- and sheddable cells………………………87 Figure 3.2.5: Absolute numbers of GFP+ mast cells with CD43/34 chimeras…………………...90 Figure 3.2.6: Absolute numbers of GFP- mast cells with CD43/34 chimeras…………………...92 Figure 3.3.1: Nuclear localization of CD43 revealved by CD43-GFP fusion…………………...95 ix  Figure 3.3.2: CD43-GFP fusion protein expression in MC/9 and CTLL2…….…………...…....98 Figure 3.3.3: Confocal microscopy of CD43…………………………………………………...100 Figure 3.3.4: Sumoylation of the cytoplasmic tail of CD43……………………………………102 Figure 3.3.5: Confocal microscopy of PML nuclear bodies……………………………………105 Figure 3.3.6: Disregulation of CD11c expression after apoptosis induction………………...…107 Figure 3.3.7: Impaired survival of CD43 null GR-1+ cells…………………………………..…109 Figure 3.3.8: TNFR signaling leading to survival or apoptosis pathway……………………....110 Figure 3.3.9: Impaired survival of CD43- cells……………………………………………..…..112 Figure 4.1.1 The mechanism of CD43 signalling……………………………………………... 122  x  LIST OF ABBREVIATIONS ADAM A disintegrin and metalloprotease AIDS Acquired immunodeficiency syndrome APP Amyloid precursor protein APC Antigen presenting cell APL Acute promyelocytic leukemia BACE β-APP-cleaving enzyme BMC Bone marrow cell C1β3GalT Core 1 β 1-3-galactosyltransferase C2GlcNAcT Core 2 β 1-6-N-acetylglucosaminyltransferase CLA Cutaneous lymphocyte-associated antigen DAG Diacylglycerol DC Dendritic cell ECM Extracellular matrix ELISA Enzyme-linked immunosorbent assay ERM Ezrin/Radixin/Moesin Gal galactose GalNAc N-acetylgalactosamine) GalT galactosyltransferase GAP GTPase activating enzyme GEF Guanine exchange factor) GlcNAc N-acetylglucosamine GlyCAM-1 Glycosylation dependent cell adhesion molecule-1 GST Glutathione-S-Transferase  xi  HA Hyaluronic acid HEC High endothelial cell HEV High endothelial venule HSF Heat shock factor IP Inositol phosphate ITAM Immunocreceptor tyrosine-based activation motif MAdCAM-1 Mucosal vascular addressin cell adhesion molecule-1 MAPK Mitogen activated protein kinase MHC Major histocompatibility class MMP Marix-associated metalloprotease MSCV Murine stem cell virus NES Nuclear export signal NLS Nuclear localization signal PBMC Primary Blood Mononuclear Cell PKC Protein kinase C PML Promyelocytic leukemia PNAD Peripheral-node addressin ppGalNAcT polypeptide N-acetylgalactosamine transferase). PS presenilins PSGL-1 P-selectin glycoprotein ligand-1 RAR Retinoic acid receptor SCF Stem cell factor ST3GalT α2,3 sialyltransferase STANK Sialophorin tail-associated nuclear kinase SUMO Small ubiquitin like modifier xii  TACE TNF-α converting enzyme TCR T cell receptor TDG Thymine DNA glycosylase TNF Tumor necrosis factor TRAIL TNF-related apoptosis-inducing ligand TPA 12- O -tetradecanoylphorbol-13-acetate VCAM-1 Vascular cell adhesion molecule-1  xiii  ACKNOWLEDGEMENT First and foremost, I would like to express my sincere gratitude to Dr. Hermann Ziltener, my supervisor. He was kind enough to accept me as a student and taught me with great enthusiasm. His ability to encourage and his wisdom to guide students were an invaluable asset. I would like to acknowledge the help of many people. I thank Dr. Douglas Carlow for guiding me through my Ph.D work and sharing his wealth of knowledge. He was my role model of a successful and confident scientist. I was so lucky to have him in the laboratory. I also would like to recognize many valuable contributions from Michael Williams. He taught the basic skills of research that I needed in the new environment and how to think scientifically. I wish to thank laboratory members with whom I have shared lab experiences. I would like to make special mention about committee members for valuable suggestion and criticism. I am deeply indebted to the little guys as we always refer them. It has been impossible to carry out my research without the sacrifice of countless mice. I can say that they have never been sacrificed for insignificant reasons. I can say that I did my best to treat them with respect. Nonetheless it was not their voluntary decision to be sacrificed. It was just my selfish choice. I wish there will one day be experimental models which can eliminate animal uses completely in research. I am sorry, mice. Last but not least, I cannot finish without saying about my wife, Kyounga Seo. I thank her for encouragement and her continuous love. She has always believed in me and stood by my side. All I can say is “I love you” and “At last, I am not a student anymore”. I dedicate this thesis to my wife.  xiv  CHAPTER 1 INTRODUCTION 1. Mucins 1.1 Mucin family Mucins are glycoproteins that are found in mucous area of epithelia on gastrointestinal, pulmonary and urinary tracts where they form a mucous barrier to provide protection from environmental stress and pathogens. Mucins are also expressed on endothelial cell surfaces as transmembrane proteins. Extracellular domains of mucins form a rigid rod-like shape and extend about 200-500 nm from the cell surface. They are tall enough to mask most of the other cell surface molecules which typically extend 10-30 nm from the surface. Mucins are heavily glycosylated with mainly simple, hydrophilic and acidic oligosaccharides and their molecular weights are usually more than 200 kDa with carbohydrates contributing to over 50% of their molecular weights. The bulky extracellular domain of mucins can be cleaved from epithelial cell surfaces and released into the surroundings as massive aggregates to form mucus. The extracellular domain typically contains several domains called apomucin domains (core protein). Apomucin domains mainly consist of tandem repeats of serine and threonine residues interspaced by prolines where O-linked glycosylation branches out via N-acetylgalactosamines attached to threonine and serine residues. This massive and generally negatively charged glycosylation gives the cells expressing mucins anti-adhesive properties. Mucins are generally divided into two families, transmembrane mucins and secreted gel-forming mucins. Secreted gel-forming mucins are released from goblet cells and go through homo-oligomerization and become the major constituent of mucus. Muc2, 5, 7, 8 and 9 belong to this family. On the other hand, the family of transmembrane 1  mucins includes Muc-1, 3, 4, 12 and 13. They have the structure of a typical transmembrane protein with an extracellular domain, a transmembrane domain and a cytoplasmic domain. In many cases, the cytoplasmic domain has several motifs that can interact with signaling molecules. Transmembrane mucins can be cleaved and homotypically associate with each other by non-covalent interaction during biosynthesis. Alternatively they can be expressed on the cell surface and then cleaved and secreted on demand. Some mucins are expressed on hematopoietic cells and cancer cells, but their functions in most cases still remain elusive.  1.2 Sialomucins Many mucins are decorated by sialic acids such as N-acetylneuraminic acids generally as terminal sugar on the O-linked glycosylation side chains, thus they are called sialomucins. Some of these transmembrane sialomucins have become an emerging family of proteins based on their structural resemblance. They include Muc-1, CD34, CD162 (PSGL-1), CD45RA, CD43 (leukosialin), GlyCAM-1, MadCAM-1, CD164 and the list is growing. Since the family members are expressed on very different cell types, they are thought to have different functions depending on where and when they are expressed. Since the extracellular domain is a mucin domain, they are generally believed to have properties resembling adhesion molecules. For some of them specific ligands have been found, but for many of them no physiological ligands are known. While the mucin ectodomain is believed to be mainly involved in adhesive process, the cytoplasmic domains are believed to be involved in various signaling mechanisms.  2  1.2.1 Muc-1 Muc-1 (CD227) was the first mucin protein to be fully sequenced and characterized in detail (Brockhausen 2006). It is expressed on almost all epithelial cells of mucous tissues and functions as a protective coat on apical cell surface. Recently it has been discovered that many hematopoietic cells such as bone marrow mononuclear cells, activated T cells, B cells, and activated dendritic cells express Muc-1 and that the cytoplasmic domain of Muc-1 can be phosphorylated on serine and threonine residues and may interact with signaling molecules (Wykes et al., 2002). However, the primary function of these mucins on hematopoietic cells is not yet understood. There has been significant interest in Muc-1 as it is overexpressed on the surface of many human adenocarcinomas including breast, ovarian, pancreatic, colorectal and prostate carcinoma as well as on hematological tumors including multiple myeloma and some B-cell nonHodgkin lymphomas (Agrawal et al., 1998; Brugger et al., 1999; Treon et al., 1999). The expression of Muc-1 on cells other than normal epithelial tissues does not show apical localization implying that Muc-1 loses its polarity and topology changes as tumor progresses. Also the glycosyl modifications are significantly changed. For example, it has been shown that Muc-1 is underglycosylated in human adenocarcinomas resulting in exposure of an antigenic peptide which is hidden on Muc-1 of normal cells. This has been an area of intense study for cancer biology and Muc-1 is now generally considered as a cancer-associated antigen. Muc-1 deficient mice appear to be healthy other than the predisposition to bacterial conjunctivitis and chronic infection of the reproductive tract due to decreased mucus (Spicer et al, 1995).  3  1.2.2 CD34 CD34 is normally expressed on hematopoietic stem cells and stromal cells in bone marrow and endothelial cells including HEV (Krause et al., 1996). More recently CD34 has also been found to be expressed on mast cells (Drew et al., 2002). Only 1.5% of bone marrow cells are CD34+, but they include most of the hematopoietic stem cell/progenitor population. Therefore, CD34 staining has been used clinically and scientifically to purify stem cells and early progenitors from bone marrow (Bhatia et al., 1997). Its presence on leukocytes and endothelial cells suggested that it might have dual functions. CD34 deficient mice have no obvious phenotypes and develop normally (Suzuki et al., 1996). They have normal hematopoiesis and peripheral blood cellularity as well as interaction between bone marrow cells and stromal cells. Known phenotypes include a delayed erythroid and myeloid differentiation, reduced frequency of colony forming cells and retarded growth of progenitors in vitro. One strong clue for the function of CD34 on HEV is that CD34 has been shown to be a strong ligand for L-selectin (CD62L) and E-selectin (CD62E) proteins in vitro (Baumhueter et al., 1993). This suggested that CD34 might mediate normal lymphocyte recirculation by acting as a HEV ligand of L- and E-selectin. However, CD34 null mice show no clear defects in lymphocyte interaction with HEV in vivo. Therefore, it is not clear whether CD34 is required as a physiological ligand for L- and E- selectins in vivo. A more recent study showing that CD34 and CD43 inhibit hematopoietic precursor reconstitution in vivo suggested that CD34 as well as CD43 might mediate hematopoietic precursor movement by acting as a non-specific anti-adhesive molecule (Drew et al., 2005). 4  1.2.3 PSGL-1 PSGL-1 (CD162, P-selectin glycoprotein ligand-1) is constitutively expressed on all leukocytes and is essential for the recruitment of leukocytes into inflamed tissue where circulating leukocytes cannot be stopped by integrins alone (Sackstein 2005). With the exception of some tissues such as lung and liver, the endothelial cells of the vascular walls in inflamed tissue express P-selectin and/or E-selectin which bind to PSGL-1 expressed on leukocytes that are flowing through the blood vessel. The PSGL-1 interactions with selectins have fast on and off kinetics resulting in a slowing down of the leukocytes visible as a “tethering and rolling” of cells along the vessel walls. PSGL-1 can also interact with L-selectin which is expressed on all leukocytes mediating leukocyte rolling “on” leukocytes that have already adhered to the vascular wall in a so called secondary tethering. To make contact with selectins, PSGL-1 needs to be properly modified by multiple steps of modification including O-glycosylation, sialylation, fucosylation and tyrosine sulfation at appropriate positions near the N-terminus (Sako et al., 1995). PSGL-1 knockout mice develop normally and show no severe phenotypes (Yang et al., 1999). A leukocyte infiltration model with induced peritonitis showed significantly delayed reaction to antigens in PSGL-1 deficient mice. Also, intravital microscopy on postcapillary venules of cremaster muscles showed decreased leukocyte rolling after trauma in vivo. These results confirmed the role of PSGL-1 on leukocyte recruitment which was suggested by earlier in vitro binding experiments. More recently, a novel experiment found that PSGL-1 might mediate leukocyte homing to secondary lymphoid tissues such as thymus and lymph nodes even in non-inflammatory conditions and that 5  PSGL-1 might play a role in leukocyte chemotaxis (Veerman et al., 2007), which could explain the observation that PSGL-1 null mice show lymphopenia.  1.2.4 CD45 CD45 is a transmembrane tyrosine phosphatase that can occur in at least nine different isoforms that result from alternative RNA splicing of a single gene (Penninger et al., 2001). The expression of a particular isoform depends on cell type and differentiation stage. CD45 is expressed on all cells of hematopoietic origin except erythrocytes and is widely used as a marker for leukocytes. CD45 deficient mice have distinctive phenotypes that include defective thymocyte development, markedly reduced numbers of mature peripheral T cells and a total lack of antigen receptor mediated activation of T and B cells (Kishihara et al., 1993). The cytoplasmic domain of CD45 has a tyrosine phosphatase that actively dephosphorylates negative-regulatory residues of protein tyrosine kinases which are involved in receptor mediated signaling. The formation of antigen-receptor complexes on T and B cells brings src kinases near the antigen-receptor complexes via lipid rafts. The default state of src kinases is inactive because their negative regulatory residues are phosphorylated. CD45 dephosphorylates inactive src kinase which then becomes activated by refolding and phosphorylation on a positive regulator residue. The src kinases then phosphorylate ITAMs (Immunoreceptor Tyrosine-based Activation Motifs) of B- and T-cell receptor complexes that serve as a recognition motif for ZAP-70 thus initiating the cascade of signaling molecules in lymphocyte activation. In summary, CD45 is an essential positive regulator of T cell and B cell antigen-receptor mediated 6  signaling pathway which is required for the development and activation of lymphocytes.  1.2.5 GlyCAM-1 and MAdCAM-1 Lymphocytes continuously and selectively home to lymph nodes from blood. In the lymph nodes they interact with APCs (Antigen Presenting Cells) and then leave lymph nodes via lymphatics to recirculate. To facilitate the entry of lymphocytes, secondary lymphoid organs including lymph nodes have special blood vessels called HEVs (High Endothelial Venules) made up with HECs (High Endothelial Cells). The constitutive expression of L-selectin on lymphocytes enables them to tether and roll on HEVs by interacting with glycoprotein receptors of L-selectins which are expressed on HECs (Rosen 1999). There are four known L-selectin receptors on HECs, which are GlyCAM-1 (GLYcosylation dependent Cell Adhesion Molecule-1), MAdCAM-1 (Mucosal vascular ADdressin Cell Adhesion Molecule-1), CD34 and podocalyxin. These mucins are called peripheral-node addressins (PNADs). All of the ligands mentioned above show a capability to bind L-selectin in vitro, but whether they constitute the physiological ligands in vivo has not yet been conclusively shown. These and other sialomucins expressed on normal venules are glycosylated differently and lack the PNAD epitopes. GlyCAM-1 is primarily expressed on HEVs of lymph nodes as a 50-60kDa GPIanchored protein and also present as a secreted protein in serum and milk (Lasky et al., 1992; Dowbenko et al., 1993). Its expression is constitutive on HEVs but inducible in inflamed tissues. When it is correctly glycosylated and sulfated, it can be recognized by L-selectin with high affinity and specificity (Hemmerich et al., 1995). How GlyCAM-1 7  mediates adhesion is not precisely known, but it can activate β2-integrins when bound on surface of a circulating neutrophils (Hwang et al., 1996). However, modulation of its secretion during immune responses could also suggest that GlyCAM could be an antiadhesive molecule (Hoke et al., 1995). MAdCAM-1 is expressed on HEVs of Peyer’s patch and the intestinal lamina propria. MadCAM-1 is structurally classified as a member of the immunoglobulin supergene family and made up with Ig-like domains and a heavily glycosylated segment. Unlike GlyCAM-1, MAdCAM-1 is involved in the firm cell adhesion by the interaction with LPAM1 and VLA4 in addition to leukocyte rolling.  2. CD43 2.1 Structure and expression of CD43 The CD43 gene is localized on chromosome 16 in human (Pallant et al., 1989) and chromosome 7 in mouse (Clare et al., 1990). In both species the gene contains no introns. CD43 is a type I transmembrane protein, and in mice has 395 amino acids including a 20 amino acid signal peptide (Figure 1.1). The 228 amino acid extracellular domain has a rigid rod-like shape that extends about 45nm from the cell surface similar to other mucins and sialomucins. The ectodomain is bulky and negatively charged due to the extensive modification by an estimated 70-80 O-linked carbohydrate chains and abundant terminal sialic acids residues (Carlsson et al., 1986; Cyster et al., 1991). The transmembrane domain has 23 amino acids and the cytoplasmic domain has a 124 amino acids. The cytoplasmic tail is highly conserved across species (70% identity between human and mouse) and contains many serine and threonines which are potential phosphorylation sites (Cyster et al., 1990). However, CD43 bears no sequence homology 8  Figure 1.1 Structure of murine CD43 CD43 has 228 amino acids extracellular domain which has a rigid rod-like shape which extends about 45nm from the cell surface. It is also heavily glycosylated, which confers the molecule bulkiness and negative charges. The 124 amino acids cytoplasmic domain is well conserved across species and contain a SH3 binding domain and many Ser/Thr, both of which can mediate signaling reactions.  9  to any other known protein even though its general structure is similar to another sialomucin CD34. CD43 is one of the most abundant proteins on leukocytes, for example, it is estimated that each T cell expresses about 100,000 molecules on the cell surface (Cyster et al., 1991). CD43 is widely expressed on all hematopoietic cells including all types of leukocytes and platelets. The exceptions are naive resting B cells and erythrocytes. This tissue specific expression of CD43 is achieved by DNA methylation in the 5’-region of the gene (Kudo et al., 1995). Considering the ubiquitous expression of CD43 on leukocytes, it is surprising that CD43 is not expressed on naïve resting B cells. It has been shown that B cells have a unique CD43 expression pattern. CD43 is expressed throughout the stem cells and B cell progenitor stages, but is lost on naïve resting B cells when they leave the bone marrow. They circulate through secondary lymphoid organs via lymphatics and blood vasculature without expressing CD43. However, once they are activated, CD43 is turned on and remains expressed on effector B cells and terminally differentiated plasma cells.  2.2 Glycosylation of CD43 Typical glycosylation of mucins is mostly O-linked glycosylation and may include a few N-linked glycosylation groups. Mucin O-linked glycosylation initiates from serine and threonine residues. The composition of O-linked glycosylation on mucins is heterogeneous, but they typically are based on core1 and core2 branching (Figure 1.2). First, GalNAc (N-acetylgalactosamine) is incorporated on the residues by a ppGalNAcT (polypeptide N-acetylgalactosamine transferase). From this point, either core1-branching 10  Figure 1.2 CD43 glycosylation CD43 is decorated by many O-linked glycosylation on the mucin domain and it isdependent on activation status and cell-types. Due to differential glycosylation of Olinked glycans, considerable molecular weight heterogeneity is observed. Two major forms of O-linked glycans are found. 115 kDa (tetrasaccharide) form is produced by core1 branching. 135 kDa (hexasaccharide) form is produced by core2 branching which is mediated by C2GlcNAcT2.  11  or core2-branching can be added. If a Gal (galactose) is added on the first GalNAc via β3 linkage by a C1β3GalT (Core 1 β 1-3-galactosyltransferase), this forms core1-branching. Core2-branching can be formed by an addition of a GlcNAc (N-acetylglucosamine) on the GalNAc of disaccharide core1 branching via β6 linkage by a C2GlcNT (Core 2 β 1-6N-acetylglucosaminyltransferase). Core1 and core2 branchings can be extended by adding more either Gal or GlcNAc. Typical mucins have terminal sialic acids and can be further modified by sulphates to add more complexity. Core1 and core2 glycosylations are the most prominent type of O-glycan structures. It is less common but the first GalNAc can go through core3- and core4-branchings. The complexity of the glycan structure gives more diverse functions to mucin. For example, the core2-branching can be elongated by additions of fucoses, sialic acids and sulfates to form sialyl Lewis X structure (sLex) which can serve as ligands for selectins. CD43 shows substantial molecular weight heterogeneity which is evident from Western blots due to differential glycosylation. The content of CD43 glycosylation can vary extensively depending on different cell types and with different cell activation status (Barran et al., 1997; Jones et al., 1994). Differential glycosylation results from the Olinked glycan branching from the serine and threonine residues of the polypeptide backbone (Cyster et al., 1990). CD43 can be decorated by either core1-branched tetrasaccharides or by core2-branched hexasaccharides (Fukuda et al., 1991). The tetrasaccharide form is composed of the core1 form (2 sialic acids, a Gal and a GalNAc) on the serine/threonine. The hexasaccharide form includes a core2 branch (a GlcNAc and a Gal) attached to the first GalNAc in addition to core1 branch. This gives rise to two major M.W. glycoforms of CD43, 115 kDa and 135 kDa. The 135 kDa form 12  (hexasaccharide form) is expressed on activated T cells, neutrophils and platelets and depends on the core2-branching mediated by C2GlcNT-I (Fukuda et al., 1991; Ellies et al., 1994). Resting T cells express the 115 kDa form of CD43 (tetrasaccharide) which has core1-branching and lacks core2 O-glycans. Our group developed the CD43 mAb 1B11 which reacts exclusively with the 135 kDa form of CD43 (Jones et al., 1994). This antibody has been an excellent tool because it enabled us to use CD43 expressed on T cells as an indicator of C2GlcNacT-I enzyme activity (Barran et al., 1997; Carlow et al., 1999). Use of 1B11 in combination with a second CD43 mAb S7 which binds exclusively to 115 kDa form of CD43 has significantly aided our studies of O-glycan modifications and in particular study of selectin ligand formation as these ligands depend on the core2 enzyme.  3. History of CD43 3.1 Discovery of CD43 CD43 was discovered 30 years ago as a unique T-lymphocyte surface glycoprotein expressed on mouse cytotoxic T cells (Kimura et al., 1978). Biochemical studies showed that it migrates to about 145 kDa (hence it was named as T145). T145 was expressed on certain types of T cells depending on the stages of T cell differentiation. CD43 was also discovered by another group that described a serine-threonine-rich galactoglycoprotein in human plasma (Schmid et al., 1980). Biochemical studies revealed that 75% of its molecular weight results from carbohydrates including large amounts of galactose, GalNAc, GlcNAc and sialic acids. Protein analysis showed that there are many serines and threonines on the protein which could be used for O-lined glycosylation. In 13  addition to this, there was evidence for several N-linked glycosylation groups as well. Because of its unusual extensive and complex glycosylation pattern compared to known other plasma glycoproteins, the group postulated that this molecule must be derived from cell surfaces. CD43 became of great interest after the observation that on lymphocytes this glycoprotein is deficient and/or defective in Wiskott-Aldrich patients, a syndrome which is an X-linked recessive disorder characterized by reduced T cell function (RemoldO’donnell et al., 1984). They found that on T lymphocytes of Wiskott-Aldrich syndrome patients a 115 kDa protein (hence they named it gpL115) could not be detected or show abnormal molecular weights with Western blots using L10, the first mAb against human CD43. Extensive biochemical studies revealed very complex patterns of glycosylations, but they failed to find a clear link between CD43 and the disease. Nonetheless this spurred much attention and set off a considerable research effort on the molecule. Cell surface staining of blood leukocytes with L10 demonstrated that CD43 (also called as sialophorin or gpL115 at that time) is expressed on all T lymphocytes, thymocytes, subpopulations of B lymphocytes, monocytes, neutrophils and platelets (Remold-O’Donnell et al., 1987). It was furthermore found that there are at least two forms of CD43 with 115 kDa and 135 kDa molecular weights due to different contents of O-linked carbohydrate units. However, the biological significances of the heterogeneity in molecular weights were not understood.  3.2 Clues about the functions of CD43 The first clue about the function of CD43 was suggested by the group who 14  developed L10. They showed that human blood T lymphocytes can be activated with soluble L10 to proliferate (Mentzer et al., 1987). Another group showed similar results by showing that L10 ligation results in the enhanced autocrine secretion of IL-2 as well as proliferation of T cells (Axelsson et al., 1988). This suggested that CD43 might regulate T cell proliferation and function through signaling. Interestingly, this second group also discovered an unexpected finding that a low amount of L10 (or L10 Fab and F(ab’)2) can induce a rapid clustering of blood T cells and T cell lines. This clustering could be blocked by the treatment of anti-CD18 antibody implying that this clustering is mediate by integrins. This result has been interpreted in two ways (Ostberg, et al., 1998). Firstly, anti-CD43 ligation can generate signaling cascades which modify cell-cell adhesive properties through other adhesion molecule. Secondly, anti-CD43 ligation can downregulate CD43 by shedding, endocytosis or capping which eliminates anti-adhesive mucin domain of CD43 allowing increased homotypic adhesion. This has led some people to believe that CD43 has to be either a signaling molecule or anti-adhesive physical barrier.  4. Function of CD43 (see Table 1) 4.1 B cells and CD43 One laboratory approached CD43 in a unique way to find its function. They investigated why CD43 was not expressed on naïve peripheral B cells while CD43 was found on all other leukocyte lineages. In an attempt to investigate the physiological role of CD43 on B lymphocytes in vivo, they generated mice transgenic for CD43 driven by the Ig heavy chain enhancer sequence forcing naïve peripheral B cell to express CD43. 15  The dysregulated expression of mouse CD43 did not affect B cell surface markers nor splenic localization of B cells, but the numbers of splenic B cells increased by 70%. The increased numbers of B cells were due to an increased longevity which seems to be derived from decreased apoptosis (Dragone et al., 1995). This suggested that CD43 might deliver a signal that can rescue B cells from going through programmed cell death. It was not clear whether CD43 functioned as a anti-apoptotic signaling molecule and was actively delivering a survival signal from stromal cells to B cells or whether CD43 merely acted as an anti-adhesive barrier molecule preventing apoptotic signals like fasfas ligand interaction from being delivered into the cells. Surprisingly, the transgenic mice had markedly decreased humoral responses in several in vivo assays (Ostberg et al., 1996). These data suggested that the defects in humoral responses were caused by the anti-adhesive properties of CD43 which might have prevented B cells from interacting with other cells. While this transgenic model has given us a unique way of looking at CD43, it still failed to give a clear answer on CD43 function and failed to resolve the question whether it functions as an adhesion molecule or a signaling molecule.  4.2. CD43 knockout mouse CD43 deficient mice were described in 1995 (Manjunath et al., 1995). The mice develop normally and are fertile. Interestingly the authors noted that T lymphocytes from CD43 null mice have a significantly increased in vitro proliferative response to ConA, anti-CD3 antibody and superantigen SEB. The group also showed that CD43 deficient mice exhibit hyper-responsiveness both in vitro and in vivo in response to TCR-CD3 16  mediated stimulation as well as TCR independent activation (Thurman et al., 1998). Moreover, CD43 deficient thymocytes and splenocytes showed increased homotypic adhesion either by their own or by PMA stimulation. These cells also showed increased adherence to fibronectin in an integrin dependent way. A second group showed that the cytoplasmic tail of CD43 is sufficient and necessary for the negative regulatory function of CD43 (Walker et al., 1999) and that GPI (GlycosylPhosphatidylInositol)-linked CD43 failed to be excluded from the T cell-APC contact sites (Tong et al., 2004) suggesting that the regulation of T cell activation through CD43 is mediated by the intracellular signaling mechanisms not by steric hindrance of the extracellular domain of CD43. It should be noted that many of these experiments with CD43 deficient mice had problems associated with mouse background issues. Our laboratory in a later study challenged these findings and showed that the T cell hyperresponsiveness was most likely due to a 129 mouse background problem. When cells from CD43 null mice that were sufficiently backcrossed (9 generations) to C57BL/6 were used, the hyperresponsiveness and altered patterns of cell migration were absent (Carlow et al., 2001). Moreover when cells from CD43+/+ and CD43-/- mice of C57Bl/6 background were compared competitively in vivo, neither neutrophils nor T cells had any differences in recruitment efficiencies (Carlow et al., 2006). With these properly backcrossed mice, T cells showed perfectly normal development and responsiveness. The only obvious phenotype of CD43 null mice that remained unchallenged was their increased homotypic adhesion of hematopoietic cells, a phenotype that was already predicted by studies of CD43 deficient CEM cell line and anti-CD43 ligation experiments. This leaves us with the question why 17  virtually all hematopoietic cells express this molecule.  4.3 CD43 as a physical barrier 4.3.1. CD43 as a physical barrier CD43 has been generally regarded as an anti-adhesive molecule because it is a mucin molecule even though it is expressed on hematopoietic cells rather than on endothelial cells. It has a typical mucin-like structure with a rigid and large rod-like shape and is decorated with negatively charged sugars. Moreover, L10 ligation increases the aggregation of T cells which could work either by masking the surface of CD43 ectodomain to block its anti-adhesive properties or by inducing down-regulation of the anti-adhesive extracellular mucin domain. Because of the wide expression of CD43, it was intriguing to see if the observed increased aggregation of T cells by L10 ligation can be applied to other cells. Indeed the group that studied T cell aggregation showed that blood monocytes can also be induced to increased homotypic adhesion (Nong et al., 1989). In addition to T cells and monocytes, DCs derived from either epidermal Langerhans Cells or from blood monocytes form much larger clusters with T lymphocytes when they are preincubated with the anti-CD43 mAb, MEM-59 (Fanales-Belasio et al., 1997). Anti-CD43-treated DCs also showed better antigen presentation to T cells and the authors suggested that this is probably due to the removal of CD43 as a physical barrier by either shedding or capping. Even before CD43 knockout mice were developed, one group generated a CD43 negative human T cell line (CEM) by targeted homologous recombination (Manjunath et 18  al., 1993). The CD43 negative cells showed enhanced homotypic adhesion which could be reversed by reintroducing the CD43 gene. Interestingly, this increased adhesiveness can be specifically blocked by anti-integrin antibodies suggesting that CD43 inhibits integrin dependent adhesion probably by acting as a physical barrier. This reconfirmed the hypothesis that CD43 might regulate leukocyte adhesion by acting as a physical barrier, but further suggested that CD43 could provide a threshold that needs to be overcome for cells to interact via integrins. CD43 has been shown to be excluded from the T cell-APC (antigen presenting cell) contact site mediated by moesin (Sperling et al., 1998; Delon et al., 2001). This observation furthermore supports the notion that CD43 is a barrier molecule that needs to be excluded from the sites of cell-cell contact. Interestingly another sialomucin CD45, which is as large and tall as CD43, was not excluded showing this to be a selective process.  4.3.2 More than a physical barrier One group developed several novel anti-CD43 antibodies to test their effects on T cell adhesion and found that some of them induced integrin-mediated T cell adhesion in vitro (Sanchez-Mateos et al., 1995). Specifically there were certain anti-CD43 antibodies that can induce T cell adhesion selectively via β1 and β2 integrins suggesting that this induction is not caused by unspecific mechanisms such as the down-regulation of physical barrier molecules upon antibody ligations. Therefore, they suggested that crosslinking of CD43 might simulate the cognate ligand binding and stimulate a signaling pathway toward integrin adhesion. This theory still allows CD43 to be a physical barrier 19  for non-specific interactions as other studies suggested, but also allows the possibility that CD43 can act as a signaling molecule. Later the same group found that CD43 is redistributed to cellular uropods of T lymphoblasts when the cells were induced to adhere to fibronectin fragments or VCAM-1 by upregulating β1 and β2 integrins after CD43 crosslinking in vitro. On the other hand, CD43 molecules appear to be distributed uniformly over T lymphoblasts surface when cells were in suspension or adhered to several β1-integrin ligands. It is postulated that this redistribution of CD43 might be a part of control mechanisms of CD43. Since the redistribution is faster than the cleavage of CD43 and is probably reversible, it could be a first and rapid regulatory step as observed on neutrophils that have been shown to redistribute CD43 to the uropods after anti-CD43 crosslinking (Seveau et al., 2000). In contrast, shedding and endocytosis could provide a more prolonged mode of the regulation of CD43.  4.3.3 CD43 and leukocyte migration CD43 as an anti-adhesive molecule could potentially affect migratory properties of leukocytes in addition to modulation of homotypic adhesion interactions. When L11, anti-CD43 antibody, was used to treat mouse T cells, it effectively blocked the binding of T cells on HEV (High Endothelial Venule) in frozen lymph node and Peyer’s patch sections in vitro (McEvoy et al., 1997 (1)). Moreover it inhibited most of the short term (1-24 hrs) homing of T cells into secondary lymphoid organs in vivo. The authors believed that L11 could block the interaction between CD43 and its cognate ligands in HEV. The attempt to find a ligand of CD43 on endothelial cells was not successful as L11 20  Table 1. Proposed functions of CD43 (table from Ostberg et al., 1998) Since its discovery, studies have suggested several functions for CD43 including proadhesion, anti-adhesion, signaling, cytoskeletal interaction, apoptosis. A definitive physiological function of CD43 has still not been found and some of them are directly contradictory.  21  failed to affect T cell binding onto purified vascular ligands such as L-selectin, α4β7, or LFA-1. This work shows that the anti-adhesive property of CD43 can go further than just homotypic interactions and that CD43 might modulate the dynamics of leukocyte migration by a not yet explained mechanism. The above study investigated the role of CD43 on T cell homing in a noninflammatory situation. Under inflammatory conditions L11 was found to inhibit the recruitment of monocytes to the inflamed peritoneal cavity indicating that CD43 might control the inflammatory infiltration of monocytes (McEvoy et al., 1997 (2)). Also, L11 has been shown to block the migration of T cells to inflamed pancreatic islets (Johnson et al., 1999). However, these experiments need to be analyzed carefully since the treatment of cells with antibodies can produce many unexpected results. For example, antibody crosslinking can bring other molecules together as well as CD43 via lipid rafts which can produce non-physiological responses even though there is still a possibility that this is a physiological process. Therefore, it was not clear if CD43 actively controls the migration of leukocytes.  4.3.4 Anti-adhesive properties of cells from CD43 null mice Artifacts from antibody crosslinking experiment can potentially and partially be overcome by using CD43 deficient mice. The group who constructed CD43 null mice later showed that T lymphocytes from CD43 deficient mice homed significantly more frequently to secondary lymphoid organs suggesting that the increased stickiness of cells might allow them to bind to other cells more efficiently (Stockton et al., 1998). This was 22  also supported by the finding that CD43 null T cells showed increased tethering and rolling on HEV. This finding has been considered as the most significant findings for the elucidation of the function of CD43 and firmly established CD43 as an anti-adhesive molecule that controls the homing of T lymphocytes by modulating the interaction between T cells and endothelial cells. The above conclusion was further supported by a study which showed that neutrophils of CD43 null mice have increased tethering and rolling on HEV just like T cells (Woodman et al., 1998). This group also showed another piece of unexpected data by demonstrating that infiltration of peritoneum by oyster glycogen-stimulated neutrophils and monocytes was severely reduced in CD43 null mice. These two conflicting data could be due to the differences of the mechanisms of different cell types or of different pathways (homing versus inflammatory infiltration). Nonetheless these studies showed that CD43 deficiency resulted in altered migratory behavior of leukocytes.  4.3.5 CD43 and ERM proteins Further evidence supporting the notion that CD43 functions as an adhesion molecule came from a study which showed that the CD43 cytoplasmic domain binds to ERM (Ezrin/Radixin/Moesin) proteins (Yonemura et al., 1998). ERM proteins interact with integral membrane proteins and actin filaments to mediate cell adhesion and migration. The authors used GST (Glutathione-S-Transferase) and CD43 cytoplasmic domain fusion protein to investigate association with ERM proteins and confirmed this interaction by coimmunoprecipitation and immunofluorescence. This showed that CD43 23  redistribution to the uropods of T cells was mediated by moesin and ezrin (Serrador et al., 1998). This study showed that CD43 might play an important role in the regulation of cell-cell integrations in lymphocyte trafficking.  4.4 CD43 as a signaling molecule 4.4.1 CD43 and leukocyte activation Crosslinking of CD43 with antibodies such as L10 can alter T cell (Axelsson et al., 1988) and monocyte (Nong et al., 1989) adhesiveness. This could imply that antibodies mask the anti-adhesive features of CD43 or alternatively that CD43 delivers a signal to other signaling molecules via the cytoplasmic tail. When L10 was added to T cell, PBMC (Peripheral Blood Mononuclear Cell) or monocyte cultures, phosphoinositide-derived second messengers, diacylglycerol (DAG) and inositol phosphates (IP), were hydrolyzed. This led to the translocation of protein kinase C (PKC) from cytosol to the plasma membrane and increased intracellular Ca2+ concentration (Silverman et al., 1989). Based on this study, CD43 appears to induce a classic signaling pathway of leukocyte activation. Especially, monocytes showed an increased hydrogen peroxide production due to CD43 signaling (Nong et al., 1989) demonstrating a functional aspect of the signaling pathway induced by CD43 ligation. Murine CD43 has also been shown to act as a T cell costimulator that functions independently of CD28 (Sperling et al., 1995) and crosslinking of CD43 has been shown to induce the maturation and activation of dendritic cell (Corinti et al., 1999) and natural killer cells (Nieto et al., 1999). Taken together, these studies illustrate that CD43 can function as a signaling molecule that induces activation in many different types of 24  leukocytes.  4.4.2 Mechanism of CD43 signaling The cytoplasmic tail of CD43 contains 22 serine and threonine residues that can potentially be phosphorylated and one SH3 binding domain. CD43 ligation has been shown to lead to tyrosine phosphorylation of Fyn and Vav in T lymphocytes through the SH3 binding domain (Pedraza-Alva et al., 1996; Pedraza-Alva et al., 1998) and these intracellular pathways seem to activate the MAP kinase pathway. CD43 ligation was also shown to induce Protein kinase C activation (Wong et al., 1990) resulting in tyrosine kinase activity of the src family kinases lyn and hck (Skubitz et al., 1998). Crosslinking of CD43 has further been shown to induce DNA binding activity of AP-1, NF-AT and NFκB (Santana et al., 2000) and nuclear translocation of ERK2 (Pedraza-Alva et al., 1998). The phosphorylation of serine residues on the cytoplasmic domain of CD43 has been shown to regulate the molecular associations and functions of the Cbl adapter protein which is known as a negative regulator of Ras (Pedraza-Alva et al., 2001). However, all theses in vitro experiments were done by ligating CD43 on primary cells from wild-type mice or cell lines. The biological significances of the signaling consequences are not well understood as exemplified by the lack of description of in vivo signaling defects in CD43 knockout mice.  4.4.3 CD43 and apoptosis Several reports showed a relationship between CD43 and apoptosis. For example, the human T cell line CEM lacking CD43 showed higher susceptibility to T lymphocyte25  mediated cytolysis in vitro than CEM cells expressing CD43 (McFarland et al., 1995). On the other hand, the overexpression of CD43 protected T cell hybridomas from TCR/CD3mediated apoptosis by inhibiting Fas-Fas ligand upregulation (He et al., 1999). The human T cell line Jukat showed a markedly increased apoptosis upon crosslinking with one specific anti-CD43 antibody while other anti-CD43 antibodies had no effect (Brown et al., 1996). The crosslinking also enhanced the DNA binding activities of the transcription factors, AP-1 and NFκB. When CD43 on human hematopoietic progenitor cells (CD34hi Lin-) was crosslinked in vitro with CD43 antibody MEM-59, the cells showed an increased apoptosis (Bazil et al., 1995). Interestingly, this phenomenon did not occur in more differentiated cells or in stem cells (Bazil et al., 1996). These data were interpreted to support that CD43 may affect actively dividing and differentiating cells but not the quiescent stem cells and mature cells and that CD43 can behave like the chemotherapeutic agent, 5-fluorouracil. This observed selective induction of apoptosis may be due to the fact that CD43 crosslinking might initiate different signaling pathways according to stage of differentiation and cell type. Based on the observation that CD43 crosslinking can cause cell proliferation (for example in T cells) or apoptosis (for example in progenitor cells), it appears that there is a window within the differentiation pathway from stem cells to mature cells, in which hematopoietic cells are sensitive to CD43 induced apoptosis and that CD43 might mediate a negative regulatory mechanism to control hematopoietic cell proliferation and possibly differentiation itself. The mechanism of CD43-mediated apoptosis was studied in more detail with TF1 cells, a myeloid progenitor-derived cell line (Cermak et al., 2002). TF-1 cells undergo 26  apoptosis when incubated with immobilized CD43 antibody MEM-59 as was observed with primary bone marrow cells. A Western blot showed that apoptosis induction by MEM-59 crosslinking induced translocation of Bad to the mitochondrial fraction implying that Bad, a pro-apoptotic protein was sequestered leading to apoptosis. Interestingly, CD43-mediated apoptosis can be inhibited by the overexpression of Daxx, an apoptotic regulator. Furthermore, they showed using the yeast two-hybrid system that the C-terminal part of Daxx interacts with the cytoplasmic tail of CD43 as well as with CD96 (Fas). It appears that Daxx, an apoptosis regulator, abrogates the apoptotic effects of CD43 ligation.  4.5 Possible ligands for CD43 At the present time it is widely accepted that there are no firmly established ligands for CD43. The molecules outlined below had been shown to interact with CD43, but the physiological relevance of interaction for many of them are questioned and require more detailed studies.  4.5.1 ICAM-1 Transfection of a murine T cell line with CD43 cDNA showed that CD43 expression on T cells enhances their response to antigen in the presence of TCR-CD3 expression (Park et al., 1991) suggesting that CD43 could function as a costimulatory molecule. The same lab later reported ICAM-1 as the ligand for CD43 based on evidence that anti-ICAM-1 antibody can block CD43 costimulatory effect (Rosenstein et al., 1991). They concluded that CD43 interferes with T cell adhesion by inhibiting the interaction of 27  LFA-1 with ICAM-1 as fewer T cells bind HeLa cells received CD43 cDNA than control HeLa cells (Ardman et al., 1992).  4.5.2 Influenza A virus A search for binding partners of influenza A viruses on human neutrophils (PMN) showed that viruses seemed to bind a protein with the molecular weights of 125 kDa and 160 kDa (Rothwell et al., 1994). Antibodies for CD43 recognized both of these bands on Western blots suggesting that influenza A viruses bind to CD43. However, it is not certain that CD43 is a docking protein for influenza A viruses in vivo since the study described above lacks critical data showing the direct binding of viruses to CD43 expressed on the cell surface of neutrophils.  4.5.3 MHC-I When T cells are activated through antigen ligation presented by APC (antigen presenting cells), a tight contact area is formed between the cells. MHC (Major Histocompatibility Class) bearing an antigen on APCs is recognized by TCR/CD3 complexes on T cells. A group found that CD43 interacts with MHC-1 molecule on DCs (Dendritic Cells) and B cell blasts by employing antibody inhibition assays of T cell-APC binding (Stockl et al., 1996). They found that the interaction between CD43 and MHC-1 can further induce a signaling pathway that leads to pro-adhesive activation of CD2mediated CD58 binding. However, the study failed to show an in vivo association of CD43 and MHC-1, a critical piece of data to claim that CD43 is a cognate ligand for MHC-1. 28  4.5.4 Sialoadhesin Sialoadhesin (Sn, Siglec-1) is a sialic acid receptor on macrophages which interacts with ligands on other leukocytes. A group reported that CD43 is a T cell counterrecptor for sialoadhesin by showing that sialoadhesin precipitates CD43 expressed on T lymphoma cell line (van den Berg et al., 2001). Since sialoadhesin is believed to function on local cell-cell interactions, CD43 might mediate the interaction between T cells and macrophages. However, neither in vivo interaction between CD43 and sialoadhesin nor any biological significance of such an interaction has been shown so far.  4.5.5 Galectin-1 Galectin-1 has been known to trigger the killing of immature thymocytes and activated peripheral T cells, and thus contribute to the modulation of immune repertoire and resolution of immune response, respectively. A group showed that galectin-1 can bind CD43 on T cells and that galectin-1 binding clusters CD43 (Hernandez et al., 2006). CD43 fusion proteins expressed in CEM cells were assayed for the binding with galectin1 and revealed that CD43 is required for optimal T cell susceptibility to galectin-1. It appeared that both core1 and core2 O-glycan branches of CD43 are recognized by galectin-1. This work showed that CD43 binding to galectin-1 along with selective binding to CD45 and CD7 might regulate T cell death.  4.5.6 E-selectin When E-selectin-IgG chimera was used to fish out possible ligands on Th1 cells, the 130 kDa glycoform of CD43 was obtained (Matsumoto et al., 2005). CHO cells 29  expressing E-selectin were then shown to roll on plates coated with CD43 under flow condition. Another group found that CD43 expresses the CLA (Cutaneous Lymphocyteassociated Antigen) epitope and that CLA on CD43 can serve as an E-selectin ligand (Fuhlbrigge et al., 2006). Memory T cells home or infiltrates skin by expressing CLA epitopes on PSGL-1 which serves as the ligands for P-selectin and E-selectin. CLA+ CD43 human T cells were shown to tether and roll on E-selectin coated blots in shear flow. However, Th1 cells of CD43 null mice had relatively minimal effects on the migration to non-inflamed and inflamed tissue in vivo compared to Th1 cells of PSGL-1 null mice (Matusumoto et al., 2007; Alcaide et al, 2007). Our laboratory also showed that neutrophils or activated T cells of CD43 null mice have no defect on migration to inflamed tissues in competitive in vivo assays (Carlow and Ziltener, 2006). More studies are required to determine whether CD43 is a physiological ligand for E-selectin.  4.6 Diseases and CD43 4.6.1 AIDS and CD43 There was a report that individuals with HIV-1 infection can make autoantibodies that bind to CD43 on thymocytes but not on mature circulating T cells because the epitope is blocked by sialic acids on mature T cells (Ardman et al., 1990). They postulated that the immunodeficiency caused by AIDS which is characterized by the increased destruction of lymphocytes might expose the autoimmunogenic epitope on CD43. Anti-CD43 autoantibodies are induced before CD4 cells are destroyed by HIV-1 viruses and these antibodies then can destroy healthy thymocytes independent of HIV-1 virus cytopathicity. 30  4.6.2 Cancer and CD43 In 1995, there was a first report that CD43 was expressed in cells other than hematopoietic cells. A group found that CD43 was expressed in the colon adenocarcinoma cell line, COLO 205 (Baeckstrom et al., 1995), and other colorectal adenomas tested subsequently (Sikut et al., 1997). In most cases, these cancer cells were shown to secret the CD43 extracellular domain with a novel glycosylation pattern which might explain why it has not been detected previously. These reports failed to show evidence for direct participation of CD43 in tumorigenesis, but suggested that CD43 might play a role in cancer possibly by providing anti-apoptotic signals. When CD43 was ectopically overexpressed in certain tumor cells, it was found that the tumor suppressor protein p53 was activated via the tumor suppressor protein ARF (Kadaja et al., 2004). In contrast, CD43 did not induce the expression of p53 when overexpressed in normal epithelial cells implying that the activation of p53 by CD43 is selective to tumor cells. The effects of CD43 on the expression of p53 appear to be at the translational or post-translational level and to be dependent on the presence of ARF which can stabilize p53.  4.6.3 Tuberculosis and CD43 It has been shown that infection of macrophages by mycobacteria (M. tuberculosis and Mycobacterium avium) involves CD43. Interestingly, the absence of CD43 reduces the binding of mycobacteria to macrophages which can be restored by the addition of the extracellular domain of CD43 directly purified from human plasma. On the other hand, the intracellular survival and replication of mycobacteria was significantly 31  enhanced in cells lacking CD43 (Fratazzi et al., 2000; Randhawa et al., 2005). These studies imply that CD43 might have an active role in pathogenesis of tuberculosis by modulating both cell entry of M. tuberculosis and intracellular proliferation of the bacteria.  5. Shedding of CD43 5.1. Shedding of CD43 CD43 shedding was first described when it was found that CD43 is downregulated from the surface of human neutrophils after TNF-α treatment (Campanero et al., 1991). The down-regulation could be blocked with several protease inhibitors implying that CD43 is down-regulated by proteolytic cleavage. Shortly after this discovery shed CD43 molecule from human neutrophils was visualized by Western blots (Rieu et al., 1992) and the identity of the sialoprotein discovered in human serum in 1980 was confirmed by amino acid sequencing to be shed CD43 molecules (Schmid et al., 1992). The amount of soluble CD43 was surprisingly high at about 10µg/ml. Shedding of human CD43 has been well established. CD43 on human leukocytes can be proteolytically cleaved from the cell surface by simply using TPA or immobilized anti-CD43 antibodies followed by FACS analysis, ELISA and Western blots (Bazil et al., 1993). The shedding process seems to involve multiple steps of proteolytic cleavage by a variety of metalloproteases, serine proteases and possibly other proteases (Bazil et al., 1994). And it seems that multiple proteolytic sites may exist resulting in multiple CD43 fragments and that there are at least two types of shedding processes, constitutive and inducible. Specifically in neutrophils, CD43 can be cleaved by neutrophil elastase which 32  can be induced by physiological stimuli (Remond-O’Donnel et al., 1995). It is still unknown why CD43 sheds. There are only two reports in the literature that address the biological significances of CD43 shedding. One report showed that albumin blocked CD43 shedding on human neutrophils via inhibition of sialidase and elastases (Nathan et al., 1993). Inhibition of shedding of this rigid and negatively charged molecule resulted in reduced neutrophil spreading on a glass coverslips indicating that CD43 shedding might regulate neutrophil spreading. The same group later showed that human neutrophils shed CD43 before adhering and migration (Lopez et al., 1998). Whether CD43 is actively participating in the spreading of neutrophils or not is still not known since the loss of function experiments have not been done using neutrophils from CD43 knockout mice.  5.2 Shedding 5.2.1 Shedding Proteolysis on the cell surface (shedding) is a crucial process for normal cellular functions because inappropriate shedding processes are associated with many diseases such as cancer, Alzheimer’s disease, and rheumatoid arthritis (Kiessling and Gordon, 1998). The cleavage of the ectodomains of transmembrane proteins on or close to the transmembrane domain is controlled by specific proteases which are called secretases or sheddases. Ectodomain shedding is involved in the regulatory function of many proteins such as TNFα, EGFR ligands, Fas, MHC-I, ErbB4, CD40, Delta, Amyloid precursor proteins, Notch, CD44, CD43, VCAM-1 and L-selectin. Shedding can have many different consequences including the release of membrane-tethered growth factors and 33  cytokines, the inactivation or activation of membrane-bound receptors, the release of the cytoplasmic portion of the proteins, and the modulation of cell adhesiveness and membrane structure. About 2-4% of all transmembrane proteins seem to make use of shedding to control their function (Blobel, 2000).  5.2.2 ADAM The ADAM (A Disintegrin And Metalloprotease) is a family of zinc-dependent membrane-associated metalloproteases, a protein family that is recognized as a key player in protein ectodomain shedding (Blobel et al., 2005). A typical ADAM family member has several protein domains; N-terminal signal sequence, pro-domain, metalloprotease domain, disintegrin domain, cysteine-rich region, EGF-like domain, transmembrane domain, and cytoplasmic domain. The pro-domain blocks catalytic activity of ADAM and the domain is removed by a furin-type pro-protein convertase or other proteases when activation is required. More than 10 ADAMs have been extensively studied and each of them has been found to have more than one substrate. For example, TACE (TNF-α Converting Enzyme), also called ADAM-17, is one of the most extensively characterized sheddases. TACE cleaves TNF-α, TGF-α, EGF, Notch receptors, TNF receptors, and β-amyloid precursor protein. TACE deficiency is embryonic lethal due to developmental defects. α-secretases that have been linked to processing of amyloid precursor protein include ADAM-9, -10 and -17. On the other hand, several ADAMs such as ADAM-2 and ADAM-15 are involved in adhesion through interactions between their disintegrin domains and integrins (Primakoff and Myles, 2000). Interestingly, several ADAM family members were shown to be highly upregulated in certain types of tumor 34  including colon, lung, prostate and breast cancer implying that ADAMs might regulate cancer cell migration and invasion (Iba et al., 1999).  5.2.3 APP Sheddases that process APP (Amyloid Precursor Protein) associated with Alzheimer’s disease have been comprehensively studied. The β-amyloid precursor protein is constitutively expressed on virtually all cell types indicating that APP itself is benign, but its physiological role in cells and tissues is not well understood. APP can be cleaved by either α- or β-secretase resulting in ectodomain release as a 90-100kDa secreted protein (sAPP) which is benign and is involved in neuroprotection and neuroplasticity as shown in Figure 1.3 (Selkoe, 1998). APP can also be cleaved by γsecretase after α- or β-secretase activity leading to the generation of a small fragment. The small piece generated by β- and γ-secretase activity is called Aβ peptide and its accumulation in so called amyloid plaques is believed to be the major cause of Alzheimer’s disease. For γ-secretase to be active, cleavage by α- or β-secretase has to occur first, implying that blockage of α- or β-secretase can inhibit γ-secretase. The β-secretase has been identified as an aspartyl protease called BACE (βAPP-cleaving enzyme), while the identification of γ-secretase has proven to be a difficult task. The γ-secretase seems to be associated with a macromolecular complex constituted of at least four proteins including presenilins (PSs), PEN-2, APH-1, and nicastrin. Several recent reports show that PSs might contain the active site of the γsecretase, but the functions of the other components remains elusive. (Steiner and Haass, 2000). 35  Figure 1.3 Processing of amyloid precursor protein (figure from Selkoe, 1998) Amyloid precursor protein (APP) can be digested either by α-secretase or βsecretase followed by γ-secretase (a). α- and γ-secretase combination yields ectodomain, p3 and C83 fragments (b) whereas β- and γ-secretase combination yields ectodomain, Aβ, and C99 fragments (c). Progressive accumulation of amyloid β-peptide (Aβ) is one of the major pathological causes of Alzheimer’s disease.  36  γ-secretases are a very unique class of proteases. They are water insoluble and are active in the lipid bilayer whereas all other known proteases are water soluble. γsecretase mediates a important cellular mechanism called RIP (Regulated intramembrane proteolysis) in which a γ secretase cleaves a α- or β-secretase cleaved product to release the cytoplasmic domain which then participates in a cellular reaction (Brown et al., 2000, Wolfe and Kopan, 2004). APP, Notch, CD44, E-cadherin, and ErbB-4 are the best known example of RIP. As with APP processing outlined above, the intramembrane cleavage does not occur unless the ectodomain is cleaved by a primary proteolytic cleavage. In general, RIP enables proteins to have dual functions, one as a receptor presented on the cell membrane and second as an intracellular soluble signaling molecule.  5.2.4 Notch Notch is a receptor protein that is extremely well conserved across animal kingdom. Vertebrates have four Notch receptors called Notch1-4. Notch receptors are involved in critical differentiation processes during embryonic and adult life (Mumm and Kopan, 2000). γ-secretase releases the cytoplasmic tail of the Notch receptor within the cell membrane after Notch binds its cognate ligand and the extracellular domain is cleaved by α-secretase TACE (Kopan and Goate, 2000) as shown in Figure 1.4. The released cytoplasmic tail translocates to the nucleus highlighting the importance of ligand-induced proteolytic cleavage for Notch signaling (Schroeter et al., 1998). The Notch cytoplasmic domain features a nuclear localization signal motif and several protein-protein interaction domains that are required for its function as a transcription factor involved in the regulation of many developmental processes (Fortini, 2001). 37  Figure 1.4 Notch synthesis and processing by proteolysis (figure from Fortini, 2001) Notch is expressed as a precursor polypeptide and cleaved by furin convertases in the trans-Golgi network, but the cleaved fragments are linked noncovalently to form a heterodimeric receptor. Once localized in the transmembrane, they can be recognized by ligands which induces proteolysis of the receptor by TACE (α-secretase) and presenilin (γ-secretase). The cleaved cytoplasmic domain migrates into the nucleus where it functions as transcriptional regulator. 38  5.2.5. CD44 CD44 is expressed on most cells types and is a major adhesion molecule for ECM (extracellular matrix) components. CD44 is involved in many physiological and pathological processes. CD44 can bind to HA (Hyaluronic acid), the major component of ECM, to mediate the adhesion of cells. For example, the shedding of CD44 plays a critical role in tumor migration (Okamoto et al., 1999). CD44 undergoes sequential proteolysis in the extracellular and transmembrane domains like other proteins subjected to RIP (Nagano and Saya, 2004; Ponta et al., 2004) as shown in Figure 1.5. The ectodomain shedding is mediated by membrane-associated metalloproteases to release Nterminal ectodomain followed by presenilin-dependent intramembranous cleavage which releases the cytoplasmic tail. The tail is translocated to the nucleus and involved in many transcriptional modulations (Okamoto et al., 2001).  6. Nuclear Localization 6.1 Nuclear localization of CD43 More recently, it was suggested that human CD43 might be a substrate of α- and γ-secretases in several human cell lines (Andersson et al., 2005). This raised a possibility that human CD43 might go though RIP–dependent signaling such as Notch-1 (Schroeter et al., 1998), APP (De Strooper et al., 2003) and CD44 (Murakami et al., 2003). Indeed, there was a follow-up study from the same group showing that the cytoplasmic domain of human CD43 might translocate to the nucleus when truncated cytoplasmic fragment of CD43 was expressed in CHO cells (Andersson et al., 2004). The group showed furthermore that the CD43 cytoplasmic domain might interact with β-catenin by 39  Figure 1.5 CD44 processing by proteolysis (figure from Nagano and Saya, 2004) Ca2+ influx, PKC activation or Ras can stimulate CD44 cleavage by matrix metalloproteases (MMPs) which releases soluble CD44. This has been reported tomodulate cell-cell adhesion. Also, this cleavage triggers the intramembranous proteolysis which releases the cytoplasmic domain by presenilin (γ-secretase). Then the cytoplasmic tail translocates to the nucleus and mediate various transcriptional activities.  40  immunoprecipitation and might affect its function such as c-myc and CyclinD1 expression. These two studies offered a unique view about CD43 function, but have a couple of issues. First, CD43 was expressed in a cell line that normally does not express CD43. Moreover, only the CD43 cytoplasmic domain fused with GFP was expressed leaving the question unanswered whether the cytoplasmic domain can be released from the full-length CD43 by proteolytic cleavage. The nuclear localization of the cytoplasmic tail of CD43 was a novel and quite surprising finding. Previously there was only one report that hinted that CD43 might be localized in other cellular regions than the cell membrane. Immunofluorescence staining for CD43 cytoplasmic tail revealed CD43 is distributed in the whole cytoplasm of colon carcinoma cells (Sikut et al., 1999). The intracellular localization of CD43 cytoplasmic domain was thus believed to be a special phenomenon associated with some cancer cells. A further possible link to the intracellular localization of CD43 came from the search for a serine/threonine kinase that interacts with the cytoplasmic domain of CD43 (Wang et al., 2000). They found a candidate and named it as STANK (Sialophorin Tail-Associated Nuclear Kinase). This is a protein homologous to a yeast kinase, YAK-1, which regulates the Ras pathway. Interestingly, STANK has been found to be localized in the both cytoplasm and nucleus.  6.2 Nuclear localization Eukaryotic cells are compartmentalized into multiple intracellular spaces that include the cytoplasm and nucleoplasm separated by the nuclear envelope. This physical separation necessitates the existence of a machinery which enables macromolecules to 41  move in and out of the nucleus. The NPC (nuclear pore complex) on the nuclear envelope controls this transport. There is a passive transport of molecules by diffusion in and out of the nucleus. The typical size exclusion limit of nuclear pore complex is about 40kDa. Evidence suggests that small molecules such as ions and small macromolecules up to 10kDa can freely diffuse through the NPC by a concentration gradient. Small macromolecules between 10-40kDa can still diffuse freely but there is a regulation by intracellular calcium levels. The transport of macromolecules larger than 40kDa is controlled by the nuclear transport machinery which recognizes and transports cargo (Lange et al., 2007). For the nuclear import machinery to function properly, it must be able to distinguish the cargo proteins from other cellular proteins. Typically proteins that are destined to the nucleus have a nuclear localization signals (NLS) even though there is evidence that alternative mechanisms could also exist (Stewart, 2007). NLS are usually bipartite consisting of two stretches of basic amino acids separated by about 10 amino acids. This signal is recognized by importin α as shown in Figure 1.6. The target protein associated to the importin α is carried into the nucleus through nuclear pores on nuclear envelope. Inside the nucleus, the complex binds Ran-GTP which induces importin α to lose its affinity for the target protein, thus releasing the cargo. Importin α-Ran-GTP then moves back to the cytoplasm where GAP (GTPase activating enzyme) hydrolyzes RanGTP into Ran-GDP. This hydrolysis modifies the conformation of Ran releasing importin α. The free importin α can bind the next target protein with NLS. Ran-GDP is recycled back into the nucleus where GEF (guanine exchange factor) regenerates Ran-GDP. Proteins that need to be exported from the nucleus into the cytoplasm have a 42  Figure 1.6 Mechanism of nuclear transport (figure from Stewert, 2007) Nuclear cargos are transported through nuclear pores by interactions between cargo proteins and transport machineries, importins and exportins. A protein destined for the nucleus and/or cytoplasm usually have a specific sequence called nuclear localization signal (NLS) which are directly or indirectly recognized by importins and exportins.  43  nuclear export signal (NES) which is made up of 5-6 hydrophobic amino acids. Much like the mechanism of nuclear import, an exportin is bound by a Ran-GTP causing a conformational change. The structural change results in the increased affinity toward a NES. Once the target protein is recognized and associated with exportin-Ran-GTP, they are exported to the cytoplasm where a GAP hydrolyzes Ran-GTP into Ran-GDP. This dissociates the exportin from Ran-GDP. Without the Ran-GDP, the exportin cannot hold the cargo protein so that it is released. Then the exportin and Ran-GDP are recycled back into the nucleus separately where a GEP changes Ran-GDP back to Ran-GTP.  7. Sumoylation 7.1 SUMO While investigating CD43 for functional domains, we discovered that CD43 has three potential sumoylation sites on the cytoplasmic domain as explained in the Result section. Sumolyation is a posttranslational modification process in which 11.5kDa SUMO (Small Ubiquitin like MOdifier) peptides are covalently attached to lysine residues of target proteins (Gill, 2004; Johnson, 2004; Muller et al., 2001). Ubiquitins and ubiquitin-like proteins such as SUMO covalently modify their target proteins similar to other modifiers such as phosphates and acetyl groups, but their function and control is far more complicated since they involve quite large peptides with complex structures. SUMO-1 has only 18% homology to ubiquitin, but the three dimensional structures of both entities are almost identical implying that their functions may be related (Bayer et al., 1998). However, unlike ubiquitination, SUMO modification does not label proteins for degradation. Instead, SUMOs have been found to alter stability, structure, activity and 44  localization of modified target proteins (Seeler and DeJean, 2003). In general they are involved in nuclear processes such as chromatin organization, transcription, RNA metabolism, nuclear body regulation, nuclear pore complex regulation. Also, unlike ubiquitins, SUMOs appear to function as a monomer in vivo even though they have a capacity to form polymers in vitro (Johnson et al., 2001).  7.2 Mechanisms of sumoylation In vertebrates, there are four SUMO paralogs called SUMO-1, -2, -3, and recently found -4. SUMO-2 and -3 differ only in three amino acids and functional differences if any are yet to be defined. Therefore, the SUMO family is generally divided into three groups, SUMO-1, SUMO-2/-3, and SUMO-4. Each SUMO has its own targets, but they also share some common target proteins. The whole process of sumoylation and desumoylation is a reversible process designed to modify protein function dynamically. The process of sumoylation is much like ubiquitination (Kim et al., 2002). SUMO is synthesized as an inactive precursor protein and processed to a mature protein by a hydrolase called SENP (Figure 1.7). Mature SUMO is conjugated to Activating Enzyme E1 by forming thioester bonds between the C-terminal carboxyl group of SUMO and the sulphydryl group of a cysteine residue of E1. Subsequently the SUMO is transferred and conjugated to a cysteine residue of Conjugating Enzyme E2. The Conjugating Enzyme E2 recognizes the substrate protein and transfers the SUMO onto the ε-amino group of lysine residue of the substrate with the help of Ligating Enzyme E3. There are several unique features where SUMO differentiates itself from ubiquitin 45  Figure 1.7 Mechanism of sumoylation (figure from Gill, 2004) SUMO (Small Ubiquitin-related Modifier) proteins are a family of small proteins (3 members in vertebrates) which is attached and detached from a target protein to modify the function and localization. It is involved in protein stability, transcriptional regulation, nuclear-cytosolic transport, apoptosis, cell cycle etc. A SUMO is usually about 100 amino acids with 12-15 kDa molecular weights. A SUMO precursor protein is cleaved by SENP to become a mature protein, then cojugated on E1. E1 transfers the SUMO to E2 which recognizes the target protein and finally transfers the SUMO to a target protein with the help of E3.  46  and ubiquitin-like proteins. First, SUMO Conjugating Enzyme E2 can directly recognize its target whereas the recognition of the target is the function of Ligating Enzyme E3 in the case of ubiquitin and ubiquitin-like proteins. A typical consensus motif of SUMO modification is ψKχE where ψ is a large hydrophobic residue and χ is any residue (Rodriguez et al., 2001).  Secondly, ubiquitin and ubiquitin-like proteins of typical  mammalian cells have a single E1 Activating Enzyme, 30 different E2 Conjugating Enzymes, and hundreds of different E3 ligating Enzymes that contribute to substrate specificity. Sumoylation utilizes only a single E2, Ubc9 in mammals and a limited number of E3 enzymes are known. The desumoylation reaction is catalyzed by the Ulp cysteine proteases that do not share homology with deubiquitination enzymes. Ulp1 which has been found in yeast is an essential protein for cell proliferation (Li et al., 1999) and it has four homologs in mammals, SENP1, SENP2, SENP3 and SENP6. SENP1, 2, and 3 are localized in and around the nucleus and SENP6 is localized in the cytoplasm. They all have been shown to have SUMO-specific protease activities and they have preferences for their target SUMOs in vivo.  7.3 Function of SUMO Ubc9, an E2 Conjugating Enzyme, is the key of sumoylation machinery and there are profound defects in Ubc9 deficient mice. Ubc9 deficiency is embryonic lethal. Embryos survive long enough to implant on the wall of uterus with the help of maternal proteins, but they fail to thrive after implantation (Nacerddine et al., 2005). The group who developed the knockout mouse used a special technique called blastocyst outgrowth 47  culture to simulate the early phase of postimplantation and found that there are severe defects in nuclear structure and function including chromosome condensation and segregation, subnuclear structure and nuclear morphology. Since most of the SUMO substrates have a NLS, it was assumed that sumoylation occurs in the nucleus (Seufert et al., 1995). Later it was found that the E1 is predominantly found inside the nucleus and the E2 and E3 are primarily found on the cytoplasmic side of the nuclear pore complex (Pichler et al., 2002). Therefore, it is now generally considered that sumoylation occurs en route to the nucleus. In some rare cases, SUMOs can modify cytoplasmic proteins because E2 and E3 are facing the cytoplasm. The existence of the SUMO has been discovered from the modification of RanGAP by SUMO-1 (Matunis et al., 1996). Since then many new SUMO targets have been found. Some of the well studied substrates are shown in Figure 1.8. SUMO has been first known for its function in regulating subcellular localization of target proteins because its first known target was Ran-GAP1. Ran, a nuclear Ras-like GTPase, is involved in the bidirectional transport of proteins in and out of the nucleus by interacting with importin α, and the function of Ran is regulated by Ran-GAP (Ran-GTP activating protein). 70kDa unmodified Ran-GAP is found in the cytoplasm and 90kDa SUMO-modified protein is found on the cytoplasmic side of NPC (nuclear pore complex). This finding suggests that sumoylation is required for Ran to be translocalized to the nucleus. Many of the SUMO targets identified so far are transcription factors, coactivators or corepressors implying that SUMO is actively involved in transcriptional regulation. It appears that SUMO modifications generally repress the activities of transcriptional regulator proteins. For example, SP3, a ubiquitous transcription factor, becomes 48  Figure 1.8 Targets for SUMO (figure from Gill, 2004) Most of SUMO targets contain a consensus sequence of tetrapeptide and substrate specificity appears to be derived from E1 enzyme (Ubc9 in mammals). Some examples from a long and growing list of SUMO target proteins are shown here. They are involved in various functions of target proteins (most known targets are nuclear proteins).  49  inactivated after SUMO modification even though modified SP3 can still bind DNA with the same specificity and affinity (Verger et al., 2003). Interestingly, SUMO also may function in a context-dependent manner (Holmstrom et al., 2003). The multiple attachments of SUMO molecules on a transcriptional regulator along with other posttranslational modifications can thus dynamically change the function of the regulator and the sum of all these modification decides the fate of the target protein. SUMO has been shown to influence the genome structure and maintenance. For example, histone H4 has been shown to be modified by SUMO (Shiio et al., 2003). Like other kinds of histone modifications, SUMO alone on H4 does not seem to modify the structure or the packing activity of histones, but promotes transcriptional repression. This implies that SUMO can add one more layer of complexity to histone regulation which is already one of the most complicated regulation systems, as histones can be phosphorylated, acetylated, methylated and ubiquitinated at the same time. It is believed that the SUMO on H4 adds one more grammatical rule for the so called “Histone Language”. Another example of genome maintenance by SUMO can be found from TDG (Thymine DNA Glycosylase) which restores mismatched G-C base pairs during base excision repair, the major DNA repair system (Hardeland et al., 2002). SUMO modification on TDG reduces the DNA binding affinity and increases the turnover of the enzyme, thus negatively regulating TDG. SUMO has also been shown to act as a regulator of signaling. For example, SUMO modification on NEMO, an IκB kinase regulator, mediates NFκB activation in response to genotoxic stress (Huang et al., 2003). Genotoxic stress induced by DNA damage causes nuclear localization of SUMO-modified NEMO that then becomes 50  ubiquitinated. This activates the NFκB survival pathway in the cytoplasm. This shows that SUMO modification can also respond to external stimuli. Phosphorylation has been known to regulate sumoylation itself. For example, phosphorylation of P53 and PML inhibits SUMO modification on these target proteins (Muller et al., 1998), whereas phosphorylation has been shown to increase sumoylation of HSF1 (Heat Shock Factor 1) (Hietakangas et al., 2003). In summary, sumoylation offers another way to regulate protein function along with other modifications such as phosphorylation. Since sumoylation targets mostly nuclear proteins and regulates nuclear activities, sumoylation can have a powerful impact on the survival of cells and organisms.  7.4 PML nuclear body Many SUMO-modified proteins and the SUMO-conjugating machinery are recruited into subnuclear domains, known as the PML (ProMyelocytic Leukemia) nuclear body, also known as POD (Presumably Oncogenic Domain), ND 10 (Nuclear domain 10) or Krüppel bodies (Borden et al., 2002; Bernardi and Pandolfi, 2003; Ching et al., 2005). Sumoylation is required for PML nuclear body formation. PML nuclear body speckles are around 0.3-0.5µm in size and mammalian cells typically have 10-30 PML nuclear bodies per nucleus as shown in Figure 1.9. As cells enter mitosis, the number of PML nuclear body drops to a few by aggregation which correlates with the loss of nuclear envelope integrity. PML bodies are made up of many proteins, but the major constituent is sumoylated PML. PML is a nuclear protein and is a member of the ring-finger proteins 51  Figure 1.9 Immunofluorescence staining of PML nuclear bodies (figure from Lallemand-Breitenbach et al., 2001) PML (PreMyelocytic Leukemia) proteins are typically concentrated in subnuclear structure called PML nuclear bodies which are a nuclear matrix-associated structure with 250-500 nm in size. There are about 10 PML nuclear bodies per nucleus, but this can vary considerably depending on the cell type, cell cycle and other factors. PML Nuclear bodies are quite immobile in interphase but become very dynamic after stresses such as heat and irradiation. Typically stresses break down PML nuclear bodies which are also linked to chromatin organization. Immortalized mouse embryo fibroblasts from PML-/- mouse were transfected with a plasmid expressing wild-type PML and stained with anti-PML antibody. Two nuclei are shown to visualize speckles of PML nuclear bodies.  52  containing a distinctive zinc finger domain termed RING. Ring-finger proteins are a family of proteins which have cysteines and histidines which form metal binding sites. PML was shown to fuse with RAR (Retinoic Acid Receptor) in acute promyelocytic leukemia (APL) (Goddard et al, 1991). APL patients with PML-RAR fusion proteins in promyelocytes have PML bodies that are disrupted and are visible as hundreds of tiny dots scattered all over the APL cells. Moreover the expression of this fusion protein was sufficient for transformation of APL cells and the development of leukemia (Altabef et al., 1996). This initial observation suggested that PML nuclear bodies may play a role in cell growth and differentiation. PML -/- mice were shown to be healthy and fertile other than being leukopenic and susceptible to spontaneous Botryomycotic infections (Wang et al., 1998 (1)). Interestingly, cultured cells of PML -/- mice showed an accelerated growth suggesting that PML might be a negative growth regulator. Detailed studies showed that PML regulates hemopoietic differentiation and controls cell growth and tumorigenesis. Also, PML was indispensible for the tumor-growth-suppressive activity of retinoic acid. Most importantly PML has an active role in apoptosis. PML was found to be essential for Fas- and caspase-dependent DNA-damage-induced apoptosis (Wang et al., 1998 (2)). The study also found that PML is essential for induction of programmed cell death by Fas, TNFα, ceramide and IFNα, β, and γ. These features made PML  -/-  mice and cultured cells to be resistant to the lethal  effects of anti-Fas antibody and irradiation. PML is the major constituent of PML nuclear bodies which are aggregates of heterogeneous proteins associated with transcription factors, chromatin modifiers, genome maintenance proteins etc. Therefore, PML nuclear bodies are implicated in a 53  variety of nuclear functions (Hofmann and Will, 2003) including genome stability (Zhong et al., 1999), transcriptional regulation (Zhong et al., 2000), apoptosis (Salomoni et al., 2002). However, the precise function and the mechanism of the regulation of PML nuclear bodies remain enigmatic. The best known example is the regulation of Daxx, a transcriptional corepressor. Upon sumoylation, Daxx is localized to PML nuclear bodies where it remains inactive. Thus PML nuclear body works as a repressor of Daxx in this example. Other proteins localized in PML nuclear bodies that have been studied in detail are shown in Table 2. Currently there are three different hypotheses that attempt to explain the regulatory mechanisms of PML nuclear bodies. The first one hypothesizes that PML nuclear bodies are storage facilities where excess nucleoplasmic proteins are stored until they are needed. In the second model, PML nuclear bodies are thought to be sites of posttranslational modification and degradation of proteins. The third model predicts that PML nuclear bodies are sites where specific nuclear activities take place such as DNA replication and transcriptional regulation. There is evidence supporting each of these three hypotheses to some degree, and it is presently not clear which ones are true. One interesting aspect about this question is whether there are specific cellular processes that only take place in PML nuclear bodies. PML nuclear bodies have a stable structure and are relatively immobile compared to other nuclear bodies like MAD (Matrix-Associated Deacetylase) (Wiesmeijer 2002). Of course, PML nuclear bodies can be mobile or disrupted under the conditions of stress, apoptosis, and transcriptional repression (Eskiw 2004). This supports the third hypothesis.  54  Table 2. Proteins that are known to be localized in PML nuclear bodies (table from Ching et al., 2005) PML protein itself is sumoylated and becomes the major constituent of PML nuclear bodies. PML nuclear bodies harbors other proteins most of which are also sumoylated. Identification of these proteins associated with PML nuclear bodies has helped to determine the function of this subnuclear structure. However, more than 60 proteins are known implicating PML nuclear bodies are probably involved in virtually every aspect of nuclear activities.  55  7. Thesis objectives The overall objective of the work presented in this thesis is the functional analysis of CD43 shedding. There has been much debate about the function of CD43, but the definitive function of CD43 still remains elusive. Since CD43 has a negatively charged and bulky extracellular mucin domain, it was suggested that CD43 might function as an antiadhesive molecule. However, there are no known physiological ligands and it does not seem to influence leukocyte migration in vivo. Nonetheless, the increased homotypic adhesion that can always be observed in all CD43 deficient leukocytes still leaves a possibility that CD43 might function as a physical barrier even though CD43 deficient mice do not have known phenotypes related to this phenomenon. Another suggested function of CD43 is a signaling through the cytoplasmic domain. There have been numerous reports showing the interaction between CD43 cytoplasmic domain and various intracellular signaling molecules. However, these reports failed to elucidate the function of CD43 because it was not known what signaling defects CD43 null mice have. One prominent feature of CD43 metabolism is the ectodomain shedding. We rationalized that by studying CD43 shedding we might be able to determine the function of CD43 because a number of molecules regulate their functions by controlling shedding. We specifically tried to find out (a) whether CD43 sheds from the cell surface of murine cells, (b) whether the inhibition of CD43 shedding has an impact on cell biology and (c) whether CD43 shedding has a physiological role.  56  CHAPTER 2 MATERIALS AND METHODS Mice and cell culture CD43-/- mice were generated on a 129 background as previously described (Manjunath et al., 1995). B6.CD43-/- mice were obtained from Dr. Anne Sperling (University of Chicago, Chicago, IL) backcrossed onto C57BL/6 for six generations and further  backcrossed to F9. Mice were kept at a pathogen-free animal facility of  Biomedical Research Centre. Mice were sacrificed with CO2 inhalation and bone marrow cells were harvested from the femur and tibia by flushing them with 10ml of sterile PBS and an 18-gauge needle. After centrifugation, the cell pellet was incubated with the red blood cell lysis buffer at room temperature for 5 minutes and washed with PBS twice. The final cell pellet was resuspended in RPMI. All cell cultures were maintained in either RPMI or DMEM supplemented with 10% FCS, 2mM L-glutamine, 100 U/ml each of penicillin and streptomycin, and 50 µM 2-mercaptoethanol at 37ºC in 5% CO2. To enrich granulocytes, the culture of bone marrow cells were supplemented with 16 U/ml IL3 and 100 ng/ml G-CSF 3-6 days. IL3 was obtained from WEHI-3B conditioned media. To mature macrophage lineage cells, the culture of bone marrow cells were supplemented with 1000 U/ml M-CSF and 50 ng/ml SCF for 6-8 days. To mature mast cells, the culture of bone marrow cells was supplemented with 16 U/ml IL3 and 50 ng/ml SCF for 4-6 weeks. To obtain T cells, 5X106 splenocytes were incubated in 5ml culture medium with 5 µg/ml concanavalin A (ConA) for 2 days and further incubated with 10 U/ml IL2 for 4-12 days. Mouse cell lines MC/9, CTLL-2, NSF-60 and WEHI274.3 were maintained in RPMI with either IL2 57  or IL3.  Antibodies S11, rat anti-pan-mouse CD43 antibody (isotype IgG2b), was purified in house from the hybridoma cell line (kindly provided by Dr. J. Kemp, Department of Pathology, University of Iowa, Iowa City, IA). H18, rabbit polyclonal anti-mouse CD43 antibody, was raised against a peptide corresponding to the C-terminus of CD43 and purified in house. M2, anti-Flag antibody, was purchased from Sigma. Anti-SUMO-1 (anti-GMP-1) antibody and anti-PML antibodies were purchased from BD Biosciences and all other secondary antibodies were purchased from BD Biosciences or Invitrogen.  Construction of chimeric and fusion proteins The retroviral vectors encoding for the wild-type CD43 and CD43/34 chimeras were constructed by subcloning them into pMX-pie-FLAG (Drew et al., 2005, see Appendix I) which is basically a MSCV retrovirus containing CD43 5’UTR, CD43 signal peptide, FLAG epitope, 2X glycine spacer, and multiple cloning sites followed by IRESeGFP (ordered as in the vector). CD43 was amplified with primers 5’ccatcgatgacagtctgcagaggacgacga and 3’-ccatcgattagagattgaggtgcggcctcatc from mouse spleen cDNA and cloned into the vector in ClaI and EcoRI. CD43 Mutant MP was constructed by swapping 180-229 amino acids of membrane proximal region of CD43 with 130-252 amino acids of membrane proximal globular region of CD34. CD43 Mutant TM was constructed by swapping the 23 amino acid transmembrane region of CD43 with the 21 amino acid transmembrane region of CD34. These two mutants were made by 58  recombinatorial PCR technique to avoid introducing foreign amino acid sequences derived from restriction sites. GFP-fusion of CD43 was made by deleting the stop codon of CD43, 3’UTR of CD43 and IRES sequences between the CD43 gene and eGFP.  Retroviral infection The retroviral vectors encoding for the wild-type CD43, CD43/34 chimeras and GFP-fusion proteins were purified from bacteria. To produce the retroviruses, BOSC cells, a retrovirus-producing cell line, were transiently transfected with 1µg of the retroviral vector by Lipofectamine (Invitrogen) according to the manufacturer’s manual. Fortyeight hours after transfection, viral supernatants were harvested, filtered through a 0.45 µm membrane and applied to target cells in 6-well dishes with 5 µg/ml polybrene (Sigma). 24 hours after infection with retrovirus, cells were washed and grown further in fresh medium. In some cases, target cells were cocultured on ψ-2 cells, a retrovirusproducing cell line that have been transfected with plasmids and selected for at least 4 weeks. A typical coculture was done with 5 µg/ml polybrene for 3 days.  Flowcytometry (FACS analysis) Flow cytometric sorting was performed on a FACS Vantage SE high speed cell sorter (Becton Dickinson). Flow cytometric analyses were performed on FACS Calibur or FACS Scan instruments using Cellquest software (Becton Dickinson) or Flowjo (Tree Star). Cells were washed twice with FACS staining buffer (PBS containing 0.5% FCS and 0.02% sodium azide), then resuspended in FACS staining buffer at 106 cells/ml. Antibodies were added at optimized concentrations and incubated on ice for 10 minutes. 59  Samples were washed twice with FACS staining buffer before analyzed.  SDS-PAGE, immunoblotting and immunoprecipitation For total lysates, cells were lysed in a non-ionic detergent buffer (0.5% TritonX100, 25 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM dithiothreitol, 1 mM EDTA and protease inhibitors) on ice for 20 minutes. Lysates were sonicated to break up insoluble matter including nucleus. For subcellular fractionation, cells were washed once with cold PBS and resuspended in cold hypotonic lysate buffer (10 mM HEPES pH 7.6, 1 mM EDTA, 20 mM NaCl, 1 mM dithiothreitol and proteinase inhibitors). After incubation on ice for 10 minutes, cells were lysed by 8 to 10 passages through a 26-gauge syringe needle. Nuclei were pelleted by centrifugation at 14,000 g at 4°C for 1 minute, the supernatant was recovered as the cellular protein fraction. Proteins were extracted from nuclei by incubation with nuclear extraction buffer (10 mM HEPES pH 7.6, 1 mM EDTA, 500 mM NaCl, 25% glycerol, 1 mM DTT and proteinase inhibitors) at 4°C for 20 minutes with vigorous shaking. Nuclear debris was pelleted by centrifugation at 14,000 g at 4°C for 5 minutes and the supernatant was collected as the nuclear protein fraction. SDS-PAGE (10 or 14%) was performed according to the conventional Laemmli method. Extracts were mixed with an equal volume of 2X protein sample buffer and heated to 100°C for 3 minutes, and loaded on precast 10 or 14% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) gels. Gels were run with MiniPROTEAN (Biorad) at constant 30 mA for about 1 hour. For immunoblotting, proteins were electrophoretically transferred from gels onto 60  nitrocellulose membranes (Amersham) by using Trans-Blot SD (Biorad) at constant 250 mA for about 2 hours. After blocking with 5% skim milk in TBS (20 mM Tris pH7.6 and 137 mM NaCl) at room temperature for 2 hours, the membranes were washed in TBS and incubated with a primary antibody with 1% skim milk and 0.1% Tween20 in TBS at room temperature for 2 hours. Then the membrane was washed extensively with TBS. The bound antibodies were visualized using species-specific secondary antibody conjugated with alkaline phosphatase or biotinylated primary antibody followed by avidinconjugated alkaline phosphatase by incubating the membranes with 1% skim milk and 0.01% Tween20 at room temperature for 1 hour. The membrane was washed extensively with TBS. Blots were developed using the enhanced chemiluminescence (ECL) detection system (Amersham). For immunoprecipitation, the soluble fraction was incubated with a primary antibody (5 µg per 10ml) at 4°C for 1 hour. Thereafter, Protein G–Sepharose previously blocked with 2% BSA in PBS was added and the samples were further incubated in PBS at 4°C for 90 minutes. The immune complexes were pelleted, washed 3 times with PBS and suspended in reducing SDS sample buffer and boiled for 3 minutes.  Sandwich ELISA Culture media were harvested 48 h following transfection of target cells and frozen at -80°C until further use. 1 μg/ml anti-Flag antibody M2 (Sigma) in PBS was immobilized onto ELISA plates at 4°C for 2 hours. After washing the plate in PBS with 2% skim milk, the culture media were incubated at room temperature for 2 hours. After washing in PBS with 2% skim milk, the plates were incubated with 1 μg/ml biotinylated 61  S11 or H18 in PBS with 2% skim milk at room temperature for 1 hour. The plates were washed vigorously in PBS with 2% skim milk and incubated with secondary antibodies conjugated to horse radish peroxidase in PBS with 2% skim milk at room temperature for 1 hour. The plates were washed vigorously in PBS with 2% skim milk and finally with distilled water. The color reaction for ELISA was developed by adding 100 µL of 1.1 mM ABTS (2,2'-azino-bis(3)-ethylbenzthiazoline-6-sulfonic acid (Sigma)) in 0.1 M citratephosphate buffer (pH 4.2) containing 0.01% H2O2 at room temperature for 10 to 20 minutes. Assays were run in triplicate, and samples were read at 450 nm in an ELISA plate reader (Molecular Devices).  Shedding assay To induce shedding, PMA was added to cell culture at 50 ng/ml for 2 hours. To induce shedding with immobilized antibodies, the cell cultures were transferred into the antibody coated plates that were coated with antibodies at 1 μg/ml in PBS at 4°C overnight.  Immunohistology with confocal and fluorescence microscopes 106 Cells were fixed and permeabilized at the same time in 200 µl of 4% paraformaldehyde and 0.4% Triton X-100 in PBS on ice at least 4 hours. Alternatively, cells were fixed and permeabilized in -20°C methanol for 15 minutes and cytospun on slides for staining. Cells were washed in PBS three times and blocked in 2% BSA and 5% skim milk in PBS on ice for 1 hour. Primary antibodies were directly added to the cells and further incubated for at least 2 hours. Cells were washed in PBS three times and 62  incubated with secondary antibodies in 2% BSA and 5% skim milk in PBS on ice for at least 2 hours. Cells were washed in PBS three times and resuspended in 10 µl PBS. After mixing with 10 µl Fluoromount (Southern Biotech), samples were mounted on slide glasses that were coated with 1% poly-L lysine and examined on Nikon Confocal microscope C1 with EZ-C1 imaging software. Fluorescent cells were directly mounted on slide glasses that were coated with 1% poly-L lysine and examined on Zeiss Axioplan 2 microscope with Openlab imaging software.  Cell purification Cells were enriched from bone marrows or spleens by positive or negative selection on autoMACS (Miltenyi Biotec) by staining with GR-1, B220, CD4 or CD8 antibodies according to the manufacturer’s instruction.  63  CHAPTER 3 RESULTS 1. Shedding of CD43 1.1 Down-regulation of CD43 1.1.1 CD43 is down-regulated from cell surface Work from several laboratories has shown that CD43 expressed on human cells is shed. To establish whether down-regulation of CD43 can also be observed on murine cells, we analyzed, in a first approach, CD43 down-regulation on several mouse cell lines. We selected 4 types of mouse cell lines that represent different hematopoietic lineages: MC/9 (mast cell line), CTLL2 (cytotoxic T cell line), NSF-60 (myelocytic cell line) and WEHI-274.3 (myelomonocytic cell line). Cells were treated with the phorbol ester TPA, a protocol shown to induce CD43 down-regulation on human cells. The cells were then subjected to FACS analysis using CD43 mAb S11 which binds the extracellular domain of CD43 independently of carbohydrate modification. As shown in Figure 3.1.1, CTLL2 and NSF-60 exhibited a significant down-regulation of CD43 upon activation by TPA, whereas MC/9 and WEHI-274.3 did not, indicating that TPA treatment can induce CD43 down-regulation on CTLL2 and NSF-60 but not on MC/9 and WEHI-274.3. Our data do not exclude the possibility that stimuli other than TPA can also induce CD43 downregulation and that these can lead to CD43 down-regulation on cell lines that did not shed CD43 by TPA. Because cell lines can often not accurately represent primary cell lineages, we repeated the CD43 down-regulation experiments using primary mouse cells, splenocytes and neutrophils obtained from peritoneal cavity. Cells were treated with TPA and then analyzed by FACS for CD43 and lineage markers, CD4+ and CD8+ for T cells or Gr-1+ 64  Figure 3.1.1 Down-regulation of CD43 on mouse cell lines MC/9, CTLL2, NSF-60 and WEHI-274.3 cell lines were activated by 10 nM TPA for 3 hours. After incubation, cells were stained with biotinylated mAb S11, anti-CD43 antibody, and analyzed by FACS. MC/9 and WEHI-274.3 do not appear to down-regulate CD43, while CTLL2 and NSF-60 do. This is a representative of several experiments.  65  for neutrophils. Splenic T cells did not down-regulate CD43 in response to TPA treatment whereas neutrophils down-regulated CD43 after TPA treatment by 20-30% (data not shown). Because primary cells might respond with different sensitivity to different activating reagents, we also used stimuli more resembling physiological ligands such as ConA for T cells and oyster glycogen for neutrophils. As shown in Figure 3.1.2, both T cells and neutrophils exhibited a significant down-regulation of CD43 after simulation, indicating that CD43 on mouse primary cells can be down-regulated upon stimulation other than a phorbol ester. Taken together these results show that CD43 down-regulation on murine leukocytes can be induced.  1.1.2 Inducers of CD43 down-regulation To investigate agonists capable of inducing down-regulation of CD43 in more detail and to obtain a sense whether there are physiological ligands, we looked into down-regulation of CD43 on T cells upon activation with variety of stimuli. We used reagents known to stimulate T cells that result in various outcomes affecting T cell biology including immobilized anti-IL2 receptor, P-selectin-hIgG chimera, anti-PSGL-1 mAb, anti-LFA-1 mAb, soluble ConA and IL2. Figure 3.1.3 illustrates the results of FACS analysis showing that these reagents can induce down-regulation of CD43 to different degrees. The most pronounced CD43 down-regulation was produced by IL2, an autocrine cytokine that stimulates T cell proliferation and by immobilized LFA-1 antibody. Why some of these stimuli can induce CD43 down-regulation and others not is unclear. Our data show that not all known agonists can induce down-regulation of CD43 suggesting that CD43 down-regulation is a selective process. 66  Figure 3.1.2 Down-regulation of CD43 on mouse primary cells Splenocytes (a) and neutrophils from peritoneal lavage (b) were activated by 4 ug/ml ConA for 24 hours and 1% glycogen for 3 hours, respectively. The cells were stained with biotinylated mAb S11, anti-CD43 antibody, and analyzed by FACS. ConA and glycogen can induce down-regulation of CD43 on splenocytes and neutrophils, respectively. This is a representative of several experiments.  67  Figure 3.1.3 Down-regulation of CD43 on mouse primary T cells Splenocytes were treated by various stimuli for 4 hours. ConA was added at 4 ug/ml and IL2 was added at 10 U/ml. The cells were stained with biotinylated mAb S11, anti-CD43 antibody, and T cell markers (CD4 and CD8). Samples were analyzed by FACS after gating on T cell (CD4+ or CD8+). Only IL2 and immobilized anti-LFA mAb treatment induced a significant down-regulation of CD43. This is a representative of several experiments.  68  1.1.3 CD43 down-regulation from different cell types One of the established methods to stimulate shedding of a given surface molecule is to stimulate cells by incubating them on plates that have been pre-coated with antibodies capable of stimulating the particular cell type. Human CD43 expressed on granulocytes has been shown to be down-regulated by several immobilized anti-CD43 antibodies. CD45 is often used as a negative control since CD45 is a non-sheddable molecule or as a positive control since CD45 can transmit signals that stimulate cells which could result in the down-regulation of other molecules. We incubated granulocytes (Gr-1+ population of total mouse bone marrow cells) on anti-CD43 mAb S11 and antiCD45 antibody 2D1 coated plates. As shown in Figure 3.1.4, CD43 expressed on GR-1+ bone marrow cells was down-regulated by an immobilized anti-CD43 antibody while the expression of CD45 was affected neither by immobilized anti-CD45 antibody nor antiCD43 antibody. This clearly shows that murine CD43 can be down-regulated by immobilized anti-CD43 antibody. We found that bone marrow cells (day 3 culture in IL3) also showed CD43 down-regulation by immobilized anti-CD43 antibodies. Interestingly CD43 down-regulation on these cells can also be stimulated by immobilized anti-CD45 antibody. This highlights that the mechanism that controls CD43 down-regulation can differ depending on the differentiation stage of cells since the same immobilized antiCD45 antibody could induce CD43 down-regulation on GR-1+ cells of day 3 bone marrow cultures but not of day 0 bone marrow cultures. As expected the expression of CD45 remained constant in all of these experiments (data not shown). Taken together the data suggest that down-regulation of CD43 can vary on the stages of cell differentiation as well as cell types. 69  Figure 3.1.4 Down-regulation of CD43 of Gr-1+ cells Total bone marrow cells (BMCs) that were freshly prepared or that were grown in granulocyte enriching medium (16 U/ml IL3 and 100 ng/ml G-CSF) for 3 days were incubated on plates coated with mAb S11 or anti-CD45 antibody as a control. The cells were stained with biotinylated mAb S11, anti-CD43 antibody, and biotinylated GR-1, a granulocyte surface marker. Samples were analyzed by FACS after gating on GR-1+ population. Day 0 BMC can down-regulate CD43 only with anti-CD43 stimulation, while day 3 BMC can down-regulate CD43 with both anti-CD43 and –CD43 stimulation. This is a representative of several experiments.  70  In an attempt to examine down-regulation of CD43 on more homogeneous cell populations, assays were also performed with more mature cells. For this, bone marrow derived mast cell progenitors (obtained after 28 days of culture in IL3 and SCF) and bone marrow derived macrophages (obtained after 7 days of culture in M-CSF) were tested for CD43 down-regulation. As shown in Figure 3.1.5, mature mast cells showed a significant down-regulation after TPA stimulation demonstrating that CD43 can be down-regulated on mature primary cells. In contrast, the surface expression of CD43 on macrophages did not change at all. While we cannot exclude that CD43 down-regulation might be inducible on macrophages with other stimuli, our data show that macrophages do not down-regulate CD43 after TPA treatment whereas mast cells do.  1.2 Secretion of CD43 from cell surface 1.2.1 CD43 secretion examined by ELISA If the down-regulation of CD43 we observed by FACS analysis is a consequence of shedding, we postulate that the cleaved extracellular domain of CD43 is highly likely to be released into the surrounding culture medium. It is notable that there are examples where the shedding of a surface molecule is not followed by secretion. For example, most of shed Muc-1 protein remains associated with the cell membrane and is retained by a non-covalent homotypic interaction (Ligtenberg, et al., 1992). There are also examples in which shedding and endocytosis occur concurrently so that the shed molecule is trapped inside endosomes and later degraded or recycled (Gutwein et al., 2003). It was thus of interest to determine whether the cleaved ectodomain of CD43 is released from the cell surface and can be detected in the medium. 71  Figure 3.1.5 Down-regulation of CD43 on mouse mast cells and macrophages Mast cells were obtained by culturing bone marrow cells with IL2 and SCF for 4 weeks. Macrophages were grown by culturing bone marrow cells with M-CSF for 1 week. Cells were stimulated with 50 uM TPA for 3 hours and stained with biotinylated mAb S11, anti-CD43 antibody, followed by FACS analysis. Mast cells but not macrophages downregulate CD43 on TPA stimulation. This is a representative of several experiments. 72  To examine this question, we transfected primary T cells from CD43 null mice and cell lines with retroviruses that express CD43 with Flag-epitope tag at the N-terminus. Flag-tagged CD43 allows us to capture CD43 in a sandwich ELISA assay regardless of whether the C-terminus was lost or retained. Infected cells were selected with antibiotics to obtain cell populations expressing the target gene, and culture supernatant was recovered and incubated on anti-FLAG antibody M2 coated ELISA plates. The amount of captured Flag-CD43 molecule was then detected with either S11 or H18, anti-CD43 antibodies. S11 binds to the extracellular domain of CD43 and H18 binds to the cytoplasmic domain of CD43. Shed CD43 molecule should have no cytoplasmic tail and can be detected by S11 but lacks a H18 signal. In contrast, full-length CD43 released by either membrane blebbing or from necrotic cells should have an ELISA signal with both antibodies S11 and H18. MC/9, CTLL2 and primary T cells were included in these experiments because MC/9 did not exhibit a down-regulation of CD43 whereas CTLL2 and T cells showed a significant loss of CD43 signal by FACS analysis. As shown in Figure 3.1.6, the supernatants from cells transfected with vector alone did not give an ELISA signal. The supernatants from CTLL2 and T cell ConA-blasts that were transfected with Flag-CD43 gave a significant signal with S11, but not H18. This clearly shows the presence of CD43 molecules in the culture supernatants that have S11 epitopes but lack H18 epitopes, consistent with soluble CD43 extracellular domain without a cytoplasmic tail. As expected, soluble CD43 could not be found in the culture supernatant from MC/9 cells. To demonstrate that anti-CD43 antibody H18 is functional in the capture ELISA, we used cell lysates of T cells transfected with Flag-CD43 and observed a good signal implying 73  Figure 3.1.6 CD43 shedding examined by ELISA Cell lines MC/9 and CTLL2 and T cells from spleen of CD43 null mice were transfected with Flag-CD43 or mock vector by retroviruses. Day 5 culture supernatants were recovered and CD43 molecules were captured by anti-Flag antibody coated on plates. Detection was performed with biotinylated S11 and H18, anti-CD43 antibodies. S11 and H18 bind to the extracellular and cytoplasmic domain of CD43, respectively. Lysates of T cells were prepared by resuspending cells in Lysis buffer (20 mM HEPES, 150 mM NaCl, 1% TritonX-100) and incubating for 30 min followed by centrifugation. Lysates were diluted 1/10 before being used as a positive control to assess H18 reactivity. MC/9 which did not induce CD43 down-regulation on previous FACS analysis did not release the extracellular domain of CD43, while CTLL2 and T cells which induced CD43 downregulation on previous FACS analysis released the extracellular domain of CD43 in to the culture medium.  74  that absence of H18 signal and presence of S11 signal is a reliable method to determine whether cytoplasmic-truncated CD43 is released from cells. Our data indicate that a significant amount of CD43 ectodomain is secreted into the culture media. However, it should be noted that we did not stimulate cells in the above experiment and that TPA stimulation of these cells did not alter the rate CD43 was shed (data not shown). It seems that CD43 shedding is constitutive and the induced down-regulation of CD43 surface expression is not followed by shedding. FACS can only detect down-regulation when there is a significant change of expression of a given cell surface molecule in situations where the methods of cell treatment are compared whereas constitutive shedding cannot be registered using FACS because regeneration keeps cell surface expression of a given target molecule at a constant level. ELISA, on the other hand, can detect molecules that have been released into the surrounding media and thus allows identification of molecules that are constitutively shed. Taken together, we conclude that under some conditions and on certain cells CD43 shedding can be either constitutive (revealed by ELISA) or inducible (revealed by FACS).  1.2.2 Unexpected finding: disappearance of endogenous CD43 Our study involved transfection of Flag-CD43 into two cell lines, MC/9 and CTLL2 that express CD43 endogenously. This was the reason we epitope-tagged CD43 with Flag to distinguish it from endogenous expression. The cell lines were infected by standard retroviral transduction and selected against the antibiotic marker for 4-5 weeks. While expanding these cell lines, we observed that the transfected cells developed a tendency to clump after 2-3 months in culture. There were very few cells that were not 75  clumped together and the clumped cells were very difficult to separate. Some of the massive clumps became totally resistant to disruption and remained as a whole mass. We used FACS analysis to determine whether the transgene continued to be expressed, as a widely accepted characteristic of CD43 on hematopoietic cells is antiadhesiveness. We thus speculated that the CD43 transgene may be lost over time, a common occurrence for transgenes introduced via retroviruses. We checked the expression of endogenous and exogenous CD43 by S11 (anti-CD43 antibody) after 4 months in culture. Surprisingly we discovered that the GFP+ population had totally disappeared suggesting that the transgene was lost because the GFP+ population contains the Flag-CD43 in our Flag-CD43-IRES-eGFP construct (Figure 3.1.7). Surprisingly, there was a population of cells that was S11- and GFP- implying that a population of CTLL2 not only down-regulated the Flag-CD43 but also lost the expression of the endogenous CD43. This suggests that over-expression of CD43 in cells that express CD43 can lead to down-regulation of CD43 altogether. This is a very unique phenomenon and has not yet been reported. It is possible that over-expression of CD43 on cells that are expressing endogenous CD43 could lead to total down-regulation of CD43 and excessive cell clumping. To dissect the mechanism that leads to such a down-regulation, western blots on cell lysates, Northern blots on cellular RNAs and southern blots on DNA may allow an insight into this novel feedback mechanism.  1.2.3 CD43 secretion examined by Western blotting To visualize and characterize the shed CD43 ectodomain, we employed Western blotting on primary cells and cell lines. CD43 from cells where C2GlcNAcT core2 76  Figure 3.1.7 Unexpected finding: disappearance of endogenous CD43 CTLL2 cell line (CD43+) was transfected with Flag-CD43 by retroviruses and selected with puromycin for 3 weeks. eGFP is expressed by IRES and puromycin gene is expressed by a separate PGK promoter. Cells were stained with S11, anti-CD43 antibody at 1month and later at 4 month point. Most cells are S11+ and GFP+ after 1 month of transfection. Over the time of 4 month culture, most cells had lost GFP and some cells are even S11- indicating CD43- populations appeared. This is a representative of several experiments.  77  glycosyltransferase is active typically migrates to about 135kDa. Since most of cells we use have active C2GlcNAcT enzyme, we can expect to see a band at around 135kDa. Shed CD43 should be detected in the culture medium with a reduced size to account for the loss of the transmembrane and the cytoplasmic tail which are about 18kDa combined, and shed CD43 ectodomain should migrate in the 120kDa range. We transfected MC/9 cells, primary mast cells (derived from CD43 null bone marrow cells) and T cells (derived from CD43 null splenocytes) with Flag-CD43 by retroviruses and then immunoprecipitated Flag-CD43 from cell lysates and cell supernatants using anti-Flag antibody M2 to concentrate CD43 regardless of its glycosylation and structure. AntiCD43 antibodies S11 and H18 were then used to detect CD43 by Western blot. As in the ELISA, shed CD43 should only be bound by S11 and is expected to have a smaller size than uncleaved full-length CD43 that can be bound by both S11 and H18 antibodies. As shown in Figure 3.1.8, cell lysate from MC/9 cells, which do not shed CD43 as determined by FACS analysis and by ELISA, showed as expected only one band recognized by both S11 and H18 at around 135kDa on cell lysate fraction. The estimated size of the band with 135kDa and the fact that it is detected by both S11 and H18 suggests that it is full-length CD43 confirming that MC/9 cells do not shed CD43. H18, a very potent polyclonal antibody, picks up a relatively weak signal on the cell supernatant fraction. We believe that this represents full-length CD43 associated with cell debris or membrane blebbing because ultracentrifugation of cell supernatant reduces the intensity of the band (data not shown). The identity of a smaller band at about 80kDa is not known but could reflect CD43 precursor in the Golgi apparatus. On the other hand, primary mast cells and T cells, shown to shed CD43 by FACS and ELISA, showed a different pattern 78  Figure 3.1.8 CD43 shedding examination by Western blotting MC/9 cell line, primary mast cell and T cell from CD43 null mice were transfected with Flag-CD43 by retroviruses. Lysates and M2 (anti-Flag antibody) immunoprecipitates from supernatants were probed with biotinylated S11 and H18, anti-CD43 antibodies. S11 and H18 bind to the extracellular and cytoplasmic domain of CD43, respectively. Bone marrow derived mast cells and T cells which induced CD43 down-regulation on previous FACS analysis and ELISA show a band which is reactive to S11 but not to H18 indicating a fragment without the cytoplasmic domain. This is a representative of several experiments.  79  of bands. In addition to the 135kDa band, they have a smaller band at around 120kDa which lacks H18 reactivity and represents shed CD43. We also observed that full-length CD43 can be detected in the medium fraction and that there is shed CD43 in the lysate fraction. It is probable that shed CD43 may remain associated with either full-length CD43 on the cell surface or possibly other cell surface molecules since there are some other sialomucins that are shed and still remain on the cell surface through homotypical interaction such as Muc-1. The retroviral infection procedure we employed was to coculture the target cells with the virus producing adherent cell line, a method that allows us to achieve very good infection rates but is associated with increased cell death. Therefore, it is very likely that membrane blebbing from dying cells spilled full-length CD43 into medium and thus full-length CD43 is detected in our assay. A more detailed study would be required to resolve these issues. Collectively our data show that CD43 can be constitutively shed on some murine cells and that CD43 shedding can be induced on some murine cells. CD43 shedding appears to be associated with non adherent cells whereas adherent cells such as macrophages do not shed CD43 even though we need to study more adherent cells to be certain.  2. Non-sheddable CD43 2.1 Construction of non-sheddable CD43 chimeras In order to study the function of CD43 shedding, we decided to make nonsheddable CD43. Shedding of human CD43 has been shown to occur at the membrane proximal region by protein sequencing and Western blots (Schmid et al., 1992; Remold80  O’Donnell and Parent, 1994). CD34 is another member of the sialomucin family, has a similar structure as CD43 and there is no evidence that CD34 is shed. To produce nonsheddable CD43, we swapped the membrane proximal region (Mutant MP) and the transmembrane region (Mutant TM) of CD43 representing two CD43 domains where shedding is most likely to occur with the corresponding regions of CD34 (Figure 3.2.1 (A)). We predicted that making these changes we will be able to block the actions of secretases, thus making CD43/34 chimeras non-sheddable. The constructs were transfected into CD43 null bone marrow cells by retroviral infection and the cells were differentiated into granulocytes with G-CSF and IL3. After 5 days when more than 90% of cells were GR-1+/Mac-1+, the cells were subjected to an assay with immobilized anti-CD43 antibody S11. As shown in Figure 3.2.1 (B), only mutant MP efficiently blocked down-regulation of CD43 whereas mutant TM transfected cells showed down-regulation that was comparable to wild-type CD43 transfected cells. This tells us that swapping membrane proximal region of CD43 with CD34 can prevent down-regulation and is consistent with down-regulation by ectodomain cleavage. Interestingly we observed there was a much reduced cell yield in both mutant MP and TM transfected cells as compared to wild-type CD43 transfected cells.  2.2 Toxicity of CD43/34 chimeras During the cultivation of cells transfected with either of the CD43/34 chimeras, mutant MP and TM, it became evident that both molecules had a negative impact on cell growth. Initially we observed that the φ-2 cells that were used as the retrovirus producing cell line, exhibited retarded growth and down-regulated CD43/34 transgenes and GFP 81  Figure 3.2.1 Construction of non-sheddable CD43 To construct non-sheddable CD43, the membrane proximal region (Mutant MP) and the transmembrane region (Mutant TM) of CD34 were swapped with the corresponding regions of CD34 (A). The constructs were transfected into bone marrow cells of CD43 null mice and differentiated into granulocytes by supplementing IL3 and G-CSF for 5 days. Cells were subjected to shedding assay by incubating them on plates coated with immobilized S11, anti-CD43 antibody (B). The wild type CD43 and mutant TM show a significant down-regulation but not from mutant MP suggesting the down-regulation occur at the membrane proximal region. This is a representative of several experiments.  82  over time when they were transfected with chimeric constructs. The down-regulation became evident usually after 4-6 weeks after transfection. We discovered that the infection of target cells with φ-2 cells transfected with mutants was problematic. As shown in Figure 3.2.2 and 3.2.3, CD43/34 chimera produced significantly less infected cells (shown as percentages of GFP+ populations) and lower viable cell numbers when they were introduced on CD43 null bone marrow cells and differentiated into granulocytes over 8 days. In contrast, wild-type CD43 was successfully introduced into cells and cell yields were significantly higher. Mutant MP and mutant TM showed only 11% and 4% of GFP+ cells and 0.06 million/ml and 0.04 million/ml total cells, respectively, whereas wild-type CD43 transfected cells had almost three times GFP+ cells with 35% and six times more cells at 0.4 million/ml total cells. FSC and SCS profiles indicated that there were less live and healthy cells in cell cultures transfected with either CD43/34 chimeras. As described above, φ-2 cells transfected with the mutants down-regulated the transgenes over time, this toxicity was only observed with freshly transfected φ-2 cells. In summary, granulocytes that can shed CD43 based on our previous data are negatively affected when transfected with either of the two CD43/34 chimeras.  2.3 Physiological relevance of toxicity of CD43/34 chimeras One explanation of the negative impact of CD43/34 chimeras on cell growth could be that these artificial constructs might not fold and/or be localized properly, thus causing non-physiological cell death. However, it was difficult to speculate that CD43 chimeras misfold as CD43 is believed to have no folded structure and is described as a 83  Figure 3.2.2 Toxicity of non-sheddable CD43 chimeras (GFP+) Bone marrow cells from CD43 null mice were transfected with CD43, mutant MP or mutant TM, and grown in granulocyte enriching condition (G-CSF and IL3) for 8 days. The constructs were expressed by LTR followed by IRES-GFP. The Shown are percentages of GFP+ cells after gating on GFP+ populations. Mutant MP and mutant TM show a severely impaired GFP expression. The graph shows average percentages from triplicates and a similar trend was observed from 3 separate experiments.  84  Figure 3.2.3 Toxicity of non-sheddable CD43 chimeras (cell numbers) Bone marrow cells from CD43 null mice were infected with CD43, Mutant MP or mutant TM, and grown in granulocyte enriching condition (G-CSF and IL3) for 8 days. The shown are the absolute numbers of live cells over 8 days in culture measured by FSC/SSC gating and bead counting. Mutant MP and mutant TM show a significantly increased cell death. The graph shows average cell numbers from triplicates and a similar trend was observed from 3 separate experiments.  85  rod-like mucin molecule (Cyster et al., 1991). Furthermore, we could detect both mutant MP and TM on the cell surface by FACS suggesting that cell surface localization does not appear to be a problem. To further investigate this issue, we infected the CD43/34 chimeric constructs into CD43 null bone marrow cells while keeping them on the stem cell/progenitor enriching condition (SCF, FLT-3 ligand, IL11 and IL3) for two days. This condition has been shown to maintain potential stem cell/progenitor cells in undifferentiated state while allowing them to divide which is essential for retroviral infection. The transfected cells were then separated into two dishes and one dish was differentiated into granulocytes using G-CSF and IL3 and the second dish into monocyte/macrophages using M-CSF and SCF. To examine cell death, cell samples were collected and stained with PI and AnnexinV to measure dead cells and apoptotic cells, respectively. As shown in Figure 3.2.4, only the cells differentiated into granulocytes have higher cell death after expressing CD43/34 chimeras based on the PI and AnnexinV staining whereas cells differentiated into macrophages did not have higher cell death with CD43/34 chimeras. If the CD43/34 chimeras are intrinsically and inherently defective in folding or in localization, they should exert toxicity regardless of cell types. However, our data clearly show that the toxicity is cell-type specific implying that the toxicity is caused by specific cellular processes. These findings correlate perfectly with shedding patterns we have established previously. Monocytes/macrophage lineage which does not shed CD43 can tolerate the CD43/34 chimeras, whereas granulocytes that normally shed CD43 cannot tolerate CD43/34 chimeras indicating that CD43 shedding might be an essential process in cells 86  Figure 3.2.4 CD43/34 chimeras on non- and sheddable cells Bone marrow cells from CD43 null mice were transfected with CD43 or CD43/34 chimeras by retroviruses and kept in stem cell/progenitor enriching culture for 2 days. Then the infected cells were split and differentiated into either granulocyte (IL3 and GCSF) or monocyte/macrophage lineage (SCF & M-CSF) for another 3 days. Granulocytes which shed CD43 showed a severely impaired growth with mutant MP and TM but not with monocytes which do not shed CD43.  87  where CD43 sheds. Both mutant MP and TM are toxic to cells even though only mutant MP is not down-regulated as established by previous FACS analysis. The observed toxicity implies that there may be a common pathway that both mutant MP and TM inhibit, but not wildtype. This made us to speculate that CD43 might go through RIP (regulated intramembrane proteolysis). As explained in the Introduction section, there are an increasing number of molecules that are subjected to RIP to control the release of their cytoplasmic domains. The controlled release of cytoplasmic domain has long been recognized to impact on many cellular functions. Classical examples are APP and Notch processing. RIP cascade is typically proteolytic processing mediated by α-secretases to release the ectodomain of the target protein. Cleavage by γ- secretases in the transmembrane domain which will release the cytoplasmic tail can only occur after release of the ectodomain. We thus speculate that mutant MP blocks the α-secretase and mutant TM inhibits the γ-secretase. Either of these will thus result in the loss of cleavage by the γ-secretase as activity of this enzyme depends on ectodomain shedding. Mutant MP and mutant TM may thus have in common a blockade of the γ-secretase, which blocks the release of the cytoplasmic tail.  2.4 Rescue of the toxicity caused by CD43/34 chimeras If our model is correct that both CD43/34 chimeras prevent γ-secretases from releasing the CD43 cytoplasmic domain from the cell membrane and that the retention of the cytoplasmic domain somehow causes cell toxicity, we should be able to, at least to some degree, prevent toxicity of the CD43/34 chimeras if they are transfected into cells 88  that co-express wild-type CD43. We thus transfected mutant MP and TM chimeras in parallel into wild-type and CD43 null bone marrow cells and differentiated them into mast cells to see whether endogenous CD43 can rescue the chimera-associated toxicity. Figure 3.2.5 shows the results of transfection experiments plotted with absolute numbers of total GFP+ cells over time. When bone marrow cells are differentiated into mast cells, there is normally massive cell death until about day 10 because the cells that are not able to become mast cells die under the culture conditions that support mast cell growth. After day 12-14, only progenitor cells that have the potential to become mast cells remain resulting in increased proliferation from this time point. Cells become mature mast cells around week 3-4. Wild-type and CD43 null bone marrow cells transfected with wild-type CD43 followed this typical growth curve of mast cell differentiation as shown in Figure 3.2.5. By day 12 when the cell numbers hit the bottom of the growth curve, there were about 20,000 GFP+ cells per ml and they start to increase from this time point. In contrast, there were less than 10,000 GFP+ cells per ml in cell cultures transfected with either mutant MP or mutant TM. This is a substantial difference as cell numbers at this initial stage of the culture will affect the future growth potential of cell cultures. This was surprising as the three retroviral constructs have an equal infection efficiency of about 35% at day 3-5 (data not shown). This suggests that the cells that express CD43/34 chimeric constructs start to die at an early phase of mast cell differentiation. About two weeks after transfection at a time when mast cell precursors normally start to divide exponentially, cells expressing mutant MP show a significantly retarded growth. These cells remain at low numbers in both wild-type and CD43 null cells. 89  Figure 3.2.5 Absolute numbers of GFP+ mast cells with CD43/34 chimeras Bone marrow cells from wild type and CD43 null mice were transfected with CD43, mutant MP and mutant TM constructs with retroviruses. Cells were differentiated into mast cells for 36 days with IL3 and SCF. At each day point, cells were stained with PI to eliminate dead cells and absolute numbers of GFP+ cells per 0.1ml were counted with counting beads. Mutant MP showed a severely impaired growth with mast cells from CD43 null mice but not from wild type mice. The graph shows average cell numbers from triplicates and a similar trend was observed from 2 separate experiments.  90  However, there is a huge 50 fold difference in cell numbers between wild type and CD43 null cells transfected with mutant MP. Mutant TM transfected cells also had significantly reduced number at days 4-8 when compared to CD43 transfected cells and the reduced number was similar to Mutant MP transfectants, but these cells appear to eventually recover indicating that mutant MP is less toxic. In conclusion, by examining cell growth of mutant MP transfectants, it is clear that wild type CD43 can rescue the toxicity associated with non-sheddable CD43, mutant MP implying that CD43 shedding is part of a cellular process. It should be noted that the picture becomes even more complicated when we examine GFP- cells, i.e. the mast cells without expressing transgenes. As shown in Figure 3.2.6, GFP- cells show comparable growth regardless of cell types and constructs used. However, there is a notable absence of GFP- mast cells in CD 43 null bone marrow transfected with mutant MP, whereas all other cultures yield significant and comparable numbers of GFP- mast cells. As the infection protocol used results in 30-40% transfection rates, we would expect to find significant numbers of GFP- cells in all cultures. Therefore, total loss of all GFP- mast cells in CD43 null transfected cells with mutant MP needs further consideration. The fact that we are losing GFP- CD43 null cells completely indicates that expression of mutant MP in early progenitors is probably very toxic and that we are likely underestimating the retroviral transfection efficiency on population of dividing cells. Since transgenes delivered by retroviruses can be expressed only in dividing cells, mast cell progenitors under our culture condition are the targets of retroviruses. Therefore, it is possible that we might have infected most of dividing mast cell progenitors even though 91  Figure 3.2.6 Absolute numbers of GFP- mast cells with CD43/34 chimeras Bone marrow cells from wild type and CD43 null mice were transfected with CD43, mutant MP and mutant TM constructs with retroviruses. Cells were differentiated into mast cells for 36 days with IL3 and SCF. At each day point, cells were stained with PI to eliminate dead cells and absolute numbers of GFP- cells per 0.1ml were counted with counting beads. The graph shows average cell numbers from triplicates and a similar trend was observed from 2 separate experiments.  92  the overall transfection efficiencies of the cultures are lower. This might explain why mutant MP induced cell death would deplete the mast cell progenitor populations. Our data also indicate that most of stem cells and progenitors in CD43 null bone marrow culture must have died in the beginning of the culture because mutant MP and mutant TM did not affect cell growth significantly in CD43 null mast cell cultures when transfection was done at day 10-12 (data not shown). Induction of apoptosis in stem cells via CD43 is supported by several publications that showed increased apoptosis of human hematopoietic progenitor cells (CD34hi LIN-) crosslinked with MEM-59 (anti-CD43 antibody) in vitro (Bazil et al., 1995). Mutant TM seems to have a negative impact on cells growth early in the cultures, but cells appear to be recovered relatively easily at later time points of culture. Taken together, we believe that dividing mast progenitors are preferentially destroyed by mutant MP (and less effectively by mutant TM) probably because CD43 shedding and/or its consequences are required for the early progenitors to differentiate.  3. Functional significance of the CD43 cytoplasmic tail 3.1 Nuclear localization of CD43 It was previously reported that the cytoplasmic tail of CD43 might translocate to the nucleus in studies performed on a CHO cell line that normally does not express CD43 (Andersson et al., 2004). To demonstrate nuclear localization of the CD43 cytoplasmic domain, the investigators used overexpression of a truncated cytoplasmic tail fused with GFP in the cytosol. They also examined Colo205, a human colon carcinoma cell line, with an anti-CD43 antibody specific to the cytoplasmic tail and showed nuclear 93  localization by confocal microscopy even though it occurred unexplainably only in low density culture condition. These investigators however failed to demonstrate nuclear localization of CD43 cytoplasmic tail in primary hematopoietic cells where full-length CD43 is normally expressed. CD43 has long been known for its signaling properties, but the mechanism by which it functions has remained as a mystery. If the cytoplasmic portion of CD43 indeed translocates to the nucleus, this might indicate a possible mechanism by which CD43 could mediate signaling. Based on our previous data, we speculate that CD43 shedding might result in the release of the cytoplasmic tail by either the α- or γ-secretase and that blockage of CD43 processing will result in cell death, as suggested by the toxicity induced by mutants MP and TM. Indeed CD43 has a very good consensus motif of the nuclear localization signal (NLS) (Figure 3.3.1). To examine whether translocation of the cytoplasmic tail of CD43 occurs, we constructed a CD43-GFP fusion protein allowing for GFP expression at the Cterminus and transfected CD43 null bone marrow cells followed by culture in IL3 and SCF for 4 weeks to differentiate into mast cells. As shown in Figure 3.3.1 (B), nuclear localization of the CD43-GFP signals can be seen in some of the cells. However we noted again that cultures of bone marrow cells transfected with CD43-GFP, cells showed a retarded growth. When we transfected ψ-2 cells with the plasmid containing CD43-GFP, only very few cells survived after antibiotic selection compared to GFP alone transfected control cells (data not shown). After initial antibiotic selection, ψ-2 cells transfected with CD43-GFP grew much more slowly when compared to GFP vector control. Also, bone marrow infection using the ψ-2 cells transfected with 94  Figure 3.3.1 Nuclear localization of CD43 revealved by CD43-GFP fusion CD43 has a nuclear localization signal that matches the consensus sequence well (A). To visualize the nuclear localization of the cytoplasmic tail of CD43, CD43-GFP fusion at the C-terminus was constructed and expressed on week 4 mast cells of CD43 null mouse (B). DAPI stains shows the location of the nucleus. This is a representative of several experiments.  95  CD43-GFP was a very inefficient process ranging from 0.1-1% of infection whereas GFP controls had over 35% infection efficiencies. After infection, granulocyte or mast cell differentiation yielded very low number of cells. The picture shown in Figure 3.3.1 was taken after a series of attempts to get the best possible infection. The toxicity we observed with the CD43-GFP fusion protein resembles our observations using cultures of ψ-2 cells transfected with mutant MP and TM even though significant cell death was not detected in this case (data not shown). CD43 null bone marrow cells transfected with CD43/34 chimeric constructs consistently had reduced cell growth and eventually resulted in massive cell death. We therefore suspected that expression of CD43-GFP fusion protein might have a similar toxic effect and that GFP fused to the C-terminus of CD43 might affect the function of the CD43 cytoplasmic tail. One possibility is that GFP fusion stabilizes the cytoplasmic tail of CD43 by protecting it from natural degradation which is common occurrence for GFP fusion proteins. CD43-GFP translocated to the nucleus might thus not be subjected to normal proteolysis and this may have a negative impact on cell growth. This explanation is supported by the fluorescence intensity of the CD43-fusion protein in Figure 3.3.1 (B). The CD43-GFP fusion protein appears to have too much brightness (much more than membrane-bound CD43) suggesting that there could be a substantial nuclear accumulation. A second explanation is that the CD43-GFP fusion protein might inhibit the release of the cytoplasmic tail of CD43 from the transmembrane by obstructing secretases. This idea is supported by the fact that we were not able to observe many cells with CD43-GFP nuclear localization. This second explanation could also explain the negative impact on cell growth as with the case of mutants MP and TM. 96  In conclusion, the CD43 cytoplasmic tail appears to translocate to the nucleus in bone marrow cells. However, GFP fusion of the CD43 cytoplasmic tail negatively affects cell growth similar to the effect of cells expressing CD43/34 chimeras that seem to block γ-secretases.  3.2 Negative impact of CD43-GFP fusion on cell growth Mutant MP and TM adversely affect cell growth when grown in culture conditions that enrich granulocytes and mast cells which normally shed CD43, whereas bone marrow cells that were differentiated into monocyte/macrophage lineage that do not shed CD43 could be cultured without difficulty. If the negative effect of CD43-GFP on cell growth is caused by the inability to be properly processed (shedding and/or RIP), CD43-GFP protein might also lose its negative impact when grown in cells that do not shed such as monocytes/macrophages or MC/9. To test this hypothesis we transfected MC/9 cells with CD43-GFP construct and CTLL2 which sheds CD43 as a control. If our hypothesis was correct, MC/9 should express CD43-GFP but there would be no nuclear GFP signal, whereas CD43-GFP should not be suitable for expression in CTLL2 and little nuclear translocation should be observed. As shown in Figure 3.3.2, our hypothesis was correct. CD43-GFP was well expressed in MC/9 but not in CTLL2 confirming that CD43-GFP can be easily expressed in a cell line which does not shed CD43. We could detect a few cells that had CD43-GFP signals in the nucleus in CTLL2 cells but the signal was weak. In contrast, MC/9 and CTLL2 cells can be transfected with similar efficiencies if they are transfected with a retroviral construct that encodes CD43 followed by IRES-GFP. Thus we conclude that 97  Figure 3.3.2 CD43-GFP fusion protein expression in MC/9 and CTLL2 CD43-GFP fusion protein was expressed in MC/9 which does not shed CD43 and CTLL2 which sheds CD43 (A) and (B). CD43-IRES-GFP was expressed in MC/9 and CTLL2 (C). The expression of CD43-GFP is severely impaired in CTLL2 where CD43 sheds but not in MC/9 where CD43 does not shed. On the other hand, CD43-IRES-GFP can be expressed properly in both CTLL2 and MC/9. This is a representative of several experiments.  98  the reason CD43-GFP is toxic is because GFP fusion at the C-terminus interferes with either CD43 shedding or CD43 nuclear function.  3.3 Nuclear localization of CD43 studied by confocal microscopy Because CD43-GFP fusion results in an impaired nuclear localization of the cytoplasmic domain of CD43 and negatively impacts cell growth due to CD43 processing, we employed confocal microscopy using staining with S11 and H18, anti-CD43 antibodies, to confirm the nuclear localization of the CD43 cytoplasmic tail. Figure 3.3.3 (A) and (B) shows confocal images of S11 and H18 staining from conA blasted T cells and culture T cells (day 7), respectively. It is obvious that S11 stains mostly the membrane and H18 stains the membrane and nucleus area. Because ConA blasted T cells have large nuclei it was difficult to decide whether CD43 is localized in the nucleus or just dispersed throughout the cytoplasm including nucleus. It should be noted that there could be a problem with H18 staining since H18 epitope on the cytoplasmic tail overlaps with one of three potential sumoylation sites we have discovered on CD43. Since sumoylation is usually associated with nuclear localization of target proteins, the cytoplasmic tail of CD43 that has been sumoylated and localized in the nucleus might not be efficiently detected by H18.  3.4 Sumoylation of CD43 To identify a possible role of the cytoplasmic tail of CD43 in the nucleus, we searched databases for potential protein domain motifs within the cytoplasmic domain, and found three potential sumoylation sites as shown in Figure 3.3.4 (A). Sumoylation is 99  Figure 3.3.3 Confocal microscopy of CD43 ConA blasted T cells (A) and day 7 culture T cells were fixed with paraformaldehyde and permeabilized with TritonX-100 before being stained with S11 and H18, anti-CD43 antibodies as well as ToPro-3, DNA stain. S11 binds to the extracellular domain of CD43 and H18 binds to the cytoplasmic domain of CD43. The picture shows that the CD43 cytoplasmic tail is clearly localized in the nucleus. This is a representative of several experiments.  100  a post-translational modification in which a SUMO peptide is covalently attached to lysine residues to modulate the stability, function and localization of nuclear proteins. The cytoplasmic domain of CD43 has three potential SUMO modification sites according to the consensus sequence ψKXE, where ψ represents a hydrophobic residue, K a lysine residue modified by SUMO-1, X can be any amino acid, and E represents glutamic acid. Previous efforts to detect the cytoplasmic tail of CD43 inside cells have been consistently unsuccessful as we have been unable to find by Western blots a 13 kDa band that would correspond to the predicted size of the cytoplasmic tail. Western blot analysis of lysates of bone marrow cells from wild type and CD43 null mice using the cytoplasmic tail specific antibody H18 revealed however several distinctive bands (Figure 3.3.4 (B) lane 2) with a M.W. range of 26-29 kDa in wild-type lysates but not in knockout lysates (Figure 3.3.4 (B) lane 1). The observed H18 reactive band may either represent degradation products of full-length CD43 as we have interpreted such bands in the past or alternatively could reflect posttranslationally SUMO modified CD43 cytoplasmic domain. One SUMO-1 modification increases the molecular weight by about 13 kDa. Therefore, we could expect bands between 13, 26, 39 or 52 kDa (from no sumoylation to 3 SUMO attachments) keeping in mind that other posttranslational modifications can also affect the migration of the bands. Indeed in Western blots from lysates of total bone marrow cells (Figure 3.3.4 (B) lane 1), we see a very intense and distinctive band at about 26 kDa suggesting that the CD43 cytoplasmic tail could be modified with one SUMO-1 moiety. Anti-SUMO-1 antibody (GMP-1) immunoprecipitation was performed from lysates of wild-type (Figure 3.3.4 (B) lane 4) and CD43 null (lane 3) bone marrow cells. 101  Figure 3.3.4 Sumoylation of the cytoplasmic tail of CD43 (A) The sequence of the cytoplasmic tail of CD43 is shown to indicate potential SUMO modification sites with the SUMO target consensus sites ψKXE, where ψ is a hydrophobic residue, K is the lysine residue modified by SUMO-1, X is any amino acid, and E is the glutamic acid. (B) Total lysates of total bone marrow cells from wild type (lane 2) and CD43 null (lane 1) mice were probed with H18, cytoplasmic domain specific anti-CD43 antibody. Lysates of CD43 null (lane 3) and of wild-type (lane 4) from total bone marrow were immunoprecipitated with anti-SUMO-1 antibody and probed with H18. This was performed only once.  102  The lysates were prepared using conditions that allows extraction of nuclear proteins (high salt, see Methods and Materials). Immunoprecipitates probed with H18 for wildtype mice lysate revealed two strong bands, one at M.W. 26 kDa and a second band at M.W. 50 kDa, whereas no H18 reactivity was observed in these regions in the corresponding immunoprecipitates from CD43 null cells. Our data thus suggest that two variants of SUMO-modified CD43 cytoplasmic tail occur, one that has a single SUMO attached and a second with three SUMO modifications. Mutational studies will be required to find out which potential SUMO sites are actually sumoylated. It will be furthermore interesting to know whether there is a difference between the one-SUMO and three-SUMO modified CD43 tail in nuclear localization. Since most of the sumoylated proteins are localized in the nucleus, our findings that CD43 is modified by SUMO peptides and localized in the nucleus fit our working model. A remaining major question is the functional significance of the CD43 cytoplasmic tail in the nucleus. Identification of binding partners of the CD43 cytoplasmic tail using approaches such as a yeast two hybrid assay or mass spectrometry will be required to resolve this puzzle.  3.5 Colocalization of the CD43 cytoplasmic tail at PML nuclear bodies Most sumoylated proteins are localized in the nucleus where many of them are found in PML nuclear bodies and influence the structure and function of PML nuclear bodies. Knocking down one of the proteins in PML nuclear bodies usually results in the disintegration of PML nuclear bodies either into much smaller pieces or reduced numbers. The discovery that CD43 is sumoylated and is localized throughout the nucleus raises the 103  question whether bone marrow cells from CD43 deficient mice show differences in either the structure or localization of PML nuclear bodies. Figure 3.3.5, shows confocal microscopy images of bone marrow cells from wild-type mice stained with H18 combined with either anti-PML antibody or anti-SUMO antibody. H18 staining of bone marrow cells from wild-type mice shows a significant number of large speckles. H18 and SUMO-1 antibody costaining shows several speckles reactive with both antibodies indicating that the CD43 cytoplasmic tail is colocalized with sumoylated proteins in subnuclear structures. H18 and anti-PML antibody staining shows that several PML nuclear bodies are H18 reactive implying that the CD43 cytoplasmic domain is indeed localized in PML nuclear bodies. Our data are consistent with the fact that most of sumoylated proteins are known to be localized in PML nuclear bodies. More studies with different types of cells and preferably subcellular fractionation will be needed in the future to determine the exact localization of CD43 cytoplasmic tail. A major function of PML nuclear bodies is regulation of apoptosis. We addressed the question whether unusual phenotypes of apoptosis can be found in CD43 deficient cells. For instance, it is known that neutrophils undergoing apoptosis make dramatic changes with respect to the expression and function of cell adhesion molecules. Typically during apoptosis, most of surface adhesion molecules including CD43 are downregulated on neutrophils to shut down their function (Hart et al., 2000). However, they also exhibit an unusual up-regulation of β2 integrins CD11b/CD18 (50%) and CD11c/CD18 (150%) even though the cells going through apoptosis show reduced adhesion to E-selectin and fibrinogen suggesting that they are functionally inert (Dransfield et al., 1995). It is still not known what mechanisms control the modulation of 104  Figure 3.3.5 Confocal microscopy of PML nuclear bodies Total bone marrow cells from wild-type mouse were fixed with paraformaldehyde and permeabilized with TritonX-100. The samples were then stained with anti-SUMO antibody, anti-PML antibody and H18 (anti-CD43 antibody) followed by confocal microscopy. H18 forms speckles which colocalized at PML nuclear bodies together with SUMOylated proteins. This illustrates that the CD43 cytoplasmic tail is localized in PML nuclear bodies. This is a representative of two experiments.  105  analysis of cell surface marker expression on neutrophils of bone marrow cells from wild-type and CD43 null mice during apoptosis. Most of the cell surface markers including B220 (not shown), GR-1 (not shown), and F4/80 (Figure 3.3.6) showed more or less similar expression patterns on wild-type and CD43 null bone marrow cells after treatment to induce apoptosis with cycloheximide and TNFα. Surprisingly, one surface molecule, CD11c/CD18, showed a dramatic difference between wild-type and CD43 knockout mice. Neutrophils from wild-type mice that were undergoing apoptosis upregulated CD11c/CD18 confirming the report mentioned above, however neutrophils from CD43 null mice failed to up-regulate CD11c/CD18 at all. At present, we do not have a mechanism to explain the failed up-regulation of CD11c/CD18 from CD43 null because the apoptosis associated dysregulation of CD11c/CD18 expression on neutrophils in general has been unexplainable.  3.6 CD43 and apoptosis regulation There are many molecules known to promote apoptosis, but for many of them the mechanisms of action are still unknown. We used FACS analysis to compare the percentages of non-apoptotic (AnnexinV-) and live (PI-) cells of bone marrow, lymph node, and spleen from wild-type and CD43 null mice after induction of apoptosis by γirradiation, IFN-γ, anti-Fas antibody, cycloheximide, TNFα and growth factor withdrawal. Many of the stimuli we applied showed no significant differences in percentages of PIand AnnexinV- cells between wild-type and CD43 deficient cells with the notable exceptions of growth factor withdrawal as well as cycloheximide and TNFα induced apoptosis. A significant difference could be observed for granulocytes (GR-1+) when 106  Figure 3.3.6 Dysregulation of CD11c expression after apoptosis induction Total bone marrow cells from wild-type and CD43 null mice were treated with 50 ug/ml cycloheximide (CHX), a protein synthesis inhibitor, and 10 ng/ml TNFα to induce apoptosis for 18 hours. Cells were stained with PI and cell surface markers, anti-F4/80 antibody, anti-CD11b antibody, and anti-CD11c antibody. The up-regulation of CD11c on wild-type bone marrow cells were not seen on bone marrow cells of CD43 null mice. This is a representative of several experiments.  107  grown in RPMI media without growth factors such as IL3 and G-CSF as shown in no treatment of Figure 3.3.7. This clearly shows that GR-1+ bone marrow cells of CD43 null mice have intrinsic defects on survival ex vivo without growth factor supplemented. Both cycloheximide (protein synthesis inhibitor and FADD –dependent apoptosis inducer) and TNFα (one of the most potent physiological promoter of apoptosis, Sidoti-de et al., 1998) induced even stronger apoptotic responses. Interestingly, cycloheximide induced wider discrepancy between wild type and CD43 null mouse, and TNFα did not induce much further apoptosis on CD43 null mice. The TNF (Tumor necrosis factor) family of cytokines includes TNF, lymphotoxinα, Fas ligand, TRAIL, and CD40 ligand. Most of the TNF family members usually induce signaling pathways that lead to the activation of the transcriptional factor NFκB which sends survival signals. On the other hand, some of the members can induce apoptosis by binding to death receptors including the TNF receptors, the Fas receptors and TRAIL receptors. Therefore, ligation between TNF family members and various receptors can induce either cell survival or apoptotic signal depending on the combinations of ligands and receptors as well as other secondary cellular signals shown in Figure 3.3.8. Ligandreceptor ligation results in the recruitment of a complex of TRADD, RIP1, and TRAF2 molecules to the receptor. If this binding is sustainable and cellular levels of FLIPL are sufficiently high, this complex leads to cell survival signal. If the TRADD, RIP1, and TRAF2 complex is not stable and FLIPL levels are low, it is dissociated from TNFR1 and binds to FADD resulting in apoptosis by recruiting Caspase-8/10. Cycloheximide is a protein synthesis inhibitor which can initiate apoptosis either acting on by its own (Martin et al., 1990) or by assisting TNFα-induced apoptosis 108  Figure 3.3.7 Impaired survival of CD43 null GR-1+ cells Bone marrow cells from wild-type and CD43 null mice were incubated with 1.5 ug/ml cycloheximide (CHX) and 100 ng/ml TNFα for 36 hours to induce apoptosis. 10 ng/ml IL3 was added in some cases to promote granulocyte differentiation. Samples were stained with granulocyte marker GR-1 and PI/AnnexinV to gate on live cells. GR-1+ cells of CD43 null mice showed intrinsic defects on survival without growth factors (IL3) which can be reversed by adding IL3. The graph shows average percentages from triplicates.  109  Figure 3.3.8 TNFR signaling leading to survival or apoptosis pathway (figure from Micheau and Tschopp, 2003) TNF receptor can send either survival or apoptotic signals depending on situation. High concentration of FLIP in the cytosol and stable association of TNFR-TRADD-TRAF2 complex lead to survival signalling. Low concentration of FLIP and unstable association of TNFR-TRADD-TRAF2 complex dissociate TRADD-TRAF2 from the receptor complex. Then the released TRADD-TRAF2 associates with FADD-Caspase to generate apoptotic signalling.  110  (Tsuchida et al., 1995). In some cases, cycloheximide can even block apoptosis (Martin et al., 1988) according to specific systems used in vitro. Its apoptosis induction has been known to involve a FADD-dependent mechanism (Tang et al., 1999). After 36 hour incubation alone without treatment, granulocytes (GR-1+) from CD43 null bone marrow show a significantly lower percentage of live and non-apoptotic (PI- and AnnexinV-) cells compared to cells from wild-type mice. When the same culture was incubated with cycloheximide, there was a three-fold reduction in PI-/AnnexinVGR-1+ cells from CD43 null bone marrow. This shows that CD43 null GR-1 cells have generally higher sensitivity toward apoptosis in vitro. Interestingly, GR-1+ cells of CD43 null bone marrow were insensitive to TNFα induced apoptosis. GR-1+ cells of wild type bone marrow showed 60% reduction of PI-/AnnexinV- cells upon TNFα treatment, but GR-1+ cells from CD43 null bone marrow showed only 20% reduction. Taken together, these data suggests that CD43 null GR-1 cells have dysregulated apoptotic signaling involving FADD and TNFα receptor. More studies will be needed to dissect the signaling pathways which are defective in apoptosis of CD43 null mice. It should be noted that when bone marrow cells were supplemented with IL3, there was no significant difference between wild-type and CD43 null GR-1+ cells indicating that the missing or defective signaling components or pathways of CD43 null mice can be compensated by IL3-IL3R signaling. Since the presence of CD43 conferred a survival advantage to GR-1+ cells, we also tested other cell types. We used magnetic bead separation to purify GR-1+ cells from bone marrow as well as B220+ and CD4+/CD8+ cells from spleen of wild-type or CD43  111  Figure 3.3.9 Impaired survival of CD43 - cells GR-1+ cells were purified form bone marrows and B220+ and CD4+/CD8+ cells were purified from spleens of wild-type mice by magnetic bead separation. GR-1+ cells were incubated in RPMI without growth factors for 36 hours, whereas B220+ and CD4/CD8+ cells were incubated in RPMI without growth factors for 16 hours. Samples were stained with PI and AnnexinV to gate on live and non-apoptotic cells. Cell types which normally express CD43 including BM GR-1+, BM B220+ and spleen CD4+/CD8+ showed differences in apoptosis induced by growth factor withdrawal, but not from spleen B220+ cells which do not normally express CD43. The graph shows average percentages from triplicates and a similar trend was observed from 4 separate experiments.  112  null mice. Cells were incubated in RPMI without growth factors for 36 hours (GR-1+) or 16 hours (B220+ and CD4+/CD8+). As shown in Figure 3.3.9, BM GR-1+, BM B220+ and spleen CD4+/CD8+ cells of wild-type mice which normally express CD43 showed a better survival. In contrast, BM GR-1+, BM B220+ and spleen CD4+/CD8+ cells of CD43 null mice as well as spleen B220+ cells of both “wild-type” and CD43 null mice which do not normally express CD43 showed poor survival. These observations indicate that the presence of CD43 gives cells an advantage in survival and that translocation of the CD43 cytoplasmic tail to the nucleus might be associated with an anti-apoptotic effect.  113  4. Conclusion In summary the work presented in this thesis supports the following conclusions. a. CD43 sheds from many murine leukocytes including T cells, neutrophils and mast cells. b. CD43 goes through regulated intramembrane proteolysis. c. Non-sheddable CD43 negatively affects cell survival. d. CD43 cytoplasmic domain is sumoylated and translocated to the nucleus. e. CD43 cytoplasmic domain in the nucleus is localized at PML nuclear bodies and regulates apoptosis responses.  114  CHAPTER 4 SUMMARY AND DISCUSSION Many cell surface molecules can be removed from the cell surface by proteolytic enzymes in a process called ectodomain shedding and this process can have several different consequences. It can result in altered cell-cell interaction if pro- or alternatively anti-adhesive receptors are removed from cells. Soluble receptors that are released by sheddases can have functionality as antagonist or agonist, for instance cytokine receptors such as gp130, TNFα and Fas. The cytoplasmic domain of the shed molecule finally can have a further function as a signaling molecule, the classical example being Notch signaling. CD43 was reported to be shed and our work was initiated to study the functional significance of CD43 shedding. CD43 shedding was initially described in human leukocytes and there were no data reported to document in murine cells. To confirm whether CD43 shedding also occurs on murine cells, we used several approaches. Down-regulation was determined by reduction of CD43 staining in FACS analysis of several mouse cell lines and primary cells after stimulation. Constitutive shedding of the CD43 extracellular domain was furthermore detected by capture ELISA and by Western blots in culture supernatants of several mouse cell lines and primary cells that were transfected with Flag-CD43. Our data conclusively show that CD43 shedding also occurs on mouse cells and that soluble CD43 corresponding to the full-length ectodomain can be easily observed, thus validating the murine model system to study function of CD43 shedding. Previous attempts in our laboratory to measure soluble CD43 in plasma of normal mice by Western blots and ELISA failed due to the lack of CD43 antibodies capable to immunoprecipitate CD43. Observation of soluble CD43 ectodomain was made possible 115  in our present study by introducing an anti-Flag epitope tag at the N-terminus of CD43, as the Flag tag allowed efficient capture and immunoprecipitation of the CD43 ectodomain. CD43 down-regulation was observed on most cell types with the notable exception of the two cell lines MC/9 and WEHI274.3. The cell line MC/9 is derived from mast cells but CD43 down-regulation from primary mast cells was confirmed. The cell line WEHI274.3 is murine myelomonocytic leukemia and interestingly we found that CD43 down-regulation was absent on primary monocyte/macrophage cells. Since cell lines cannot truly represent the cell types that they were derived from, we put heavy emphasis on primary cells and concluded that CD43 down-regulation is cell-type dependent. Since CD43 was mainly believed to mediate cell-cell interaction, it was of particular interest to determine whether CD43 shedding can modulate cell adhesion. To address the question whether CD43 shedding plays a role in the control of cell-cell interaction, we had planned to perform in vitro rolling assays with leukocytes expressing non-sheddable CD43/34 chimeric molecules. However, we could not conduct these experiments due to the toxicity associated with such molecules. To address which part of CD43/34 chimeras cause the toxicity, we constructed numerous CD43/34 chimeric constructs as shown in Appendix II, which did not give us any conclusive answer as our earlier working models did not include the possibility of nuclear localization of CD43 and test for this translocation. Interestingly the toxicity with CD43/34 chimeras was observed in stem cell and/or progenitor populations where CD43 can be shed but not in cells where CD43 does not 116  shed such as macrophages suggesting that CD43 shedding may be an essential process in cells where CD43 can shed. Previously there have been several reports showing that proliferating stem cells and progenitors go though apoptosis when ligated with anti-CD43 antibody but no mechanism or physiological function for this apoptosis was proposed (Bazil et al., 1995; Bazil et al., 1996). Our data are consistent with these earlier data and may point to an important function of CD43. Indeed the toxicity of the CD43/34 chimeras could be reversed to a certain degree when the mutants were expressed in wild type CD43 positive cells supporting our hypothesis that CD43/34 chimeras were intervening with a specific cellular process and led us to speculate that CD43 shedding might deliver signals intracellularly. Since both mutant MP and mutant TM were toxic even though only mutant MP was non-sheddable, we speculated that there must be a common pathway that is inhibited by both chimeras. The model we propose is that CD43 goes through regulated intramembrane proteolysis (RIP) in which an α-secretase releases CD43 ectodomain and a γ-secretase then releases the membrane anchored CD43 cytoplasmic tail into the cytosol. We have however no direct evidence to prove that murine CD43 is a substrate of α- and γ-secretases. There are no known consensus motifs for α- and γ-secretases and the identity of the specific secretases involved in CD43 shedding is largely unknown. There are also multiple α- and γ-secretases in cells and they appear to be in large protein complexes (Marjaux and Strooper, 2004) ruling out approaches such as SiRNA or specific inhibitors to characterize involvement of these proteases in CD43 processing. Our hypothesis that CD43 is subjected to proteolysis by α- and γ-secretases is based on three observations. First, ectopic expression of both membrane proximal and 117  transmembrane mutants have a negative impact on cell survival even though only the membrane proximal mutant cannot be shed suggesting that the two chimeras negatively affect a common downstream process. We are postulating that this common downstream process constitutes the blockage of the γ-secretase processing step in RIP. Secondly we found that the CD43 cytoplasmic domain is released from the membrane and translocates to the nucleus which is an important requirement for RIP. Unfortunately, the phenotype associated with ectopic expression of these chimeras was cell death making it very difficult to investigate the signaling pathway affected by these chimeras. Thirdly it has been shown that human CD43 is a substrate of γ-secretase even though the experiments were performed in human adenoma cell lines (Andersson et al., 2005). We found that the cytoplasmic domain was localized in the nucleus, consistent with our hypothesis that CD43 shedding is followed by the release of the CD43 cytoplasmic tail using CD43-GFP fusion proteins and confocal imaging. However, the toxicity associated with this GFP fusion protein also prevented us from using this construct to study nuclear translocation in more detail. The most suitable reagent available to us to study nuclear translocation of the CD43 cytoplasmic tail was the rabbit polyclonal antibody H18 that was raised against a peptide corresponding to the last 16 amino acids of the CD43 cytoplasmic tail. However, interpretation of data obtained with this antibody must take into consideration that the antibody was raised against a CD43 peptide that contains one of the three sumoylation consensus motifs we have identified. It is thus possible that CD43 detection in the nucleus may be limited in that one of the SUMO modifications at this site may result in reduced signal with H18. Nonetheless we used confocal microscopy with H18 and visualized the cytoplasmic tail in the nucleus 118  consistent with GFP construct data. In future studies, experiments should be done using a construct of CD43 tagged with the Flag epitope at the C-terminus to allow better immunostaining of the cytoplasmic tail. This may provide better signals to determine more clearly the localization of the cytoplasmic tail within the nucleus by confocal microscopy. Our discovery that CD43 cytoplasmic tail might contain up to three SUMO modifications on the cytoplasmic tail is of significant interest. Since almost all SUMOmodified proteins are localized in the nucleus, this observation is in agreement with data that show nuclear translocation of the CD43 cytoplasmic tail. Mutations of the CD43 SUMO modification sites will be required to confirm this process in more detail. Many SUMO substrates are colocalized with PML nuclear bodies and we have also been able to find good evidence that the CD43 cytoplasmic tail is colocalized with PML nuclear bodies. A major function of PML nuclear bodies is believed to be in the regulation of apoptosis. In agreement with this notion, we have found that growth factor withdrawal results in an increased apoptotic responses in CD43 deficient cells indicating that CD43 might be an anti-apoptotic molecule. Even though we have backcrossed CD43 null mice 9 generations onto the C57Bl/6 background to eliminate possible influences of the 129 genomic background of the original CD43 null mice, we need to remain careful with interpretation of our data as earlier studies in our laboratory have shown that phenotypes ascribed to CD43 deficiency can be due to influence of 129 background genes. An influence of 129 background genes on the outcome of the experiments performed with CD43/34 chimeras can however be excluded as we carried out comparative analysis of cells transfected with chimeras versus 119  wild type CD43 transfectants. In contrast to this, we cannot yet completely rule out that 129 background genes contribute to the dysregulation of the localization of PML nuclear bodies and to the regulation of apoptosis. Some of these experiments should be repeated by putting wild-type CD43 back into CD43 null cells to see if we can rescue the observed phenotypes since these were done by directly comparing phenotypes of cells from wild type and null mice. Without such rescue experiments, any work done with CD43 null mice should be deemed as preliminary no matter how reproducible and evident they appear to be. It is of great interest to determine how the CD43 cytoplasmic tail in the nucleus influences the apoptosis responses. We believe that identification of a binding partner for CD43 may be required to resolve this interesting question. The nuclear translocation and sumoylation of the CD43 cytoplasmic tail that we observed to occur after ectodomain shedding are novel observations that point to a possible role of CD43 in the regulation of apoptosis. With this CD43 function in mind we performed bioinformatics analysis for known protein binding domain on CD43. Our results showed that the CD43 cytoplasmic tail has a consensus binding motif for TRAF2 indicating that the CD43 cytoplasmic tail may interact with this signaling molecule that has been shown to be involved in TNFαinduced anti-apoptotic signaling via NFκB. TRAF2 is a member of TRAF (TNF receptor associated factor) family and directly interacts with TNF receptor. Yeast two-hybrid screening of a Jurkat cDNA library has furthermore shown that the CD43 cytoplasmic tail might interact with Daxx, a nuclear apoptosis regulator (Cermak et al., 2002). The group showed that apoptosis of TF1 cells, a myeloid progenitor-derived cell line, mediated by immobilized anti-CD43 antibody was reversed by overexpression Daxx. Daxx is known 120  for its pro-apoptotic role as a Fas death domain-associated protein and as a PML nuclear body-associated protein indicating it shuttles between the cytoplasm and nucleoplasm. The fact that we observed a dysregulation (most notably anti-apoptotic) of apoptosis in CD43 null cells make TRAF2 and/or Daxx perfect candidates as CD43 binding partners. Our observation that CD43 can have anti-apoptotic effects by interfering with TRAF2 and/or Daxx could provide an explanation why it is beneficial for some cancer cells to express CD43. Binding assays by yeast two-hybrid system, by mass spectrometry or by immunoprecipitation would be excellent ways to examine whether CD43 interacts with TRAF. Daxx binding to human CD43 has already been shown in contrast to the possibility of a TRAF2-CD43 interaction. However at present, we do not know which amino acid residues of the CD43 cytoplasmic tail are critical for interaction with Daxx. In conclusion, our data support the following model. The CD43/34 chimeras cannot be cleaved by γ-secretase and CD43 cytoplasmic domain remains anchored in the cell membrane. The CD43/34 chimeric mutants consequently cause TRAF and/or Daxx to accumulate on the cellular membrane making them non-functional and causing cell toxicity. Our model could also explain why CD43 null mice have mild phenotypes in stark contrast to the cell toxicity we observe with the CD43/34 chimeras, as the absence of CD43 should not inhibit the function and/or localization of TRAF2 and/or Daxx because these apoptosis regulators can freely mobilize when CD43 is not present. Our study shows that CD43 goes through RIP to release the cytoplasmic tail which is sumoylated and translocates to the nucleus as outlined in Figure 4.1.1. The CD43 cytoplasmic tail influences the structure of PML nuclear bodies and apoptotic responses of leukocytes presumably by interacting with TRAF-2 and/or Daxx. 121  Figure 4.1.1 The mechanism of CD43 signalling Our model in this thesis proposes that CD43 goes through Regulated Intramembrane Proteolysis (RIP) to release the cytoplasmic tail which is then sumoylated and translocates to the nucleus. Our data suggest that CD43 cytoplasmic tail in the nucleus influences the structure of PML nuclear bodies, thus regulating apoptosis responses. Mutant MP and TM cannot be shed, so that they cannot release the cytoplasmic tail. We hypothesize that proteins that are interacting with the cytoplasmic tail might be accumulated on the cellular membrane making them non-functional and causing cell toxicity.  122  References Agrawal, B., Krantz, M .J., Parker ,J. and Longenecker, B. M. 1998. Expression of MUC1 mucin on activated human T cells: implications for a role of MUC1 in normal immune regulation. Cancer Res. 58: 4079–4081. Alcaide, P., King, S. L., Dimitroff, C. J., Lim, Y.-C., Fuhlbrigge, F. and Luscinskas, F. W. 2007. The 130kDa Glycoform of CD43 functions as an E-selectin ligand for activated Th1 cells in vitro and in delayed-type hypersensitivity reactions in vivo. J. Invest. Derma. 127: 1964-1972. Altabef, M., Garcia, M., Lavau, C., Bae, S. C., Dejean, A. and Samarut, J. 1996. A retrovirus carrying the promyelocyte-retinoic acid receptor PML-RARalpha fusion gene transforms haematopoietic progenitors in vitro and induces acute leukaemias. EMBO J. 15 : 2707-2716. Andersson, C. X., Fernandez-Rodriguez, J., Laos, S., Sikut, R., Sikut, A., Baeckstrom, D. and Hansson, G. C. 2004. CD43 has a functional NLS, interacts with β-catenin, and affects gene expression. Biochem. Biophys. Res. Commun. 316: 12-17. Andersson, C. X., Fernandez-Rodriguez, J., Laos, S., Baeckstrom, D., Haass, C. and Hansson, G. C. 2005. Shedding and γ-secretase-mediated intramembrane proteolysis of the mucin-type molecule CD43. Biochem. J. 387: 377-384. Ardman, B., Sikorski, M. A., Settles, M. and Staunton, D. E. 1990. Human Immunodeficiency Virus Type 1-infected individuals make autoantibodies that bind to CD43 on normal thymic lymphocytes. J. Exp. Med. 172: 1151-1158. Ardman, B., Sikorski, M. A. and Staunton, D. E. 1992. CD43 Interferes with T-Lymphocyte Adhesion. Proc. Natl. Acad. Sci. U. S. A. 89: 5001–5005. Axelsson, B., Youseffi-Etemad, R., Hammarstrom, S. and Perlmann, P. 1988. Induction of aggregation and enhancement of proliferation and IL-2 secretion in human T cells by antibodies to CD43. J. Immunol. 141: 2912-2917. Baecher, C. M., Dorfman, K. S., Mattei, M-C. and Frelinger, J. G. 1990. cDNA cloning and localization of the mouse leukosialin gene (Ly48) to chromosome 7. Immunogen. 31: 307-314. Baecher-Allan, C. M., Kemp, J. D., Dorfman, K. S., Barth, R. K. and Frelinger, J. G., 1993. Differential epitope expression of Ly-48 (mouse leukosialin). Immunogenetics 37: 183–192. Baeckstrom, D., Zhang, K., Asker, N, Ruetschi, U., Ek, M. and Hansson, G. C. 1995. Expression of the leukocyte-associated sialoglycoprotein CD43 by a colon carcinoma cell line. J. Biol. Chem. 270: 13688-13692. Barran, P., Fellinger, W., Warren, C. E., Dennis, J. W. and Ziltener, H. J. 1997. Modification of CD43 and other lymphocyte O-glycoproteins by core 2 N-acetylglucosaminyltransferase. 123  Glycobiology 7: 129. Baumhueter, S., Singer, M. S., Henzel, W., Hemmerich, S., Renz, M., Rosen, S. D. and Lasky, L. A. 1993. Binding of L-selectin to the vascular sialomucin, CD34. Science. 262: 436-438. Bayer, P., Arndt, A., Metzger, S., Mahajan, R., Melchior, F., Jaenicke, R., and Becker, J. 1998. Structure determination of the small ubiquitin-related modifier SUMO-1. J. Mol. Biol. 280: 275286. Bazil, V. and Strominger, J. L. 1993. CD43, the major sialoglycoprotein of human leukocytes, is proteolytically cleaved from the surface of stimulated lymphocytes and granulocytes. Proc. Natl. Acad. Sci. U. S. A. 90: 3792-3796. Bazil, V. and Strominger, J. L. 1994. Metalloprotease and serine protease are involved in cleavage of CD43, CD44, and CD16 from stimulated human granulocytes. J. Immunol. 152: 1314-1322. Bazil, V., Brandt, J., Tsukamoto, A. and Hoffman, R. 1995. Apoptosis of human hematopoietic progenitor cells induced by crosslinking of surface CD43, the major sialoglycoprotein of leukocytes. Blood. 86: 502-511. Bazil, V., Brandt, J., Chen, S., Roeding, M., Luens, K., Tsukamoto, A. and Hoffman, R. 1996. A monoclonal antibody recognizing CD43 (leukosialin) initiates apoptosis of human hematopoietic progenitor cells but not stem cells. Blood. 87: 1272-1281. Bernardi, R. and Pandolfi, P. P. 2003. Role of PML and the PML-nuclear body in the control of programmed cell death. Oncogene. 22: 9048-9057. Bhatia, M., Wang, J. C. Y., Kapp, U., Bonnet, D. and Dick, J. E. 1997. Purification of primitive human hematopoietic cells capable of repopulating immune-deficient mice. P.N.A.S. 94: 53205325. Blobel, C. P. 2000. Remarkable roles of proteolysis on and beyond the cell surface. Curr. Opion. Cell. Biol. 12: 606-612. Blobel, C. P. 2005. ADAMs: Key components in EGFR signaling and development. Nat. Rev. Mol. Cell. Biol. 6: 32-43. Borden, KLB. 2002. Pondering the Promyelocytic leukemia protein (PML) puzzle: possible functions for PML nuclear bodies. Mol. Cell Biol. 22:5259-5269. Brockhausen, I. 2006. Mucin-type O-glycans in human colon and breast cancer: glycodynamics and functions. EMBO reports. 7: 599-604. Brown M. S., Ye, J., Rawson, R. B. and Goldstein, J. L. 2000. Regulated intramembrane proteolysis: a control mechanism conserved from bacteria to humans. Cell. 100: 391-398. 124  Brown, T. J., Shuford, W. W., Wang, W-C., Nadler, S. G., Bailey, T. S., Marquardt, H. and Mittler, R. S. 1996. Characterization of a CD43/leukosialin-mediated pathway for inducing apoptosis in human T-lymphoblastoid cells. J. Biol. Chem. 271: 27686-27695. Brugger, W., Bühring, H.-J., Grünebach, F., Vogel, W., Kaul, S., Müller, R., Brümmendorf, T. H., Ziegler, B. L., Rappold, I., Brossart, P., Scheding, S. and Kanz, L. 1999. Expression of MUC1 epitopes on normal bone marrow: implications for the detection of micrometastatic tumor cells. J. Clin. Oncol. 17: 1535–1544. Carlow, D. A., Ardman, B. and Ziltener, H. J. 1999. A novel CD8 T cell-restricted CD45RB epitope shared by CD43 is differentially affected by glycosylation. J. Immunol. 163: 1441-1448. Carlow, D. A., Corbel, S. Y. and Ziltener, H. J. 2001. Absence of CD43 fails to alter T cell development and responsiveness. J. Immnunol. 166.: 256-261. Carlow, D. A. and Ziltener, H. J. 2006. CD43 deficiency has no impact in competitive in vivo assays of neutrophil or activated T cell recruitment efficiency. J. Immunol. 177: 6450-6459. Carlsson, S. R., Sasaki, H. and Fukuda, M. 1986. Structural variations of O-linked oligosaccharides present in leukosialin isolated from erythroid, myeloid, and T-lymphoid cell lines. J. Biol. Chem. 261: 12787. Campanero, M. R., Pulido, R., Alonso, J. L., Pivel, J. P., Pimentel-Muinos, F. X., Fresno, M. and Sanchez-Madrid, F. 1991. Down-regulation by tumor necrosis factor-α of neutrophil cell surface expression of the sialophorin CD43 and the hyaluronate receptor CD44 through a proteolytic mechanism. Eur. J. Immunol. 21: 3045-3048. Cermak, L., Simova, S., Pintzas, A., Vaclav, H. and Andera, L. 2002. Molecular mechanisms involved in CD43-mediated apoptosis of TF-1 cells: Roles of transcription, Daxx expression, and adhesion molecules. J. Biol. Chem. 277: 7955-7961. Ching, R.W., Dellaire, G., Eskiw, C.H. and Bazett-Jones, D.P. 2005. PML bodies: a meeting place for genomic loci? J. Cell Sci. 118:847-854. Corinti, S., Fanales-Belasio, E., Albanesi, C., Cavani, A., Angelisova, P. and Girolomoni, G. 1999. Cross-linking of membrane CD43 mediates dendritic cell maturation. J. Immunol. 162: 6331-6336. Cyster, J. G., Somoza, C., Killeen, N. and Williams, A. F. 1990. Protein sequence and gene structure for mouse leukosialin (CD43), a T lymphocyte mucin without introns in the coding sequence. Eur. J. Immunol. 20: 875. Cyster, J. G., Shotton, D. M. and Williams, A. F. 1991. The dimensions of the T lymphocyte glycoprotein leukosialin and identification of linear protein epitopes that can be modified by glycosylation. EMBO J. 10: 893. 125  Daniel, M.T., Koken, M., Romagné, O., Barbey, S., Bazarbachi, A., Stadler, M., Guillemin, M.C., Degos, L., Chomienne, C. and de Thé, H. 1993. PML protein expression in hematopoietic and acute promyelocytic leukemia cells. Blood. 82: 1858-1867. Delon, J., Kaibuchi, K. and Germain, R. N. 2001. Exclusion of CD43 from the immunological synapse is mediated by phosphorylation-regulated relocation of the cytoskeletal adaptor moesin. Immunity. 15: 691-701. Desterro, J. M., Rodriguez, M. S. and Hay, R. T. 1998. SUMO-1 modification of IκBα inhibits NF- κB activation. Mol. Cell. 2: 233-239. Dowbenko, D., Kikuta, A., Fennie, C., Gillett, N. and Lasky L. A. 1993. Glycosylationdependent cell adhesion molecule 1 (GlyCAM 1) mucin is expressed by lactating mammary gland epithelial cells and is present in milk. J. Clin. Invest. 92: 952-960. Dransfield, I., Stocks, S. C. and Haslett, C. 1995. Regulation of cell adhesion molecule expression and function associated with neutrophil apoptosis. Blood. 85: 3264-3273. Drew, E., Merkens, H., Chelliah, S., Doyonnas, R. and McNagny, K. 2002. CD34 is a specific marker of mature murine mast cells. Exp Hematol. 30: 1211-1218. Drew, E., Merzaban, J. Seo, W., Ziltener, H. and McNagny, K. 2005. CD34 and CD43 Inhibit Mast Cell Adhesion and Are Required for Optimal Mast Cell Reconstitution. Immunity. 22: 4357. De Strooper, B. 2003. Aph-1, Pen-2, and nicastrin with presenilin generate an active gammasecretase complex. Neuron. 38: 9-12. Dragone, L. L., Barth, R. K., Sitar, K. L., Disbrow, G. L. and Frelinger, J. G. 1995. Disregulation of leukosialin (CD43, Ly48, sialophorin) expression in the B-cell lineage of transgenic mice increases splenic B-cell number and survival. P.N.A.S. 92:626-630. Ellies, L. G., Jones, A. T., Williams, M. J. and Ziltener, H. J. 1994. Differential regulation of CD43 glycoforms on CD4+ and CD8+ T lymphocytes in graft-versus-host disease. Glycobiology. 4: 885-893. Eskiw, C. H., Dellaire, G. and Bazett-Jones, D. P. 2004. Chromatin contributes to structural integrity of promyelocytic leukemia bodies through a SUMO-1-independent mechanism. J. Biol. Chem. 279: 9577-9585. Fanales-Balasio, E., Zambruno, G., Cavani, A. and Girolomoni, G. 1997. Antibodies against sialophorin (CD43) enhance the capacity of dendritic cells to cluster and activate T lymphocytes. Fortini, M. E. 2001. Notch and presenilin: a proteolytic mechanism emerges. Cur. Opin. Cell Biol. 13: 627-634.  126  Fratazzi, C., Manjunath, N., Arbeit, R. D., Carini, C., Gerken, T. A., Ardman, B., RemoldO’Donnell, E. and Remold, H. G. 2000. A macrophage invasion mechanism for mycobacteria implicating the extracellular domain of CD43. J. Exp. Med. 192: 183-191. Fuhlbrigge, R. C., King, S. L. and Sackstein, R. and Kupper, T. S. 2006. CD43 is a ligand for Eselectin on CLA+ human T cells. Fukuda, M. 1991. Leukosialin, a major O-glycan-containing sialoglycoprotein defining leukocyte differentiation and malignancy. Glycobiology. 1: 347-356. Gill, G. 2004. SUMO and ubiquitin in the nucleus: different functions, similar mechanism? Genes & Dev. 18: 2046-2059. Goddard, A. D., Borrow, J., Freemont, P. S. and Solomon, E. 1991. Characterization of a zinc finger protein gene disrupted by the t(15;17) in acute promyelocytic leukemia. Science. 254: 1371-1374. Gostissa, M., Hengstermann, A., Fogal, V., Sandy, P., Schwarz, S. E., Scheffner, M. and Del Sal, G. 1999. Activation of p53 by conjugation to the ubiquitin-like protein SUMO-1. EMBO J. 18: 6462-6471. Gutwein, P., Mechtersheimer, S., Riedle, S., Stoeck, A., Gast, D., Joumaa, S., Zentgraf, H., Forgel, M. and Altevogt, D. P. 2003. ADAM10-mediated cleavage of L1 adhesion molecule at the cell surface and in released membrane vesicles. FASEB J. 17: 292-294. Hardeland, U., Steinacher, R., Jiricny, J. and Schar, P. Modification of the human thymine-DNA glycosylase by ubiquitin-like proteins facilitates enzymatic turnover. 2002. EMBO J. 21: 14561464. Hart, S. P., Ross, J. A., Ross, K., Haslett, C. and Dransfield, T. 2000. Molecular characterization of the surface of apoptotic neutrophils: implications for functional downregulation and recognition by phagocytes. Cell. Death. Differ. 7: 493-503. He, Y-W. and Bevan, M. J. 1999. High level expression of CD43 inhibits T cell receptor/CD3mediated apoptosis. J. Exp. Med. 190: 1903-1908. Hemmerich, S., Leffler, H. and Rosen, S. D. 1995. Structure of the O-glycans in GlyCAM-1, an endothelial-derived ligand for L-selectin. J. Biol. Chem. 270: 12035-12047. Hernandez, J. D., Nguyen, J .T., He, J., Wang, W., Ardman, B., Green, J. M., Fukuda, M. and Baum, L .G. 2006. Galectin-1 binds different CD43 glycoforms to cluster CD43 and regulate T cell death. J. Immunol. 177: 5328-5336. Hietakangas, V., Ahlskog, J.K., Jakobsson, A.M., Hellesuo, M., Sahlberg, N.M., Holmberg, C.I., Mikhailov, A., Palvimo, J.J., Pirkkala, L., and Sistonen, L. 2003. Phosphorylation of serine 303 is a prerequisite for the stress-inducible SUMO modification of heat shock factor 1. Mol. Cell 127  Biol. 23: 2953-2968. Hofmann, T. G. and Will, H. 2003. Body language: the function of PML nuclear bodies in apoptosis regulation. Cell Death & Differentiaiton. 10: 1290-1299. Hoke, D., Mebius, R. E., Dybdal, N., Dowbenko, D., Gribling, P., Kyle, C., Baumhueter, S. and Watson, S. R. 1995. Selective modulation of the expression of L-selectin ligands by an immune response. Curr. Biol. 5: 670-678. Holmstrom, S., Van Antwerp, M.E., and Iniguez-Lluhi, J.A. 2003. Direct and distinguishable inhibitory roles for SUMO isoforms in the control of transcriptional synergy. Proc. Natl. Acad. Sci. 100: 15758-15763. Huang, T.T., Wuerzberger-Davis, S.M., Wu, Z.H., and Miyamoto, S. 2003. Sequential modification of NEMO/IKK by SUMO-1 and ubiquitin mediates NF-B activation by genotoxic stress. Cell 115: 565-576. Hwang, S. T., Singer, M. S., Giblin, P. A., Yednock, T. A., Bacon, K. B., Simon, S. I. and Rosen, S. D. 1996. GlyCAM-1, a physiologic ligand for L-selectin, activates 2 integrins on naive peripheral lymphocytes. J. Exp. Med. 184:1343-1348. Iba, K., Albrechtsen, R., Gilpin, B. J., Loechel, F. and Wewer, U. M. 1999. Cysteine-rich domain of human ADAM 12 (meltrin) supports tumor cell adhesion. Am. J. Pathol. 154: 1489– 1501. Johnson, G. G., Mikulowska, A., Butcher, E. C., McEvoy, L. M. and Michie, S. A. 1999. AntiCD43 monolonal antibody L11 blocks migration of T cells to inflamed pancreatic islets and prevents development of diabetes in nonobese diabetic mice. J. Immunol. 163: 5678-5685. Johnson, E.S. and Gupta, A.A. 2001. An E3-like factor that promotes SUMO conjugation to the yeast septins. Cell 106: 735-744. Jones, A. T., Federsppiel, B., Ellies, L. G., Williams, M. J., Burgener, R., Duronio, V., Smith, C. A., Takei, F. and Ziltener, H. J. 1994. Characterization of the activation-associated isoform of CD43 on murine T lymphocytes. J. Immunol. 153: 3426. Johnson, E. S. 2004. Protein modification by SUMO. Ann. Rev. Biochem. 73: 355-382. Kadaja, L., Laos, S. and Maimets, T. 2004. Overexpression of leukocyte marker CD43 causes activation of the tumor suppressor proteins p53 and ARF. Oncogene. 23: 2523-2530. Kim, K.I., Baek, S.H., and Chung, C.H. 2002. Versatile protein tag, SUMO: Its enzymology and biological function. J. Cell Physiol. 191: 257-268. Kimura, A. K. and Wigzell, H. 1978. Cell surface glycoproteins of murine cytotoxic T lymphocytes. J. Exp. Med. 147: 1418-1434.  128  Kishihara, K. et al. 1993. Normal B lymphocyte development but impaired T cell maturation in CD45- exon6 protein tyrosine phosphatase-deficient mice. Cell. 74: 143−156. Kiessling, L. L. and Gordon, E. J. 1998. Tranforming the cell surface through proteolysis. Curr. Biol. Chem. Biol. 5: R49-R62. Kopan, R. and Goate, A. 2000. A common enzyme connects Notch signaling and Alzheimer’s disease. Genes Dev. 15: 2799-2806. Krause, D. S., Fackler, M. J., Civin, C. I. and May, W. S. 1996. CD34: structure, biology, and clinical utility. Blood. 87: 1-13. Kudo, S and Fukuda, M. 1995. Tisssue-specific transcriptional regulation of human leukosialin (CD43) gene is achieved by DNA methylation. J. Biol. Chem. 270: 13298-13302. Lallemand-Breitenbach, V., Zhu, J., Puvion, F., Koken, M., Honore, N., Doubeikovsky, A., Duprez, E., Pandolfi, P. P., Puvion, E., Freemont, P. and de The, H. 2001. Role of promyelocytic leukemia (PML) SUMOlation in nuclear body formation, 11S proteasome recruitment, and As2O3-induced PML or PML/retinoic acid receptor alpha degradation. J. Exp. Med. 193: 1361– 1372. Lange, A., Mills, R. E., Lange, C. J., Stewart, M., Devine, S. E. and Corbett, A. H. 2007. Classical nuclear localization signals: definition, function, and interaction with importin α. J. Biol. Chem. 8: 5101-5105. Lasky, L. A., Singer, M. S., Dowbenko, D., Imai, Y., Henzel, E. J., Fennie, C., Gillett, N., Watson, S. R. and Rosen, S. D. 1993. An endothelial ligand for L-selectin is a novel mucin-like molecule. Cell. 69: 927-938. Li, S.J. and Hochstrasser, M. 1999. A new protease required for cell-cycle progression in yeast. Nature 398: 246-251. Ligtenberg, M. J. L., Kruijshaar, L., Buijs, F., van Meijer, M., Litvinov, S. V. and Hilkens, J. 1992. Cell-associated episialin is a complex containing two proteins derived from a common precursor. J. Biol. Chem. 267: 6171-6177. Lopez, S., Seveau, S., Lesavre, P., Robinson, M. K. and Halbwachs-Mecarelli, L. 1998. CD43 (sialophorin, leukosialin) shedding is an initial event during neutrophil migration, which could be closely related to the spreading of adherent cells. Cell. Adhes. and Commun. 5: 151-160. Mahajan, R., Delphin, C., Guan, T., Gerace, L. and Melchior, F. 1997. A small ubiquitin-related polypeptide involved in targeting RanGAP1 to nuclear pore complex protein RanBP2. Cell. 88: 97-107. Manjunath, N., Johnson, R. S., Staunton, D. E., Pasqualini, R. and Ardman, B. 1993. Targeted disruption of CD43 gene enhances T lymphocyte adhesion. J. Immunol. 151: 1528-1534. 129  Manjunath, N., Correa, M., Ardman, M. and Ardman, B. 1995. Negative regulation of T-cell adhesion and activation by CD43. Nature 377: 535. Marjaux, E. and Strooper, B. D. 2004. γ-secretase inhibitors: still in the running as Alzheimer’s therapeutics. Drug Discovery Today. 1: 1-6. Martin, D. P., Schmidt, R. E., DiStefano, P. S., Lowry, O. H., Carter, J. G. and Johnson, E. M. 1988. J. Cell Biol. 106: 829-844. Martin, S. J., Lennon, S. V., Bonham, A. M. and Cotter, T. G. 1990. J. Immunol. 145: 1859-1867. Matsumoto, M., Atarashi, K., Umemoto, E., Furukawa, Y., Shigeta, A., Miyasaka, M. and Hirata, T. 2005. CD43 functions as a ligand for E-selectin on activated T cells. J. Immunol. 175: 80428050. Matsumoto, M., Shigeta, A., Furukawa, Y., Tanaka, T., Miyasaka, M. and Hirata, T. 2007. CD43 collaborate E-selectin-dependent T cell migration into inflamed skin. J. Immunol. 178: 24992506. Matunis, M.J., Coutavas, E., and Blobel, G. 1996. A novel ubiquitin-like modification modulates the partitioning of the Ran-GTPase-activating protein RanGAP1 between the cytosol and the nuclear pore complex. J. Cell Biol. 135: 1457-1470. McEvoy, L. M., Sun, H., Frelinger, J. G. and Butcher, E. C. 1997. Anti-CD43 inhibition of T cell homing. J. Exp. Med. 185: 1493-1498. (1) McEvoy, L.M., Jutila, M. A., Tsao, P. S., Cooke, J. P. and Butcher, E. C. 1997. Anti-CD43 inhibits monocyte-endothelial adhesion in inflammation and atherogenesis. Blood. 90: 35873594. (2) McFarland, T. A., Ardman, B., Manjunath, N., Fabry, J. A. and Lieberman, J. 1995. CD43 diminishes susceptibility to T lymphocyte-mediated cytolysis. J. Immunol. 154: 1097-1104. Mentzer, S. J., Remold-O’Donnell, E., Crimmins, M. A. V., Bierer, B. E., Rosen, F. S. and Burakoff, S. J. 1987. Sialophorin, a surface sialoglycoprotein defective in the Wiscott-Aldrich Syndrome, is involved in human T lymphocyte proliferation. J. Exp. Med. 165: 1383-1392. Micheau, O. and Tschopp, J. 2003. Induction of TNF receptor I-mediated apoptosis via two sequential signaling complexes. Cell. 114: 181-190. Muller, S., Matunis, M.J., and Dejean, A. 1998. Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. EMBO J. 17: 61-70. Muller, S., Hoege, C., Pyrowolakis, G. and Jentsch, S. 2001. SUMO, ubiquitin’s mysterious cousin. Nat. Rev. Mol. Cell Biol. 2: 202–210.  130  Mumm, J. S. and Kopan, R. 2000. Notch signaling: from the outside in. Dev. Biol. 228: 151-165. Murakami, D., Okamoto, I., Nagano, O., Kawano, Y., Tomita, T., Iwatsubo, T., De Strooper, B., Yumoto, E. and Saya, H. 2003. Presenilin-dependent gamma-secretase activity mediates the intramembranous cleavage of CD44. Oncogene. 22: 1511-1516. Nacerddine, K., Lehembre, F., Bhaumik, M., Artus, J., Cohen-Tanoudji, M., Babinet, C., Pandolfi, P. P. and Dejean, A. 2005. The SUMO pathway is essential for nuclear integrity and chromosome segregation in mice. Dev. Cell. 9: 769–779 Nathan, C., Xie, Q-W., Halbwachs-Mecarelli, L. and Jin, W. W. 1993. Albumin inhibits neutrophil spreading and hydrogen peroxidase release by blocking the shedding of CD43 (sialophorin, leukosialin). J. Cell. Biol. 122: 243-256. Nieto, M., Rodriguez-Fernandez, J. L., Navarro, F., Sancho, D., Frade, J. M. R., Mellado, M., Martinez-A, C., Cabanas, C. and Sanchez-Madrid, F. 1999. Signalling through CD induces natural killer cell activation, chemokine release, and PYK-2 activation. Nong, Y., Remold-O’Donnell, E, Lebien, T. W. and Remold, H. G. 1989. A monoclonal antibody to sialophorin (CD43) induces homotypic adhesion and activation of human monocytes. J. Exp. Med. 170: 259-267. Nacerddine, K., Lehembre, F., Bhaumik, M., Artus, J., Cohen-Tannoudji, M., Babinet, C., Pandolfi, P. P. and Dejean, A. 2005. The SUMO pathway is essential for nuclear integrity and chromosome segregation in mice. Dev. Cell. 9: 769-779. Nagano, O. and Saya, H. 2004. Mechanism and biological significance of CD44 cleavage. Cancer Sci. 95: 930-935. Okamoto, I., Kawano, Y., Tsuiki, H., Sasaki, J-I., Nakao, M., Matsumoto, M., Suga, M., Ando, M., Nakajima, M. and Saya, H. 1999. CD44 cleavage induced by a membrane-associated metalloprotease plays a critical role in tumor cell migration. Oncogene. 1999a. 18: 1435-1446. Okamoto, I., Kawano, Y., Murakami, D., Sasayama, T., Araki, N., Miki, T., Wong, A. J. and Saya, H. 2001. Proteolytic release of CD44 intracellular domain and its role in the CD44 signaling pathway. J. Cell. Biol. 155: 755-762. Ostber, J. R., Dragone, L. L., Driskell, T., Moynihan, J. A., Phipps, R., Barth, R. K. and Frelinger, J. G. 1996. Disregulated expression of CD43 (leukosialin, sialophorin) in the B cell lineage leads to immunodeficiency. J. Immunol. 157: 4876-4884. Ostberg, J. R., Barth, R. K. and Frelinger, J. G. 1998. The Roman god Janus: a parakigm for the function of CD43. Immunol. Today. 19: 546-550. Pallant, A., Eskenazi, A., Mattei, M. G., Fournier, R. E. K., Carlsson, S. R., Fukuda, M. and Frelinger, J. G. 1989. Characterization of cDNAs encoding human leukosialin and localization of 131  the leukosialin gene to chromosome 16. P. N. A. S. 86: 1328-1332. Park, J. K., Rosenstein, Y. J., Remold-O’Donnell, E., Bierer, B .E., Rosen, F. S. and Burakoff, S. J. 1991. Enhancement of T-cell activation by the CD43 molecule whose expression in WiskottAldrich syndrome. Nature. 350: 706-709. Pedraza-Alva, G., Merida, L. B., Burakoff, S. J. and Rosenstein, Y. 1996. CD43-specific activation of T cells induces association of CD43 to Fyn kinase. J. Biol. Chem. 271: 2756427568. Pedraza-Alva, G., Merida, L. B., Burakoff, S. J. and Rosenstein, Y. 1998. T cell activation through the CD43 molecule leads to Vav tyrosine phosphorylation and mitogen-activated protein kinase pathway activation. J. Biol. Chem. 273: 14218-14224. Pedraza-Alva, G., Merida, L. B., Burakoff, S. J. and Rosentein, Y. 1998. T cell activation through the CD43 molecule leads to Vav tyrosine phosphorylation and mitogen-activated protein kinase pathway activation. J. Biol. Chem. 273: 14218-14224. Pedraza-Alva, G., Sawasdikosol, S., Liu, Y. C., Merida, B. M., Cruz-Munoz, M. E., OcegueraYanez, F., Burakoff, S. J. and Rosenstein, Y. 2001. Regulation of Cbl molecular interactions by the co-receptor molecule CD43 in human T cells. J. Biol. Chem. 276: 729-737. Penninger, J. M., Irie-Sasaki, J., Sasaki, T. and Oliveira-dos-Santos, A. J. 2001. CD45: new jobs for an old acquaintance. Nat. Immunol. 2: 389-396. Pichler, A., Gast, A., Seeler, J. S., Dejean, A. and Melchior, F. 2002. The nucleoporin RanBP2 has SUMO1 E3 ligase activity. Cell. 108: 109-120. Ponta, H., Sherman, L. and Herrlich, P. A. 2004. CD44: from adhesion molecules to signaling regulators. Nat. Rev. Mol. Cel. Biol. 4: 33-45. Remold-O’Donnell, E., Kenney, D. M., Parkman, R., Cairns, L., Savage, B. and Rosen, F. S. 1984. Characterization of a human lymphocyte surface sialoglycoprotein that is defective in Wiskott-Aldrich Syndrome. J. Exp. Med. 159: 1705-1723. Randhawa, A. K., Ziltener, H. J., Merzaban, J. S. and Stokes, R. W. 2005. CD43 is required for optimal growth inhibition of Mycobacterium tuberculosis in macrophages and in mice. J. Immunol. 175: 1805-1812. Remold-O’Donnell, E., Zimmerman, C., Kenney, D. and Rosen, F. S. 1987. Expression on blood cells of sialophorin, the surface glycoprotein that is defective in Wiskott-Aldrich Syndrome. Blood. 70: 104-109. Remold-O’Donnell, E. and Parent, D. 1994. Two proteolytic pathways for down-regulation of the barrier molecule CD43 of human neutrophils. J. Immunol. 152: 3595-3605.  132  Remold-O’Donnell, E and Parent, D. 1995. Specific sensitivity of CD43 to neutrophil elastase. Blood. 86: 2395-2402. Rieu, P., Porteu, F., Bessou, G., Lesavre, P. and Halbwachs-Mecarelli, L. 1992. Human neutrophils release their major membrane sialoprotein, leukosialin (CD43), during cell activation. Eur. J. Immunol. 22: 3021-3026. Rodriguez, M.S., Dargemont, C., and Hay, R.T. 2001. SUMO-1 conjugation in vivo requires both a consensus modification motif and nuclear targeting. J. Biol. Chem. 276: 12654-12659 Rosen, S. D. 1999. Endothelial ligands for L-selectin: from lymphoncyte recirculation to allograft rejection. J. Pathol. 155: 1013-1020. Rosenstein, Y., Park, J. K., Hahn, W. C., Rosen, F. S. and Burakoff, S. J. 1991. CD43, a molecule defective in Wiskott-Aldrich syndrome, binds ICAM-1. Nature. 354: 233-235. Rothwell, S. W. and Wright, D. G. 1994. Characterization of influenza A virus binding sites on human neutrophils. J. Immunol. 152: 2358-2367. Primakoff, P. and Myles, D. G. 2000. The ADAM gene family: surface proteins with adhesion and protease activity. Trends Genet. 16: 83-87. Sako, D., Comess, K. M., Barone, K. M., Camphausen, R. T., Cumming, D. A. and Shaw, G. D. 1995. A sulfated peptide segment at the amino terminus of PSGL-1 is critical for P-selectin binding. Cell. 83: 323–331. Salomoni, P. and Pandolfi, P. P. 2002. The role of PML in tumor suppression. Cell. 108: 165-170. Sanchez-Mateos, P., Campanero, M. R. del Poze, M. A. and Sanchez-Madrid, F. 1995. Regulatory role of CD43 leukosialin on integrin-mediated T cell adhesion to endothelial and extracellular matrix ligands and its polar redistribution to a cellular uropod. Blood. 86: 22282239. Sackstein, R. 2005. The lymphocyte homing receptors: gatekeepers of the multistep paradigm. Curr. Opin. Hematol. 12: 444–450. Satana, M. A., Pedraza-Alva, G., Olivares-Zavaleta, N., Madrid-Marina, V., Horejsi, V., Burakoff, S. J. and Rosenstein, Y. 2000. CD43-mediated signals induce DNA binding activity of AP-1, NFAT, and NFκB transcription factors in human T lymphocytes. J. Immunol. 275: 31460-31468. Schmid, K., Mao, S. K. Y., Kimura, A., Hayashi, S. and Binette J. P. 1980. Isolation and characterization of a serine-threonine-rich galactoglycoprotein from normal human plasma. J. Bio. Chem. 255: 3221-3226. Schmid, K., Hediger, M. A. and Brossmer R., 1992. Amino Acid Sequence of Human Plasma Galactoglycoprotein: Identity with the Extracellular Region of CD43 (Sialophorin). Proc. Natl. 133  Acad. Sci. U. S. A. 89: 663–667 Schroeter, E. H., Kisslinger, J. A. and Kopan, R. 1998. Notch-1 signaling requires ligand-induced proteolytic release of intracellular domain. Nature. 393: 382-386. Seeler, J-S. and Dejean, A., 2003. Nuclear and unclear functions of SUMO. Nat. Rev. Mol. Cell. Biol. 4: 690-699. Selkoe, D. J. 1998. The cell biology of beta-amyloid precursor protein and presenilin in Alzheimer's disease. Trends Cell. Biol. 8: 447–453. Serrador, J. M., Nieto, M., Alonso-Lebrero, J. L., del Pozo, M. A., Calvo, J., Furthmayr, H., Schwartz-Albiez, R., Lozano, F., Gonzalez-Amaro, R., Sanchez-Mateos, P. and Sanchez-Madrid, F. 1998. CD43 interacts with moesin and ezrin and regulates its redistribution to the uropods of T lymphocytes at the cell-cell contacts. Blood. 91: 4632-4644. Seufert, W., Futcher, B. and Jentsch, S. 1995. Role of a ubiquitin-conjugating enzyme in degradation of S- and M-phase cyclins. Nature. 373: 78-81. Seveau, S., Keller, H., Maxfield, F. R., Piller, F. and Halbwachs-Mecarelli, L. 2000. Neutrophil polarity and locomotion are associated with surface redistribution of leukosialin (CD43), an antiadhesive membrane molecule. Blood. 95: 2462-2470. Shiio, Y. and Eisenman, R.N. 2003. Histone sumoylation is associated with transcriptional repression. Proc. Natl. Acad. Sci. 100: 13225-13230. Sidoti-de, F. C., Rincheval, V., Risler, Y., Mignotte, B. and Vayssiere, J. L. 1998. TNF activates at least two apoptotic signaling cascades. Oncogene 17: 1639-1651. Sikut, R., Nilsson, O., Baeckstrom, D. and Hansson, G. C. 1997. Colon adenoma and cancer cells aberrantly express the leukocyte-associated sialoglycoprotein CD43. Biochem. Biophys. Res. Comm. 238: 612-616. Sikut, R., Andersson, C. X., Sikut, A., Fernandez-Rodriguez, J., Karlsson, N. G. and Hansson, G. C. 1999. Detection of CD43 (leukosialin) in colon adenoma and adenocarcinoma by novel monoclonal antibodies against its intracellular domain. Int. J. Cancer. 82: 52-58. Silverman, L. B., Wong, R. C. K., Remold-O’Donnell, E., Vercelli, D., Sancho, J., Terhorst, C., Rosen, F., Geha, R. and Chatila, T. 1989. Mechanism of mononuclear cell activation by an antiCD43 (Sialophorin) agonistic antibody. J. Immunol. 142: 4194-4200. Skubitz, K. M., Campbell, K. D. and Skubitz, A. P. N. 1998. CD43 is associated with tyrosine kinase activity in human neutrophils. J. Leuk. Biol. 64: 800-802. Sperling, A. I., Green, J. M., Mosley, R. L., Smith, P. L., DiPaolo, R. J., Klein, J. R., Bluestone, J. A. and Thompson, C. B. 1995. CD43 is a murine T cell costimulatory receptor that functions 134  independently of CD28. J. Exp. Med. 182: 139-146. Sperling, A. I., Sedy, J. R. Manjunath, N., Kupfer, A., Ardman, B. and Burkhardt, J. k. 1998. TCR signaling induces selective exclusion of CD43 from the T cell–antgen-presenting cell contact site. J. Immunol. 161: 6459-6462. Spicer, A. P., Rowse, G. J., Lidner, T. K. and Gendler, S. J. 1995. Delayed mammary tumor progression in Muc-1 null mice. J. Biol. Chem. 270: 30093-30101. Steiner, H. and Haass, C. 2000. Intramembrane proteolysis by presenilins. Nat. Rev. Mol. Cel. Biol. 1: 217-224. Stewart, M. 2007. Molecular mechanism of the nuclear protein import cycle. Nat. Rev. Mol. Cell. Biol. 8: 195-208. Stockl, J., Majdic, O., Kohl, P., Pickl, W. F., Menzel, J. E. and Knapp, W. 1996. Leukosialin (CD43)-Major Histocompatibility Class I molecule interactions involved in spontaneous T cell conjugate formation. J. Exp. Med. 184: 1769-1779. Stockton, B. M., Cheng, G., Manjunath, N., Ardman, B. and von Andrian, U. H. 1998. Negative regulation of T cell homing by CD43. Immunity 8: 373. Suzuki, A., Andrew, D. P., Gonzalo, J. A., Gonzalo, J. A., Fukumoto, M., Spellberg, J., Hashiyama, M. Takimoto, H., Gerwin, N., Webb, I., Molineux, G., Amakawa, R., Tada, Y., Wakeham, A., Brown, J., McNiece, I., Ley, K., Butcher, E. C., Suda, T., Gutierrez-Ramos, J. C. and Mak, T. W. 1996. CD34-deficient mice have reduced eosinophil accumulation after allergen exposure and show a novel crossreactive 90-kD protein. Blood. 87: 3550–3562. Tang, D., Lahti, J. M., Grenet, J. and Kidd, V. J. 1999. Cycloheximide-induced T-cell death is mediated by a Fas-associated death domain-dependent mechanism. J. Biol. Chem. 274: 72457252. Thurman, E. C., Walker, J., Jayaraman, S., Manjunath, N., Ardman, B. and Green, J. M. 1998. Regulation of in vitro and in vivo T cell activation by CD43. Intern. Immunol. 10: 691-701. Treon, S. P., Mollick, J. A., Urashima, M., Teoh, G., Chauhan, D., Ogata, A., Raje, N., Hilgers, J., Nadler, L., Belch, A. R., Pilarski, L. M. and Anderson,K.C. 1999. MUC1 core protein is expressed on multiple myeloma cells and is induced by dexamethasone. Blood. 93: 1287–1298. Tong, J., Allenspach, E. J., Takahashi, S. M., Mody, P. D., Park, C., Burkhardt, J. K. and Sperling, A. I. 2004. CD43 regulation of T cell activation is not through steric inhibition of T cell-APC interactions but through an intracellular mechanism. J. Exp. Med. 199: 1277-1283. Tsuchida, H., Takeda, Y., Takei, H., Shinzawa, H., Takahashi, T. and Sendo, F. 1995. J. Immunol. 157: 2403-2412.  135  Veerman, K. M. Williams, M. J., Uchimura, K., Singer, M. S., Merzaban, J. S., Naus, S., Carlow, D. A., Owen, P., Rivera-Vieves, J., Rosen, S. D. and Ziltener, H. J. 2007. Interaction of the selectin ligand PSGL-1 with chemokines CCL21 and CCL19 facilitates efficient homing of T cells to secondary organs. Nat. Immunol. 8: 532-539. Walker, J. and Green, J. M. 1999. Structural requiremens for CD43 function. J. Immunol. 162: 4109-4114. Wang, W., Link, V. and Green, J. M. 2000. Identification and cloning of a CD43-associated Serine/Threonine kinase. Cell. Immunol. 205: 34-39. Wang, Z. G., Delva, L., Gaboli, M., Rivi, R., Giorgio, M., Cordon-Cardo, C., Grosveld, F. and Pandolfi, P. P. 1998. Role of PML in cell growth and the retinoic acid pathway. Science. 279: 1547-1551. (1) Wang, Z.-G., Ruggero, D., Ronchetti, S., Zhong, S., Gaboli, M., Rivi, R. and Pandolfi, P.P., 1998. PML is essential for multiple apoptotic pathways. Nat. Genet. 20: 266–271. (2) Wiesmeijer, K., Molenaar, C., Bekeer, I. M., Tanke, H. J. and Dirks, R. W. 2002. Mobile foci of Sp100 do not contain PML: PML bodies are immobile but PML and Sp100 proteins are not. J. Struct. Biol. 140: 180-188. Wolfe, M. S. and Kopan, R. 2004. Intramembrane proteolysis: theme and variations. Science. 305: 1119-1123. Wong, R. C., Remold-O’Donnell, E., Vercelli, D., Sancho, J., Terhorst, C., Rosen, F., Geha, R. and Chatila, T. 1990. Signal transduction via leukocyte antigen CD43 (sialophorin). Feedback regulation by protein kinase C. J. Immunol. 144: 1455-1460. Woodman, R. C., Johnson, B., Hickery, M. J., Teoh, D., Reinhardt, P., Poon, B. Y. and Kubes, P. 1998. The functional paradox of CD43 in leukocyte recruitment: a study using CD43-deficient mice. J. Exp. Med. 188: 2181-2186. Wykes, M., MacDonald, K. P. A., Tran, M., Quin, R. J., Xing, P. X., Gendler, S. J., Hart, D. N. J. and McGuckin, M. A. 2002. Muc-1 epithelial mucin (CD227) is expressed by activated dendritic cells. J. immunol. 72: 692-701. Van den Berg, T. K., Nath, D., Ziltener, H. J., Vestweber, D., Fukuda, M., van Die, I. and Crocker, P. R. 2001. CD43 functions as a T cell counterreceptor for the macrophage adhesion receptor sialoadhesin (siglec-1). J. Immunol. 166: 3637-3640. Verger, A., Perdomo, J.. and Crossley, M. 2003. Modification with SUMO. A role in transcriptional regulation. EMBO Rep. 4: 137-142. Yang, J., Hirata, T., Croce, K., Merrill-Skoloff, G., Tchernychev, B., Williams, E., Flaumenhaft, R., Furie, B. C. and Furie, B. 1999. Targeted gene disruption demonstrates that P-selectin 136  glycoprotein ligand 1 (PSGL-1) is required for P-selectin-mediated but not E-selectin-mediated neutrophil rolling and migration. J Exp Med 190: 1769–1782. Yonemura, S., Hirao, M., Doi, Y., Takahashi, N., Kondo, T., Tsukita, S. and Tsukita, S. 1998. Ezrin/radixin/moesin (ERM) proteins bind to a positively charged amino acids cluster in the juxta-membrane cytoplasmic domain of CD44, CD43, and ICAM-2. J. Cell. Biol. 140: 885-895. Zhong, S., Hu, P., Ye, T. Z., Stan, R., Ellis, N. A. and Pandolfi, P. P. 1999. A role for PML and the nuclear body in genomic stability. Oncogene 18: 7941-7947. Zhong, S., Salomoni, P. and Pandolfi, P. P. 2000. The transcriptional role of PML and the nuclear body. Nat. Cell Biol. 2: E85-E90.  137  Appendix I: Map of pMPSF  138  Appendix II: Constructs made to dissect the mechanisms of shedding of CD43  139  Appendix III: LIST OF PUBLICATIONS 1. Morozova, T., Seo, W. and Zimmerly, S. 2002. Non-cognate Template Usage and Alternative Priming by a Group II Intron-encoded Reverse Transcriptase. J. Mol. Biol, 315: 951-963. 2. Robart, A. R., Seo, W. and Zimerly, S. 2007. Insertion of group II intron retroelements after intrinsic transcriptional terminators. P.N.A.S. 104: 6620-6625. 3. Drew, E., Merzaban, J. S., Seo, W., Ziltener, H. J. and McNagny, K. M. 2004. CD34 and CD43 inhibit mast cell adhesion and are required for optimal mast cell reconstitution. Immunity, 22: 43-57. The first two publications were from M.Sc works. The third publication is not directly related to this thesis.  140  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items