PREFACE
To some people, the study of cancer and autoimmunity may be something only slighdy less sterile than a Johnson...
34 downloads
1456 Views
34MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
PREFACE
To some people, the study of cancer and autoimmunity may be something only slighdy less sterile than a Johnson & Johnson gauze pad, a pursuit followed by dilettanti and pseudo-scholastic professors. To others, the word connotates the exotic, the esoteric and the desirable. To us, it embodies the hopes and ambitions of generating an anti-tumor response in the patient. It is somewhat ironic that over forty years ago there were two disciplines that started with "TI". There was tumor immunology and there was transplantation immunology. The latter thrived and has led to some of the most critical discoveries in immunobiology. The former continues to thwart bench scientists and clinicians alike. In fact, the original hope that cancer cells would each contain a novel antigen that might be recognized by the immune system has proven, for the most part, to be naive. On the other hand, it was perhaps equally naive to assume that a process as biologically conserved as neoplasia, would lead to the production of something as simple as a unique antigen that would be common to all patients. The work, however, on tumor immunology has been productive and has led to interrelationships between the molecular processes of neoplastic development and the understanding of the phenotypic changes which occur. These changes which were once considered to be only involved in cell surface markers, now encompass the disciplines of signal transduction, apoptosis, and differentiation. As immunologists, our goal is to develop a simple and effective means to manipulate cancer in vivo. This manipulation can encompass several venues. First, it might be as direct as the original hope and aspiration of identifying a phenotypic marker and the use of either active or passive immunization. Second, it might include the use of passive reagents carrying "warheads" to selectively destroy cancer cells. Third, it might include altering the basic process of cell survival, be it via nucleic acid or protein biosynthesis, or programmed cell death. The list goes on and on as the black box gets bigger and bigger. In fact, we used to teach our students that the immune system was little more than a large black box, except that upon opening the box, one only discovered multiple smaller black boxes, and so on. This volume is an attempt by a collection of workers in many disciplines, to present a theme which has not been well described before. The papers include both basic and clinical science and range from sophisticated molecular biology to little more than phenomenology (e.g., the increased association of cancer in some autoimmune diseases and increased presentation of autoimmune phenomena in malignant conditions). This, however, is state-of-the-art. Our hope is that this collection of themes will be of use not only to bench scientists, but also to clinicians who treat patients. We also expect that as we enter the millenium, that much of this work will become an anachronism. The latter of course would be a great success and would imply real progress. In fact, as we finish this volume, the editors realize more than anything else the need to update this book 5-10 years hence. We greatly appreciate the help of our contributors. We have done our best to edit the manuscripts. The errors which remain are ours alone. Y.S. and M.E.G.
Vll
List of Contributors
Mahmud Abu-Shakra Rheumatic Diseases Unit and Department of Medicine 'B' & 'D' Soroka Medical Center and Ben-Gurion University P.O. Box 151 Beer-Sheva 84101 Israel Donato Alarcon-Segovia Department of Immunology and Rheumatology Institute Nacional De La Nutricion Salvador Zubiran Mexico City Mexico C. Alessandri University di Roma "La Sapienza" Policlinico Umberto I Clinica Medica I 00161 Roma Italy
Antonio Bandeira Instituto Gulbenkian de Ciencia Oeiras Portugal Yaron Bar-Dayan INSERM Unite 430 Hopital Broussais 96, Rue Didot 75014 Paris Cedex 14 France Yosefa Bar-Dayan INSERM Unite 430 Hopital Broussais 96, Rue Didot 75014 Paris Cedex 14 France
Anabel Aron-Maor Department of Medicine 'B' Research Unit of Autoimmune Diseases Chaim Sheba Medical Center (Affiliated to Tel-Aviv University) Tel-Hashomer 52621 Israel
Eytan R. Barnea The S.I.E.P Division of Research Scientific Secretariat and Registration 1697 Lark Lane, Cherry Hill New Jersey 08003-3157 USA
Ronald A. Asherson The Rheumatic Diseases Unit Department of Medicine University of Cape Town School of Medicine The Groote Schuur Hospital Observatory 7925 Cape Town 8001 South Africa
Narayan K. Bhat Center for Molecular and Structural Biology HoUings Cancer Center Medical University of South Carolina 17 Ashley Avenue Charleston, S.C. 29425-2213 USA
Emmanuelle Bonnin INSERM Unite 430 Hopital Broussais 96, Rue Didot 75014 Paris Cedex 14 France
Dan Buskila Rheumatic Disease Unit Department of Medicine 'B' Soroka Medical Center and Faculty of Health Science Ben-Gurion University of the Negev P.O. Box 151 Beer-Sheva 84101 Israel
Mary C. Cantrell Division of Rheumatology/ Allergy and Clinical Immunology University of California at Davis TB 192, School of Medicine Davis, CA 95616-8660 USA
Carlos A. Casiano Department of Microbiology and Molecular Genetics Loma Linda University School of Medicine Loma Linda, CA 922350 USA Ricard Cervera Unitat de Malalties Autoimmunes Sistemiques Hospital Clinic, Villarroel 170 Barcelona 08036 Catalonia Spain Karsten Conrad Institute for Immunology Medical Faculty Technical University of Dresden Karl Marx Str. 3 PO. Box 8001 15 D-01101 Dresden Germany
Fabrizio Conti University di Roma "La Sapienza" Policlinico Umberto I Clinica Medica I 00161 Roma Italy Antonio Coutinho Department of Immunology Pasteur Institute 25 Rue du Docteur Roux 75724 Paris Cedex 15 France Sidney Croul Center for Neurovirology and Neurooncology MCP Hahnemann University 245 N. 15th Street, Mail Stop #406 Philadelphia, PA 19102 USA David D'Cruz Bone & Joint Research Unit The Royal London Hospital 25-29 Ashfield Street Whitechapel London El 2AD England Jocelyne Demengeot Unite du Developpement des Lymphocytes CNRS URA 1961 Institut Pasteur Paris France Luis Del Valle Center for Neurovirology and Neurooncology MCP Hahnemann University 245 N. 15th Street, Mail Stop #406 Philadelphia, PA 19102 USA Smruti A. Desai New York Medical College Department of Microbiology & Immunology Valhalla, New York 10505 USA
Guilliam Dighiero Institut Pasteur Unite d'Immuno-Hematologie et d' Immunopathologie 28 rue du Dr Roux F-75724 Paris Cedex 15 France
Josep Font Unitat de Malalties Autoimmunes Sistemiques Hospital Clinic, Villarroel 170 Barcelona 08036 Catalonia Spain
Lea Eisenbech Department of Immunology Weizmann Institute of Science Rheovot76100 Israel
Mario Garcia-Carrasco Unitat de Malalties Autoimmunes Sistemiques Hospital Chnic, Villarroel 170 Barcelona 08036 Catalonia Spain
Khaled M. El-Shami Department of Immunology Weizmann Institute of Science Rheovot76100 Israel
Jacob George Department of Medicine 'B' and The Research Unit of Autoimmune Diseases Chaim-Sheba Medical Center Tel-Hashomer 52621 Israel
Felix Fernandez-Madrid Department of Internal Medicine Division of Rheumatology and Center for Molecular Medicine and Genetics Wayne State University 4707 St. Antoine Detroit MI 48201 USA Soldano Ferrone New York Medical College Department of Microbiology & Immunology Valhalla, New York 10505 USA Claire Fieschi Medecin des Hopitaux Groupe Hospitalier Pitie—Salpetiere 47-83 Bd de I'Hopital 75651 Paris Cedex 13 France Heiko T. Flammann Institute of Immunology Pathology and Molecular Biology Lademannbogen 61 D-22339 Hamburg Germany
Panagiotis Georgiou Center for Molecular and Structural Biology Hollings Cancer Center Medical University of South Carolina 17 Ashley Avenue Charleston, S.C. 29425-2213 USA Eric M. Gershwin Division of Rheumatology/ Allergy and Clinical Immunology University of California at Davis TB 192, School of Medicine Davis, CA 95616-8660 USA Pascal Godmer Department of Internal Medicine Hopital Avicenne Universite Paris-Nord 125, Rue de StaUngrad 93000 Bobigny France Jennifer Gordon Center for Neurovirology and Neurooncology MCP Hahnemann University 245 N. 15th Street, Mail Stop #406 Philadelphia, PA 19102 USA
XI
Wolfgang L. Gross University Poliklinik fiir Rheumatologie, Lubeck Ratzeburger Allee 160 23538 Lubeck And Akad. Lehrkrankenhaus der Rheumaklinik B.B. Postfach 1488 Bramstedt 24572 Germany
Michel D. Kazatchkine INSERM Unite 430 Immunopathologie Humaine Hopital Broussais 96 Rue Didot 75014 Paris Cedex 14 France
Loic Guillevin Department of Internal Medicine Hopital Avicenne Universite Paris-Nord 125, Rue de Stalingrad 93000 Bobigny France
Kamel Khalili Center for Neurovirology and Neurooncology MCP Hahnemann University 245 N. 15th Street, Mail Stop #406 Philadelphia, PA 19102 USA
Michael Heike Johannes Gutenberg-Universitat Mainz Klinikum I. Medizinische Klinik und Poliklinik Das Klinikum befindet sich in der LangenbeckstraBe 1 55131 Mainz Germany
Arnoldo Kraus Department of Immunology and Rheumatology Institute Nacional De La Nutricion Salvador Zubiran Mexico City Mexico
Shohei Hori Department of Immunology Pasteur Institute 25 Rue du Docteur Roux 75724 Paris Cedex 15 France C. Jamin Centre Hospitaluer Universitaire Laboratoire d'Immunologic B.R 824 F-29609 Brest Cedex France Viggo J0nsson Department of Autoimmunology Statens Serum Institiit 5 Artillerivej 2300 Copenhagen 5 Denmark Srinivas Kaveri INSERM Unite 430 Hopital Broussais 96, Rue Didot 75014 Paris Cedex 14 France
xn
Barbara Krynska Center for Neurovirology and Neurooncology MCP Hahnemann University 245 N. 15th Street, Mail Stop #406 Philadelphia, PA 19102 USA Hella-Monika Kuhn Israelitic Hospital Orchideenstieg 14 D-22297 Hamburg Germany Nitza Lahat Immunology Research Unit Carmel Medical Center 7, Michal Str. Haifa 34362 Israel Steven P. Levine The S.I.E.P Division of Research Scientific Secretariat and Registration 1697 Lark Lane, Cherry Hill New Jersey 08003-3157 USA
Peter M. Lydyard Centre Hospitaluer Universitaire Laboratoire d'Immunologie B.R 824 F-29609 Brest Cedex France
Jessica Otte Center for Neurovirology and Neurooncology MCP Hahnemann University 245 N. 15th Street, Mail Stop #406 Philadelphia, PA 19102 USA
loanna Maroulakou Center for Molecular and Structural Biology Rollings Cancer Center Medical University of South Carolina 17 Ashley Avenue Charleston, S.C. 29425-2213 USA
Takis S. Papas Center for Molecular and Structural Biology HoUings Cancer Center Medical University of South Carolina 17 Ashley Avenue Charleston, S.C. 29425-2213 USA
Karl-Herman Meyer zum Buschenfelde Office: I. Med. Klinik und Poliklinik Universitat Mainz D55131 Germany
J.O. Pers Centre Hospitaluer Universitaire Laboratoire d'Immunologie B.P 824 F-29609 Brest Cedex France
Ariel Miller Immunology Research Unit Carmel Medical Center 7, Michal Str. Haifa 34362 Israel Mathias Montenarh Universitat des Saarlandes Medizinische Biochemie und Molekularbiologie Geb. 44 66421 Hamburg Germany Arnon Nagler Department of Bone Marrow Transplantation Hadassah University Hospital Ein Karem Jerusalem 91120 Israel Elvyra J. Noronha New York Medical College Department of Microbiology & Immunology Valhalla, New York 10505 USA
Jean-Charles Piette Medecin des Hopitaux Groupe Hospitaller Pitie—Salpetiere 47-83 Bddel'Hopital 75651 Paris Cedex 13 France Miloslav Pospisil Academy of Science of Czech Republic Prague Czech Republic Sonja Praprotnik University Medical Center Ljubliana Department of Rheumatology Vodnikova 62 1000 Ljubliana Slovenia Nagenda Prasad INSERM Unite 430 Hopital Broussais 96, Rue Didot 75014 Paris Cedex 14 France
Xlll
Roberta Priori University di Roma "La Sapienza" Policlinico Umberto I Clinica Medica I 00161 Roma Italy
Marc Scbmitz Institute of Immunology Medical Faculty Technical University of Dresden Karl-Marx-Str. 3 D-01109 Dresden Germany
Thomas P. Prindiville Division of Rheumatology/ Allergy and Clinical Immunology University of California at Davis TB 192, School of Medicine Davis, CA 95616-8660 USA
Yaniv Sberer Department of Internal Medicine 'B' Sheba Medical Center Tel Hashomer 52621 Israel
O. Pritsch Institut Pasteur Unite d'Immuno-Hematologie et d'Immunopathologie 28 rue du Dr Roux F-75724 Paris Cedex 15 France
Yehuda Shoenfeld Department of Medicine 'B' Research Unit of Autoimmune Diseases Chaim Sheba Medical Center (Affiliated to Tel-Aviv University) Tel-Hashomer 52621 Israel
Michal A. Rabat Immunology Research Unit Carmel Medical Center 7, Michal Str. Haifa 34362 Israel
Emanuel Sikuler Department of Medicine 'B' Research Unit of Autoimmune Diseases Chaim Sheba Medical Center (Affiliated to Tel-Aviv University) Tel-Hashomer 52621 Israel
Manel Ramos-Casals Unitat de Malalties Autoimmunes Sistemiques Hospital Clinic, Villarroel 170 Barcelona 08036 Catalonia Spain Ernst Peter Rieber Institute of Immunology Medical Faculty Technical University of Dresden Karl-Marx-Str. 3 D-01109 Dresden Germany Jozef Rovensky Research Institute of Rheumatic Diseases Nabrezie I. Krasku 4 921 01 Piestany Slovak Rep.
XIV
Shimon Slavin Department of Bone Marrow Transplantation Hadassah University Hospital Ein Karem Jerusalem 91120 Israel Sooryanarayana INSERM U430 Hospital Broussais Sheba Medical Center Tel Hashomer Israel Renata Stepankova Academy of Science of Czech Republic Prague Czech Republic
Efstratios Tatsis University Poliklinik ftir Rheumatologie, Lubeck Ratzeburger Allee 160 23538 Lubeck And Akad. Lehrkrankenhaus der Rheumaklinik B.B. Postfach 1488 Bramstedt 24572 Germany Moshe Tishler Department of Rheumatology Tel Aviv Souraski Medical Center Tel-Aviv University Sackler School of Medicine 6 Weiman St. Tel Aviv 64239 Israel Helena Tlaskalova Academy of Science of Czech Republic Prague Czech Republic Ludmila T\ickova Academy of Science of Czech Republic Prague Czech Republic Yaron Tomer Department of Endocrinology Box 1055 Mount Sinai Medical Center One Gustave L. Levy Place New York, N.Y. 10029 USA
Alena Tbchyiiova Research Institute of Rheumatic Diseases Nabrezie I. Krasku 4 921 01 Piestany Slovak Rep. Guido Valesini University di Roma "La Sapienza" Policlinico Umberto I Clinica Medica I 00161 Roma Italy F. Viganego University di Roma "La Sapienza" PoHcHnico Umberto I CUnica Medica I 00161 Roma Italy Antonio R. Villa Department of Immunology and Rheumatology Institute Nacional De La Nutricion Salvador Zubiran Mexico City Mexico Xinhui Wang New York Medical College Department of Microbiology & Immunology Valhalla, New York 10505 USA
J. Tomkiel Department of Internal Medicine Division of Rheumatology and Center for Molecular Medicine and Genetics Wayne State University 4707 St. Antoine Detroit MI 48201 USA
Dennis K. Watson Center for Molecular and Structural Biology Rollings Cancer Center Medical University of South Carolina 17 Ashley Avenue Charleston, S.C. 29425-2213 USA
M. Tomsic University Medical Center Ljubliana Department of Rheumatology Vodnikova 62 1000 Ljubliana Slovenia
Allan Wiik Department of Autoimmunology Statens Serum Institiit 5 Artillerivej 2300 Copenhagen 5 Denmark
XV
Joerg Willers New York Medical College Department of Microbiology & Immunology Valhalla, New York 10505 USA Pierre Youinou Centre Hospitaluer Universitaire Laboratoire d'Immunologic B.R 824 F-29609 Brest Cedex France
XVI
Dongsheng Zhang New York Medical College Department of Microbiology & Immunology Valhalla, New York 10505 USA
(c) 2000 Elsevier Science B.VAll rights reserved. Cancer and Autoimmunity Y. Shoenfeld and M. E. Gershwin, editors
Introduction: The Immune System, the Autoimmune State and Autoimmune Disease Jacob George and Yehuda Shoenfeld Department of Medicine 'B' and the Research Unit of Autoimmune Diseases, Sheba Medical Center, Tel-Hashomer, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel
1. INTRODUCTION Autoimmune diseases stand as important causes of morbidity and mortality in western society and as such, they impose a heavy burden in financial terms. The significance of autoimmune diseases can be demonstrated by the study showing that half of the patients with RA are unemployed due to medical disabilities resulting from their illness [1]. Similarly, patients with insulin dependent diabetes mellitus, an additional autoimmune disease, who today have a longer life expectancy, increasingly require supporting medical facilities such as dialysis. The pathogenesis and the relative importance of factors leading to autoimmunity are not defined precisely. Key questions include: what is the origin of autoantibodies, thought to play a detrimental role in autoimmunity, their interrelations with the autoantigens to which they are directed and the influence of this interplay on the immune system. The enigma is further intensified by the detection of natural autoantibodies found in healthy organisms and thought to possess regulatory and protective properties. Moreover, the cellular immune response has similarly been shown to participate in the evolvement of autoimmunity through its principal effector—the T-cell. However, the initiating events rendering these cells autoreactive and therefore capable of precipitating damage to cell-structures and the precise nature of this reaction still await comprehensive elucidation.
2. THE ESSENTIALS OF THE IMMUNE RESPONSE The principal role of the immune system is to confer protection on the organism against foreign invading pathogens, which can gain access to the body by different routes. The targeted (specific) immune response is generated by the combined interaction of the cellular and humoral responses, both coordinated by the production of active substances—the cytokines. Cellular immunity refers to the immune mechanisms mediated by T lymphocytes, regardless of immunoglobulin molecules, whereas humoral immunity denotes secretion of antibodies by B lymphocytes. The humoral and cellular arms of the immune system should be viewed, not as two independent mechanisms of self defense, but rather as acting in an orchestrated and synergistic manner to accomplish protection [2]. 2.1. Humoral Immune Response B-cells stem from bone marrow precursors and later localize in the circulation as well as in folHcles of peripheral lymphoid tissues. They are responsible for the production of antibodies, once they have differentiated into plasma cells [3]. B-lymphocytes are also involved in antigen presentation to T-cells, secretion of immunoregulating cytokines and establishment of 'memory' towards antigenic determinants [4]. Direct activation of B-cells by distinct antigens is facilitated by binding to antigen receptors located within the membrane of the B-lymphocytes and under the influence of cytokines. This complex interac-
tion activates B-cells after which they proliferate and produce the appropriate antibody. The final function of the B-cell (i.e., memory cell, plasma cell or cytokine secreting cell) is determined by the profile of the cytokines present, and the mechanisms of activation (through B-cell receptors for the Fc region of IgG, or for complement components) [4]. The end product following antigenic stimulation is a population of B-cells producing and secreting one specific antibody against the introduced antigen. Immunoglobulins are glycoproteins forming 9 classes of isotypes: IgG, divided to 4 subclasses (IgGI-4), IgM, IgA comprising 2 subclasses (IgAl-2), IgD and IgE (Tables 1) [5]. The basic structure of immunoglobulins (similar in all five isotypes) consists of two identical heavy chains (MW 50,000-75,000) combined with two identical light chains (MW 25,000). Antigen specificity is determined by variable areas containing the antigen binding site, whereas the constant region (as can be inferred from its name), is common to all immunoglobulins of a certain class. The hypervariable region is located in the variable region, representing the closest relationship to the epitope (its corresponding site on the antigen). The idiotypes, located in the variable region are the antigenic determinants (defining antigen binding) of the immunoglobulins themselves. The diversity of antibody response is formulated due to encoding of the heavy and light chains by multiple genetic elements. As such, light chains are generated following pairing of VK and JK genes, whereas heavy chains exhibit greater diversity since they are created following the assembly of three germline genes (VH, DH, JH). 2.2. Cellular Immune Response A T-cell cycle is initiated in hematopoietic stem cells, differentiating in the thymus and subsequently wandering to the lymphoid tissue in the periphery [2, 6]. T-cells are heterogeneous by virtue of their different functions (lysis of foreign cells, modulation of the interaction between B and T cells, regulation of monocyte functions). The peripheral T-cells are discerned by their expression of antigenic markers. As such, Tcells carrying CD4 molecules (T-helpers) interact with antigen associated with MHC class II on the surface of the antigen presenting cell, and T-cells expressing
CDS molecules (cytotoxic T-cells) engage in suppression of the immune response. The T-cell receptor is a molecule present on the surface of the T-cell, responsible for recognition of the complex antigen-MHC II molecule [7]. The variety of T-cell receptors is immense, thus accounting for its ability to recognize diverse antigens. The immune response is mounted following presentation of the antigen to the lymphocytes by antigen presenting cells, examples of which are: macrophages, Langerhans cells and dendritic cells. The process of presentation requires the participation of MHC class II molecules on the surface of the antigen presenting cell. The antigen, prior to its presentation to the T-cell is processed and degraded and later associated with the MHC class II molecule to form a complex reacting with the T-cell receptor. It should be outlined that the APCs are capable of secreting cytokines that act to facilitate the interaction described above. 2.3. Coordination of the Immune Response Several intrinsic factors belonging to the immune system itself are responsible for the modulation and regulation of the immune response. Cytokines are small proteins (MW 8000-30,000) produced and secreted by a diverse population of cells (i.e., macrophages, monocytes, T and B cells, as well as nonlymphoid cells) [8]. Cytokines elicit different actions (Table 2) including proinflammatory (TNF, IL-1, IL-2) and anti-inflammatory (TGF, IL-4, IL6, IL-10) functions and stimulation of lymphocyte proliferation (IL-2, IL-7). The regulation of cytokines is under the supervision of genetic factors (capable of generating corresponding inhibitors) and by the liberation of soluble forms of cytokine receptors. The complement system consists of circulating glycoprotein constituents that can be triggered and activated in two major patterns to initiate a cascadic chain of events, the consequence of which leads to diverse influences on the progression and perpetuation of the immune response [9]. This cascade can, therefore, be activated by the classical pathway (immune complexes comprising IgM and IgG) or by the alternative pathway—independent of antibodies (by bacterial LPS). The idiotypic system—will later be reviewed in detail.
Table 1. Characteristics of human immunoglobuhn subclasses Characteristic
IgG
IgM
IgA
IgD
IgE
Molecular form Molecular weight Subclasses Serum half life(days) Valence Serum concentration (mg/dl) Sedimentation constant Percentage of serum immunoglobulins Placental transfer
Monomer 160,000 1,2,3,4 23 2 1000-1500 7S 75-85
Pentamer, hexamer 900,000 None 5.1 10,12 100-150 19S 5-10
Monemer, dimer 170,000 1,2 5.8 2,4 250-300 7S(9, 11, 13) 7-15
Monomer 180,000 None 2.8 2 0.3-30 7S 0.3
Monomer 190,000 None 2.3 2 0.0015-0.2
-H
-
-
-
-
8S 0.0003
Table 2. Representative cytokins and their corresponding biological activities Cytokine
Activities
lL~a, IL-^
Lymphocyte acivation; bone resorption, induction of fibroblasts synovial cells and endothelial cells; prostaglandin liberation. T and B growth factor; increased secretion of several cytokines; activation of cytotoxic cells. Proliferation of marrow stem cells; growth factor for: macrophages, eosinophils, mast cells. Activation of B-cells and macrophages; stimulated proliferation of T-cells and mast cells; Induce secretion of IgE by B-cells. Induce antibody production and acute phase protein production by the hepatocytes. Inhibit production of several cytokines Decrease cell replication; Increases MHC class I replication; disrupts viral replication. Activate NK cells, cytotoxic T cells, endothelial cells and macrophages; anti-tumoral effects; Increase expression of MHC class I and II. Acute phase reactant; anti-tumoral; activate macrophages; increase expression of MHC class I; bone resorption. Inhibit IL-1; enhance tissue repair; suppress lymphocyte proliferation.
IL-2 IL-3 IL-4 IL-6 IL-10 IFN-a IFN-y TNF-a TGF-)^
Suppressor T-cells. As can be recalled, suppressor T-cells constitute a distinct subset of T-cells in charge of down-regulating the expression of either T cells and immunoglobulin secreting cells [10].
3. EVOLUTION OF THE AUTOIMMUNITY CONCEPT Paul Ehrlich was the first to coin the term autoimmunity [11] with regard to the harmful aspects of immunity, namely—the emergence of autoantibodies directed against the organism's own antigens. However, the expression used ('horror autotoxicus') has served to denote a mechanism avoiding autoimmunization, exemplified in goat models (producing alloantibodies but not autoantibodies).
The revolutionary ideas expressed by Ehrlich were subsequently abandoned for a century although anecdotal works confirming his notions were sporadically reported. The turning point, leading to the general acceptance of the autoimmunity concept was the experiments by Witebski & Rose (reviewed in Reference [12]) showing that rabbits immunized with rabbit thyroglobulin developed thyroiditis following production of anti-thyroglobulin autoantibodies. These observations were supported by the models of autoimmune hemolytic anemia and thrombocytopenia in which anti-red blood cell antibodies were detected and had been shown to be associated with bouts of hemolysis and thrombocytopenia [13]. The discovery of the NZB mouse (a strain which develops spontaneous autoimmune disease) provided a new tool for the study of autoimmmunity, con-
firming the previous evolving notions regarding the abiUty of the immune system to attack its own inherent constituents. Further progress towards better understanding of autoimmunity has been achieved by the reaUzation that the abihty to distinguish between self and non-self during fetal life is a prerequisite for normal function of the immune system. The term tolerance was introduced to signify the lack of autoreactivity. These ideas were later extended by Burnet [14], assuming that autoimmunization results from the emergence of 'forbidden' clones (lymphocytes possessing receptors for autoantigens escaping 'normal' deletion by the thymus during fetal life). Although this idea was later neglected, it represents the basis for the modern approach regarding autoimmunity as failure of the immune system to recognize its intrinsic components as its own. Different methods of classification have been proposed for autoimmune diseases. An accepted method is based upon the organs afflicted (Table 3). Accordingly, diseases involving multiple organs include: SLE, rheumatoid arthritis, Sjogren's syndrome and scleroderma, whereas examples of organ-specific diseases encompass: Graves' disease, Hashimoto's thyroiditis, myasthenia gravis, polymyositis, pernicious anemia, Addison's disease, IDDM, primary billiary cirrhosis, autoimmune hemolytic anemia and pemphigus.
4. DYNAMIC PRESERVATION OF IMMUNOLOGIC TOLERANCE 4.1. The Nature of Lymphocyte Repertoire Selection The appearance of tolerance, a key concept in autoimmunity, is not yet fully elucidated. It is probably established by processes which involve the selection of lymphocytes in the thymus. The precise nature of the signal, determining the selection of lymphocytes is likely to reside in a small peptide (9 amino acid) either endogenous or extrinsic [15] which is contained in the MHC molecule expressed by thymic epithelial cells or macrophages. The importance of this peptide lies in its ability to direct the lymphocye to the 'negative selection' pathway (leading to its death or to its entry into a dormant state)
and subsequently to the 'positive selection' destiny terminating in its differentiation into a mature T cell. The immunoregulatory properties of the peptide are expressed by its ability to influence (by the mechanism previously described) the population of helper and cytotoxic T cells, thereby defining the nature and aim of the immune response. One of the key questions regarding the developmental aspects of the immune system relates to the signaling events which determine the fate of selection of its inherent constituents (positive versus negative selection). Moreover, it is clear that in comparison to an adult animal, the fetus and newborn are highly susceptible to the induction of immunological tolerance. A factor contributing to the quality of selection is the affinity of the interaction TCR-MHC-peptide complex. Thus, progression of the T-cell towards apoptotic death (negative selection), or to differentiation into a mature T-lymphocyte is governed by the strength of the complex (T-cell-epithelial cell) affinity. Furthermore, it has been suggested that macrophages exert clonal deletion, whereas epithelial cells enhance clonal anergy (entry of the T-cell into a dormant state). 4.2. Programmed Cell Death (Apoptosis) The high turnover of human lymphocytes necessitates the existence of a control mechanism that would provide means of deleting these cells (the autoreactive cells in particular) in a coordinated manner. Apoptosis indeed fulfiUs such requirements by leading to nuclear fragmentation and ingestion by macrophages. This process of rapid cell elimination does not result in tissue inflammation thereby differentiating it from cell death by necrosis which evokes a considerable inflammatory response [16]. The recognition of apoptosis provided a compelling insight into the possible emergence of autoimmune disorders as a result of its defective development. Two genes play major roles regarding the evolution of autoimmunity in association with programmed cell death: the bcl-2 gene producing a mitochondrial protein prevents the apoptotic death by the B-cell. However, this 'rescue operation' depends on sensitization by an antigen, which renders the B-cell long lived by virtue of its escape from apoptosis. The population of B-cells is thus selected in a highly specialized manner, defined by its antigenic reactivity.
Table 3. Classification of autoimmune diseases and corresponding autoantigens Disease
Corresponding autoantibody towards:
(1) Multisystem disease Systemic lupus erythematosus Antiphospholipid syndrome Rheumatoid arthritis Sjogren's syndrome Goodpasture's syndrome Scleroderma
DNA (other polynuclotides), phospholipids, cellular ribonucleoproteins, histones. Phospholipids, y0-2-glycoprotein 1 Fc region of IgG collagen components Ro (SS-A) and La (SS-B) Alveolar and kidney basement membrane Centromere, Topoisomerase l(Scl-70)
(2) Organ specific disease Myasthenia gravis Polymyositis Graves' disease Hashimoto's thyroiditis Addison's disease Pernicious anemia Insulin dependent diabetes Primary billiary cirrhosis Autoimmune hemolytic anemia Idiopathic thrombocytopenic purpura Pemphigus
Acetylcholine receptor Jo-1 TSH receptor Thyroglobulin Adrenocortical cytoplasmic antigen Gastic parietal cells Pancreatic islet cells Mitochondrial antigens I antigens Platelet antigens Desmosomes
Bcl-2 also influences the T-cell repertoire. It has been observed that the location of lymphocytes within the thymus determines its faith. Thus, in the medullar regions, enhanced clonal deletion is observed following low gene transcription, while the opposite is so in the thymic cortex. The second gene is Fas (Apo-I), which specifies for a 48kD transmembrane protein [17]. It has been observed that mice with mutations in Fas or the Fas ligand (Ipr, Iprcg, gld) have defective apoptosis and develop massive lymphadenopathy and autoantibodies characteristic of lupus. The lupus like disease was however more obvious when the mutations occurred on an autoimmune background, such as MRL and NZB. 4.3. Cytokines The existence of these molecules had been predicted by Bretcher & Cohn in 1970. They postulated that lymphocyte activation follows a two signal mechanism. As such, the binding of a lymphocyte receptor to the antigen was not sufficient for its activation and therefore requires a 'second signal'. This role has been found to be possessed by a special 'messenger
molecule'. Indeed, it was later realized that provision of the first signal (the mere ligation of the antigen to the receptor) resulted in a state of 'clonal anergy' (inactivation). The CO-stimulatory molecules constituting signal 2 were identified as bacterial products (lypopolysacchrides, Freund's adjuvant), lymphokines and adhesion molecules. It has been shown that the existence of B or T-cells lacking stimulation by these factors did not result in autoimmunization.
5. NATURAL AUTOIMMUNITY The existence of B and T-cells harboring selfreactivity complicates the concept of autoimmunity, since it goes to show that this property is not necessarily associated with disease states [18]. Moreover, the abundant existence of anti-self reactivity implies that representation of self within the immune system might even have teleological roles in terms of protection or immune modulation. As such, it has been initially proposed by Grabar [19] that natural autoantibodies (NAA) could act as transporters of catabolic products serving to clear the organism of harmful self as well as
foreign substrates. This view has later been extended by suggesting that by low affinity binding to autoantigens, natural autoantobodies could function as filters preventing the induction of autoimmunity [20]. An elaborate description of the biological roles of NAA is provided in Table 4. A compelling theory regarding the essentials of this network has been proposed by I. Cohen designating the expression 'immunologic hommunculus' to indicate the capacity of the brain to "imagine" itself. NAA are bound to various structures found in the organism's body (i.e., serum proteins, cytokines and hormones) amongst which are also highly conserved self antigens (DNA, intracellular structures). One of the unique properties of NAA is their independent production by the immune system, namely— they do not require antigenic stimuli as do other antibodies. NAA are predominately of the IgM isotype, although some consist of IgG and IgA. Another characteristic property is polyreactivity (widespread reactivity with infectious antigens and organic chemical substances). The clinical significance of NAA has been questioned by some authors owing to their low avidities to self-antigens. Furthermore, the presence of NAA in huge amounts among patients with monoclonal gammopathies (Reference reviewed in [21]) without any corresponding clinical manifestations was inconsistent with a presumed pathogenic potential of these autoantibodies. This view was recently challenged by a set of studies in which active immunization with monoclonal antibodies (human IgM antibodies obtained from a healthy subject immunized with diphtheria and tetanus) resulted in the emergence of a clinical picture resembling human SLE and antiphospholipid syndrome [22]. Another relevant finding is the increased occurrence of NAA with ageing, which may seem paradoxical, owing to the well documented decline in immunologic functions accompanying the ageing process. Moreover, attempts to induce experimental autoimmune disease in aged animals are fraught with heightened resistance. Thus, the age-related increase in the incidence of autoantibodies can be considered a physiological process that improves the capacity of the individual to handle tissue damage. NAA are probably produced by CD5 positive Bcells [23] by using selected unmutated germline genes that encode conserved sequences for large binding
sites, capable of reacting with various autoantigens. This characteristic of NAA could explain the low affinities for different autoantigens.
6. AVOIDANCE OF IMMUNOLOGIC TOLERANCE 6.1. Autoantigens The fundamental study of any autoimmune disease is initiated by attempts to characterize and define the autoantigen towards which the autoantibody is directed. Research in this field is headed by immunohistochemical studies of recombinant proteins obtained by screening expression libraries with autoantibodies. Examples of autoantigens detected by this method consist of myelin basic protein (MBP) in mice and acetylcholine receptors in patients with myasthenia gravis. The role of autoantigen recognition has been exemplified in a classic study by which sera from diabetic children were found to contain elevated titers of antibodies, specific for A-17 residue bovine serum peptide differing in sequence from that of human albumin. Cross-reaction of these antibodies with the p-69 (a pancreatic ^-cell surface protein) which is considered the target autoantigen in autoimmune diabetes, had been noticed [24]. It should be emphasized that even the most highly conserved and basic self structures can potentially induce autoantibody production. As such, immunizing rabbits with cytochrome C resulted in the production of antibodies against mouse specific domains of the enzyme as well as against rabbit cytochrome C. The latter was shown to react with conserved residues found in all mammalian cytochromes C [25]. 6.2. Aberrant Expression of HLA HLA are classified to two groups (I and II) based on structural and functional attributes. These antigens are characterized by a wide variance between unrelated individuals (allotypic polymorphism). Class I antigens are located in chromosome 6 and consist of three subclasses (HLA A, B and C). These molecules are found in membranes of nucleated cells and blood platelets. Class II molecules are mainly detected on macrophages, dendritic cells and other antigen pre-
Table 4. Biologic and physiologic functions of natural autoantibodies NAA
Action
IgG and IgM Anti-a galacosyl Anti-band 3 IgG Anti-keratin IgG IgM NAA
Clearance of altered self constituents Phagocytosis of senescent erythrocytes form in the circulation Clearance of cellular debris from the circulation as well as enhancement of phagocytosis Disposal of keratin following death of keratinocytes Increased resistance to tumors Protection against microbial infections Protection against parasitic infections Regulation of the immune system by increased IgG binding following tissue damage Inhibition of IgG binding to self-antigens Possessing proteolytic activities on vasoactive intestinal peptides (VIP) Regulation of the immune system through suppression or induction of antibody synthesis Preventing the harmful interaction of autoreactive B-cells with self-antigens
IgG NAA IgM anti-IgG F(ab02 Various NAA's Anti-idiotyopic NAA Various NAA
senting cells. Binding of HLA class II by certain antigens is an absolute requirement for recognition by T-helper cells, whereas cytotoxic T-cells identify antigens associated with class I HLA molecules. It has been suggested that avoidance of autoimmunization of self peptides, presented by the HLA class II molecules, stems from the failure of the cell to express these antigens. Aberrant expression of HLA glycoproteins (provoked by stimuli such as interferon administration) could play a key role in the evolution of autoimmunity [26]. This concept has been supported by studies showing that thyroid epithelial cells acquired antigen presenting properties following viral infections or stimulation by y-interferon. The study of the association between HLA and autoimmune diseases furnished clues to understanding their etiopathogenesis. It was noticed that several autoimmune diseases are more prevalent among humans with HLA DR 3/4. For example, distinct subtypes of DR4 are associated with different susceptibilities to contract rheumatoid arthritis (RA) in different ethnic groups. Evidence for the association between HLADR haplotypes and several AI diseases is presented in Table 5. Furthermore, knowledge of the HLA haplotype may provide tools for recognition of specific subgroups. For example, RA patients who are DR3 positive are more likely to develop gold induced nephropathy, whereas DR4 positive subjects tend to have a severe form of the disease, complicated by extra-articular manifestations [27].
It should be stated, however, that the value of HLA haplotype at the level of the individual, is of limited application, either in predicting the disease course or as a tool for genetic counseHng. 6.3. Polyclonal Activation Most autoimmune diseases are antigen driven. An elaborate activation of B-cells repertoire in an antigen dependent manner represents a possible mechanism, interfering with self-tolerance, although an established link to organ specific autoimmune damage has not been proven. The stimuli leading to polyclonal B cell activation could be bacterial lipopolysacharide (or other bacterial mitogen) as well as a specific 'atmosphere' of cytokines. It has been demonstrated that mice stimulated with LPS continuously produce antiDNA and rheumatoid factor which were subsequently detected within immune complexes in their kidneys. An additional important factor is the product of the bcl-2 gene (responsible for blocking apoptosis), which following its enhanced expression, results in the production of high levels of immunoglobolins in transgenic mice immunized with sheep red blood cells, finally leading to their death due to immune deposit nephritis. The role of polyclonal T-cell activation in triggering autoimmunity has been observed following the introduction of high doses of IL-2 in thymectomized animals and in humans. Moreover, thymectomy of MRL Ipr/lpr mice early in life prevented
Table 5. Association of autoimmune diseases with HLA-DR haploypes HLA-DR2
HLA-DR3
HLA-DR4
HLA-DR5
Subacute thyroiditis Multiple sclerosis Systemic lupus erythematosus Graves's disease Goodpasture's syndrome
Multiple sclerosis Myasthenia gravis Sjogren's syndrome Graves' disease Addison's disease Insulin dependent diabetes mellitus Systemic lupus erythematosus Dermatitis herpetiformis
Rheumatoid arthritis Insulin dependent diabetes mellitus Pemphigus vulgaris
Pernicious anemia Hashimoto's disease
the emergence of splenomegaly, nephritis and massive lymphadenopathy in these animals [28].
diseases consistent with APS and SLE in humans [22, 31].
6.4. Idiotypes and Idiotypic Connectivities
6.5. Environmental Factors
Idiotypes are phenotypic markers of the V genes used to encode immunoglobulin molecules (as soluble antibody molecules or as lymphocyte receptors). It was initially suggested by Jerne [29] that recognition of self by the organism formulates an immune equilibrium in addition to identification of foreign antigens. This theory, by which a complementary set of interconnecting idiotypes form to establish a well orchestrated network has been evidenced in mice and considerable support is present for its existence in humans. The principle of the network is founded on the presence of complementary pairs of antibodies consisting of the idioype (Abl) and its anti-idiotype (Ab2) and correspondingly, the anti-idiotype and its anti-anti-idiotype (Ab3). It can be viewed (Figure 1) that structural resemblance exists between the antigen and Ab2, as well as between Abl and Ab3. Since these two sets of idiotypes display apparently opposing influences, the system is capable of modulating the intensity of the targeted immune response by changing the titers of the idiotypes. Probably the most soUd proof to support the possible pathogenic potential of disruption of the idiotypic network resides in a set of studies by which autoimmune diseases have been induced by introduction of the pathogenic idiotype (reviewed in [30]). Accordingly, it has been suggested that immunizing mice with Abl leads to production of Ab2 which in turn elicits Ab3, bearing structural homology with the 'original' Abl. The production of Ab3 is associated with the emergence of clinical manifestations of autoimmune
Although not established unequivocally, infectious agents are regarded as highly probable etiologic factors leading to disruption of immune regulation resulting in autoimmune diseases [32]. Evidence for the role of infections in autoimmunity consists of: * Onset of autoimmune diseases following distinct infections (rheumatic fever after streptococcal infections and insulin dependent diabetes following mumps or Cocksackie infections) * Structural antigenic similarities between infectious agents and self-antigens. * Viral, bacterial, and parasitic infections are associated with increased titers of antibodies in the host (summarized in Reference [33]). The mechanisms by which infections may induce autoimmunity are still debated. However, an acceptable one is molecular mimicry initially suggested by George Snell in 1968, referring to the antigenic similarities between infecting agents and self structures following which cross reactivity is triggered between the resulting antibodies and self-antigens. Several examples exist which could support this concept, one of the most demonstrative of which is the proposed relationship between mycobacteria and autoimmunity [34]. Epidemiologic data also incriminates Klebsiella, Yersinia and Shigella infections in triggering ankylosing spondylitis and Reiter's syndrome in subjects who are HLA-B27 positive. The model of adjuvant arthritis in rats provides a strong argument favoring molecular mimicry as a key mechanism. Clinically it is evident that rats injected subcutaneously with oil suspension
AB3 anti-anti-Id « autoantibody
ADJUVANT Bacterial wall Superandgen
1-2 MONTHS
AB2
ABl
aiiti*Id
2-3 WEEKS
T HELPER Figure 1. Immunization with an autoantibody carrying a pathogenic idiotype (Abl) results in the appearance of anti-idiotypic antibodies (Ab2), 2-3 weeks afterwards. Ab2 may induce the production of complementary anti-anti-idiotypic antibodies (Ab3), six to 12 weeks after the initial immunization (1-2 months following the detection of Ab2). Ab3 bears structural resemblance to Abl, which is expressed by similar binding properties and therefore comparable pathogenic potential. This idiotypic manipulation could also account for emergence of AI disease following infections. As such, the infectious agent would stand as the triggering element leading to the production of Ab2.
of Mycobacterium tuberculosis, develop characteristic inflammation of the joints. It has been found that Tcell clones from these mice recognized the heat shock protein-65 (HSP65) antigen present both in mycobacteria and in the synovial fluid [35]. Concomitantly, an autoimmune attack is generated, that was initially triggered by the foreign mycobacterial agent. Other mechanisms of autoimmune disease induction, thought to be initiated by infections include: * Polyclonal activation (discussed earlier). * The 'altered self hypothesis according to which the release of bacterial or viral products changes the structure of self constituents rendering them immunogenic. * Aberrant or excessive expression of HLA class II resulting in autoantigen presentation to autoareactive T-cells (discussed previously). Apart from infections which probably constitute the commonest environmental factors, several other extrinsic insults have been incriminated as contributing to autoimmune diseases (reviewed in Reference [36]). Drug induced autoimmunity is a well characterized phenomenon, the classic prototype of which is lupus like syndrome [37] occurring in the context of hydralazine, isoniazide, procainamide (and probably other medications) ingestion and is manifested either by the appearance of various clinical pictures (rheumatic symptoms, fever, pleuropulmonary involvement) and by serologic markers (anti-histone antibodies, ect.). The mechanisms proposed to account for these drug related effects consist of cross reactivity or interaction with nuclear antigens. Alternatively, the medication may impair immune function by interacting with lymphocytes or by eliciting anti-lymphocytic antibodies. Similarly the role of various toxins and chemicals have been noticed as possible factors antedating the occurrence of autoimmune phenomena. As such, cigarette smoking has been observed to increase the risk of developing autoimmune conditions such as Graves' ophthalmopathy and to exacerbate pulmonary hemorrhage in patients with Goodpasture's syndrome. Ultraviolet radiation has been shown to precipitate flares of SLE (in particular, its skin lesions).
6.6. Hormonal Factors The influence of the hormonal profile on the evolution of autoimmune conditions can be deduced from their higher prevalence among women and exacerbations during puberty and the postpartum period. Furthermore, several observations provide additional support for the significance of hormones with regard to autoimmunity. As such, it has been shown that the clinical manifestation of SLE worsened following the introduction of sex hormones in experimental models [38]. The role of estrogen in autoimmune disease is probably bimodal, namely, it could either inhibit intrathymic differentiation, or in other circumstances, extrathymic pathways. Testosterone had opposite effects regarding the immune system and the initiation of autoimmunity. Recent studies also support the existence of a link between prolactin, the immune system and autoimmune diseases. 6.7. Immunologic Factors Deficiencies in early complement components (CI, C2, C4) have been associated with the evolvement of SLE and other autoimmune conditions. Moreover, patients with SLE or RA were found to harbor, more often, a complement receptor deficiency (CRl-normally present on red blood cell membranes). Various immune deficiency states have also been linked with autoimmune diseases. The classic example is IgA deficiency (one of the most common immune deficiency states) which has been found to co-exist with conditions such as SLE, RA, thyroiditis, polymyositis, and other autoimmune conditions. One of the most acceptable explanations for this well documented association relates to the continuous 'burden' of infections imposed on the compromised immune system resulting in its elaborate activation to produce high titers of antibodies resulting in frequent allergic and autoimmune states.
7. AUTOANTIBODY MEDIATED TISSUE DAMAGE In general, although not always fully defined, autoantibodies can be subclassified into those which were
10
definitely proved as pathogenic, others, for which the pathogenicity has not been defined and those whose pathogenic properties are unhkely (the principal group of which are NAA). The following fines will engage in a short description of the mechanisms by which autoantibodies have been shown or presumed to precipitate organ damage. 7.1. Direct Cytotoxicity This mechanism follows binding of the autoantibody to surface membranes resulting in cell destruction. This action is well documented and is modulated by several mediators: ^Antibody mediated cell mediated cytotoxicity (ADCC) This mechanism involves lysis of the tissue mediated by mononuclear cells carrying receptors for the Fc portion of IgG (K-cells). The binding of the autoantibodies to the target cells supplemented by attachment of the Fc receptor of the K-cell, results in the liberation of hydroxyl radicals and hydrogen peroxide, which exert direct cytotoxic effects. The mechanism is targeted owing to the specificity of the IgG antibodies. Conflicting data is provided regarding the nature of ADCC in SLE during in vitro studies (showing decreased ADCC) and in vivo ones (by which ADCC is apparently enhanced). Concomitant findings are lymphopenia and anti-lymphocyte antibodies explained by binding of the latter to lymphocyes via F(ab02 fragment and to K-cells through their Fc portion. * Phagocytosis Human macrophages dispersed throughout the body formulate the mononuclear system (formerly regarded as part of the reticuloendothelial system). These cells stem from progenitors (promonocytes) contained within the bone marrow. Following the process of maturation monocytes wander and deposit at the basement membrane around the small blood vessels as mature macrophages, obviating their role as protecting guards of penetrating pathogens. The principle location in which macrophages are situated are however: the alveoli, the liver (Kuppfer cells) and the splenic sinusoids. The mechanisms of protection conferred by the macrophages are related to their ability to engulf the foreign pathogen following its adherence to the surface of the cell. This attachment triggers a contractile system (composed of actin and myosin fibers)
extending pseudopods that surround the target substance/pathogen forming a phagosome. This process is followed by fusion of the macrophage cytoplasmic granules with the phagosome, externalizing their lytic content which leads to the destruction of the engulfed element. It has been shown that this process may be responsible for some autoimmune phenomena due to phagocytosis of self constituents by the macrophages. For example, SLE patients may develop anemia as a result of the possession of Fc receptors (in their activated macrophages), which can attach autoantibodies bound to erythrocytes. * Complement mediated cytotoxicity The complement system can trigger tissue damage by a process eventuating in a net influx of sodium and water through transmembranal channels generated in the target cell leading to its destruction. Activation of the complement cascade can be initiated by the binding of autoantibodies to cell membranes. This can be demonstrated in cases of anti-red blood cell antibodies present in patients with autoimmune hemolytic anemia (AIHA) who have anti-red blood cell antibodies and some individuals with Hashimoto's disease having anti-thyroid antibodies. 7.2. Cell Surface Receptor Binding (No Cytolysis) Autoantibodies can bind cell surface receptors after which alteration of their biologic activities occurs. Several mechanisms have been described as responsible for these modifications: * Binding of the antibody to the cell surface receptor may reduce the expression of the receptors. This process of effective down-regulation of receptors results from their realignment causing their disappearance from the cell surface. The classic example of this phenomena is impairment of neuromuscular function in myasthenia gravis due to the presence of anti-acetyl choline receptors. * Different effects may be accomplished by antibodies which block binding of a physiologic ligand resulting in inhibition of cell binding. This influence is exemplified by antibodies to type I intrinsic factor preventing binding of B12 to intrinsic factor molecules leading to the development of pernicious anemia. * Opposing effect can be evidenced by the existence of long acting thyroid stimulating (LATS) antibodies which precipitate thyrotoxicosis by virtue of
11
their stimulatory action culminating in activation of TSH receptors on thyroid cells.
7.3. Immune Complex Mediated Damage One of the principle defense mechanisms engaged in normal homeostasis preservation stems from concomitant elaboration and disposal of immune complexes (consisting of antibodies bound to autoantigens) from circulation by the mononuclear phagocyte system. This well-orchestrated and dynamic equilibrium can be offset in certain autoimmune conditions with resultant tissue damage. The mechanisms which might divert this tightly equilibrated response to evolve in tissue damage include: (1) formation of antibodies reacting with distinct self autoantigens (i.e., Goodpasture's syndrome) and (2) in situ formation of immune complexes (Farmer's lung) and lodging in specific tissues (seen in certain glomerulonephropathies). The site of immune complex deposition could either be determined by physical conditions (i.e., local vascular permeability) or by the relevant properties of the antigen and the existence of corresponding receptors in the tissue. For example, it has been suggested that in RA, the local elaboration of rheumatoid factor reacting with IgG in the joint is essential for the development of the inflammatory process. Another possible mechanism could be inferred from the observation of C3b receptors within normal tissue, the presence of which may designate a process of local entrapment of immune complexes. The net biologic effect achieved following the formation of immune complexes depends largely on its properties, namely, its complement binding (classic or alternative) potential. The local immune mediated inflammation within the tissue is potentiated by the production of various modulators (e.g., cytokines) which may amplify the response. Finally, it has been argued that some of the effects of immune complexes may depend on its size which may grow to contain 2 or 3 bound IgG molecules. An additional mechanism of antibody mediated damage includes its penetration to the tissues to produce deleterious effects. This is suggested by the observation of IgG in epithelial cells of skin biopsy from lupus patients. Several cell populations (predominantly neurons) have been shown to engulf IgG by pinocytosis through in vitro studies.
12
Exposure of cryptic (hidden) epitopes following cell injury may result in presentation of an intracellular antigen, previously not exposed to the influence of the immune system. This mechanism has been demonstrated in Goodpasture's syndrome patients exposed to cigarette smoking (which probably mediate alveolar and glomerular exposure of the cryptic epitopes towards which the corresponding autoantibody react).
8. CANCER AND AUTOIMMUNITY Many reports exist regarding the association of autoimmune diseases and neoplasms. Autoimmune phenotype and various autoantibodies have been found in patients with diverse malignancies. For example: antinuclear antibodies have been described in a prevalence of 19%-31% by different authors (reviewed in Reference [39]). Unlike the autoantibodies found in prototypic autoimmune diseases (i.e., SLE), autoantibodies in various malignancies exhibit a more restricted pattern of reactivity. Similarly, rheumatoid factor has also been detected in higher prevalence in patients with malignancies in comparison with control subjects (between 11% and 85% in different reports). The presence of RF was found to correlate with a poorer prognosis in malignancies such as transitional cell carcinoma, melanoma and gastrointestinal malignancies. Malignant transformation is an important process occurring in patients with autoimmune diseases. The reasons for the increased tendency to malignancies is not always apparent but probably derives both from genetic factors and from environmental determinants (i.e., chemotherapy). Examples of common neoplasms in autoimmune diseases include: lymphoproliferative malignancies in patients with rheumatoid arthritis, lymphomas in patients with SLE and Sjogren's syndrome, epithelial malignancies in patients with dermatomyositis and polymyositis, lung cancer in scleroderma and lymphomas as well as thyroid papillary carcinomas in patients with autoimmune thyroid diseases. Thus, the relationship between autoimmunity and cancer is dual and a cause and effect association can not be determined with certainty. However, regardless of the initial triggering factor, it appears that the immune system plays a detrimental role in the pathogenesis of both autoimmunity and cancer suggesting
AUTOIMMUNE STATE/NORMAL HEALTH ENVIRONMENTAL: Infection Stress Drugs, UV Smoking
GENETIC: HLA, Gm allotypes idiotypes C deficiency
IMMUNE DEFICIENCIES: IgA deficiency fl
HORMONAL: Estrogen i\ Testosterone U Prolactin 11 Thymic hormones ftU
TsU C2,C4 deficiency i\ NKU
1 1 1 1 1 1
t MffiCO^IMUNE DISEASl Figure 2. Evolution of an autoimmune disease from a "state of autoimmunity" may require the participation of diverse elements consisting of environmental, hormonal, immunological and genetic effects. These factors may act in concert or each one a part, to render an individual prone to autoimmune disease following an inciting 'triggering event'. It is probably the combined action of these factors, rather than the existence of each, that results in an autoimmune disease.
it may be implemented in the future for targeted immunomodulation of neoplasms.
9. CONCLUDING REMARKS-THE "MOSAIC" OF AUTOIMMUNITY As could be viewed from the above, the ability to distinguish self from non-self is a key inherent property of the immune system engaging in normal preserva-
tion of immunological homeostasis [40]. However, an interplay of predisposing conditions which indeed have been found to occur frequently in autoimmmue diseases render an individual prone to experience an attack of his immune system on his own self constituents [41]. No consistent condition has been shown to accompany all autoimmune states, yet it may be so that a unique set of terms are fulfilled in an individual, combining in a 'mosaic' pattern that results in diverting
13
an immune mediated inflammatory process to damage self-structures. Thus, incriminated predisposing factors include hormonal, environmental, immunologic and genetic elements (Figure 2), which collectively constitute a milieu which enable a seemingly innocent triggering event to activate a vicious cascade resulting in autoimmune disease. Thus, for example, two members of the same family (first degree relatives) indeed ^inherit' the same preponderance to develop autoimmunity as determined by their HLA status. However, the different frequencies of AI diseases among these apparently 'equally prone' subjects could well be explained by the existence of additional factors (environmental, hormonal and immunologic). These observations lend further support to the concept of mosaicism as predominating in autoimmunity.
11.
12. 13.
14.
15.
16.
17. 18.
REFERENCES 1.
2. 3.
4.
5.
6.
7. 8. 9.
10.
14
Pincus T, Mitchell JM, Burkhauser RV. Substantial work disabilities and earnings losses in individuals less than 65 with osteoarthritis: comparison with rheumatoid arthritis. J Clin Epidemiol 1989; 42: 449457. Nossal GJ. Immunology: the basic concepts of the immune system. N Eng J Med 1987; 8: 227-235. Hamaoka T, Ono S. Regulation of B-cell differentiation: interaction of factors and corresponding receptors. Annu Rev Immunol 1986; 4: 167-204. Cambier JC, Ransom JT. Molecular signaling in Blymphocytes. Annu Rev Immunol 1987; 5: 175199. Natvig JB, Kunkel HG. Human immunoglobulins: classes, subclasses, genetic variants and idioypes. Adv Immunol 1973; 16: 1-59. Romain PL, Schlossman SE. Human T lymphocyte subsets: functional heterogenicity and surface recognition sturctures. J Clin Invest 1984; 74: 15591565. Acuto O, Reinherz EL. The human T-cell receptor. N Eng J Med 1985; 312: 1100-1 111. Balkwill PR, Burk F. The cytokine network. Immunol Today 1989; 10: 299-302. Muller-Eberhard HJ. Molecular organization and function of the complement system. Annu Rev Biochem 1988; 57: 321-335. Tada TM, Tanaguch TM, Takemori T. Properties of primed suppressor T cells and their products. Transplant Rev 1975; 26: 106-129.
19.
20.
21.
22.
23.
24.
25.
26.
27.
Ehrlich P. In : Himmelweit F (Ed) The collected papers of Paul Ehrlich, Vol 2, London, Pergamon Press, 1957; 246-255. Rose NR. Autoimmunity: a personal memoir. Autoimmunity 1988; 1: 15-21. Coombs RRA, Mourant AE, Race RR. A new test for the detection of weak and incomplete Rh agglutinins. Br J Exp Pathol 1945; 26: 255. Burnet FN. The clonal selection theory of acquired immunity. Cambridge University Press, London, 1959. Elliot T How do peptides associate with MHC class I -restricted processing pathway. Immunol Today 1991; 12: 447-155. Cohen J J, Duke RC, Fadok VA, Sellins KS. Apoptosis and programmed cell death in immunity. Annu Rev Immunol 1992; 10: 267-293. Nagata S, Golstein P. The fas death factor. Science 1995;267:1449-1456. Shoenfeld Y, Isenberg DA, eds. Natural autoantibodies. Boca Ranton: CRC Press, Inc, 1992, 1-550. Grabar P. Autoantibodies and the physiological roles of immunoglobulins. Immunol Today 1983; 4: 387342. Cohen IR, Cooke A. Natural autoantibodies might prevent autoimmune disease.Immunol Today 1986; 7: 363-364. George J & Shoenfeld Y. Natural Autoantibodies. In: "Autoantibodies". Eds. Peter JB & Shoenfeld Y. Elsevier, Ltd, 1996. Bakimer R, Fishman P, Blank M, Srendi B, Djaldetti M, Shoenfeld Y. Induction of primary antiphospholipid syndrome in naive mice by immunization with a human monoclonal anti-cardiolipin antibody J Clin Invest 1992; 89 1558-1663. Casali P, Burastero SE, Nakamura M, et al. Human lymphocytes making rheumatoid factor and antibody to ss-DNA belong to Leu-1+ B cell subset. Science 1987; 236: 77-81. Karjalainen J, Martin JM, Knip M, Ilonen J, Robinson BH, Savilahti E, et al. A bovine albumin peptide as a possible trigger of insulin-dependent diabetes mellitus. N Eng J Med 1992; 327: 302. Jemmerson R, Morrow PR, Klinman NR, et al. Analyses of an evolutionarily conserved antigenic site on mammalian cytochrome c using synthetic peptides. Proc Natl Acad Sci USA 1982; 82: 1508. Botazzo GF, Pujol-Borrel R, Hanfusa T, et al. Role of aberrant HLA-DR expression and antigen presentation in induction of endocrine autoimmunity. Lancet 1983; 2:1115. Ferraccioli G, Savi M. Association between DR antigens, rheumatoid arthritis with and without extraar-
28.
29. 30.
31.
32.
33.
34.
ticular features and SLE in Northern Italy. Ital J Rheumatol 1988 15:51-53. Thofilopoulus AN, Dixon FJ. Murine models of systemic lupus erythematusus. Adv Immunol 1985; 37: 269. Jeme NK. Towards a network theory of the immune system. Annales D'immunologie 1974 125c; 373-89. Shoenfeld Y. Idiotypic induction of autoimmunity: a new aspect of the idiotypic network. FASEB J 1994; 8: 1296-1303. Mendlovic S, Brocke S, Shoenfeld Y et al. Induction of SLE-like disease in mice by a common anti-DNA idiotype. Proc Natl Acad Sci USA 1988; 85: 2260-64. Behar MB, Porcelli SA. Mechanisms of autoimmune disease induction. The role of the immune response to microbial pathogens. Arthritis Rheum 1995; 38: 45876. Shoenfeld Y, Cohen IR. Infection and autoimmunity. In: Sela M, ed. The Antigens. CRC Press, Boca Raton, FL1988: 430-88. Shoenfeld Y, Isenberg DA. Mycobacteria and autoimmunity. Immunol Today 1988; 9: 178-182.
35.
36.
37. 38.
39.
40.
41.
Van Eden W. Heat shock proteins as immunogenic bacterial antigens with the potential to induce and regulate autoimmune arthritis. Immunol Rev 1991; 121: 5. Shoenfeld Y, Isenberg DA. Environmental Factors and Autoimmunity. In: The Mosaic of Autoimmunity, Elsevier, Amsterdam, 1989, 321-410. Hess E. Drug related lupus. N Eng J Med 1988; 318: 1460. Blank M, Mendlovic S, Fricke H, Moses E, Talal N, Shoenfeld Y Sex hormone involvement in the induction of SLE by a pathogenic anti-DNA idiotype in naive mice. J Rheumatol 1990; 17: 311-319. Tomer Y, Sherer Y, Shoenfeld Y. Autoantibodies, autoimmunity and cancer. Oncology Rep 1998; 5: 753-761. Schwartz RS. Autoimmunity and autoimmune disease. In: Fundamental Immunology. Ed:Paul WE. Raven Press, Ltd. New York, 1993: 10331098 Shoenfeld Y, Isenberg DA. The mosaic of autoimmunity. Immunology Today. 1989; 10: 123-126.
15
© 2000 Elsevier Science B.V All rights reserved. Cancer and Autoimmunity Y. Shoenfeld and M. E. Gershwin, editors
Rheumatoid Arthritis and Cancer Mahmoud Abu-Shakra^ Dan Buskila^ and Yehuda Shoenfeld^ ^Soroka Medical Center and Ben-Gurion University, Beer-Sheva, Israel; ^Sheba Medical Centre and Sackler Faculty of Medicine, Tel-Aviv, Israel
1. INTRODUCTION Rheumatoid arthritis (RA) is a chronic systemic inflammatory disease characterized by proliferative synovitis of diarthrodial joints, serositis, lymphocytic infiltration in various tissues, vasculitis of small vessels and the production of anti-immunoglobulin autoantibodies (rheumatoid factor) [1]. The etiology of the disease remains unknown. However, environmental and immunogenetic factors may have a role in the development of the disease [2]. Furthermore, the genetic background of the patients may determine the severity of the disease. Genetic analyses have shown that certain HLA DRBl genotypes (0401/0401)are associated with more severe, erosive and deforming form of RA [3]. The disease affects both females and males, but it is 2-3 times more common in women than men. The articular features of the disease include, morning stiffness and swelling, erosions and deformities of the small joints of the hands and feet as well as the large joints. Extra-articular features are also common and include malaise, fatigue, rheumatoid nodules, Sjogren's syndrome, lymphadenopathy, splenomegaly, vasculitis of the skin and internal organs, pleuritis, perecarditis and others [1]. The management of RA include the use of nonsteroidal anti-inflammatory drugs, steroids, and the disease modifying anti-rheumatic drugs (DMARDs). Methotrexate, gold salts and azathioprine are commonly used in the treatment of the disease. Other treatment modalities include the use of cyclosporine A and biologic agents [4]. Various studies have suggested an increased risk of malignancy among patients with RA and malig-
nant diseases contribute significantly to the morbidity and mortality of the disease [5-9]. In the present chapter, we review the association between RA and malignancies.
2. PREVALENCE OF CANCER IN RA The link between RA and cancers is based on numerous case reports of RA patients who developed solid and lymphoproliferative cancers and on large population-based studies. Cases of non-Hodgkin's lymphoma (NHL) [10], acute and chronic leukemia [11-12], multiple myeloma [13], and lung cancers [14] were all reported in association with RA patients. Over the last three decades, several studies have reported the rate of cancer among large cohorts of RA patients [5-7]. Furthermore, the contribution of various DMARDs to the overall risk of cancer and to the risk of lymphoproliferative tumors was also reported [15-19]. Mortality and autopsies studies have indicated a low rate of cancer-related deaths among patients with RA compared to the general population [20-22]. In three mortality studies, the rate of death as result of cancer was between 7-11%. In a cohort of 1000 patients with RA followed over 3 years between 19701975, cancer was the cause of death in 9% of the total deaths compared to 27% in the general population [21]. In a more recent study [8] from four North American centers, the observed rate of cancer death among the patients with RA was 9.4% compared with 27.7% in the general population. Epidemiologic studies have shown the frequency of cancers in patients with RA to be between 2-15%.
19
In three large population-based studies, the prevalence of cancer was 2.6% among 46,101 patients with RA included in the nationwide registry in Finland [5], 7.2% among 11,683 RA patients from Sweden [6], and 9.7% within a cohort of 20,699 patients recorded in the Danish Hospital Discharge Register [7]. Higher rates of cancer were identified among RA patients living in Rochester, Minnesota, USA (16%) [23], among RA patients living in Saskatchewan, Canada (15.8%) [24] among men with Felty's syndrome (15%) [26], and among RA patients who received cyclophosphamide (16-30%) [27-28]. Table 1 shows the frequency of cancer within various cohorts. Tennis et al. [29] estimated the incidence of cancer among 1210 patients with RA who followed over 6539 patient-years in the province of Saskatchewan, Canada. The age adjusted incidence densities of cancer in RA was 8.2 cases/1000 person-years for women and 26.1 cases/1000 person-years for men. In the same study, incidence densities in the general population were, 10.4 cases/1000 persons in women and 16.1 cases/1000 persons per year. Similarly, in a large population-based study [6] of 11,683 Swedish RA patients followed over 101,000 patient-years, the estimated incidence rate of cancer in that cohort was 8.3 cases/1000 person-years.
In three large population-based long-term studies of RA patients living in Scandinavia, the overall risk of cancer was not increased in two cohorts [5,6], and a significant 11% increase in all cancers was observed within the third cohort [7]. The studies populations comprised 46,101 RA patients living in Finland [5], 11,683 RA patients from Sweden [6] and 20,699 RA patients living in Denmark [7]. The standardized incidence ratios (SIR) (observed/expected) for cancer were 1.06 in the first study [5], 0.95 in the second [6], and 1.11 (1.7 for hematopoeitic cancers, 1.08 for nonhematopoietic cancers) in the third [7]. A significantly increased risk of cancer compared with the general population was observed in RA patients treated with cyclophosphamide (SIR = 3.7) [28], and in RA patients with Felty's syndrome (SIR = 2.09) [25]. Taken together, the data suggest that the overall risk of cancer among patients with RA is not increased, or is slightly increased. However, patients with severe RA and particularly those with Felty's syndrome [25], and those who received cyclophosphamide [28], have a clinically and statistically significant increased risk for development of cancers.
4. SPECIFIC CANCERS AND RA 3. RISK OF CANCER IN RA COMPARED TO THE GENERAL POPULATION
4.1. Lymphoproliferative Cancers
Several studies were designed to determine whether, or not, RA is associated with: (a) an increased risk for the development of all cancers; (b) an increased risk for the development of specific cancer types; (c) no link occurs between RA and malignancy; and (d) low risk for the development of certain cancers. The risk of cancer among patients with RA was studied in small and large cohorts of RA patients followed at certain rheumatology clinics and in large population-based studies. Table 1 shows the relative risk of cancer among patients with RA compared to the general population. The table shows that while some of the studies indicate that the overall risk of cancer in RA is not increased [5, 6, 15, 23, 24], others have shown a significantly increased risk of cancer in patients with RA compared with the general population [7, 25, 27, 28].
The data from several cohorts of RA patients who developed malignant neoplasms indicate that RA is associated with an increased risk of lymphoproliferative cancers including lymphoma [5-7, 16, 25, 30], leukemia [5, 6, 24, 25, 27] and myeloma [5]. Table 2 shows the relative risk for development of lymphoproliferative diseases in patients with RA compared to the general population. A 1.5- to 8.7-fold increased risks for all lymphoproliferative cancers were observed in several studies [6, 7, 15, 27]. In two large population-based studies, the risks of developing all lymphoproliferative cancers were 1.52 (1.2-1.9) [6] and 1.7 (1.5-2.0) [7]. While the relative risks for haematopoietic cancers among men and women with RA were similar (1.7 each) in a large cohort of patients with RA [7], other studies reported higher risks for lymphoproliferative cancers among men with RA [6, 27] and among men with Felty's syndrome [25]. Prior [27] found the
20
Q
^
C/5
i-
U
H
-u
^
D
C
toX)
u
cd
s ;-( O !^
^
li
li
B
o
(N
^-^
^
5
S
o
II
21
vo' OS ^^ r- ^ ^ 1 1
9 U
(N ON
o ^
VO
(N (N
00
O lO 400) of tumor antigens, approximately one third of these are novel, have hitherto been identified with autoantibodies of cancer patients by various groups using this new methodology called SEREX (serological analysis of recombinant cDNA
expression libraries of human tumors with autologous serum) [12]. Antibodies directed against self antigens (proteins, glycoproteins, lipoproteins, gangliosides, nucleoproteins, nucleic acids) can be classified into two main categories: natural occurring and nonnatural (pathologic) AAb. Some characteristic differences between these two are: (1) mechanisms of induction (germline-coded versus induced); (2) specificity (polyspecific/polyreactive versus monospecific); and (3) affinity (low versus high affinity). Both AAb groups are implicated in tumor immunology. Natural occurring autoantibodies (NOA) are germline-encoded polyreactive antibodies probably involved in the "first line defence", in the clearance of senescent products and in immune regulatory mechanisms. Among the first functions described for NOA was the reactivity with tumor cells [13]. This reactivity is most probably a manifestation of polyreactivity rather than an interaction between a tumor-specific antigen and a corresponding antibody [14]. Tumorreactive NOA may have different biological effects ranging from enhancement of, to surveillance of tumor development. Furthermore, the modulation of host responses by NOA may influence tumorigenesis [15]. The induction and presence of nonnatural autoantibodies in tumor patients is mostly associated with aberrant, de novo or overexpression of the autoantigenic targets in the tumor (Table 1). The detection of such AAb may be important to find new markers for the improvement of diagnosis and prognosis of tumors as well as for the development of new vaccine strategies for the treatment of tumor patients.
159
Table 1. Autoantigens that may elicit or augment immune responses in tumor patients Type of autoantigens
Expression and possible factors involved in the induction of autoantibodies
Examples of (tumorspecific) autoantibodies
Oncoproteins
Expressed as proto-oncogenes in normal tissues; (over)expression of mutated proteins in tumors
L-myc HER2/neu
Tumor suppressor proteins
Expressed in normal tissues; accumulation of (mutated)proteins in tumors
p53
Cell-cycle/mitosisassociated proteins
Overexpression in tumors?
CENP-F SG2NA
Differentiation antigens
Expressed lineage-specific in tumors and also in normal cells of the same origin
tyrosinase
Cancer/testis class of tumor antigens (CTA)
Expressed in a variable proportion of a wide range of cancers; the expression in normal tissues is restricted to the testis; the (neo/de novo) expression in tumors might result by gene activation or derepression [12]
NY-ESOl MAGE-1 MAGE-3 SSX
Onconeural antigens (ONA)
Normally restricted to the nervous system but aberrantly expressed in a number of tumors (by gene activation or derepression or post-transcriptional regulatory mechanisms?)
Hu, Ri, Yo Amphiphysin
voce
Expressed both in neoplastic and normal tissues, but elicit antibody responses only in cancer patients possibly by tumorassociated post-translational modifications and changes in the antigen processing and/or presentation in tumor cells
HOM-MEL-2.4
Cancer-independent autoantigens
Expressed in normal cells or tissues; specific targets of autoantibodies in autoimmune diseases; increased incidence of tumors in a variety of autoimmune diseases
Topoisomerase I ACA
Carbohydrate antigens (mucins, gangliosides)
Modified expression in tumors (underglycosylation, sialylation)
MUCl GM1,GM2, GD2
Cancer-related autoantigens
2. AUTOANTIBODIES AGAINST ONCOPROTEINS Proteins encoded by oncogenes and tumor suppressor genes are involved in the control of cell growth and differentiation and, if inappropriately expressed or mutated, in the genesis of tumors. Recently, AAb against oncoproteins and the tumor suppressor p53 have been reported in patients with malignant tumors. Such immune responses may be of special interest in the early diagnosis and therapy of cancer. There are a growing number of reports regarding p53 AAb, but only a few data about autoimmunity against oncogene products in tumor patients (summarized in Table 2) are available to date [7, 16-23]. AAb to the HER-
160
[11]
2/neu product, a 185-kDa transmembrane protein with extensive homology to the epidermal growth factor receptor have been found in 11-55% of breast cancer patients [7, 16]. The AAb response correlated with HER-2/neu protein overexpression in the patient's primary tumor, but could also be found in some women with HER-2/neu negative breast cancer, suggesting an active immunoselection for HER-2/neu negative variants [16]. This possibility is also underlined by the higher frequency and the higher titers of pjg5HER-2/neu p^J^^^ -j^ ^^^ ^^^^^ ^^^g^ of disease [16]. AAb against the p2F^^ protein, a member of the GTPase complex whose transforming activity evolves by point mutations, has been found in 32% of patients with colon cancer [17]. Although p21^^^ is activated
Table 2. Autoantibodies against proteins involved in the pathogenesis of maHgnant tumors Autoantibody against
Frequency in tumor patients
Remarks
plg^HER—2/neu
Breast cancer 11-55% [7]
Frequency in healthy volunteers 0-5% [16]
p2ira^
Colon cancer 32% [17]
Frequency in healthy volunteers 3% [17]
c-myc
Colorectal cancer 57% [18] Hematological maUgnancies 2.7-5.2% [20] AIDS related lymphoma 15.4% [20]
Different results in healthy volunteers (1.1-17%) and SLE patients (0.8-35%) [18-20]; significantly higher titers in African Burkitt lymphoma patients and Ghanian normals compared to American normals [20]
L-myc
Lung cancer 10% [21]
No association to histology or staging; not detected in normal volunteers [21]; detectable in pleural effusions [22]
c-myb
Breast cancer 43% [23] Colon cancer 40.5% [23] Ovary cancer 33.3% [23]
Not cancer-specific (24% in controls), but frequency significantly elevated in breast cancer patients compared to controls; not found in neuroblastoma patients [23]
by point mutations, most AAb detect epitopes near the carboxyl terminus of the wild-type protein [17]. In addition to these AAb directed against growth factor receptors (pi85™^"^/"^") or GTP binding proteins (p2r^^), AAb to another group of oncoproteins have been described in patients with solid tumors (colorectal, breast, ovary, lung cancer) and patients with leukemias/lymphomas. These AAb are directed against nuclear regulatory proteins such as myb and myc [18-23]. However, with the exception of AAb to the L-myc protein, these antibodies have been shown to be relatively unspecific for tumors in some studies (see Table 2). Furthermore, the frequencies of c-myc AAb in healthy volunteers and SLE patients varied greatly in the different studies [18-20]. In general, different methods of AAb determinations and differences in the populations studied may account for varying results. The source and purification of autoantigens and the assays used for AAb determination may influence results dramatically. For example, c-myc AAb were determined with ELISAs using 31mer c-myc peptides [19] or human prokaryotically expressed recombinant c-myc protein [20] or with immunoblotting using the recombinant protein [18]. Furthermore, variations in the definition of standards of detectability led to different frequencies as has been shown for pjg5HER-2/neu ^^13 -^^ ^^^^^^ cancer patients (1121%) and in healthy volunteers (0-1%) [16]. But also ethnic differences and different influences of endogeneous and exogenous factors in the populations studied may be relevant for the variation in results.
For example, the frequency of pi85™^"^/"^" AAb is highest in women with premenopausal breast cancer because there is also highest frequency of HER-2/neu protein overexpression [7]. In conclusion, there is a further need for studies of the clinical and biological nature concerning humoral autoimmune responses to oncoproteins such as: (a) the evaluation of diagnostic relevance (diagnostic sensitivity and specificity) and the prognostic significance (correlation with the stage of the disease and survival) in defined patient groups using optimized and standardized methods; (b) the search for associations of antibody titers with disease progression or relapse and therapeutic effects; (c) the search for possible mechanisms of AAb induction (correlation with protein overexpression, mutations or presence of oncoproteins in the circulation); and (d) the search for possible effects of AAb on tumor cells.
3. AUTOANTIBODIES AGAINST THE TUMOR SUPPRESSOR PROTEIN P53 As the "guardian of the genome", p53 is the main regulator preventing carcinogenic and teratogenic lesions [24, 25]. Therefore, it is not surprising that the loss of p53 function(s) is linked to the development of human cancer. Mutations of this tumor suppressor gene are the most frequently reported gene alteration in human cancers [26, 27]. In 1982, it was shown for the first time that p53 may become immunogenic in cancer pa-
161
Table 3. Autoantibody response against p53 in different groups of former uranium miners Group of uranium miners or control subjects
(n)
p53 antibody response (in%) B
A-C
Miners with lung cancer Serum analysis - before manifestation - at time of manifestation - after manifestation and therapy
(73)
6.8
4.1
1.4
12.3
(39) (18) (16)
10.3 0 6.3
2.6 5.5 0
5.1 0 0
18.0 5.5 6.3
Miners with probable lung cancer
(30)
6.7
3.3
3.3
13.3
Miners with SLE or scleroderma, but without tumor
(51)
11.8
2.0
0
13.8
(463)
4.1
1.3
1.3
7.8
1.7 2.7 0.8 0
1 1.8 0 0
0 0 0 0
4.5 0.8 0
Healthy miners Women with idiopathic SLE (50 years old - men 18-40 years old - women 18-62 years old
(37) (415) (225) (128) (62)
tients [28]. In the years that followed p53 AAb were detected in sera from a variety of cancer patients in frequencies between 3 and 65% depending on tumor type and method of antibody detection, whereas the prevalence of such AAb in normal populations was very low [29, 30]. In most studies of serum p53 AAb, patients with a relatively late stage of cancer were screened. Lubin et al. [31] were the first to describe that the humoral anti-p53 response may be an early event during tumorigenesis and can be detected before clinical manifestation of the disease. In two studies Trivers et al. [32, 33] showed that p53 AAb were present months to years before the manifestation of tumors: (1) angiosarcoma of the liver in workers occupationally exposed to vinyl chloride; and (2) lung cancer in heavy smokers with chronic obstructive pulmonary disease (Table 8). To further evaluate the predictive value of p53 AAb we analysed the AAb responses in a group with a high risk of developing lung cancer, former uranium miners occupationally exposed to alpha radiation of radon and radon daughter products [34]. Our objective was to determine the prevalence and time of appearance of these AAb during the pathogenesis of lung cancer. We determined p53 AAb by three different enzyme immunoassays using eukaryotically and prokaryotically expressed human p53 as
162
2.7
well as p53 of tumor cell preparations. Positive results were also tested by immunoblotting with recombinant p53. Sera of 73 uranium miners with lung cancer, 30 miners with radiographically suspected lung cancer, 51 miners with connective tissue diseases (CTD) and 463 healthy miners were studied. In 39 lung cancer patients, sera could be analyzed up to 7 years before clinical disease manifestation. We observed that the different ELISAs used (PharmaCell Paris, France; Dianova, Hamburg, Germany; ORGA-MED, Bad Heilbrunn, Germany) showed some different results that may in part be explained by the detection of p53 AAb which differ in their epitope recognition. Therefore, anti-p53 responses were grouped into three categories: (A) positive in only one ELISA; (B) positive in at least two different ELISA; and (C) strongly positive in at least two different ELISA as well as positive in immunoblot. The results are shown in Table 3. Assessing all categories (A-C) as true positive, the frequency of p53 AAb in lung cancer patients is similar to the frequency described in other studies. Nearly the same prevalence can be observed in miners with probable lung cancer and in miners with scleroderma and SLE suggesting a higher risk of developing cancer or of having "silent" cancer in these groups. Indeed, p53 AAb can be detected in 17- 47 months before clini-
Table 4. Autoantibodies to proliferation associated antigens in tumor patients Autoantibodies against Cyclins and CDKs -CyclinBl - Cyclin A - Cyclin-dependent kinase 2 (CDK2) SG2NA (S/G2 nuclear antigen)
Autoantigen expression and function
Relevance
Involved in cell cycle progression (Gl/S; G2/M); Increased expression in many cancers [39^2]
Hepatocellular carcinoma: anticyclin Bl in 15%, anticyclin A in 1%, anti-CDK2inl%[42]
Member of the TLE (tranducin-like enhancer of split) protein family, many of which take part in regulating nuclear functions associated with the cell cycle [45]
Found in a patient with bladder and metastatic lung cancer [44]
Centromere protein F (CENP-F) Cell cycle-regulated centromere protein that appears to play an (p330^, mitosin, MSA 3) important role in mitosis [46, 47].
Retrospective study of patients selected on the basis of a particular autoantibody specificity: association with cancer and other disorders involving increased cell proliferation [49]
DNA Topoisomerase II
Nuclear enzyme that catalyzes the interconversion of topological Found in hepatocellular carcinoma [55] forms of dsDNA (required for DNA replication, recombination and chromosome segregation); marker for proliferating cells [53]; target of antitumor drugs [54]; might be directly involved in oncogenesis [55]
NOR-90/hUBF (human upstream binding factor)
Nucleolus organizer region proteins (89 and 93 kDa) involved in RNA polymerase I transcription
Found in a patient with HCC [56]
Fibrillarin
Protein of the nucleolar U3-RNP involved in pre-ribosomal RNA processing
Found in a patient with HCC [56]
B23/nucleophosmin
Nucleolar protein involved in ribosome maturation and cell proliferation
Found in patients with HCC, lung cancer and dysgerminoma [56]
HCCl
64 kDa nuclear protein probably involved in splicing of pre-mRNA [57]
ANA specificity changed to HCCl in a patient with liver cirrhosis who progressed to HCC [57]
cal tumor manifestation or diagnostic detection. This has been shown in 7 of the 39 patients whose serum samples were available. In 4 patients only response A could be seen. In one patient response B and in two patients response C was detected 41,24 and 25 months before disease manifestation. Furthermore, we found a significantly higher frequency of p53 AAb in uranium miners compared with blood donors (p = 0.008). Taken together, there are important hints of a predictive value of p53 AAb regarding tumor development: (1) these relatively cancer-specific AAb are detectable months to years before disease manifestation [31-34]; (2) the frequency of p53 AAb is significantly higher in risk groups for lung cancer compared to healthy blood
donors as has been shown for uranium miners heavily exposed to alpha radiation [34]. The higher frequency in older male blood donors compared to younger men or women (Table 3) may also be explained in part by a higher prevalence of "silent" tumors (prostate, lung) in older men. Unfortunately, we had no information about the smoking habits of these blood donors; (3) in the risk group of uranium miners without detectable tumor at time of serum analysis, the highest frequency of p53 AAb could be found in patients with a further increase of risk: scleroderma patients and miners with large silicotic opacities [35, 36]; and (4) in the followup of the autoimmune response in two miners, signs of epitope spreading could be observed.
163
Therefore, p53 AAb should be determined in larger risk groups (e.g., smokers) for early diagnosis and therapy of cancer. On account of the low frequency, other possible autoimmune tumor markers, e.g., AAb against L-myc, pi85^^^"^/"^" or some of those described below, should be further evaluated.
4. Autoantibodies to proliferation associated antigens other than oncoproteins Proteins involved in cell-cycle regulation/progression and mitosis, but also other proteins involved in cellular processes that might be increased in unregulated cell growth, may drive autoimmune responses in tumor patients (Table 4). AAb against nuclear and cytoplasmic cell-cycle regulated or regulating proteins and proteins involved in splicing processes and ribosome biosynthesis could be detected in tumor patients. This supports previous observations that AAb responses against intracellular antigens are often directed at molecules involved in cellular biosynthetic or proliferative functions [37]. Cyclins and cyclin-dependent kinases (CDK) are a group of cell-cycle regulating proteins acting at different points of the cell-cycle progression. They are amplificated and overexpressed in many tumors [3841]. Covini et al. [42] showed that AAb to cyclin Bl, cyclin A and CDK2 are present in sera of patients with hepatocellular carcinomas (HCC) in 15, 1 and 1%, respectively. Furthermore, anticyclin Bl antibodies could be found in patients with a higher risk of HCC development, e.g., in patients with chronic hepatitis (in 1 out of 70 cases) and cirrhosis (in 3 out of 70 cases), suggesting a predictive relevance of these AAb. To date, there are no reports about aberrant expression of cyclin Bl in HCC tissue. If an antigen-driven process caused by cyclin B1 overexpression results in AAb production, cyclin B1 antibodies should be also found in other tumors, such as leukemias, breast and colorectal cancers [39-41]. Sera of cancer patients have been shown to be useful reagents for identifying new cellular proteins possibly involved in tumor development. A new cellcycle-specific DNA-binding nuclear protein has been identified using autoimmune serum of a patient with bladder and metastatic lung cancer [43, 44]. This serum produced a previously undescribed cell-cyclerelated staining pattern on HEp-2 cells. According
164
to the cell-cycle distribution the detected antigen was provisionally named SG2NA (S/G2 nuclear antigen). The structural analyses reveal that SG2NA belongs to the TLB (tranducin-like enhancer of split) protein family, many of which take part in regulating nuclear functions associated with the cell cycle [44]. The centromere protein F is another novel proliferation associated and cell-cycle-dependent protein detected by autoimmune sera. Casiano et al. identified a centromere protein provisionally designated p330^ (doublet polypeptide of 330 kDa), which accumulates in the nuclear matrix during S phase, reaching maximum levels during G2 phase and localized at the centromeres during prophase and metaphase and at the central spindle and midbody regions during anaphase and telophase [45]. The same protein, designated centromere protein F (CENP-F), was identified by Rattner et al. [46] using a serum from a lung cancer patient. Already in 1986, an AAb called MSA 3 with a staining pattern of p330^/CENP-F was described by Humbel [47]. A retrospective analysis of the clinical features of the 36 CENP-F autoantibody positive patients showed that 22 of these (=61%) had neoplasms of various types [48]. The predominant types were breast (9 cases) and lung cancer (5 cases). Other types of cancer included stomach (2 cases), tracheal, tonsilar, nasopharyngeal, ovarial, HCC, and Waldenstrom's macroglobulinemia (1 case each). The other diseases were also associated with abnormal cell proliferation: chronic inflammatory diseases of liver, intestinum, pancreas and joints (8 cases), chronic renal allograft rejection (2 cases), SLE (2 cases) and undifferentiated connective tissue disease (1 case). The findings of this retrospective study of patients selected on the basis of a particular AAb specificity point to the probability that anti-CENP-F autoimmunity could be related to increased or abnormal cell proliferation, but important conclusions regarding disease associations cannot be drawn. A screening of several hundred sera of unselected cancer patients by indirect immunofluorescence on HEp-2 cells showed that the prevalence of CENPF antibodies detectable with this method is very low ( 1:320. We compared the results of the cancer patients without CTD symptoms (all cancers and separately lung cancer) with the results of the following groups of miners or control subjects:
167
Table 7. CTD typical autoantibodies in sera of different groups of uranium miners Group of uranium miners or control subjects
(n)
Autoantibodies (in %) against Topocentromere isomerase I proteins
nucleolar proteins
Ro/SS-A
dsDNA
At least one CTD AAb
All miners with definite and probable cancer: - without CTD symptoms
(128) (99)
2.40 2.02
2.40 1.01
0.80 0
5.60 4.04
6.40 4.04
14.40 9.09
Miners with lung cancer: - without CTD symptoms
(73) (53)
4.10 3.77
2.73 1.89
1.37 0
5.48 3.77
6.84 3.77
16.44 11.32
(311)
1.61
1.93
1.61
3.86
6.11
14.46
(1304)
0.92
0.69
0.46
1.69
2.15
5.67
0
0
0.5
0.5
0
1.0
Miners without cancer, but with possible CTD development Healthy miners^ Healthy older men^
(200)
^ Miners without cancer and symptoms of possible CTD development. ''Age and gender related control group for miners with cancer: men from the same geographical region older than 55 years, no cancer, no CTD symptoms and no environmental exposure to carcinogeneous substances.
(1)311 miners without tumor, but with a possible CTD development: symptoms of a possible development of SSc (Raynaud, diffuse lung fibrosis) or symptoms of possible development of other CTD, esp. SLE: one of the clinical ACR criteria for SLE and/or two or more "minor" signs of possible CTD development [82, 83]; (2) 1304 "healthy" miners regarding cancer and CTD symptoms; and (3)200 healthy men older than 55 years, from the same geographical region as the miners, but without any environmental exposure to carcinogeneous substances (=age and gender related control group of miners with cancer). The results are shown in Table 7. The prevalence of CTD typical AAb in miners with tumors or probable tumors is significantly higher compared to healthy miners (p < 0.03) and to a gender and age related control group (p < 0.0002), even if the tumor patients have no CTD symptoms. Particularly antitopoisomerase I (ATA) and anticentromere antibodies (ACA) may be associated with tumor development. The higher prevalence of tumors and p53 AAb in miners with CTD compared to those without CTD, the higher prevalence of CTD AAb in miners with cancer regardless of CTD symptoms and the development of cancer in patients with CTD AAb but without CTD symptoms suggest that CTD antibody positive miners should not only be followed-up for the development
168
of CTD, but also for the manifestation of tumors, especially for lung cancer.
8. AUTOANTIBODIES AS PARAMETERS IN THE PREDICTION AND EARLY DIAGNOSIS OF CANCER? The sooner cancer is diagnosed, the better the therapeutic outcome. However, an early diagnosis is still a problem with many kinds of tumors, e.g., with lung, pancreas, gallbladder and colon cancer. Therefore, there is a need for parameters which are specific for tumors and are detectable in preclinical stages. The ideal tumor marker should be (1) highly sensitive and (2) highly specific for tumors. (3) The tumor sensitivity should be higher than that of other diagnostic methods and (4) the earlier diagnosis should lead to an improvement of therapy. Tumor-associated antigens present in sera of cancer patients (e.g., CEA, NSE, s e c , CA50 etc.) can be useful markers for prognosis and for monitoring cancer therapy but have a limited value for diagnosis, esp. for the early diagnosis of cancer. Furthermore, other diagnostic methods—although improving—do not allow an early diagnosis of most tumors. Theoretically, the autoimmune response to antigens (proteins, glycoproteins) involved in the tumorigenesis and/or aberrantly expressed in tumors, may be an early event leading to detectable parameters
•H. E^^
<