Protagonists of Medicine
Domenico Ribatti
Protagonists of Medicine
123
Prof. Domenico Ribatti Dipartimento di Anatomia Umana e Istologia Piazza G. Cesare, 11 Policlinico Università degli Studi di Bari 70124 Bari Italy
[email protected] ISBN 978-90-481-3742-8 e-ISBN 978-90-481-3741-1 DOI 10.1007/978-90-481-3741-1 Springer Dordrecht Heidelberg London New York Library of Congress Control Number: 2009944150 © Springer Science+Business Media B.V. 2010 No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
I am indebted for their invaluable knowledge and wisdom to my colleagues and friends, Enrico Crivellato, Marco Presta and Angelo Vacca.
Contents
Part I
Immunology and Pathology
1 Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
2 Paul Ehrlich’s Doctoral Thesis: A Milestone in the Study of Mast Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
3 The Contribution of Gianni Bonadonna to the History of Chemotherapy . . . . . . . . . . . . . . . . 3.1 Historical Background . . . . . . . . . . . . . . . . . . 3.2 The Cyclophosphamide, Methotrexate and Fluorouracil (CMF) Regimen in the Treatment of Breast Cancer . . . 3.3 The Adriamycin, Bleomycin, Vinblastine and Dacarbazide (ABVD) Regimen in the Treatment of Hodgkin’s Disease . . . . . . . . . . . . . . . . . . . . 4 Sir Frank Macfarlane Burnet and the Clonal Selection Theory of Antibody Formation . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 4.2 Biographical Profile . . . . . . . . . . . . . . . . . 4.3 The Instructive Theories of Antibody Production . . 4.4 The Selective Theories of Antibody Production . . . 4.5 The Clonal Selection Theory . . . . . . . . . . . . . 4.6 Evidence Supporting the Theory . . . . . . . . . . . 4.7 Concluding Remarks . . . . . . . . . . . . . . . . .
. . . . . . . .
5 The Contribution of Bruce Glick to the Definition of the Role Played by the Bursa of Fabricius in the Development of the B-Cell Lineage . . . . . . . . . . . . . . . . . . . . . 5.1 Fabricius ab Aquapendente . . . . . . . . . . . . . . . 5.2 The BF Plays a Major Role in the Development of Antibody-Mediated Immunity . . . . . . . . . . . . 5.3 BSX Do Not Abrogate the Antibody Response to Cellular Antigens . . . . . . . . . . . . . . . . . . 5.4 Immunoglobulin Synthesis Regulation . . . . . . . . .
. . . . . . . .
13 13
. . . .
13
. . . .
15
. . . . . . . .
. . . . . . . .
19 19 19 21 22 23 24 25
. . . . . . . . . .
27 27
. . . . .
28
. . . . . . . . . .
29 30
. . . . . . . .
. . . . . . . .
. . . . . . . .
vii
viii
Contents
5.5 5.6
Delineation of the Thymic and Bursal Lymphoid Systems in the Chicken . . . . . . . . . . . . . . . . . . . . . . BF Equivalent in Mammals and Other Vertebrates . . . . . . .
6 Miller’s Seminal Studies on the Role of Thymus in Immunity 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 The Thymus in Mouse Leukaemia . . . . . . . . . . . . . 6.3 The Thymus Is Essential for Normal Development of the Immune System . . . . . . . . . . . . . . . . . . . 6.4 If the Immune System Was Destroyed in the Adult, Would the Thymus Still Be Involved in Lymphopoiesis? . 6.5 Cell Transfer Studies . . . . . . . . . . . . . . . . . . . . 6.6 The Functional Anatomy of the Thymus . . . . . . . . . . 6.7 Thymocyte Positive and Negative Selection . . . . . . . . 6.8 The Medulla and Central Tolerance . . . . . . . . . . . .
30 32
. . . . . . . . .
33 33 33
. . .
34
. . . . .
36 36 37 38 38
. . . . .
. . . . .
7 The Fundamental Contribution of Robert A. Good to the Discovery of the Crucial Role of Thymus in Mammalian Immunity 7.1 Idiopathic Acquired Agammaglobulinaemia Associated with Thymoma . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 The Role of Thymus in Immunity . . . . . . . . . . . . . . . . 7.3 The Effects of Neonatally Thymectomy . . . . . . . . . . . . . 7.4 The Restorative Effect of Thymomas or Thymus Grafts on Thymic Function . . . . . . . . . . . . . . . . . . . . . . . 7.5 Removal of Either the Thymus or Bursa of Fabricius . . . . . . 7.6 The Importance of a Functional Thymus in Human . . . . . . . 8 The Fundamental Contribution of Jan C. Waldenström to the Discovery and Study of the So-Called Waldenström Macroglobulinaemia . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 Biographical Notes . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Waldenström Involvement in the Discovery and Study of the Waldenström Macroglobulinaemia . . . . . . . . . . . .
41 41 42 42 44 44 45
47 47 48
Part II Vascular Biology and Angiogenesis 9 Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Giulio Bizzozero and the Discovery of Platelets . . . . . . . . . 10.1 Biographical Notes . . . . . . . . . . . . . . . . . . . . . . 10.2 The Discovery of Platelets . . . . . . . . . . . . . . . . . . 10.3 The Role of Platelets in Thrombosis and Blood Coagulation
. . . .
59 59 60 60
11
A milestone in the Study of the Vascular System: Wilhelm Roux’s Doctoral Thesis on the Bifurcation of Blood Vessels . . . . .
63
Stephen Paget and the “Seed and Soil” Theory of Metastatic Dissemination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.2 Paget’s Theory . . . . . . . . . . . . . . . . . . . . . . . . . .
65 65 66
12
. . . .
55
Contents
12.3 12.4 12.5 12.6 12.7 12.8 13
14
ix
Ewing’s Viewpoint . . . . . . . . . . . . . . . . . . The Contribution of Experimental Pathology to the Study of the Process of Metastasis . . . . . . . . . . Metastasis Can Result from Survival of Only a Few Tumour Cells . . . . . . . . . . . . . . . . . . . . . Experimental Evidence of Metastatic Heterogeneity of Tumours . . . . . . . . . . . . . . . . . . . . . . Studies of Experimental Brain Metastasis . . . . . . Concluding Remarks . . . . . . . . . . . . . . . . .
. . . . . .
67
. . . . . .
68
. . . . . .
68
. . . . . . . . . . . . . . . . . .
69 70 70
The Contribution of Roberto Montesano to the Study of Interactions Between Epithelial Sheets and the Surrounding Extracellular Matrix . . . . . . . . . . 13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 13.2 Mechanisms Underlying Angiogenesis . . . . . . . . . 13.3 Collagen Matrix Promotes the Organization of Endothelial Cells into Capillary-Like Tubules and the Induction of the Invasive Phenotype . . . . . . . . . . . 13.4 Unbalanced Proteolysis Results in Aberrant Vascular Morphogenesis . . . . . . . . . . . . . . . . . 13.5 Mechanisms Underlying Tubulogenesis . . . . . . . . . 13.6 Fibroblast-Derived Soluble Factors Induce Morphogenesis of Branching Tubules by Kidney Epithelial Cells . . . . . . . . . . . . . . . . . . . . . . 13.7 Hepatocyte Growth Factor Is a Paracrine Mediator of Morphogenetic Epithelial–Mesenchymal Interactions in MDCK Cells . . . . . . . . . . . . . . . . . . . . . . 13.8 The Role of the Transcription Factor Snail in Modulation of Permeability and Tight Junction Proteins in MDCK Cells . . . . . . . . . . . . . . . . . . . . . . 13.9 Epithelial Tubulogenesis Is Dependent on Extracellular Plasmin-Dependent Proteolysis . . . . . . . . . . . . . 13.10 Paracrine Epithelial–Mesenchymal Interactions Induce Tubulogenesis in Other Types of Epithelial Cells . . . . 13.11 HGF and TGF-β1 Play a Role in Mammary Gland Morphogenesis In Vivo . . . . . . . . . . . . . . . . . . 13.12 Retinoids Induce Lumen Formation, Whereas Tumour Necrosis Factor Alpha and Bone Morphogenetic Protein-4 Confer an Invasive and Transformed Phenotype to Cultured Mammary Epithelial Cells . . . . 13.13 Concluding Remarks . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
73 73 74
. . . .
75
. . . . . . . .
76 77
. . . .
77
. . . .
78
. . . .
78
. . . .
79
. . . .
79
. . . .
80
. . . . . . . .
80 81
The Seminal Work of Werner Risau in the Study of the Development of the Vascular System . . . . . . . . . . . . . . 14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2 Changes in Extracellular Matrix During Embryonic Vasculogenesis and Angiogenesis . . . . . . . . . . . . . . . .
83 83 84
x
Contents
14.3 14.4 14.5 14.6 14.7 14.8 14.9
Characterization of Flk-1 . . . . . . . . . . . . . . . . . . . Embryonic Brain Angiogenesis . . . . . . . . . . . . . . . The Role of Vascular Endothelial Growth Factor in Brain Angiogenesis . . . . . . . . . . . . . . . . . . . . Characterization of the Blood–Brain Barrier . . . . . . . . . Angiogenesis in Glioma . . . . . . . . . . . . . . . . . . . The Role of Platelet-Derived Growth Factor in Angiogenesis Hypoxia and Angiogenesis . . . . . . . . . . . . . . . . . .
. . . .
84 85
. . . . .
86 87 88 89 90
. . . . .
15
Judah Folkman, A Pioneer in the Study of Angiogenesis . . . . . . 93 15.1 Early Evidence of Tumour Cells Releasing Specific Growth Factor for Blood Vessels . . . . . . . . . . . . . . . . . 93 15.2 Tumours in Isolated Perfused Organs: Absence of Angiogenesis 94 15.3 Hypothesis: Tumour Growth Is Angiogenesis Dependent . . . . 94 15.4 Evidence that Tumours Are Angiogenesis Dependent . . . . . . 94 15.5 Isolation of the First Angiogenic Tumour Factor . . . . . . . . 96 15.6 First Evidence of the Existence of the Avascular and Vascular Phases of Solid Tumour Growth . . . . . . . . . . 97 15.7 Dormancy of Micrometastases May Be Governed by Angiogenesis 98 15.8 Prognostic Significance of Tumour Vascularity . . . . . . . . . 98 15.9 Antiangiogenesis . . . . . . . . . . . . . . . . . . . . . . . . . 99 15.9.1 Interferon Alpha . . . . . . . . . . . . . . . . . . . . . 100 15.9.2 Platelet Factor 4/Protamine . . . . . . . . . . . . . . . 100 15.9.3 Angiostatic Steroids . . . . . . . . . . . . . . . . . . . 100 15.9.4 Fumagillin . . . . . . . . . . . . . . . . . . . . . . . . 101 15.9.5 Angiostatin and Endostatin . . . . . . . . . . . . . . . 101 15.9.6 Thalidomide . . . . . . . . . . . . . . . . . . . . . . . 102 15.9.7 Cleaved Antithrombin III . . . . . . . . . . . . . . . . 103 15.9.8 Caplostatin . . . . . . . . . . . . . . . . . . . . . . . 103 15.9.9 Antiangiogenic Chemotherapy . . . . . . . . . . . . . 103 15.10 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . 104
16
The Contribution of Harold F. Dvorak to the Study of Tumour Angiogenesis and Stroma Generation Mechanisms 16.1 Biographical Notes . . . . . . . . . . . . . . . . . . . . . 16.2 Tumour Blood Vessels Are Hyperpermeable to Plasma Proteins and to Other Circulating Macromolecules . . . . 16.3 Vascular Permeability Factor Activity Is Present in Tumour Culture Supernatants . . . . . . . . . . . . . . . 16.4 The Discovery of Vascular Permeability Factor/Vascular Endothelial Growth Factor . . . . . . . . . . . . . . . . . 16.5 Tumours: Wounds That Not Heal . . . . . . . . . . . . . 16.6 Expression of VEGF-A-164 In Vivo . . . . . . . . . . . . 16.7 The New Blood Vessels Generated by VEGF-A-164 . . .
. . . . . .
105 105
. . .
106
. . .
106
. . . .
107 108 108 109
. . . .
. . . .
Contents
17
18
xi
Napoleone Ferrara and the Saga of Vascular Endothelial Growth Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.2 The Isolation of VEGF . . . . . . . . . . . . . . . . . . . . . . 17.3 Role of VEGF in Embryonic Vasculogenesis and Angiogenesis 17.4 Endocrine Gland-Derived VEGF . . . . . . . . . . . . . . . . . 17.5 VEGF Expression in the Development of Pancreas . . . . . . . 17.6 VEGF Expression in Skeletal Growth and Endochondral Bone Formation . . . . . . . . . . . . . . . . . . . . . . . . . 17.7 VEGF Expression in Ovary . . . . . . . . . . . . . . . . . . . 17.8 Role of VEGF in Intracular Neovascular Syndromes . . . . . . 17.9 Towards the Discovery of Avastin . . . . . . . . . . . . . . . . Pietro M. Gullino and Angiogenesis . . . . . . . . . . . . . . 18.1 Biographical Notes . . . . . . . . . . . . . . . . . . . . 18.2 Tumour Pathophysiology: Interstitial Pressure and Blood Dissemination . . . . . . . . . . . . . . . . . 18.3 Angiogenesis as a Marker for Neoplastic Transformation 18.4 Neoplastic Cell Populations Release Molecules Able to Stimulate Angiogenesis into the Surroundings: The Role of Prostaglandin E-1 . . . . . . . . . . . . . . . . 18.5 Microenvironment and Tumour Angiogenesis: The Role of Copper and Gangliosides . . . . . . . . . . . . . . . 18.6 Concluding Remarks . . . . . . . . . . . . . . . . . . .
111 111 113 116 117 118 118 119 119 120
. . . . . . . .
123 123
. . . . . . . .
124 124
. . . .
125
. . . . . . . .
126 127
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
129
Part I
Immunology and Pathology
Chapter 1
Foreword
The science of immunology grew from the common knowledge that those who survived many of the common infectious diseases rarely contracted that disease again. Immunology as a distinctive subject developed in the middle of the twentieth century as researchers started to understand how the adaptive immune system aids in defence against pathogens. Since that time it has grown in importance at a steadily increasing rate and has become diversified into special fields such as immunohistochemistry, immunogenetics, and immunopathology. Lower animal forms possess the so-called innate or non-specific immune mechanisms, such as phagocytosis of bacteria by specialized cells, which afford them protection against infecting organism. Higher animals have evolved an adaptive or acquired immune response which provides a specific and more effective reaction to different infections. Specificity was mentioned as a fundamental feature of the adaptive immune response. For the past century, immunology has fascinated and inspired some of the greatest scientists of our time and numerous Nobel Prizes have been awarded for fundamental discoveries in immunology, from Paul Ehrlich’ work on antibodies (1908) to the studies of Zinkernagel and Doherty (1977) elucidating mechanisms of cell-mediated immunity. The idea of cells directly involved in the defence of the body was first suggested by the zoologist Ilya Metchnikoff in 1884. Based upon purely Darwinian evolutionary principles, Metchnikoff proposed in his famous book entitled “Immunity in the Infective Diseases” published in 1905 that phagocytic cell is the primary element in natural immunity (the first line of defence against infection) and is critical also for acquired immunity. Another notable contribution of the phagocytic theory was to the field of general pathology. In fact, at the time, most believed that inflammation was a damaging component of the disease itself; Metchinikoff, on the other hand, suggested that the inflammatory response was an evolutionary mechanism designed to protect the organism. In 1888, Louis Pasteur invited Metchnikoff to join him at the newly constructed “Pasteur Institute” in Paris, where Metchnikoff spent the next decades working productively to verify and extend the cellular theory of immunity.
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_1, C Springer Science+Business Media B.V. 2010
3
4
1 Foreword
In 1908, the Swedish Academy conferred the Nobel Prize in medicine jointly to Metchnikoff, the leading exponent of cellular theory, and to Paul Ehrlich, the leading exponent of humoral theory. Ehrlich identified for the first time mast cells when he was a medical student. As it has been reported in the article entitled “Paul Ehrlich’s doctoral thesis: a milestone in the study of mast cells” Ehrlich described these cells in 1878 in his doctoral thesis on the basis of their selective staining properties and the presence of large cytoplasmatic granules. In his doctoral thesis, which was an admirable piece of forerunning insight, Ehrlich developed basic concepts about the microscopical features, histochemical behaviour and functional properties of these cells. Such studies led Ehrlich to formulate the concept of molecules that specifically bind to cell receptors and this principle was at the basis of the formulation of the side chain theory of antibody formation. Then Ehrlich, decided to revisit small molecules with the aim of finding a “magic bullet” to kill microbial pathogens. During this work with dyes, Ehrlich tested the effects of methylene blue on malarial plasmodia. In 1909, Ehrlich discovered the first effective cure for syphilis, the “compound 606” (also called Salvarsan). Salvarsan was first tried on rabbits that had been infected with syphilis and then on patients with the dementia associated with the final stages of the disease. Salvarsan was used in the treatment of syphilis during the first half of the last century until it was superseded by penicillin. For this insight and this achievement Ehrlich is known as the founder of chemotherapy. The history of cancer chemotherapy and of the discipline of medical oncology has been that of drug discovery. The pioneering discoveries of the early days of chemotherapy have allowed the development of a paradigm for drug discovery that persists, with modifications to the present day. The article entitled “The contribution of Gianni Bonadonna to the history of chemotherapy” summarizes the seminal work of the Italian scientist Gianni Bonadonna, working at the Department of Medical Oncology of the “Istituto Nazionale Tumori” in Milan, Italy, who introduced extremely innovative protocols for the treatment of breast cancer and Hodgkin’s disease. In 1896, Ehrlich elaborated his side chain theory to explain the appearance of antibodies in the circulation. He suggested that cells capable of forming antibodies possessed on their surface membranes specific side chains, which were receptors for antigens. He proposed that binding of antigen to the side chains provoked new synthesis of these side chains, which were liberated into serum as antibodies. The cellular aspects of the same problem of antibody formation, as envisaged by Niels Jerne and Sir Macfarlane Burnet, are encompassed in the clonal selection theory. As it is outlined in the article entitled “Sir Frank Macfarlane Burnet and the clonal selection theory of antibody formation”, Burnet, director of the Water and Eliza Hall Institute in Melbourne, Australia, in the years 1941–1956, focused his effort on the elaboration of a theory of antibody synthesis, the clonal selection theory, capable of bridging the gap between physiological findings, such as the kinetics of antibody
1
Foreword
5
production, self-tolerance and immunological memory on the one hand, and the newest ideas on synthesis of proteins, on the other hand. The bursa of Fabricius, which is present in birds, is a lymphoepithelial organ located near the cloaca. Just as the thymus appears to act as a central lymphoid organ controlling the maturation of lymphocytes concerned largely with cellmediated immunity, so the bursa of Fabricius is responsible for the development of immunocompetence in cells destined to make humoral immunity. The article entitled “The contribution of Bruce Glick to the definition of the role played by the bursa of Fabricius in the development of the B cell lineage” outlines the contribution of Bruce Glick to the definition of the role played by the bursa of Fabricius in the development of the B-cell lineage. In 1956, for the first time Glick and Timothy Chang reported that the bursa of Fabricius plays an important role in the antibody production. Their demonstration that antibody responses are suppressed in the majority of bursectomized chickens become the cornerstone of modern immunology. After the discovery of Glick, it has been demonstrated that the mammalian equivalent of the bursa of Fabricius is the bone marrow. The thymus is one of the two primary lymphoid organs. It is responsible for the provision of T lymphocytes to the entire body and provides a unique microenvironment in which T-cell precursors (thymocytes) undergo development, differentiation and clonal expansion. The article entitled “Miller’s seminal studies on the role of thymus in immunity” summarizes the seminal work of the Australian scientist Francis Albert Pierre Miller concerning the description for the first time of the crucial role of the thymus for normal development of the immune system. The article entitled “The fundamental contribution of Robert A. Good to the discovery of the crucial role of thymus in mammalian immunity” outlines the contribution of Robert Alan Good, a pioneer in the field of immunodeficiency diseases. He and his colleagues in Minnesota defined the cellular basis and functional consequences of many of the inherited immunodeficiency diseases. He contributed to the discovery of the pivotal role of the thymus in the immune system development and defined the separate development of the thymus-dependent and bursa-dependent lymphoid cell lineages and their responsibilities in cell-mediated and humoral immunity. There are few areas in haematology that are not significantly affected by immunologic processes. A large group of haematologic disorders, the plasma cell dyscrasias, lymphocytic leukaemias and lymphomas, represent abnormal proliferations of primary cells of the immune system. Lymphoid cells at almost any stage in their differentiation or maturation may become malignant and proliferate to form a clone which is virtually “frozen” in the developmental stage of the parental cell and bear the markers of the normal cell type from which they are derived. The monoclonal gammopathies are a group of disorders characterized by the proliferation of a single clone of plasma cells that produces a homogeneous monoclonal M protein. Waldenström macroglobulinaemia may be said to be characterized chiefly by two features: (1) a striking increase in IgM globulin of the “monoclonal gammapathy” type and (2) an abnormal bone marrow characterized by either a
6
1 Foreword
leukaemic or leukaemic-like picture, in which lymphocytes of various types are prominent. All the other features, haemorrhagic, haemolytic, rheologic (viscosity syndrome), must be considered as secondary manifestations. The article entitled “The fundamental contribution of Jan C. Waldenström to the discovery and study of the so called ‘Waldenström macroglobulinemia’” summarizes the fundamental contribution of the Swedish scientist Jan Costa Waldenström to the discovery and study of this lymphoproliferative disorder.
Chapter 2
Paul Ehrlich’s Doctoral Thesis: A Milestone in the Study of Mast Cells
The history of mast cell research begins with a name and a date. The name is that of Paul Ehrlich (Fig. 2.1); the date is 17 June 1878. That day, the 24-year-old medical student from Strehlen (Schlesien) presented his doctoral thesis at the Medical Faculty of the Leipzig University. The title of his dissertation was “Contribution to the theory and practice of histological dyes” (“Beiträge zur Theorie und Praxis der histologischen Färbung”) (Ehrlich, 1878). This work is a beautiful and admirable
Fig. 2.1 A portrait of Paul Ehrlich
Published in collaboration with Enrico Crivellato, Carlo Alberto Beltrami and Franco Mallardi in “British Journal of Haematology”, 123:19–21, 2003
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_2, C Springer Science+Business Media B.V. 2010
7
8
2 Paul Ehrlich’s Doctoral Thesis
example of analytical experimental method and foresight. Ehrlich’s thesis is organized into two parts. In the first part he overviews the chemical bases of many important histological reactions and in the second part, he discusses the chemical, technological and histological properties of aniline dyes. In the chapter dedicated to the histological applications of this class of chemical compounds, he presents his personal point of view about a type of cell, which he named as “mast cell” (“Mastzelle”). He stated that “aniline dyes display an absolutely characteristic behaviour toward the protoplasmic deposits of certain cells” that were “chemically so sharply” distinguished from the group of Waldeyer’s “Plasmazellen”. With this term, he referred to a broad and heterogeneous category of cells previously described by Waldeyer. Among Waldeyer’s “Plasmazellen”, there was a group of connective tissue cells exhibiting large dimension and round shape, which could be distinguished “from white blood cells on the basis of their significantly large size and lack of contractile activity”. Ehrlich emphasized his assumption that most of those cells that he had described in connective tissues as reactive to aniline staining did not correspond to Waldeyer’s description, which was otherwise based on purely morphological criteria, not chemical. “Anilophilic cells should be strongly separated from ‘Plasmazellen’”, he insisted. These anilinereactive cells “represent sui generis elements and must be distinguished from Waldeyer’s ‘Plasmazellen’ by a different denomination”. He then came to the central part of his presentation. “From the descriptive point of view”, he said, aniline-positive cells should be “most conveniently described as ‘granular cells of the connective tissue’ (‘granulierte Bindgewebezellen’); from the physiological standpoint, these cells may provisionally be indicated as mast cells (‘Mastzellen’) because, like fat cells, they represent a further development (‘Weiterentwicklung’) of the fixed cells of connective tissue”. Ehrlich’s concept is absolutely remarkable in that, although mast cells “are localized with extremely high frequency around blood vessels in the loose connective tissues”, “it seems not justified to regard them as members of a perivascular system”. He also provided a notable explanation to support his view: aniline-reactive cells indeed “have a tendency to collect around developing preformed structures in connective tissues”. In discussing this point, he added that “in certain acinar glands (goat parotid), the pattern of mast cell accumulation [inside the organ] is not determined by the branching of the vascular system but by the ramification of the gland excretory ducts”. In the course of his dissertation Ehrlich underlined the concept that mast cells must be principally distinguished on the ground of their reactivity to aniline dyes, not simply by their shape and morphological appearance. “Granular cells are characterized by the presence of a still undetermined chemical substance”, “which is bound to the granular storages in the protoplasm” and which reacts to aniline dyes giving typical metachromasia. The binding of this chemical substance to aniline dyes shows different staining: red-violet with cyanine, orange with fuchsin and red with dahlia and gentian. Finally, he provides us with an extremely precise description of mast cell microscopical features. “The typical aspect of ‘granular cells’ is as follows. The mostly stainless protoplasm appears as being filled by more or less numerous grains of
2
Paul Ehrlich’s Doctoral Thesis
9
variable size. These granules exhibit subtle nuances specific for each staining procedure. The nucleus is mostly not stainable, even in samples which otherwise display beautiful nuclear staining reactions. In flattened cells, the nucleus appears as a characteristically clear spot, due to the absence of the coloured granules and this picture nearly gives the impression of a lacuna in the cell body.” Many interesting aspects of Ehrlich’s dissertation deserve some comments. He first coined the term “Mastzellen” to describe the aniline-reactive granular cells he found in connective tissues. The German word “Mast” (from the Greek μαστ´oς = breast) implies a nourishing and “suckling” function for these cells. Certainly, mast cells do not provide nutrients in a strict sense; however, they are deeply involved in the “trophism” of tissues. Mast cells are increasingly being recognized as key cells for connective tissue homeostasis, remodelling and repair. They also express relevant angiogenic activity. Their granules indeed contain proteases and cytokines which are known to exert “trophic” effects (survival, growth and chemotactic) on different cells, such as fibroblasts, myofibroblasts, smooth muscle cells, neurons and endothelial cells. Therefore, the “provisional” term “Mastzellen” seems more and more appropriate for describing these cells. Ehrlich also observed that mast cells did not strictly belong to a diffuse perivascular system (according to Waldeyer’s concept of “Perivaskuläresysteme”), despite their characteristic arrangement close to capillaries. This is indeed an absolutely correct statement. Mast cells often localize far from blood vessels and also express a series of biological properties that are not related to microvascular functions. He argued that mast cells could also be found around areas of developing tissues. The close relationship between mast cells and tumour growth is of extremely actual interest in the sense proposed by Ehrlich. In addition, he pointed out that the use of aniline dyes was of the utmost importance for identifying mast cells. Reactivity of aniline with a “still undetermined chemical substance” stored in the granules was the sole reliable procedure that would enable the microscopist to recognize these cells with certainty. We now know that aniline dyes interact with the highly acidic glycosaminoglycan residues contained within mast cell granules. This reaction, in turn, determines the characteristic metachromasia of such structures. We acknowledge that his advice, not simply to consider cell morphology but to base cell identification upon a specific histochemical reaction, was an extremely modern concept. As to the origin of mast cells, we now know that they do not differentiate from fibroblasts, as suggested by Ehrlich. He could not imagine, however, that these cells derive from precursors of the haematopoietic lineage and complete their differentiation in peripheral tissues. This was certainly more than he could determine with the simple support of a light microscope and some histological dyes. On 17 January 1879, the Physiological Society of Berlin heard a remarkable paper by Paul Ehrlich about the mast cells that he had discovered as a medical student 2 years previously. Ehrlich pointed out that the granules of mammalian mast cells not only do display great avidity for basic dyes but also tend to alter the shade of the dye (metachromasia). Later (with one his pupils), he stressed a second characteristic feature of the mast cell granules in many species, their solubility in water
10
2 Paul Ehrlich’s Doctoral Thesis
(Westphal, 1891). Michels (1938) wrote that “uncounted pages of useless and misleading research have been the result of the failure on the part of many investigators to heed the admonition originally given by Ehrlich and Westphal, that the mast cell granules are soluble in water and that to preserve them tissues must be fixed in 50% alcohol and stained in alcoholic thionine”. Ehrlich then went on to study the staining reactions of blood cells, laying the foundations of modern haematology on the basis of the specific affinities of the leucocytes for various dyes (Ehrlich, 1891; Ehrlich and Lazarus, 1898). He encountered cells with basophilic, metachromatic granules, and thus came to recognize two types of mast cells: the first – derived from, and living in, the connective tissues (tissue mast cells) and the second – the counterpart of the neutrophil polymorph and eosinophil leucocyte – with its origin in the bone marrow and habitat is in the peripheral blood (blood mast cell, basophil or mast leucocyte). Meanwhile, Ehrlich (1891) had discovered basophilic granular cells in human blood, though so far only in myeloid leukaemia. Nevertheless, with characteristic insight he at once perceived that, in higher vertebrates, the blood mast cells are true leucocytes stemming from precursors in the bone marrow. By the time that his textbook (Ehrlich and Lazarus, 1898) was revised in 1909, the evidence for the myeloid origin of the blood mast cell was complete (Jolly, 1900). Later work established that mast cells and basophils share several notable features besides staining properties. Both cell types represent a major source of histamine and other potent chemical mediators implicated in a wide variety of inflammatory and immunological processes. To study the presence and significance of mast cells in pathological conditions is again to acknowledge our debt to the pioneer observations of Ehrlich who described two situations in which connective tissue may be overnourished, in chronic inflammation and the environs of tumours. Here there exist a lymph stasis, a damming up of tissue fluid rich in nutriment, whereby certain fixed connective tissue cells are stimulated to become mobile, to multiply and to convert some of the abundant extracellular materials into specific intracellular granules. According to Ehrlich, the mast cells were “indices of the nutritional state of the connective tissue”, increasing during periods of hypernutrition, diminishing during periods of relative starvation. Ehrlich found many mast cells in tumours, especially carcinoma, but it was left to his pupil Westphal (1891) to recognize that the cells tend to accumulate at the periphery of carcinomatous nodules rather than within the substance of the tumour (Westphal, 1891). The number of mast cells within the perivascular and interstitial connective tissue of different neoplasias has been reported to be increased. In some cases this phenomenon is a characteristic feature of the lesion. During the 60 years that followed Ehrlich’s discovery, the research on the mast cell was almost entirely histological. Controversies arose but their resolution for the most part merely emphasized the soundness of Ehrlich’s original work. However, the functional biology of mast cells yet resisted clarification until recently, as their role in promoting the non-specific inflammatory reaction and in different immune responses could be elucidated. Also neoplasias arising from mast cells have been elusive to clinicians, haematologists and pathologists.
2
Paul Ehrlich’s Doctoral Thesis
11
The origin of mast cells remained obscure for many years. It is now accepted that mast cells arise from pluripotential haematopoietic cells in the bone marrow that express CD34, c-kit and CD13 (Kirshenbaum et al., 1999). This was demonstrated for the first time by Kitamura et al. (1978), who performed in vivo experiments using genetically mast cell-deficient mutant mice. However, in contrast to other cells of the haematopoietic stem cell lineage, which differentiate in the bone marrow before being released into the circulation, mast cells do not circulate as mature cells, but in small numbers as committed progenitors. The progenitors complete their maturation with concomitant phenotypic diversity after moving into diverse peripheral tissues. The concept of “mast cell heterogeneity” has represented a focal point in recent discussion of mast cell biology and it emphasizes that different mast cell populations exhibit significant variation in multiple potentially important aspects of their phenotype. Mast cells from different species, from different sites in the same species and even from the same organ in one species can vary in their response to stimuli and inhibitors of mediator release. Observations of histochemical and functional heterogeneity of mast cells, first given a sound basis in the 1960s by Enerbäck (1966, 1986), are now receiving increasing attention (Galli, 1990). Enerbäck reported that, in contrast to mast cells in rat skin, mast cells in the intestinal mucosa were sensitive to routine formalin fixation and could not be identified in standard histological sections. However, after appropriate fixation and sequential staining with Alcian blue and safranin, the mucosal mast cells stained blue in comparison to the connective tissue mast cells which stained with safranin and were red. There is no disease, biological condition or animal model yet identified that exhibits an absolute lack of mast cell from which or in which their biological role might be inferred. Mast cells are most commonly regarded as key effectors in the pathogenesis of allergic diseases. However, an exciting development in the study of mast cell biology was the discovery that mast cells can generate or release various cytokines, which indicate a key role played by mast cells also in diverse pathophysiological processes, such as chronic inflammatory processes, wound healing, angiogenesis, fibrosis and tumours. We wish to conclude this historical note with the remark that all scientists involved in the field of mast cell research should acknowledge their debt to Ehrlich’s pioneering observations. By reading his doctoral thesis, in particular, we cannot help admiring the precocious scientific debut of a far-seeing genius who won the Nobel Prize 30 years later.
Chapter 3
The Contribution of Gianni Bonadonna to the History of Chemotherapy
3.1 Historical Background The German scientist Paul Ehrlich (1845–1915) coined the word “chemotherapy” in reference to the systemic treatment of both infectious diseases and neoplasms. Ehrlich’s use of in vivo rodent model systems to develop antibiotics for treatment of infectious diseases led Georges Lowes, at Roswell Park Memorial Institute in Buffalo, New York, in the early 1900s to develop inbred rodent lines bearing transplanted tumour that could be used to screen potential anticancer drugs. This in vivo system provided the foundation for mass screening of novel compounds (Marchall, 1964). Alkylating agents represent the first class of chemotherapeutic drugs to be used in clinical settings. The introduction of chemotherapy in the fifth and sixth decades of the twentieth century has resulted in the development of curative therapeutic interventions for patients with several types of advanced solid tumours and haematologic neoplasms. In the past four decades, the clinical oncologist Gianni Bonadonna (Fig. 3.1), working at the Department of Medical Oncology of the “Istituto Nazionale Tumori” in Milan, Italy, introduced two extremely innovative protocols for the treatment of breast cancer and Hodgkin’s disease. Here, an overview of these two milestones in the history of chemotherapy is traced.
3.2 The Cyclophosphamide, Methotrexate and Fluorouracil (CMF) Regimen in the Treatment of Breast Cancer Up to the mid-1970s, the treatment strategy for primary breast cancer based on radical (and even superadical) resection of the primary tumour, en bloc with full dissection of axillary content, represented the unchallenged curative method (Fisher and Gebhardt, 1978).
Published in “Cancer Chemotherapy and Pharmacology”, 60:309–312, 2007
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_3, C Springer Science+Business Media B.V. 2010
13
14
3
The Contribution of Gianni Bonadonna to the History of Chemotherapy
Fig. 3.1 A portrait of Dr. Gianni Bonadonna
Over the past three decades, revolutionary changes have occurred in the locoregional management of primary breast cancer. As a result, radical and extended radical mastectomy has been relegated in the archives of surgical history and today there are few, if any, indications for radical mastectomy. The first trials of adjuvant chemotherapy in the treatment of breast cancer were launched in the 1950s, but it was not until the late 1960s that the first modern trials of combination chemotherapy were initiated. Since the 1970s, randomized trials have addressed many fundamental questions related to adjuvant chemotherapy. In May 1972, Paul P. Carbone of the National Cancer Institute (NCI) showed to Bonadonna the Medicine Branch Annual Report, including the initial NCI data on a quadruple drug regimen, cyclophosphamide, methotrexate, fluorouracil and prednisone (CMFP). The results of the study conducted in clinical-disseminated breast cancer and published later by Canellos and co-workers (1974, 1976a) showed a remarkable response rate (complete remission, 20% – partial remission, 40%) with a median duration of response of 8 months. Combination chemotherapy was developed based on the rationale that combining agents with different mechanisms of action and non-overlapping toxicities would increase treatment benefit, prevent or delay the emergence of drug resistance without significantly worsening morbidity or quality of life. Using the large case series of breast cancer patients available at the Milan Cancer Institute, Bonadonna drafted two CMF protocols, one for clinically advanced breast cancer and one for surgically resectable tumours with histologically positive axillary nodes. Carbone together with Pietro Becalossi and Umberto Veronesi in Milan approved and supported Bonadonna’s proposal. Bonadonna and co-workers first started the trial on advanced breast cancer and their findings with CMF (De Lena et al., 1975; Brambilla et al., 1976) were similar
3.3
(ABVD) Regimen in the Treatment of Hodgkin’s Disease
15
to those obtained by the Eastern Cooperative Oncology Group (ECOG) (Canellos et al., 1976b). In 1976, Bonadonna et al. presented the first report on the efficacy of CMF as adjuvant treatment for node-positive breast cancer (Bonadonna et al., 1976). These results, along with those reported in a similar population of patients by the National Surgical Adjuvant Breast Project (Fisher et al., 1975), raised hopes that chemotherapy could have a more central role in the primary management of breast cancer. The ease of administration and the virtual absence of severe acute toxicity made CMF the most frequently used combination of drugs in clinical practice in oncology, as well as the regimen against which all new systemic adjuvant treatments were tested. In 1995, Bonadonna et al. reported the results of 20 years of follow-up of their original series of women who had a radical mastectomy and who were randomly assigned to receive no further treatment of CFM chemotherapy for 12 monthly cycles. The long-term results continued to show a significant overall benefit for adjuvant chemotherapy (Bonadonna et al., 1995). In 2005, Bonadonna et al. reported the results of 30 years of follow-up and confirmed that the effects of such a regimen are long lasting and may benefit patients with favourable and unfavourable prognostic indicators, at the cost of minimal long-term sequelae. Moreover, the poor prognosis associated with unfavourable indicators in patients treated locoregionally alone was improved by administration of adjuvant CMF (Bonadonna et al., 2005).
3.3 The Adriamycin, Bleomycin, Vinblastine and Dacarbazide (ABVD) Regimen in the Treatment of Hodgkin’s Disease Among the unexpected results of both world wars was the development of the high toxic, but therapeutically useful mustard gas derivative, nitrogen mustard [methyl bis(beta-chloroethyl) amine]. The observation made during the First World War that mustard gas poisoning caused leucopenia and the exposure of military seamen to mustard gas in the Second World War, as a consequence of the explosion of a ship containing material manufactured for use in chemical warfare, led to the observation that alkylating agents caused marrow and lymphoid hypoplasia (Hersh, 1968; Infield, 1971). Gilman and Philips (1946) conducted the first clinical trial with nitrogen mustard in patients with malignant lymphomas at Yale University in 1942. This material was the first modern antitumour drug to regularly produce significant response in Hodgkin’s disease (Goodman et al., 1946). Before 1960, chemotherapeutic agents to treat Hodgkin’s disease were used only for palliation. The first chemotherapy study to have an impact on the management of patients with Hodgkin’s disease was published in 1963 (Scott, 1963). Of the 89
16
3
The Contribution of Gianni Bonadonna to the History of Chemotherapy
patients with advanced Hodgkin’s disease who received a conventional induction course of nitrogen mustard, 40 patients with satisfactory response were randomized to receive either no further treatment or continuous treatment with the newly developed oral alkylating agent, chlorambucil. In 16 patients who received chlorambucil, the time for relapse averaged 35 weeks compared with 11.7 weeks without further treatment. The treatment of Hodgkin’s disease with multiple chemotherapeutic agents was developed from sequential studies defining curative treatment for acute lymphoblastic leukaemia in children (Skipper et al., 1964). In 1963, a pilot study was initiated to test the feasibility of using a combination of chemotherapeutic agents, cyclophosphamide, vincristine, methotrexate, prednisone, followed by radiation therapy for the treatment of Hodgkin’s disease (Moxley et al., 1976). In 1964, the nitrogen mustard, vincristine, prednisone, procarbazine (MOPP, “M” for nitrogen mustard, “O” for oncovin, the brand name for vincristine, and “PP” for prednisone and procarbazine) scheme was conceived, followed by the initial report of Lacher and Durant (1965) employing chlorambucil and vinblastine in combination with the first regimen that achieved cure in proportion of patients with advanced lymphoma (De Vita et al., 1970). An 80% complete remission rate was noted. This was a fourfold increase over results achieved with the best use of single agents of the day, and those remissions proved durable and appeared to influence survival. The impressive survival curves in MOPP-treated patients published in 1970 indicated that it was possible to cure advanced Hodgkin’s disease with combination chemotherapy. In 1975, investigators at the NCI (De Vita et al., 1975) reported the cure of a small number of patients with advanced stage diffuse large cell lymphoma with the C-MOPP (cyclophosphamide, vincristine, prednisone, procarbazine) drug combination. The next year, the first or many reports appeared attesting to the efficacy of the doxorubicin-containing CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) regimen in intermediate- and high-grade lymphomas. The observation that approximately 20% of the treated patients failed to achieve complete remission of their lymphoma, coupled with the relative insensitivity of the tumour patients who experienced short remissions, suggested that the primary cause of treatment failure was the presence and overgrowth of cells resistant to the drugs in the MOPP regimen. Following the introduction of MOPP for the treatment of Hodgkin’s disease, other effective drug combinations were reported of which the ABVD programme was developed by Bonadonna (1975a). This four-drug regimen included adriamycin a new anticancer antibiotic available for clinical use in the summer of 1968, bleomycin, vinblastine and dacarbazide. The selection of the four agents was based on the evidence of the anti-lymphoma properties of each individual drug and on their non-overlapping sensitivity profiles with MOPP. A randomized trial was mounted in 1973 to test whether ABVD chemotherapy could induce a complete remission comparable with that of MOPP chemotherapy (Bonadonna et al., 1975b). Overall, six cycles of either regimen yielded a comparable incidence of complete remission, and this trend had an influence on the
3.3
(ABVD) Regimen in the Treatment of Hodgkin’s Disease
17
5-year freedom from progression- and relapse-free survival rates. This study showed that ABVD was as effective as MOPP in inducing durable remission in advanced Hodgkin’s disease. Later, a larger randomized study, which also included radiation therapy, proved that ABVD was able to improve long-term treatment outcome compared to MOPP (Bonadonna, 1982a). The higher therapeutic activity of ABVD, which is easy to administer, devoid of severe side effects and well tolerated by the patients, was confirmed in many other studies. Salvage treatment with ABVD in patients failing during or soon after MOPP yielded higher complete remission rates (46%) compared with the opposite sequence, i.e. salvage MOPP in ABVD-resistant patients. In the first study, Bonadonna and Santoro (1982a) compared the efficacy of ABVD versus MOPP in advanced Hodgkin’s disease previously untreated with chemotherapy and, through a cross-over design, they tested either regimen in resistant patients (Bonadonna and Santoro, 1982b). The following study, aimed at assessing the relative efficacy and long-term complications of a combined modality approach, with three cycles of either MOPP or ABVD delivered before and after extensive irradiation in patients with stages II B and III disease (Santoro et al., 1987). In the third study, Bonadonna et al. (1986) designed an effective light-drug programme alternating cycles of MOPP and ABVD. The findings demonstrated a superiority of the alternating regimen over MOPP alone in the achievement of complete remission. This regimen was particularly effective for those with advanced and symptomatic Hodgkin’s disease (stages III B and IV B). A further study reported the results of a new trial aimed at assessing whether a more rapid alternation of the eight drugs could improve treatment outcome (Viviani et al., 1996). These treatment findings, along with those reported in randomized studies conducted by the Cancer and Leukaemia Group (CALBG) (Canellos et al., 1992) and by the North American Intergroup Study in which the MOPP/ABVD hybrid regimen was tested against ABVD (Duggan et al., 2003), confirm that ABVD should be considered the standard regimen for the treatment of Hodgkin’s disease. Over the past 25 years, the ABVD regimen has gradually replaced the MOPP regimen as the most commonly used chemotherapeutic regimen for all stages in Hodgkin’s disease. In head-to-head trials, it has been found to be superior to MOPP in both early stage and advanced stage disease (Canellos et al., 1992). Although the regimen carries some risk of cardiac and pulmonary toxicity, it is not associated with leukaemogenesis or sterility, the two major complications associated with MOPP.
Chapter 4
Sir Frank Macfarlane Burnet and the Clonal Selection Theory of Antibody Formation
4.1 Introduction Many accounts have been previously published concerning the ontogeny of clonal selection theory of antibody formation. In 1988, J. Lederberg published his reflections on Darwin and Ehrlich entitled “Ontogeny of the clonal selection of antibody formation” (Lederberg, 1988). In 1995, The “FASEB Journal” published a special article of D.R. Forsdyke entitled “The origins of the clonal selection theory of immunity as a case study for evaluation in science” (Forsdyke, 1995). In 2002, “Nature Immunology” published a commentary of A.M. Silverstein entitled “The clonal selection theory: what it really is and why modern challenges are misplaced” (Silverstein, 2002). In 2007, on the occasion of the 50th anniversary of the publication of Frank Macfarlane Burnet’s clonal selection theory, “Nature Immunology” published an editorial, an essay signed by G.J.V. Nossal, a historical commentary containing the reproduction of the original version of the 1957 Burnet’s paper (Editorial, 2007; Nossal, 2007; Hodgkin et al., 2007). In the same year “Nature Reviews in Immunology” published a viewpoint entitled “Reflections on the clonalselection theory” (Cohn et al., 2007). Finally, in 2008, G. Ada published a landmark entitled “The enunciation and impact of Macfarlane Burnet’s clonal selection theory of acquired immunity” (Ada, 2008). My proposal in this chapter, after my previous historical review articles dedicated to other protagonists of the history of immunology of the twentieth century, namely R.A. Good, F.A.P. Miller and B. Glick (Ribatti, 2006; Ribatti et al., 2006a, b), is to outline the seminal work of F. M. Burnet.
4.2 Biographical Profile Sir Frank Macfarlane Burnet was born in Traralgon, in eastern Victoria, Australia, on 3 September 1899. He was educated at the Victoria State Schools and at Geelong College, completing his medical course at the University of Melbourne, where he Published in “Clinical and Experimental Medicine”, 9: 253–258, 2009
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_4, C Springer Science+Business Media B.V. 2010
19
20
4
Sir Frank Macfarlane Burnet and the Clonal Selection Theory of Antibody Formation
Fig. 4.1 A portrait of Sir Frank Macfarlane Burnet
graduated as a Medical Doctor in 1923. In 1926, he was awarded a Beit Fellowship for Medical Research and worked for a year at the Lister Institute, London. In 1932, he spent a year at the National Institute for Medical Research, Hampstead, London. Otherwise, he has worked continuously at the Hall Institute in Melbourne and from 1944 to 1965 he was director of this institute and professor of experimental medicine in the University of Melbourne (Fig. 4.1). From 1951 to 1956, Burnet concentrated on studies of the genetics of influenza virus. In parallel with his work on virology, Burnet had always been interested in the immune response and in 1941 he published a monograph analysing the nature of antibody formation (Breinl and Haurowitz, 1930). In 1948, he proposed a new hypothesis on antibody production based on analogies with adaptive enzymes (Mudd, 1932). In 1957, Burnet changed the direction of his own work and that of the institute, abandoning virology and concentrating on immunology. In 1960, Burnet received the award of the Nobel Prize in physiology or medicine jointly with Sir Peter Medawar, for an immunological discovery, acquired immunological tolerance. In 1965, Burnet retired from directorship of the institute, and Doctor Gustav Nossall was appointed as director. After retirement, he moved into the School of Microbiology in the University of Melbourne. During the 12 years he was at the University of Melbourne, Burnet produced 13 books, initially on immunology and subsequently on human biology, aging and cancer, as well as a fourth edition of his first book. In 1969 and again in 1974, international symposia were organized by Nossal to celebrate his 70th and 75th birthdays. In 1978, at the age of 78, Burnet left the School of Microbiology and moved to his home, where he produced two more books.
4.3
The Instructive Theories of Antibody Production
21
In November 1984, he was operated on for cancer on the rectum, but secondary lesion was discovered early in August 1985, and he died on 31 August 1985.
4.3 The Instructive Theories of Antibody Production During the 1930s Breinl and Haurowitz (1930) and Mudd (1932) proposed a hypothesis to account for antibody production which was clarified and reformulated by Pauling (Pauling, 1940). The “instructive” theory, as it was proposed by Pauling, arose from the universally accepted concept that the antibody repertoire had to be transcendental to protect the animal. As the number of genes had to be lower than the number of antigens that are recognized, the only solution seemed to be the “instructive” theory. According to the “instructive” theory, the antibody was synthesized, or according to Pauling folded, in specific ways in spatial contact with the antigenically significant parts of the antigen, which acted as a template. No supporting findings were found for this concept, but much contrary evidence accumulated. In 1941, Burnet summarized his views on antibody production in a monograph (Burnet et al., 1941). Because of the apparently almost infinite variety of antibodies, Burnet accepted an instructive hypothesis, but suggested that the antigen impressed a complementary pattern not on the globulin molecule, but on some cellular components. In Section 4.1, Burnet identified the problem of antibody synthesis as linked to key biological questions: the conditions governing protein synthesis in the living cell and the capacity of biological systems to be modified in reaction. This second point, the capacity of an organism to remember its first encounter with a given antigen and to elaborate an accelerated secondary immune response, had brought Burnet to reflect on immunological behaviour of young animals. Burnet’s interpretation of the finding that chick embryo was able to tolerate foreign tissues was that “in some ways the embryonic cells seem to be unable to recognize and resent contact with foreign material in the way adults cells do.” In a second edition of the monograph (Burnet and Fenner, 1949), Burnet developed a new hypothesis for the process of antibody production itself based on an analogy with adaptive enzyme process in bacteria, in which bacteria produce large amounts of an enzyme that is capable of breaking down to specific sugar that is placed in culture medium. Burnet suggested that an antigen might induce modifications of those enzymes involved in globulin synthesis so that a protein with the required specificity might be formed. This hypothesis concerned the manner in which the body normally failed to make antibodies to its own components, the “self-marker” concept, according to which the distinction between the “self” and the “not-self” is based on a limited number of recognizable components in each cell, the combination of which constitute the specific “self-pattern” of the organism. Burnet reported data concerning the effect that mice and calves exposed continuously to antigens during embryonic life failed to produce antibodies if exposed to these antigens in adult life. He concluded that “If in embryonic life expandable cells form a genetically distinct race are implanted and established, no antibody response should develop against the foreign cell antigen when the animal takes on independent existence.”
22
4
Sir Frank Macfarlane Burnet and the Clonal Selection Theory of Antibody Formation
In a book published in 1953, Burnet proposed that antigens induced the formation of stable cytoplasmic synthetic units, “almost a case of inheritance of acquired characteristics” (Burnet, 1953). In 1956, recognizing the growing importance of nucleic acids in protein synthesis, Burnet proposed a complicated model in which antigens associated with either cytoplasmic RNA or with DNA served as a template for synthesis of a new protein (Burnet, 1956).
4.4 The Selective Theories of Antibody Production In 1900, Paul Ehrlich published a selective theory of antibody formation, called the “side chain theory” (Ehrlich, 1900). The theory proposed that the antibody located on cell surface could serve as a receptor for antigen. Following reaction with a foreign antigen, the receptor/antigen complex would be discarded from the cell surface. The affected cell overproduces more side chains which would become circulating antibodies. Ehrlich differed from his contemporary Elie Metchinkoff who ascribed the production of antibodies to macrophages. Ehrlich suggested that this function might be a specialized characteristic of “haematopoietic tissue”. The key feature of Ehrlich’s model was that there was a pre-existing repertoire of specificities for a variety of antigens. The antigens would act to select from among the specificities. He understood that antibody molecules have a distinct structure, and that parts of the molecule that react with complement might differ from parts reacting with specific antigens. He also recognized that antibodies themselves are potential antigens and that distinct anti-antibodies might be raised against different parts of the antibody molecule and introduced the idea of a mechanism of self/notself-discrimination (“horror autotoxicus”). The long time interest of Burnet in the theory of antibody production had been stimulated by a paper published by Jerne in 1955 (Jerne, 1955) that proposed a “natural selective theory” for the process, rather than the “instructive” theory. In his paper entitled “The natural selection theory of antibody formation”, Jerne wrote that “Among the comparatively small number, perhaps a few thousand, of antigen– antibody systems investigated, cross-reactions are by no means rare, suggesting that the number of specific configurations which globulin molecule can exhibit is large but limited. Since normal mammalian serum contains more than 1017 globulin molecules per millimetre, these may include a million 1011 fractions of different specificity. This would seem an amply sufficient number” (Jerne, 1955). According to Jerne, the function of an antigen was to combine with those globulins with which it made a chance fit and to transport the selected globulins to antibody-producing cells, which would then make mainly identical copies of the globulin presented to them. Jerne proposed that antigen–antibody complexes are taken into cells where the antibody is then replicated. Jerne’s concept that antibodies were natural globulins was attractive, but his concepts of globulin randomization and replication, like nucleic acids, were incompatible with new information that was available on DNA. Jerne’s 1955 “natural selective theory” was basically a revision of Ehrlich’s “side chain theory”. Both theories held that antibodies, not cells, were selected by antigens.
4.5
The Clonal Selection Theory
23
Burnet rejected Jerne’s theory based on the implausibility of a self-replicating antibody molecule, and instead suggested that the B-cell receptor was located on cells that replicate. David Talmage, replacing randomly diversified globulins with randomly diversified cells, provided a “cell selection theory” of antibody formation. Talmage’s working hypothesis favoured a selective model and proposed that the unit responsible for expansion was the antibody-producing cell itself. In a review of 1957, he wrote “. . .it is tempting to consider that one of the multiplying units in the antibody response is the cell itself. According to this hypothesis, only those cells are selected for multiplication whose synthesized product has affinity for the antigen injected. This would have the disadvantage of requiring a different species of cell for each species of protein produced, but would not increase the total amount of configurationally information required on the hereditary process (. . .). The cellular hypothesis is compatible with current concepts that the configuration of a protein molecule is determined solely by information contained in the hereditary units of the cell, the nucleic acid” (Talmage, 1957). According to Talmage, to have selection there must be diversity so the concept was based on the idea that there were many different cells in the body that made antibodies, each making a different molecule or species of antibody. So, the clonal selection theory is basically a combination of the selection hypothesis and the cell hypothesis. Talmage discussed supporting evidence from the kinetics of antibody response, from immunological memory and from the fact that myeloma tumours often result in a massive production of one globulin randomly selected from the family of normal globulins. Talmage never received the recognition he deserved from his seminal contribution.
4.5 The Clonal Selection Theory Following on Jerne’s and Talmage’s ideas, Burnet suggested the clonal selection approach. The basis of clonal selection theory is that the specific capacity of a cell to react immunologically either as a cell or as a producer of antibody is conferred on by genetic processes and not by the intrusion of a pattern from the antigen. The clonal selection theory advanced the concept that antibodies were natural globulins that possessed an affinity for antigens and are selected from a large group of pre-existing globulins. Burnet published a short paper dated 21 October 1957, published in the “Australian Journal of Science”, which was little more than a newsletter published by the “Australian and New Zealand Association for the Advancement of Science”, and described as a “preliminary account”, where he cited Talmage’s paper “Talmage has suggested that Jerne’s view is basically an extension of Ehrlich’s side chain theory of antibody production and that replicating elements essential to any such theory were cellular in character ab initio rather than circulating protein which can replicate only when taken into an appropriate cell. Talmage does not elaborate
24
4
Sir Frank Macfarlane Burnet and the Clonal Selection Theory of Antibody Formation
this point of view but clearly accepts it as the best basis for the future development of antibody theory. Before receiving Talmage’s review we had adopted virtually the same approach but had developed it from what might be called a ‘clonal’ point of view” (Burnet, 1957). Burnet asserted that in a population of mesenchymal cells that are a certain number of different clones and that it is characteristic of any such clone that is composed of cells which are immunologically competent towards a certain antigenic determinant. If the antigenic determinant makes effective contact with such a cell, it is stimulated to proliferate (clonally expand) and under appropriate circumstances some of the descendant cells will be converted into plasma cells and produce antibodies that combine with the antigen. If the antigen is part of the surface of a virus or bacterium, then the antibody labels that organism as foreign (“not self”). The organism is then ingested by phagocytic cells and degraded. In this “preliminary account” but in the meantime a “fundamental paper”, Burnet wrote that “Among [antibodies] are molecules that can correspond probably with varying degrees of precision to all, or virtually all, the antigenic determinants that occur in biological material other than characteristic of the body itself. Each type of pattern is a specific product of a clone [lymphocytes] and it is the essence of a hypothesis that each cell automatically has available on its surface representative reactive sites equivalent to those of the globulin they produce. (. . .) It is assumed that when an antigen enters the blood or tissue fluids it will attach to the surface of any lymphocyte carrying reactive sites which corresponds to one of its antigenic determinants. (. . .) It is postulated that when antigen-[antibody] contact takes place on the surface of a lymphocyte the cell is activated to settle in an appropriate tissue. (. . .) and there undergo proliferation to produce a variety of descendents. In this way, preferential proliferation will be initiated of all those clones whose reactive sites correspond to the antigenic determinants on the antigen used. The descendents will [be] capable of active liberation of soluble antibody and lymphocytes which can fulfil the same functions as the parental forms” (Burnet, 1957). In the book published in 1959, Burnet clarified several points in the theory and expanded upon the earlier hint on the importance of somatic mutation. He developed further the subsidiary hypothesis of clonal abortion during fetal life to explain tolerance to self-antigens (Burnet, 1959). He described in detail how the clonal selection theory could explain a broad range of immunological phenomena, including immunological memory, original antigenic sin, the effects of adjuvants, mucosal immunity, natural antibodies and autoimmunity.
4.6 Evidence Supporting the Theory Initially, there was opposition to the idea that the body could make enough different natural globulins to react specifically with every conceivable antigen. In response to this opposition, in 1959, Talmage demonstrated how an almost unlimited number of different combinations of approximately 50,000 different globulins might explain immunological specificity (Talmage, 1959).
4.7
Concluding Remarks
25
Gordon Ada and Nossal collaborate to device a microscopic assay to test Burnet’s concept. It was based on the ability of flagella-specific antibody, on addition to bacteria cultures, to prevent bacterial migration as observed under the microscope. Individual rats were immunized with two serological distincta Salmonella flagella to establish if single plasma cells from the immunized rats produced antibodies to both flagella (double producers) or only to one? Nossal and Lederberg published a paper in 1958 with results from 62 positive cells showing that no double producers had yet been found and that the single antibody-producing cells in culture made only one antibody (Nossal and Lederberg, 1958). Nossal confirmed these data in his Ph.D. thesis, reporting that following examination of nearly 1,500 anti-flagellaproducing cells, no double producers were found (Nossal, 1960). In 1958, White found no double producers in his system (White, 1958). Attardi et al. (1958) using a method based on the inactivation of bacterial phages demonstrated that about 10% of active cells were double producers. A new technology introduced by Jerne, the plaque assay (Jerne and Nordin, 1963), allowed several groups to show that individual cells from donors who were immunized with multiple antigens made antibodies specific for only one of those antigens. Colonies of antibody-forming cells could be found in the spleen of animals given whole-body gamma radiation and injected with small numbers of spleen cells (Palyfair et al., 1965). Raff and co-workers showed that incubation of lymphocytes with antigen could aggregate all the surface immunoglobulin on antigen-binding cells and this indicated that only immunoglobulin on the surface of these cells was antibody of a single specificity (Raff et al., 1973). In 1970s cellular immunology itself exploded and molecular immunology also advanced rapidly. The structure of immunoglobulin and its genetic basis were completely developed. In 1976, following the development of the hybridoma technology it was possible to develop a technique to make large amounts of a monoclonal antibody (Kohler and Milstein, 1975). In this process, the antibody-secreting cells, which have a short half-life, are fused with myeloma cells, resulting in an immortal line of antibody-secreting cells. Jerne, George J.F. Kohler and Cesar Milstein shared the 1984 Nobel Prize in physiology or medicine for this achievement.
4.7 Concluding Remarks At a symposium on antibodies held at Cold Spring Harbor in 1967, Burnet delivered the opening address with a paper outlining the historical background to the evolution of his clonal selection theory. He wrote “Jerne postulated that gammaglobulin molecules are continuously being synthesized in an enormous variety of different configurations. The origin of the diversity is left unexplained. When an antigen intrudes into the body, sooner or later globulin molecules of the appropriate natural pattern will become attached to the antigenic molecules or particles. The complex is then taken up by phagocytic cell where, by hypothesis, the globulin can be released from the antigen. Such globulin molecules either in the macrophage of
26
4
Sir Frank Macfarlane Burnet and the Clonal Selection Theory of Antibody Formation
after transfer to another cell were said to serve as a ‘signal for the synthesis or reproduction of molecules identical to those introduced, i.e. of specific antibodies.’ Even in 1955 this seemed wholly inadmissible. Most other aspects of the new theory were highly acceptable but the basic flaw seemed to be a fatal one” (Burnet, 1967). In the same symposium, Jerne acknowledged how Burnet’s scientific originality had contributed to the modern immunology: “Sir Macfarlane Burnet must have been pleased not only to witness at this symposium the vindication of his clonal selection theory of acquired immunity, but also to see how his stimulating ideas have led to a great proliferation of immunologists and to know that the fate of immunology is deposited in so many capable hands” (Jerne, 1967). In a paper published in 1988 and dedicated to the memory of Burnet, Joshua Lederberg wrote that “By the 1967 Cold Spring Harbor Symposium, the clonal selection theory was an undeniable fundament for almost every investigation of the chemistry of antibodies or the biology of immunocytes. It was also clear that further progress would depend on the propagation of antibody-forming cells as clones” (Lederberg, 1988). It was primarily for the formulation of the clonal selection theory that Burnet was included in the London Sunday Times 1969 list of the 1,000 people who had made the twentieth century. His lasting contribution was expressed in this form: “By turning immunological theory upside down he has kept the world’s immunologists busy and happy for the last decade.” In 1967, Jerne explained that there were two kinds of immunologists, who hardly spoke to each other: the “trans-immunologists”, who were interested in the structure of the antibody molecule and its binding to the antigen, and the “cisimmunologists”, who were interested in the events following antigen exposure. The consequence of acceptance of the clonal theory of antibody synthesis on the other hand, and the progress in understanding the mechanism of protein synthesis and the role of cellular receptors in the regulation of immune phenomena on the other hand, reduced the distance between the cis- and trans-immunologists.
Chapter 5
The Contribution of Bruce Glick to the Definition of the Role Played by the Bursa of Fabricius in the Development of the B-Cell Lineage
5.1 Fabricius ab Aquapendente Girolamo Fabrici or Fabrizio (Hieronymus Fabricius ab Aquapendente) (Fig. 5.1) was the student and successor of Andreas Vesalius (1514–1564) and Gabriele Fallopius (1523–1562). He practiced and taught anatomy at Padova for more than 50 years (Smith et al., 2004). Harvey was one of his pupils. In addition to his demonstration of the valves of the vein, Fabricius is best known for his description of the bursa that bears his name. A manuscript titled De Formatione Ovi et Pulli found among his lecture notes was published in 1621. It contains the first description of the bursa (Adelman, 1967) “The third thing which should be noted in the podex is the double sac [bursa] which in its lower portion projects toward the pubic bone and appears visible to the observer as soon as the uterus already mentioned presents it self to view” (p. 147). The sac-like organ has ever since been known as the bursa of Fabricius (BF). It overlies the dorsal surface of the terminal portion of the gut in birds. In the 5- to 6-day-old chick embryo, it arises as a dorso-caudal outpouching near the cloaca that takes the form of a median lamina of endodermal epithelium permeated with spherical vacuoles of various sizes that eventually coalesce to create a lumen (Hamilton, 1952). The bursa grows considerably during development and changes from round to oval. Hypertrophy of the mesoderm surrounding the bursal epithelium produces longitudinal plicae that project into its lumen (Romanoff, 1960). Between the 13th and 15th day, epithelial cells lining the plicae thicken and extend into the tunica propria as epithelial buds. These buds then separate from the epithelium. Lymphopoiesis is active in those that form the medullus of the bursal follicle (Ackerman and Knouff, 1959; Ackerman, 1962). Follicles may be present during late embryonic development (after 16 days). Even so, they are best observed by light microscopy at hatching and during the early growth of the BF (Frazier, 1974). The BF has 8,000–12,000 total follicles,
Published in collaboration with Enrico Crivellato and Angelo Vacca in “Clinical and Experimental Immunology”, 145:1–4, 2006
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_5, C Springer Science+Business Media B.V. 2010
27
28 5 Contribution of Bruce Glick to the Definition of the Role Played by the Bursa of Fabricius Fig. 5.1 A portrait of Fabricius ab Aquapendente
each comprising a cortex, medulla, corticomedullary border and follicle-associated epithelium (Glick, 1983). The presence of M cells within this epithelium explains the movement of antigen from the lumen into the medulla, where immature B cells develop (Sayesh et al., 2000).
5.2 The BF Plays a Major Role in the Development of Antibody-Mediated Immunity In December 1952 Bruce Glick, at the Ohio State University, demonstrated that the BF grows most rapidly during the first 3 weeks after hatching. He thus became convinced that functional investigation of the BF would be successful only if it was removed (bursectomy, BSX) within this period.
5.3
BSX Do Not Abrogate the Antibody Response to Cellular Antigens
29
In 1954 Timothy S. Chang, a graduate student, needed birds to develop antibody against Salmonella; the only birds available were those of Glick. He therefore injected 6-month-old pullets with Salmonella-type O antigen to obtain serum with a high antibody titre. Several pullets died and none of those that survived produced antibodies. It was then found that the entire batch had been BSX during the period of rapid bursa growth. Glick deduced that the absence of BF was responsible for this failure, since non-BSX pullets produced normal antibody titres (Glick, 1955), and designed two experiments to substantiate this conclusion. Equal numbers of male and female white leghorns were BSX at 12 days of age and injected six times at intervals of 4 days with S. typhimurium O antigen. At 7 weeks, 7 of 10 BSX birds and 2 of 10 controls failed to produce antibody (Glick, 1955). The second experiment employed larger numbers of birds and two breeds: 89.3% of the BSX birds failed to produce antibody compared with only 13.7% of the controls (Chang et al., 1955; Glick et al., 1956).
5.3 BSX Do Not Abrogate the Antibody Response to Cellular Antigens Next, Chang et al. (1957) showed that BSX at 2 weeks BSX was more effective in suppressing antibody production than at 5 or 10 weeks of age. Failure of BSX to eliminate all antibody production suggested the existence of a brief period in embryo development during which the BF could be functional. The first experiments to evaluate the existence of a functional period for the BF (Meyer et al., 1959) took advantage of the regressive influence of androgens on the post-hatched BF (Glick, 1957; Kirkpatrick and Andrews, 1944). Treatment of 9- to 12-day-old embryo with testosterone prevented immunoglobulin production and lymphoid development, and presumably destroyed the stem cells which are necessary for B-cell production. Subsequent injection of bovine serum albumin (BSA) into chicks hatched from eggs injected on day 5 of incubation revealed complete immunoglobulin elimination, while chicks from eggs injected with testosterone on day 12 or 13 possessed significantly reduced levels of antibody (Mueller et al., 1960, 1962). The BF was generally absent in 19-day embryos that had received testosterone prior to the 8 day (Warner and Burnet, 1961). Various BSX methods cause more or less complete B-cell defect and agammaglobulinaemia. They include testosterone treatment (Glick, 1957; Glick and Sadler, 1961; Glick, 1964; Meyer et al., 1959; Papermaster et al., 1962b; Warner and Burnet, 1961), cyclophosphamide administration (Eskola and Toivanen, 1974; Lerman and Weidanz, 1970), colchicine treatment (Romppanen and Sorvari, 1980), X-irradiation (Cooper et al., 1966) and surgical operations (Fitzsimmons et al., 1973; Van Alten et al., 1968).
30 5 Contribution of Bruce Glick to the Definition of the Role Played by the Bursa of Fabricius
Cooper et al. (1966) showed that chickens irradiated at hatching and also subjected to total BSX develop peripheral small lymphocytes in a normal fashion, reject skin syngeneic grafts and display normal graft-versus-host reactions. They are, however, prevented from developing the two clearly definable immunoglobulins and are completely unable to form circulating antibodies, even in response to strong antigenic stimulation.
5.4 Immunoglobulin Synthesis Regulation The BF as a site of antibody synthesis was investigated by Glick and his co-workers in the early 1960s. Two experiments gave conflicting results. In the first, pheasant bursa cells produced antibodies to bovine immunoglobulins (Kerstetter et al., 1962), whereas in the second the BF was unable to produce plaque-forming cells to sheep red blood cells (SRBC) (Dent and Good, 1965). The reason that pure B cells did not produce antibodies against SRBC became obvious only later, when Henry Claman’s group discovered the requirement of B–T-cell cooperation for antibody production against this and other T-dependent antigens (Claman and Chaperon, 1969). These authors investigated the participation of both T and B cells in the in vitro response of spleen cells from mice immunized with the hapten NIP coupled to a nonimmunogenic isologous gamma globulin carrier (MGG) (Claman and Chaperon, 1969). Glick failed to identify antibody to BSA in the BF from 3-week-old intravenously immunized chickens (Glick and Whatley, 1967). The B cell differentiates in the BF and is able to produce immunoglobulins on the 14th day of embryo development. The first immunoglobulin is the large 1,000,000 molecular weight molecule called immunoglobulin M (IgM), followed by IgG on the 20th day and then by IgA (Cooper et al., 1969; Kincade and Cooper, 1971). Two equally plausible explanations of this sequence were advanced. One held that IgM producing B cells give rise to the IgG and IgA B cells, the other proposed sequential intrabursal development of isotype-committed sublineages. Kincade and Cooper (1973) found that the antiμ-mediated inhibition of IgM producing B cells also inhibited development of the IgG and IgA B cells. Moreover, the combination of embryonic anti-μ administration and post-hatching BSX resulted in permanent agammaglobulinaemia. These experiments indicated that while all chicken B cells express IgM initially, they can switch to the production of other isotypes. Neonatal anti-μ antibody treatment also inhibited mouse B-cell development and antibody production of all Ig isotypes (Lawton et al., 1972).
5.5 Delineation of the Thymic and Bursal Lymphoid Systems in the Chicken Functional dissociation of the chicken immune system based on differences in thymic and bursal influences was suggested originally by Szenberg and Warner (Szenberg and Warner, 1962).
5.5
Delineation of the Thymic and Bursal Lymphoid Systems in the Chicken
31
Following the Glick’s demonstration of the crucial function of the BF in development of antibodies and the immune responses related to their production, in 1958 Francis A.P. Miller in Australia discovered the role of thymus-derived cells for cellular immunity (Ribatti et al., 2006b). Miller’s experiments indicate that (a) thymectomy is associated generally with a diminution in the lymphocyte population and (b) the earlier in life the thymectomy is performed, the greater the deficiency of lymphocytes in other lymphoid organs (Ribatti et al., 2006b). Robert Good and his collaborators (notably Max D. Cooper) developed the idea of the B and T cell concepts, demonstrating the essential role of the thymus in development of cellular immunity functions other than antibody production in the chickens (Cooper et al., 1965, 1966). Chickens were thus the first source of the two-component concept of immunity. Sublethal X-irradiation of newly hatched chickens was needed to clarify the roles of the thymus and the BF in development of the two separate and functionally different lymphoid systems (Cooper et al., 1966). The BSX and irradiated birds were completely devoid of germinal centres, plasma cells and the ability to produce antibodies, yet they had perfectly normal development of thymocytes and lymphocytes elsewhere in the body that mediated cellular immune reactions, while the thymectomized and irradiated birds were deficient in lymphocytes that mediated cellular immunity as assessed by skin graft rejection, delayed-type hypersensitivity and graft-versus-host reaction, but still produced germinal centres, plasma cells and circulating Igs. Van Alten et al. (1968) used BSX within the eggs to show that the two-component concept was clearly evident even in the absence of X-irradiation. The BF and the thymus are central lymphoid organs in the chicken, essential for the ontogenetic development of their adaptive immunity. Surgical removal of one or both of organs in the newly hatched chickens, followed by sublethal X-irradiation, led to the recognition of two morphological distinct cell systems in the peripheral lymphoid tissues of the spleen, gut and other organs, and clear differentiation of their functions. The thymus controls development of all cell-mediated immunities, including delayed reactions, allograft immunities and other immunologic functions. In addition to being a basic immunologist, Good also held the position of professor of pediatrics at the School of Medicine in Minneapolis, Minnesota. He thus had access to the various cases of immunodeficiencies that had led him to recognize similarities between Bruton’s agammaglobulinaemia and Glick’s BSX, on the one hand, and Di George syndrome and Miller’s thymectomized mice on the other hand. In fact, removal of the BF from the egg inhibits germinal centres and plasma cells and prevents antibody production (Perey and Good, 1968). BSX chickens are strikingly similar to patients with Bruton’s X-linked agammaglobulinaemia (Peterson et al., 1965) and in ovo thymectomized chicks are strikingly similar to those with Di George syndrome, while with severe combined immunodeficiency disease (SCID) are similar to chickens bursectomized and thymectomized in the newly hatched period (Peterson et al., 1965). The major immunodeficiencies, Bruton’s disease, Di George syndrome and SCID are thus mimicked by BSX or thymectomy in ovo.
32 5 Contribution of Bruce Glick to the Definition of the Role Played by the Bursa of Fabricius
5.6 BF Equivalent in Mammals and Other Vertebrates The BF is present in all avian orders but is absent in mammals. Several structures, however, have been identified as “bursa equivalents”, such as gut-associated lymphoid tissues in rabbits and ungulates and bone marrow in rodents and primates, including humans. Archer et al. (1963) found that the rabbit sacculus rotundus located at the ileocoecal valve, like the BF, develops within follicular outpouchings of the lower gut. Immediate extirpation of this organ in neonates resulted in an impressive and lifelong immunodeficiency of antibody production (Archer et al., 1963; Perey and Good, 1968). Knight and Crane (1994) have since demonstrated that the BF and the appendix-sacculus rotundus mediate very similar influences on the humoral system. However, the sacculus rotundus has not emerged as the BF equivalent organ. Owen et al. (1974) found that Ig-bearing cells first appear in the liver during mouse embryogenesis and employed fetal liver organ cultures to show that B cells are generated in haematopoietic tissue. Moreover, Owen et al. (1975) found that, after their colonization with haematopoietic stem cells, fetal long bones can also generate B-cell ex vivo. This finding suggested that mammalian B-cell generation is a multi-focal process that shifts from one haematopoietic environment to another during development, to continue throughout life in the bone marrow. It is now clear that in mammals, B cells remain and differentiate in the bone marrow, a most convenient etymological coincidence, as the nomenclature for B cells as bone marrow-derived cells does not change.
Chapter 6
Miller’s Seminal Studies on the Role of Thymus in Immunity
6.1 Introduction The thymus was recognized as such by the Greeks. Its purpose, however, has been rendered clear in only recent times, since immunologists began to elucidate the origin and function of peripheral lymphocytes in disease. The word itself may be derived from a Greek word, thýmos, meaning the heart or soul. The ancient Greeks used very young animals for their sacrifices. They noted a large mass of tissue in the chest above the heart that extended into the neck and concluded that it must be the seat of the soul. Galen first described the morphology of the gland and many medieval students regarded it as being at the heart of good health. Abnormalities of the human thymus are associated with several syndromes. The first of these associations was recognized by Weigert in 1901, in his classical description of the relationship between myasthenia gravis and thymic tumours. By the 1950s, recognition of the thymus as the site of production of lymphocytes had been well established. Their immunological competence was demonstrated unequivocally by Billingham et al. (1956) and Gowans et al. (1961). Circulating lymphocytes were divided eventually into T and B cells, following the identification by Glick et al. (1956) of the bursa of Fabricius as the source of antibody-producing cells (Szenberg and Warner, 1962). Interest in the thymus for the generation of T cells was aroused when mammals were found to be the same in this respect.
6.2 The Thymus in Mouse Leukaemia In 1951, Gross injected filtered material extracted from the tissues of mice with spontaneous leukaemia into mice with low leukaemic strains which, provided
Published in collaboration with Enrico Crivellato and Angelo Vacca, in “Clinical and Experimental Immunology”, 144:371–375, 2006
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_6, C Springer Science+Business Media B.V. 2010
33
34
6 Miller’s Seminal Studies on the Role of Thymus in Immunity
Fig. 6.1 A portrait of Dr. Francis Albert Pierre Miller
newborns were injected, then developed a high incidence of leukaemia (Gross, 1951). In 1958, Francis Albert Pierre Miller (Fig. 6.1), after his medical degree and an internship at the Royal Prince Alfred Hospital in Sydney, received a research fellowship that enabled him to read for his PhD at the Chester Research Institute, an Institute of Cancer Research, in South Kensington, London, where Dr R.J.C. Harris suggested that he should investigate the pathogenesis of lymphocytic leukaemia in mice. This form, whether spontaneous or induced by irradiation or chemical agents, was known to involve the thymus, and adult thymectomy had prevented its development (Miller, 1961b). The role of the thymus in virus-induced leukaemia, on the other hand, was unknown. Miller thymectomized 4- to 5-week-old C3Hf/Gs mice that he had injected at birth with leukaemic extracts. None developed leukaemia (Miller, 1959a). Implantation of syngeneic thymus from normal mice into these thymectomized mice restored their potential for leukaemia development (Miller, 1959b). They were healthy until some time after weaning, when many lost weight and died prematurely, whether or not inoculated with virus. Miller thus suggested that “The thymus at birth may be essential for life” (Miller, 1960).
6.3 The Thymus Is Essential for Normal Development of the Immune System Miller’s neonatally thymectomized mice showed a marked deficiency of lymphocytes both in the circulation and in the lymphoid tissues. The lymphocyte/polymorph ratio did not increase significantly during the first 8 days of life
6.3
The Thymus Is Essential for Normal Development of the Immune System
35
and at 6 weeks of age was not much higher than at birth. Involution of the lymphoid tissues was a characteristic feature. At 6 weeks, the spleen was greatly reduced in size (Miller, 1962c) and displayed ill-defined, inactive follicles, with little basophilia, poor cellularity and few mitoses (Miller, 1961a, b). There were few germinal centres. The lymph nodes were also considerably smaller and displayed inactive follicles and poor cellularity. Peyer’s patches were present, but smaller and less cellular than in the controls. Circulating lymphocytes had been shown by Billingham et al. (1956) and Gowans et al. (1961) to be immunologically competent and able to reject skin grafts. Miller found that thymectomized mice failed to reject skin from foreign mouse strains (Miller, 1962b). Female C57BL mice thymectomized at 2 weeks of age developed no immune response to syngeneic male skin grafts, but rejected allogeneic grafts (Miller, 1961a). Thymectomy after 3 weeks of age was unaccompanied by any significant impairment of this response. In Miller’s words, therefore “During embryogenesis the thymus would produce the originators of immunologically competent cells, many of which would have migrated to other sites at about the time of birth. This would suggest that lymphocytes leaving the thymus are specially selected cells” (Miller, 1962b). The time was by no means ripe, however, for acceptance of the view that the thymus was endowed with an immune function. By contrast with the small lymphocytes taken by thoracic duct cannulation and with spleen and lymph node cells, thymocytes in general displayed little ability to initiate immune reactions after adoptive transfer. Thoracic duct lymphocytes homed from blood into lymphoid tissues, the only exception being the thymus, in which very few small lymphocytes lodge (Gowans et al., 1962). The production of antibody-forming plasma cells and the formation of germinal centres, so prominent in spleen and lymph nodes, were absent in normal thymus tissue. Defects in immune responsiveness had never been documented in mice whose thymus had been removed during adulthood, and hence it was thought that “the thymus gland does not participate in the control of the immune response” (MacLean et al., 1956). The immune defects observed after neonatal thymectomy were soon confirmed independently by other investigators (Arnason et al., 1962; Martinez et al., 1962a). When mice were thymectomized 1 or more weeks after birth, i.e. when their lymphoid system and the immune mechanisms had partially developed, only negligible effects were observed. The foreign experiments indicate that (a) thymectomy is associated generally with a diminution in the lymphocyte population and (b) the earlier in life the thymectomy is performed, the greater the deficiency of lymphocytes in other lymphoid organs. Furthermore, while diminished lymphocyte production continues in mice thymectomized during adulthood, it eventually stops in those thymectomized at birth. Two mechanisms may account for this defect: (a) the thymus, through cell migrations, populates and continually replenishes other lymphoid tissues. This would be of major importance in very early life and decrease with age. (b) The
36
6 Miller’s Seminal Studies on the Role of Thymus in Immunity
thymus produces a non-cellular or humoral factor which regulates lymphocyte production and maturation, particularly during early life. Metcalf (1956) claimed to have demonstrated in the thymus a specific lymphocytosis stimulating factor (LFS) whose activity was apparently associated with the epithelial reticular cell complex of the medulla. This factor was heat-labile and filtrable, but non-dialysable.
6.4 If the Immune System Was Destroyed in the Adult, Would the Thymus Still Be Involved in Lymphopoiesis? Grégoire and Duchateau (1956) reported that implants of thymus tissue depleted lymphocytes by irradiation, and this comprised mainly radioresistant epithelial stroma-stimulated lymphopoiesis, whereas lymph node and muscle implants had no such effect. Because total body irradiation destroys lymphoid tissues, Miller predicted that recovery of immune functions following irradiation would be thymus dependent. Adult mice were thymectomized and subjected to total body irradiation. To prevent death they were given bone marrow cells, a source of haematopoietic stem cells. Control, non-thymectomized mice, treated in this way recovered normal lymphoid functions within 6–8 weeks. The thymectomized mice did not (Miller, 1962c; Miller et al., 1963). In the adult, therefore, the thymus is still required to re-establish defence mechanisms depleted as a result of some accident or disease.
6.5 Cell Transfer Studies Immune functions can be restored to animals thymectomized at birth or thymectomized and irradiated in adult life by infusing lymphocytes or implanting thymus tissue. Miller investigated the effect of injecting lymphoid cells into neonatally thymectomized mice and found the following: 1. Syngeneic thymus cells from 1-day-old mice given intravenously to newborn mice immediately after thymectomy did not prevent runting, lymphoid atrophy or immunological failure (Miller, 1962b). 2. Syngeneic lymphoid cells from 8-week-old mice presensitized against Ak skin, on injection into 10-week-old neonatally thymectomized C3H mice carrying healthy Ak skin grafts, for more than 1 month conferred adoptive immunity. The Ak skin was rejected within 12 days and the mice showed evidence of immunity to a second-set graft (Miller et al., 1964). 3. Allogeneic lymphoid cells from 2-month-old mice caused a severe graftversus-host (GVH) reaction when injected intravenously into newborn mice immediately after thymectomy (Miller, 1962c).
6.6
The Functional Anatomy of the Thymus
37
Lymphocytes restored immune capabilities, but only if the donor was syngeneic (Miller, 1964). If it was of a different strain the thymectomized host wasted and died, because injected lymphocytes, being immunologically competent, reacted against the foreign tissues of their host and brought about a fatal GVH reaction. Neonatally thymectomized mice, implanted with syngeneic thymus tissue soon after birth, developed a normal immune system. When grafted with foreign thymus tissue, they were specifically tolerant of thymus donor-type skin only (Miller, 1962b, 1963). This finding led Miller to postulate that “When one is inducing a state of immunological tolerance in a newly born animal, one is in effect performing a selective or immunological thymectomy” (Miller, 1962b). In other words, the precursors of thymic lymphocytes differentiating in the presence of foreign cells and with specificities for the foreign antigens would be deleted, implying that the thymus might be the site in which self–non-self discrimination occurs and self-tolerance is imposed. This idea received strong support from Macfarlane Burnet, who in a lecture in June 1962 at the University of London stated “If, as I believe, the thymus is the site where proliferation and differentiation of lymphocytes into clones with definable immunological functions occurs, we must also endow it with another function, the elimination or inhibition of self-reactive clones” (Burnet, 1962). Experiments combining the techniques of thymectomy and injection of marked thymus cells led to the conclusion that thymus-derived cells were small lymphocytes, able to circulate in blood and lymph for many months in rodents and years in man (Miller and Osoba, 1967).
6.6 The Functional Anatomy of the Thymus The thymus is an encapsulated gland that undergoes remarkable age-related changes. It is replenished eventually with fatty areolae that replace its normal lymphoid tissue. Disappearance of thymic structures, however, is not complete and some islands of functionally competent tissue are still recognizable in senility. The gland displays a lobuled pattern, with distinct cortical and medullary compartments that is related strictly to its function, namely the production of fully competent circulating T cells bearing the form of the T-cell receptor. Two main cell populations are recognizable (Anderson and Jenkinson, 2001). The stromal population consists of fixed ectodermal-derived, keratin-positive epithelial cells, which form a three-dimensional network occupying the cortex and the medulla. These cells are comprehensively referred to as thymic epithelial cells. The second population constitutes the parenchyma and is composed of thymocytes plus a variety of antigen-presenting cells, including interdigitating cells, macrophages and small amounts of B cells. Thymocyte precursors reach the thymus from the blood. In the embryo they come first from the rudimentary liver and then from the bone marrow. On entering
38
6 Miller’s Seminal Studies on the Role of Thymus in Immunity
the gland they undergo proliferation, lineage commitment and selection, which is largely under the control of thymic epithelial cells.
6.7 Thymocyte Positive and Negative Selection Two selective processes accompany thymocyte migration, proliferation and differentiation (Sprent and Kishimoto, 2002). The final result is the apoptosis of about 96% of thymocytes and only 3–5% become fully competent T cells, i.e. cells able to recognize foreign antigens, but unresponsive towards self-antigens, that eventually enter the circulation as naive T cells. Until a few years ago it was believed that negative selection occurred in the thymic cortex, whereas positive selection occurred in the medulla. Today, both compartments are thought to provide selective signals leading to cell survival or death. Uncommitted haemopoietic progenitors therefore enter the gland through postcapillary venules at the cortical–medullary junction. They move first towards the subcapsular region and acquire T lineage commitment. Subcapsular thymocytes express both helper/inducer and suppressor/cytotoxic phenotypes (CD4+ CD8+ or “double-positive” thymocytes). These cells then return to the medulla and, during their passage through specific cortical zones, undergo either positive or negative selection under the guidance of both contact and paracrine signals from the epithelium (Petrie, 2002). Epithelial cells present thymocytes with an enormous repertoire of self-peptides conjugated to major histocompatibility complex moieties. Positive selection is obtained when these complexes on the surface of thymic cells are recognized by the T-cell receptor located on the thymocyte surface. This interaction generates survival signals that rescue thymocytes from apoptosis. By contrast, negative selection occurs when such interactions are too strong or too weak, and apoptosis is promoted by the absence of survival factors. This is called “apoptosis by neglect” (Klein and Kyewski, 2000). Only about 4% of double-positive cells are selected positively to generate mature CD4+ or CD8+ cells. Once this phenotype is acquired, thymocytes enter the medulla where they remain for a few days before being released into the peripheral lymphoid pool.
6.8 The Medulla and Central Tolerance The structure of the medulla is substantially different from that of the cortex. Its cell population is very composite and the stromal epithelium itself is more compact and less arborized. A close relationship has been detected recently between the architecture of the medullary stroma and the emergence of autoimmune disorders in the mouse. The correct expression of the product of the autoimmune regulator (AIRE) gene correlates with a normal organization of the medullary stroma
6.8
The Medulla and Central Tolerance
39
(Zuklys et al., 2000). By contrast, mutations in the AIRE gene are responsible for an autoimmune syndrome called APECED (autoimmune polyendocrinopathy– candidiasis–ectodermal dystrophy), characterized by the loss of self-tolerance to multiple organs and abnormal structure of the thymic medulla (Ramsey et al., 2002). This compartment thus seems essential for instructing thymocytes to self-tolerance (central tolerance) (Durkin and Waksman, 2001). Some 1–5% of epithelial cells in the medulla express a mosaic of “ectopic” tissue-specific molecules, such as parathyroid hormone, thyroglobulin, insulin, Creactive protein (Farr and Rudensky, 1998). This has led to the formulation of the theory of “promiscuous” gene expression, wherein the medulla represents a collection of antigenic structures that do not strictly belong to the thymus, but pertain to peripheral epithelial tissues (Derbinski et al., 2001). Anatomists have long recognized that the medulla contains groups of cells displaying the organization, morphology and functional activity of other epithelial tissues, namely Hassal’s bodies (squamous epithelial cells that resemble epidermal epithelium), cystic “organoid” structures with the morphological and phenotypic features of respiratory epithelium, neuroendocrine cells, myoid cells and solitary thyroid follicles. It is thus a highly specialized structural and antigenic environment that recapitulates the spectrum of an epithelial “self” by creating a type of “immunological homunculus” (Derbinski et al., 2001). This epithelial organization, along with its numerous antigen-presenting cells (interdigitating cells, macrophages, B cells), makes the medulla the best candidate for the accomplishment of functions such as deletion of auto-reactive clones of thymocytes and induction of immunological tolerance.
Chapter 7
The Fundamental Contribution of Robert A. Good to the Discovery of the Crucial Role of Thymus in Mammalian Immunity
7.1 Idiopathic Acquired Agammaglobulinaemia Associated with Thymoma Robert A. Good (Fig. 7.1) began his intellectual and experimental queries related to the thymus in 1952 at the University of Minnesota, initially with paediatric patients. However, his interest in the plasma cell, antibodies and the immune response began in 1944, while still in Medical School at the University of Minnesota in Minneapolis, with his first publication appearing in 1945 (Good and Campbell, 1945). In 1953, his colleague Richard Varco asked Good to consult on a 54-year-old male patient who had initially presented to his chest clinic in June 1951. The patient complained of having suffered at least 17 bouts of pneumonia during the previous 8 years and a pronounced susceptibility to infection, which had increased, concomitant with the appearance and extirpation of a benign thymoma, occupying almost the entire thymic gland (MacLean et al., 1956).
Fig. 7.1 Robert A. Good with two young patients Source: www. robertagoodarchives.com
Published in “Immunology”, 119:291–295, 2006
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_7, C Springer Science+Business Media B.V. 2010
41
42
7 Discovery of the Crucial Role of Thymus in Mammalian Immunity
The interesting thing to Good about this patient was that he also carried a diagnosis of “acquired agammaglobulinaemia”, a markedly deficient ability to produce antibodies and significant deficits of all or most of the cell-mediated immunities. Surgical removal of the tumour, which was primarily an epithelial stromal overgrowth of the thymus, did not correct the immunodeficiencies in this patient. Since then seven cases of the combined occurrence of these two disorders have been reported (Martin et al., 1956; Ramos, 1956; Lambie et al., 1957; Gafni et al., 1960) and in no instance did removal of the thymic tumour restore immunological function or correct the protein deficit. Good described a new syndrome that would carry his name: “Good syndrome: thymoma with immunodeficiency” (Good and Varco, 1955). The clinical characteristics of Good syndrome are increased susceptibility to bacterial infections with encapsulated organisms and opportunistic viral and fungal infections. Subsequently, Good saw several patients with thymic tumours, which regularly presented with immunodeficiencies, leucopenia, lymphopenia and eosinophylopenia. Plasma cells, however, were not completely absent: the patient was severely hypogammaglobulinaemic rather than agammaglobulinaemic.
7.2 The Role of Thymus in Immunity The association of thymoma with the profound and broadly based immunodeficiency provoked Good’s group to ask what role the thymus plays in immunity. Good (Good, 1954; Bridges and Good, 1960) and others (Janeway et al., 1953; Gitlin et al., 1956) found that the patients lacked all of the subsequently described immunoglobulins (Bridges and Good, 1960). These patients were found not to have plasma cells or germinal centres in their haematopoietic and lymphoid tissues. They possessed circulating lymphocytes in normal numbers (Good, 1955). Good decided to investigate the possibility that the thymus had something to do with adaptive immunity, and under his direction, Zak and MacLean performed thymectomies on 4- to 5-week-old rabbits, but they found that thymectomy had no demonstrable effects on the antibody response (MacLean et al., 1956, 1957). In the discussion of the second paper the authors noted that, although their laboratory investigation had not led to discover the exact function of thymus, they believed that their patient represented an experiment of nature that suggested that the thymus does, indeed, play a crucial role in immunity.
7.3 The Effects of Neonatally Thymectomy In the mouse and other rodents, immunologic depression is profound after thymectomy in neonatal animals, resulting in considerable depression of antibody production, plus deficient transplantation immunity and delayed-type hypersensitivity
7.3
The Effects of Neonatally Thymectomy
43
(Waksman et al., 1962). Speculation on the reason for immunologic failure following neonatal thymectomy has centred in the thymus as a source of cells or humoral factors essential to normal lymphoid development and immunologic maturation. At the University of Wisconsin, a second group of investigators was engaged in endocrinologic studies which led to the first experiments on neonatally thymectomy in rabbits. Three independent groups of experiments showed that neonatal thymectomy has a significant effect on immunologic reactivity: (i) the studies of Fichtelius et al. (Fichtelius et al., 1961) in young guinea pigs showed that the depression of antibody response is slight, but significant; (ii) the experiments of Archer, Good and co-workers in rabbits (Archer and Pierce, 1961; Archer et al., 1962; Good et al., 1962) and mice (Good et al., 1962; Martinez et al., 1962a, b; Dalmasso et al., 1963); and (iii) the studies by Miller at the Chester Beatty Research Institute in London (Miller, 1961a, 1962a, b). In rabbits, the effects of neonatal thymectomy on antibody production were variable both from animal to animal and antigen to antigen (Archer et al., 1962). In the mouse transplantation immunity was sufficiently affected by neonatal thymectomy to permit skin transplants across the H2 histocompatibility barrier and even across species barriers in some instances, and production of antibodies to certain antigens was almost entirely eliminated (Martinez et al., 1962a, b; Miller, 1962a; Papermaster et al., 1962a). Good wrote that “the simultaneous occurrence of acquired agammaglobulinemia and benign thymoma in a human being, suggested that the thymus might participate in the control of antibody formation. [...] It still seems likely that some essential relationship exists between the thymic tumor and the acquisition of an acquired agammaglobulinemia. A second case of acquired agammaglobulinemia with thymoma present itself and strengthens the conviction that the two phenomena are related in some essential manner.” Parrott and East (1964) showed that the effect on antibody production was a quantitative one, and that with potent antigens such as haemocyanin and pneumococcal polysaccharides, many of the neonatally thymectomized mice produced significant amounts of antibody. Neonatally thymectomized mice were particularly vulnerable to homologous disease when injected with parent strain lymphoid cells (Parrott, 1962). Their lymphoid tissues showed minimal lymphoid development and their circulating lymphocyte levels were greatly reduced, although most strains have adequate levels of gamma globulins and appreciable, if not entirely normal, number of plasma cells. The cells of spleen and lymph nodes of neonatally thymectomized mice had very low immunologic activity and splenomegaly and other manifestations of graft-versushost reactivity were not evident unless the dosage of cells was multiplied several times (Dalmasso et al., 1963). Rabbits thymectomized during the neonatal period revealed gross deficiencies in the distribution of T lymphocytes. Parrot et al. (1966) showed that thymically derived lymphocytes occupied the paracortical regions in the lymph nodes and
44
7 Discovery of the Crucial Role of Thymus in Mammalian Immunity
periarterial regions in the spleen, the so-called thymus-dependent regions (Cooper et al., 1966). Other specialized areas in the lymph nodes are located in the far cortical regions, the B-zones, where germinal centre developed. Overall these experimental data indicate that the mouse has one primary central lymphoid organ, the thymus, and that this is a key source of cells or humoral substances, or both, that are necessary to the normal maturation of the peripheral lymphoid tissues and to normal development of immunological capabilities. The fact that neonatally thymectomized mice retained a small measure of immunologic reactivity suggests that the influence of the thymus on immunological development may already have been excised to some extent before birth.
7.4 The Restorative Effect of Thymomas or Thymus Grafts on Thymic Function Stutman, in Good’s laboratory, demonstrated that nonlymphoid thymomas induced restoration of immunological functions in neonatally thymectomized mice (Stutman et al., 1967) and that when thymomas were grafted into allogenic hosts, immunological restoration was mediated by lymphoid cells of host type (Stutman et al., 1968). Comparable results were obtained with free thymus grafts (Stutman et al., 1969a). When the treatment of neonatally thymectomized host was delayed after neonatal thymectomy, a decrease in the restorative capacity was observed (Stutman et al., 1969b). These results indicate that a population of cells in thymectomized hosts capable of responding to the action of thymus or thymomas decreased progressively with time after neonatal thymectomy (Stutman et al., 1969b).
7.5 Removal of Either the Thymus or Bursa of Fabricius The bursa of Fabricius and the thymus are “central lymphoid organs” in the chicken, essential to the ontogenetic development of adaptive immunity in that species. Studies by Papermaster and co-workers in Good’s laboratory (Papermaster et al., 1962a, b) indicated that bursectomy in the newly hatched chicks did not completely abolish immunological potential in the adult animal but rather produces a striking quantitative reduction insufficient to eliminate the homograft reaction. The failure of thymectomy in newly hatched chicks to alter the immunological potential of the maturing animal probably only reflected the participation of the bursa of Fabricius in the development of full immunological capacity. In 1963, Max Cooper started a long series of experiments in Good’s laboratory. He removed either the thymus or bursa of Fabricius from some newly hatched chicks, and both from others. Then, to destroy all peripheral lymphoid components, he subjected the chicks to intense x-ray irradiation just below the lethal level, to destroy cells that might have seeded earlier from the thymus and bursa or that could have been influenced by postulated thymic and bursal humoral factors. Finally,
7.6
The Importance of a Functional Thymus in Human
45
Cooper waited several weeks until the experimental and irradiated control animals recovered from the irradiation effects. Results showed that the bursectomized and irradiated birds were completely devoid of germinal centres, plasma cells and the capacity to make antibodies yet they had perfectly normal development of thymocytes and lymphocytes elsewhere in the body that mediated cellular immune reactions (Cooper et al., 1965, 1966). On the other hand, thymectomized and irradiated animals were deficient in lymphocytes that mediated cellular immunity as assessed by skin graft rejection, delayed-type hypersensitivity and graft-versus-host assays, but they still produced germinal centres, plasma cells and circulating immunoglobulins (Cooper et al., 1965, 1966). Like thymectomized mice, the thymectomized and irradiated birds had impaired antibody responsiveness to antigens. Birds subjected to combined thymectomy, bursectomy and irradiation had severe cellular and humoral immune system deficit (Cooper et al., 1965, 1966). Cooper et al. (1965) postulated that a lymphoid stem cell population exists that is induced to differentiate along two distinct and separate cell lines related to two central lymphoid organs. In birds this developmental influence is exercised by the thymus and the bursa of Fabricius. Removal of one or both in the early post-hatching period has strikingly different influences on immunologic function in the maturing animals (Cooper et al., 1965, 1966). The thymus in the chicken functions exactly as does the thymus of the mouse. It represents the site of differentiation of a population of lymphocytes that subserve largely the functions of cell-mediated immunity. Overall, these data indicate that at some point differentiation along two distinctly different pathways occurs within the lymphoid system and that the critical point seems to focus about two separate central lymphoid organs.
7.6 The Importance of a Functional Thymus in Human There are several clinical observations which serve to underline the importance of a functional thymus to immunological development in man. Good said that one morning in 1952 he “opened his green pediatric journal” and found an article by Odgen Bruton describing the electrophoretic pattern of an 8-year-old patient, seemingly, devoid of gamma globulin, i.e. agammaglobulinaemia. Bruton’s initial reports were published in 1952 (Bruton, 1952; Bruton et al., 1952). He described a patient for whom life was one severe life-threatening bacterial infection after another. Immediately after Bruton’s description of the association of immunodeficiency with agammaglobulinaemia, Janeway’s group in Boston (Janeway et al., 1953; Gitlin et al., 1956) and Good’s group in Minneapolis (Good and Varco, 1955; Good and Zak, 1956) launched studies with one series of patients with X-linked agammaglobulinaemia. This X-linked agammaglobulinaemia was one of the first, if not the first, recognized primary immunodeficiency of man. Most of these patients exhibited normal delayed-type hypersensitivity and rejected primary as well as secondary skin allografts. In these patients thymus was
46
7 Discovery of the Crucial Role of Thymus in Mammalian Immunity
found to be almost normal and thymus-dependent lymphocytes were also normal as well as their functions. Life of these agammaglobulinaemic children involved a succession of life-threatening episodes of infection usually caused by Streptococcus pneumonii, Haemophilus influenzae, meningococci or other high-grade encapsuled pyogenic pathogens including Pseudomonas aeruginosa. Furthermore, those patients who did not produce antibodies did not have normal adaptive immune responses that involved a normal ability to express, terminate and resist recurrence childhood exanthems, e.g. chicken pox and vaccinia virus. They also resisted bacillus Calmette–Guérin infection quite normally (Good and Varco, 1955; Good, 1955, 1954; Good and Zak1956). However, a number of these patients had vestigial thymus tissue weighing less than a gram in most instances, and almost completely lacking in lymphoid cells (Good et al., 1964). In several of these patients, the thymus was not only very small and very deficient in lymphoid development but also showed a failure of migration. The Di George third and fourth pharyngeal pouch syndrome patients had absence or deficiency of all cell-mediated immunological functions. They were also somewhat deficient in antibody production, but not so deficient as the patients with Bruton’s X-linked agammaglobulinaemia. They had low-set, crumpled, abnormally rotated ears and characteristic face with a small mandible, short filtrum and a small low-shaped mouth (Conley et al., 1979). Di George patients had severe deficiencies of small T lymphocytes and profound deficiencies of all cell-mediated immunities, including delayed allergies, deficient allograft immunities and deficiencies in resistance to viruses, fungi and opportunistic infections. Having defined the Di George syndrome as selective deficiency of T-cell development as the result of failure of differentiation of thymus, it seemed likely that the abnormality should be correctable by thymus transplants (Cleveland et al., 1968; August et al., 1970; Biggar et al., 1972). The athymic children described by Di George, who lacked lymphoid cells in the deep cortical areas of the nodes but not at the peripheral areas, seemed the equivalent of the neonatally thymectomized mice and chickens.
Chapter 8
The Fundamental Contribution of Jan C. Waldenström to the Discovery and Study of the So-Called Waldenström Macroglobulinaemia
8.1 Biographical Notes Jan Costa Waldenström was born in Stockolm, Sweden, on 17 April 1906. His grandfather, Johan Anton Waldenström (1839–1879) was professor of internal medicine in Uppsala and his father, Johan Henning Waldenström (1877–1972), was professor of orthopedic surgery in Stockholm. After obtaining his M.D. at the University of Uppsala, Waldenström studied organic chemistry in the laboratory of the Nobel Prize Hans Fischer at the Technische Hochschule in Munich. This led to his classical monograph entitled Studien Über Porphyric in which acute phorphyria for the first time was manifested biochemically by the excretion of uroporphyrin III in the urine (Waldenström, 1937). It was Waldenström who, with Bo Vahlquist, introduced the term “porphobilinogen” in 1939. Returning to the University of Uppsala after 1 year in Munich, he studied diseases associated with an elevation of erythrocyte sedimentation rate (ESR) (Waldenström, 1943). Waldenström’s interest in serum proteins emerged from follow-up studies of patients with constantly elevated ESR. Of utmost importance for these studies was the unique opportunity Waldenström had in Uppsala to collaborate with members of the Institute of Physical Chemistry. In collaboration with Pedersen sedimentation constants and electrical mobility of serum proteins in different diseases with elevated ESR were determined by means of the Swedberg’s ultracentrifuge and Arne Tiselius’s electrophoresis method. Waldenström became professor of theoretical medicine at the University of Uppsala in 1941 and 3 years later was appointed professor of practical medicine at the University of Lund and physician in chief at Malmö General Hospital. He was head of the Department of Internal Medicine at Malmö General Hospital until his retirement in 1972 (Fig. 8.1). Retiring from the university, Waldenström meant more time for work as consultant of the most important Swedish and international institutions, as well as chief editor of Acta Medica Scandinavica. Waldenström received honorary degrees from Published in “Leukemia Research”, 31:435–438, 2007
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_8, C Springer Science+Business Media B.V. 2010
47
48
8
The Discovery and Study of the So-Called Waldenström Macroglobulinaemia
Fig. 8.1 Jan Waldenström at a conference in New York in 1963
the Universities of Oxford, Freiburg, Olso, Paris, Dublin, Mainz, London, Innsbruck and Poitiers. He died on 1 December 1996 at the age of 90.
8.2 Waldenström Involvement in the Discovery and Study of the Waldenström Macroglobulinaemia In 1944 Waldenström described two patients with oronasal bleeding, lymphadenopathy, normochromic anaemia, increased ESR, thrombocytopenia, hypoalbuminaemia, low serum fibrinogen and increased numbers of lymphoid cells in bone marrow (Waldenström, 1944). The oronasal bleeding could have been related to hyperviscosity and the normochromic anaemia could be related to increased plasma volume and reduced erythropoiesis. Hyperviscosity is only observed in 15% of patients at diagnosis and is related to the increasing serum concentrations of the IgM parameters which result in aggregation of red cells and increased viscosity (Gertz and Kyle, 1995). As Waldenström pointed out “It must kept in mind that hyperviscosity syndrome with a characteristic fundus, diffuse bleedings and an enlarged plasma volume with pseudoanemia also exists in a few cases of myeloma with increase in 7S IgG or preferably 12S to 13S, dimeric” (Waldenström, 1986). Prolonged bleeding after lymph node biopsy and bone marrow aspiration, lobar pneumonia and retinal haemorrhages were observed. The bleeding after biopsies could be attributed to IgM interference with platelet function and/or coagulation factors. Neurological complications ranging from mild symptoms (e.g. headache, lightheadedness and fatigue) to severe symptoms (e.g. mental confusion, stroke and
8.2
Waldenström Involvement in the Discovery
49
focal neurological deficits) may be present alongside hyperviscosity (Waldenström, 1948). A number of observations have been published of patients with severe motor disease, mostly resembling motoneuritis multiplex and macroglobulinaemia (Waldenström, 1986). Excess cells in the bone marrow in these patients were lymphoid, not plasma cells as in patients with multiple myeloma. Thus, although the condition had many resemblances to multiple myeloma, the main difference, at least from the standpoint of the marrow, lays in the lymphocytosis as opposed to the morphology of multiple myeloma, i.e. plasmocytosis. The lack of bone disease does distinguish macroglobulinaemia from multiple myeloma, because only 2% of the patients with the former and 75% of the patients with the latter have manifestations of disease in bone. As Waldenström pointed out “The cellular basis of macroglobulinemia is the proliferation of one clone of cells that produces one definitive macroglobulin molecule. These cells may appear in different histological patterns. The most common is diffuse infiltration of the bone marrow but many patients also have enlargement of lymph glands (‘lymphoma’). In some instances the spleen is also invaded and rare patients have a lymphoma picture in the lungs and even in the brain. The characteristic finding is that these cells contain IgM with the same light chain type, either kappa or lamba, as the increased IgM in the serum” (Waldenström, 1986). Although the lymphocytosis in the marrow is usually striking, there has apparently been a reluctance to call the cells concerned typical lymphocytes or to designate the condition present lymphocytic leukaemia even in the presence of the characteristic blood picture of that disease. As Waldenström pointed out “A patient with 200,000 lymphocytes in the blood and only a trace of monoclonal macroglobulin in the blood may be diagnosed as lymphatic leukemia and the IgM is forgotten. On the other hand, a patient with the hyperviscosity syndrome and a very high IgM may have considerable ‘lymphocytosis’ without being labelled lymphatic leukemia” (Waldenström, 1986). Waldenström noticed that most normal serum globulins sedimented with a coefficient of 7S, while in his patients he observed an abnormally large amount of a homogeneous globulin with sedimentation coefficients of 19S and 20S, corresponding to a molecular weight of more than 1,000,000. He postulated that the protein consisted of a giant molecule rather than an aggregation of smaller globulin molecules. The laboratory features described before the development of techniques for paper electrophoresis of plasma proteins allowed to detection of monoclonal immunoglobulins; so the concept of a monoclonal serum peak did not exist. Immunoelectrophoresis classifying serum heavy chains into G, A and M was not generally available. Not until 1937 did Teselius described the separation of serum globulins into α, β and γ and the concept of immunoglobulins was not proposed until late 1950 s. This protein is now known to be a member of the immunoglobulins and is designed as IgM. As we know, many blood disorders can raise the IgM levels. In addition to multiple myeloma, the list included chronic lymphocytic leukaemia, non-Hodgkin’s lymphoma, amyloidosis and monoclonal gammapathies of undetermined significance
50
8
The Discovery and Study of the So-Called Waldenström Macroglobulinaemia
(MGUS). In addition, infections like hepatitis, AIDS, and various rheumatological disorders can also raise IgM levels. In 1962, in view of MacFarlane Burnet’s “Clonal selection theory”, Waldenström suggested that conditions giving rise to unusual concentrations of immunoglobulins be called gammapathies, and that they be subdivided into polyclonal or monoclonal gammapathies, according to their electrophoretic pattern, i.e. whether the immunoglobulin region was diffuse (polyclonal) or characterized by a narrow spike (monoclonal). “Monoclonal gammapathy” is believed to be characterized by the existence of a single clone of immunologically competent cells such as lymphocytes or plasma cells, responsible for the production of one given type of protein (as manifested by the presence of a narrow-banded hyperglobulinaemia or a paraprotein fraction in the electropherogram). In Waldenström’s macroglobulinaemia, the studies of Zucker-Franklin et al. (1962) using both C14-labelled lysine cultures of lymphoid tissues together with immunofluorescence techniques have clearly indicated that medium and large lymphocytes were synthesizing macroglobulin. In the Harvey lecture series in 1961, Waldenström clearly presented the concept of monoclonal versus polyclonal gammopathies (Waldenström, 1961). He described patients with a narrow band of hypergammaglobulinaemia as having a monoclonal protein. Although many of these patients had multiple myeloma, others had no evidence of malignancy and were considered as having “essential hypergammaglobulinaemia” or benign monoclonal gammapathy. Waldenström further correctly regarded the broad band in hypergammaglobulinaemia as a polyclonal increase in proteins. This distinction is extremely important clinically because patients with a monoclonal gammapathy already have or may develop a neoplastic process, whereas patients with a polyclonal gammapathy have an inflammatory or reactive cause of their hypergammaglobulinaemia. Finally, Waldenström emphasized that diffuse or polyclonal hypergammaglobulinaemia was most pronounced and most common in systemic lupus and cirrhosis of the liver and was sometimes seen in rheumatoid arthritis, chronic sialoadenitis, thyroiditis and discoid lupus, i.e. conditions more or less closely related to each other and to the two first mentioned diseases. For 50 years Waldenström has had a never warning interest in the myeloma disease, not only from the point of protein disturbance but also with respect to its clinical manifestations, cellular morphology, prognosis and treatment. His experience of the biological nature of this disease was eminent and his therapeutic program for cytostatic treatment was probably one of the most promising in malignant disorders. As Waldenström pointed out “Treatment with cytostatics is the only way of obtaining constantly low and lasting levels of macroglobulin and of treating the proliferation of lymphatic cells. During the last decades I have personally treated and followed about 40 patients with severe macroglobulinemia. The large majority have been treated with Alkeran only or with modest doses of steroids. The fact that alkylating agents may induce acute leukemia should not be forgotten. We have seen one such patient and another with development of what was called reticular cell
8.2
Waldenström Involvement in the Discovery
51
sarcoma” (Waldenström, 1986). Alkylating agents are the mainstay of treatment for Waldenström macroglobulinaemia (Bayrd, 1961). A study conducted at Mayo Clinic and published in 2000 showed that oral chlorambucil, whether given daily or in intermittent pulses, resulted in an objective response in 75% of patients with a median duration of survival of 4–5 years (Kyle et al., 2000). Plasmapheresis with total plasma exchange can lead to a dramatic control of the symptoms associated with hyperviscosity (Avnstorp et al., 1985). As Waldenström pointed out “Removal of very large quantities of plasma without any transfusion of red cells may normalize the level of red cells and hemoglobin. It is probable that the hyperviscosity of the plasma becomes less detrimental when the concentration of red cells is diminished because this leads to decreased viscosity of the whole blood” (Waldenström, 1986). Waldenström continued to document the features of the disease over a period of 50 years, publishing his last observations shortly before his death (Bjorkholm et al., 1995). Today, the International Waldenström’s Macroglobulinemia Foundation exists to encourage and support Waldenström macroglobulinaemia patient’s and their families. Their web site at http://www.uwmf.com is a suitable place to gain access to the organization.
Part II
Vascular Biology and Angiogenesis
Chapter 9
Foreword
Advances in vascular biology in the last century have contributed to the development of many therapies and preventive strategies. The late 1960 s and early 1970 s were particularly exciting times to be involved in studying the biology of haemostasis. New ideas were revolutioning the concepts of blood coagulation physiology and biochemistry. The article entitled “Giulio Bizzozero and the discovery of platelets” emphasizes the pioneeristic work of the Italian pathologist Giulio Bizzozero, who discovered the platelets in 1882. Recent findings indicate that the activated platelets are crucial regulators of tumour vascular homeostasis in that they prevent tumour haemorrhage. This effect is independent of platelets’ capacity to form thrombi and instead relies on the secretion of their granule content. Targeting platelets secretory activities may represent a new approach to specifically destabilize tumour vasculature. Endothelial cells derived from human umbilical veins were first successfully cultured in vitro in 1973 by Eric A. Jaffe and co-workers. Electron microscopy, Weibel-Palade bodies and the von Willebrand factor antigen were used as morphological, immunohistochemical and functional markers to identify these cells (Jaffe et al., 1973a, b). Angiogenesis is an important process of new blood vessel growth that occurs in the body, both in health and in disease. Over the past 25 years, the number of MEDLINE publications dealing with angiogenesis has risen in a nonlinear fashion with almost 40,000 publications in the year 2009, that included “angiogenesis” as a key word, reflecting the interest among basic scientists and clinicians in this field. Angiogenesis is recognized as one of the critical events required for tumour progression, where cancerous growth is dependent on vascular induction and the development of a vascular supply. The work and name of Wilhelm Roux are tightly associated with modern developmental biology. As it is outlined in the article entitled “A milestone in the study of the vascular system: Wilhelm Roux’s doctoral thesis on the bifurcation of blood vessels” already in his doctoral thesis “On the Bifurcation of Blood Vessels”, Roux addressed two important issues: angiogenesis and developmental physiology. Roux was convinced that causal explanations for structures and functions of organism could be deduced from the laws of physics.
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_9, C Springer Science+Business Media B.V. 2010
55
56
9 Foreword
In 1889, the English surgeon Stephen Paget published his “seed and soil” explanation of the non-random pattern of metastasis and was the first to suggest that interactions between tumour cells and host cells in the microenvironment are critical in regulating tumorigenesis. Certain favoured tumour cells (the “seed”), he said, had a special affinity for the growth-enhancing milieu within specific organs (the “soil”), and hence metastasis only occurred when the “seed” and “soil” were compatible. The importance of several components of the “soil” in regulating tumour growth has since been emphasized: (1) the extracellular matrix; (2) stromal cells and their growth factors and inhibitors; (3) microvessels and angiogenic factors; and (4) inflammatory cells. There is now substantial evidence that tumour growth and progression depend on the cross-talk between malignant cells and their stromal compartment. The article entitled “Stephen Paget and the ‘seed and soil’ theory of metastasis dissemination” summarizes the most important literature data about this matter. The article entitled “The contribution of Roberto Montesano to the study of interactions between epithelial sheets and the surrounding extracellular matrix” outlines the contribution of the Italian scientist Roberto Montesano and colleagues at the University Medical Center of Geneva, Switzerland, who have extensively investigated the mechanisms underlying two morphogenetic processes: angiogenesis and the generation of branching epithelial tubules (tubulogenesis), which are crucial events in the development of most parenchymal organs. Dr. Montesano has contributed to clarify some cellular and molecular mechanisms of angiogenesis and tubulogenesis using an original three-dimensional cell culture system that replicates key events of angiogenesis and tubulogenesis, thereby facilitating molecular analysis. A major advance of this technique over conventional monolayer cultures is that cells can be embedded within a lattice of reconstituted collagen fibrils and it mimics the three-dimensional organization of connective tissue matrices. The results of these studies support the notion that cell interactions with the surrounding extracellular matrix are crucial determinants of cell responses to growth factors and that epithelial tissues morphogenesis is governed by the interplay of two different classes of signalling molecules, i.e. paracrine-acting growth factors and insoluble extracellular matrix components. The article entitled “The seminal work of Werner Risau in the study of the development of the vascular system” is dedicated to a retrospective analysis of the most important contributions of the German scientist Werner Risau in the field of angiogenesis during embryonic development and in the post-natal life, in both physiological and pathological conditions. Risau’s work had a decisive impact on defining the overall nature of neoascularization processes during development and had the capability to integrate different directions in the field of endothelial cell biology research. In particular, he had a special interest in understanding the development, differentiation and maintenance of the blood–brain barrier. Risau very successfully propagated the concept that the same factors, such as basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF), which are essential for the formation of blood vessels during embryonic development, also influence pathological angiogenesis during tumour growth.
9
Foreword
57
The article entitled “Judah Folkman, a pioneer in the study of angiogenesis” outlines the fundamental role played by Judah Folkman in the study of tumour angiogenesis. More than 30 years ago, Folkman found a revolutionary new way to think about cancer. He postulated that in order to survive and grow, tumours require blood vessels and that by cutting off that blood supply a cancer could be starved into remission. What began as a revolutionary approach to cancer has evolved into one of the most exciting areas of scientific inquiry today. Over the years, Folkman and a growing team of researchers have isolated the proteins and unravelled the processes that regulate angiogenesis. Meanwhile, a new generation of angiogenesis research has emerged as well, widening the field into new areas of human disease and deepening it to examine the underlying biological processes responsible for those diseases. VEGF/vascular permeability factor (VPF) is one of the best studied angiogenic cytokines, as evidenced by more than 2,000 published reports dealing with VEGF biology and it appears to play a critical role in the regulation of blood vessel growth as well as having a considerable therapeutic potential. VEGF/VPF induces an extremely heterogeneous response that includes hyperpermeability, extravascular fibrin deposition, oedema, angiogenesis and lymphangiogenesis. The article entitled “The contribution of Harold F. Dvorak to the study of tumor angiogenesis and stroma generation mechanism” outlines the important role of the American scientist Harold Dvorak and his colleagues, who in 1983 were the first to show that tumour cells secreted VPF and that a blocking antibody to VPF could prevent the oedema and fluid accumulation that is characteristic of human cancers. In 1986, Dvorak went on to demonstrate that VPF was secreted by a variety of human tumour cell lines and proposed that VPF was in part responsible for the abnormal vasculature seen in human tumours. As a result, he and other investigators demonstrated that VPF was capable of stimulating endothelial cell growth and angiogenesis. In 1986, Dvorak proposed that by secreting VPF, tumours induce angiogenesis by turning on the wound healing response. He noted that wounds, like tumours, secrete VPF, causing blood vessels to leak plasma fibrinogen which stimulates blood vessel growth and provides a matrix on which they can spread. Unlike wounds, however, that turn off VPF production after healing, tumours did not turn off their VPF production and instead continued to make large amounts of VPF, allowing malignant cells to continue to induce new blood vessels and so to grow and spread. Thus, tumours behave like wounds that fail to heal. This work is again extremely significant for patients worldwide, as Dvorak’s scientific research is leading his colleagues all over the world to examine how to treat a tumour through its blood supply. The article entitled “Napoleone Ferrara and the saga of vascular endothelial growth factor” remarks the contribution of the Italian scientist Napoleone Ferrara and his colleagues at Genentech in the USA, who in 1989 were the first to isolate and clone VEGF. His laboratory has investigated many aspects of VEGF biochemistry and molecular biology. In 1993, Ferrara reported that inhibition of VEGF-induced angiogenesis by specific monoclonal antibodies resulted in dramatic suppression of the growth of a variety of tumours in vivo. These findings provided an important
58
9 Foreword
evidence that inhibition of angiogenesis may suppress tumour growth and blocking VEGF action could have therapeutic value for a variety of malignancies and validate the notion introduced in 1971 by Judah Folkman that inhibition of tumour angiogenesis might be a valid approach to control tumour growth. A further development was the design in a rational fashion in 1997 of a humanized anti-VEGF monoclonal antibody (Avastin), now in clinical trials as a treatment for several solid tumours and also outside of cancer, for example, in the treatment of age-related macular degeneration. The article entitled “Pietro M. Gullino and angiogenesis” outlines the important contributions of the Italian scientist Pietro M. Gullino in the study of tumour microenvironment and its relationship with tumour angiogenesis. His pioneer work introduced several, now widely accepted, concepts in the angiogenesis field. Among these are (1) the relationship between acquisition of angiogenic capacity and neoplastic transformation of a cell population; (2) the concept of high tumour interstitial pressure; and (3) the modifications of tissue composition at the onset of angiogenesis.
Chapter 10
Giulio Bizzozero and the Discovery of Platelets
10.1 Biographical Notes Giulio Bizzozero (Fig. 10.1) was born in Varese in 1846. In 1861, after his highschool classical studies, he enrolled at the University of Pavia as a medical student (Gravela, 1989; Barbiero, 1996). He began to carry out histological and histopathological research under the direction of Paolo Mantegazza who, in 1861, had founded the Laboratory of Experimental Pathology and was the most influent teacher of Bizzozero. In June 1866 at the age of 20, Bizzozero graduated in medicine and received the Mateucci Prize, which was awarded to the student who had achieved the highest grade in all courses. Bizzozero travelled abroad, visiting the laboratories of the histologist Heinrich Frey in Zurich and founder of cellular pathology, Rudolf Virchow in Berlin. In 1867, backing in Pavia, he began his academic career as a deputy professor of general pathology and lecturer of histology. In 1872, Bizzozero at the age of 26 was appointed professor of general pathology at the University of Turin, where he remained active until his death in 1901, aged 55 years, of pneumonia.
Fig. 10.1 A portrait of Dr. Giulio Bizzozero
Published in collaboration with Enrico Crivellato in “Leukemia Research”, 31:1339–1341, 2007
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_10, C Springer Science+Business Media B.V. 2010
59
60
10
Giulio Bizzozero and the Discovery of Platelets
10.2 The Discovery of Platelets Bizzozero for many years studied the haematopoietic function of the bone marrow (Bizzozero, 1869) and the usefulness of blood transfusion in anaemia (Bizzozero and Golgi, 1880). Schulze (1865) published the first accurate description of platelets as a part of a study devoted to the white blood cells. According to his description, blood of healthy individuals contains irregular aggregates consisting of small, uncoloured spherules or granules for which he proposed the term granular masses. These granule aggregates were considered by Hayem (Hayem, 1878), the product of alteration of discoid elements recognized in fresh blood. He reached the conclusion that the blood contains, besides red and white cells, additional storage elements which are rapidly altered, the so-called Hayem haemoblasts. Bizzozero was the first to describe the platelets as a third morphological element of the blood, unrelated to erythrocytes and leucocytes. He published a first short account of its discovery in Italian (Bizzozero, 1881), a second further in French (Bizzozero, 1882a) and the definitive in German (Bizzozero 1882b). He called the granules piastrine in Italian, petit plaques, later plaquettes in French and Blutplattchen in German. In English, they were later named as platelets. Bizzozero started his paper with this sample statement “The existence of a constant blood particles, differing from red and white blood cells, has been suspected by several authors for some time” (Bizzozero 1882b). “It is astonishing – he continued – that none of the previous investigators made use of the observation of circulating blood in living animals.” As wrote Mazzarello in his fundamental contribution “Before his investigation, several researches had observed platelets in the blood, but they regard these particles either as degenerated and disintegrated leukocytes, or as clots of fibrin or a particular kind of microbe” (Mazzarello et al., 2001). Bizzozero described platelets as discoid corpuscles without a nucleus, consisting of a membrane and a matrix in which there were a few dispersed granules. He wrote “If one investigates the contents of such vessels with an immersion objective (regardless of whether they are veins or capillaries) so one achieves the surprising result that actually, in addition to the red and white blood corpuscles, a third sort of morphological element circulates in the blood vessels. In form, they are very thin platelet, disc-shaped, with parallel surfaces or rarely lens-shaped structures, round or oval and with a diameter 2–3 times smaller than the diameter of the red cells” (Bizzozero, 1882b).
10.3 The Role of Platelets in Thrombosis and Blood Coagulation Bizzozero was the first to clearly demonstrate the role of platelets in promoting thrombosis in vivo through blood coagulation in vitro. He exerted light pressure with a fine needle on a point in the wall of an artery in the mesentery of anaesthetized rabbits and guinea pigs and observed that “Blood platelets, swept along
10.3
The Role of Platelets in Thrombosis and Blood Coagulation
61
by the blood stream, are held up at the damaged spot as soon as they arrive at it. At first one sees only 2-4-6 (platelets); very soon the number climbs to hundreds. Usually some white blood cells are held up amongst them. Little by little the volume increases and soon the thrombus fills the lumen of the blood vessels, and impedes the blood stream more and more” (Bizzozero, 1882b). In the opinion of Bizzozero the coagulation of blood, flowing from a wound, is probably due to the presence of aggregates of blood platelets which have formed shortly after the time of incision, on the surface of vascular lesions and the margins of the incision. These aggregates induce coagulation of the blood which passes through them at a later time. Bizzozero wrote that “Whereas under normal conditions the platelets float isolated in the plasma, when subject to an influence that leads to thrombosis, they adhere to one another to form an accumulation. The blood platelets, free in the blood stream and being hurried along are held up by other platelets that they come into contact with as they become stickier than they are under normal condition” (Bizzozero, 1882b). Bizzozero devised a small chamber on a microscope slide in which he observed the platelets being deposited on a thread “. . .after an interval of time, which varies from fractions to more than a minute. . .in addition to a few red and white cells, numerous blood platelets adhere to the threads and coat it with a thick layer. After that fibrous material is deposited on the layer of platelets and as the flow continues it is laid down predominantly in long bundles of threads” (Bizzozero, 1882b). Using this system, Bizzozero observed that, initially, only a few platelets adhered to the thread. Later on, numerous platelets, besides a few red and white corpuscles, stick to the thread, covering it with thick layers. Thereafter, fibrin was deposited on top of the platelet layer as elongated fibrillary bundles. The latter acquired the property of a foreign surface so that they were rapidly covered by increasing layers of platelets and became active as new centre for blood coagulation. Bizzozero concluded his paper by the following sentences “Thus one recognizes that the barely initiated study of the blood platelets already provides a significant contribution for explaining the phenomena of thrombosis and blood coagulation. However, we have not as yet obtained information concerning the physiological role of the blood platelets, since both thrombosis and coagulation take place exclusively under abnormal conditions. It is hardly permitted to assume that elements, represented in the blood in such a constant fashion and great number, as it is the case for blood platelets, are active only under abnormal or pathological conditions. Their physiologic significance, therefore, remains to be investigated, as well as their origin and their possible relationship with other elements of the blood. It is not necessary to emphasize the difficulty of these tasks. . . . In the future, one will have to consider this new blood constituent in studies of pathological events as well. Thus, it is probable that the blood platelets are involved not only in thrombosis and coagulation but also in other vital reactions of the blood and the blood vessels associated with disease states. . . . One may assume that their increased number alters the conditions of blood circulation. Alternatively, it is equally likely that in such conditions the slightest alteration
62
10
Giulio Bizzozero and the Discovery of Platelets
of vessel may lead to widespread thrombosis. Thus, a broad field for new research has been opened.” As pointed out by De Gaetano “Despite the excellency of these and other contribution, the platelets remained for many years the neglected stepchild in the family of blood cells” (De Gaetano, 2001). Only during the 1960s new discoveries in this field were announced and the interest of many experts moved from the interaction of platelets with the process of blood coagulation to the interactions of these cells with the vascular wall and each other.
Chapter 11
A milestone in the Study of the Vascular System: Wilhelm Roux’s Doctoral Thesis on the Bifurcation of Blood Vessels
Jonathan Bard wrote that “If the science of embryology has an hero, it is probably Wilhelm Roux because he, through the force of its thinking, writing an experimentation, changed the direction of embryology from its interest in evolution and teleology to a concern with mechanisms, or in the language of those times, from final to efficient causes” (Bard, 1990). Roux (Fig. 11.1) (1850–1924) inaugurated his program of developmental mechanisms (Entwicklungsmechanik), the physiological approach to embryology. He was one of the first to attempt a causal analysis of early development. With a hot needle, he killed one of the two cells of the frog embryo after the first cleavage and then watched the development of the surviving cell. A typical half embryo was seen to emerge just as if an older embryo has been sliced in two with a razor. Only very few embryos survived as far as the gastrula stage, a finding that he thought lend support to the idea of qualitative cell division. Conversely, Hans Dreisch (1867–1941) discovered that when he separated blastomers of sea-urchin eggs by shaking, they developed into half-sized embryos, some of which reached the larval stage. It seemed after all that each cell retained its totipotency enabling it to develop into any part of the organism as the occasion demanded. In 1878, Roux discussed his doctoral thesis entitled On the bifurcation of blood vessels. A morphological study. As underlined by Kurz “Roux realized that an enormous number of detailed studies would be needed to untangle the molecular and regulatory complexity of the vascular system, and that he did not have the tools to cope with this enterprise” (Kurz et al., 1997). Since the early work of Roux, diameter relations and branching at bifurcations were studied by anatomists, physiologists, mathematicians and theoretical biologists (Thompson, 1917). More recently, Kurz et al. (1998) demonstrated that an optimum value exists for the bifurcation exponent in the avian extraembryonic
Published in “Haematologica”, 87:675–676, © 2002 Ferrata Storti Foundation, Italy. Printed with permission
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_11, C Springer Science+Business Media B.V. 2010
63
64
11
A milestone in the Study of the Vascular System
Fig. 11.1 A portrait of Dr. Wilhelm Roux
circulation. Moreover, they speculated whether this minimum mass condition influenced the evolution of developmental mechanisms such that a minimum of genetic information is needed for realizing a vascular network. The bifurcation of pre-existing blood vessels takes place during the process of arteriogenesis, defined as the development of collateral arteries from pre-existing arteriolar connections by growth, requiring the proliferation of endothelial cells and smooth muscle cells (Buschmann and Schaper, 2000). In fact, it is established that this process is not a consequence of a passive dilatation, but it is characterized by an active proliferation and remodelling. In the case of acute or chronic occlusion of a major artery, collateral arteries can ameliorate the ensuing detrimental effects in many regions of the body, such as hindlimb, heart, brain and kidney. The morphological substrates of arteriogenesis are pre-existing collateral arterioles and the hallmarks of this process are increased levels of shear forces and the invasion of circulating monocytes. Arteriogenesis differs from angiogenesis in several aspects, the most important being the dependence of angiogenesis on hypoxia, while arteriogenesis depends on inflammation. Whereas, angiogenesis can be largely explained by the action of vascular endothelial growth factor (VEGF), arteriogenesis is probably a multifactorial process in which several growth factors co-operate. Deindl et al. (2001) recently demonstrated that neither endogenous nor exogenous VEGF contribute directly to arteriogenesis. They also showed that the continuous infusion of VEGF did not improve collateral formation, yet the administration of monocyte chemotactic protein-1 (MCP-1) significantly improved collateral conductance. The up-regulated expression of MPC-1 by the endothelium attracts monocytes that adhere to and invade arteriolar collaterals, the first visible morphological change during arteriogenesis.
Chapter 12
Stephen Paget and the “Seed and Soil” Theory of Metastatic Dissemination
12.1 Introduction Cancer metastasis represents the major cause of morbidity and death for cancer patients. In fact, whereas the primary tumour is in most cases susceptible of eradication by combined surgical and radiochemical treatments, its metastases, when distributed throughout the body, are most difficult to treat by any therapeutic means, and finally cause patient’s death. It has long been accepted that most malignant tumours show an organ-specific pattern of metastasis. For example, colon carcinomas metastasize usually to liver and lung but rarely to bone, skin or brain and almost never to kidneys, intestine or muscle. In contrast, other tumour entities, such as breast carcinomas, frequently form metastases in most of these organs. This specific formation of secondary tumours at distant sites appears to require the successful completion of a number of steps by metastasizing tumour cells (Chambers et al., 2002). Various explanations have been proposed for the site selectivity of blood–borne metastases, including tumour cell surface characteristics (Turner, 1982; Reading and Hutchins, 1985; Raz and Lotan, 1987), response to organ-derived chemotactic factors (Hujanen and Terranova, 1985), adhesion between tumour cells and the target organ components (Nicolson, 1988a, b) and response to specific host tissue growth factors (Nicolson and Dulski, 1986). The relative importance of pre-existing tumour subpopulations with specific metastatic properties and the organ environment characteristics in determining metastatic homing have been debated (Nicolson, 1988a; Weiss, 1979; Talmadge and Fidler, 1982; Fidler, 1986). An alternative explanation for the different sites of tumour growth involves interactions between the metastatic cells and the organ environment, possibly in terms of specific binding to endothelial cells and responses to local growth factors. Endothelial cells in the vasculature of different organs express different cell surface receptors and growth factors that influence the phenotype of the corresponding
Published in collaboration with Giuseppe Mangialardi and Angelo Vacca in “Clinical and Experimental Medicine”, 6:145–149, 2006
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_12, C Springer Science+Business Media B.V. 2010
65
66
12
Stephen Paget and the “Seed and Soil” Theory of Metastatic Dissemination
metastases. Greene and Harvey (1964) first suggested that the organ distribution patterns of metastatic foci were dependent on the formation of sufficient adhesive bonds between arrested tumour cells and endothelial cells, and they hypothesized that these interactions were similar to lymphocyte/endothelial cells at sites of inflammation. The development of organ-derived microvascular endothelial cell cultures has allowed more specific studies on the preferential homing of tumour cells. Auerbach and co-workers (Auerbach et al., 1987) found that different tumours showed differences in their adhesive propensity and preference for different endothelial cells and in a few cases preferential adhesion was observed to the endothelial cells derived from the organ of origin and the target organ.
12.2 Paget’s Theory Stephen Paget (Fig. 12.1) (1855–1926) was an English surgeon, son of the famed surgeon Sir James Paget. He trained at St. Bartholomew’s Medical School, and then practised surgery in London, where he developed a strong interest in supporting cancer research. In 1908 he founded the Royal Defense Society to provide scientific input into animal welfare debate and to support experimental research for the benefit of cancer patients. In 1889, Paget proposed that the processes of metastasis did not occur by chance but, rather, that certain favoured tumour cells with metastasis activity (the “seed”) had a special affinity for the growth-enhancing milieu within specific organs (the “soil”, i.e. organs providing a growth advantage to the seeds). He concluded that metastases developed only when the seed and soil were compatible. In other words,
Fig. 12.1 A portrait of Dr. Stephen Paget
12.3
Ewing’s Viewpoint
67
Paget suggested that the site of metastasis depended on the affinity of the tumour for the microenvironment (Paget, 1889). Paget analysed autopsy records of 735 women with breast cancer. His analysis documented a non-random pattern of metastasis to visceral organs and bones, suggesting that the process was not due to chance but rather that certain tumour cells had a specific affinity for the milieu of certain organs. Auerbach (1988), in his comments about organ selectivity of metastasis, wrote “Paget is almost apologetic as he contrasts the work of those that study the ‘seed’ to his own work on the ‘soil’: the best work in the pathology of cancer is done by those studying the nature of the ‘seed’. They are like scientific botanists; and he who turns over the records of cases of cancer is only a ploughman, but his observations of the properties of the ‘soil’ may also be useful.” Auerbach then added “Those individuals who study the properties of the host environment should not be ignored. Not only are the observations of the ‘soil’ useful, they provide essential information without which we will not be able to understand the nature of the metastatic process.”
12.3 Ewing’s Viewpoint In 1928, James Ewing challenged Paget’s “seed and soil” theory and hypothesized that metastatic dissemination occurs by purely mechanical factors that are a result of the anatomic structure of the vascular system (Ewing, 1928). Thus, it would be completely accounted for by the vascular connections of the primary tumour: intravasating tumour cell emboli are much more likely to be mechanically trapped in the circulatory network of the first connected organ, which will then sustain the highest burden of metastatic colonization. Other organs receive less tumour cells and develop fewer metastatic colonies. Ewing’s viewpoint prevailed for several decades. His proposal, however, does not explain the observation that some organs, such as brain, bone and adrenals, are served by a very small fraction of the circulatory system, yet they are often involved by metastatic deposits of certain cancers. Moreover, other organs, such as heart, muscle, skin, kidney and spleen, each receiving a considerable supply of blood, are only sporadically colonized by cancers. Sugarbaker (1979) pointed out that common regional metastatic involvements could be attributed to anatomical or mechanical considerations, such as afferent venous circulation or lymphatic drainage to regional lymph nodes, but that metastasis to distant organs by metastatic cell from numerous types of cancers had a different pattern of site specificity. This specificity has been demonstrated by Tarin and co-workers (1984). Patients with incurable abdominal ascitic cancer were treated with peritoneal-venous shunting in order to alleviate the abdominal pain and distension. In this procedure, the abdominal effusion is returned to the circulation via an anastomosis, containing a one-way valve, between the peritoneal cavity and the lungs. Therefore, a large number of tumour cells are directly infused into the circulation. Despite this huge
68
12
Stephen Paget and the “Seed and Soil” Theory of Metastatic Dissemination
tumour load, many patients did not develop evident metastases and among those who did, the distribution of secondary deposits was unexpected, in that metastases did not form in the organ containing the first capillary bed encountered, i.e. the lungs.
12.4 The Contribution of Experimental Pathology to the Study of the Process of Metastasis Zeidman and co-workers (1950) reported that the number of metastases was directly proportional to the number of tumour cells injected intravenously, but that most injected tumour cells still failed to form tumours. Coman and co-workers (1951) reported that the direct intravascular injection of tumour cells into animals produced metastases in some, but not all, visceral organs. The authors found that in those organs, circulating tumour cells were lodged in the capillaries, whereas in organs that were rare sites of metastasis, circulating cells lodged in arterioles. This observation indicated that the distribution of metastases was largely dependent on mechanical factors, that is, on the arrest of emboli in capillaries of secondary organs. Lucke and co-workers (1952) compared carcinoma metastases in the livers and lungs of rabbits and found that liver metastases were larger and more numerous. Human cancer patients also develop a larger number of liver, rather than lung, metastases, so both mechanical and local “soil” factors are likely to determine whether or not a metastasis will develop after the arrest of tumour emboli. Zeidman and Buss (1952) used cinephotomicrography to observe the incidence of emboli arrest in mesenteric capillaries of rabbits. They found that some tumour cells become distorted and passed through the marrow capillary tube, whereas others appeared more rigid and were trapped. The incidence of arrest varied with the type of tumour studied. This work established the morphological foundation for previous indirect demonstrations that some tumour cell emboli could pass immediately through the vascular bed of organs.
12.5 Metastasis Can Result from Survival of Only a Few Tumour Cells As a whole, metastasis favours the survival and growth of a few subpopulations of cells that pre-exist within the parent neoplasm. So, metastases can have a clonal origin and different metastases can originate from the proliferation of different single cells (Fidler, 1990). Fidler (1970) showed that within 24 h after entry into the circulation, less than 0.1% of tumour cells are still viable and that less than 0.01% of these cells, when introduced into the circulation, survive to produce metastases. Therefore, only a few cells in a primary tumour can give rise to metastasis. Although it has been
12.6
Experimental Evidence of Metastatic Heterogeneity of Tumours
69
determined that less than 0.01% of tumour cells that enter the circulation have the potential to form secondary tumours (Fidler, 1970; Butler and Gullino, 1975; Mayhew and Glaves, 1984), still hundreds of viable tumour cells each day have the possibility to lodge into distant organs. Cells with different metastatic properties have been isolated from the same parent tumour, indicating that not all the cells in a primary tumour have the same potential to disseminate. Tumour cells were implanted subcutaneously, intramuscularly, directly into tissues or injected intravenously into mice. Tumours were then harvested and the recovered cells expanded in culture. The behaviour of the expanded cells was compared to that of the cells of the parent tumour to determine whether the selection process enhanced metastatic capacity. This procedure was originally used to isolate the B16-F10 line from B16 melanoma (Fidler, 1973). In a second approach, cells were selected for the development of a phenotype that was associated with the metastatic sequence, and then they were tested in animal models to determine whether concomitant metastatic potential was increased or decreased. This method has been used by Nicolson (1988a) and Poste (1982) to determine whether properties such as adhesive characteristics, invasive capacity, lectin resistance and resistance to natural killer cells were required for metastasis.
12.6 Experimental Evidence of Metastatic Heterogeneity of Tumours Experimental data to support Paget’s “seed and soil” hypothesis were derived from studies provided by Fidler and Kriple (1977) using mouse B16 melanoma cells. They showed that different tumour cell clones, each derived from individual cells isolated from a parent tumour, vary markedly in their ability to form pulmonary nodules following intravenous inoculation of B16 melanoma cells into syngeneic C57BL/6 mice. Tumour growth developed in the lungs and in fragments of pulmonary or ovarian tissue that were implanted intramuscularly. By contrast, metastatic lesions did not develop in implanted renal tissue or at the site of surgical trauma. A detailed analysis of experimental metastasis in syngeneic mice indicated that mechanical arrest of tumour cells in the capillary bed of distant organs could indeed occur, but that subsequent proliferation and growth into secondary lesions was influenced by specific organ cells (Hart and Fidler, 1980). Controlled subcloning procedures showed that the observed diversity was not a consequence of the cloning procedures. This indicates that the sites of metastasis are determined not only by the characteristics of the neoplastic cells but also by the microenvironment of the host tissue (Hart and Fidler, 1980). To exclude the possibility that the metastatic heterogeneity of B16 melanoma cells might have been introduced as a result of the lengthy cultivation, studies on the biological and metastatic heterogeneity of spontaneous tumours were carried out. Melanomas were induced in mice by chronic exposure to ultraviolet B irradiation and the tumour-promoting agent Croton oil, and tumour metastases were found to
70
12
Stephen Paget and the “Seed and Soil” Theory of Metastatic Dissemination
differ greatly from each other and from the parent tumour. In addition to differences in the number of metastases that developed from each tumour, there was also significant variability in the size and pigmentation of the metastases. Metastases to the lymph nodes, brain, heart, liver and skin were found in addition to lung metastases. Those growing in the brain were uniformly pigmented, whereas those growing in other organs generally were not (Fidler et al., 1981). Other observations relating the “seed and soil” hypothesis were made by Pilgrim (1969), using a transplantable reticulum cell sarcoma, which selectively metastasized to the mouse spleen. When equal numbers of cells were injected into the kidney and the spleen, growth in the spleen was always considerably greater than in the kidney. However, in no case was the mitotic index higher in the spleen than in the kidney. Pilgrim therefore considered that cell loss in the kidney was greater than in the spleen; however, his emphasis was on cell migration rather than cell death within the target organ. Regardless of mechanism, compared with the spleen, the kidney was therefore unfavourable “soil” for this tumour.
12.7 Studies of Experimental Brain Metastasis Schackert and Fidler (1988a) described the development of a mouse model to study cerebral metastasis after injection of syngeneic tumour cells into the internal carotid artery of mice, which stimulates the haematogenous spread of tumour emboli in the brain. This technique can be used to examine the last steps of the metastatic process, such as release of tumour cells into the circulation, arrest of tumour cells in capillaries, penetration and extravasation of the tumour cells into the brain through the blood–brain barrier and continuous growth of the cells in the tissue. This procedure was used to study metastases of two different murine melanomas. The two melanomas differed in their brain metastatic patterns: the K1735 melanoma produced lesions only in the brain parenchyma, whereas the B16 melanoma grew only in the meninges and ventricles (Schackert and Fidler, 1988a). Similarly, different human melanomas or carcinomas injected into the internal carotid artery of nude mice produce unique patterns of brain metastasis. These results demonstrate specificity for metastatic growth in different regions within a single organ (Schackert and Fidler, 1988b).
12.8 Concluding Remarks Paget postulated that microenvironment provides a fertile “soil” for cancer cells endowed with a capacity to grow under specific conditions provided by the “soil”. A current definition of the “seed and soil” hypothesis consists of three principles. First, neoplasms are biologically heterogeneous and contain subpopulations of cells with different angiogenic, invasive and metastatic properties. Second, the process of metastasis is selective for cells that succeed in invasion, embolization, survival in the
12.8
Concluding Remarks
71
circulation, arrest in a distant capillary bed, and extravasation into and multiplication within the organ parenchyma. Third, the outcome of metastasis depends on multiple interactions of metastatic cells with homeostatic mechanisms, which the tumour cells can escape. In 1989, in his introductory remarks to the symposium commemorating the anniversary of Paget’s “seed and soil” hypothesis, George Poste pointed out that “There are few scientists, historically or contemporary, whose work will stand 100 years of scrutiny and not succumb to the depressing trend of modern publications – to ignore papers published more than 5 years ago.”
Chapter 13
The Contribution of Roberto Montesano to the Study of Interactions Between Epithelial Sheets and the Surrounding Extracellular Matrix
13.1 Introduction The interactions between epithelial cells and the surrounding extracellular matrix (ECM) are a central issue in research on morphogenesis. Since the studies by Grobstein (1954), it has become increasingly clear that the ECM not only serves as an inert physical scaffolding for the developing tissues but also provides environmental signals to the cells involved, playing a guiding role in morphogenetic processes. Moreover, the ECM is the natural substratum through which various cell types migrate during embryogenesis. Different epithelia, such as those of the thyroid follicles, kidney tubules and the branching ducts in the lung and exocrine glands, acquire diverse forms correlated to their specific functions. The development of most glandular organs begins with the invagination of an existing epithelial sheet into the underlying mesenchyma. The primary bud thus generated subsequently undergoes a series of morphogenetic events that culminate in the formation of branching tubes (excretory ducts) and hollow spheres (follicles, alveoli, acini). A number of studies have led to the identification of mesenchyme-derived diffusible messengers that promote the elongation and branching of epithelial tubes. In the last 30 years, working at the University Medical Center of Geneva, Switzerland, Roberto Montesano (Fig. 13.1) has extensively investigated the mechanisms underlying two morphogenetic processes: the formation of new capillary blood vessels from pre-existing ones (angiogenesis) and the generation of branching epithelial tubules (tubulogenesis), which are crucial events in the development of most parenchymal organs.
Published in “International Journal of Developmental Biology”, 54:1–6, 2010
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_13, C Springer Science+Business Media B.V. 2010
73
74
13 The Contribution of Roberto Montesano
Fig. 13.1 A portrait of Dr. Roberto Montesano (courtesy of Dr. Montesano)
13.2 Mechanisms Underlying Angiogenesis The term angiogenesis, applied to the formation of capillaries from pre-existing vessels, i.e. capillaries and post-capillary venules, is based on endothelial sprouting or intussusceptive (non-sprouting) microvascular growth (Risau, 1997; Burri and Tarek, 1990). During angiogenesis, microvascular endothelial cells focally degrade their investing basement membrane and subsequently migrate into the interstitial matrix of the surrounding connective tissue (Ausprunk and Folkman, 1977). The sprout elongates by further migration and by endothelial cell proliferation proximal to the migrating front, and a lumen is gradually formed proximally to the region of proliferation. Contiguous tubular sprouts anastomose to form functional capillary loops, and vessel maturation is accomplished by means of reconstitution of the basement membrane (Ausprunk and Folkman, 1977). Under physiological conditions, angiogenesis is dependent on the balance of positive and negative angiogenic modulators within the vascular microenvironment (Hanahan and Folkman, 1996) and requires the functional activities of a number of molecules, including angiogenic factors, extracellular matrix proteins, adhesion receptors and proteolytic enzymes. As a consequence, angiogenic endothelial cells have a distinct gene expression pattern that is characterized by a switch of the cell proteolytic balance towards an invasive phenotype, as well as by the expression of specific adhesion molecules (Pepper et al., 1996). The process of angiogenesis remained largely inaccessible to experimental analysis until the development of bioassays for angiogenesis during the 1970s. These included the long-term culture of vascular endothelial cells, the development of the
13.3
Collagen Matrix Promotes the Organization of Endothelial Cells
75
chick embryo chorioallantoic membrane and in vivo rabbit/murine assays. These assays have been used to describe the morphological events occurring in angiogenesis, to identify stimulators and inhibitors of angiogenesis and to quantitate the neovascular response to test compounds.
13.3 Collagen Matrix Promotes the Organization of Endothelial Cells into Capillary-Like Tubules and the Induction of the Invasive Phenotype In 1983, Montesano and co-workers demonstrated that when a monolayer of microvascular endothelial cells on the surface of a collagen gel is covered with a second layer of collagen, it reorganizes within a few days into a network of branching and anastomosing tubules, without invading the underlying matrix, demonstrating that a three-dimensional interaction with collagen fibrils plays an important role in driving capillary morphogenesis (Montesano et al., 1983). Next, Montesano and co-workers demonstrated that when confluent monolayers of microvascular endothelial cells on collagen gels were treated with phorbol myristate acetate (PMA), a tumour promoter that markedly stimulated the production of collagenase and plasminogen activators (PAs) whereas control endothelial cells were confined to the surface of the gels, PMA-treated endothelial cells invaded the underlying collagen matrix, where they formed capillary-like tubular structures (Montesano and Orci, 1985). This experimental system has become a widely used in vitro assay of angiogenesis (reviewed in Vailhé et al., 2001), and the underlying mechanisms have been investigated in detail by the group of George Davis. More recent studies have revealed that genes regulating the matrix–integrin–cytoskeletal signalling axis are differentially expressed during endothelial cell morphogenesis. It has been demonstrated that endothelial cell interactions with extracellular matrices establish signalling cascade downstream of integrin ligation leading to activation of the Rho GTPase, CDC42 and Rac1, which are required for lumen formation (reviewed in Davis et al., 2007). Current research analyses how specific molecules integrate signalling information to induce endothelial cell lumen formation, pericyte recruitment and stabilization processes to control vascular morphogenesis. Montesano and co-workers demonstrated that also the basic fibroblast growth factor (bFGF) induced the endothelial cells to invade the underlying collagen matrix and to form capillary-like tubules and stimulated the endothelial cells to produce PAs (Montesano et al., 1986). These data, together with other evidence gained at the Montesano lab demonstrating an increased uPA and urokinase plasminogen activator receptor (uPAR) expression in endothelial cells migrating from the edge of an experimental wound in vitro (Pepper et al., 1987, 1993), supported the role of the PA/plasmin system in angiogenesis. In 1992, Montesano and co-workers demonstrated that the vascular endothelial growth factor (VEGF) induces microvascular endothelial cells grown on collagen gels to invade the underlying matrix, where they form capillary-like tubules (Pepper
76
13 The Contribution of Roberto Montesano
Fig. 13.2 A scheme of three-dimensional cell culture system showing that VEGF induces microvascular endothelial cells grown on collagen gels to invade the underlying matrix (courtesy of Dr. Montesano)
et al., 1992a). Moreover, when added simultaneously, VEGF and bFGF induced an in vitro angiogenic response which was greater than the sum of the two and which occurred with greater rapidity than the response to either cytokine alone (Pepper et al., 1992a) (Fig. 13.2). These results demonstrate that, by acting in concert, these two cytokines have a potent synergistic effect on the induction of angiogenesis in vitro. In 1993, Montesano and co-workers demonstrated a biphasic action of transforming growth factor beta-1 (TGF-β1) on in vitro angiogenesis: at low concentrations it decreased bFGF- or VEGF-induced invasion, whereas at high concentrations it increased invasion (Pepper et al., 1993). At the TGF-β1 concentration which potentiates bFGF- or VEGF-induced invasion, an optimal balance between proteases and protease inhibitors may be achieved at the cell surface, which allows focal pericellular matrix degradation, while at the same time protecting the matrix against excessive degradation (Pepper et al., 1994).
13.4 Unbalanced Proteolysis Results in Aberrant Vascular Morphogenesis In 1990, Montesano and co-workers developed an in vitro model of endothelioma formation by embedding middle-T oncogene-expressing endothelial cells into three-dimensional fibrin gels. In contrast to normal endothelial cells, which formed a network of capillary-like tubules, middle-T-expressing endothelial cells formed large cyst-like structures which resembled the endotheliomas observed in vivo (Montesano et al., 1990). When studying the proteolytic properties of middle-T-expressing endothelial cells, Montesano and co-workers found that these
13.6
Fibroblast-Derived Soluble Factors Induce Morphogenesis
77
cells displayed increased PA activity when compared to non-middle-T-oncogeneexpressing endothelial cells, and this could be accounted for by an increase in uPA and a decrease in plasminogen activator inhibitor-1 (PAI-1) activity (Montesano et al., 1990). Finally, they found that when serine protease inhibitors were added to the cultures, the middle-T-expressing endothelial cells, instead of forming cysts, now formed branching capillary-like tubules (Montesano et al., 1990). Overall, these data demonstrated that an excessive proteolytic activity is not compatible with normal capillary morphogenesis, but that by reducing this activity through the addition of protease inhibitors, normal morphogenetic properties can be restored.
13.5 Mechanisms Underlying Tubulogenesis The formation of branched tubes from initially unbranched epithelial buds is a fundamental morphogenetic process in the development of many organs, including the pancreas, liver, mammary gland, lung and kidney (Affolter et al., 2003; Lubarsky and Krasnow, 2003). In vivo, this complex process relies on the interaction of different cell types and various environmental factors, resulting in tubular structures that contain multiple cell types with different functions. Tubules can arise through two main mechanisms: the invagination of cells from an epithelial sheet, as occurs in the formation of the neural tube, and through the organization of initially unpolarized cells into cord-like structures that invade the surrounding mesenchyme, forming branched, hollow tubules lined by polarized cells (Hogan and Kolodziej, 2002). Using tissue dissociation techniques, it has been demonstrated that the embryonic epithelia fail to undergo branching morphogenesis if separated from the adjacent mesenchyme and that morphogenesis resumes when the components are recombined in vitro. Elucidation of the mechanisms responsible for epithelial tubulogenesis is a complex task in view of the myriad, complex cell interactions occurring in vivo, involving the concerted action of a number of external positive and inhibitory stimuli including various growth factors and ECM components.
13.6 Fibroblast-Derived Soluble Factors Induce Morphogenesis of Branching Tubules by Kidney Epithelial Cells Montesano and co-workers investigated whether the morphogenetic properties of the Madin–Darby canine kidney (MDCK) cells might be influenced by diffusible factors released by neighbouring mesenchymal or stromal cells. MDCK cells have been shown to retain several anatomical and functional properties of the kidney distal tubule or collecting duct epithelium and to generate tubule-like structures in vivo. When grown within three-dimensional collagen gels, MDCK cells form spherical cysts (McAteer et al., 1987). Montesano and co-workers demonstrated that (1) MDCK cells suspended within a collagen gel contiguous to a fibroblast-populated gel layer form branching tubules
78
13 The Contribution of Roberto Montesano
instead of the spherical cysts that develop in the absence of fibroblasts; (2) MDCK cells grown as a monolayer on a cell-free collagen gel cast layer invade the underlying collagen matrix, within which they form a network of branching tubules; (3) fibroblast-conditioned medium mimics the effect of co-culture by eliciting tubule formation by MDCK cells (Montesano et al., 1991a).
13.7 Hepatocyte Growth Factor Is a Paracrine Mediator of Morphogenetic Epithelial–Mesenchymal Interactions in MDCK Cells With the aim of identifying the factor which mediates the tubulogenic effect of fibroblast-conditioned media, Montesano and co-workers demonstrated that (1) MDCK cells grown in collagen gels in the presence of the hepatocyte growth factor (HGF) formed linear or branching tubular structures; (2) MDCK cells grown in the presence of fibroblast-conditioned medium that had been preincubated with specific anti-HGF antibodies exclusively formed spherical cysts similar to those observed in the absence of conditioned medium; (3) anti-HGF antibodies suppressed tubulogenesis in co-cultures of MDCK cells and fibroblasts (Montesano et al., 1991b). Overall, these data demonstrated that the fibroblast-derived factor that induces tubule formation by MDCK cells is HGF. HGF is a pleiotropic factor, capable of inducing cell motility, proliferation, anchorage-independent growth and morphogenesis. It was identified as the fibroblast growth factor that stimulates epithelial cells derived from a variety of different organs to form tubule-like extensions when seeded in three-dimensional matrices (Weidner et al., 1993; Brinkmann et al., 1995). The epithelial morphogenesis assay developed by Montesano has been utilized by many investigators to dissect the mechanisms of HGF-induced tubulogenesis. These tubes can form elaborate networks in the lung, kidney, reproductive organs and vascular tree, as well as the many glands branching from the digestive tract, such as the liver, pancreas and salivary glands. In particular, a thorough analysis of this phenomenon by the group of Keith Mostov has provided a detailed documentation of the sequential steps involved and has demonstrated the requirement for partial and transient epithelial–mesenchymal transition (EMT) in the tubulogenic process (reviewed in O’Brien et al., 2002; Zegers et al., 2003). These authors proposed that growth factors such as HGF promote tubular development by inducing a transient, partial EMT that alters the arrangement of cells at the onset of the three-surface pursuit.
13.8 The Role of the Transcription Factor Snail in Modulation of Permeability and Tight Junction Proteins in MDCK Cells Montesano and co-workers investigated the role of the transcription factor Snail on epithelial properties of MDCK cells (Carrozzino et al., 2005). To this purpose, they expressed mouse Snail cDNA in MDCK cells with a tetracycline-inducible
13.10
Paracrine Epithelial–Mesenchymal Interactions Induce Tubulogenesis
79
expression system with the aim to identify novel downstream targets of Snail. They demonstrated that the inducible expression of Snail does not result in overt epithelial–mesenchymal transition, but selectively reduces the expression of claudin-3, claudin-4 and claudin-7 and parallely increases paracellular ionic conductance without affecting tight junction permeability to uncharged solutes (Carrozzino et al., 2005). Overall, these data suggest that in addition to its well-established role in epithelial–mesenchymal transition during embryogenesis and tumour progression, Snail may act as a regulator of epithelial permeability.
13.9 Epithelial Tubulogenesis Is Dependent on Extracellular Plasmin-Dependent Proteolysis Montesano and co-workers observed that when MDCK cells were grown in fibrin gels instead of collagen gels, HGF-induced tubule formation was prevented by the addition of serine proteinase inhibitors (Montesano et al., 1991a). Moreover, Montesano and co-workers found that (1) conditioned medium from MCR-5 fibroblasts increased uPA activity and mRNA by about fivefold and this effect was completely inhibited by preincubation of conditioned medium with anti-HGF antibodies; (2) exogenously added recombinant HGF induced a comparable increase in uPA activity and mRNA in MDCK cells; (3) both MRC-5-conditioned medium and HGF induced a more than 30-fold increase in uPAR mRNA in MDCK cells (Pepper et al., 1992b). Overall, these data suggest that epithelial tubulogenesis is dependent on extracellular plasmin-dependent proteolysis, which results from a concomitant increase in uPA and uPAR expression.
13.10 Paracrine Epithelial–Mesenchymal Interactions Induce Tubulogenesis in Other Types of Epithelial Cells Montesano investigated the mechanisms underlying tubulogenesis and lumen formation in the mammary gland, which is one of the few organs that undergo cycles of growth, morphogenesis, differentiation, functional activity and involution in post-natal life (Daniel and Silberstein, 1987). First, Montesano and co-workers investigated whether diffusible factors released by fibroblasts could promote the formation of duct-like structures by mammary gland epithelial cells embedded in collagen gels. They used a mammary glandderived cell line (TAC-2) and found that fibroblast-conditioned medium stimulated the development of an extensive system of highly arbourized duct-like structures (Soriano et al., 1995). Moreover, the effect of fibroblast-conditioned medium was completely abrogated by antibodies to HGF, whereas the addition of exogenous HGF to the cultures mimicked the tubulogenic activity of conditioned medium. Finally, the effect of HGF was markedly boosted by the simultaneous addition of hydrocortisone, which also enhanced lumen formation (Soriano et al., 1995).
80
13 The Contribution of Roberto Montesano
13.11 HGF and TGF-β1 Play a Role in Mammary Gland Morphogenesis In Vivo Montesano and co-workers studied the expression of HGF and its receptor, c-met, in the rat mammary gland during pregnancy, lactation and involution (Pepper et al., 1995). They demonstrated that the levels of both HGF and c-met mRNA progressively reduced during pregnancy, were undetectable during lactation, but increased during the involution phase up to prepregnancy levels. Moreover, after 3 days of lactation both HGF and c-met transcripts were once again reduced to undetectable levels in the mothers. Finally, because the levels of HGF/c-met were inversely correlated with the levels of prolactin, Montesano and co-workers investigated the effects of prolactin on c-met expression in TAC-2 cells. They demonstrated that prolactin significantly reduced the levels of c-met mRNA in TAC-2 cells, thus providing a possible mechanism for c-met downregulation in the rat mammary gland during lactation. Montesano and co-workers demonstrated that low concentrations of TGF-β1 promote the elongation and branching of TAC-2 cells, whereas high levels have inhibitory effects (Soriano et al., 1996). They employed an in vitro system in which J3B1A mammary epithelial cells grown in collagen gels in chemically defined medium form spherical cysts. Montesano and co-workers demonstrated that the addition of acidified fetal calf serum (FCS) to the defined medium induced the formation of branching tubes (Montesano et al., 2007). Using a pharmacological inhibitor of TGF-β receptor signalling and a neutralizing antibody to TGF-β1, Montesano and co-workers identified the active component in acidified FCS as TGF-β1. Moreover, the effect of acidified FCS was replicated by the addition of exogenous TGF-β1. Overall, these findings demonstrate that, at low concentrations, TGF-β1 can activate a morphogenetic programme resulting in the formation of epithelial tubes. To elucidate the mechanisms responsible for TGF-β1-induced tubulogenesis, Montesano and co-workers assessed the effect of TGF-β1 on the production of the matrix metalloproteinases (MMP) involved in collagen turnover and demonstrated a dose-dependent increase in matrix metalloproteinase-9 (MMP-9) following TGF-β1 treatment. Tube formation was suppressed by a synthetic broadspectrum MMP inhibitor, by a recombinant tissue inhibitor of MMP-2 and by a selective inhibitor of MMP-9, indicating that this morphogenetic process requires the activity of MMP-9.
13.12 Retinoids Induce Lumen Formation, Whereas Tumour Necrosis Factor Alpha and Bone Morphogenetic Protein-4 Confer an Invasive and Transformed Phenotype to Cultured Mammary Epithelial Cells Montesano and Soulié (2002) demonstrated that retinoic acid induces the formation of lumen-containing colonies (cysts) in cultured mammary epithelial cells.
13.13
Concluding Remarks
81
Moreover, using gelatin zymography, they observed a dose-dependent increase in the latent and active forms of MMP-9 following retinoic acid treatment. Finally, lumen formation was abrogated by the addition of the synthetic MMP inhibitor BB94, indicating that this morphogenetic process likely requires MMP activity. Montesano and co-workers reported that tumour necrosis factor alpha (TNF-α) causes multicellular colonies of mammary epithelial cells to disaggregate and induces cells grown on top of a collagen gel to invade the underlying matrix (Montesano et al., 2005). Moreover, they showed that TNF-α confers to mammary epithelial cells several additional properties that are characteristic of malignantly transformed cells, including proliferation in the absence of exogenously added growth factors, anchorage-independent growth and the loss of contact-mediated inhibition of proliferation (Montesano et al., 2005). These data provide a mechanistic basis for the reported ability of TNF-α to promote tumour progression and cancer cell dissemination (Szlosarek and Balkwill, 2003). More recently, Montesano reported that bone morphogenetic protein-4 disrupts cyst organization in a concentration-dependent manner, causing lumen obliteration, the extension of invading cell cords and three-dimensional cell scattering (Montesano, 2007). Montesano concluded that these bone morphogenetic protein4-induced biological responses may play a role in the progression of breast cancer.
13.13 Concluding Remarks Epithelial–mesenchymal interactions are inductive interactions occurring among cells or tissues of the same or a different embryonic origin. In specific developing settings, epithelial cells can transiently lose cell–cell adhesion and exhibit a migratory behaviour (Hayed, 1995). The process by which polarized epithelial cells are converted into motile cells is called the epithelial–mesenchymal transition (Eguchi and Kodama, 1993). Tubulogenesis is a complex cellular response that requires the interplay of growth factor signalling induced by adhesion to the ECM and neighbouring cells. The ECM does not only serve as an inert physical scaffolding for the developing tissues, and the notion that ECM components mediate, at least in part, the inducing effect of mesenchyme on epithelial morphogenesis is now supported by a large body of experimental evidence. Dr. Montesano has contributed to clarify some cellular and molecular mechanisms of angiogenesis and tubulogenesis using an original three-dimensional cell culture system that replicates key events of angiogenesis and tubulogenesis, thereby facilitating molecular analysis (Fig. 13.3). A major advance of this technique over conventional monolayer cultures is that cells can be embedded within a lattice of reconstituted collagen fibrils, and it mimics the three-dimensional organization of connective tissue matrices.
82
13 The Contribution of Roberto Montesano
Fig. 13.3 Principal factors studied by Roberto Montesano and collaborators. Schematic drawing of molecules involved in angiogenesis and tubulogenesis of Madin–Darby canine kidney (MDCK) cells and mammary epithelial cells by using an original three-dimensional cell culture system. (PMA, phorbol myristate acetate; uPA, urokinase plasminogen activator; bFGF, basic fibroblast growth factor; VEGF, vascular endothelial growth factor; TGF-β1, transforming growth factor beta 1; HGF, hepatocyte growth factor; RA, retinoic acid)
The results of these studies support the notion that cell interactions with the surrounding ECM are crucial determinants of cell responses to growth factors and that epithelial tissue morphogenesis is governed by the interplay of two different classes of signalling molecules, i.e. paracrine-acting growth factors and insoluble ECM components.
Chapter 14
The Seminal Work of Werner Risau in the Study of the Development of the Vascular System
14.1 Introduction Werner Risau was born on 18 December 1953. He studied chemistry and biochemistry at the University of Münster and University of Tübingen, respectively, before he went in 1984 to the laboratory of Judah Folkman at the Children’s Hospital of the Harvard Medical School, Boston. In 1999, Folkman wrote that “He had written to me that he wished to learn the arcane world of endothelial cells and the growth factors that regulate them. I answered that there was only one such factor, bFGF, and the existing supply was only a few micrograms, but that he was welcome to come nevertheless. He did. He quickly became an expert and then a pioneer on the role of this protein in neovascularization, and he also became intrigued with studying the development of the vascular system in the chick embryo (. . .). After he completed his fellowship and had left our lab, we watched with great pride his meteoric rise as a distinguished scientist known throughout the world. His seminal studies established the molecular mechanisms of VEGF and its receptors, pioneered the identification of cell lineages in the vascular system, and laid the ground work for a molecular distinction between vasculogenesis and angiogenesis.” Risau worked initially as a junior group leader at the Max Planck Institute for Experimental Biology in Tübingen and thereafter at the Max Planck Institute for Psychiatry, Department of Neurochemistry, at Martinsried close to Munich (Fig. 14.1). In 1993 he was appointed to the position of a director at the Max Planck Institute for Physiology and Clinical Research in Bad Nauheim, until his death at the age of 44 on 13 December 1998. Risau’s work had a decisive impact on defining the overall nature of neovascularization process during embryonic development. He propagated the concept that the same factors, which are essential for the formation of blood vessels during embryonic development, also influence pathological angiogenesis during tumour growth.
In press in “International Journal of Developmental Biology”
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_14, C Springer Science+Business Media B.V. 2010
83
84
14
The Seminal Work of Werner Risau
Fig. 14.1 A portrait of Dr. Werner Risau
14.2 Changes in Extracellular Matrix During Embryonic Vasculogenesis and Angiogenesis The extracellular matrix of the developing blood vessels modifies its composition in terms of expression of fibronectin, laminin, collagen type IV and distribution of specific glycosaminoglycans (Ausprunk, 1986). Risau demonstrated that blood islands in the yolk sac produced high levels of fibronectin but not laminin, whereas all intraembryonic blood vessels are surrounded by a fibronectin-rich extracellular matrix early in the development, which is subsequently remodelled into a basement membrane-like matrix (Risau and Lemmon, 1988). Capillary sprouts invading the neuroectoderm migrated in a fibronectinrich matrix devoid of laminin. Ultrastructural immunolocalization demonstrated the presence of fibronectin exclusively on the abluminal site of endothelial cells. Around larger vessels (e.g. dorsal aorta) several layers of fibronectin-expressing cells can be observed early on, but are devoid of laminin. Overall, Risau and Lemmon proposed that laminin is an early marker for vascular maturation.
14.3 Characterization of Flk-1 Flk-1 is the first endothelial receptor tyrosine kinase known to be expressed on endothelial cell precursors and plays a central role in regulating embryonic vascular development and tumour angiogenesis. Mice deficient in flk-1 died in utero between days 8.5 and 9.5 postcoitum as a result of an early defect in the development of haematopoietic and endothelial cells and a complete lack of vasculature (Shalaby et al., 1995). Risau and co-workers (Kabrun et al., 1997) investigated the expression pattern of flk-1 in mouse embryonic and foetal haematopoietic tissues as well as on haematopoietic tissues derived from these tissues and demonstrated that flk-1
14.4
Embryonic Brain Angiogenesis
85
expression provides a marker for the earliest detectable haematopoietic precursors. RNA analysis indicated that flk-1 was expressed in the yolk sac at days 10 and 12 of gestation, in the liver throughout foetal life and embryoid bodies. Flk-1 was also detected in erythroid and macrophage colonies generated from precursors of yolk sac, foetal liver, adult marrow and embryo bodies’ origin. Risau and co-workers (Rönicke et al., 1996) isolated genomic clones that encompass the promoter region of the murine flk-1 gene and performed a functional analysis of the flk-1 promoter in vivo. They demonstrated that flk-1 5 -flanking sequences confer endothelium-specific expression in transfected endothelial cells. In a paper published in 1999, after Risau’s death, Kappel et al. (1999) characterized regulatory sequences in transgenic mice. Despite their activity in cultured endothelial cells, 5 -flanking sequences alone could not target expression of a LacZ reporter gene to the endothelium of murine embryos. However, in combination with sequences form the first intron of the flk-1 gene, the flk-1 promoter could specifically drive reporter gene expression in endothelial cells. Risau and co-workers (Kappel et al., 2000) identified by mutational analysis binding sites for SCL/Tal-1, GATA and Ets transcription factors as critical elements for the endothelium-specific flk-1 gene expression in transgenic mice, providing the first evidence that SCL/Tal-1, GATA and Ets transcription factors act up-stream of flk-1 in a combinatorial fashion to determine embryonic blood vessel formation and are key regulators not only of hematopoiesis but also of vascular development.
14.4 Embryonic Brain Angiogenesis Risau focused his interest on the vascularization of the brain, where a primary plexus surrounds the neural tube. From this plexus capillary sprouts invade the neuroectoderm. The endothelial cells migrate deep into the neuroectoderm and form capillary branches into the subependymal layer. This process is regulated by the developing organ itself and seems to be dependent on the growth and differentiation of cells present in the brain. Risau characterized endothelial cell growth and angiogenic factors from embryonic chick brain extracts, two of which were identified as acidic and basic fibroblast growth factor (aFGF and bFGF) (Risau, 1986; Risau et al., 1988). The mitogenic activity in chick brain was found to increase several hundred fold during embryonic chick brain development and reached a plateau at around day 14. The mRNA levels of the chick aFGF gene showed a similar pattern during brain development (Schnürch and Risau, 1991). However, growth factor activity and aFGF and bFGF gene transcription are not down-regulated concomitantly with cessation of endothelial cell growth in post-natal brain, and these factors are most abundant in the adult brain when angiogenesis has ceased (Schnürch and Risau, 1991). This suggests that either the activity of these factors is inhibited in the later stages of development or that these factors are not involved in embryonic brain angiogenesis. Furthermore, both aFGF and bFGF do not possess a hydrophobic signal sequence needed for efficient secretion from their producer cells, and their mode of secretion is unknown.
86
14
The Seminal Work of Werner Risau
14.5 The Role of Vascular Endothelial Growth Factor in Brain Angiogenesis Over the years, five VEGF-related genes have been identified (VEGF-A, VEGF-B, VEGF-C, VEGF-D and VEGF-E). There are five characterized VEGF-A isoforms of 121, 145, 165, 189 and 206 amino acids in mammals, generated by alternative splicing of the mRNA from a single gene comprising eight exons. They display differential interactions with related receptor tyrosine kinases VEGFR-1/Flt-1, VEGFR-2/Flk-1, VEGFR-3/Flt-4 and neuropilin-1 and -2. As a result of the receptor activation and subsequent signal transduction, VEGF target cells may proliferate, migrate or alter gene expression, e.g. of matrix metalloproteinases or cytokines (Ribatti, 2005, 2008). Risau cloned the gene-encoding mouse VEGF and found that the spatial and the temporal expression patterns of VEGF mRNA correlated with angiogenesis during embryonic development in the mouse brain (Breier et al., 1992). By in situ hybridization, Risau and co-workers showed that in day 17 mouse embryos VEGF mRNA was expressed in the ventricular layer of the developing neuroectoderm. A concentration gradient of the diffusible VEGF isoforms declining from the ventricular layer towards the perineural vascular plexus may cause the radial ingrowth of capillaries from the vascular plexus towards the angiogenic stimulus provided by VEGF-secreting ventricular epithelial cells (Breier et al., 1992). In contrast to the transient expression in ependymal cells, VEGF expression persisted in the choroid plexus epithelium of the brain of adult mice, when vascularization of these structures is complete. Risau and co-workers isolated a high-affinity VEGF receptor from proliferating endothelial cells of post-natal day 4–10 mouse brain, when there was maximal endothelial cell proliferation (Millauer et al., 1993). This VEGF receptor belongs to the family of receptor tyrosine kinases and had previously been designated foetal liver kinase-1 (flk-1) (Matthews et al., 1991). In situ hybridization in the embryonic day 14.5 mouse embryo revealed that flk-1 was restricted to capillaries and blood vessels. At day 11.5 in the mouse embryo when the first vascular sprouts begin to radially invade the neuroectoderm from the perineural plexus, expression of flk-1 was high in the perineural vascular plexus and in invading vascular sprouts (Millauer et al., 1993). At embryonic day 14.5, when the neuroectoderm is already highly vascularized, numerous radial vessels as well as branching vessels of the intraneural plexus contained large amounts of flk-1 mRNA (Millauer et al., 1993). Finally, in the adult brain, when angiogenesis had ceased, flk-1 expression was very low and restricted to the choroid plexus endothelial cells. Risau and co-workers demonstrated that flk-1 was highly expressed in early postnatal mouse brain but was down-regulated at post-natal day 15 and was hardly detectable at post-natal day 30 (Kremer et al., 1997). Moreover, they observed that hypoxia up-regulated flk-1 in post-natal day 30 mouse brain slices, suggesting the presence of a hypoxia-inducible factor in the murine neuroectoderm that up-regulated flk-1.
14.6
Characterization of the Blood–Brain Barrier
87
The specific role of VEGF as an embryonic angiogenesis factor was further supported by experiments in which VEGF was overexpressed in the limb of chick embryos, resulting in hypervascularization without alteration of limb morphogenesis (Flamme et al., 1995). Risau described an additional endothelial cell-specific receptor tyrosine kinase, tie 2 (tek), in developing vasculature of the brain, which is down-regulated in adult organisms (Schnürch and Risau, 1993).
14.6 Characterization of the Blood–Brain Barrier Although there has been considerable controversy since the observation by Ehrlich more than 100 years ago that the brain did not take up dyes from the vascular system, the concept of an endothelial blood–brain barrier (BBB) was confirmed by the unequivocal demonstration that the passage of molecules from blood to brain and vice versa was prevented by endothelial tight junctions. There are three major functions implicated in the term BBB: (1) protection of the brain from the blood milieu, (2) selective transport and (3) metabolism or modification of blood- or brain-borne substances. The BBB phenotype develops under the influence of associated brain cells, especially astrocytic glia, and consists of complex tight junctions and a number of specific transport and enzyme systems which regulate molecular traffic across the endothelial cells. The development of the BBB is a complex process that leads to endothelial cells with unique permeability characteristics due to high electrical resistance and the expression of specific transporters and metabolic pathways. Risau had a special interest in understanding the development, differentiation and maintenance of the BBB. Risau and co-workers characterized a monoclonal antibody named HT7 that recognizes a highly glycosylated 48 kDa protein belonging to the immunoglobulin superfamily, which is expressed specifically on cerebral endothelium (Albrecht et al., 1990; Seulberger et al., 1990, 1991). The HT7 antigen is specifically expressed on chicken BBB endothelium but not on other endothelial cells (Risau et al., 1986; Albrecht et al., 1990). It is absent from the fenestrated choroid plexus endothelium but is present on choroid plexus epithelium, the site of the cerebrospinal fluid barrier. Risau and co-workers investigated the expression of HT7 in the brain circumventricular organs which lack a BBB. Using immunohistochemical techniques they found that the protein was absent from the vascular system of pituitary, median eminence, subfornical organ, pineal gland, the organum vasculosum lamina terminalis and sinusoid blood vessels of the area postrema. Embryonic mouse brain tissue transplanted on the chick embryo chorioallantoic membrane induced the expression of this protein in endothelial cells derived from the chick chorioallantoic vessels, which normally do not express this protein. Several homologues in different species have been described (Seulberger et al., 1992).
88
14
The Seminal Work of Werner Risau
In collaboration with Hartwig Wolburg’s laboratory, Risau performed a quantitative analysis of the structure and function of tight junctions in primary cultures of bovine brain endothelial cells using quantitative freeze-fracture electron microscopy and ion and inulin permeability (Wolburg et al., 1994). By freeze-fracture analysis, Risau demonstrated that P-face-associated tight junction particles rather than the number of branching frequency of tight junction strands correlate with BBB function. The complexity of tight junctions, defined as the number of branch points per unit length of tight junctional strands, decreased 5 h after culture but thereafter remained almost constant. In contrast, the association of tight junction particles with the cytoplasmic leaflet of the endothelial membrane bilayer (P-face) decreased continuously with a major drop between 16 and 24 h of culture. The P-face association of tight junctions could be restored to a certain extent by co-culture of endothelial cells with astrocytes or astrocyte-conditioned medium. Co-culture with fibroblasts had no effect on P-face association. This study demonstrated that the cytoplasmic anchoring of the tight junctions plays an important role in the functioning of the BBB. Occludin is an integral membrane protein specifically associated with tight junctions. Although it has been demonstrated not to be necessary for the induction of functional junctions, it was found by establishing transfectants with an N-terminal truncated occludin that the N-terminal half of occludin plays an important role in tight junction assembly and maintenance of the barrier function. Risau and co-workers (Klingler et al., 2000) demonstrated that following removal of calcium from the culture medium of epithelial cells in vitro, protein kinase A is activated and subsequently is involved in the disruption of tight junctions. Engelhardt and Risau (1995) proposed two phases of endothelial– neuroectodermal interactions leading to BBB differentiation. Their model is based on the observation that early during brain angiogenesis induction of specific genes can be observed in brain endothelial cells. In a subsequent phase, secondary interaction of “committed” brain endothelial cells in the developing neuroectodermal elements induces further endothelial differentiation, which leads to the fully functional BBB.
14.7 Angiogenesis in Glioma Glioblastoma is the most common and most malignant tumour of astrocytic origin in human adults. It comprises about 50% of all glial tumours in humans. Glioblastoma multiforme is one of the most vascularized tumours, characterized by numerous and abnormal blood vessels, which rapidly proliferate and invade the brain tissue. Tumour blood vessels are characterized by several alterations consisting of an increase in the number of endothelial caveolae and fenestrations, prominent pinocytotic vesicles, lack of perivascular glial endfeet, as well as opening, loss and/or abnormal morphology of tight junctions, leading to an alteration of the vascular permeability and the loss of the BBB properties. Risau and co-workers observed in low-grade glioma up-regulation of VEGF in some tumour cells, whereas in high-grade gliomas, in particular in glioblastoma,
14.8
The Role of Platelet-Derived Growth Factor in Angiogenesis
89
they observed a significant up-regulation of VEGF mRNA in certain tumour areas (Plate et al., 1992, 1994). Histological analysis revealed a striking association of VEGF mRNA producer cells with areas of necrosis (Plate et al., 1992, 1994). Moreover, Risau and co-workers observed a tumour-stage up-regulation of VEGFR-1 and VEGFR-2 during tumour development and progression. VEGFR-2 transcripts were not detected in normal human adult brain vascular cells. VEGFR-2 transcripts were detected only in vascular cells in high-grade gliomas, but not within low-grade gliomas (Plate et al., 1994). The availability of a suitable rat glioma model has allowed Risau and co-workers to study the mechanisms of glioma angiogenesis in vivo (Plate et al., 1993). Rat C6 glioma cells, when transplanted into the brain of syngeneic recipient rats, form brain tumours which resemble human malignant glioma. In these experimental brain tumours, the expression of VEGF, VEGFR-1 and VEGFR-2 was strongly up-regulated during tumour development, as in malignant human glioma. Risau and co-workers have shown that VEGFR-2 is necessary for glioma angiogenesis. A dominant-negative mutant of VEGFR-2, which lacked most of the intracellular domain, was able to inhibit wild-type VEGFR-2 activity in vitro (Millauer et al., 1994). In mice, tumour growth was significantly inhibited when cells producing a virus encoding the VEGFR-2 mutant were cotransplanted with tumour cells. Histological analysis revealed that tumour growth inhibition was due to inhibition of angiogenesis (Millauer et al., 1994). Since mutations in the tumour suppressor gene p53 represent a common genetic alteration during glioma progression, Risau and co-workers analysed whether p53 may be a negative regulator of VEGF expression, i.e. whether a p53 loss of function may up-regulate VEGF expression in vivo, since they observed a selective up-regulation of VEGF mRNA in certain glioma cells and hypothesized that these cells could be deficient in p53 (Plate et al., 1992). However, subsequent studies failed to provide evidence that p53 loss of function may up-regulate VEGF. In fact, the lack of co-expression of mutant p53 protein and VEGF in tissue sections of malignant glioma does not support this hypothesis (Plate et al., 1994).
14.8 The Role of Platelet-Derived Growth Factor in Angiogenesis Platelet-derived growth factor (PDGF) is secreted by endothelial cells, presumably in response to VEGF, and facilitates recruitment of mural cells. Mutation of PDGF caused failure of recruitment of pericytes (Lindahl et al., 1997, 1998). A detailed analysis of the vessel development in both PDGF and PDGF-receptor mutant embryos showed that smooth muscle cells and pericytes initially form around vessels but as vessels sprout and enlarge, PDGF signalling is required for co-migration and proliferation of supporting cells (Hellstrom et al., 1999). Risau demonstrated that endothelial cells cultured on nitrocellulose membranes secreted a potent PDGF-like chemotactic factor almost exclusively onto the abluminal compartment (Zerwes and Risau, 1987).
90
14
The Seminal Work of Werner Risau
Risau contributed to the protein sequencing, cDNA cloning and expression of functionally active platelet-derived endothelial cell growth factor (PD-ECGF) and demonstrated that PD-ECGF has chemotactic activity for endothelial cells in vitro and angiogenic activity in vivo in the chick embryo chorioallantoic membrane assay (Ishikawa et al., 1989).
14.9 Hypoxia and Angiogenesis Oxygen tension is a crucial factor in new vessel growth, with regions of hypoxia which induce the expression of angiogenic molecules, such as VEGF-A, and hypoxia stimulates vasculogenesis (Hoper and Jahn, 1995). Hypoxia-inducible factor-1α (HIF-1α) is a transcription factor that is selectively stabilized and activated under hypoxic conditions and that coordinates the adaptive response of tissue to hypoxia (Semenza, 1999). Hypoxia and the subsequent alteration of HIF-1α can induce VEGF-A and stromal cell-derived factor-1 (SDF-1), which stimulate migration of endothelial precursors during vasculogenesis (Ramirez-Bergeron et al., 2006). One of the major drives to tumour angiogenesis is hypoxia (Harris, 2002) that characterizes the tumour microenvironment and is a negative prognostic factor for cancer patient response to treatment and survival (Harris, 2002). Production of proangiogenic mediators by tumour cells and inflammatory leucocytes in response to hypoxia was demonstrated both in vitro and in vivo in hypoxic areas of tumours of different origin (Ryan et al., 2000). In particular, VEGF is one of the key angiogenic factors implicated in hypoxia-induced angiogenesis of a wide variety of tumours (Roftsad and Danielsen, 1999; Shweiki et al., 1992; Mizukami et al., 2004). Risau and co-workers (Damert et al., 1997) asked whether the mechanism defined for hypoxia-induced VEGF expression in vitro is similarly involved and sufficient for up-regulation of VEGF gene expression in vivo. To this purpose, they used a LacZ reporter gene under the control of VEGF regulatory sequences in an experimental glioma model and demonstrated that deletion of the HIF-1 binding sites abolished reported gene expression in a special tumour cell subtype, the socalled perinecrotic palisading (PNP) cells that flank necrotic regions within the tumours, indicating that transcriptional activation of VEGF expression in gliomas is mediated by HIF-1. Inclusion of 3 -untranslated sequences from the VEGF gene in the reporter constructs resulted in an increased β-galactosidase expression in the PNP cells, suggesting that mRNA stabilization also contributes to VEGF up-regulation in glioblastoma cells. Moreover, combination of the 5 -flanking region including the HIF-1 site along with 3 -untranslated sequences produced increased levels of β-galactosidase expression in PNP cells. Overall, these data provide experimental evidence that VEGF gene expression is activated in PNP cells by two distinct hypoxia-driven regulatory mechanisms. Marti and Risau (1998) demonstrated that VEGF is induced in vivo by exposing mice to systemic hypoxia. VEGF induction was highest in brain, but also occurred in kidney, testis, lung, heart and liver. In situ hybridization analysis revealed that
14.9
Hypoxia and Angiogenesis
91
a distinct subset of cells within a given organ, such as glial and neurons in brain, tubular cells in kidney and Sertoli cells in testis, responded to the hypoxic stimulus with an increase in VEGF expression. Furthermore, expression of VEGFR-1 was induced by hypoxia in endothelial cells of lung, heart, brain, kidney and liver. Risau and co-workers (Marti et al., 2000) demonstrated that 48 and 72 h after permanent middle cerebral artery occlusion in mouse, a strong increase in the number of newly formed blood vessels was recognizable at the border of infarction. Expression of VEGF and of VEGFRs was strongly up-regulated in the ischaemic border and finally, HIF-1 and HIF-2 were increased in the ischaemic border after 72 h, suggesting a regulatory function for these factors.
Chapter 15
Judah Folkman, A Pioneer in the Study of Angiogenesis
15.1 Early Evidence of Tumour Cells Releasing Specific Growth Factor for Blood Vessels In 1939, Ide et al. (1939) were the first to suggest that tumours release specific factors capable of stimulating the growth of blood vessels. In 1945, Algire and Chalkley (1945) were the first to appreciate that growing malignancies could continuously elicit new capillary growth from the host. They used a transparent chamber implanted in a cat’s skin to study the vasoproliferative reaction secondary to a wound or implantation of normal or neoplastic tissues. They showed that the vasoproliferative response induced by tumour tissues was more substantial and earlier than that induced by normal tissues or following a wound. They concluded that the growth of a tumour is closely connected to the development of an intrinsic vascular network. In 1956, Merwin and Algire (1956) found that the vasoproliferative response of normal or neoplastic tissues transplanted into muscle was not significantly different with respect to the time of onset of new blood vessels, though it was stronger when the implantation was performed in a resection area. In addition, while normal tissues induced a vasoproliferative response confined to the host, tumour tissues induced the formation of neovessels that pierced the implant. Lastly, the intensity of the response seemed to be influenced by the distance between the implant and the host’s vessels: normal tissue was unable to induce a response if placed more than 50 μm away, whereas tumour tissue had a longer activity range. In 1968, Greenblatt and Shubik (1968) implanted Millipore chambers (pore size 0.45 μm) into a hamster’s cheek pouch and placed some tumour fragments around them. In a few days, the growing tumour mass engulfed the whole chamber, whose pores were permeable to the tumour interstitial fluid but not to the tumour cells. New blood vessels, however, were formed in any case very likely through the release of a diffusible factor that could pass through the pores. Ehrman and Knoth (1968) confirmed these data with tumour fragments laid on Millipore filters planted on the chick embryo chorioallantoic membrane (CAM).
Published in “Angiogenesis”, 11:3–10, 2008
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_15, C Springer Science+Business Media B.V. 2010
93
94
15
Judah Folkman, A Pioneer in the Study of Angiogenesis
15.2 Tumours in Isolated Perfused Organs: Absence of Angiogenesis In 1963, Folkman and Becker were studying haemoglobin solutions as potential substitutes for blood transfusion. To test which solution was optimal for tissue survival, they perfused these solutions through the vasculature of canine thyroid glands, by using an apparatus with a silicone rubber oxygenator. The glands survived for about 2 weeks. They could distinguish different haemoglobin preparations by histologic analysis of the thyroid glands after a week or more of continuous arterial perfusion. To determine whether these isolated organs could support growth, they injected them with adult mouse melanoma cells. Tiny tumours developed but stopped growing at 1- to 2-mm diameter and never became vascularized (Folkman, et al., 1963). Endothelial cells swelled and could not proliferate in the presence of free haemoglobin solutions lacking platelets (Gimbrone et al., 1969). However, the tumours were not dead. When they were transplanted to their host mice, they rapidly vascularized and grew to more than 1 cm3 . Folkman et al. observed that when tumour cells were inoculated into isolated perfused organs, tumours were limited in size to 1–2 mm3 (Folkman, et al., 1963). Subsequently, they found that neovascularization does not occur in isolated perfused organs and that tumours transplanted from these organs to syngeneic mice became vascularized and grew rapidly to 1–2 mm3 . This was the first evidence that the absence of neovascularization correlated with severe restriction of tumour growth. The data were consistent with work from Harry Green, who had shown long before that growth of rabbit tumours transplanted into the anterior chamber of the guinea pig coincided with the growth of new blood vessels. Tumours that remained viable, but did not grow, had no visible blood vessels (Green, 1941).
15.3 Hypothesis: Tumour Growth Is Angiogenesis Dependent In 1971, Folkman published in the “New England Journal of Medicine” a hypothesis that tumour growth is angiogenesis dependent and that inhibition of angiogenesis could be therapeutic (Folkman, 1971). This chapter also introduced the term antiangiogenesis to mean the prevention of new vessel sprout from being recruited by a tumour. The hypothesis predicted that tumours would be enable to grow beyond a microscopic size of 1–2 mm3 without continuous recruitment of new capillary blood vessels. This concept is now widely accepted because of supporting data from experimental studies and clinical observations carried out over the intervening years.
15.4 Evidence that Tumours Are Angiogenesis Dependent Folkman and collaborators provide evidence for the dependence of tumour growth on neovascularization:
15.4
Evidence that Tumours Are Angiogenesis Dependent
95
(1) Tumour growth in the avascular cornea proceeds slowly at a linear rate, but after vascularization, tumour growth is exponential (Gimbrone et al., 1974). (2) Tumours suspended in the aqueous fluid of the anterior chamber of the rabbit eye and observed for a period up to 6 weeks remain viable, avascular and of limited size (less than 1 mm3 ) and contain a population of viable and mitotically active tumour cells. These tumours induce neovascularization of the iris vessels but are too remote from these vessels to be invaded by them. After implantation contiguous to the iris, which had abundant blood vessels, the tumours induced neovascularization and grow rapidly, reaching 16,000 times the original size within 2 weeks (Gimbrone et al., 1972). This experiment introduced the concept of tumour dormancy brought about by prevention of neovascularization. In a parallel study tumours were suspended in the aqueous humour of the anterior chamber, placed at various distances from the iris vessels, and compared with tumours implanted directly on the iris and with those implanted in the cornea (Gimbrone et al., 1973). Moving the distant, dormant tumours closer to the iris jump started their growth. This suggested that this type of tumour dormancy was caused not by cell cycle arrest or immune control but by a lack of blood supply. (3) B-16 mouse melanoma, V-79 Chinese hamster lung cells and L-5178 Y murine leukaemia cells were plated in soft agar (Folkman and Hochberg, 1983). After 6–7 days of incubation, spheroid colonies of 0.1 mm were visible. All spheroids first enlarged exponentially for a few days and then continued on a linear growth curve for 5–23 weeks before reaching a diameter beyond which there was no further expansion. This was termed the dormant phase. After the dormant diameter was reached, these spheroids remained viable for 3–5 months or as long as they were frequently transferred to new medium. Cells in the periphery of the spheroid incorporated 3 H-labelled thymidine while cells in the centre died. This is a form of population dormancy in which the proliferating cells near the surface of the spheroid just balance those dying cells deep in the centre of the spheroid. (4) Tumours implanted on the CAM of the chick embryo do not exceed a mean diameter of 0.93±0.29 mm during the prevascular phase (approximately 72 h). Rapid growth begins, however, within 24 h after vascularization and tumours reach a mean diameter of 8.0 ± 2.5 mm by 7 days (Knighton et al., 1977). (5) Tumours grown in the vitreous of the rabbit eye remain viable but attain diameters of less than 0.50 mm for as long as 100 days. Once such a tumour reaches the retinal surface, it becomes neovascularized and within 2 weeks can undergo a 19,000-fold increase in volume over the avascular tumour (Brem et al., 1976). (6) The CAM appears at day 5 during development of the chick embryo. The 3 H-thymidine labelling index of its vascular endothelium decreases with age, with an abrupt reduction at day 11 (Ausprunk et al., 1974). Prior to day 11, labelling index is approximately 23%; during 11 days, the labelling index decreases to 2.8% and, subsequently, the cells begin to acquire the structural characteristics of matured, differentiated endothelium. One-millimetre fragments of fresh Walker 256 carcinoma were implanted on the CAM from day 3
96
15
Judah Folkman, A Pioneer in the Study of Angiogenesis
to day 16 (Knighton et al., 1977). The size of the tumours was measured daily, and the onset of vascularization of each tumour was determined in vivo with a stereomicroscope and confirmed with histological sections. Proliferation of chick capillaries occurred in the neighbourhood of the tumour graft by 24 h after implantation, but capillary sprouts did not penetrate the tumour graft until approximately 72 h. During the avascular phase, tumour diameter did not exceed 1 mm. Small tumour implants of 0.5 mm or less grew to 1 mm and stopped expanding. Larger tumour implants of 2 or 3 mm shrank until they reached 1 mm diameter. During the first 24 h after penetration by capillaries, there was a rapid tumour growth. Neovascularization was not grossly observable with the stereomicroscope until after day 10 or 11. Tumours implanted on the CAM after day 11 grow at slower rate in parallel with the reduced rates of endothelial growth. (7) When tumour grafts of increasing size (from 1 to 4 mm) are implanted on the 9-day CAM, grafts larger than 1 mm undergo necrosis and autolysis during the 72-h prevascular phase. They shrink rapidly until the onset of neovascularization, when rapid tumour growth resumes (Knighton et al., 1977). In another study (Ausprunk et al., 1975), the behaviour of tumour grafts on the CAM was compared to grafts of normal adult and embryonic tissues. In tumour tissue, pre-existing blood vessels within the tumour graft disintegrated by 24 h after implantation. Neovascularization did not occur until after at least 3 days, and only by penetration of proliferating host vessels into the tumour tissue. There was marked neovascularization of host vessels in the neighbourhood of the tumour graft. By contrast, in embryonic graft, pre-existing vessels did not disintegrate. They reattached by anastomosis to the host vessels within 1–2 days, but with minimal or almost no neovascularization on the part of the host vessels. In adult tissues, the pre-existing graft vessels disintegrated, there was no reattachment of their circulation with the host and adult tissues did not stimulate capillary proliferation. These studies suggest that only tumour grafts are capable of stimulating formation of new blood vessels in the host. (8) In transgenic mice that develop carcinomas of the β cells in the pancreatic islets, large tumours arise only from a subset of preneoplastic hyperplastic islets that have become vascularized (Folkman et al., 1989).
15.5 Isolation of the First Angiogenic Tumour Factor Until the early 1970s it was widely assumed that tumours did not produce specific angiogenic proteins. The conventional wisdom was that tumour vasculature was an inflammatory reaction to dying or necrotic tumour cells. Previous studies had shown that tumour-stimulated vessel growth did not require direct contact between tumour and host tissue (Ehrman and Knoth, 1968; Greenblatt and Shubik, 1968). This made sense to Folkman, who reasoned that a soluble factor would be more likely to reach nearly than distant blood vessels. He and
15.6
First Evidence of the Existence of the Avascular and Vascular Phases
97
his colleagues isolated an angiogenic factor in 1971 (Folkman et al., 1971). The homogenate of a Walker 256 carcinoma – a breast tumour of Sprague-Dawley rats – was fractionated by gel filtration on Sephadex G-100. The fraction that exhibited the strongest angiogenic activity had a molecular weight of about 10,000 Da and consisted of 25% RNA, 10% proteins and 58% carbohydrates, plus a possible lipid residue. It was inactivated by digestion with pancreatic ribonuclease or by heating at 56◦ C for 1 h and was not modified when kept at 4◦ C for 3 months, nor when treated with trypsin for more than 3 days. This active fraction was subsequently called “tumour angiogenesis factor” (TAF) (Folkman et al., 1971). Both the cytoplasmic and the nuclear fractions of tumour cells stimulated angiogenesis. In the nuclear fraction, this was found to be associated with nonhistonic proteins (Tuan et al., 1973). Tumour angiogenesis factor has since been nondestructively extracted from several tumour cell lines, and several low molecular weight angiogenic factors have been isolated, again from the Walker 256 carcinoma. These factors induced a vasoproliferative response in vivo when tested on rabbit cornea or chick CAM, and in vitro on cultured endothelial cells (Fenselau et al., 1981; Mc Auslan and Hoffman, 1979; Weiss et al., 1979).
15.6 First Evidence of the Existence of the Avascular and Vascular Phases of Solid Tumour Growth The earliest evidence of the existence of the two phases was obtained by Folkman and collaborators in 1963, who perfused the lobe of a thyroid gland with plasma and inoculated a suspension of melanoma B16 tumour cells through the perfusion fluid. These cells grew into small, clearly visible black nodules. The nodules did not exceed 1 mm in diameter and did not connect with the host’s vascular network. Their outer third generally remained vital, while the interior portion underwent necrosis. Reimplanted nodules, on the other hand, equipped themselves with a vascular network and grew very rapidly. The conclusion was thus drawn that the absence of vascularization limits the growth of solid tumours. Further research by Folkman’s group resulted in an experimental system in which the tumour, or its extracts, could be separated from the vascular bed (Cavallo et al., 1972, 1973). This system was based on subcutaneous insufflation to lift the skin of a rat and form a poorly vascularized region below it. Millipore filters containing Walker 256 cancer cells or their cytoplasmic or nuclear extracts (TAF) were implanted into the fascial floor of the dorsal air sac. At intervals thereafter, 3 H-labelled thymidine was injected into the air sac and the tissues were examined by autoradiography and electron microscopy. Autoradiographs showed thymidine-3 H labelling in endothelial cells of small vessels, 1–3 mm from the site of implantation, as early as 6–8 h after exposure to tumour cells. DNA synthesis by endothelium subsequently increased and within 48 h new blood vessels formation was detected. The presence of labelled endothelial nuclei, endothelial mitosis and regenerating endothelium was confirmed by electron microscopy. Tumour angiogenesis factor
98
15
Judah Folkman, A Pioneer in the Study of Angiogenesis
also induced neovascularization and endothelial DNA synthesis after 48 h. Further ultrastructural autoradiographic studies were carried out with the same model (Cavallo et al., 1973). It was apparent that by 48 h there was ultrastructural evidence of regenerating endothelium, including marked increase in ribosomes and endoplasmic reticulum, scarce or absent pinocytotic vesicles and discontinuous basement membrane. Labelled endothelial cells were seen along newly formed sprouts as well as in parent vessels. Furthermore, pericytes were also shown to synthesize DNA. In another series of experiments, 1 mm fragments from Brown-Pearce and V2 carcinomas were implanted into the avascular stroma of a rabbit cornea 1–6 mm away from the limbic vessels and observed the tumour growth daily with a stereomicroscope (Gimbrone et al., 1974). After 1 week, new blood vessels had invaded the cornea starting from the edge closer to the site of implantation and developed in that direction at 0.2 mm and then about 1 mm/day. Once the vessels reached the tumour, it grew very rapidly to permeate the entire globe within 4 weeks.
15.7 Dormancy of Micrometastases May Be Governed by Angiogenesis Folkman and collaborators found that metastases were suppressed when a primary tumour was implanted and allowed to grow in nude mice, whereas they underwent neovascularization and became clinically evident when primary neoplasm was removed. In the absence of angiogenesis, micrometastases rarely exceeded 0.2-mm diameter and contained many proliferating tumour cells balanced by many apoptotic cells. When they were allowed to become angiogenic, they grew rapidly. Dormancy may be generalizable to a variety of tumours in which blocked angiogenesis results in balanced tumour cell proliferation and apoptosis (Holmgren et al., 1995).
15.8 Prognostic Significance of Tumour Vascularity In 1972, Brem in the Folkman laboratory reported the first quantitative method for histologic grading of tumour angiogenesis. He correlated neovascularization in human brain tumours with tumour grade (Brem et al., 1972). In the early 1990s, Weidner and collaborators (Weidner et al., 1971, 1992, 1993) showed that measurement of microvascular density within isolated regions of high vessel concentration (i.e. hot spots) was a prognostic indicator for human breast and prostate carcinomas. Microvascular density counting protocols have become the morphological gold standard to assess the neovasculature in human tumours. This method requires the use of specific markers to vascular endothelium and of immunohistochemical procedures to visualize microvessels. Microvascular density determined in primary tumours is significantly associated with metastasis and prognosis in several solid and haematological tumours.
15.9
Antiangiogenesis
99
15.9 Antiangiogenesis The existence of specific angiogenesis inhibitors was first postulated by Folkman 1971 in an editorial. No angiogenesis inhibitors existed before 1980, and few scientists thought at that time that such molecules would ever be found. From 1980 to 2005, Folkman’s laboratory (Fig. 15.1) reported the discovery of 12 angiogenesis inhibitors (Table 15.1). The first angiogenesis inhibitor was found in cartilage, an avascular tissue that resists invasion by many tumours (Eisenstein et al., 1973). Brem and Folkman demonstrated that tumour-induced vessels were inhibited by a diffusible factor from neonatal rabbit cartilage (Brem and Folkman, 1975). The partially purified inhibitor suppressed tumour growth when it was infused into the vascular bed of murine and rabbit tumours (Langer et al., 1980). 2-Methoxyoestradiol was first reported by Fotsis et al. (1994), and the article reporting 2-methoxyoestradiol and its molecular mechanism as an inhibitor of tubulin polymerization by acting at the colchicine site was published a month later (D’Amato et al., 1994a).
Fig. 15.1 A portrait of Dr. Judah Folkman in his laboratory
Table 15.1 Angiogenesis inhibitors discovered in Folkman’s laboratory from 1980 to 2005
1980 1982 1985 1990 1994 1994 1994 1997 1999 2005
Interferon alpha–beta Platelet factor 4/protamine Angiostatic steroids Fumagillin Angiostatin Thalidomide 2-Methoxyoestradiol Endostatin Cleaved antithrombin III Caplostatin
100
15
Judah Folkman, A Pioneer in the Study of Angiogenesis
15.9.1 Interferon Alpha Interferon alpha was first shown to inhibit endothelial cell migration in a dosedependent and reversible manner in 1980 by Zetter in Folkman’s laboratory (Brouty-Boye and Zetter, 1980). Since 1988, interferon alpha has been used successfully to cause complete and durable regression of life-threatening pulmonary haemangiomatosis, haemangiomatosis of the brain, airway and liver in infants, recurrent high-grade giant cell tumours refractory to conventional therapy and angioblastomas (Ezekowitz et al., 1972; Folkman, 2002; Kaban et al., 1999). A 12-year-old boy with fatal pulmonary haemangiomatosis had a complete remission and recovered completely after 7 months therapy with interferon alpha. Therapy was continued for 7 years. This led to the successful use of low-dose daily interferon alpha therapy administered subcutaneously to infants with sightthreatening or life-threatening haemangiomas and haemangioendotheliomas of the heart, airway and liver (Ezekowitz et al., 1972; Folkman, 2002; Kaban et al., 1999).
15.9.2 Platelet Factor 4/Protamine Protamine was shown to be an angiogenesis inhibitor (Taylor and Folkman, 1972), but cumulative toxicity from prolonged administration and a narrow window of angiostatic efficacy prevented its consideration for clinical use. Platelet factor 4 was first tested for antiangiogenic activity because its method of binding and neutralizing heparin is similar to that of protamine (Taylor and Folkman, 1972). Recombinant human platelet factor 4 (rHuPF4) has been produced (Maione et al., 1990). It specifically inhibited endothelial proliferation and migration in vitro (Sharpe et al., 1990). The inhibitory activities are associated with the carboxy-terminal region of the molecule. The growth of human colon carcinoma in athymic mice, as well as the growth of murine melanoma, was markedly inhibited by intralesional injections, whereas tumour cells were completely insensitive to rHuPF4 in vitro at levels that inhibited normal endothelial cell proliferation. Systemic administration of rHuPF4 has so far been ineffective against tumour growth, perhaps because of rapid inactivation or clearance of the peptide.
15.9.3 Angiostatic Steroids Folkman had begun to use the CAM of the chick embryo to detect angiogenic activity in fractions being purified from tumour extracts. The addition of heparin increased the speed of development of the angiogenic reaction so that it could be read 1–2 days later (Taylor and Folkman, 1972). But one problem with this assay is that occasionally eggshell dust falls on the CAM and causes background inflammation. Folkman guessed that adding cortisone to the CAM might eliminate the irritation from the shell dust but not abolish the tumour angiogenic reaction.
15.9
Antiangiogenesis
101
As expected, cortisone alone prevented shell dust inflammation without interfering with angiogenesis induced by tumour extracts. The surprise was that when heparin and cortisone were added together tumour angiogenesis was inhibited (Folkman et al., 1983). Furthermore, when this combination of heparin and steroid was suspended in a methylcellulose disc and implanted on the young (6-day) CAM, growing capillaries regressed leaving in their place, 48 h later, an avascular zone. The antiangiogenic effect was specific for growing capillaries. Mature nongrowing capillaries in older membranes were unaffected. Nonanticoagulant heparin had the same effect. A hexasaccharide fragment with a molecular weight of approximately 16,000 was found to be the most potent inhibitor of angiogenesis (in the presence of a corticosteroid). The combination of the heparin hexasaccharide fragment and cortisone also inhibited tumour-induced angiogenesis in the rabbit cornea. The regression of a growing vessel exposed to heparin–steroid combinations begins with endothelial cell rounding and is followed by cessation of endothelial proliferation, desquamation of endothelial cells and retraction of the capillary sprout (Ingber et al., 1986). These events occur as 24–48 h and are accompanied by dissolution of the basement membrane of the new capillary vessels.
15.9.4 Fumagillin Fumagillin was found by Ingber in the Folkman laboratory to inhibit endothelial cell proliferation without causing endothelial cell apoptosis, when a tissue culture plate of endothelial cells became contaminated with a fungus Aspergillus fumigatus Fresenius (Ingber et al., 1990). A conditioned medium from fungal cultures contained an inhibitor of endothelial cell proliferation and angiogenesis, which, upon purification, was found to be fumagillin, a polyene macrolide. When capillary endothelial cells were stimulated by fibroblast growth factor-2 (FGF-2), half-maximal inhibition was observed with fumagillin at 100 pg/mL. This antiproliferative effect appeared to be relatively specific for endothelial cells because inhibition of nonendothelial cells, including tumour cells, was observed at up to 1,000-fold higher concentrations. Scientists at Takeda Chemical Industries (Osaka, Japan) made a synthetic analogue of fumagillin, called TNP-470, which inhibits endothelial proliferation in vitro at a concentration of 3 logs lower than the concentration necessary to inhibit fibroblasts and tumour cells.
15.9.5 Angiostatin and Endostatin They were discovered by M. O’Reilly in Folkman laboratory based on Folkman’s hypothesis of a mechanism to explain the phenomenon that surgical removal of certain tumours leads to rapid growth of remote metastases. This hypothesis said that if tumours produce both stimulators and inhibitors of angiogenesis, an excess
102
15
Judah Folkman, A Pioneer in the Study of Angiogenesis
of inhibitors could accumulate within an angiogenic tumour. In the circulation, however, the ratio would be reversed. Angiogenesis inhibitors would increase relative to stimulators, because of rapid clearance of stimulators from the blood. Folkman formulated this hypothesis after reading Bouck’s first report in 1989 that the emergence of tumour angiogenesis was the result of a shift in balance between positive and negative regulators of angiogenesis in a tumour (Rastinejad et al., 1989). Bouck reported that the switch to angiogenesis during tumorigenesis of transformed hamster cells was associated with downregulation of an inhibitor of angiogenesis, thrombospondin. She suggested that the switch to the angiogenic phenotype could be the result of a shift in the net balance of positive and negative regulators of angiogenesis. In 1991, O’Reilly began to screen a variety of transplantable murine tumours for their ability to suppress metastases. A Lewis lung carcinoma was the most efficient. When the metastasis-suppressing primary tumour was present in the dorsal subcutaneous position, micrcoscopic lung metastases remained dormant at a diameter of less than 200 μm surrounding a pre-existing microvessel, but revealed no new vessels. Within 5 days after surgical removal of the primary tumour, lung metastases became highly angiogenic and grew rapidly, killing their host by 15 days (O’Reilly et al., 1994). This striking evidence that primary tumour could suppress angiogenesis in its secondary metastases by a circulating inhibitor was further supported by the demonstration that a primary tumour could also suppress corneal angiogenesis by an implanted pellet of FGF-2. O’Reilly then succeeded in purifying this inhibitor from the serum and urine of tumour-bearing animals. It was a 38-kDa internal fragment identical in amino acid sequence to the first four kringle structures of plasminogen and it was named angiostatin. Angiostatin specifically inhibited the proliferation of growing vascular endothelial cells and had no effect on resting confluent endothelial cells or on other cell types, including smooth muscle cells, epithelial cells, fibroblasts and tumour cells. It also inhibited growth of primary tumours by up to 98% (O’Reilly et al., 1996) and was able to induce regression of large tumours (1–2% of body weight) and maintain them at a microscopic dormant size. Based upon the same rationale and strategy, O’Reilly isolated and purified another angiogenesis inhibitor from a murine haemangioendothelioma. This inhibitor called endostatin (O’Reilly et al., 1997) is a 20-kDa protein with an N-terminal amino acid sequence identical to the carboxyterminus of collagen XVIII. It was purified directly from tumour cell-conditioned medium. Endostatin is also a specific inhibitor of endothelial proliferation and has no effects on resting endothelial cells or on other cell types. It is slightly more potent than angiostatin and also causes regression of large tumours to a microscopic size.
15.9.6 Thalidomide In 1994, D’Amato in the Folkman laboratory reported that thalidomide is an angiogenic inhibitor (D’Amato et al., 1994b). Corneal neovascularization in rabbits induced by FGF-2 or VEGF was blocked by orally administered thalidomide. This
15.9
Antiangiogenesis
103
activity of thalidomide was mainly the result of its direct effect on inhibiting new blood vessel formation and not on suppression of infiltrating host inflammatory cells. Histologic sections of the pretreated neovascularized corneas were virtually free of inflammatory cells. Thalidomide also inhibited corneal neovascularization in mice, but it was necessary to give the drug by the intraperitoneal route and at high doses, because mice do not metabolize thalidomide effectively. 15.9.6.1 2-Methoxyoestradiol In 1994, D’Amato in Folkman laboratory demonstrated that a metabolite of oestradiol, 2-methoxyoestradiol inhibited angiogenesis in the chick CAM (D’Amato et al., 1994a). Moreover, since 2-methoxyoestradiol causes mitotic perturbations, they examined its interactions with tubulin and showed that 2-methoxyoestradiol bound to colchicine site of tubulin and, depending on reaction conditions, either inhibited assembly or seems to be incorporated into a polymer with altered stability properties.
15.9.7 Cleaved Antithrombin III A human small cell lung carcinoma suppressed angiogenesis and tumour growth at remote sites in immunodeficient mice. These cells generated an enzyme in vitro that converted the 58-kDa conformation of circulating antithrombin III to a 53-kDa form of the protein (O’Reilly et al., 1999) in which the externally configured stressed loop of antithrombin was retracted into the body of the molecule. The 53-kDa form is a specific endothelial inhibitor and a potent angiogenesis inhibitor and has no thrombin-binding activity.
15.9.8 Caplostatin Caplostatin is a nontoxic synthetic analogue of fumagillin conjugated to a watersoluble-N-(2-hydroxypropyl)methacrylamide (HPMA) copolymer (Satchi-Fainaro et al., 2004, 2005). Caplostatin has a similar broad antitumour spectrum of TNP470 without any toxicity. In addition to its antiangiogenic activity, caplostatin is the most potent known inhibitor of vascular permeability (Satchi-Fainaro et al., 2005). Caplostatin prevents vascular leakage induced by VEGF, bradykinin, histamine and platelet-activating factor and prevents pulmonary oedema induced by interleukin 2.
15.9.9 Antiangiogenic Chemotherapy Browder in the Folkman laboratory was the first to demonstrate the concept that by optimizing the dosing schedule of conventional cytotoxic chemotherapy to achieve
104
15
Judah Folkman, A Pioneer in the Study of Angiogenesis
more sustained apoptosis of endothelial cells in the vascular bed of a tumour, it is possible to achieve more effective control of tumour growth in mice, even if the tumour cells are drug resistant (Browder et al., 2000). Conventional chemotherapy is administered at maximum-tolerated doses followed by off-therapy intervals of 2–3 weeks to allow the bone marrow and gastrointestinal tract to recover. In contrast, antiangiogenic chemotherapy is administered more frequently at lower doses, without long interruptions in therapy and with little or no toxicity. During antiangiogenic chemotherapy, endothelial cell apoptosis and capillary dropout precede the death of tumour cells that surround each capillary.
15.10 Concluding Remarks Currently, several compounds with angiostatic activity are approved for clinical use, and many are in late-stage clinical development. However, the results from clinical trials have not shown the antitumour effects which were expected following preclinical studies. It appears that clinical applications of antiangiogenic therapy are more complex than originally thought. The main problem in the development of antiangiogenic agents is that multiple angiogenic molecules may be produced by tumours, and tumours at different stages of development may depend on different angiogenic factors for their blood supply. Therefore, blocking a single angiogenic molecule was expected to have little or no impact on tumour growth. Current development of targeted antiangiogenic agents include their use in adjuvant settings and the combination of different antiangiogenic inhibitors to take a more comprehensive approach in blocking tumour angiogenesis. Advancing insights into fundamental mechanisms will be necessary in the development of novel anticancer strategies based on inhibition of angiogenesis.
Chapter 16
The Contribution of Harold F. Dvorak to the Study of Tumour Angiogenesis and Stroma Generation Mechanisms
16.1 Biographical Notes Harold F. Dvorak stepped down as chair of pathology at Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, USA, in July 2006 after 26 years of devoting his life to basic science cancer research; he is emeritus Mallinckrodt professor of pathology at Harvard Medical School. He has served on the Harvard Medical School Faculty since 1967 and at Beth Israel since 1979 and has written over 220 original scientific works (Fig. 16.1). Dvorak is a fellow of the American Association for the Advancement of Science and of the National Foundation for Cancer Research and has served as president of
Fig. 16.1 A portrait of Harold F. Dvorak
Published in “Endothelium”, 14:131–135, 2007
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_16, C Springer Science+Business Media B.V. 2010
105
106
16
Study of Tumour Angiogenesis and Stroma Generation Mechanisms
the American Society for Investigative Pathology. He was awarded the 2002 RousWhipple Award by the American Society for Investigative Pathology. Educated at Princeton University and Harvard Medical School, he did residency training in pathology at the Massachusetts General Hospital and postdoctoral research training at the National Institutes of Health.
16.2 Tumour Blood Vessels Are Hyperpermeable to Plasma Proteins and to Other Circulating Macromolecules Back in the 1970s, Dvorak investigated the cellular composition of delayed-type hypersensitivity reactions to soluble protein antigen in guinea pigs and discovered that basophilic leucocytes were a prominent component (Dvorak et al., 1970). Then, Dvorak used as antigens two guinea pig tumour cell lines and demonstrated that the immune response elicited by these tumours included basophils and macrophages (Dvorak et al., 1973). Moreover, within days of transplant, the tumours were organized into clumps of cells that were separated by spaces containing thin strands of fibrillary material, constituted by cross-linked fibrin, as demonstrated by electron microscopy, immunohistochemistry and biochemistry (Dvorak et al., 1979a, b). Once deposited, cross-linked fibrin behaves as a gel that causes oedema by trafficking extravasated plasma and provides a proangiogenic stroma. In fact, endothelial cells, fibroblasts and inflammatory cells synthesize and secrete the matrix proteins, proteoglycans and glycosaminoglycans that comprise mature tumour stroma and express adhesion molecules whose interaction with fibrin allows them to move freely in tumour stroma. Finally, the fibrin matrix supports proliferation of the tumour cells. Dvorak demonstrated that vascular hyperpermeability to fibrinogen and other plasma proteins, as well as fibrin deposition, is a common feature of many animal and human tumours, both transplantable and autochthonous (Brown et al., 1988; Dvorak et al., 1981, 1983, 1984; Harris et al., 1982). Hyperpermeable vessels were especially prominent at the tumour–host interface and it was therefore not certain whether tumour cells were permeabilizing normal host microvessels and/or were generating the formation of new, abnormal blood vessels that were intrinsically permeable.
16.3 Vascular Permeability Factor Activity Is Present in Tumour Culture Supernatants Testing cell-free supernatants from a variety of human and animal tumour cells in the Miles assay (Miles and Miles, 1952), Dvorak found that supernatants from nearly all of them generated an intense blue spot due to extravasated Evans blue, whereas those from several normal cells did not (Dvorak et al., 1979a). Dvorak called this tumour supernatant permeabilizing activity vascular permeability factor
16.4
The Discovery of Vascular Permeability Factor/Vascular Endothelial Growth Factor
107
(VPF). With a potency some 50,000 times that of histamine (Dvorak et al., 1992; Senger et al., 1983), VPF is effective at concentrations well below 1 nM in the Miles assay. Dvorak showed that VPF was non-dialysable and therefore likely to be a macromolecule, while inhibition of protein synthesis profoundly depressed its secretion and heat and proteases largely inactivated its activity (Senger et al., 1983). VPF does not itself provoke mast cell degranulation or induce a significant inflammatory cell infiltrate. The permeabilizing action of VPF is not blocked by inhibitors of inflammation, including those that block histamine, thrombin and platelet-activating factor (Dvorak et al., 1979a; Senger et al., 1993). VPF increases the permeability of microvessels, primarily post-capillary venules and small veins, to circulating macromolecules. VPF permeabilizes a number of vascular beds, including those of the skin, subcutaneous tissues, peritoneal wall, mesentery and diaphragm (Dvorak et al., 1979b; Collins et al., 1993; Nagy et al., 1995).
16.4 The Discovery of Vascular Permeability Factor/Vascular Endothelial Growth Factor Senger purified VPF to homogeneity with heparin-Sepharose and hydroxylapatite chromatography and demonstrated that VPF is a 34- to 43-kDa dimeric protein whose activity was lost by reduction but was unaffected by deglycosylation (Senger et al., 1983; Yeo et al., 1991). However, the affinity of VPF for heparin is substantially lower than that of other typical heparin-binding growth factors, such as basic fibroblast growth factor (Senger et al., 1983). Senger sequenced the N-terminus and made use of this sequence to prepare a rabbit antibody against a peptide corresponding to the first 24 amino acids of VPF (Senger et al., 1983). This antibody abolished all of the permeability-increasing activity present in culture medium from several guinea pig and rat tumours and prevented circulating albumin from accumulating in tumour ascites fluid. Subsequently, Connolly and co-workers at Monsanto Company showed that VPF is an endothelial mitogen in vitro and an angiogenic factor in vivo (Connolly et al., 1989). Independently, Ferrara and co-workers at Genentech showed that VPF is an endothelial cell mitogen in vitro and gave it the name vascular endothelial growth factor (VEGF) (Ferrara and Henzel, 1989; Gospodarowicz et al., 1989; Leung et al., 1989), more recently known as VEGF-A. The VEGF-A gene is alternatively spliced to yield major isoforms of 189, 165 and 121 amino acids. The murine homologues are in each case one amino acid shorter. The 164/5 isoform is generally most abundantly expressed. All the activities of VEGF-A 164/5 are mediated through two receptor tyrosine kinases (VEGFR-1, flt-1) and VEGFR-2 (flk, KDR) and through a non-tyrosine kinase receptor, neuropilin. Therefore, much of the endothelial cell proliferation induced by VPF/VEGF in vivo must be mediated indirectly by the activation of clotting, generation of
108
16
Study of Tumour Angiogenesis and Stroma Generation Mechanisms
thrombin, deposition of fibrin, release of platelet growth factors and extensive reprogramming of endothelial cell gene expression patterns. At low levels VPF/VEGF does not increase vascular permeability or induce angiogenesis and it might have other functions in normal physiology such as acting as an endothelial cell survival factor (Benjamin and Keshet, 1997) or preventing endothelial cell apoptosis and senescence (Watanabe et al., 1997; Benjamin et al., 1999). Fluid accumulation results from VPF-induced leakage of plasma through hyperpermeable microvessels but is also favoured by the fact that tumours in general lack lymphatic vessels and hence are unable to drain extravasated proteinaceous fluid effectively.
16.5 Tumours: Wounds That Not Heal Dvorak pointed out that similarities exist between tumour stroma generation and wound healing. He noted that wounds, like tumours, secrete VPF, causing blood vessels to leak plasma fibrinogen which stimulates blood vessel growth and provides a matrix on which they can spread. Unlike wounds, however, that turn off VPF production after healing, tumours did not turn off their VPF production and instead continued to make large amounts of VPF, allowing malignant cells to continue to induce new blood vessels and so to grow and spread. Thus, tumours behave as wounds that fail to heal (Dvorak, 1986). Wounds in rodent skin, like tumours, secrete VPF/VEGF: within 24 h of wounding VPF/VEGF mRNA expression increases in epidermal keratinocytes at the wound edge (Brown et al., 1992). VPF/VEGF overexpression reaches a peak at 2–3 days and persists at an elevated level for about 1 week, the time required for granulation tissue to form and migrating keratinocytes to cover the wound defect. In contrast to tumours, VPF/VEGF expression was down-regulated as healing progressed and, parallel with the decreased expression of VPF/VEGF, vascular permeability returned to normal. In contrast to normal mice, congenitally diabetic db/db mice have elevated endogenous levels of VPF/VEGF mRNA in their nude skin, which increase transiently after wounding. However, the rise of VPF/VEGF is not sustained and, as granulation tissue forms, VPF/VEGF expression plummets to barely detectable levels, thus associating decreasing VPF/VEGF expression with defective wound healing (Peters et al., 1993).
16.6 Expression of VEGF-A-164 In Vivo To investigate the capacity of VEGF-A-164 to induce tumour-like blood vessels and stroma, Dvorak and co-workers sought to achieve a constant rate of expression of VEGF-A-164 in normal mouse tissues for a sufficient period of time to induce angiogenesis and new stroma, by using a non-replicating adenoviral expression
16.7
The New Blood Vessels Generated by VEGF-A-164
109
vector (Ad-VEGF-A-164) (Pettersson et al., 2000). Upon introduction into mouse tissues, infected host cells expressed VEGF-A-164 protein within a few hours and continued to secrete it at fairly constant levels for approximately 2 weeks. With hours of injection into any of several tissues of nude mice or rats, host cells began to express VEGF-164 mRNA and continued to do so at steady levels for 10–14 days, generating a strong angiogenic response (Pettersson et al., 2000). The initial response (1–3 days) was similar in all tissues studied and consisted of vascular hyperpermeability, fibrin deposition and oedema. Enlarged lymphatics were first recognized at about 3 days after injection of AdVEGF-A-164; they increased in size and number over the course of several weeks and persisted in some instances for more than 1 year. They lack pericytes and a well-developed basement membrane and enlarge in response to oedema (Feng et al., 2002). The finding that VEGF-A-induced lymphangiogenesis was unexpected. In fact, VEGF-C and VEGF-D were thought to be responsible for lymphatic development. Thereafter, VEGF-A-164 expression declined gradually and, by 4–6 weeks, was no longer detectable by in situ hybridization; by about 8 weeks the angiogenic response had largely resolved and microvascular density had returned to near-normal levels.
16.7 The New Blood Vessels Generated by VEGF-A-164 The angiogenic response induced by Ad-VEGF-A-164 leads to the formation of four distinct types of new vessels, namely mother vessels, daughter capillaries, glomeruloid bodies and vascular malformations. Within 18 h of Ad-VEGF-A-164 injection, thin-walled, hyperpermeable and strongly VEGFR-2-positive sinusoids appeared, defined as “mother vessels”, which arose from the enlargement of pre-existing microvessels, primarily venules (Pettersson et al., 2000). For the first 48 h, mother vessels formed without significant endothelial cell or pericyte division; thereafter, both cell types proliferated extensively. “Mother” vessels continued to develop for about 5 days following injection of Ad-VEGF-A-164 into tissues. Mother vessel formation was accompanied by progressive endothelial cell thinning and a reduction in vesiculo-vacuolar organelles (VVOs) and vacuoles. VVOs are grape-like clusters of uncoated, largely para-junctional, cytoplasmic vesicles and vacuoles that traverse the endothelial cytoplasm from the lumen to the ablumen (Dvorak et al., 1996). VVOs provided an intracellular store of membrane that could be rapidly mobilized to the plasma membrane to increase the cell surface area and permit vessel enlargement. VVO provided a VEGF-A-regulated trans-cellular pathway for macromolecule extravasation (Feng et al., 1996). Many “mother” vessels underwent a process of bridging, in which cells expressing endothelial markers projected processes into and across mother vessel lumens and formed transluminal bridges that divided “mother” vessels into smaller sized channels to form individual smaller caliber daughter capillaries.
110
16
Study of Tumour Angiogenesis and Stroma Generation Mechanisms
Glomeruloid bodies (GB) are poorly organized vascular structures that resemble renal glomeruli and are commonly found in glioblastoma multiforme and other human tumours (Lantos et al., 1997). GB precursors were first recognized as focal collections of large, primitive cells in the endothelial lining of mother vessels. These cells proliferated rapidly and extended both into the vascular lumen and outwards into the extravascular tissue. As they expanded, GBs severely impaired the “mother” vessel in which they had arisen. After about 14 days, as VEGF-A 164 expression declined, GBs progressively modified through a process that involved both apoptosis and reorganization of endothelial cells and pericytes to form capillaries with a normal appearance. Vascular malformations in mouse form as “mother” vessels acquire an irregular coating of smooth muscle cells and/or perivascular fibrosis (Pettersson et al., 2000), resembling the vascular malformations found in patients (McKee, 1996). Unlike mother vessels and GBs, vascular malformations persist indefinitely in the mouse long after VEGF-A-164 expression has ceased.
Chapter 17
Napoleone Ferrara and the Saga of Vascular Endothelial Growth Factor
17.1 Introduction Napoleone Ferrara (Fig. 17.1) was born on 26 July 1956 in Catania, Italy. He studied medicine and obtained a medical degree from the University of Catania Medical School in 1981. Ferrara developed research interests in the fields of neuroendocrinology and reproductive biology, working with Professor Umberto Scapagnini. Ferrara joined Genentech in 1988 after postdoctoral training at the University of California at San Francisco, where he performed research under a fellowship in the Department of Obstetrics, Gynecology, and Reproductive Sciences in Doctor
Fig. 17.1 A portrait of Dr. Napoleone Ferrara Published in “Endothelium”, 15:1–8, 2008
D. Ribatti, Protagonists of Medicine, DOI 10.1007/978-90-481-3741-1_17, C Springer Science+Business Media B.V. 2010
111
112
17
Napoleone Ferrara and the Saga of Vascular Endothelial Growth Factor
Richard Wiener’s laboratory. There he was handed the task of exploring the basic science of pituitary gland. He was able to isolate and culture for the first time follicular cells, a population of non-hormone secreting cells from the anterior pituitary of cows (Ferrara et al., 1987). These cells had unclear function, but, intriguingly, their cytoplasmic projections were known to have intimate relations with the perivascular spaces, suggesting the possibility that these cells may play some role in the regulation of the pituitary vasculature. One day Ferrara mixed supernatants from cultures of follicular cells with endothelial cells. To his great surprise, the endothelial cells started proliferating rapidly. Ferrara theorized that the pituitary cells were secreting an angiogenic protein. In those days, the rate-limiting step in the discovery of regulatory proteins was isolating the protein to near homogeneity and then obtaining an NH2 -terminal amino acid sequence. The amino acid sequence could enable the design of an oligonucleotide probe suitable for screening cDNA libraries. Emphasizing the difficulties and challenges associated with this process, Senger et al. described in 1983 the activity and partial purification of vascular permeability factor (VPF), a permeability-enhancing factor that eventually proved to be the same molecule as vascular endothelial growth factor (VEGF) (Senger et al., 1983). However, it took these investigators 7 additional years of work until they were able to report the full purification and NH2 -terminal amino acid sequencing of VPF (Senger et al., 1990). Very little progress in understanding the role of this protein was possible in the meanwhile. Ferrara and his colleagues at Genentech were the first to isolate and clone vascular endothelial growth factor (VEGF). Ferrara said “I worked on the isolation of VEGF in my spare time during my first 6 months to a year at Genentech. Once we cloned VEGF in 1989, the company became interested and this became more and more my full time pursuit.” Ferrara’s laboratory has investigated many aspects of VEGF biochemistry/molecular biology, including the identification and characterization of its receptors (Flt-1 and Flk/KDR), regulation of VEGF activity by alternative RNA splicing and by extracellular proteolytic mechanisms, structure/function studies on the factor and its receptors, elucidation of its role in angiogenesis in bone and reproductive system. In 1993, Ferrara reported that inhibition of VEGF-induced angiogenesis by specific monoclonal antibodies resulted in a dramatic suppression of the growth of a variety of tumours in vivo. These findings provided the first direct evidence that inhibition of angiogenesis may suppress tumour growth and blocking VEGF action could have therapeutic value for a variety of malignancies. Ferrara said “We were surprised to find out that an antibody that selectively targeted human VEGF-A substantially inhibited the growth of several human tumour cell lines transplanted in nude mice. Considering that these cell lines were known to produce several other angiogenic factors, it was truly unexpected that blocking VEGF alone could have such a profound impact on tumor growth.” The first antiangiogenic agent approved by the Food and Drug Administration (FDA) is bevacizumab (Avastin), a recombinant humanized anti-VEGF monoclonal antibody developed in Ferrara’s laboratory, now in phase III clinical trials as a
17.2
The Isolation of VEGF
113
treatment for several solid tumours, including colorectal, non-small cell lung and breast cancer. Patients with colorectal cancer treated with Avastin in such a trial showed a highly significant increase in time to progression and survival. Ferrara said that “Avastin is derived from a murine anti-VEGF monoclonal antibody that had been engineered in such way that approximately 7% of the murine residues, including the complementary determining regions (involved in antigen binding), are inserted in a human antibody framework. The purpose of a humanization is to avoid the immune responses that are frequently associated with the administration of murine and sometimes chimeric antibodies.”
17.2 The Isolation of VEGF In 1979, Harold D. Dvorak testing cell-free supernatants from a variety of human and animal tumour cells in the Miles assay found that supernatants from nearly all of them generated an intense blue spot due to extravasated Evans blue, whereas those from several normal cells did not (Dvorak et al., 1979). Dvorak called this tumour supernatant permeabilizing activity vascular permeability factor (VPF). With a potency some 50,000 times that of histamine (Dvorak et al., 1992; Senger et al., 1983), VPF is effective at concentrations well below 1 nM in the Miles assay. Dvorak showed that VPF was non-dialysable and therefore likely to be a macromolecule, while inhibition of protein synthesis profoundly depressed its secretion and heat and proteases largely inactivated its activity (Senger et al., 1983). VPF does not itself provoke mast cell degranulation or induce a significant inflammatory cell infiltrate. VPF increases the permeability of microvessels, primarily post-capillary venules and small veins, to circulating macromolecules. VPF permeabilizes a number of vascular beds, including those of the skin, subcutaneous tissues, peritoneal wall, mesentery and diaphragm (Dvorak et al., 1979b; Collins et al., 1993; Nagy et al., 1995). In 1983, Senger purified VPF to homogeneity with heparin-Sepharose and hydroxylapatite chromatography and demonstrated that VPF is a 34- to 43-kDa dimeric protein whose activity was lost by reduction but was unaffected by deglycosylation (Senger et al., 1983). However, no amino acid sequence data were obtained, precluding the establishment of the identity of VPF at that time and the affinity of VPF for heparin was substantially lower than that of other typical heparin-binding growth factors, such as basic fibroblast growth factor (bFGF) (Senger et al., 1983). Subsequently, Connolly and co-workers at Monsanto Company showed that VPF is an endothelial mitogen in vitro and an angiogenic factor in vivo (Connolly et al., 1989). Independently in 1989, Ferrara and Henzel reported the isolation of a diffusible endothelial cell-specific mitogen from medium conditioned by bovine pituitary follicular cells, which they named “vascular endothelial growth factor” (VEGF) to reflect the restricted target cell specificity of this molecule. NH2 -terminal amino acid sequencing of purified VEGF proved that this protein was distinct from the known endothelial cell mitogens such as acidic FGF (aFGF) and bFGF and indeed did not
114
17
Napoleone Ferrara and the Saga of Vascular Endothelial Growth Factor
match any known protein in available databases (Ferrara and Henzel, 1989). By the end of 1989, Ferrara reported the isolation of cDNA clones for bovine VEGF 164 and three human VEGF isoforms: VEGF 121, VEGF 165 and VEGF 189 (Leung et al., 1989). Subsequent studies indicated that these isoforms had markedly different properties in terms of diffusibility and binding to heparin. VEGF 121, which lacked heparin binding, was highly diffusible, whereas VEGF 189, a highly basic and heparin-binding protein, was almost completely sequestered in the extracellular matrix, VEGF 165 had intermediate properties (Houck et al., 1991). Additionally, some proteases like plasmin were found to cleave heparin-binding VEGF isoforms in the COOH terminus and thus generating a non-heparin-binding diffusible fragment (Houck et al., 1992). These early studies suggested that both alternative RNA splicing and extracellular proteolysis regulate the activity of VEGF. Over the years, five VEGF-related genes have been identified (VEGF-A, VEGFB, VEGF-C, VEGF-D and VEGF-E). There are five characterized VEGF-A isoforms of 121, 145, 165, 189 and 206 amino acids in mammals, generated by alternative splicing of the mRNA from a single gene comprising eight exons (Fig. 17.2). They display differential interactions with related receptor tyrosine kinases VEGFR-1/Flt-1, VEGFR-2/Flk-1, VEGFR-3/Flt-4 and neuropilin-1 and neuropilin-2 (NRP-1 and NRP-2) (Fig. 17.3). As a result of the receptor activation and subsequent signal transduction, VEGF target cells may proliferate, migrate or alter gene expression, e.g. of matrix metalloproteinases or cytokines. VEGFR-1 and VEGFR-2 are restricted largely to vascular endothelium in their expression, accounting for the specificity of action of this growth factor family. In
Fig. 17.2 A threedimensional structure of VEGF-A Source: Protein Data Bank
17.2
The Isolation of VEGF
115
Fig. 17.3 VEGF growth factors and receptors family
1992 in a collaborative study between Ferrara’s laboratory and Lewis Williams’s group at the University of California at San Francisco, VEGFR-1 was shown to be a high-affinity VEGF receptor (de Vries et al., 1992). Ferrara also demonstrated that VEGFR-1 expression is up-regulated by hypoxia via a hypoxia-inducible factor (HIF)-1-dependent mechanism (Gerber et al., 1997) and that VEGFR-1 binds not only VEGF-A but also placental growth factor (PlGF) (Park et al., 1994). Ferrara initially proposed that VEGFR-1 may be not primarily a receptor transmitting a mitogenic signal, but rather a “decoy” receptor, able to regulate in a negative fashion the activity of VEGF on the vascular endothelium, by sequestering and rendering this factor less available to VEGFR-2 (Park et al., 1994). Thus, the observed potentiation of the action of VEGF by PlGF could be explained, at least in part, by displacement of VEGF from VEGFR-1 binding (Park et al., 1994). VEGFR-3 is restricted largely to the lymphatic endothelium (Kukk et al., 1996). VEGFR-3 may play a role in disorders involving the lymphatic system and angiogenesis and may be of potential use in drug targeting, in vivo imaging of the lymphatic vessels and in therapeutic lymphangiogenesis. VEGF-C binds to VEGFR-3, expressed on lymphatic endothelium, and has been implicated in lymphangiogenesis. Like VEGF-C, to which is structurally related, VEGF-D is an endothelial cell mitogen and interacts with VEGFR-2 and VEGFR-3. VEGF-E, encoded by the open reading frame (ORF) virus, induces angiogenesis through an interaction with VEGFR-2 (Meyer et al., 1999). Overexpression of VEGF-C and VEGF-D in transgenic mice induces the formation of hyperplastic lymphatic vessels. Conversely, inhibition of VEGF-C and/or VEGF-D by overexpression of a soluble form of VEGFR-3 in the skin of transgenic mice leads to inhibition of lymphatic vessel growth (Jussila and Alitalo, 2002). Transgenic inactivation of both VEGF-C alleles results in prenatal death: endothelial cells commit to the lymphatic lineage, but do not sprout from veins (Karkkainen et al., 2004).
116
17
Napoleone Ferrara and the Saga of Vascular Endothelial Growth Factor
NRP-1 is important for both blood vessel development and development of the nervous system and is a receptor for semaphorin-3A, which acts as an axonalrepellent factor. NRP-1 form complexes with either VEGFR-1 or VEGFR-2 and is an enhancer of VEGFR-2 activity (Soker et al., 1998; Wihitaker et al., 2001). In this way, NRP-1 contributes to the sum of proangiogenic functions mediated by VEGFR-2, and it might also participate in endothelial guidance and vascular patterning (Gerhardt et al., 2004). Moreover, stimulation by nerve growth factor (NGF) and VEGF activates two common intracellular signalling cascades in endothelial cells, the Ras/ERK and P13K/Akt pathways, both of which are involved in cell proliferation and survival, suggesting that NGF, acting in concert with VEGF, plays a role in controlling angiogenic processes (Nico et al., 2008). Recent findings revealed that VEGF has direct effects on neural cells (Raab and Plate, 2007). Genetic studies showed that mice with reduced VEGF levels develop adult-onset motor neuron degeneration, reminiscent of the human neurodegenerative disorder amyotrophic lateral sclerosis (ALS) (Bogaert et al., 2006). VEGF may also affect neuron death implicated in other neurological disorders such as diabetic and ischemic neuropathy, nerve regeneration, Parkinson’s disease, Alzheimer’s disease and multiple sclerosis (Greenberg and Jin, 2005). These findings have raised growing interest in assessing the therapeutic potential of VEGF for neurodegenerative disorders (Lambrechts and Carmeliet, 2006).
17.3 Role of VEGF in Embryonic Vasculogenesis and Angiogenesis In 1996, Ferrara’s laboratory (Ferrara et al., 1996) and a collaborative effort between Peter Carmeliet in Leuven, Werner Risau in Martinsried and Andras Nagy in Toronto (Carmeliet et al., 1996) demonstrated an essential role of VEGF in embryonic vasculogenesis and angiogenesis in the mouse. Inactivation of a single VEGF allele resulted in embryonic lethality between days 11 and 12. The VEGF+/– embryos exhibited a number of developmental anomalies. The forebrain region appeared significantly underdeveloped. In the heart, the outflow region was grossly malformed; the dorsal aortas were rudimentary and the thickness of the ventricular wall was markedly decreased. The yolk sac revealed a substantially reduced number of nucleated red blood cells within the blood islands, indicating that VEGF regulated both vasculogenesis and haematopoiesis. Also, the vitelline veins failed to fuse within the vascular plexus of the yolk sac. Significant defects in the vasculature of other tissues, including placenta and nervous system, were evidenced. For example, in the nervous system of heterozygous embryos at day 10.5, vascular elements could be demonstrated in the mesenchyme, but not in neuroepithelium and the failure of blood vessels ingrowth was accompanied by apoptosis and disorganization of neuroepithelial cells.
17.4
Endocrine Gland-Derived VEGF
117
17.4 Endocrine Gland-Derived VEGF In 2001, Ferrara identified an endocrine gland-derived VEGF (EG-VEGF) as a novel human endothelial mitogen, through a bioassay assessing the ability of a library of purified human-secreted proteins to promote the growth of primary adrenal cortex capillary endothelial cells (Le Couter et al., 2001). EG-VEGF is not structurally related to VEGF but belongs to a unique gene family having distant homology to Dickoff, an inhibitor of Wnt signaling (Le Couter et al., 2002). Whereas VEGF mRNA is strongly expressed in early stage corpus luteum, coincident with the initial development of a capillary plexus, its expression is markedly reduced by mid-luteal phase. In contrast, EG-VEGF is expressed later than VEGF but persists throughout mid- and early–late luteal phase, suggesting that EG-VEGF may be important for the persistence and adequacy of luteal function (Ferrara et al., 2003). This indicates that other angiogenic factors are also likely to have organ-specific angiogenic effects. The ability of certain tumours to escape anti-VEGF strategies might be due, at least in part, to the expression of such molecules. In this context, initial evidence indicates that VEGF and EG-VEGF are coexpressed by adrenal carcinomas, raising the possibility that a complete inhibition of angiogenesis in these tumours might require blocking both VEGF and EG-VEGF signaling pathways (Le Couter et al., 2002). The idea developed by Ferrara of an organ-specific regulation of angiogenesis has been ascribed to genetic predisposition and “microenvironmental” influences (Cleaver and Melton, 2003). For instance, the vasculature of liver, spleen and bone marrow sinusoids is highly permeable because vessels are lined by discontinuous endothelial cells. Conversely, endothelial cells in the brain and retinal capillaries present many tight junctions that contribute to blood–tissue barrier. Continuous endothelial cell capillaries are also found in the bone tissue, skeletal muscles, myocardium, testes and ovaries. Endothelial cells in endocrine glands and kidney are fenestrated and fenestration in the capillary endothelium seems to depend on VEGF secretion (Esser et al. 1998). Endothelial cells release in a paracrine fashion and express to the cell surface many signaling molecules that can affect the destiny of developing tissue cells intimately associated with them. The emerging scenario is that of a general developmental model whereby cross-talk between endothelial and tissue cells would be responsible for a series of sequential inductive and differentiating events. It has been speculated that endothelial cell–tissue interactions may “offer the opportunity to control organ development and growth systematically, rather then individually for each organ” (Lammert et al. 2003). Further investigations should be addressed to understand whether, upon adequate stimulation, the endothelium can be instructed to produce a series of mitogenic/survival factors that can protect parenchymal cells from injury and initiate regeneration. In the liver, for instance, the VEGF-driven paracrine supply of hepatocyte growth factor (HGF) by sinusoid endothelial cell represents a promising model for hepatocyte rescue and survival after toxic events (Le Couter et al., 2003). Indeed, HGF induction by VEGF or VEGFR-1 agonists
118
17
Napoleone Ferrara and the Saga of Vascular Endothelial Growth Factor
may form the basis of a future therapeutic scheme aimed towards protection of liver parenchyma in at least some liver disorders.
17.5 VEGF Expression in the Development of Pancreas Ferrara contributed to the study of the role of VEGF in the development of pancreas (Lammert et al., 2003). Deletion of VEGF in the mouse pancreas reveals that endocrine cells signal back to the adjacent endothelial cells to induce the formation of a dense network of fenestrated capillaries in islets. Moreover, glucose tolerance tests reveal that the VEGF-induced capillary network is not strictly required for blood glucose control but is essential for fine-tuning blood glucose regulation (Lammert et al., 2003).
17.6 VEGF Expression in Skeletal Growth and Endochondral Bone Formation Cartilage, an avascular tissue, is replaced by bone in a process named endochondral ossification. During this process, the epiphyseal growth plate undergoes morphogenesis. A region of resting chondrocytes differentiates into a zone of proliferating chondrocytes that then hypertrophies and finally undergoes apoptotic cell death while being replaced by bone. The net result is lengthening of the bone, whereas the thickness of the growth plate remains relatively constant. During this process, blood vessel invasion from the metaphysis coincides with mineralization of the extracellular matrix, apoptosis of hypertrophic chondrocytes, extracellular matrix degradation and bone formation. Ferrara and collaborators demonstrated that VEGF mRNA is expressed by hypertrophic chondrocytes in the epiphyseal growth plate, suggesting that a VEGF gradient is needed for directional growth and cartilage invasion by metaphyseal blood vessels (Gerber et al., 1999). Following VEGF blockade with a soluble VEGFR-1 chimaeric protein or an anti-VEGF monoclonal antibody, blood vessel invasion is almost completely suppressed, concomitant with impaired trabecular bone formation, in mice and primates (Gerber et al., 1999). Although proliferation, differentiation and maturation of chondrocytes were apparently normal, resorption of hypertrophic chondrocytes was inhibited, resulting in a marked expansion of the hypertrophic chondrocytes zone. These findings indicate that VEGF-dependent blood vessel invasion is essential for coupling cartilage resorption with bone formation. Following VEGF inactivation, hypertrophic chondrocytes fail to undergo apoptotic cell death and cessation of the anti-VEGF treatment is followed by capillary invasion, restoration of bone growth and normalization of the growth plate architecture.
17.8
Role of VEGF in Intracular Neovascular Syndromes
119
17.7 VEGF Expression in Ovary Angiogenesis is a key aspect of normal cyclical ovarian function. Ferrara’s laboratory demonstrated that VEGF mRNA expression is temporally and spatially related to the proliferation of blood vessels in the normal rat, mouse and primate ovary, suggesting that VEGF may be a mediator of the cyclical growth of blood vessels that occurs in the female reproductive tract (Phillips et al., 1990; Ravindranath et al., 1992). Administration of VEGF inhibitors suppresses luteal angiogenesis in a rat model of hormonally induced ovulation (Ferrara et al., 1998). This effect was associated with inhibition of corpus luteum development and progesterone release. Also, failure of maturation of the endometrium was observed, probably reflecting suppression of ovarian steroid production plus a direct inhibition of locally produced VEGF. Areas of ischemic necrosis were demonstrated in the corpus luteum of treated animals. However, no effects on the pre-existing ovarian vasculature were observed. These findings indicate that VEGF-mediated angiogenesis is essential for corpus luteum development and endocrine function. More recently, Ferrara’s laboratory demonstrated that EG-VEGF might play a cooperative role with VEGF in the regulation of angiogenesis in the human ovary (Le Couter et al., 2001). A sequential activation of the two genes occurs in the corpus luteum formation (Ferrara et al., 2003). Although VEGF is strongly expressed in early stage corpus luteum, its expression is reduced by mid-luteal phase. In contrast, EG-VEGF starts being expressed later than VEGF but persists through mid- and early–late luteal phase (Ferrara et al., 2003).
17.8 Role of VEGF in Intracular Neovascular Syndromes Ferrara’s laboratory contributed to the description of elevations of VEGF levels in aqueous and vitreous of eyes with proliferative retinopathy (Aiello et al., 1994). In a large series where ocular fluids from 165 patients were examined, a strong correlation was found between levels of immunoreactive VEGF in the aqueous and vitreous humours and active proliferative retinopathy (Aiello et al., 1994). VEGF levels were undetectable or very low (
