Handbook of Biomineralization

Handbook of Biomineralization Edited by Matthias Epple and Edmund Ba¨uerlein Handbook of Biomineralization. Edited by ...

42 downloads 1551 Views 37MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Handbook of Biomineralization Edited by Matthias Epple and Edmund Ba¨uerlein

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

Handbook of Biomineralization Medical and Clinical Aspects

Edited by Matthias Epple and Edmund Ba¨uerlein

The Editors Prof. Dr. Matthias¨Epple University of Duisburg-Essen Inorganic Chemistry Universitätsstr. 2 45141 Essen Germany

Prof. Dr. Edmund Ba¨uerlein Max-Planck-Institute for Biochemistry Department of Membrane Biochemistry Am Klopferspitz 18 A 82152 Planegg Germany ¨ Cover Illustration (designed by Felix Bäuerlein) (Top right, Bottom left and Bottom right designed by Felix Baeuerlein) Top left: Heterodont molar of carnivorous mammals (wolf, no functional wear) with no exposed dentin and very small pulp chamber. (P. Gaengler, W. H. Arnold, Chap. 14. Fig. 14.1c). Top right: Apatite formed on TiO2 gel in simulated body fluid (SBF). (T. Kokubo, H. Takadama. Chap. 7. Fig. 7.4(2)). Bottom left: An implant manufactured by hot pressing and gas flushing for cranial reconstruction with gradients in composition and spatially different porosity. (M. Epple, Chap. 6) Bottom right: Calcified lung metastases of a primary colorectal adenocarcinoma. Light microscopic image, HE-stain, metastases with typical structure of colon (*). typical lung structure is not present, ossification (**). (Inge Schmitz. Chap. 18, Fig. 18.7) Handbook of Biomineralization Biological Aspects and Structure Formation: ISBN 978-3-527-31804-9 Biomimetic and Bioinspired Chemistry: ISBN 978-3-527-31805-6 Medical and Clinical Aspects: ISBN 978-3-527-31806-3 Set (3 volumes): ISBN 978-3-527-31641-0

9 All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.: applied for British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Bibliographic information published by the Deutsche Nationalbibliothek Die Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliograﬁe; detailed bibliographic data are available in the Internet at hhttp://dnb.d-nb.dei. 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microﬁlm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not speciﬁcally marked as such, are not to be considered unprotected by law. Printed in the Federal Republic of Germany Printed on acid-free paper Typesetting Asco Typesetters, Hong Kong Printing betz-druck GmbH, Darmstadt Binding Litges & Dopf GmbH, Heppenheim Wiley Bicentennial Logo Richard J. Paciﬁco ISBN 978-3-527-31806-3

V

Contents Preface Foreword

XVII XIX

List of Contributors

XXI

Part I

Bone 1

1

Mineralization of Bone: An Active or Passive Process? Thorsten Schinke and Michael Amling

1.1 1.2 1.3 1.4 1.5 1.6

2

2.1 2.2 2.3 2.4 2.5 2.6

3

Abstract 3 Physiological and Pathological Mineralization 3 Inhibitors of Pathological Mineralization 5 Activators of Physiological Mineralization 6 The Key Role of Pyrophosphate 9 The Mysterious Role of the Endopeptidase Phex 11 Concluding Remarks 13 References 14 Bone Morphogenetic Proteins 19 Walter Sebald, Joachim Nickel, Axel Seher, and Thomas D. Mu¨ller

Abstract 19 Introduction 19 What is a Bone Morphogenetic Protein? 21 BMP Receptors are Composed of Diverse Type I and Type II Receptor Chains 23 The Basic Signaling Mechanism is the Same for BMPs and other TGF-b-like Proteins 24 Biochemistry and Cell Biology of Receptor Speciﬁcity 25 Structural Basis for Speciﬁcity and Aﬃnity of BMP Receptor Interaction 27

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

VI

Contents

2.7

What We Can Do with BMPs: The Engineering of BMP-2 and GDF-5 Variants 29 References 32

3

Biomechanics of Bones: Modeling and Computation of Bone Remodeling 35 Udo Nackenhorst

3.1 3.2 3.3 3.3.1 3.3.2 3.3.3 3.3.4 3.4

4

4.1 4.2 4.2.1 4.2.2 4.3 4.3.1 4.3.2 4.3.3 4.4 4.5

5

5.1 5.1.1 5.1.1.1 5.1.1.2

Abstract 35 Introduction 35 The Biomechanical Equilibrium Approach 36 A Computational Multi-Scale Approach for Cortical Bone 42 Closed Nano-to-Meso Control Circuit Approach 43 Sub-Cellular Length-Scale 44 Micro-Scale Model (Single Osteon) 45 Meso-Scale Model of Cortical Bone 45 Conclusions 46 References 47 Direct X-Ray Scattering Measurement of Internal Stresses and Strains in Loaded Bones 49 Stuart R. Stock and Jonathan D. Almer

Abstract 49 Introduction 49 Background 50 X-Ray Scattering 50 Strains and Stresses 51 Methods 51 Specimens and Geometry 51 Analysis of Two-Dimensional (2-D) Scattering Patterns 53 X-Ray Elastic Constants and Strain–Stress Conversion 55 Examples of Data and Analysis 55 Discussion and Future Directions 56 References 57 Osteoporosis and Osteopetrosis Adele L. Boskey

59

Abstract 59 Introduction: Two Distinct Diseases with Common Features Comparisons of Clinical Features of Osteoporosis and Osteopetrosis 60 Histology 60 Radiography 61

59

Contents

5.1.2 5.2 5.2.1 5.2.1.1 5.2.1.2 5.2.2 5.2.2.1 5.2.2.2 5.3 5.3.1 5.3.2 5.4

6

6.1 6.2 6.3 6.4 6.4.1 6.4.2 6.5 6.6 6.7 6.8 6.9 6.10

7

7.1 7.2 7.3 7.4

Comparisons of Bone Mineral Properties in Osteoporosis and Osteopetrosis 62 Animal Models of Osteoporosis and Osteopetrosis 63 Osteoporosis 63 Rodent Models 63 Non-Rodent Models 66 Osteopetrosis 66 Rodent Models 67 Other Osteopetrotic Models 68 The Cellular and Molecular Bases of Osteopetrosis and Osteoporosis 70 Osteoporosis 70 Osteopetrosis 72 Biomineralization in Osteopetrosis and Osteoporosis 74 References 75 Biomimetic Bone Substitution Materials Matthias Epple

81

Abstract 81 The Clinical Need for Bone Substitution Materials 81 Synthetic Materials Used for Bone Substitution 82 Ceramics and Bone Cements 84 Polymers 86 PMMA-Based Materials 86 Polyester-Based Materials 86 Metals 87 Composites 87 Bone Substitutes of Biological Origin 87 Biological Functionalization of Synthetic Materials 90 An Example of a Synthetic Biomimetic Bone Substitution Material 90 Conclusions and Future Developments 91 References 92 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants 97 Tadashi Kokubo and Hiroaki Takadama

Abstract 97 Introduction 97 Qualitative Correlation of Bone-Bonding Bioactivity of a Material with Apatite Formation on its Surface in SBF 98 Quantitative Correlation of Bone-Bonding Bioactivity and ApatiteForming Ability in SBF 100 Ion Concentrations of SBF 101

VII

VIII

Contents

7.5 7.6 7.7 7.8 7.9

Materials Able to Form Apatite 102 Composition and Structure of Apatite 103 Mechanism of Bonding of Bioactive Material to Bone 104 Mechanisms of Apatite Formation 105 Summary 106 References 106

8

Stimulation of Bone Growth on Implants by Integrin Ligands Mo´nica Lo´pez-Garcı´a and Horst Kessler

8.1 8.1.1 8.1.2 8.1.3 8.1.4 8.2 8.2.1 8.2.2 8.2.3 8.2.4 8.2.4.1 8.2.4.2 8.2.4.3 8.2.4.4 8.3

9

9.1 9.2 9.2.1 9.2.2 9.2.3 9.2.4 9.2.5

109

Abstract 109 Introduction 109 Biomimetic Materials for Implant Technology 109 Integrins and RGD Sequence 110 Natural Proteins or Synthetic Peptides as Cell-Adhesive Molecules? 111 Integrin-Mediated Cell Adhesion 112 Improvements in Implant-Osseointegration by Surface Modiﬁcation with Integrin Ligands 115 Mechanisms of Bone Grafting 115 Modiﬁcations on Implant Surfaces to Improve its Osseointegration 116 Structure of the Coating Molecules 117 Stimulation of Osteoblasts Adhesion and Proliferation on Implants Promoted by Integrin Ligands 118 Poly(methyl methacrylate) 118 Silks 120 Titanium 120 RGD Mimetics 121 Conclusions 123 References 123 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles from Titanium and other Materials 127 Fumio Watari, Kazuchika Tamura, Atsruro Yokoyama, Kenichiro Shibata, Tsukasa Akasaka, Bunshi Fugetsu, Kiyotaka Asakura, Motohiro Uo, Yasunori Totsuka, Yoshinori Sato, and Kazuyuki Tohji

Abstract 127 Introduction 128 Materials and Methods 128 Specimens 128 Dissolution Testing of Ti Particles 129 Probe Cells 129 Biochemical Analyses of Cellular Reactions to Materials Animal Experiments 129

129

Contents

9.3 9.3.1 9.3.2 9.3.2.1 9.3.2.2 9.3.2.3 9.3.2.4 9.3.2.5 9.3.3 9.3.4 9.3.5 9.3.6 9.3.6.1 9.3.6.2 9.3.6.3 9.4 9.4.1 9.4.2 9.4.3 9.4.4 9.4.5

10

10.1 10.2 10.2.1 10.2.2 10.2.3 10.2.4 10.2.5 10.2.6 10.2.7 10.2.7.1 10.2.7.2 10.3 10.3.1 10.3.2

Results 130 Dependence of Tissue Reaction In Vivo on Material Macroscopic Size 130 Eﬀect of Particle Size on Biocompatibility 130 Size Distribution of the Abraded Particles 130 Particle Size Dependence In Vitro 131 Particle Size Dependence In Vivo 133 Material Dependence of the Particle Size Eﬀect In Vitro 135 Material Dependency of Tissue Reaction to Particles In Vivo 135 Shape Eﬀect 136 The Origin of the Particle Size Eﬀect 137 Toxicity Level of Particle Size Eﬀect for Bioactive and Bioinert Materials 138 Nanotoxicology 139 Size-Dependent Stimulus Down to Nanometer Size 139 Internal Diﬀusion of Nanoparticles 139 Toxicity-Enhancing Eﬀects of Biostimulatory Materials by Nanosizing 140 Discussion 140 Particle Size-Dependence of Cytotoxicity 140 Particle Size-Dependence in Soft Tissues 141 Comparison of Ti, Fe, and Ni Particles 141 The Eﬀect of Micro-/Nanosizing on Biological Reactions 142 Terminology on ‘‘Nanotoxicology’’ 143 References 143 Tissue Engineering of Bone 145 Hans-Peter Wiesmann, Beate Lu¨ttenberg, and Ulrich Meyer

Abstract 145 Tissue Engineering: What Does it Mean? 145 Components of Bone Tissue Engineering 147 Osteoblasts 147 Bone Marrow Cells 148 Marrow-Derived Stem Cells 148 Vascular Cells 149 Scaﬀold Design and Cell Compatibility 149 Bioreactors 150 In-Vitro Cell Stimulation 150 Biophysical Stimulation 150 Biochemical Stimulation 151 Bone Biomineralization in Tissue Engineering Ex Vivo and In Vivo 151 Principles of ECM Biomineralization 151 Principles of Bone Formation 152

IX

X

Contents

10.3.3 10.4 10.5

Particular Features of Extracorporeal Biomineralization Clinical Demands 153 Future Aspects 154 References 155

Part II

Teeth

11

Formation of Teeth 159 Katharina Reichenmiller and Christian Klein

11.1 11.2 11.2.1 11.2.2 11.3 11.3.1 11.3.2 11.3.3 11.4 11.5 11.5.1 11.5.2 11.5.3 11.5.4 11.6

12

12.1 12.2 12.3 12.4 12.5

13

13.1 13.2

153

157

Abstract 159 Introduction 159 Odontogenesis 163 Genes Involved in Odontogenesis 165 Stem Cells 165 Dentinogenesis 165 Mantle and Circumpulpal Dentin 166 Intertubular Dentin 168 Peritubular Dentin 168 Amelogenesis 170 Cementogenesis 172 Acellular Extrinsic Fiber Cementum (AEFC) 172 Cellular Intrinsic Fiber Cementum (CIFC) 173 Cellular Mixed Stratiﬁed Cementum (CMSC) 174 Acellular Intrinsic Fiber Cementum (AIFC) 174 Acknowledgments 174 References 174 The Structure of Teeth: Human Enamel Crystal Structure Fre´de´ric Cuisinier and Colin Robinson

177

Abstract 177 Introduction 177 HRTEM Observations 178 AFM Observations 179 Discussion 181 Conclusions 182 References 182 Design Strategies of Human Teeth: Biomechanical Adaptations Paul Zaslansky and Steve Weiner

Abstract 183 Introduction 183 Deformation of Whole Teeth under Load

185

183

Contents

13.3 13.4 13.5 13.6

Mechanical Behavior of the Enamel Cap 191 The Role of Crown Dentin During Load Bearing 194 The Role of the Root and Supporting Structures 196 Broader Implications and Conclusions 198 References 200

14

Clinical Aspects of Tooth Diseases and their Treatment Peter Ga¨ngler and Wolfgang H. Arnold

14.1 14.2 14.2.1 14.2.2 14.3 14.4 14.5 14.5.1 14.5.2

15

15.1 15.2 15.3 15.4 15.4.1 15.4.2 15.5 15.5.1 15.5.2

16

16.1 16.2 16.2.1 16.2.2

203

Abstract 203 Introduction 203 Tooth Development and Developmental Anomalies 206 Developmental Features and Elemental Analysis of Early Mineralization 207 Developmental Anomalies 210 Dental Caries 212 Periodontal Diseases 216 Dental Trauma 220 Acute Dental Trauma 220 Chronic Dental Trauma 220 References 221 Dental Caries: Quantifying Mineral Changes Susan M. Higham and Philip W. Smith

223

Abstract 223 Introduction 223 Enamel Caries 224 Dentine Caries 224 Analyzing Mineral Changes in Dental Caries 225 Transverse Microradiography 226 TMR Studies 227 Quantitative Light-Induced Fluorescence 229 In Vitro QLF Studies 231 In Vivo QLF Studies 234 References 236 Periodontal Regeneration 239 Hom-Lay Wang and Lakshmi Boyapati

Abstract 239 Deﬁnitions 239 Periodontal Wound Healing 240 Wound-Healing Principles 240 Compartmentalization 241

XI

XII

Contents

16.2.3 16.3 16.3.1 16.3.1.1 16.3.2 16.3.2.1 16.3.2.2 16.3.2.3 16.3.2.4 16.3.3 16.3.3.1 16.3.3.2 16.3.4 16.3.4.1 16.3.4.2 16.3.4.3 16.3.4.4 16.4 16.4.1 16.4.2 16.4.3 16.4.4 16.5 16.5.1 16.5.2 16.5.3 16.5.4 16.6

Evaluating Regeneration 241 Techniques Used for Regeneration 241 Root Surface Biomodiﬁcation 241 Root Surface Conditioning 241 Bone Replacement Grafts 242 Autografts 243 Allografts 243 Xenografts 245 Alloplasts 245 Guided Tissue Regeneration 246 Non-Absorbable Membranes 246 Absorbable Membranes 246 Biologic Modiﬁers 248 Growth Factors/Cytokines 248 Bone Morphogenetic Proteins (BMPs) 248 Pep-Gen p-15 249 Enamel Matrix Derivative (EMD) 249 Factors Inﬂuencing GTR Success 249 Patient Factors 251 Defect/Local Factors 251 Treatment Factors 252 Postoperative Care 252 Surgical Principles 253 Furcation Defects 253 Intrabony Defects 253 Root Coverage 253 Surgical Techniques 255 Conclusions 258 References 258

17

Tissue Engineering of Teeth Misako Nakashima

17.1 17.2 17.2.1 17.2.1.1 17.2.1.2 17.2.1.3 17.2.2 17.2.3 17.3 17.3.1

265

Abstract 265 Introduction 265 The Triad 266 Pulp Stem/Progenitor Cells 266 Isolation 266 Self-Renewal 268 Multipotential Diﬀerentiation 269 Morphogenetic Signals, BMPs 269 Scaﬀold 270 Dentin Regeneration 271 Protein Therapy 271

Contents

17.3.2 17.3.2.1 17.3.2.2 17.4 17.4.1 17.4.2 17.5 17.6

Gene Therapy 272 In-Vivo BMP Gene Therapy 272 Ex-Vivo BMP Cell Therapy and Gene Therapy 273 Pulp Regeneration 276 Vasculogenesis 276 Neurogenesis 276 Whole-Teeth Regeneration 277 Conclusions and Future Perspectives 278 References 278

Part III

Pathological Calciﬁcations

18

Aspects of Pathological Calciﬁcations Inge Schmitz

18.1 18.1.1 18.1.2 18.2 18.2.1 18.2.2 18.2.2.1 18.3 18.3.1 18.3.1.1 18.3.1.2 18.3.2 18.3.3 18.4 18.4.1 18.4.2 18.5

19

19.1 19.2 19.2.1 19.2.2 19.2.3

283 285

Abstract 285 Introduction 285 Examples of Pathological Calciﬁcation 286 Regulation of Calciﬁcations 287 Heterotopic Ossiﬁcation 288 Calciﬁcation in Ulcera of Patients with Paraplegia 288 Calciﬁcations of the Lung 289 Metastatic Pulmonary Calciﬁcations 291 Vascular Calciﬁcations: Arteriosclerosis 291 Calciﬁcations of Arteries 291 Calciﬁcation of the Tunica Media (Mo¨nckeberg’s Arteriosclerosis) 292 Calciﬁcation of the Tunica Intima (Arteriosclerosis) 293 Ossiﬁcations of Arteries 294 Characterization of Atherosclerotic Plaques of the Human Aorta 294 Calciﬁcation of Synthetic Vascular Grafts 296 Chronic Kidney Disease-Dialysis and Vascular Calciﬁcation of Arteries and Arteriovenous Shunts 296 Ossiﬁcation of Synthetic Grafts 298 Conclusions 299 References 299 Atherosclerosis: Cellular Aspects 301 Diane Proudfoot and Catherine M. Shanahan

Abstract 301 Introduction 301 Role of VSMCs in Vascular Calciﬁcation 303 Release of Apoptotic Bodies and Vesicles 303 Phagocytosis 305 VSMC Osteo/Chondrocytic Conversion 306

XIII

XIV

Contents

19.2.4 19.3 19.3.1 19.3.2 19.4 19.5

20

20.1 20.2 20.3 20.4

21

Role of Calcifying Vascular Cells and Pericytes 309 Role of Inﬂammatory Cells 310 Macrophages 310 Dendritic Cells, Mast Cells and T Lymphocytes 312 The Role of Osteoclasts: Is there a Possibility for CalciﬁcationRegression? 312 Conclusions 313 References 313 The Biological and Cellular Role of Fetuin Family Proteins in Biomineralization 317 Cora Scha¨fer and Willi Jahnen-Dechent

Abstract 317 Osteogenesis and Bone Mineralization versus Calciﬁcation 317 a2 -HS Glycoprotein/Fetuin-A is a Systemic Inhibitor of Ectopic Calciﬁcation 320 The Mechanism of Fetuin-A Inhibition of Calciﬁcation 322 The Fate of Calciprotein Particles 322 References 325 Stone Formation 329 Pierfrancesco Bassi

Abstract 329 Urinary Stones 329 Pathogenesis 330 Inhibitors of Stone Formation 331 Classiﬁcation of Urinary Stones 333 Calcium Stones 333 Uric Acid Stones 336 Magnesium Ammonium Phosphate Stones, Struvite or Infection Stones 337 21.1.2.4 Cystine Stones 338 21.1.3 Risk Factors 338 21.1.3.1 Non-Genetic Factors 338 21.1.3.2 Genetic Factors 341 21.2 Other Urological Stones: Testicular Microlithiasis 343 21.3 Biliary and Gallbladder Stones 343 21.4 Miscellaneous Stones 344 21.4.1 Sialolithiasis 344 21.4.2 Dental Stones 344 21.4.3 Pancreatic Stones 345 21.4.4 Broncholithiasis and Pulmonary Alveolar Microlithiasis 345 References 346 21.1 21.1.1 21.1.1.1 21.1.2 21.1.2.1 21.1.2.2 21.1.2.3

Contents

22

Ectopic Mineralization: New Concepts in Etiology and Regulation Cecilia M. Giachelli

349

Abstract 349 Introduction 349 Regulators of Ectopic Mineralization 350 Circulating Factors that Regulate Ectopic Mineralization 350 Ion Transporters and Homeostatic Enzymes that Regulate Ectopic Mineralization 352 22.2.2.1 Role of Sodium-Dependent Phosphate Co-Transporters in Ectopic Mineralization 353 22.2.3 Extracellular Matrix Molecules that Regulate Ectopic Mineralization 354 22.2.3.1 Role of Osteopontin in Ectopic Mineralization 355 22.2.4 Cell Signaling Pathways that Regulate Ectopic Mineralization 355 22.2.5 Roles of Cell Death and Bone Remodeling in Ectopic Mineralization 357 22.3 Conclusions and Implications for Therapeutic Control of Ectopic Mineralization 358 References 358 22.1 22.2 22.2.1 22.2.2

23

23.1 23.2 23.3 23.4 23.5 23.6 23.7 23.8

24

24.1 24.2 24.3 24.4 24.4.1

Pathological Calciﬁcation of Heart Valve Bioprostheses Birgit Glasmacher and Martin Krings

361

Abstract 361 Introduction 361 In-Vitro Calciﬁcation Models 364 Heart Valve Bioprostheses 364 Calciﬁcation Hypotheses and Study Design 364 Calciﬁcation Imaging Methods 365 Calciﬁcation Patterns 367 Description of Findings 369 Conclusions and Future Research 370 References 371 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials, Tissue Engineering and Biomineralization 373 Pedro L. Granja, Jose´ Paulo Pereira, and Ma´rio A. Barbosa

Abstract 373 The Internet as a Major Healthcare Resource 373 Impact of Biomaterials Science in Modern Society 375 Biomat.net as a Biomineralization Resource 376 The Biomaterials Network (Biomat.net) 383 An Overview 383

XV

XVI

Contents

24.4.2 24.4.3 24.4.4 24.4.4.1 24.4.4.2 24.4.4.3 24.4.4.4 24.4.4.5 24.4.4.6 24.4.4.7

Objectives 384 Team 385 Functionalities 386 Site Map 387 Membership 387 Links Lists 387 Directory of Researchers 387 Jobs 388 Newsletter 388 Endorsement of Scientiﬁc Meetings References 389 Index 391

389

XVII

Preface When I began my academic career in Hamburg, I worked on the solid-state chemistry of organic compounds. By chance, our group identiﬁed a solid-state reaction which led to a polymer (polyglycolide), a class of biodegradable polyesters which a literature survey showed to have many applications in clinical medicine. This ﬁnding led in turn to my ﬁrst contacts with physicians and their ideas about implants, biomaterials, and bone. As a member of a graduate school on bone in the medical faculty, I learned that biomaterials usually depend on biomineralization – that is, the cellular action and similarity to the corresponding hard tissue. I also became aware for the ﬁrst time of the great importance of pathological calciﬁcations in our society. The process of biomineralization is based on the formation of inorganic crystals, and is strongly controlled by organic molecules which are themselves controlled by cells. As a chemist, I was familiar with the ﬁrst part of this story, but I had much to learn about biomolecules, the diﬀerent cell types, and also the clinical treatments of diseases. Fortunately, in Hamburg, as well as in the subsequent institutions where I worked – at Bochum and Essen – I met people from the ﬁeld of medicine who were interested in conducting joint studies. When all parties had learned to communicate – due to the diﬀerent vocabulary of their disciplines – many fruitful collaborations developed, and in this respect I am deeply indebted to my colleagues (listed in alphabetical order), Prof. M. Amling, Prof. G. Delling, Prof. S. A. Esenwein, Prof. H. Euﬁnger, Rof. P. Ga¨ngler, Prof. W. JahnenDechent, Dr. A. Klocke, Prof. M. Ko¨ller, Dr. P. Lanzer, Dr. W. Linhart, Prof. G. Muhr, Prof. J. M. Rueger, Prof. W. Ruether, Dr. I. Schmitz, and Dr. S. Weihe, from whom I learned much about the clinical aspects of biomineralization. Therefore, I was very excited when Edmund Ba¨uerlein asked me jointly to edit a volume of the Handbook of Biomineralization. We quickly selected the three main topics which are relevant in medicine, namely bone, teeth, and pathological calciﬁcations, and also were fortunate to ﬁnd many competent authors from all over the world who agreed to contribute. Many aspects of biomineralization in medicine are highlighted in this book, and I sincerely hope that it will contribute to our understanding of this ﬁeld of research, as it is not only of academic inter-

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

XVIII

Preface

est but also covers many general aspects of biomineralization. Indeed, it can be said that good health is dependent on correct biomineralization! Essen, Germany February 2007

Matthias Epple

XIX

Foreword During the past decade, biomaterials research has undergone tremendous change. Ten years ago, the ﬁrst dialogues were made between physicians searching for artiﬁcial materials that could be used for regenerative therapies (for example, in bone surgery), and materials scientists oﬀering a variety of structural and/or functional materials which originally were developed for engineering applications. Today, we can take advantage of the impressive advances in modern biology and biochemistry, and as a consequence we have the chance to follow completely new pathways for solving such problems. As will be shown in this volume, our current understanding of biomineralization opens up new approaches not only for physicians but also for materials scientists in many medical applications. The key here is provided by the exploration of genetically controlled mechanisms of the biomineralization of hard tissues. The activation and inhibition of biomineralization are two complementary processes which lead to such wonderful structures as bone or teeth, and based on the phenotypic analyses of several mouse models and various diseases causing calciﬁcation of soft tissues, we have learned that there is a variety of noncollagenous proteins that control this interplay of activation and inhibition. The following examples will indicate how this knowledge can be used for biomimetic implant development. As the overall age of our modern-day society continues to rise, bone diseases will become increasingly important, and consequently the successful treatment of the pathological mineralization of bone – for example, in the case of osteoporosis – represents one of the major challenges of the next decade. This general problem in bone research is discussed in Chapter 5, wherein it is clear that, based on an understanding of biomineralization in such pathological situations, new strategies could be derived for biomedical treatments as well as for new materials that may be used in the regenerative therapy of disturbed tissue. Today, although a large variety of bone substitution materials is applied on a practical basis, an evolutionary process can be foreseen in which a group of artiﬁcial engineering materials will be completed by more biologically functionalized materials. Here, one favored strategy is the stimulation of bone growth on implants, and today the development of scaﬀolds suited to the immobilization of

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

XX

Foreword

living bone cells has been boosted by the recent progress made in stem cell research. Another exciting feature relating to the biomineralization of bone and teeth is the formation of hierarchical or graded structures. This leads to basic questions concerning structural evolution on the mesoscopic scale. The guiding principle for understanding such structures is related to the biomechanical adaptation of living tissue. In Chapter 13, the author explains how, in the case of teeth, evolutionary pressure has led to the creation of highly optimized structures with respect to the complex mechanical loading situations in a living organism. Based on this theory, many attempts have been made to model such structures, whereby the models describe the formation of the mineralized tissue by combining cellular activity with acting mechanical stresses. However, the predictive power of such numerical simulations based on Finite Element codes remains limited. In particular, the uncertainty of the constitutive laws for materials behavior sets such restrictions, and consequently new experimental approaches are required that will allow the appropriate measurement of the properties of these materials on the micro- and mesoscale. Today, there are growing numbers of promising methods available to perform just this task, and this is demonstrated throughout this volume. As mentioned above, the unwanted pathological calciﬁcation of soft tissue or of vascular systems is an issue which is closely connected with the biomineralization of hard tissues. Today, we know that there are no signiﬁcant diﬀerences between both phenomena, and thus another broad ﬁeld of research activity is opening up with relevance for biological tissue characterization as well as for the development of artiﬁcial materials, for example in vascular prostheses. In the ﬁnal part of this volume, we show that this phenomenon is also dominated by active cell-mediated processes and not by the simple precipitation of minerals at a given substrate. In summarizing, it can be said that the exciting interdisciplinary cooperation of biologists, biochemists, materials scientists, and physicians has led us to a challenging new research ﬁeld of biomineralization, with wonderful perspectives for a better understanding of the beauty of the evolution of living organisms, whilst at the same time making signiﬁcant contributions to human healthcare. Prof. Dr. rer. nat Wolfgang Pompe Technische Universita¨t Dresden Institut fu¨r Werkstoﬀwissenschaft Professur fu¨r Materialwissenschaft und Nanotechnik Dresden Germany

XXI

List of Contributors Tsukasa Akasaka Graduate School of Dental Medicine Hokkaido University Kita 13, Nishi 7 Kita-Ku, Sapporo Japan

Kiyotaka Asakura Hokkaidou University Catalysis Research Center Kita 21, Nishi 10 Kita-ku, Sapporo Japan

Jonathan D. Almer Building 431 Advanced Photon Source Argonne National Laboratory 9700 South Cass Ave. Argonne, IL 60439 USA

Ma´rio A. Barbosa BIOMATERIALS NETWORK (Biomat.net) Biomaterials Laboratory INEB (Instituto de Engenharia Biomedica) University of Porto 823, Rua di Campo Alegre 4150-180 Porto Portugal

Michael Amling Clinics of Trauma-, Hand-, and Reconstruction Surgery University Medical Center Hamburg-Eppendorf Martinistraße 52 20246 Hamburg Germany Wolfgang H. Arnold Abteilung fu¨r Konservierende Zahnheilkunde Fakulta¨t fu¨r Zahn-, Mund- und Kieferheilkunde Universita¨t Witten/Herdecke Alfred-Herrhausen-Straße 50 58448 Witten Germany

Pierfrancesco Bassi Universita` Cattolica del Sacro Cuore Policlinico Universitario ‘A. Gemelli’ Largo A. Gemelli 00168 Rome Italy Adele L. Boskey Hospital for Special Surgery Weill Medical College of Cornell University 535 East 70th Street New York, NY 10021 USA

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

XXII

List of Contributors

Lakshmi Boyapati School of Dentistry Department of Periodontics and Oral Medicine University of Michigan 1011 North University Avenue Ann Arbor MI 45109-1078 USA Fre´de´ric Cuisinier UFR Odontologie 545 Avenue du Professeur Jean-Louis Viala 34193 Montpellier Cedex 5 France Matthias Epple Institute of Inorganic Chemistry University of Duisburg-Essen Universita¨tsstraße 5–7 45117 Essen Germany Bunshi Fugetsu Graduate School of Environmental Science Hokkaido University Kita 10, Nishi 5 Kita-Ku, Sapporo Japan Peter Ga¨ngler Abteilung fu¨r Konservierende Zahnheilkunde Fakulta¨t fu¨r Zahn-, Mund- und Kieferheilkunde Universita¨t Witten/Herdecke Alfred-Herrhausen-Straße 50 58448 Witten Germany

Cecilia M. Giachelli Bioengineering Foege Hall Box 35506-1 University of Washington 1705 NE Paciﬁc Street Seattle, WA 98195 USA Birgit Glasmacher Institute for Multiphase Processes and Center for Biomedical Engineering Gottfried Wilhelm Leibniz University Hannover Callinstraße 36 30167 Hannover Germany Pedro L. Granja BIOMATERIALS NETWORK (Biomat.net) Biomaterials Laboratory INEB (Instituto de Engenharia Biomedica) University of Porto 823, Rua di Campo Alegre 4150-180 Porto Portugal Susan M. Higham University of Liverpool Department of Clinical Dental Sciences Edwards Building Daulby Street Liverpool L69 3GN United Kingdom Willi Jahnen-Dechent Dept. of Biomedical Engineering Biointerface Laboratory RWTH Aachen University Hospital Pauwelsstraße 30 52074 Aachen Germany

List of Contributors

Horst Kessler Department of Chemistry Institut fu¨r Organische Chemie und Biochemie Technische Universita¨t Mu¨nchen Lichtenbergstr. 4 85747 Garching Germany Christian Klein School of Dental Medicine Department of Operative Dentistry and Periodontology Osianderstraße 2–8 72076 Tu¨bingen Germany Tadashi Kokubo College of Life and Health Sciences Dept. of Biomedical Sciences Chubu University 1200 Matsumoto-cho Kasugai city, Aichi 487-8501 Japan Martin Krings Institute for Multiphase Processes and Center for Biomedical Engineering Gottfried Wilhelm Leibniz University Hannover Callinstraße 36 30167 Hannover Germany Mo´nica Lo´pez-Garcı´a Department of Chemistry Institut fu¨r Organische Chemie und Biochemie Technische Universita¨t Mu¨nchen Lichtenbergstr. 4 85747 Garching Germany

¨ttenberg Beate Lu Department of Cranio-Maxillofacial Surgery University of Mu¨nster Waldeyerstr. 30 48149 Mu¨nster Germany Ulrich Meyer Clinic for Maxillofacial and Plastic Facial Surgery University of Du¨sseldorf Moorenstr. 5 40225 Du¨sseldorf Germany ¨ller Thomas Dieter Mu Universita¨t Wu¨rzburg, Biozentrum Physiologische Chemie II Am Hubland 97074 Wu¨rzburg Germany Udo Nackenhorst Institut fu¨r Baumechanik und Numerische Mechanik (IBNM) International Center for Computational Engineering Sciences (ICCES) Appelstraße 9A 30167 Hannover Germany Misako Nakashima Laboratory for Oral Disease Research National Institute for Longevity Sciences National Center of Geriatry and Gerontology 36-3 Gengo, Morioka, Obu Aichi 474-8522 Japan

XXIII

XXIV

List of Contributors

Joachim Nickel Universita¨t Wu¨rzburg, Biozentrum Physiologische Chemie II Am Hubland 97074 Wu¨rzburg Germany Jose´ Paulo Pereira BIOMATERIALS NETWORK (Biomat.net) Biomaterials Laboratory INEB (Instituto de Engenharia Biomedica) University of Porto 823, Rua di Campo Alegre 4150-180 Porto Portugal Diane Proudfoot University of Cambridge Division of Cardiovascular Medicine ACCI Building Level 6, Box 110 Addenbrooke’s Hospital Hills Road Cambridge CB2 2QQ United Kingdom Katharina Reichenmiller School of Dental Medicine Department of Operative Dentistry and Periodontology Osianderstraße 2–8 72076 Tu¨bingen Germany Colin Robinson Leeds Dental Institute Clarendon Way Leeds LS29LU United Kingdom

Yoshinori Sato Graduate School of Environmental Studies Tohoku University 6-6-20, Aramaki Aza Aoba Aoba-ku, Sendai Miyagi Japan Cora Scha¨fer Dept. of Biomedical Engineering Biointerface Laboratory RWTH Aachen University Hospital Pauwelsstraße 30 52074 Aachen Germany Thorsten Schinke Clinics of Trauma-, Hand- and Reconstruction Surgery University Medical Center HamburgEppendorf Martinistraße 52 20246 Hamburg Germany Inge Schmitz Institute of Pathology and German Mesothelioma Register Ruhr University Bochum Bergmannsheil Clinic Buerkle-de-la-Camp-Platz 1 44789 Bochum Germany Walter Sebald Universita¨t Wu¨rzburg, Biozentrum Physiologische Chemie II Am Hubland 97074 Wu¨rzburg Germany

List of Contributors

Axel Seher Universita¨t Wu¨rzburg, Biozentrum Physiologische Chemie II Am Hubland 97074 Wu¨rzburg Germany

Hiroaki Takadama College of Life and Health Sciences Deptartment of Biomedical Sciences Chubu University 1200 Matsumoto-cho Kasugaicity, Aichi 487-8501 Japan

Catherine M. Shanahan University of Cambridge Division of Cardiovascular Medicine ACCI Building Level 6, Box 110 Addenbrooke’s Hospital Hills Road Cambridge CB2 2QQ United Kingdom

Kazuchika Tamura Graduate School of Dental Medicine Hokkaido University Kita 13, Nishi 7 Kita-Ku, Sapporo Japan

Kenichiro Shibata Graduate School of Dental Medicine Hokkaido University Kita 13, Nishi 7 Kita-Ku, Sapporo Japan Philip W. Smith University of Liverpool Department of Dental Sciences Pembroke Place Liverpool L3 5PS United Kingdom Stuart R. Stock Department of Molecular Pharmacology and Biological Chemistry Mail code S-215 Northwestern University 303 East Chicago Avenue Chicago, IL 60611-3008 USA

Kazuyuki Tohji Graduate School of Environmental Studies Tohoku University 6-6-20, Aramaki Aza Aoba Aoba-ku, Sendai Miyagi Japan Yasunori Totsuka Graduate School of Dental Medicine Hokkaido University Kita 13, Nishi 7 Kita-Ku, Sapporo Japan Motohiro Uo Graduate School of Dental Medicine Hokkaido University Kita 13, Nishi 7 Kita-Ku, Sapporo Japan

XXV

XXVI

List of Contributors

Hom-Lay Wang School of Dentistry Department of Periodontics and Oral Medicine University of Michigan 1011 North University Avenue Ann Arbor, MI 45109-1078 USA Fumio Watari Graduate School of Dental Medicine Hokkaido University Kita 13, Nishi 7 Kita-Ku, Sapporo Japan Steve Weiner Department of Structural Biology Faculty of Chemistry Weizmann Institute of Science Kimmelman Building (13) Rehovot 76100 Israel

Hans-Peter Wiesmann Department of Cranio-Maxillofacial Surgery University of Mu¨nster Waldeyerstr. 30 48149 Mu¨nster Germany Atsruro Yokoyama Graduate School of Dental Medicine Hokkaido University Kita 13, Nishi 7 Kita-Ku, Sapporo Japan Paul Zaslansky Dept. Biomaterials Max-Planck-Institute of Caloies and Interfaces Wissenschaftspark Golm Ann Mu¨hlenburg 1 14476 Golm Germany

Part I Bone

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

3

1 Mineralization of Bone: An Active or Passive Process? Thorsten Schinke and Michael Amling

Abstract

The most speciﬁc feature of bone tissue is the presence of a mineralized extracellular matrix (ECM) produced by osteoblasts. However, compared to our steadily increasing knowledge on the regulation of osteoblast proliferation and diﬀerentiation, our understanding of the underlying mechanisms which regulate bone ECM mineralization remain incomplete. Moreover, the phenotypic analysis of several mouse models and human patients with ectopic calciﬁcation of soft tissues has led to the hypothesis that bone mineralization might be rather a passive process, whereas ECM mineralization is actively inhibited in other tissues. Although this hypothesis is in line with the fact that extracellular concentrations of calcium and phosphate are far above the solubility product for spontaneous precipitation, there is accumulating evidence demonstrating that this view of ECM mineralization is rather simpliﬁed. This chapter provides a review of the most important ﬁndings from the genetic analysis of bone mineralization, using mouse models and human patients. Key words: Phex.

extracellular matrix, mineralization, osteoblast, Mgp, pyrophosphate,

1.1 Physiological and Pathological Mineralization

Physiological mineralization is restricted to bones, teeth and the hypertrophic zone of growth plate cartilage, whereas pathological mineralization – which more often is referred to as ‘‘ectopic calciﬁcation’’ – can be found in any tissue [1]. Whereas bone extracellular matrix (ECM) mineralization is the result of the cellular activity of osteoblasts, pathological ECM mineralization occurs mostly in the absence of osteoblasts, although there are some rare conditions where ectopic bone formation has been observed [2, 3]. With few exceptions (e.g., calcium oxaHandbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

4

1 Mineralization of Bone: An Active or Passive Process?

late crystals in nephrolithiasis), the mineral found not only in bone but also ectopically is composed mainly of calcium and phosphate, with a ratio indicative of crystalline hydroxyapatite [4, 5]. Thus, it is not surprising that low serum levels of calcium and/or phosphate can be one cause for defective skeletal mineralization [6, 7]. Likewise, elevated serum levels of calcium and/or phosphate are associated with a higher risk of ectopic calciﬁcation [8]. There is, however, increasing evidence that the so-called Ca P product is not the only factor inﬂuencing ECM mineralization, but that several gene products are involved in the local control of ECM mineralization in various tissues [9]. The importance of understanding the molecular mechanisms underlying ECM mineralization is underscored by the high prevalence of diseases associated with ectopic calciﬁcation or defects of skeletal mineralization. These include renal failure and atherosclerosis, where accompanying vascular calciﬁcations are frequently observed, osteoarthritis, where mineral deposition occurs in the joints, as well as several inherited diseases, where the functional inactivation of speciﬁc genes leads to locally restricted pathological mineralization [10–14]. Likewise, defects of skeletal mineralization – such as rickets and osteomalacia – are not infrequent, and there is increasing evidence that not only the bone matrix volume but also the degree of bone matrix mineralization has a signiﬁcant impact on the mechanical stability of the skeleton [15, 16]. In addition to the apparent clinical relevance of mineralization-related research, there is also an important biological question that needs to be addressed, as it is still virtually unknown why, under physiological conditions, ECM mineralization is restricted to the skeleton. In a simpliﬁed view, there are two possible explanations for this speciﬁcity: Only osteoblasts, odontoblasts and hypertrophic chondrocytes could produce factors inducing ECM mineralization – that is, skeletal mineralization is actively promoted. Only non-skeletal cell types could produce factors preventing ECM mineralization; that is, pathological mineralization is actively inhibited, whereas physiological mineralization is explained by the absence of inhibition. Although general opinion, which is reﬂected by several textbooks chapters on bone mineralization, was rather postulating an active mechanism, there is increasing evidence in favor of the second possibility. First, the extracellular concentrations of calcium and phosphate exceed their solubility product by several orders of magnitude, thus making all extracellular ﬂuids metastable solutions in terms of calcium phosphate precipitation and mineral formation [17]. Second, several genes encoding inhibitors of pathological mineralization have been identiﬁed, based on the ﬁnding that their functional inactivation in mice and/or humans results in ectopic calciﬁcation [12–14, 18, 19]. Third, there is increasing evidence that tissue-non-speciﬁc alkaline phosphatase – one of the few gene products required for physiological mineralization of the bone ECM – acts by the removal of pyrophosphate, an inhibitor of mineral forma-

1.2 Inhibitors of Pathological Mineralization

tion [7, 20, 21]. However – and not surprisingly – biology is more complex, and there are still many observations that cannot be ﬁtted into such a simpliﬁed concept of ECM mineralization. Therefore, at this point it is worth summarizing our current knowledge on ECM mineralization, and highlighting some of the questions that remain to be clariﬁed by future experiments.

1.2 Inhibitors of Pathological Mineralization

Our current view of ECM mineralization was probably most inﬂuenced by the phenotypic analysis of two mouse models published ten years ago, where the genes encoding the two skeletal g-carboxyglutamate-containing proteins were speciﬁcally deleted [18, 22]. g-Carboxyglutamate (Gla) residues, resulting from a vitamin K-dependent post-translational modiﬁcation step, are well known from factors of the blood coagulation system, where they increase binding to calciumloaded, negatively charged phospholipid surfaces [23]. Gla residues were also found in two proteins of the skeletal ECM, osteocalcin (also called bone Gla protein) and Mgp (matrix Gla protein). Whereas osteocalcin is speciﬁcally expressed by osteoblasts and represents a major constituent of the bone ECM, Mgp is expressed by growth plate chondrocytes, and also by vascular smooth muscle cells [18]. In both cases the negatively charged Gla residues are responsible for a highaﬃnity binding to hydroxyapatite, which made both osteocalcin and Mgp excellent candidates for a regulation of physiological ECM mineralization [24, 25]. Thus, it was very surprising that mice lacking osteocalcin did not display any defect of bone matrix mineralization, although their increased osteoblast activity resulted in a high bone mass phenotype [22]. In contrast, Mgp-deﬁcient mice did have a mineralization defect of growth plate cartilage, though unexpectedly ECM mineralization was not diminished but extended into the prehypertrophic zone [18]. Moreover, the requirement of Mgp to inhibit ECM mineralization was much more obvious in the vasculature. In fact, all Mgp-deﬁcient mice died around the age of 6 to 8 weeks due to rupture of the aorta, which became, like other major arteries, completely calciﬁed over time [18]. Taken together, these results showed that the presence of Mgp in the ECM of the arterial wall, and also in the ECM of prehypertrophic cartilage, is required to prevent pathological mineralization. This is true not only for mice, but also for humans, where the functional inactivation of MGP in patients with the Keutel syndrome leads to similar phenotypic abnormalities [26]. In the meanwhile, the properties of the two skeletal Gla proteins were further compared using several transgenic mouse models [27]. While the presence of a transgene restoring Mgp expression in vascular smooth muscle cells completely rescued the vascular calciﬁcation of Mgp-deﬁcient mice, the same approach with osteocalcin did not inﬂuence their phenotype at all. Moreover, while overexpression of Osteocalcin in osteoblasts had no inﬂuence on bone matrix mineralization, the same approach with Mgp led to severe defects, namely that the amount of

5

6

1 Mineralization of Bone: An Active or Passive Process?

non-mineralized bone matrix (osteoid) was increased more than 10-fold [27]. These results demonstrate that only Mgp, and not Osteocalcin, has an inhibitory inﬂuence on ECM mineralization. Moreover, as Mgp is not a component of the bone ECM, and since its ectopic expression in osteoblasts interferes with bone mineralization, it must be concluded that the absence of mineralization inhibitors (e.g., Mgp) from the bone matrix is one prerequisite for physiological mineralization. The importance of Mgp in arteries and prehypertrophic cartilage is only one example of the need for speciﬁc gene products to prevent ectopic calciﬁcation. A second good example is the serum protein a2-HS-glycoprotein (Ahsg), also known as fetuin-A. As discussed in Chapter 22 of this volume, Ahsg is a systemically acting inhibitor of pathological mineralization, as the deﬁciency of Ahsg in calciﬁcation-prone mice leads to severe ectopic calciﬁcation of several organs, including kidney, lung, and skin [19]. Like Mgp, Ahsg has a high aﬃnity for hydroxyapatite, and both proteins have been shown to be components of a high molecular-weight complex containing calcium and phosphate that was initially found in the serum of rats treated with etidronate [28]. In-vitro studies have further demonstrated that Ahsg acts by forming soluble colloidal spheres with calcium and phosphate in a manner comparable to the action of apolipoproteins that are required to solubilize lipids [29]. Besides these two examples, there is in fact further genetic evidence of the need for speciﬁc inhibitors of unwanted mineralization, although their mechanism of action is less well understood. Osteoprotegerin, for example, is a tumor necrosis factor (TNF)-receptor-like molecule which inhibits bone resorption, yet its deﬁciency in mice also results in arterial calciﬁcation [30]. In contrast, pseudoxanthoma elasticum – an inherited human disorder that is characterized by progressive calciﬁcation of the skin and blood vessels – is caused by a deﬁciency of the ATP-binding transmembrane protein ABCC6 through mechanisms that are still not fully understood [13]. Two other inhibitors of pathological mineralization, Ank and Enpp1, the deﬁciency of which causes ectopic calciﬁcation in mice and humans, act by raising the extracellular level of pyrophosphate, an inhibitor of ECM mineralization (this is discussed in detail below) [14, 31]. Regardless of the underlying mechanisms, however, there is indeed one obvious conclusion that must be drawn from this overwhelming genetic evidence, namely that pathological ECM mineralization needs to be prevented by active mechanisms – that is, by the expression of speciﬁc genes. Such a statement is completely in line with the astonishing fact that extracellular concentrations of calcium and phosphate are far above their solubility product; this means that, in theory, every organ should be calciﬁed.

1.3 Activators of Physiological Mineralization

Compared to the number of mouse models and human diseases associated with pathological mineralization, there are fewer examples for genes that have been

1.3 Activators of Physiological Mineralization

Fig. 1.1 Three examples of skeletal mineralization defects. Undecalciﬁed sections from the tibia of Vdr-deﬁcient mice (left), Phex-deﬁcient Hyp mice (middle), and from a bone biopsy of a TNAP-deﬁcient hypophosphatasia patient (right) were stained using the von Kossa/van Gieson technique. Mineralized matrix is stained black; non-mineralized bone matrix (osteoid) is stained red. All three genetic defects cause a pathological enrichment of osteoid.

shown to be required for physiological mineralization of the skeleton. The most established ones encode the vitamin D receptor (Vdr), tissue-non-speciﬁc alkaline phosphatase (Tnap), and a bone-speciﬁc endopeptidase named Phex [32–34] (Fig. 1.1). Surprisingly, none of these proteins is a component of the bone ECM, and two of these genes (Vdr and Tnap) are not even expressed in a bone-speciﬁc fashion – which raises the question of how they can contribute to bone-speciﬁc ECM mineralization. In the case of Tnap and Phex, the molecular mechanisms underlying their functions are still being investigated, and are discussed below. In contrast, the requirement for Vdr in bone mineralization can already be explained, based mainly on the analysis of a mouse model with a targeted deletion of the Vdr gene. Vdr-deﬁcient mice display all the characteristics of vitamin D-dependent rickets, including alopecia, as known from human patients with inactivating VDR mutations [35]. In addition to the striking defect of growth plate calciﬁcation in Vdrdeﬁcient mice, their bone-speciﬁc histomorphometric analysis further revealed an osteoidosis; that is, an enrichment of non-mineralized bone matrix [36]. In order to provide an explanation for this phenotype, we utilized this mouse model and fed them a high-calcium diet, thereby fully normalizing their hypocalcemia, which is another phenotypic aspect of Vdr-deﬁciency, in mice and humans. Following histologic and histomorphometric analysis of these non-hypocalcemic Vdr-deﬁcient mice, no defects of cartilage and bone ECM mineralization were observed; equally important, no changes in any of the parameters of bone formation and bone resorption were observed [36]. Thus, unlike the situation in alopecia, the severe mineralization defect associated with Vdr-deﬁciency is fully rescued by normalizing the serum calcium levels. This in turn indicates that the major physiologically relevant action of Vdr lies in the intestinal uptake of calcium, and not in any direct stimulation of bone ECM mineralization.

7

8

1 Mineralization of Bone: An Active or Passive Process?

These ﬁndings were not especially unexpected, and did not challenge the classic concept that bone ECM mineralization is actively promoted. In fact, it was still possible that the bone ECM contains proteins that are speciﬁcally expressed by osteoblasts, and are needed to induce the formation of hydroxyapatite crystals. The major protein component of the bone matrix is type I collagen. As it has been shown that mutations in one of the two genes encoding this heterotrimeric protein are the cause of osteogenesis imperfecta, there is no doubt that type I collagen provides a molecular scaﬀold that is important for the ordered deposition of hydroxyapatite [37]. However, the presence of type I collagen in the bone ECM cannot by itself explain the speciﬁcity of the mineralization process, especially as it is also expressed by ﬁbroblasts. Moreover, two recently described transgenic mouse models with ectopic Tnap expression provided evidence that ECM mineralization can be induced in the skin, when Tnap is expressed in ﬁbroblasts, but not when it is expressed in keratinocytes [7]. Taken together, these ﬁndings demonstrate that type I collagen is required, but is not suﬃcient for ECM mineralization. Other major protein components of the bone ECM include the above-described Ahsg, which is enriched from the serum, but not expressed by osteoblasts [38], osteocalcin, osteopontin and bone sialoprotein (Bsp); the latter three components are strongly, if not speciﬁcally, expressed by osteoblasts [39]. As discussed above, neither the deﬁciency nor the overproduction of osteocalcin in transgenic mice has any inﬂuence on mineralization of the bone ECM. In fact, although osteocalcin-deﬁcient mice display a high bone mass phenotype, their increased amount of bone matrix is normally mineralized [22]. Neither do osteopontindeﬁcient mice display any defects of bone ECM mineralization [40]. Moreover, a combination of osteopontin- and Mgp-deﬁciency in mice leads to a further enhancement of the vascular calciﬁcation observed in the Mgp-deﬁcient mice alone, demonstrating that osteopontin rather acts as an inhibitor of ECM mineralization, and not as an activator [41]. Among the known protein components of the mineralized bone ECM, Bsp was always considered to be the best candidate for an activator of ECM mineralization, as it is able to promote hydroxyapatite formation in vitro [42]. Thus, it was especially surprising that the targeted deletion of the Bsp gene in mice also does not result in obvious defects of skeletal mineralization [43]. Therefore, until recently, the only ECM protein that has been shown to inﬂuence mineralization in vivo was Mgp. Moreover, given the striking ectopic mineralization of the Mgpdeﬁcient mice and the ability of Mgp to interfere with bone ECM mineralization when ectopically expressed in osteoblasts, it had to be concluded that there is no need for speciﬁc activators of physiological mineralization residing in the bone ECM [27]. However, as the functions of an increasing number of genes have been studied in vivo – mostly by their inactivation in mice – it was not too surprising that one bone ECM protein which is required for proper skeletal mineralization has ﬁnally been identiﬁed, namely dentin matrix protein 1 (Dmp1) [44]. As the name implies, Dmp1 was originally identiﬁed in teeth, but subsequently was also found to be expressed by diﬀerentiated osteoblasts [45, 46]. Dmp1, like Bsp and osteo-

1.4 The Key Role of Pyrophosphate

pontin, belongs to a family of integrin-binding acidic glycoproteins, and has been shown to induce hydroxyapatite formation in vitro [47]. Consistent with this activity, Dmp1-deﬁcient mice display not only a severe hypomineralization of dentin but also a severe osteoidosis; that is, a pathological enrichment of non-mineralized bone matrix [44, 48]. However, even in the case of Dmp1-deﬁcient mice, the explanation for the skeletal mineralization defect is not as simple as initially anticipated, as these mice have marked reductions in serum calcium and phosphate levels, for reasons that are still unknown [44]. It is not yet clear, therefore, whether it is the direct action of Dmp1 in the bone ECM that is required for its proper mineralization.

1.4 The Key Role of Pyrophosphate

As ECM proteins – with the exception of Mgp, and possibly Dmp1 – do not appear to play the most dominant roles in ECM mineralization, it is not surprising that other cellular activities participate in regulating physiological and pathological mineralization. One mechanism by which cells control ECM mineralization involves the regulation of extracellular pyrophosphate levels. The importance of such regulation has only recently been demonstrated by the functional analysis of three gene products, the deﬁciency of which in mice and humans leads to defects in physiological or pathological mineralization, namely Tnap, Enpp1, and Ank. There is hallmark evidence demonstrating that the activity of Tnap is required for skeletal mineralization, as several inactivating mutations within the human TNAP gene have been identiﬁed as the cause of hypophosphatasia, an inherited disorder characterized by defective mineralization of the bone ECM [33]. Likewise, Tnap-deﬁcient mice recapitulate the phenotype of the human patients and develop severe osteoidosis [49]. Both, the patients with hypophosphatasia and the Tnap-deﬁcient mice, have elevated serum levels of three phosphocompounds – phosphoethanolamine, pyridoxal phosphate and inorganic pyrophosphate – that therefore appear to be natural substrates for Tnap [49, 50]. In particular, the increased pyrophosphate levels (not only based on the genetic evidence discussed below) provided important information with regards to a possible mechanism for the physiologic action of Tnap. Inorganic pyrophosphate (PPi), which structurally is composed of two phosphate ions linked by an ester bond, has the ability to bind to nascent hydroxyapatite crystals, thereby preventing their further growth [51, 52]. Tnap, an ectoenzyme linked to the cell membrane via a GPI-anchor [53], has been shown to hydrolyze PPi, thereby lowering its extracellular concentration and producing phosphate ions (Pi) in turn [54, 55]. Modulation of the PPi/Pi ratio by Tnap appears to be the most important function of this enzyme, as the cell-autonomous mineralization defect of Tnap-deﬁcient osteoblasts can be counteracted by the exogenous addition of Pi [7]. Likewise, lowering the PPi levels in Tnap-deﬁcient osteoblasts by combining Tnap-deﬁciency with a deﬁciency of Enpp1 or Ank also rescues their mineralization defect, thus demonstrating that the PPi-hydrolyzing

9

10

1 Mineralization of Bone: An Active or Passive Process?

activity of Tnap is indeed the predominant action, which explains its requirement for bone ECM mineralization [21, 56]. In contrast to Tnap, Enpp1 and Ank act by increasing the extracellular PPi concentration through diﬀerent mechanisms. Enpp1 (also known as PC-1; plasma cell membrane glycoprotein 1) is a PPi-generating enzyme which resides in the cell membrane not only of osteoblasts but also of other cell types [57, 58]. The importance of Enpp1 in the regulation of physiological and pathological mineralization was ﬁrst demonstrated by the genetic analysis of human patients suﬀering from OPLL (ossiﬁcation of the posterior longitudinal ligament of spine), and a corresponding mouse model termed Ttw (tiptoe walking). In both cases, inactivating mutations within the Enpp1 gene cause ectopic calciﬁcation of the spinal ligaments, thereby providing another example of an inhibitor of pathological mineralization [31, 59]. Moreover, certain mutations of the human ENPP1 gene have recently been shown to cause IIAC (idiopathic infantile arterial calciﬁcation), which is another inherited disease with severe ectopic calciﬁcation of arteries [60]. As Enpp1-deﬁcient mouse models have been shown to display increased bone formation and hypermineralization, it appears that the activity of Enpp1 is also required for the regulation of physiological mineralization [21]. Likewise, the importance of the Ank gene was ﬁrst identiﬁed by the demonstration that its mutational inactivation in mice causes progressive ankylosis, characterized by ectopic calciﬁcation within the synovial ﬂuid [61]. The Ank gene encodes a transmembrane protein which shuﬄes intracellular PPi to the extracellular space, thereby also leading to an increased PPi/Pi ratio in the ECM. ANK mutations in humans have been shown to cause either chondrocalcinosis, or craniometaphyseal dysplasia, the latter condition being characterized by hyperostosis of the calvarial and facial bones [14, 62]. Thus, like Enpp1, Ank is required for the regulation of pathological and physiological mineralization, where both proteins have an inhibitory action mediated through the elevation of extracellular PPi concentrations. In conclusion, hallmark genetic evidence is available showing that the extracellular PPi/Pi ratio plays a key role in ECM mineralization, and that at least three enzymes are physiologically involved in this regulation. As mentioned above, the best demonstration of the interactions of Tnap, Enpp1 and Ank came from the combination of the corresponding mouse deﬁciency models. This analysis demonstrated that the mineralization defect of the Tnap-deﬁcient mice is rescued by an absence of the antagonistic actions of either Enpp1 or Ank [21, 56]. It is clear, therefore, that the requirement of Tnap for physiological mineralization can be explained by the removal of PPi, an inhibitor of mineral formation. Moreover, these ﬁndings also support the concept that ECM mineralization is mainly controlled by inhibitory mechanisms that must be released in the skeletal microenvironment. The question remains, however, of how these enzymes contribute to the tissuespeciﬁcity of the mineralization process, as they are all produced by osteoblasts, but not only. While Enpp1 and Ank are virtually expressed in ubiquitous fashion, there is at least some restriction in the case of Tnap, which is expressed in bone, kidney, liver, and testes [7]. Interestingly, the co-expression of Tnap with genes

1.5 The Mysterious Role of the Endopeptidase Phex

encoding type I collagen can only be found in bone – and this may even be the solution to the problem. As mentioned above, ectopic expression of Tnap in the skin of transgenic mice leads to ectopic mineralization only when the gene is coexpressed with type I collagen [7]. Thus, it is possible that physiological mineralization of the skeleton is due simply to the co-existence of a collagenous molecular scaﬀold and the activity of an enzyme that reduces the local levels of PPi. There is, however, at least one further mechanism clearly involved in regulating physiological mineralization, and this includes the activity of Phex.

1.5 The Mysterious Role of the Endopeptidase Phex

X-linked hypophosphatemic rickets (XLH) is the most common inherited disease in humans that is associated with defects of skeletal mineralization [63]. As the name implies, XLH patients also display low circulating phosphate levels that result from increased urinary phosphate excretion. The genetic defect underlying this phenotype was identiﬁed in 1995, and the aﬀected gene subsequently termed Phex (phosphate-regulating gene with homologies to endopeptidases on the Xchromosome) [34]. Further analysis revealed that this gene encodes a transmembrane protein speciﬁcally expressed by osteoblasts (and not by kidney cells) with an extracellular zinc-binding domain sharing homologies with a family of endopeptidases. This family includes neutral endopeptidase (NEP), the endothelinconverting enzymes (ECE-1 and ECE-2) and the KELL antigen, which act by cleaving speciﬁc substrates such as substance P and enkephalin in the case of NEP, or the endothelin precursor protein in the case of ECE-1 and ECE-2 [64]. Thus, it appeared that the activity of Phex could lie in the activation or inactivation of osteoblast-produced substrate(s) regulating bone ECM mineralization and/or phosphate homeostasis. Until now, the major problem is that no physiological Phex substrate has yet been identiﬁed, although data are accumulating with regards to the pathophysiology of XLH. Most of this information has been derived from analyses of a Phex-deﬁcient XLH mouse model (the Hyp mouse), but also from the genetic analysis of human patients with autosomal dominant hypophosphatemic rickets (ADHR) or tumor-induced osteomalacia (TIO) (see below). The Hyp mouse was established and already characterized as a model of XLH long before the demonstration that their genetic defect lies in an inactivating deletion of the Phex gene [65–67]. Several experiments using these mice led to the hypothesis that two substrates of Phex might exist – one regulating bone ECM mineralization in an autocrine/paracrine manner, and the other one regulating renal phosphate handling [68, 69]. As Phex is not expressed in the kidney, the latter substrate was postulated to be a circulating factor and hypothetically termed ‘‘phosphatonin’’. The existence of such a factor was conﬁrmed by parabiosis experiments, where the surgical connection of the vasculature from wild-type and Hyp mice led to an increased urinary phosphate excretion of the parabiosed wild-type mice [70]. Moreover, cross-transplantation experiments showed that

11

12

1 Mineralization of Bone: An Active or Passive Process?

transplanting the kidney from Hyp mice into wild-type mice did not aﬀect phosphate homeostasis, whereas the opposite experiment led to hypophosphatemia [71]. Several biochemical experiments further demonstrated that the increased urinary phosphate excretion is caused by down-regulation of the expression of Npt2, a sodium-dependent phosphate transporter localized in the proximal tubule [72, 73]. In contrast to the pathogenesis of the hypophosphatemia, controversy persists regarding the question of whether the defects of skeletal mineralization in the absence of Phex are the result of decreased serum phosphate levels, or whether they are caused by cell-autonomous defects. In favor of the ﬁrst possibility, it was recently reported that the severe osteoidosis observed in Hyp mice can almost be normalized by feeding them a high-phosphate diet [7]. There is, however, also convincing evidence in support of an intrinsic defect of bone ECM mineralization associated with the Phex-deﬁciency that is independent of the hypophosphatemia. First, transplanting bone cells from Hyp mice into the gluteal muscle of wild-type mice showed that their ability to form mineralized bone is reduced compared to transplanted bone cells from wild-type mice [74]. Second, at least some groups have found that primary osteoblast cultures derived from Hyp mice have a reduced ability to form mineralized bone nodules ex vivo [75]. Third – and probably most convincing – mice lacking Npt2, the renal phosphate transporter downregulated by phosphatonin, display hypophosphatemia, but not the characteristic defects of skeletal mineralization associated with deletion of the Phex gene [76]. Based on these ﬁndings, a second Phex substrate was postulated and termed ‘‘minhibin’’, as it is hypothetically involved in the local control of bone ECM mineralization. Whereas the identity of minhibin is still not clear, it appears that phosphatonin – the phosphaturic factor elevated in XLH – has already been identiﬁed as Fgf23 [69]. The importance of Fgf23 in the pathogenesis of XLH was ﬁrst discovered by the parallel analysis of two human diseases sharing striking similarities to XLH, namely autosomal dominant hypophosphatemic rickets (ADHR) or tumorinduced osteomalacia (TIO). ADHR is an inherited disease, and its genetic analysis revealed that it is caused by mutations of the human FGF23 gene that lead to increased stability of the circulating protein [77, 78]. TIO is an acquired disorder, where certain tumors of mesenchymal origin lead to hypophosphatemia and defects of skeletal mineralization [79]. In a screen for genes that are diﬀerentially expressed in a bony tumor inducing osteomalacia, Fgf23 was found to be strongly up-regulated compared to the adjacent bone tissue [80]. Based on these ﬁndings, subsequent experiments were performed which showed that the injection of recombinant Fgf23 into wild-type mice leads to decreased renal phosphate reabsorption, and that the implantation of Fgf23-producing Chinese hamster ovary (CHO) cells into nude mice induces a phenotype reminiscent of XLH. Further experiments by several other groups showed that Fgf23 has the ability to down-regulate the expression of Npt2 in kidney cells, and that the elevation of Fgf23 serum levels correlates with the degree of hypophosphatemia in XLH patients [81–84]. Based on all this genetic and experimental evidence, it was specu-

1.6 Concluding Remarks

lated that Fgf23 is indeed the long-sought phosphatonin, which is physiologically inactivated by Phex-mediated cleavage and therefore enriched in the absence of functional Phex. Thus, several groups attempted to show that Fgf23 is indeed a Phex-substrate, but the ﬁnal conclusion from these experiments was that Fgf23 is not cleaved by the endopeptidase activity of Phex [85, 86]. In contrast, the comparison of osteoblast cultures derived from wild-type and Hyp mice revealed an increased expression of Fgf23, suggesting that its up-regulation is caused by transcriptional mechanisms, and not by a diﬀerence in protein degradation. Thus, if Phex really acts as an endopeptidase, a physiological substrate still remains to be identiﬁed that can ﬁll the ‘‘black box’’ between Phex-deﬁciency and an increased expression of the phosphaturic factor Fgf23. One good candidate for such a function was an ECM protein termed Mepe (matrix extracellular phosphoglycoprotein), which is speciﬁcally expressed in terminally diﬀerentiated osteoblasts and strongly upregulated in TIO tumors [87, 88]. However, like Fgf23, Mepe is not cleaved by the endopeptidase activity of Phex, and the combination of both deﬁciencies achieved by crossing Mepe-deﬁcient mice with Hyp mice, did not alter the hypophosphatemia or the defects of skeletal mineralization observed in the latter animals [89, 90]. Thus, despite the eﬀorts of many diﬀerent laboratories, Phex is still an endopeptidase in search of a physiological substrate. However, there is no doubt that Fgf23 plays a major role in phosphate homeostasis, which was ﬁnally conﬁrmed by the generation of a Fgf23-deﬁcient mouse model. As expected, Fgf23-deﬁcient mice display a hyperphosphatemia caused by increased renal phosphate reabsorption, thus demonstrating that the phosphaturic activity of Fgf23 is indeed of physiological importance [91]. In contrast, it was completely unexpected that these mice were severely growth-retarded and displayed a striking increase in the amount of osteoid, which was almost comparable to the skeletal mineralization defect observed in Hyp mice. The fact that both, the Fgf23-deﬁcient mice and the Hyp mice display an osteoidosis – despite having oppositely regulated serum phosphate levels – underscores that the defects of skeletal mineralization cannot be simply explained by pathological alterations of phosphate homeostasis. Moreover, given the fact that both mouse models display a similar bone phenotype, which is still evident when both deﬁciencies are genetically combined [92], it must be concluded that Fgf23 can not be the only factor involved in the pathophysiology associated with the deﬁciency of Phex in mice and humans. Thus, it is very clear that there is at least one factor missing before the important role of this axis in bone ECM mineralization can be fully understood.

1.6 Concluding Remarks

Why does bone mineralize, and not every other tissue? This simple question can now be partially answered, although some open questions remain. It should be

13

14

1 Mineralization of Bone: An Active or Passive Process?

pointed out that the ability to create genetically modiﬁed mouse models, and to pinpoint the genetic defects of inherited human disorders, were absolute requirements when answering the above question. Thus, our knowledge on physiological and pathological mineralization has begun to explode only during the past decade – which means that the ﬁnal issues will most likely be solved very soon. These issues include the identiﬁcation of those molecular mechanisms explaining the phenotypic abnormalities associated with the Phex-deﬁciency, as well as the interaction between the Phex/Fgf23 system and the enzymes regulating local concentrations of pyrophosphate. Concerning the role of the ECM proteins it remains to be clariﬁed, whether Dmp1 is a direct activator of physiological mineralization, or whether the mineralization defects associated with its deﬁciency are caused by hypocalcemia. It is surprising that, to date, Mgp has been the only ECM protein shown to be required to inhibit ectopic calciﬁcation – which raises the question of whether the arterial wall is especially protected, or whether other tissues have their own ECM inhibitors of mineralization. Regardless of these remaining questions, it is already possible to draw a conclusion concerning the philosophical question, whether bone mineralization is an active or passive process. As the ectopic expression of Mgp in osteoblasts leads to a severe reduction in bone ECM mineralization, it is quite clear that an absence of potent mineralization inhibitors from the bone ECM contributes to its ability to mineralize. Moreover, as Tnap clearly acts by removing the mineralization inhibitor pyrophosphate, it appears that ECM mineralization is generally regulated by inhibitory mechanisms. Thus, it would not be too surprising, if the results of future experiments were to indicate that even the endopeptidase Phex acts by inactivating an inhibitor of mineralization. Although this latter issue is at present speculative, a current concept of ECM mineralization can already be established: it is actively inhibited in extraskeletal tissues, but this is counter-acted by speciﬁc gene products in the skeleton.

References 1 T. Schinke, M.D. McKee, G. Karsenty, 2

3

4 5

Nat. Genet. 1999, 21, 150–151. K. Bostrom, K.E. Watson, W.P. Stanford, L.L. Demer, Am. J. Cardiol. 1995, 75, 88B–91B. K.M. Galvin, M.J. Donovan, C.A. Lynch, R.I. Meyer, R.J. Paul, J.N. Lorenz, V. Fairchild-Huntress, K.L. Dixon, J.H. Dunmore, M.A. Gimbrone, Jr., D. Falb, D. Huszar, Nat. Genet. 2000, 24, 171–174. F.L. Coe, A. Evan, E. Worcester, J. Clin. Invest. 2005, 115, 2598–2608. A.S. Posner, Clin. Orthop. Relat. Res. 1985, 200, 87–99

6 Y.C. Li, M. Amling, A.E. Pirro,

M. Priemel, J. Meuse, R. Baron, G. Delling, M.B. Demay, Endocrinology 1998, 139, 4391– 4396. 7 M. Murshed, D. Harmey, J.L. Millan, M.D. McKee, G. Karsenty, Genes Dev. 2005, 19, 1093–1104. 8 G.A. Block, T.E. Hulbert-Shearon, N.W. Levin, F.K. Port, Am. J. Kidney Dis. 1998, 31, 607–617. 9 M. Ketteler, V. Brandenburg, W. Jahnen-Dechent, R. Westenfeld, J. Floege, Nephrol. Dial. Transplant. 2005, 20, 673–677.

References 10 W.G. Goodman, J. Goldin, B.D.

11 12

13

14

15 16 17

18

19

20

21

Kuizon, C. Yoon, B. Gales, D. Sider, Y. Wang, J. Chung, A. Emerick, L. Greaser, R.M. Elashoﬀ, I.B. Salusky, N. Engl. J. Med. 2000, 342, 1478–1483. D. Hamerman, N. Engl. J. Med. 1989, 320, 1322–1330. F. Rutsch, N. Ruf, S. Vaingankar, M.R. Toliat, A. Suk, W. Hohne, G. Schauer, M. Lehmann, T. Roscioli, D. Schnabel, J.T. Epplen, A. Knisely, A. Superti-Furga, J. McGill, M. Filippone, A.R. Sinaiko, H. Vallance, B. Hinrichs, W. Smith, M. Ferre, R. Terkeltaub, P. Nurnberg, Nat. Genet. 2003, 34, 379–381. B. Struk, L. Cai, S. Zach, W. Ji, J. Chung, A. Lumsden, M. Stumm, M. Huber, L. Schaen, C.A. Kim, L.A. Goldsmith, D. Viljoen, L.E. Figuera, W. Fuchs, F. Munier, R. Ramesar, D. Hohl, R. Richards, K.H. Neldner, K. Lindpaintner, J. Mol. Med. 2000, 78, 282–286. A. Pendleton, M.D. Johnson, A. Hughes, K.A. Gurley, A.M. Ho, M. Doherty, J. Dixey, P. Gillet, D. Loeuille, R. McGrath, A. Reginato, R. Shiang, G. Wright, P. Netter, C. Williams, D.M. Kingsley, Am. J. Hum. Genet. 2002, 71, 933–940. B. Wharton, N. Bishop, Lancet 2003, 362, 1389–1400. G. Boivin, P.J. Meunier, Osteoporosis Int. 2003, 14, S19–S24. W.F. Neuman, in: M.R. Urist (Ed.), Fundamental and Clinical Bone Physiology. Lippincott Co., Philadelphia, 1980, pp. 83–107. G. Luo, P. Ducy, M.D. McKee, G.J. Pinero, E. Loyer, R.R. Behringer, G. Karsenty, Nature 1997, 386, 78–81. C. Schafer, A. Heiss, A. Schwarz, R. Westenfeld, M. Ketteler, J. Floege, W. Muller-Esterl, T. Schinke, W. JahnenDechent, J. Clin. Invest. 2003, 112, 357–366. R. Gijsbers, H. Ceulemans, W. Stalmans, M. Bollen, J. Biol. Chem. 2001, 276, 1361–1368. D. Harmey, L. Hessle, S. Narisawa, K.A. Johnson, R. Terkeltaub, J.L. Millan, Am. J. Pathol. 2004, 164, 1199–1209.

22 P. Ducy, C. Desbois, B. Boyce, G.

23 24 25 26

27

28

29

30

31

32

33 34 35

36

Pinero, B. Story, C. Dunstan, E. Smith, J. Bonadio, S. Goldstein, C. Gundberg, A. Bradley, G. Karsenty, Nature 1996, 382, 448–452. B. Furie, B.C ¸ . Furie, Ann. N. Y. Acad. Sci. 1991, 614, 1–10. Q.Q. Hoang, F. Sicheri, A.J. Howard, D.S. Yang, Nature 2003, 425, 977–980. M.E. Roy, S.K. Nishimoto, Bone 2002, 31, 296–302. P.B. Munroe, R.O. Olgunturk, J.P. Fryns, L. Van Maldergem, F. Ziereisen, B. Yuksel, R.M. Gardiner, E. Chung, Nat. Genet. 1999, 21, 142– 144. M. Murshed, T. Schinke, M.D. McKee, G. Karsenty, J. Cell. Biol. 2004, 165, 625–630. P.A. Price, G.R. Thomas, A.W. Pardini, W.F. Figueira, J.M. Caputo, M.K. Williamson, J. Biol. Chem. 2002, 277, 3926–3934. A. Heiss, A. DuChesne, B. Denecke, J. Grotzinger, K. Yamamoto, T. Renne, W. Jahnen-Dechent, J. Biol. Chem. 2003, 278, 13333–13341. N. Bucay, I. Sarosi, C.R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H.L. Tan, W. Xu, D.L. Lacey, W.J. Boyle, W.S. Simonet, Genes Dev. 1998, 12, 1260–1268. I. Nakamura, S. Ikegawa, A. Okawa, S. Okuda, Y. Koshizuka, H. Kawaguchi, K. Nakamura, T. Koyama, S. Goto, J. Toguchida, M. Matsushita, T. Ochi, K. Takaoka, Y. Nakamura, Hum. Genet. 1999, 104, 492–497. M.R. Hughes, P.J. Malloy, D.G. Kieback, R.A. Kesterson, J.W. Pike, D. Feldman, B.W. O’Malley, Science 1988, 242, 1702–1705. E. Mornet, Hum. Mutat. 2000, 15, 309–315. The HYP Consortium, Nat. Genet. 1995, 11, 130–136. Y.C. Li, A.E. Pirro, M. Amling, G. Delling, R. Baron, R. Bronson, M.B. Demay, Proc. Natl. Acad. Sci. USA 1997, 94, 9831–9835. M. Amling, M. Priemel, T. Holzmann, K. Chapin, J.M. Rueger, R. Baron, M.B. Demay, Endocrinology 1999, 140, 4982–4987.

15

16

1 Mineralization of Bone: An Active or Passive Process? 37 B. Sykes, D. Ogilvie, P. Wordsworth, 38

39 40

41

42

43

44

45

46

47 48

49

50 51 52 53

J. Anderson, N, Lancet 1986, 2, 69–72. T. Schinke, C. Amendt, A. Trindl, O. Poschke, W. Muller-Esterl, W. JahnenDechent, J. Biol. Chem. 1996, 271, 20789–10796. J.E. Aubin, Biochem. Cell. Biol. 1998, 76, 899–910. S.R. Rittling, H.N. Matsumoto, M.D. McKee, A. Nanci, X.R. An, K.E. Novick, A.J. Kowalski, M. Noda, D.T. Denhardt, J. Bone Miner. Res. 1998, 13, 1101–1111. M.Y. Speer, M.D. McKee, R.E. Guldberg, L. Liaw, H.Y. Yang, E. Tung, G. Karsenty, C.M. Giachelli, J. Exp. Med. 2002, 196, 1047–1055. G.K. Hunter, H.A. Goldberg, Proc. Natl. Acad. Sci. USA 1993, 90, 8562– 8565. J. Aubin, A. Gupta, R. Zinrgbl, J. Rossant, J. Bone Miner. Res. 1996, 11, S102. Y. Ling, H.F. Rios, E.R. Myers, Y. Lu, J.Q. Feng, A.L. Boskey, J. Bone Miner. Res. 2005, 20, 2169–2177. A. George, B. Sabsay, P.A. Simonian, A. Veis, J. Biol. Chem. 1993, 268, 12624–12630. R.N. D’Souza, A. Cavender, G. Sunavala, J. Alvarez, T. Ohshima, A.B. Kulkarni, M. MacDougall, J. Bone Miner. Res. 1997, 12, 2040–2049. G. He, T. Dahl, A. Veis, A. George, Nat. Mater. 2003, 2, 552–558. L. Ye, Y. Mishina, D. Chen, H. Huang, S.L. Dallas, M.R. Dallas, P. Sivakumar, T. Kunieda, T.W. Tsutsui, A. Boskey, L.F. Bonewald, J.Q. Feng, J. Biol. Chem. 2005, 280, 6197–6203. K.N. Fedde, L. Blair, J. Silverstein, S.P. Coburn, L.M. Ryan, R.S. Weinstein, K. Waymire, S. Narisawa, J.L. Millan, G.R. MacGregor, M.P. Whyte, J. Bone Miner. Res. 1999, 14, 2015–2026. M.P. Whyte, Endocr. Rev. 1994, 15, 439–461. H. Fleisch, S. Bisaz, Nature 1962, 195, 911. R.A. Terkeltaub, Am. J. Physiol. Cell. Physiol. 2001, 281, C1–C11. R. Jemmerson, M.G. Low, Biochemistry 1987, 26, 5703–5709.

54 D.W. Moss, R.H. Eaton, J.K. Smith,

55

56

57

58

59

60

61 62

63 64 65

66

67

L.G. Whitby, Biochem. J. 1967, 102, 53–57. A.A. Rezende, J.M. Pizauro, P. Ciancaglini, F.A. Leone, Biochem. J. 1994, 301, 517–522. L. Hessle, K.A. Johnson, H.C. Anderson, S. Narisawa, A. Sali, J.W. Goding, R. Terkeltaub, J.L. Millan, Proc. Natl. Acad. Sci. USA 2002, 99, 9445–9449. R. Terkeltaub, M. Rosenbach, F. Fong, J. Goding, Arthritis Rheum. 1994, 37, 934–941. K. Johnson, A. Moﬀa, Y. Chen, K. Pritzker, J. Goding, R. Terkeltaub, J. Bone Miner. Res. 1999, 14, 883–892. A. Okawa, I. Nakamura, S. Goto, H. Moriya, Y. Nakamura, S. Ikegawa, Nat. Genet. 1998, 19, 271–273. F. Rutsch, N. Ruf, S. Vaingankar, M.R. Toliat, A. Suk, W. Hohne, G. Schauer, M. Lehmann, T. Roscioli, D. Schnabel, J.T. Epplen, A. Knisely, A. Superti-Furga, J. McGill, M. Filippone, A.R. Sinaiko, H. Vallance, B. Hinrichs, W. Smith, M. Ferre, R. Terkeltaub, P. Nurnberg, Nat. Genet. 2003, 34, 379–381. A.M. Ho, M.D. Johnson, D.M. Kingsley, Science 2000, 289, 265–270. P. Nurnberg, H. Thiele, D. Chandler, W. Hohne, M.L. Cunningham, H. Ritter, G. Leschik, K. Uhlmann, C. Mischung, K. Harrop, J. Goldblatt, Z.U. Borochowitz, D. Kotzot, F. Westermann, S. Mundlos, H.S. Braun, N. Laing, S. Tinschert, Nat. Genet. 2001, 28, 37–41. B. Bielesz, K. Klaushofer, R. Oberbauer, Bone 2004, 35, 1229–1239. A.J. Turner, K. Tanzawa, FASEB J. 1997, 11, 355–364. E.M. Eicher, J.L. Southard, C.R. Scriver, F.H. Glorieux, Proc. Natl. Acad. Sci. USA 1976, 73, 4667–4671. L. Beck, Y. Soumounou, J. Martel, G. Krishnamurthy, C. Gauthier, C.G. Goodyer, H.S. Tenenhouse, J. Clin. Invest. 1997, 99, 1200–1209. T.M. Strom, F. Francis, B. Lorenz, A. Boddrich, M.J. Econs, H. Lehrach, T. Meitinger, Hum. Mol. Genet. 1997, 6, 165–171.

References 68 H.S. Tenenhouse, Nephrol. Dial. 69 70 71

72

73 74

75

76

77 78

79 80

81

82

Transplant. 1999, 14, 333–341. P.S. Rowe, Crit. Rev. Oral Biol. Med. 2004, 15, 264–281. R.A. Meyer, Jr., M.H. Meyer, R.W. Gray, J. Bone Miner. Res. 1989, 4, 493–500. T. Nesbitt, T.M. Coﬀman, R. Griﬃths, M.K. Drezner, J. Clin. Invest. 1992, 89, 1453–1459. H.S. Tenenhouse, A. Werner, J. Biber, S. Ma, J. Martel, S. Roy, H. Murer, J. Clin. Invest. 1994, 93, 671–676. H.S. Tenenhouse, L. Beck, Kidney Int. 1996, 49, 1027–1032. B. Ecarot, F.H. Glorieux, M. Desbarats, R. Travers, L. Labelle, J. Bone Miner. Res. 1992, 7, 215–220. Z.S. Xiao, M. Crenshaw, R. Guo, T. Nesbitt, M.K. Drezner, L.D. Quarles, Am. J. Physiol. 1998, 275, E700–E708. A. Gupta, H.S. Tenenhouse, H.M. Hoag, D. Wang, M.A. Khadeer, N. Namba, X. Feng, K.A. Hruska, Bone 2001, 29, 467–476. The ADHR Consortium, Nat. Genet. 2000, 26, 345–348. K.E. White, G. Carn, B. LorenzDepiereux, A. Benet-Pages, T.M. Strom, M.J. Econs, Kidney Int. 2001, 60, 2079–2086. N. Weidner, D. Santa Cruz, Cancer 1987, 59, 1442–1454. T. Shimada, S. Mizutani, T. Muto, T. Yoneya, R. Hino, S. Takeda, Y. Takeuchi, T. Fujita, S. Fukumoto, T. Yamashita, Proc. Natl. Acad. Sci. USA 2001, 98, 6500–6505. T. Shimada, I. Urakawa, Y. Yamazaki, H. Hasegawa, R. Hino, T. Yoneya, Y. Takeuchi, T. Fujita, S. Fukumoto, T. Yamashita, Biochem. Biophys. Res. Commun. 2004, 314, 409–414. K.B. Jonsson, R. Zahradnik, T. Larsson, K.E. White, T. Sugimoto, Y. Imanishi, T. Yamamoto, G.

83

84

85

86

87

88

89

90

91

92

Hampson, H. Koshiyama, O. Ljunggren, K. Oba, I.M. Yang, A. Miyauchi, M.J. Econs, J. Lavigne, H. Juppner, N. Engl. J. Med. 2003, 348, 1656–1663. Y. Yamazaki, R. Okazaki, M. Shibata, Y. Hasegawa, K. Satoh, T. Tajima, Y. Takeuchi, T. Fujita, K. Nakahara, T. Yamashita, S. Fukumoto, J. Clin. Endocrinol. Metab. 2002, 87, 4957– 4960. T.J. Weber, S. Liu, O.S. Indridason, L.D. Quarles, J. Bone Miner. Res. 2003, 18, 1227–1234. R. Guo, S. Liu, R.F. Spurney, L.D. Quarles, Am. J. Physiol. Endocrinol. Metab. 2001, 281, E837–E847. S. Liu, R. Guo, L.G. Simpson, Z.S. Xiao, C.E. Burnham, L.D. Quarles, J. Biol. Chem. 2003, 278, 37419– 37426. P.S. Rowe, P.A. de Zoysa, R. Dong, H.R. Wang, K.E. White, M.J. Econs, C.L. Oudet, Genomics 2000, 67, 54– 68. L. Argiro, M. Desbarats, F.H. Glorieux, B. Ecarot, Genomics 2001, 74, 342–351. R. Guo, P.S. Rowe, S. Liu, L.G. Simpson, Z.S. Xiao, L.D. Quarles, Biochem. Biophys. Res. Commun. 2002, 297, 38–45. S. Liu, T.A. Brown, J. Zhou, X.S. Xiao, H. Awad, F. Guilak, L.D. Quarles, J. Am. Soc. Nephrol. 2005, 16, 1645–1653. T. Shimada, M. Kakitani, Y. Yamazaki, H. Hasegawa, Y. Takeuchi, T. Fujita, S. Fukumoto, K. Tomizuka, T. Yamashita, J. Clin. Invest. 2004, 113, 561–568. S. Liu, J. Zhou, W. Tang, X. Jiang, D.W. Rowe, L.D. Quarles, Am. J. Physiol. Endocrinol. Metab. 2006, 291, E38–E49.

17

19

2 Bone Morphogenetic Proteins Walter Sebald, Joachim Nickel, Axel Seher, and Thomas D. Mu¨ller

Abstract

Bone morphogenetic proteins (BMPs) and growth and diﬀerentiation factors (GDFs) determine multiple processes in early embryonic development and organogenesis, including the formation of the bones and joints of the axial and appendicular skeleton. BMP-2 and BMP-7 (which is also known as OP-1) are approved as drugs and medical devices for the treatment of non-union fractures and for spinal fusion. Possible future indications for BMPs, GDFs or variants of these proteins include osteoporosis, osteoarthritis, ﬁbrosis, parodontosis, and sinus lift. BMPs and GDFs are members of the TGF-b superfamily which signal into the cell by using two types of single-span membrane receptor chains that both have a cytosolic serine/threonine protein kinase domain. The extracellular ligand-binding domains are small, rich in disulﬁde bonds, and their fold is related to the three-ﬁnger toxins as, for example, are some of the conotoxins. The crystal structure of binary and ternary complexes between BMPs and the ectodomains of type I and type II receptors reveals the mechanism of receptor activation and the important determinants (hot spots) for binding speciﬁcity and aﬃnity. Structure-based design of BMP and GDF variants yields proteins with new and potentially useful properties. Key words: bone morphogenetic protein (BMP), growth and diﬀerentiation factor (GDF), TGF-b superfamily, BMP receptor, X-ray structure, BMP signaling, bone formation and repair, mutants and variants.

2.1 Introduction

Orthopedic surgeons have long known that bone has an enormous capacity not only to heal fractures but also to regenerate defects up to a critical size. Furthermore, critical size defects that do not heal spontaneously can be repaired by Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

20

2 Bone Morphogenetic Proteins

implants of autologous or heterologous donor bone. Bone fragments that have been processed to diﬀerent extents still result in the repair of critical size defects by processes of osteoconduction and de-novo osteoinduction. Most remarkably, bone fragments that have been demineralized and extracted for immunogenic proteins retain their osteoinductive properties. Mainly as a result of the investigations conducted by Marshal Urist [1], which commenced during the 1960s, it became clear that the osteoinductive properties of bone can be dissociated and extracted on a preparative scale from the collagenous bone matrix by chaotropic agents such as 6 M urea. This orthopedic research converged during the late 1980s with biochemical and recombinant DNA techniques. Starting with 100 kg of bovine bone, Wang et al. were able to isolate a few micrograms of protein that, after proteolytic digestions, yielded a few peptide sequences. After translation into DNA probes, genomic DNA and cDNAs of several BMP isoforms (BMP-2/-3/-4/-5/-6/-7/-8) could be obtained ﬁrst from bovine and later from human sources [2, 3]. In some slightly earlier studies, Reddi, Sampath and colleagues succeeded in isolating pure OP-1 [4], which turned out to be the same protein as BMP-7. Pure natural OP-1, as well as recombinant BMP-2 to -7 produced in Chinese hamster ovary (CHO) cells possessed high osteoinductive activity when implanted with a carrier into a critical size defect, or into an ectopic site of an animal. It is clear today that single BMPs can induce in vivo the whole process of endochondrial or desmal bone formation, including vasculature and bone marrow. Recently, recombinant BMP-2 and OP-1/BMP-7 have been approved as drugs for orthopedic indications (non-union fractures, spinal fusion), and have acquired a considerable share of the bone reconstruction and tissue engineering market within a few years. BMP sequences are related to that of the transforming growth factor (TGF)-bs [5], the activins [6], and anti-Mu¨llerian hormone (AMH) [7], which had been identiﬁed prior to 1989. Together with the growth and diﬀerentiation factors (GDFs) [8] and glial-derived neurotrophic factor (GDNF) [9], these proteins now form the TGF-b superfamily which comprises more than 30 members. The TGFb-like proteins play pivotal roles during early and all later stages of embryonal development. In the adult organism, they regulate homeostasis and the repair of many tissues and organs. Despite these numerous functions of BMPs in many diverse tissues, the signaling machinery including ligands, receptors, and intracellular signaling proteins is remarkably conserved within this superfamily [10–13]. The role of BMPs in regenerative medicine [14], developmental biology [15], as well as their genetics in mouse mutants [16] and their role in the pathomechanism of human diseases [17], have been described in several excellent reviews. Recently an entire issue of Cytokine and Growth Factor Reviews [Volume 16(3)] has been devoted to BMPs, and today additional data on the structural basis of BMP signaling, on BMP regulation by modulator proteins, and on engineered BMP variants allow the subject to be discussed on a more molecular and chemical perspective with regard to the role(s) of BMPs, and what can be done with them.

2.2 What is a Bone Morphogenetic Protein?

2.2 What is a Bone Morphogenetic Protein?

The role of BMPs within the organism and their eﬀects when provided externally are not the same. For example, whilst some BMPs are termed BMPs they have no bone-inducing capacity in vivo. BMP-1 is a procollagen proteinase [18], while BMP-3 is an inhibitor of bone formation [19]. Likewise, some members of the TGF-b superfamily are not called BMPs but rather GDFs, and they are (e.g., GDF-5) nevertheless supposed to be bone inducers [20]. The phylogenetic tree of the superfamily (Fig. 2.1A) shows that several subfamilies exist. In the following sections, the members of BMP-2/-4 subgroups will be referred to as BMP-2s, those of the BMP-5/-6/-7/-8 subgroups as BMP-7s, and those of the GDF-5/-6/-7 subgroups as GDF-5s. Some of the proteins have several designations, as they have been identiﬁed more than once, sometimes in diﬀerent organisms. Thus, osteogenin 1 (OP-1) was described ﬁrst by Reddi and co-workers [4], and independently as BMP-7 by Wozney and colleagues [2, 21]. GDF-5 was ﬁrst described by Pohl and co-workers [22], but designated ‘‘cartilagederived morphogenetic protein 1’’ by another group [23]. The BMP-2s, BMP-7s and GDF-5s have been shown to induce ectopic bone formation and to induce alkaline phosphatase as a marker for osteoblastic diﬀerentiation in several cell lines. The sequences of BMP-2, OP-1 and GDF-5 are protected by patents which provided the basis for activities of three biotechnological companies in the orthopedic and dental markets. As detailed below, however, the proteins have distinct biochemical properties and receptor usage, and they are expressed at distinct times in speciﬁc tissues and cells. In addition to their bone-inducing capabilities, it is not yet clear how similar the function and activities of these proteins as therapeutics are within the organism. Members of the TGF-b superfamily are primarily synthesized as larger proproteins which initially dimerize, and are then cleaved at a RXXR site by a Furin-type protease. The C-terminal approximately 100 amino acid residues, as a dimer, represent the functional mature protein. Although the BMPs are normally homodimers, there is some indication that heterodimers between BMP-2 and BMP-6 or BMP-7 are signiﬁcantly more active in vitro and in vivo [24]. The propiece as a dimer remains associated with some mature BMPs, without inhibiting their activity [25]. In contrast, the propiece of GDF-8 is a strong inhibitor of GDF-8 activity both in vitro and in vivo [26]. The BMPs and other members of the TGF-b superfamily have a very similar backbone fold and dimer assembly [27, 28]. The backbone forms a cystine knot, where the cysteine side chains of the signature CxGxC and CXC sequence elements constitute a ring, with a third disulﬁde bond threading through the opening of the ring. Further elements of the monomer fold are ﬁnger 1 and ﬁnger 2, each of which comprises a two-stranded b-sheet, a central a-helix and a prehelix loop, and ﬁnally an N-terminal peptide preceding the ﬁrst cysteine of the knot (Fig. 2.1B,C). The whole monomer is usually compared to an open left hand.

21

22

2 Bone Morphogenetic Proteins

Fig. 2.1 (A) Similarity tree of the human TGF-b superfamily, employing aligned sequences of the mature proteins (SwissProt; MultAlin). The average distance tree was constructed by means of the program ‘‘Jalview’’. The designations are BMP (bone morphogenetic protein), GDF (growth and diﬀerentiation factor), CDMP (cartilagederived morphogenetic protein), TGF-b (transforming growth factor beta), Act

(activin), Inh (inhibin), AMH/MIS (antiM€ ullerian hormone), GDNF (glial-derived neurotrophic protein), NRTN (neurturin), Mic (macrophage inhibitory cytokine). (B) Ribbon model of dimeric human BMP-2 [28]. Type I receptors bind to the ‘‘wrist epitope’’, and type II receptors to the ‘‘knuckle epitope’’ [53]. (C) Secondary structures and topology of the BMP-2 monomer.

The dimer is formed by mutually inserting the a-helix as the wrist region of one monomer into the concave open-hand region of the other monomer. An intermonomer disulﬁde bond stabilizes the dimer, and an axis of rotation runs between the two monomers. The BMPs are highly stable, and resistant to denaturation by chaotropic agents; this is most likely due to the extensive disulﬁde binding within and between the monomers. The BMP-2s and BMP-7s – but not the GDF-5s – are glycosylated at multiple sites. Glycosylation has only a minor eﬀect on the solubility of the proteins, as

2.3 BMP Receptors are Composed of Diverse Type I and Type II Receptor Chains

non-glycosylated proteins – either naturally or after Escherichia coli expression – are insoluble under physiological conditions at concentrations higher than 0.1 to 0.25 mM. Glycosylated proteins, for example BMP-4 or BMP-7, remain soluble at two- to fourfold higher concentrations.

2.3 BMP Receptors are Composed of Diverse Type I and Type II Receptor Chains

In mammals, seven type I receptors and ﬁve type II receptors exist for TGF-b-like ligands (for reviews, see [10, 11, 29]). A structure-based alignment of the amino acid sequences of the extracellular domains [30] shows that the members of each type fall into one branch of an average distance sequence tree (Fig. 2.2). The established BMPs receptor chains BMPR-IA and BMPR-IB form a sub-branch, whereas ActR-I (Alk2) shows a relationship only to Alk1 (which is a TGF-b receptor). The dual-speciﬁcity receptors ActR-II and ActR-IIB which signal with both activines and BMPs form a sub-branch, whereas BMPR-II is only distantly related to all other type II chains. The ectodomains of the receptor chains are small and consist of only one domain of about 100 residues. They are rich in disulﬁde bonds, are connected to the membrane-spanning segment by a short linker, and they contain one to three putative N-glycosylation sites. The similarities among the amino acid sequences are very low. The pattern of the cysteines and disulﬁde bridges, however, is conserved and allows sequence alignment based on the established structures of ActR-II [31], TbR-II [32], BMPR-IA [30], and ActR-IIB [33]. The fold of the type I and type II ectodomains (Fig. 2.2) is characterized by one two-stranded and a three-stranded b-sheet, that are held together by ﬁve disulﬁde bonds. The type II ectodomains contain a third two-stranded sheet, whereas the

Fig. 2.2 Similarity tree of the human receptor chains for BMPs and other TGF-b-like proteins, employing structure-based sequence alignment of the ectodomains [30]. Left side: Ribbon models of the ectodomains of BMPR-IA [30] and ActR-II [31].

23

24

2 Bone Morphogenetic Proteins

corresponding segment of the type I ectodomain is folded into a short a-helix. The ectodomains exhibit similarities with the cysteine-rich toxins, as for example conotoxins [34].

2.4 The Basic Signaling Mechanism is the Same for BMPs and other TGF-b-like Proteins

BMPs signal to cells by recruiting two types of single-span membrane receptors into a transphosphorylating complex (Fig. 2.3) (for reviews, see [35–37]. Both types of receptor chain have a cytosolic serine/threonine kinase domain. The type II kinase seems to be constitutively active and transphosphorylates a glycine/serine-rich region (GS box) of the type I chain at multiple sites, thereby activating its kinase. The type I receptors for BMPs (i.e., BMPR-IA, BMPR-IB and ActR-I) activate intracellular signaling Smad proteins which translocate into the nucleus to regulate the expression of target genes. In addition, non-Smad pathways can be activated [13]. The dimeric BMP ligand most likely binds two type I chains and two type II chains, and a hexameric complex consisting of

Fig. 2.3 Signaling BMP receptor complex consisting of type I and type II receptor chains. For further details, see text.

2.5 Biochemistry and Cell Biology of Receptor Speciﬁcity

BMP and two pairs of receptor chains must be assembled for eﬃcient signal transduction [38]. BMPs share this conserved activation mechanism with the other members of the TGF-b superfamily. Several variations of the basic mechanism exist, however, as described below. For example, the BMP-2, BMP-7, and GDF-5 subgroups all use Smad 1, 5, or 8, whereas activines and TGF-bs use Smad 2 or 3 as the intracellular signaling protein. All receptor-restricted Smads (rSmads), with the exception of rSmad2, can directly bind to DNA, although the binding aﬃnity is relatively low and cooperation with sequence-speciﬁc transcription factors is critical for the formation of a stable DNA-binding complex. For example, Smads 1 and 5 interact with bone-speciﬁc transcription factor Runx2 and activate the transcription of target genes [39]. About 500 target genes are estimated according to gene array data to be regulated by BMP signaling in the whole organism (see [10]). About 100 genes responsive to BMP signaling are found in C2C12 cell line upon osteoblastic diﬀerentiation [40].

2.5 Biochemistry and Cell Biology of Receptor Speciﬁcity

The basic signaling machinery and mechanism described above appears to be operating in all receptors for BMPs, and for the other TGF-b-like proteins. However, it is still not clear in all cases which BMPs or GDFs do functionally interact with which receptor chains, although a deﬁned set of receptors has been identiﬁed as Smad1, 5, 8 activating and therefore as bona ﬁde BMP receptors Pioneering experiments conducted a decade or so ago showed that BMPs can be chemically crosslinked in whole cells to the three type I receptors BMPR-IA, BMPR-IB and ActR-IA, and to the three type II receptors BMPR-II, ActR-II, and ActR-IIB [41–44]. The ActR-II and ActR-IIB receptors were originally identiﬁed as receptors for activin [45]. Later, these proteins were shown to exhibit dual specificity and to function also as BMP receptors [46]. ActR-I (Alk2) was ﬁrst postulated to be an activin receptor, but later it was shown that ActR-I functions in BMP and not in activin signaling and activates Smads 1, 5, and 8 [46, 47]. Crosslinking experiments with transfected and non-transfected cell lines demonstrate that receptor preferences exist for BMPs from diﬀerent subgroups, or even for members of the same subgroup. In the cell, BMP-2 and -4 use BMPRIA and possibly BMPR-IB as high-aﬃnity receptors, but the role of ActR-I as a receptor for BMP-2s is uncertain [41]. Among the type II receptors, BMPR-II and ActR-II function with BMP-2s, but whether ActR-IIB is a receptor for BMP-2 or BMP-4 in vivo is unclear. The main GDF-5 receptor is BMPR-IB [48]. GDF-5 is much less eﬃcient with BMPR-IA, and this is supported by the observation that, for example in C2C12 cells which contain BMPR-IA but are devoid of signiﬁcant BMPR-IB levels,

25

26

2 Bone Morphogenetic Proteins Table 2.1 Dissociation constants (KD ) for the 1:1 interaction of receptor ectodomains with biosensor-immobilized BMPs/GDFs (Biacore analysis).

Receptor

BMPR-IA BMPR-IB ActR-I BMPR-II ActR-II ActR-IIB

KD [mM] BMP-2

BMP-7

GDF-5

0.015* 0.095* 420 54 14 8.5

10 1.1* 55 32 0.43 2.7

3.3* 0.3* – 60 22 4.7

* Dissociation

constants were calculated from the kinetic constants as KD ¼ koﬀ =kon . All other KD values were evaluated form the dosedependence of equilibrium binding.

GDF-5 is inactive. GDF-5 signaling is enabled by transfection with BMPR-IB; GDF-5 can also use BMPR-II or ActR-II as type II receptors. BMP-7 and BMP-6 can signal eﬃciently in cells that are devoid of BMPR-IB or of both BMPR-IA and BMPR-IB, whereas BMP-4 cannot, thereby indicating that ActR-I is a functional receptor for these ligands. Interaction with BMPR-II and ActR-II seems to be possible. Detailed investigations established that large diﬀerences in binding aﬃnities exist for diﬀerent ligand/receptor constellations. The apparent dissociation constants, KD , for the interaction of receptor ectodomains with immobilized ligands as determined by biosensor analysis (Biacore) are listed in Table 2.1. BMP-2 binds preferentially to BMPR-IA, and six- to 10-fold less tightly to BMPR-IB. The aﬃnity for ActR-I and for all type II receptors is over 100-times lower. GDF-5 as the representative of another subgroup binds with intermediate aﬃnity to BMPR-IB, and 10-fold less tightly to BMPR-IA. All type II ectodomains bind to GDF-5 with very low aﬃnity. BMP-7, as a member of a third subgroup, exhibits no clear highaﬃnity interaction with any of these receptors. All three type II chains bind with about micromolar aﬃnity, and a medium aﬃnity is seen with BMPR-IB. Remarkably, BMP-6 and -7, as members of the same subgroup, show diﬀerences in binding aﬃnity for some chains, such as BMPR-IA, BMPR-IB, or ActR-II. The in-situ crosslinking and the ectodomain binding data for BMP-2 can be reconciled on the basis of the following two-step mechanism. The activation of these receptors is a sequential process: ﬁrst, the ligand binds to its high-aﬃnity receptor chain – that is, BMPR-IA (or if necessary BMPR-IB) for BMP-2. Second, in the membrane the low-aﬃnity type II chain is recruited into the signaling complex. Collisions – and therefore also on-rates – occur more frequently in the membrane

2.6 Structural Basis for Speciﬁcity and Aﬃnity of BMP Receptor Interaction

than in solution, and therefore solute KD s in excess of 1 mM can be kinetically competent for productive transactivation by reduction of dimensionality. One essential feature of the sequential two-step mechanism is the high-aﬃnity binding of solute BMP-2. However, the aﬃnities of GDF-5, BMP-6 or -7 for any of the receptor chains being discussed here appear to be much too low for the eﬃcient binding of a solute ligand. In particular, the extreme low aﬃnity of ActR-I for BMP-6 and BMP-7 certainly requires some high-aﬃnity interaction with solute ligand. The in-vitro data in Table 2.1 were acquired by measuring 1:1 interactions between ectodomains and the ligands. However, it remains uncertain as to how the aﬃnities change when the dimeric ligand binds to the membrane receptors, which exist to a large percentage in oligomeric form (avidity eﬀects) [49]. As another possibility, accessory or modulatory proteins might cooperate with type II receptors to generate high-aﬃnity binding. (A similar situation exists in the TGFb receptors, where betaglycan or endoglin converts the low-aﬃnity interaction between TGF-b2 and TbRII into a high-aﬃnity signaling-competent interaction.)

2.6 Structural Basis for Speciﬁcity and Aﬃnity of BMP Receptor Interaction

Binary and ternary complexes between BMP-2 or BMP-7 and ectodomains of BMPR-IA, ActR-II and ActR-IIB, have been crystallized and their structures elucidated [30, 50–52]. These data yield important insights into the assembly mechanism of, and molecular recognition in, BMP receptor complexes. The structures also provide the basis for the design of BMP-2 variants with new and potentially useful properties. BMPs are homodimeric and consequently can bind two type I and two type II receptors, thereby forming a ternary complex with a pseudohexameric 1:2:2 composition. Because BMP-2 is a dimer of two identical monomers, the complete complex consists of six polypeptides chains (Fig. 2.4A). BMP-2, with its two hand-like monomers, binds its BMPR-IA receptor at the so-called ‘‘wrist epitope’’ [30, 53] (Fig. 2.4B). This epitope is composed of segments from two monomers; one monomer contributes the central a-helix plus its preceding loop, and the other contributes the concave side of the two twostranded b-sheets. The resultant surface forms a hydrophobic patch with a deep hole on one side. The interface with BMPR-IA shows that 10 hydrogen bridges exist between receptor and ligand. Mutational analysis of contact residues [54] revealed that the two central hydrogen bridges involving main-chain NH- and COgroups are important binding determinants. Interestingly, these polar bonds originate in the pre-helix loop of BMP-2, and one of them ends in the Gln86 side chain of the receptor located in the short a-helix. In addition to these polar bonds, many hydrophobic interactions are found at the interface. The most conspicuous hydrophobic contact exists between the hole of BMP-2 and the Phe85 side chain protruding from the a-helix of the receptor (‘‘knob-into-hole’’ motif ). Thus, the

27

28

2 Bone Morphogenetic Proteins

Fig. 2.4 (A) Ribbon model of the crystal structure of the ternary complex consisting of BMP-2 and two ectodomains of BMPR-IA and two ectodomains of ActR-IIB [52]. (B) Open-book view of the interface between BMPR-IA and BMP-2 wrist epitope. (C) Interface between ActR-IIB and BMP-2 knuckle epitope. Residues are color-coded: gray ¼ aliphatic; green ¼ polar; red ¼ acidic; blue ¼ basic side chains.

2.7 What We Can Do with BMPs: The Engineering of BMP-2 and GDF-5 Variants

a-helix of the type I receptor carries the most important polar and hydrophobic determinants for binding. Until now, the BMP-2 wrist epitope for BMPR-IA binding is the only example of type I receptor interaction. However, it is of interest to note that large knob-like side chains exist in all type I receptors of the TGF-b superfamily, with the exception of Alk1 at a position equivalent to BMPR-IA Phe85. Furthermore, deep holes are seen in the structures of all known TGF-b-like proteins, with the possible exception of BMP-9 [55]. It is therefore conceivable that all of these type I receptors bind their ligands at a site equivalent to the BMP-2 wrist epitope. BMP-2 binds the type II ActR-II and ActR-IIB receptors at the so-called ‘‘knuckle’’ epitope [51, 52] (Fig. 2.4C). Activin A, another TGF-b-like protein, binds ActR-IIB at the same site [33, 56]. Because BMP-7 also uses this site for ActR-II interaction [50], it seems safe to conclude that the knuckle epitope is the common site for BMP and activin interaction with the type II receptors ActR-II and ActR-IIB. Certain mutations in this epitope enhance BMPR-II binding to BMP-2 [53]; thus, all type II receptors are likely to bind to this epitope. (The situation is diﬀerent in TGF-b interaction with TbRII, where the binding epitope of TGF-b is located at the ﬁnger tips; that is, at the terminal loop regions of the b-sheets [57].) The knuckle epitope occurs at the convex side of the BMP ﬁnger region, and is formed from residues of one monomer only. The contact is predominantly hydrophobic, with side chains of aromatic residues at the receptor side and of aliphatic residues at the ligand side. A conserved hydrogen bond at the core of the knuckle epitope contact provides an enlightening example of how high- and low-aﬃnity interactions are generated in the various receptor systems [51, 52]. When this hydrogen bond is mutationally disrupted in ActR-IIB, high-aﬃnity binding to activin A is lost, with the aﬃnity being reduced to the low levels observed in BMP-2 interaction. In contrast, disruption of the corresponding interaction of BMP-2 has only a marginal eﬀect on BMP binding. The geometry of the hydrogen bond is similar in both circumstances, but a conspicuous diﬀerence exists for the two side chains ﬂanking the bond at the periphery. A Lys/Asp pair provides perfect sealing from the solute in the ActA/ActR-IIB contact, whereas a Leu/Asn pair seems to allow some solute access. Transferring the ActA side chains to BMP-2 generates high-aﬃnity binding to ActR-IIB. X-ray structure investigations reveal, indeed, that the Lys/Asp side chains of mutated BMP-2 are immobilized by forming an ion pair, and that the hydrophobic part of these residues provides a perfectly sealed contact [52].

2.7 What We Can Do with BMPs: The Engineering of BMP-2 and GDF-5 Variants

The N-terminal segment of BMP-2 carries a heparin-binding epitope [42, 58]; these segments are ﬂexible and present twice in the dimeric BMP, and therefore

29

30

2 Bone Morphogenetic Proteins

Fig. 2.5 Direct reconstitution of the mandible bone of a mini-pig. (A) A critical-size 5-cm defect in the mandible was treated with carrier material plus recombinant BMP-2 [67]. Full regeneration of the mandible with a mechanically stable bone is visible in the

X-rays taken after 8 weeks. (B) The control defect treated with carrier alone formed a pseudarthrose, i.e. the defect was ﬁlled with connective tissue. (Reproduced from Ref. [67], with permission of Springer Sciences and Business Media.)

they can provide tight links to glycosaminoglycans present in the extracellular matrix (ECM) or the plasma membrane. This interaction likely localizes BMP activity for autocrine or paracrine functions. A BMP-2 variant in which the heparin-binding epitope has been removed [58] showed reduced activity during ectopic bone formation [59]. In contrast, BMP-2 variants with a duplicated heparin-binding epitope showed stronger binding to heparin in vitro and a more eﬃcient bone formation in vivo [59]. The functional reconstitution of large bone defects can be accomplished with recombinant BMP2, as shown in Figure 2.5. Another type of BMP-2 variant with mutations in the knuckle epitope shows a drastic loss in type II receptor binding aﬃnity [53]. The A34D variant has retained very low biological activity, but inhibits the activity of normal BMP-2; this variant is thus an antagonist that competes with normal BMP2 for BMPR-IA receptor binding. Remarkably, the inhibitory activity (IC50 ) of the antagonist is similar to the activity (EC50 ) of normal BMP-2, which indicates that the type II receptor contributes very little – if at all – to the receptor aﬃnity in whole cells. This variant represents a receptor antagonist, whereas natural BMP inhibitors (e.g., Noggin [60], which has been used to inhibit BMP activity in vivo (e.g., [61]), bind directly to the ligand. The wrist epitope of BMP-2, which determines type I receptor binding, can be inactivated by a substitution of Leu51 by proline [54]. The resultant L51P variant

2.7 What We Can Do with BMPs: The Engineering of BMP-2 and GDF-5 Variants

Fig. 2.6 (A) Familial symphalangism caused by a gain-of-function mutation in GDF-5 [66]. Joints are replaced by bone in ﬁnger V and defective in ﬁnger IV (see arrows). The gene exhibits a mutation R438L located in the wrist epitope of GDF-5 (R57L in the mature

protein). The mutant GDF-5 has a severalfold increased aﬃnity for the BMPR-IA receptor. (B) A similar phenotype is produced by heterozygous mutations in the Noggin gene. (Reproduced from Ref. [66], with kind permission.)

is receptor-dead, as it binds only very weakly to the type I receptors. The defect produced by this substitution is local and allows the remainder of the BMP-2 surface to interact with a variety of other proteins, including many BMP modulator proteins. Indeed, the L51P variant releases the inhibition by Noggin and other BMP inhibitors for BMP-2-dependent functions. The L51P variant could therefore be useful during conditions such as fractures [62], osteoporosis [63], and osteoarthritis [64], in which BMP activity might to be limited by modulator proteins as Noggin, CTGF, CHL2, and others. Finally, a GDF-5 variant has been created that exhibits altered receptor speciﬁcity [65]. Normal GDF-5 binds preferentially to the BMPR-IB receptor [48], and only with a much lower aﬃnity to BMPR-IA. A gain-of-function variant of GDF5 could be created by substituting Arg57 by alanine, with the resultant R57A variant binding with identical high aﬃnity to both BMPR-IB and BMPR-IA. It might be of interest to determine if this change in receptor speciﬁcity will inﬂuence the in-vivo activity of GDF-5, for example during bone induction. Remarkably, a similar gain-of-function mutation R57L has been identiﬁed in a form of familiar symphalangism characterized by missing joints in the distal phalanges (Fig. 2.6) [66]. Here, the phenotype is similar to that produced by inactivating mutations in the noggin gene. Inactivating the inhibitor Noggin will also lead to increased BMP signaling, similar to that seen with the gain-of-function GDF-5 mutant. The crystal structures of wild-type and variants of BMPs and GDFs show that the backbone fold and dimer assembly of these proteins is largely resilient to the type of side chains in the receptor-binding wrist and knuckle epitopes. Even proline substitutions are tolerated. The dimeric cystine-knot scaﬀold seems to stabilize these proteins so that speciﬁcities and aﬃnities for diverse receptors can be created by exposing diverse side-chain patterns on an invariant backbone.

31

32

2 Bone Morphogenetic Proteins

References 1 M.R. Urist, Science 1965, 150, 893. 2 E.A. Wang, V. Rosen, P. Cordes, R.M.

3

4

5

6

7

8 9

10 11

12 13 14 15 16 17 18

19

Hewick, M.J. Kriz, D.P. Luxenberg, B.S. Sibley, J.M. Wozney, Proc. Natl. Acad. Sci. USA 1988, 85, 9484. J.M. Wozney, V. Rosen, A.J. Celeste, L.M. Mitsock, M.J. Whitters, R.W. Kriz, R.M. Hewick, E.A. Wang, Science 1988, 242, 1528. F.P. Luyten, N.S. Cunningham, S. Ma, N. Muthukumaran, R.G. Hammonds, W.B. Nevins, W.I. Woods, A.H. Reddi, J. Biol. Chem. 1989, 264, 13377. R. Derynck, J.A. Jarrett, E.Y. Chen, D.H. Eaton, J.R. Bell, R.K. Assoian, A.B. Roberts, M.B. Sporn, D.V. Goeddel, Nature 1985, 316, 701. A.J. Mason, J.S. Hayﬂick, N. Ling, F. Esch, N. Ueno, S.Y. Ying, R. Guillemin, H. Niall, P.H. Seeburg, Nature 1985, 318, 659. R.L. Cate, R.J. Mattaliano, C. Hession, R. Tizard, N.M. Farber, A. Cheung, E.G. Ninfa, A.Z. Frey, D.J. Gash, E.P. Chow, et al., Cell 1986, 45, 685. S.J. Lee, Mol. Endocrinol. 1990, 4, 1034. L.F. Lin, D.H. Doherty, J.D. Lile, S. Bektesh, F. Collins, Science 1993, 260, 1130. Y. Shi, J. Massague, Cell 2003, 113, 685. S. Souchelnytskyi, A. Moustakas, C.H. Heldin, Trends Cell Biol. 2002, 12, 304. P. ten Dijke, C.S. Hill, Trends Biochem. Sci. 2004, 29, 265. A. Moustakas, C.H. Heldin, J. Cell Sci. 2005, 118, 3573. A.H. Reddi, Nat. Biotechnol. 1998, 16, 247. B.L. Hogan, Curr. Opin. Genet. Dev. 1996, 6, 432. G.Q. Zhao, Genesis 2003, 35, 43. X. Cao, D. Chen, Gene 2005, 357, 1. S.W. Li, A.L. Sieron, A. Fertala, Y. Hojima, W.V. Arnold, D.J. Prockop, Proc. Natl. Acad. Sci. USA 1996, 93, 5127. A. Daluiski, T. Engstrand, M.E. Bahamonde, L.W. Gamer, E. Agius,

20

21

22

23

24

25

26 27

28 29

30 31 32

33 34 35 36 37 38

S.L. Stevenson, K. Cox, V. Rosen, K.M. Lyons, Nat. Genet. 2001, 27, 84. R.C. Spiro, L. Liu, M.A. Heidaran, A.Y. Thompson, C.K. Ng, J. Pohl, J.W. Poser, Biochem. Soc. Trans. 2000, 28, 362. A.J. Celeste, J.A. Iannazzi, R.C. Taylor, R.M. Hewick, V. Rosen, E.A. Wang, J.M. Wozney, Proc. Natl. Acad. Sci. USA 1990, 87, 9843. G. Hotten, H. Neidhardt, B. Jacobowsky, J. Pohl, Biochem. Biophys. Res. Commun. 1994, 204, 646. S.C. Chang, B. Hoang, J.T. Thomas, S. Vukicevic, F.P. Luyten, N.J. Ryba, C.A. Kozak, A.H. Reddi, M. Moos, Jr., J. Biol. Chem. 1994, 269, 28227. D.I. Israel, J. Nove, K.M. Kerns, R.J. Kaufman, V. Rosen, K.A. Cox, J.M. Wozney, Growth Factors 1996, 13, 291. D.I. Israel, J. Nove, K.M. Kerns, I.K. Moutsatsos, R.J. Kaufman, Growth Factors 1992, 7, 139. S.J. Lee, A.C. McPherron, Proc. Natl. Acad. Sci. USA 2001, 98, 9306. P.R. Mittl, J.P. Priestle, D.A. Cox, G. McMaster, N. Cerletti, M.G. Grutter, Protein Sci. 1996, 5, 1261. C. Scheuﬂer, W. Sebald, M. Hulsmeyer, J. Mol. Biol. 1999, 287, 103. P. ten Dijke, J. Fu, P. Schaap, B.A. Roelen, J. Bone Joint Surg. Am. 2003, 85-A (Suppl. 3), 34. T. Kirsch, W. Sebald, M.K. Dreyer, Nat. Struct. Biol. 2000, 7, 492. J. Greenwald, W.H. Fischer, W.W. Vale, S. Choe, Nat. Struct. Biol. 1999, 6, 18. C.C. Boesen, S. Radaev, S.A. Motyka, A. Patamawenu, P.D. Sun, Structure (Camb.) 2002, 10, 913. T.B. Thompson, T.K. Woodruﬀ, T.S. Jardetzky, EMBO J. 2003, 22, 1555. H. Terlau, B.M. Olivera, Physiol. Rev. 2004, 84, 41. J. Massague, Annu. Rev. Biochem. 1998, 67, 753. C.H. Heldin, K. Miyazono, P. ten Dijke, Nature 1997, 390, 465. P. ten Dijke, K. Miyazono, C.H. Heldin, Curr. Opin. Cell Biol. 1996, 8, 139. P. Knaus, W. Sebald, Biol. Chem. 2001, 382, 1189.

References 39 K.S. Lee, H.J. Kim, Q.L. Li, X.Z. Chi,

40

41

42

43 44

45 46

47

48

49

50

51

52

53 54

C. Ueta, T. Komori, J.M. Wozney, E.G. Kim, J.Y. Choi, H.M. Ryoo, S.C. Bae, Mol. Cell. Biol. 2000, 20, 8783. O. Korchynskyi, K.J. Dechering, A.M. Sijbers, W. Olijve, P. ten Dijke, J. Bone Miner. Res. 2003, 18, 1177. P. ten Dijke, H. Yamashita, T.K. Sampath, A.H. Reddi, M. Estevez, D.L. Riddle, H. Ichijo, C.H. Heldin, K. Miyazono, J. Biol. Chem. 1994, 269, 16985. B.B. Koenig, J.S. Cook, D.H. Wolsing, J. Ting, J.P. Tiesman, P.E. Correa, C.A. Olson, A.L. Pecquet, F. Ventura, R.A. Grant, et al., Mol. Cell. Biol. 1994, 14, 5961. F. Liu, F. Ventura, J. Doody, J. Massague, Mol. Cell. Biol. 1995, 15, 3479. B.L. Rosenzweig, T. Imamura, T. Okadome, G.N. Cox, H. Yamashita, P. ten Dijke, C.H. Heldin, K. Miyazono, Proc. Natl. Acad. Sci. USA 1995, 92, 7632. L.S. Mathews, W.W. Vale, J. Biol. Chem. 1993, 268, 19013. H. Yamashita, P. Ten Dijke, C.H. Heldin, K. Miyazono, Bone 1996, 19, 569. M. Macias-Silva, P.A. Hoodless, S.J. Tang, M. Buchwald, J.L. Wrana, J. Biol. Chem. 1998, 273, 25628. H. Nishitoh, H. Ichijo, M. Kimura, T. Matsumoto, F. Makishima, A. Yamaguchi, H. Yamashita, S. Enomoto, K. Miyazono, J. Biol. Chem. 1996, 271, 21345. L. Gilboa, A. Nohe, T. Geissendorfer, W. Sebald, Y.I. Henis, P. Knaus, Mol. Biol. Cell 2000, 11, 1023. J. Greenwald, J. Groppe, P. Gray, E. Wiater, W. Kwiatkowski, W. Vale, S. Choe, Mol. Cell 2003, 11, 605. G.P. Allendorph, W.W. Vale, S. Choe, Proc. Natl. Acad. Sci. USA 2006, 103, 7643. D.Weber,A.Kotzsch,J. Nickel,S.Harth, A. Seher, K. Mueller, W. Sebald, T.D. Mueller, BMC Struct. Biol. 2007, 7, 6. T. Kirsch, J. Nickel, W. Sebald, EMBO J. 2000, 19, 3314. S. Keller, J. Nickel, J.L. Zhang, W. Sebald, T.D. Mueller, Nat. Struct. Mol. Biol. 2004, 11, 481.

55 M.A. Brown, Q. Zhao, K.A. Baker, C.

56

57

58 59

60

61

62

63

64

65

66

67

Naik, C. Chen, L. Pukac, M. Singh, T. Tsareva, Y. Parice, A. Mahoney, V. Roschke, I. Sanyal, S. Choe, J. Biol. Chem. 2005, 280, 25111–25118. J. Greenwald, M.E. Vega, G.P. Allendorph, W.H. Fischer, W. Vale, S. Choe, Mol. Cell 2004, 15, 485. P.J. Hart, S. Deep, A.B. Taylor, Z. Shu, C.S. Hinck, A.P. Hinck, Nat. Struct. Biol. 2002, 9, 203. R. Ruppert, E. Hoﬀmann, W. Sebald, Eur. J. Biochem. 1996, 237, 295. K.K. Wurzler, J. Emmert, F. Eichelsbacher, N.R. Kubler, W. Sebald, J.F. Reuther, Mund. Kiefer. Gesichtschir. 2004, 8, 83. J. Groppe, J. Greenwald, E. Wiater, J. Rodriguez-Leon, A.N. Economides, W. Kwiatkowski, M. Aﬀolter, W.W. Vale, J.C. Belmonte, S. Choe, Nature 2002, 420, 636. E. Canalis, A.N. Economides, E. Gazzerro, Endocr. Rev. 2003, 24, 218. Y. Yoshimura, S. Nomura, S. Kawasaki, T. Tsutsumimoto, T. Shimizu, K. Takaoka, J. Bone Miner. Res. 2001, 16, 876. G. Turgeman, Y. Zilberman, S. Zhou, P. Kelly, I.K. Moutsatsos, Y.P. Kharode, L.E. Borella, F.J. Bex, B.S. Komm, P.V. Bodine, D. Gazit, J. Cell Biochem. 2002, 86, 461. N. Nakayama, C.Y. Han, L. Cam, J.I. Lee, J. Pretorius, S. Fisher, R. Rosenfeld, S. Scully, R. Nishinakamura, D. Duryea, G. Van, B. Bolon, T. Yokota, K. Zhang, Development 2004, 131, 229. J. Nickel, A. Kotzsch, W. Sebald, T.D. Mueller, J. Mol. Biol. 2005, 349, 933–947. P. Seemann, R. Schwappacher, K.W. Kjaer, D. Krakow, K. Lehmann, K. Dawson, S. Stricker, J. Pohl, F. Ploger, E. Staub, J. Nickel, W. Sebald, P. Knaus, S. Mundlos, J. Clin. Invest. 2005, 115, 2373. K.K. Wurzler, M. Heisterkamp, H. Bohm, N.R. Kubler, W. Sebald, J.F. Reuther, Mund. Kiefer. Gesichtschir. 2004, 8, 75.

33

35

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling Udo Nackenhorst

Abstract

Bones are living organs that have the ability to adapt themselves to their mechanical demands. This phenomenon is of major importance in endoprosthetics. Following an artiﬁcial implant, the bone is stressed in a non-physiological manner, and this causes bone remodeling. Computational methods are available to predict this behavior, which in turn allows the optimization of prosthesis design such that the surgeon can identify the best available implant for an individual patient’s condition. However, many uncertainties are encountered when quantifying the mechanical loading conditions and the overall mechanical properties of bone tissue. The concept of statically equivalent loads is stated, where the boundary conditions are computed by an inverse simulation from computed tomography data. The mechanical properties of cortical bone are obtained using a micromechanical approach, with several stages of homogenization. Moreover, the process of mechanotransduction may be simulated by using this multi-scale approach. Key words: stress-adaptive bone remodeling, ﬁnite element techniques, hip– joint endoprosthetics, multi-scale methods.

3.1 Introduction

Bone remodeling describes the mechanically driven process of changes in bone constitution with regards to their geometry and internal architecture. The most famous citation on this phenomenon is by Wolﬀ [1], who stated that the structure of bones can be determined by mathematical rules, depending on their mechanical demands. Although Wolﬀ did not propose any written formulae, his statement has been underlined by several investigations, whilst in the past mathematical models have also been derived. Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

36

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling

Inspired by this simple idea, somewhat phenomenological mathematical models and related computational techniques have been developed, enforced by the exponentially increasing availability of computational recourses, and beginning during the 1980s. The phenomenology relates to the growth mechanism, which is assumed to be driven by a rather simple mechanical stimulus. The state of the art of these phenomenological approaches will be reviewed in the ﬁrst section of this chapter, where it will be shown that today such models permit the conduct of qualitative studies on bone remodeling caused by changes in stress conditions following endoprosthesis implant. Later, fully three-dimensional (3-D) analyses are performed, taking individual environmental conditions into account. For example, patient-speciﬁc geometry as well as equivalent muscle forces and joint loads can be derived from computed tomography (CT) data. These computational techniques may help in identifying optimized and biomechanically more compatible designs for prostheses, as well as determining the best implant and treatment procedure for an individual patient’s condition. These approaches will be of qualitative nature, mainly because in-vivo validation is not possible. A major question here relates to the mechanism of mechanical stimulus, which has been the subject of much controversy in reports [2, 3]. This process – which is referred to as called ‘‘mechanosensation’’ and ‘‘mechanotransduction’’ [4] – is driven by the bone cells, the so-called osteocytes that are embedded between the dense bone and connected by a network of numerous processes. Rather simple experiments with cell cultures have been conducted to investigate the nature of mechanosensation [5]. Recently, computations on the behavior of bone cells have been reported, on a cellular scale and with detailed modeling of the cytoskeleton and nucleus, down to a scale where the response of proteins due to mechanical forces has been studied [6, 7]. However, in order to obtain the complete picture, the cells must be studied in their natural environment. Because of the hierarchical architecture of bone (e.g., see [8]), a multi-scale analysis is necessary for these investigations. A ﬁrst approach on such as computational multi-scale analysis is outlined in Section 3.2, and takes into account the osteonal structure of cortical bone, and the laminar architecture of the osteons with anisotropic material properties obtained on a third smaller length-scale, depending on the grade of mineralization. Rather simple cell models have been modeled between the lamellae for sensing strains and signaling the ongoing mineralization and growth of existing osteons, as well as for the creation of new osteons which are dependent upon mechanical demand.

3.2 The Biomechanical Equilibrium Approach

For single-scale macroscopic investigations of stress-adaptive bone remodeling, a continuum approach is suitable. In this way, the microstructure is smeared and expressed by an averaged bone mass density distribution, for cortical as well as spongious bone, which is related to constitutive equations (cf. [9–11]). Because

3.2 The Biomechanical Equilibrium Approach

remodeling phenomena appear to function on long time scales in relation to an individual motion, a quasi-static and isothermal treatment is justiﬁed. This approach is referred to as the theory of ‘‘biomechanical equilibrium’’, because the target is a mechanical equilibrium state with the subsidiary condition that no change of bone mass appears anywhere. Based on this assumption, the problem can be deﬁned mathematically as: (a) the mechanical equilibrium conditions of a continuous solid body; (b) a constitutive law which incorporates the remodeling phenomena; and (c) boundary conditions. The mechanical equilibrium (a) is described by elliptic partial diﬀerential equations, for which the ﬁnite element method (FEM) has been proven to provide efﬁciently approximate solutions. Problems on numerical stability have been discussed in the literature – and solved – with regards to the simulation of bone remodeling during its early stages (e.g., [12]). The constitutive law (b) for the biomechanical problem consists of several ingredients, which include the basic stress–strain relationships, a constitutive relationship between the mechanical properties and the bone mass density distribution, and an evolutionary rule for the mechanically driven local rate of change of bone mass density. With regards to the stress–strain relationship, linear elasticity is amicably accepted, where macroscopic anisotropic mechanical behavior might be taken into account [10, 13, 14]. With regards to the second aspect – namely the relationship between elastic coeﬃcients and the scalar valued bone mass density – there is no consensus in literature to date. Here, often-cited studies include the empirical ﬁndings after Carter and Hayes [15], which relate Young’s modulus and apparent bone mass density by a power law. Other authors prefer piecewise power laws with fractional exponents which represent ﬁts to measurements [9, 16]. By a statistical analysis based on much experimental data reported previously in the literature, Rice et al. [17] have shown that the quadratic term in a polynomial approximation behaves in dominant fashion, which means that Young’s modulus is proportional to the square of bone mass density. This latter ﬁnding has been underlined by theoretical investigations within a framework of consistent continuum theory of materials reported elsewhere [14]. Based on the entropy principle, it is argued that the relationship between Young’s modulus E and bone mass density r is of the form E @ r 2 . The diﬀerence in physical dimensions is treated by a dimensional constant, to be adjusted from experimental results. The isotropic approach sketched here, is easily extended to a more sophisticated theory, where anisotropic mechanical behavior is taken into account [14]. The ﬁnal constituent of the continuum constitutive model is on the formulation of adequate rules for the stress-driven evolution of bone mass density. Rather heuristically, suggestions have been made where often strong non-linear, piecewise continuous evolution rules have been introduced [9, 16]. However, this subject can also be treated within the consistent framework of continuum theory of

37

38

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling

materials, where it can be classiﬁed as locating between continuum damage mechanics, because of the explicit changes in elastic properties as a result of mechanical treatment (e.g., [10]), and visco-elasticity because of the explicit time dependency of the processes (see [18]). Within this theoretical framework, the evolutionary rule is derived from the entropy-principle, where the mechanical stimulus function is obtained as conjugate thermodynamic force which drives bone growth [18]. A remaining challenge here is to describe the boundary conditions (c) with regards to the muscle forces and joint loads. Despite the fact, that much excellent data have been based on measured hip-joint loads for speciﬁc activities [19] and related muscle force collectives have been computed [20], these data cannot be used directly for the remodeling simulation because they reﬂect short-term action only. The time scale for bone remodeling processes extends to month and years, and such an argument has already has been stressed in order to justify the quasistatic equilibrium approach. Consequently, a statically equivalent load set must be described which, in an ideal case, reﬂects the individual conditions. Related computational strategies have now been under development for more than 10 years (see [21]), and recently have been made available for practice on 3-D bone remodeling simulations [22]. When applying these approaches, which are based on measured bone mass density distributions (e.g., CT data), the statically equivalent muscle forces and joint loads are computed by solving the inverse problem, which is deﬁned mathematically as follows: Find the muscle-forces and joint loads, such that a given bone mass density distribution is obtained in the biomechanical equilibrium state. A combination of genetic and gradient-type algorithms have been developed in order to compute the statically equivalent load sets for this ill-posed problem. A result of this schema is depicted in Figure 3.1. The measured bone mass densities for some horizontal slices are shown at the left of the ﬁgure. These results are mapped onto the ﬁnite element discretization of the femoral bone (as depicted in the center of the ﬁgure). With these data, the statically equivalent load sets are computed by solving the inverse problem. Finally, a straightforward iteration is performed to generate an equilibrated model. The resulting bone mass density distribution obtained for this example, taking into account ﬁve muscle forces, is depicted at the right of the ﬁgure. The correlation with the originally measured data was quite good, the diﬀerences being due mainly to the imaging based on the ﬁnite element discretization and the pixel-based CTimages. With regards to the sensitivity of the results on the ﬁnite element discretization, it has been shown in numerical experiments [22] that rather rough models lead to qualitatively good results. For example, in the presentation shown in Figure 3.1, fewer than 4000 linear tetrahedral elements were used, a fact which also underlines the robust nature of the computational approach. One point should be made with regards to the relevant muscle groups that have been taken into account. During the early investigations, it was shown that the joint load and subsumption of the abductor muscles acting at the trochanter major (a simple equilibrium of a statically determined, one-leg stand condition)

3.2 The Biomechanical Equilibrium Approach

Fig. 3.1 Mapping of computed tomography (CT) data (left) onto the ﬁnite element mesh (middle) and reverse-computed bone mass density distribution (right) for the biomechanical equilibrium state.

leads to qualitatively good results [23]. The optimization techniques for solving the inverse problem provide clear hints on the relevance of diﬀerent muscle groups in the bone remodeling simulation. For the example depicted in Figure 3.1, the statically equivalent load set listed in Table 3.1 has been computed. For comparison, maximal forces obtained for a walking sequence are listed. From these results it is clearly seen that, in addition to the joint force, the gluteus group provides the most dominant inﬂuence, followed by the psoas major and the

Table 3.1 Computed statically equivalent load sets in comparison with

maximal forces. (Modiﬁed from Ref. [20].) Indication

Statically equivalent forces [N]

Maximal forces during walking [N]

Joint force Gluteus minimus Gluteus medius Gluteus maximus Psoas major, illiacus Vastus lateralis Vastus intermedialis Biceps femoris, caput breve Vastus medialis Adductor longus

1371 589 272 223 222 139 58 21 7 4

2190 284 306 91 174 228 63 92 9 7

39

40

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling

Fig. 3.2 Bone remodeling caused by a standard stem prosthesis. Compared to the immediate postoperative state (left), a clear loss of bone mass is observed in the cortical bone surrounding the stem in the biomechanical equilibrium state (right).

vastus group. In comparison with the peak forces reported for a walking sequence, the values diﬀer signiﬁcantly, although this is not a surprising result as the statically equivalent load set reﬂects an average long-term median. However, the tendency in the ordering scheme seems to be the same. From these results it can be concluded that at least ﬁve muscle groups must be taken into account for a suﬃcient bone remodeling simulation (cf. [24]). With these careful preparations – which are based on a rather phenomenological but nonetheless physically consistent approach – tools are available to study bone remodeling caused by medical treatment, for which total hip-joint arthroplasty is the most famous example. The predicted bone remodeling due to treatment with a standard stem endoprosthesis is shown in Figure 3.2. At the left of the ﬁgure the bone mass density distribution in the immediate postoperative state is depicted for a frontal plane cut and for horizontal slices, where the remaining structure of the natural bone has not changed at all. In this gray-scaled presentation, the increasing bone mass density is indicated from bright to dark color, with the exception of black, which indicates regions of vanishing bone mass. The computed biomechanical equilibrium state is depicted at the right of the ﬁgure, and indicates a clear but drastic change in bone mass density in the cortical regions surrounding the stem. This result is in agreement with the socalled ‘‘stress-shielding theorem’’, which postulates that bone resorption occurs in unsuitably stressed domains. These results are underlined by clinical observations, with the predicted bone loss being judged as a referable complication in revisionary treatment. Despite these promising results, the computational techniques available today enable only qualitative prognoses to be made. It is, however, possible to rate

3.2 The Biomechanical Equilibrium Approach

Fig. 3.3 Upper: Simulated time series of bone remodeling caused by a novel metaphyseal-anchored hip-joint endoprosthesis. Lower: Radiographs of a follow-up series.

more or less biomechanically compatible implant designs. In Figure 3.3, for example, the computationally predicted bone remodeling behavior of a new prosthesis model, which has been designed for the treatment of younger patients with a high risk of revision, is compared with X-radiographs from a follow-up series. The computational results clearly show, in qualitative terms, the same tendencies, where the focus is laid on the clearly signed mass density evolution at the tip of the prosthesis. Overall, a stable osseointegration is concluded from the computational results. Additionally, by comparison with Figure 3.1, the innovation is clearly signed, and the bone stock in distal regions remains unchanged because these regions are unaﬀected by this metaphyseal-anchored device. Therefore, this prosthesis provides a good basis for a revision using traditional techniques. As outlined in this section, computational methods based on phenomenological (but physically consistent constitutive) assumptions are available for the qualitative prediction of the remodeling behavior of bones as result of altered mechanical conditions. Today, these techniques can be used to help identify optimal implant designs in terms of their biomechanical compatibility. In the near future, a supportive system for the treatment of individual patients might become available, providing surgeons with clear advice as to the best choice of available implant, the surgical strategy, and the rehabilitation treatment. At present, although such quantitative predictions are not yet possible, it is not possible to validate these models by using in-vivo measurements. One open question that arises with this macroscopic length-scale relates to the mechanism of stimulation. In the case of this phenomenon – which we refer to

41

42

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling

as mechanosensation and mechanotransduction [4] – the nature of the responsible bone cells must be considered. The problem may be solved in part by using a computational multi-scale analysis, a ﬁrst approach for which is outlined in the following section.

3.3 A Computational Multi-Scale Approach for Cortical Bone

The overall functional behavior of bone must be explained on several lengthscales because of its hierarchical architecture. Distinctions between spongious and cortical bone are observable at the macroscopic length-scale, with the former being found in metaphyseal regions and short bones such as vertebra, where it consists of a porous, framework-like structure. In spongious bone, the orientation of the rods and their density is clearly correlated to the mean stress trajectories. From the mechanical point of view, the duty of the spongious bone is to provide smooth load transition between joints. Some micromechanical approaches to compute the average mechanical properties and the stress-driven adaptation of spongious bone have been reported (e.g., [11, 25, 26]), and therefore investigations into this topic appear to be well advanced. Cortical bone has a dense structure and is used to build the tube-like bones. When observed microscopically, it is constructed from an amorphous matrix which is reinforced by cylindrical structures, known as osteons or Haversian systems. The osteons are composed of concentric layers of the basic bone material, which is a composite of collagen matrix and hydroxyapatite crystals. The basic composite is organized in ﬁbrils and ﬁbers, which in each layer of an osteon build up a more or less homogeneous but orthotropic structure. The orientation of ﬁbrils changes from layer to layer; thus osteons must be regarded as cross-ply constructions. This hierarchical organization is an essential ingredient for the overall mechanical resistance of bones. This speciﬁc construction of osteons is very important from a biological point of view. Between the concentric layers bone cells are embedded the osteocytes; these cells are thought to be responsible for mechanosensation, although the mechanisms involved in this functional ﬂow are the subject of controversy (see [2, 3, 7, 27, 28]). However, knowledge concerning mechanical load transfer in those cross-ply systems can help to explain the local shear stress ampliﬁcation that is needed by the cells for load detection. Computational methods based on well-established theorems from natural science can contribute to a deeper understanding of this biological self-organizing process. Knowledge of these processes – which without doubt are driven by mechanical stimulation – will be valuable for medical treatment, and they may also open the door to goal-oriented drug treatment. A ﬁrst approach, highlighting the capabilities of the current state of computational techniques in this ﬁeld, is presented in the following section.

3.3 A Computational Multi-Scale Approach for Cortical Bone

Fig. 3.4 Schematic diagram of the overall closed-loop control circuit on the osteonal dynamics in cortical bone.

3.3.1 Closed Nano-to-Meso Control Circuit Approach

In order to provide and develop a better understanding of the mechanically driven growth and adaptation of bones, a closed-loop multi-scale algorithm (see Fig. 3.4) has proved invaluable [29]. At the largest length-scale, a section of cortical bone is modeled to include roughly discretized models of osteons (for a detailed description, see Section 3.3.4). The cortical section is loaded by axial compression. At the next smaller length-scale (micrometer scale), detailed ﬁnite element models of each individual osteon are created (see Section 3.3.3) which are loaded by the displacement conditions obtained from the cortical section analysis. Each layer of the osteon is modeled with orthotropic material properties that are computed at a subcellular length-scale which depends on the grade of mineralization (see Section 3.3.2). The osteon models are processed for two purposes: (i) to compute the homogenized material properties used in the analysis of the cortical section; and (ii) to compute the mechanical stimulus based on the strains detected by the cell models embedded between the lamellae. Depending on the local strain detected by the cell models, a decision tree is activated for ongoing mineralization, the growth of existing osteons in height, or the creation of new osteons. A sequence of results from this simulation procedure is shown in Figure 3.5. The simulation started with one single osteon in a partially mineralized state. The stage of mineralization is depicted with gray-scales, where black indicates complete mineralization with a 70% hydroxyapatite content. The use of this se-

43

44

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling

Fig. 3.5 Mechanically driven development and mineralization of osteons in a cortical bone section. The grade of mineralization is visualized by the gray scale.

quence illustrates how new osteons are created, how they grow in height, and how the mineralization process occurs continuously. It must be emphasized that from a computational point of view, this multiscale algorithm is easily scalable, because each osteon can be treated independently; for example, each one can be placed on a single personal computer in a network. However, the solution of the overall problem becomes more expensive with increasing density of osteons within a cortical section. The treatment of a complete bone within this multi-scale approach remains visionary, as scale bridging between these models is not yet available. However, the micro-mechanical investigations discussed here may provide a deeper insight into the biomechanical interactions and help in the identiﬁcation of more reliable constitutive properties for macroscopic approaches. 3.3.2 Sub-Cellular Length-Scale

The basic constituents of bone are collagen molecules reinforced by hydroxyapatite crystals, which form ﬁbrils and ﬁbers. A number of analytical approaches derived from the theory of elasticity can be applied in order to monitor the eﬀective linear-elastic behavior of such a composite. An example is the Mori–Tanaka technique [30], which represents a lower bound for the eﬀective elastic coeﬃcients. When using this method approach it is assumed that the shape of the hydroxyapatite crystals can be approximated by ellipsoids which are embedded in a homogeneous collagen matrix. The obtained elastic properties of the mixture are depicted in Figure 3.6 for the orthotropic mean axis system, depending upon the grade of mineralization. The values for fully mineralized bone agree well with those reported elsewhere [8]. At this point it should be emphasized that the volume fraction of the hydroxyapatite

3.3 A Computational Multi-Scale Approach for Cortical Bone

Fig. 3.6 Young’s modulus of bone tissue in the orthotropic mean axis depending on the grade of mineralization, as computed using the Mori–Tanaka approach. The mean axes have been chosen according to the typical size of hydroxyapatite crystals as a1 ¼ 10 nm, a2 ¼ 2 nm, and a3 ¼ 1 nm.

phase is variable when using this approach, and therefore the time-dependent grade of mineralization of bone-tissue can be represented. 3.3.3 Micro-Scale Model (Single Osteon)

Single osteons, when analyzed at the micrometer-length-scale, show the ﬁnite element models to consist of several concentric orthotropic layers, with the mean direction of the transversal isotropy changing from layer to layer. Osteocytes are modeled by assigning randomly distributed elements with isotropic, but much softer, material properties. Due to the orthotropic properties of the layers (which are obtained by using the approach described above) and the embedded osteocyte elements, axis-symmetry does not exist. However, the computational eﬀort can be reduced by the analysis of a representative section with periodic boundary conditions. These computations clearly indicate the strain-ampliﬁcation caused by the kinematics of orthotropic cross-ply constructions. The absolute values that have been computed even when using this geometrically simple approach correlate well with those obtained from cell experiments [3]. 3.3.4 Meso-Scale Model of Cortical Bone

On the next (centimeter) length-scale, a small section of cortical bone saturated with osteons is modeled by using a ﬁnite element approach. At this length-scale

45

46

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling

the osteons cannot be modeled as detailed (see above), and therefore a further step of homogenization must be carried out. Computational techniques are preferred in order to identify the homogenized elastic properties of a cylindrical cross-ply (see [31]). A representative volume element (RVE), which is cut from the osteon model described in Section 3.3.3, serves for the computation of elastic properties observed on the millimeter-length-scale. Besides homogenization, an additional challenge arises with regards to geometric modeling and meshing techniques. The general goal is to compute the dynamic behavior of cortical bone, namely of the creation and growth of new osteons. Thus, at each time step a new geometry and mesh must be created.

3.4 Conclusions

In this chapter, we have reviewed the numeric simulation of stress adaptive boneremodeling phenomena, and outlined the computational methods aimed at optimizing medical treatment and further investigations. A continuum model based on phenomenological assumptions regarding boneremodeling processes was presented, with special emphasis placed on the consistent formulation within the framework of the constitutive theory of continuous media and biomechanical equilibrium. With regards to mechanical loading, the conditions due to joint-forces and muscle interaction and statically equivalent load sets have been computed, based on CT-data and using an inverse simulation technique. The ability of these methods to predict, in qualitative terms, bone remodeling has been demonstrated with a full 3-D analysis of hip-joint endoprostheses. Today, such computations enable distinctions to be made between (biomechanically) compatible implants, and for the identiﬁcation of optimal prosthesis designs. These computations also help surgeons to choose the best prosthesis, to study the eﬀects of the surgery itself, and to optimize remobilization treatment. Although the continuum modeling approach mainly reﬂects clinical observations, several uncertainties persist. One problem relates to the constitutive parameters of bone tissue, and another to the mechanism of mechanosensation and mechanotransduction, though neither problem can be resolved at the macroscopic length-scale due to the hierarchical architecture of bone. Today, much eﬀort is expended into modeling approaches on smaller lengthscales, including computations of spongious bone portions [11, 25] and stress analyses on single-cell models [32] to a point where the mechanical responses of single proteins are analyzed [6]. Whilst these investigations are important, a single-scale analysis will not suﬃce to explain the complex biomechanical interactions involved; this also touched on experimental investigations on cells, and from which the importance of ﬂuid shear on cell stimulation was deduced [2, 5]. Computational methods can help to bridge these gaps, as proposed by the ﬁrstorder approach (see Section 3.2), whereby a closed-loop controlled multi-scale computational technique was introduced for the stress-driven osteonal develop-

References

ment of cortical bone, with three typical length-scales being encountered. On the smallest scale, a constitutive model of the composite, comprising collagen molecules and hydroxyapatite crystals, was introduced and served for the analysis of mechanical responses of single osteons, with their time-dependent grades of mineralization. A further homogenization step was needed when analyzing cortical bone in which, driven by the mechanical environment, new osteons are created, grow in height, and become mineralized with time. This overall process is controlled by the strains that are detected at elements and which simulate the mechanical behavior of osteocytes embedded between the orthotropic layers of the osteons. Although this ﬁrst-order approach to simulating bone growth and aging by bridging several length-scales does not reﬂect the complicated biology in detail, it does demonstrate the ability of computational mechanics in developing a deeper understanding of these processes. It is clear that, as our computational resources increase, much more detail will become available for each particular length-scale, with the provision of more realistic simulations.

References 1 J. Wolﬀ, Das Gesetz der Transformation

2 3 4

5 6

7 8

9

10 11

12

der Knochen. Hirschwald, Berlin, 1892. E.H. Burger, J. Klein-Nulend, FASEB J. 1999, 13, 101–112. H. Frost, Bone Min. 1992, 19, 257– 271. S.C. Cowin (Ed.), Bone Mechanics Handbook. CRC Press, Boca Raton, 2001. T.D. Brown, J. Biomech. 2000, 33, 3–14. H. Huang, R.D. Kamm, R.T. Lee, Am. J. Physiol. Cell Physiol. 2004, 287, C1–C11. V. Vogel, Annu. Rev. Biophys. Biomol. Struct. 2006, 35, 459–488. J.Y. Rho, L. Kuhn-Spearing, P. Zioupos, Med. Eng. Phys. 1998, 20, 92–102. G.S. Beaupre, T.E. Orr, D.R. Carter, J. Orthop. Res. 1990, 8, 662– 670. M. Doblare, J.M. Garcia, J. Biomech. 2002, 35, 1–17. R. Ruimerman, P. Hilbers, B. van Rietbergen, R. Huiskes, J. Biomech. 2005, 38, 931–941. C.R. Jacobs, M.E. Levenston, G.S. Beaupre, J.C. Simo, D.R. Carter, J. Biomech. 1995, 28, 449–459.

13 C.R. Jacobs, J.C. Simo, G.S. Beaupre,

14

15 16

17 18

19

20

21 22 23

D.R. Carter, J. Biomech. 1997, 30, 603–613. N. Krstin, U. Nackenhorst, R. Lammering, Techn. Mech. 2000, 20, 31–40. D.R. Carter, W.C. Hayes, J. Bone Joint Surg. 1977, 59, 954–962. H. Weinans, R. Huiskes, H.J. Grootenboer, J. Biomech. 1992, 25, 1425–1441. J.C. Rice, S.C. Cowin, J.A. Bowman, J. Biomech. 1988, 21, 155–168. U. Nackenhorst, Proceedings, International Conference on Computer Methods in Mechanics (CMM05), June 21–24, 2005, Czestochowa, Poland. G. Bergmann, G. Deuretzbacher, M. Heller, F. Graichen, A. Rohlmann, J. Strauß, G.N. Duda, J. Biomech. 2001, 34, 859–871. G. Duda, M. Heller, G. Bergmann, Theoret. Issues Ergonom. Sci. 2005, 6, 287–292. K.J. Fischer, C.R. Jacobs, D.R. Carter, J. Biomech. 1995, 28, 1127–1135. B. Ebbecke, PhD Thesis. IBNM, University of Hanover, 2006. U. Nackenhorst, Techn. Mech. 1997, 17, 31–40.

47

48

3 Biomechanics of Bones: Modeling and Computation of Bone Remodeling 24 J.A. Simoes, M.A. Vaz, S. Blatcher, M.

25

26

27 28

Taylor, Med. Eng. Phys. 2000, 22, 453– 459. T. Adachi, K. Tsubota, Y. Tomita, S.J. Hollister, J. Biomech. Eng. 2001, 123, 403–409. T.M. Keaveny, E.F. Morgan, G.L. Niebur, O.C. Yeh, Annu. Rev. Biomed. Eng. 2001, 3, 307–333. L.A. Taber, Appl. Mech. Rev. 1995, 48, 487–545. T.H. Smit, E.H. Burger, J. Bone Miner. Res. 2000, 15, 301–307.

29 C. Lenz, PhD Thesis. IBNM,

University of Hanover, 2005. 30 T. Mori, K. Tanaka, Acta Metall. 1973,

21, 571–575. 31 T.I. Zohdi, P. Wriggers, Introduction

to Computational Micromechanics. Spinger, 2004. 32 J.G. McGarry, J. Klein-Nulend, M.G. Mullender, P.J. Prendergast, FASEB J. 2004, 10.1096/fj.04-2210fje (express article).

49

4 Direct X-Ray Scattering Measurement of Internal Stresses and Strains in Loaded Bones Stuart R. Stock and Jonathan D. Almer

Abstract

High-energy X-ray scattering oﬀers a unique, non-destructive method for quantifying stress in the interior of bones during in-situ loading. The mineral phase and collagen phase of the composite material bone can be studied independently using wide angle X-ray scattering (WAXS or diﬀraction) and small angle X-ray scattering (SAXS), respectively. In this chapter, X-ray scattering procedures and stress determinations are brieﬂy reviewed, after which the methods used for the studies are summarized and data from several loading experiments presented. Herein, two main results are featured: (i) an independent determination of Young’s modulus in the mineral phase and in the collagen phase of bone via in-situ loading, and comparison with the composite modulus derived from an attached strain gage; and (ii) stress gradients studied in an inhomogeneously loaded rat tibia. Key words:

stress, strain, X-ray scattering, synchrotron radiation, bone.

4.1 Introduction

Most measurements of stress in bone have been made with attached strain gages, and include some human and animal in-vivo studies (for a summary, see [1]). However, in practical terms the number and spacing of such attached gages are very restricted and the data limited to the bone surface. While many remain under the mistaken impression that X-rays are only useful for studying thin surface layers in bone, or for bones reduced to powder, scattering methods employing high-energy X-rays are, in fact, a very attractive means of quantifying internal strains and the related stresses in the interior of mineralized tissues such as bone. High-energy synchrotron X-radiation sources such as the Advanced Photon Source (APS) provide photons that can penetrate millimeters of mineralized Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

50

4 Direct X-Ray Scattering Measurement of Internal Stresses and Strains in Loaded Bones

tissue. Most neutron and X-ray-scattering studies of bone, a composite of collagen reinforced with a high density of carbonated apatite (cAp) nanocrystallites, have concentrated, however, on the crystalline texture or aspects of crystal quality or stoichiometry (see [2] and references therein), and not on internal stress and strain measurements. In this chapter we describe how high-energy X-ray scattering (E > 60 keV) can be used in situ to measure internal stress in intact bones under applied stress. In this method, wide-angle X-ray scattering (WAXS) or diﬀraction is used to measure the cAp response to applied stress, and small-angle X-ray scattering (SAXS) to monitor collagen response. The results obtained at the APS illustrate methods of data collection and analysis, and suggest proﬁtable directions for future research.

4.2 Background 4.2.1 X-Ray Scattering

X-rays scatter from the electron clouds of atoms and from nanoparticles and ﬁbrils. An assembly of scatterers with characteristic size or spacing d reinforces scattered intensity in speciﬁc directions according to Bragg’s well-known relationship, l ¼ 2dhkl sin y, where l is the X-ray wavelength, dhkl is the crystal lattice spacing, and 2y is the angle between the incident and diﬀracted beam directions. For bone, the collagen D-period (@67 nm) along the ﬁbril axis produces SAXS peaks for scattering vectors q ¼ 2p=D, and the Angstrom-level periodicities of apatite crystallites produce diﬀraction peaks in the WAXS regime. Specimens containing many small crystallites with diﬀerent crystal axis orientations produce cones of diﬀracted intensity for monochromatic X-rays; Debye cones from diﬀerent hkl exist simultaneously and produce rings of increased intensity on area detectors [3]. Force applied to a specimen distorts the unit cells and alters the Debye cones (Fig. 4.1). Hydrostatic stresses (those with equal magnitude in all directions) uniformly alter the diameter of cones, whereas deviatoric stresses (those with directionality) change the shape of diﬀraction rings. Most diﬀractometers for polycrystalline specimens are equipped with Cu X-ray tubes and utilize 8 keV photons (Cu Ka line). Only @1% of 8 keV X-rays are

Fig. 4.1 Schematic showing: (A) changes in unit cell dimensions under compression; and (B) the corresponding distortion of Debye rings. The solid lines show the unit cell (Debye rings) prior to loading; the dashed lines show the situation after compression.

4.3 Methods

transmitted through 400 mm of cortical bone [4] and, as most researchers are only familiar with standard diﬀractometers, it is not surprising that many do not realize that options exist for collecting diﬀraction patterns from the interior of intact, centimeter-sized bones. High-energy photons, available at synchrotrons such as the APS, provide good transmission through bone (at 60 keV there is 10% transmission through @14 mm of cortical bone [4]). 4.2.2 Strains and Stresses

X-ray scattering measures quantities such as dhkl in cAp or the D in collagen, and changes in these quantities deﬁne the internal strain imposed during loading – that is, the strain in cAp is ecAp ¼ ðd d initial Þ=d initial and in collagen is ecollagen ¼ ðD Dinitial Þ=Dinitial . Internal stress is a quantity derived from internal strain, and stress sij and strain e kl are second-rank tensors related through the fourth-rank elastic constants Cijkl (i.e., sij ¼ Cijkl e kl ). For a single crystal, the numerical calculation of stress components from strains is straightforward. For data obtained from crystallites with diﬀerent orientations (i.e., polycrystalline samples), the strains measured by X-rays are averages, and to determine an average stress from these strains requires average elastic constants derived from Cijkl according to one of several approximations. The Reuss approximation assumes that all crystallites experience the same stress; the Voigt approximation assumes that all grains within the sample are subjected to the same uniform strain. The Kro¨ner–Eshelby limit, which is calculated for anisotropic precipitates coupled to an isotropic matrix, yields values of elastic moduli close to those observed experimentally; that is, near the mean of the Reuss and Voigt limits [5]. The Kro¨ner constants are used below.

4.3 Methods 4.3.1 Specimens and Geometry

Figure 4.2 illustrates the transmission diﬀraction geometry for WAXS and SAXS of loaded bone specimens as is typically used at station 1-ID, APS. The use of area detectors and high-energy X-rays (80.7 keV) and associated small y allows the scattering from the loading (x2 ) and transverse (x1 ) directions to be collected simultaneously. Two X-ray area detectors were used for rapid data collection: a CCD detector for SAXS positioned over 4 m from the specimen and an image plate (IP) detector for WAXS just over 1 m from the specimen. These separations produced adequate SAXS resolution and WAXS angular range, including the cAp 00.2, 22.2 and 00.4 reﬂections. X-ray exposure times were typically 5 s and 1 s for

51

52

4 Direct X-Ray Scattering Measurement of Internal Stresses and Strains in Loaded Bones

Fig. 4.2 (A) Schematic of the experimental set-up for SAXS and WAXS data collection. Directions x2 and x1 (along the loading direction and transverse to it, respectively) are indicated. (B) MicroCT-derived 3-D image of a 10-mm section of canine ﬁbula loaded in situ. A typical WAXS pattern is shown in the same orientation as the specimen; the lighter pixels on the pattern indicate higher diﬀracted intensities.

WAXS and SAXS, respectively. Capturing full Debye rings in WAXS required that this detector be centered on the incident beam, and SAXS patterns could not be recorded unless the WAXS detector was translated out of the beam along xIP , an operation requiring @30 s to complete. Both detectors were nominally normal to the incident X-ray beam. Loading in three- or four-point bending is frequently used to study the mechanical properties of rodent long bones [6], but this is less than ideal for high-energy diﬀraction studies. The large, rapidly varying stress gradients, ﬁnite X-ray beam dimensions and averaging over the X-rays’ path through the specimen all degrade the sensitivity of WAXS and SAXS analyses when bending geometries are employed. Therefore, the authors’ experiments have employed uniaxial compression of bone. In order to attach entire long bones or bone sections to the load frame, the ends of the specimens are cast in stiﬀ, water-resistant plastic cylinders (Fig. 4.2) that ﬁt snugly into recesses in the ends of the load train. Small strain gages (one or two per specimen) are glued to the specimen surface so that the macroscopic strain in the gage volume emacro can be compared with those in the two phases of the composite (i.e., ecAp and ecollagen ). The screw-driven load frame used here was constructed at APS for high-energy X-ray scattering determinations of internal stresses, and the details are described elsewhere [2]. The frame’s load cell measures the force applied to the specimen, and laboratory microcomputed tomography (microCT) is used to measure the bone cross-sectional area for use in computing the applied stress, sapplied . Careful calibration (specimen–detector separations, detector tilts relative to the incident X-ray beam) is essential for accurate internal stress/strain analysis. Reference samples with well-deﬁned scattering peaks are used for this purpose (ceria NIST Standard Reference Material SRM-674a for WAXS, and silver behenate for SAXS). Accuracy is further improved by using a laser distance gage to correct for small specimen shifts during loading and for curvature of the bone surface.

4.3 Methods

Fig. 4.3 (A) WAXS pattern showing deﬁnition of aximuthal angle h and radius from the pattern center. The darker the pixel, the higher the diﬀracted intensity. (B) Experimental 00.4 peak position versus azimuthal angle for a section of canine ﬁbula at the indicated applied compressive stresses (MPa). The quantities r and h are deﬁned in the text. (Figures from Ref. [2].)

4.3.2 Analysis of Two-Dimensional (2-D) Scattering Patterns

The scattering pattern analyses are performed using FIT2D [7] and MATLAB programs developed at Sector 1 of APS. Figure 4.3 shows a typical 2-D WAXS pattern and the variation of 00.4 peak position versus azimuthal angle h for several applied compressive stresses, but this is not visible at the scale shown. In order to see any diﬀerences, each Debye ring is divided into 72 azimuthal bins, and intensity versus radius is plotted. The resulting 1-D radial plots are ﬁtted with pseudoVoigt functions to provide radial peak position rh that were converted into absolute lattice plane spacings d h using known ceria d-spacings. For the experimental conditions of Figure 4.2, estimates of absolute error in d h are below 104 [8]. For each reﬂection studied, the proﬁles of radius versus h obtained at the diﬀerent applied stresses intersected at a single radius, the invariant radius r , and the corresponding azimuthal orientation was h (see Fig. 4.3B). Measured radii were referred to this r value to provide orientation-dependent (deviatoric) strain values: eh ¼ ðrh r Þ=r . These strain proﬁles, eh , were ﬁtted to a biaxial strain model [9] to account for sample geometry, and provided values of the deviatoric strain components exx and eyy . These strain components, along with the X-ray elastic constants, were then used to calculate stress along the loading direction (see Section 4.3.4). The WAXS patterns show azimuthal intensity variations – that is, diﬀracted intensity around the 00.2 Debye ring varies substantially. In the case of the long bones described here, the diﬀraction patterns show that the cAp crystallites have c-axes oriented primarily along the axes of the long bones, which was an expected

53

54

4 Direct X-Ray Scattering Measurement of Internal Stresses and Strains in Loaded Bones

Fig. 4.4 (A) SAXS pattern from mineralized turkey tendon. The lighter the pixel, the greater the scattered intensity. The black arrows point to 3rd and 5th order D-period peaks, and the longitudinal (L) and transverse (T) directions for the tissue are shown. (B) SAXS intensity as a function of scattering vector q ¼ 2p/D for the pattern shown in using G9 azimuthal integration to improve the signal-to-noise ratio. Data along the tendon L and T directions are shown.

result. Although cAp texture in bones is not the subject of this chapter, it is important to realize that extreme texture produces incomplete Debye rings and can seriously interfere with curve ﬁtting (e.g., of the data in Fig. 4.3B). Crystallite size and microstrain analyses can also be performed using the radial shapes of the diﬀraction peaks in WAXS patterns, such as Fig. 4.3A (see [2]). Analyses can be as simple as the peak width analysis of the Williamson and Hall method (see [3]), or as complex as the peak shape analysis of the Warren– Averbach method (see [5]). Figure 4.4 shows typical SAXS patterns from cAp-mineralized tissue; this example is from hydrated turkey tendon. A 2-D pattern is shown in Figure 4.4A, while Figure 4.4B shows I(q) plots for transverse T and longitudinal L tendon directions. At least 12 peaks are seen, and the use of multiple peak positions (on both sides of the incident beam) improves the precision with which D can be determined. In Figure 4.4, D ¼ 67.4 nm, which is about the value expected. It should also be noted that if no mineral were present, the peak intensities would be much weaker. Variation of peak intensity with order (here, the odd-order peaks are more intense than the even) can be used to determine the fraction of D-period occupied by mineral [10]. This type of calculation is fairly straightforward, and is used not only in SAXS analysis but also in the analysis of structures such as multiple quantum well structures grown by molecular beam epitaxy [11]. The shape of the I(q) curve along the transverse direction can be analyzed to provide information concerning mineral crystallite size and shape [12].

4.4 Examples of Data and Analysis

4.3.3 X-Ray Elastic Constants and Strain–Stress Conversion

For the biaxial geometry used in experiments described below, the transverse strains exx and ezz are assumed to be equal, and the stress and strain tensors are related through [13]: syy ¼

1 S1 eyy ðeyy þ 2exx Þ ; S2 =2 S2 =2 3S1

ð1Þ

where S1 ðhk:lÞ ¼ ðn=EÞaverage and S2 =2ðhk:lÞ ¼ ðð1 þ nÞ=EÞaverage are the average X-ray elastic constants [5] computed from values of Cijkl using the Kro¨ner– Eshelby model via the computer program Hauk.exe (available from the authors upon request): this mathematical formulation is too lengthy to reproduce here [13, 14]. In other words, measured strains exx and eyy (where x and y are along the horizontal and vertical axes of the 2-D diﬀraction pattern, respectively, and the X-ray beam is along z) are used to calculate internal stresses along the loading (i.e., y) direction. The change of notation (exx for e11, etc.) emphasizes the coordinate axes of the detector as opposed to the specimen.

4.4 Examples of Data and Analysis

Measurement of the strains in cAp and collagen (from WAXS and SAXS, respectively) as a function of sapplied (measured by the load cell of the mechanical testing apparatus) allows one to determine Young’s modulus for the individual constituent phase of bone. The strain gage data provides a measure of Young’s modulus (i.e., the macroscopic value). Figure 4.5A shows ecAp , ecollagen and emacro as a function of sapplied for a section of canine ﬁbula kept in a hydrated state by dripping phosphate-buﬀered saline onto the specimen’s surface. The points for the cAp phase are the average of values for the 22.2 and 00.4 reﬂections. The resulting moduli (90% conﬁdence limit) are: Emacro ¼ 24:7ð0:2Þ GPa, EcAp ¼ 41ð1:0Þ GPa, and Ecollagen ¼ 18ð1:2Þ GPa [15]. The value for Emacro is in good agreement with moduli of similar bone types reported elsewhere. The modulus for cAp is about one-third of that of inorganic apatite [16], and this presumably reﬂects the nanocrystalline form of cAp found in bone in intimate contact with collagen. The value of Ecollagen is at least nine times higher than one would expect [17, 18]; even though the collagen D-periodicity is the basis for this determination, most of the scattering power is from the templated cAp, and straight-jacketing of the collagen [18] may be the reason for the large experimental modulus. The collagen straight-jacket model treats bone as a composite of a very high-volume fraction of stiﬀ rods (mineralized collagen ﬁbrils which themselves are nanoparticulate reinforced structures) glued together by a low-volume fraction of non-collagenous

55

56

4 Direct X-Ray Scattering Measurement of Internal Stresses and Strains in Loaded Bones

Fig. 4.5 (A) Longitudinal strain measured in cAp (WAXS), in collagen (SAXS) and by the strain gage (macroscopic) as a function of applied compressive stress for a 10-mm-long section of canine ﬁbula. The WAXS data are the average for 22.2 and 00.4 reﬂections. (Plot is adapted from Ref. [15].) (B) Longitudinal strain as a function of applied

compressive stress in a rat tibia. A 3-D rendering of the specimen is inset at the lower left, and data from the strain gage are shown as well as that for the 00.2 cAp reﬂection measured at ﬁve positions across the specimen (along x). The strain gage data is oﬀset by 7 104 e (i.e., vertically) for clarity.

proteins; this is quite diﬀerent from a simple, uniform, discontinuously reinforced composite. From this perspective, what is measured by SAXS probably reﬂects the stiﬀ rod response, whilst what is measured by WAXS reﬂects the cAp response and Emacro is the combined response of rods and inter-ﬁbrillar mineral and proteins. Figure 4.5B shows data illustrating how the high-energy diﬀraction methods can be used to investigate stress gradients in bones, or assemblies of bones. Strain versus sapplied is shown for a rat tibia loaded in compression. The intrinsic curvature of this long bone results in a signiﬁcant amount of bending when the bone is compressed, a situation that mimics the situation in vivo. The bone’s axis of curvature was positioned parallel to the incident X-ray beam direction so that simple lateral translation of the specimen allowed volumes under diﬀerent stress states to be probed. Data from the strain gage and from ﬁve positions across the specimen are shown in Figure 4.5B.

4.5 Discussion and Future Directions

It should be noted that others have used diﬀraction to measure stresses in thin sections of bone [19], and recent studies combining SAXS with in-situ loading of thin bone specimens have contributed much to our understanding of bone deformation [20, 21]. The studies summarized in this chapter focus on intact bone cross-sections (canine ﬁbula) or entire bones (rat and mouse tibiae), which is con-

References

siderably diﬀerent from thin-section studies. The results presented herein show that it should be quite simple to extend the approach to undissected assemblies of bones and to the in-vivo loading of animal models, such as the rodent ulnar [22] or tibial [23] loading models. It is diﬃcult – if not impossible – to obtain these data by other means.

Acknowledgments

The authors thank Dr. W. Landis and his group (Northeastern Ohio University College of Medicine) for providing the mineralized turkey tendon, Dr. R. Sumner and his group (Rush Medical College) for providing the canine ﬁbula, and Dr. K. Igarashi (Tohoku University, Japan) for providing the rat tibiae. Use of the APS was supported by the US Department of Energy, Oﬃce of Science, Oﬃce of Basic Energy Science, under contract No. W-109E-ENG-38.

References 1 S.P. Fritton, C.T. Rubin, in: S.C.

2 3

4

5

6

7

Cowin (Ed.), Bone Mechanics Handbook, 2nd edn. CRC Press, Boca Raton, 2001, pp. 8.1–8.41. J.D. Almer, S.R. Stock, J. Struct. Biol. 2005, 152, 14–27. Fundamentals of X-ray scattering and diﬀraction appear in texts such as: B.D. Cullity, S.R. Stock, Elements of Xray Diﬀraction, 3rd edn. Prentice-Hall, New York, 2001. NIST, July 2001. Tables of X-Ray Mass Attenuation Coeﬃcients and Mass Energy Absorption Coeﬃcients: from 1 keV to 20 MeV for Elements Z ¼ 1 to 92 and 48 Additional Substances of Dosimetric Interest, NISTIR 5632. I.C. Noyan, J.B. Cohen, Residual stress: Measurement by diﬀraction and interpretation. Springer, New York, 1987. C.H. Turner, D.B. Burr, in: S.C. Cowin (Ed.), Bone Mechanics Handbook, 2nd edn. CRC Press, Boca Raton, 2001, pp. 7.1–7.35. (a) A.P. Hammersley, S.O. Svensson, A. Thompson, Nucl. Instrum. Methods 1994, A346, 312–321; (b) A.P. Hammersley, S.O. Svensson, M. Hanﬂand, A.N. Fitch, D. Ha¨user-

8

9

10

11

12

13

14 15 16

mann, High-Press. Res. 1996, 14, 235– 248. J. Almer, U. Lienert, R.L. Peng, C. Schlauer, M. Ode´n, J. Appl. Phys. 2003, 94, 697–702. B.B. He, K.L. Smith, in: Society for Experimental Mechanics Annual Conference and Exposition, Houston, TX, 1998, pp. 217–220. A. Ascenzi, A. Bigi, M.H.J. Koch, A. Ripamonti, N. Roveri, Calcif. Tissue Int. 1985, 37, 659–664. P.C. Huang, S.R. Stock, A. Torabi, C.J. Summers, Adv. X-ray Analysis 1990, 33, 67–74. W. Tesch, T. Vandenbos, P. Roschgr, N. Fratzl-Zelman, K. Klaushofer, W. Beertsen, P. Fratzl, J. Bone Miner. Res. 2003, 18, 117–125. V. Hauk, Structural and residual stress analysis by nondestructive methods: Evaluation, application, assessment. Elsevier, New York, 1997. Computer program Hauk.exe; available from J. Almer, upon request. J.D. Almer, S.R. Stock, J. Struct. Biol. 2007, 157, 365–370. (a) R.S. Gilmore, J.L. Katz, J. Mater. Sci. 1982, 17, 1131–1141; (b) T.N. Gardner, J.C. Elliott, Z. Sklar, G.A.D.

57

58

4 Direct X-Ray Scattering Measurement of Internal Stresses and Strains in Loaded Bones

17 18 19 20

Briggs, J. Biomech. 1992, 251, 265– 1277; (c) M.C. Sha, Z. Li, R.C. Bradt, J. Appl. Phys. 1994, 75, 7784–7787. N. Sasaki, S. Odajima, J. Biomech. 1996, 29, 655–658. I. Ja¨ger, P. Fratzl, Biophys. J. 2000, 79, 1737–1746. K.S. Borsato, N. Sasaki, J. Biomech. 1997, 30, 955–957. H.S. Gupta, W. Wagemaier, G.A. Zickler, D. Raz-Ben Aroush, S.S. Funari, P. Roschger, H.D. Wagner,

P. Fratzl, Nano Lett. 2005, 5, 2108– 2111. 21 H.S. Gupta, W. Wagemaier, G.A. Zickler, J. Hartmann, S.S. Funari, P. Roschger, H.D. Wagner, P. Fratzl, Int. J. Fract. 2006, 139, 425–436. 22 S.P. Kotha, Y.F. Hsieh, R.M. Strigel, R. Mu¨ller, M.J. Silva, J. Biomech. 2004, 37, 541–548. 23 T.S. Gross, S. Srinivasan, C.C. Liu, T.L. Clemens, S.D. Bain, J. Bone Miner. Res. 2002, 17, 493–501.

59

5 Osteoporosis and Osteopetrosis Adele L. Boskey

Abstract

Osteoporosis, a common bone disease which generally aﬀects the elderly, and osteopetrosis, a much rarer disease which appears early in life, share features of defective osteoclast activity, abnormal osteoblast activity, increased tendency to fracture, and altered bone mechanical properties. These features in both cases are associated with geometric and material abnormalities. In this chapter we review these diseases, their molecular and cellular bases, and the mineral and matrix properties as determined by X-ray diﬀraction and vibrational spectroscopy of bone from humans and animals with these conditions. The features of animal models of each of these conditions are compared with the presentation in humans. Key words: bone formation, bone remodeling, osteoporosis, osteopetrosis, FTIR microspectroscopy, Raman spectroscopy, X-ray diﬀraction, electron microscopy, osteoclasts, osteoblasts.

5.1 Introduction: Two Distinct Diseases with Common Features

Although patients with osteoporosis or osteopetrosis are all at an increased risk of fracture, these diseases – in clinical terms (see Table 5.1) – are very diﬀerent. For example, osteoporosis is quite prevalent, and aﬀects four in ﬁve women and one in eight 8 men [1]. In contrast, osteopetrosis is much rarer, occurring in from 1 in 200 000 individuals in the United States [2], while estimates from an earlier population-based study in Finland suggested that it occurs in 11 in 200 000 individuals [3, 4]. In terms of patient age, osteopetrosis occurs most often in childhood, although there is a rarer adult form [5]; by contrast, osteoporosis does have some childhood variants [6–8], but it is generally a disease of older individuals [9]. Both, osteopetrosis (which is also referred to as ‘‘marble bones’’ or Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

60

5 Osteoporosis and Osteopetrosis Table 5.1 Comparison of clinical features of osteopetrosis and osteoporosis.

Clinical feature

Osteopetrosis

Osteoporosis

Skeletal shape

Short, thicker bones

No growth abnormalities, thinner cortices

Bone mineral density

Elevated

Decreased

Tissue connectivity

Increased

Decreased

Fracture incidence

Elevated

Elevated

Failure to heal fractures

High

Low

Albers–Scho¨nberg disease) and osteoporosis have a variety of forms [10]. Moreover, although the processes of mineralization and skeletal maintenance in each condition are distinct, the diseases share common features. In the healthy individual the processes of bone formation (by osteoblasts) and bone remodeling or turnover (by osteoclasts) are coupled, but in both osteopetrosis and osteoporosis this coupling is lost. The distinctive feature of osteoporosis, by deﬁnition, is the presence of more-porous (less-dense) mineralized bone with an increased tendency to fracture [1]. Osteopetrosis, in contrast, is characterized by an increased amount of calciﬁed cartilage, which results in extremely dense bones that also tend to fracture [10]. The altered properties of both the mineral and the matrix in each of these conditions contribute to the bone fragility. 5.1.1 Comparisons of Clinical Features of Osteoporosis and Osteopetrosis 5.1.1.1 Histology In the bones of patients with either osteoporosis or osteopetrosis, the osteoclasts – the cells that remove mineral and matrix from bone in response to signals (for reviews, see [11, 12]) – do not function properly. In general, in osteopetrosis the osteoclasts have a decreased activity, whereas in osteoporosis their activities are often increased. In osteopetrosis, the impaired osteoclastic activity results in an accumulation of calciﬁed cartilage (Fig. 5.1a) that, in contrast to the normal situation, is not replaced by bone. Bone formation by osteoblasts is generally also impaired, but in certain cases it may be normal. In osteoporosis, there is generally an imbalance between osteoblastic and osteoclastic activity, with too little new bone formation and too much bone resorption, leading to the loss of connectivity (Fig. 5.1b), which contrasts with the situation in healthy, control bone (Fig. 5.1c).

5.1 Introduction: Two Distinct Diseases with Common Features

Fig. 5.1 Histologic and radiographic characteristics of cancellous bone in (a) osteopetrosis, (b) osteoporosis and (c) normal bone. Note the persistence of calciﬁed cartilage within the woven bone in the osteopetrotic tissue, and the thin struts in the osteoporotic bone. (Reproduced, with permission, from Peter G. Bullough, Orthopaedic Pathology, 4th edn. Mosby, New York, 2004, Figures 7.10a, 7.47, and 7.48a.)

5.1.1.2 Radiography Similar to the two-dimensional (2-D) histology of bone sections, whole-bone 2-D radiographic images of osteoporosis show a loss of bone mass. In contrast, radiographs of the bones of patients with osteopetrosis demonstrate an increase in tissue density due to the accumulated calciﬁed cartilage. Many osteopetrotic patients are dwarfed and anemic, some are blind and deaf as a result of pressure on the relevant nerves, and most have thickened skulls [13]. It is believed that the artist Toulouse-Lautrec was a victim of some form of osteopetrosis [14]. Despite the diﬀerence in radiographic appearance when patients with osteopetrosis and osteoporosis are compared, there is an increased risk of fracture in both conditions [4]. In osteopetrosis there is also a high incidence of non-unions (failure to heal fractures), because a key feature of the normal fracture-healing process is the removal of calciﬁed cartilage by osteoclasts, followed by replacement with bone. The fragility fractures in osteoporosis generally occur through the thinning trabeculae, but generally heal without non-union complications.

61

62

5 Osteoporosis and Osteopetrosis

Clinically, bone mineral density (BMD) measurements are used for the diagnosis of both, osteoporosis and osteopetrosis. Two-dimensional values (areal BMD) are calculated from dual photon absorptiometry (DXA); the resultant value is generally reported as a T-score, the standard deviation (SD) of the patient’s BMD compared to a healthy young-adult reference population. The T-score is calculated as: T ¼ (patient BMD mean young adult BMD)/1 SD young adult BMD; with BMD and SD expressed in units of g cm2 .

Z-scores, which compare similarly aged individuals, are less frequently used. In the diagnosis of osteoporosis, individuals with a T-score less than 1.5 are at risk for fracture, individuals with a T-score less than 2.5 are deﬁned as osteoporotic, and those with positive score are considered not at risk of fracture [1]. Patients with osteopetrosis have very dense bones, and high values of BMD, with positive T-scores; such people are also at high risk of fracture. 5.1.2 Comparisons of Bone Mineral Properties in Osteoporosis and Osteopetrosis

Diﬀerences have been identiﬁed between the mineralization processes and the properties of the bone mineral and matrix in osteoporosis and osteopetrosis when contrasted with age- and gender-matched, disease-free controls. These differences have been elucidated by analyses of the bones themselves, and also from analyses of cell and organ cultures derived from these bones. Reviews of these methods can be found elsewhere [15]. The strength of bones, or their ability to resist fracture, is determined by the amount of mineral present (BMD), the geometry (shape) and architecture of the bones, the composition of both the mineral and the matrix, and the presence of micro-cracks. Methods for evaluating whole-bone properties range from mechanical tests to micro-computed tomography [16], and from histochemistry to in-situ hybridization. X-ray diﬀraction and chemical analyses, nuclear magnetic resonance (NMR), energy-dispersive X-ray analysis (EDX), and vibrational spectroscopic techniques provide insight into mineral and matrix properties [15, 17]. The vibrational spectroscopy parameters have been validated by comparison with independent methods, and include information on mineral content, mineral crystal size, mineral crystal composition, and matrix maturity [17–19]. The availability of array detectors has enabled the rapid detection of spectra in sections of tissues with a spatial resolution under 7 mm. The calculation of peak area ratios or intensity ratios in these multispectral ﬁles permits the generation of hyperspectral images where the x- and y-axes correspond to locations in the tissue and the z-axis to the value of the parameter in question. Using these techniques to assay bone biopsies from patients with these diseases, in osteopetrosis the bones are found to be more dense, and the mineral and matrix to be less mature than in control bones. The crystals in the osteope-

5.2 Animal Models of Osteoporosis and Osteopetrosis

trotic bone tend to be smaller and have more imperfections (inclusions, adsorbed ions, etc.) than those in bones from age- and gender-matched controls. In contrast, in osteoporosis the average mineral and matrix properties in the moreporous bone appears similar to that of bones from older individuals than of the age-matched controls. In other words, osteoporotic bone tends to contain larger, more perfect crystals, with a higher carbonate content and less acid phosphate than control bones. In both osteopetrotic bone and osteoporotic bone, the distribution of mineral properties diﬀers from that seen in the bones of healthy individuals. Most of the detailed information on mineral properties in these diseases has been obtained studies with animal models, and consequently it important to discuss whether these are valid models of the human condition. In the following sections the characteristics of animal models of osteoporosis and osteopetrosis will be compared to the bone properties observed in humans with these conditions. Animal models provide the advantage of being able to compare similar genetic backgrounds, of obtaining larger amounts of tissues for analysis, and allowing controlled studies to be conducted of pharmaceutical therapies.

5.2 Animal Models of Osteoporosis and Osteopetrosis

A variety of naturally occurring drug-induced and surgically induced ‘‘models’’ of osteoporosis [20–65] and osteopetrosis [66–103] are available, in addition to more recently developed genetically engineered mutants (Table 5.2). Although these models all provide insights into the human conditions, they do not all totally resemble the diseases in humans. Thus, in the following discussion it must be recognized that analyses of human tissue, where possible, may yield diﬀerent information about the mineralization process and mineral properties than would the animal models. This occurs, in part, because the diseases are so heterogeneous, in part because they are inﬂuenced by complex factors (multiple genes and genetic interactions, diet, exercise, etc.), and in part because modifying gene expression in the animal may not have the same eﬀect as altering the gene in humans. 5.2.1 Osteoporosis 5.2.1.1 Rodent Models As osteoporosis has long been recognized as a disease which is more prevalent in postmenopausal women and elderly men, the early animal models were based on the ovariectomy and castration of rats [20, 21] and later of mice [22]. Ovariectomized/castrated rats had bones that, during mechanical testing, broke more easily [23], although they did not fracture during normal activities. The immobilization of rodent bones has been used to mimic the bone loss associated

63

64

5 Osteoporosis and Osteopetrosis

with bed-rest or space ﬂight in humans. The most popular model – hind-limb suspension – causes a rapid loss of bone in the hind limbs [24]. Similarly, steroid-induced osteoporosis in rodents [25] did not result in spontaneous fractures, yet the bones were weaker, less dense, and had altered mineral properties, especially around the osteocytes [26]. The ovariectomized rat is an accepted model for testing anti-osteoporotic drugs [21], while the ovariectomized mouse provides the additional advantage of being able to modulate gene expression during these studies. However, because in rodents the cortical bones thicken during aging, whilst in humans the cortices become thinner with age, the mineral properties and mineralization processes in rodents may not always be the best model for comparison to human bone. A limited number of rodent models do fracture spontaneously; examples include the accelerated aging mouse (SAM) [27], in which the mechanical properties are distinct from those of age-matched controls [28]; and the spontaneously fracturing mouse, sfx [29], which lacks an enzyme needed to process vitamin C, a cofactor required for collagen synthesis. The SAM mouse can lose up to 60% of its bone mass during aging, [30]. Interestingly, while the long bones of these mice are mechanically weaker, the vertebrae – though reduced in trabecular volume, number, and thickness – do not show the reduced mechanical strength associated with osteopenia in other models, or in humans [28, 31]. The bones of SAM are weak and brittle, the matrix contains less collagen, and the collagen ﬁbrils are not properly organized. However, the mineral properties, as determined by Raman analyses, do not diﬀer from those in controls [32]. Mice in which genes have been ablated (knockouts), inserted (knockins), or otherwise modiﬁed (transgenics) while not fracturing, often develop phenotypes with decreased bone density (osteopenia), and these may provide important insights into the mineralization process. For example, the osteonectin knockout loses trabecular bone with age and shows extensive cortical thinning [33]. Osteonectin is a matrix glycoprotein that inter alia regulates collagen ﬁbrillogenesis. An early phenotype of these animals is the development of cataracts [34]. As these animals age, their bones become weaker in torsion (i.e., they fracture more easily) than wild-type controls, and the mineral content of their bones is increased, as is the average crystal size and maturity of the collagen [35]. The increase in crystal size and matrix maturity is similar to what is seen in humans with osteoporosis, but in contrast to the frequently noted decreased mineral content in osteoporotic humans [36]. The biglycan knockout mouse also loses bone with age [37]. Biglycan is a small leucine-rich proteoglycan that regulates collagen ﬁbrillogenesis and binds growth factors within the matrix. In vitro, and in the absence of ﬁbrillar collagen, biglycan can act as a bone mineral (hydroxyapatite) nucleator [38]. The biglycan knockout mouse has weaker bones than its wild-type control, decreased numbers of trabeculae with decreased mineral content, and increased bone mineral crystal size [37]; this situation is more analogous to that seen in osteoporotic humans. The

5.2 Animal Models of Osteoporosis and Osteopetrosis

biglycan knockout phenotype is dependent on the background of the mouse in which the gene is ablated [38]. The above models indicate the importance of collagen organization for the mechanical properties of bone. Humans and animals with osteogenesis imperfecta (OI), or ‘‘brittle bone disease’’, further illustrate this. OI is a rare birth defect which is due, for the most part, to a variety of diﬀerent mutations aﬀecting the formation of type I collagen, the principal matrix component of bone [39]. The animal models of OI all show increased bone brittleness (reduced energy required to break the bones and decreased elasticity), with the severity and number of spontaneous fractures depending on the nature of the genetic mutation [40– 44]. Some regions of bones of OI patients and animal models have mineral crystals outside the collagen matrix [45]. Additionally, in OI bones there are fewer mineral crystals than in healthy age- and gender-matched bone, and in general the crystals are smaller and have a composition which is distinct from that of the age-matched controls [46, 41]. As in humans with OI, the fracture incidence in mice (if the disease is not perinatal lethal) diminishes with age, which is quite the opposite of what happens in osteoporosis. In fact, in osteoporotic humans the probability of having a second or third fracture after the ﬁrst fragility fracture is markedly increased during the ﬁrst year [47]. A number of other transgenic and knockout mice that have osteopenic phenotypes for which mineral properties also exist, but have not been reported. For example, mice that over-express interleukin-4 (IL-4), a cytokine which has multiple eﬀects on a variety of cell types, have decreased osteoblastic activity [48]. These mice (both sexes) develop a ‘‘hump-back’’ with age (reminiscent of the dowager’s hump in osteoporotic women), but the mice bones do not fracture spontaneously. The cortical thickness in the long bones of these mice decreases with age, as did the trabeculae in the vertebral bodies. Histological studies have revealed no evidence of osteomalacia or other bone diseases, but the bones of the transgenics had reduced mechanical strength. Similarly, mice overexpressing noggin (one of many antagonists of bone morphogenetic proteins; BMPs) in bone cells [49] show signiﬁcantly decreased osteoblastic activity with age. At the age of 8 months these mice had increased marrow space, a signiﬁcantly lower BMD in all bones, decreased bone formation rates, and decreased osteoblastic and osteoclastic activities [50]. Another BMP antagonist, sclerostin (a product of the SOST gene expressed by osteoblasts and osteocytes), when overexpressed under a bone-speciﬁc promoter resulted in mice that had less bone, less mineral within the existing bones, and bones which were mechanically weaker [51]. As reviewed elsewhere, knockout of the high bone density gene, LRP5, or disruption of the factors with which it interacts, also produces osteopenia in mice [52]. A stem-cell antigen cell-surface protein, Sca-1, when knocked out, causes development of osteopenia in older mice associated with a failure to replenish osteoblasts rather than a failure for pre-osteoblasts to undergo osteogenesis [53]. Factors such as osteoprotegrin (OPG), which regulates osteoclastogenesis, when

65

66

5 Osteoporosis and Osteopetrosis

knocked out can also provide models of osteoporosis; the OPG-knockout has a decreased bone density [54]. Additional mouse models with osteopenic phenotypes have not been analyzed for mineral properties (for a review, see [55]). 5.2.1.2 Non-Rodent Models Larger animals mimic more closely the loads placed on human bone and, when ovariectomized, they have been used as osteoporotic models. Thus, ovariectomized dogs [56], ewes [57], mini-pigs [58] and non-human primates [59] have been analyzed. Dietary manipulation has also been used, with or without ovariectomy, to mimic human osteoporosis in these animals. In an on-going study, bones from adult sheep treated with a diet to induce metabolic acidosis, ovariectomized sheep, and ovariectomized sheep given the metabolic acidosis diet showed no signiﬁcant diﬀerence in mineral content relative to controls, but all had increased crystallinity, increased carbonate content, and increased collagen maturity relative to controls, similar to the situation seen in humans with osteoporosis. The non-human primates are the only animals that fracture in the wild. Such animals have menstrual cycles, and Haversian systems similar to those in humans. In Caribbean monkeys, cortical mineral density and porosity increased with age, and vertebral density and cortical area increased with animal weight [60]. Ovariectomized cynomolgus monkeys develop accelerated bone loss, increase bone turnover, and have reduced bone strength [59] relative to shamoperated controls. Density fractionation and mineral analyses by X-ray diﬀraction of the jaws of immobilized monkeys show a signiﬁcant shift toward higher density fractions indicative of the presence of a greater than normal content of mature highly mineralized bone and a parallel decrease in crystallite size [61]. Young ovariectomized and sham-operated cynomolgus monkeys have less-dense bones than control animals; crystal size in these animals was not altered [62]. By using infrared (IR) microspectroscopy, it was reported [63] that trabecular bone from ovariectomized monkeys had signiﬁcantly lower mineral-to-matrix ratios, a parameter directly related to ash weight [64] with values of 5.8 G 0.2 compared to controls (6.2 G 0.2; p a 0.05) and contained larger/more perfect apatite crystals (increased crystallinity) with increased carbonate: phosphate ratios. Similarly, using synchrotron-based IR microspectroscopy, Miller et al. [65] found an increased acid phosphate content and diﬀerent collagen structure in ovariectomized monkeys at 2 years post ovariectomy. Reduced rates of mineralization were also found in these animals. These results are in good agreement with ﬁndings in osteoporotic humans, even when taking into account the population variation. 5.2.2 Osteopetrosis

In humans, the osteopetroses are a heterogeneous group of bone-remodeling disorders characterized by an increase in bone density due to defects in bone remodeling (osteoclastic resorption), and with an increased incidence of fracture [66].

5.2 Animal Models of Osteoporosis and Osteopetrosis

These diseases are usually classiﬁed based on inheritance, age of onset, severity, and clinical symptoms. They include infantile malignant autosomal recessive osteopetrosis, an intermediate (milder) autosomal recessive form, adult benign autosomal dominant osteopetrosis type I, and autosomal dominant osteopetrosis type II. Another variant, known as pycnodysostosis, has also been reported. This disease is due to a deﬁciency in cathepsin K activity, and may involve the impaired formation of osteoclasts, the impaired activity of these bone-resorbing cells, or both. Various naturally occurring and genetically modiﬁed animal models mimicking diﬀerent forms of osteopetrosis have been studied in eﬀorts to elicit further understanding of the pathogenesis of the disease and to evaluate potential treatments. In addition, some drugs that block osteoclast action have been found to induce an osteopetrotic phenotype in animals and humans. 5.2.2.1 Rodent Models The ﬁrst recognized rodent models of osteopetrosis were the mutant animals described by Marks and colleagues [67]. The earliest observed phenotype in many of these rodents is failure of the teeth to erupt, as bone must be removed by osteoclasts for this to occur. While the functional activity of the osteoclasts in these rodents is impaired for a variety of reasons, the resulting bone phenotype is quite similar, with a persistence of calciﬁed cartilage, and the presence of excessive mineralized tissue containing smaller crystals than comparable tissue in normal controls. For example, the toothless rat (tl/tl) [68], has a decreased bone mineral crystal size relative to normal controls. The collagen matrix in these rats is characterized by slight decreases in reducible cross-links and increases in the content of the stable cross-links, pyridinoline, and deoxy-pyridinoline, reﬂecting the persistence of more mature tissue. The incisor-absent (ia/ia) rat, similarly, has a higher mineral content in both their calvaria and long-bone metaphyses than age-matched controls [69]. The ia/ia mineralized tissue contains smaller mineral crystals relative to controls. The metaphyses in the ia/ia rats also has an elevated hexosamine content, deriving from cartilage proteoglycans, and verifying the persistence of cartilage. A naturally occurring mouse model (op/op) lacks one of the factors essential for osteoclast diﬀerentiation [70]. These mice also have decreased bone-forming (osteoblastic) activity. The presence of an altered osteoblast phenotype was also suggested in studies of tl/tl rats [71]. These data suggest that the coupling of osteoblasts and osteoclasts [72] may be disturbed in these models. While these rodent models have increased bone mineral density and cross-sectional geometries relative to their controls, their bones are weaker than their normal littermates, and their cortices signiﬁcantly thinner [73]. The analysis of a necropsy sample from a human infant with the lethal form of osteopetrosis similarly showed the presence of smaller crystals [74]. In contrast to these models, the gray lethal (gl/gl) osteopetrotic mouse, with the most severe form of osteopetrosis [75, 76], and @10% of patients with malignant infantile osteopetrosis, have a defective bicarbonate-chloride transport system [77,

67

68

5 Osteoporosis and Osteopetrosis

78]. This chloride transport is essential for the osteoclast-mediated acidiﬁcation of the extracellular matrix necessary to remove the mineral, and for the activity of other lysosomal enzymes. In this severe variant, while osteoclast activity is impaired, osteoblast activity is normal, which leads to the suggestion that the chloride transport mechanism might be targeted for the treatment of osteoporosis [79]. Other less-severe rodent models include the op/op mouse that fails to resorb bone due to a defect in macrophage colony-stimulating factor [70], the congenitally osteosclerotic oc/oc mouse that lacks a functional H(þ)-ATPase a3 subunit (also important for acidifying the extracellular matrix) [80], and the microphthalmic mi/mi mouse [81]. The op/op mouse has defective osteoblast activity and impaired mineralization [70], while the small-eyed mi/mi mouse has a defect in the microphthalmia-associated transcription factor (MITf ). These mice have decreased osteoclast formation [81], while their marrow cells over-express receptor activator of nuclear factor kappa B (RANK) ligand (RANKL), a factor which, when binding to its receptor, activates osteoclastogenesis [82]. Another rodent model of osteopetrosis mimics the condition known as pycnodysostosis, which is linked to a defect in cathepsin-K, a lysosomal cysteine protease elevated in activity in active osteoclasts [83]. Detailed structural analyses of the bones of one young patient and one older patient with this disease showed thickening of the bone mineral particles, poorly aligned crystals associated with collagen ﬁbrils, and abnormal trabecular structure [84]. The cathepsin K knockout mouse [85] had abnormal matrix turnover, but no histochemically apparent defects in either mineral removal or accretion. Infrared imaging of the bones of cathepsin K null animals (kindly provided by Drs. Gelb and Schaﬄer of Mount Sinai Medical School) showed not only the persistence of highly mineralized calciﬁed cartilage, but also a decreased crystallinity and decreased matrix maturity similar to that reported in a limited number of human biopsies (Fig. 5.2). These mice also show disorganized matrices and increased bone fragility associated with increased osteoclast recruitment [86]. 5.2.2.2 Other Osteopetrotic Models While bovine models of osteopetrosis have been reported [87–89], and avians infected with viruses develop an osteopetrosis-like disease [90, 91], only the bovine mechanical properties have been shown to be similar to those of humans [88]. Mineral properties have not been characterized. Other models that have an osteopetrotic phenotype, most often associated with genetic modiﬁcations leading to impaired osteoclast activity, are listed in Table 5.2. Mineral properties in the osteocalcin knockout [92] demonstrated the presence of smaller crystals, increased calciﬁed cartilage, and increased bone mass, like that found in osteopetrotic models. The only other knockout in which mineral properties have been described is the c-fos knockout [95], which has a ﬁvefold increase in bone volume, and a decreased, but more homogeneous, distribution of mineral.

5.2 Animal Models of Osteoporosis and Osteopetrosis

Fig. 5.2 Fourier transform infra-red (FTIR) imaging of the cathepsin K knockout (KO) mouse bones and age- and gender-matched wild-type (WT) controls. (a) A typical spectrum illustrating the peaks of interest from one pixel in the center of WT cortical bone. Parameters calculated for each pixel are mineral/matrix (integrated area of phosphate/amide I bands), crystallinity (peak height ratio of 1030 cm1 /1020 cm1 subbands), and collagen maturity (peak height

ratio of 1660 cm1 /1690 cm1 sub-bands). (b) Hyperspectral images of the mineral/ matrix ratio in growth plate of WT (top; width height ¼ 250 mm 74 mm) and KO (lower; 250 mm 50 mm). (c) Hyperspectral images of crystallinity in same growth plate sections; WT (top) and KO (lower). (d) Mineral/matrix in trabecular bone of WT (top; 2000 mm 1395 mm) and KO (lower; 370 mm 155 mm). (e) Crystallinity in cortical bone of WT (top; 310 mm 250 mm) and KO

In addition to these models, a limited number of examples exist of druginduced osteopetrosis associated with an inhibition of osteoclastic activity. Thus, osteopetrosis has been reported in association with prolonged high dosages of anti-resorptive bisphosphonates (in both rodents [102, 103] and humans [104]), with high prolonged doses of a phytoestrogen in rats [105], and with excessive ﬂuoride consumption [106].

69

70

5 Osteoporosis and Osteopetrosis Table 5.2 Animal models of osteoporosis and osteopetrosis.

Osteoporosis

Reference

Osteopetrosis

Reference(s)

Ovariectomized mouse

22

Spontaneous fracture mouse Steroid-induced rodent Immobilized rodent Senescence-accelerated mouse

29 25 24 27

Naturally occurring rats tl/tl ia/ia

68 69

Ovariectomized monkey Aging monkey Ovariectomized ewe Ovariectomized mini-pig Ovariectomized canine Biglycan KO Osteonectin KO LRP5 KO Sca-1 KO IL-4 transgenic Osteoprotegrin KO Noggin transgenic Sclerostin transgenic Osteogenesis imperfecta mice Mov 13 oim/oim Brtl fro/fro Type I transgenic

58 60 57 59 56 37 33 52 53 48 55 49 51

Naturally occurring mice op/op oc/oc mi/mi gl/gl Cathepsin K KO PU.1 KO c-src KO c-FOS KO Bcl-2 KO Osteocalcin KO RANK Ligand KO Fra-1 transgenic Klotho mutant mice TGF-beta binding protein KO

80 81 82, 83 75, 76 85 93 94 95 96 98 99 97 100 101

40 41 43 44 42

Osteopetrotic cows Avian osteopetrosis OI mouse with excess BP treatment

87–89 90, 91 102, 103

KO ¼ knockout; Klotho ¼ thread of life, an aging model; BP ¼ bisphosphonate.

5.3 The Cellular and Molecular Bases of Osteopetrosis and Osteoporosis 5.3.1 Osteoporosis

Osteoporosis occurs when there is an imbalance of the activities of osteoblasts and osteoclasts. Factors including environment, age, lifestyle, hormonal status, and genetics contribute to this heterogeneous group of diseases. Genetic segregation analysis, family and population studies, and evaluation of congenic mice

5.3 The Cellular and Molecular Bases of Osteopetrosis and Osteoporosis

[107, 108] have identiﬁed polymorphisms associated with osteoporosis, including those in the vitamin D receptor gene, a collagen type I gene, the interleukin gene, the Alox gene, and the estrogen receptor alpha gene, among others. Studies of congenic mice have linked candidate genes to bone fragility and high density [108–111] and associated candidate genes with bone fragility [109], but those genes identiﬁed in humans account for only @5% of the heritability of osteoporosis [112]. Many of the genes identiﬁed are involved in the regulation of the WNT pathway (Fig. 5.3), aﬀecting the activities of both osteoblasts and osteoclasts [113, 114]. As osteoporosis is a complex trait, however, it is unlikely that for most cases a single genetic defect will be identiﬁed.

Fig. 5.3 Osteoblast-osteoclast coupling and the role of the Wnt canonical pathway. Cartoon illustrating key factors coupling osteoblast and osteoclast activity. In osteoblasts and pre-osteoblasts, Wnt proteins bind to the transmembrane domain spanning Frizzled receptor (fzR) and LRP5/LRP6 coreceptors. This activates the Dishevelled protein (dsh) by over-phosphorylating it, leading to phosphorylation of b-catenin and degradation of the scaﬀold complex (GSK-3; glycogen synthase kinase 3), APC (adenomatous polyposis coli), and Axin. Phosphorylation of b-catenin by GSK3 stimulates the degradation of the complex. Stabilized b-catenin accumulates in the

cytosol, translocates to the nucleus (shaded), where it interacts with T-cell factor/lymphoid enhancer binding factor (TCF/LEF) transcription factors to mediate gene transcription, leading to osteoblastogenesis and the inhibition of both osteoblast and osteocyte apoptosis, an increased ratio of osteoprotegrin (OPG) to RANKL, and represses osteoclastogenesis. LRP5/6 coreceptor activity is inhibited by sclerostin (SOST gene product) and (Dkk). Interaction of the Dkk/LRP complex with kremen internalizes the complex for degradation, thus diminishing the number of Wnt coreceptors for signaling.

71

72

5 Osteoporosis and Osteopetrosis

A limited number of cases exist where a simple (Mendelian) genetic mutation leads to an altered bone phenotype. For example, a variety of mis-sense mutations leading to single amino acid changes in the amino terminus of the low-density lipoprotein receptor-related protein 5 (LRP5) results in an autosomal dominant high bone mass phenotype [115], without clinical features. Mice in which this gene is knocked out have a severe form of osteoporosis characterized by low bone mass, implying LRP5 might be involved in osteoporosis [116]. Mice lacking both LRP5 and LRP6 have severe limb deformities, and even lower bone mass [117]. A recent large population study of men with osteoporotic fractures has associated variations in both LRP5 and LRP6 [118]. The study of the LRP5/6 mutations is stimulating the analyses of the elements of the WNT-beta catenin pathway for additional osteoporosis-associated genetic defects. The eﬀects of many of these mutated or altered genes have been identiﬁed not only from the knockout and transgenic models, but also from studies of osteoblasts and osteoblast progenitor cells derived from these models. As reviewed elsewhere [119], focusing on the factors that regulate the formation of the osteoblast, the abnormalities in these factors, and the coupling of the osteoblast and osteoclast using cell culture systems, provides additional insights into the multitude of factors leading to osteoporosis.

5.3.2 Osteopetrosis

Osteopetrosis is associated with several genes [66, 120]. The genes associated with a human osteopetrosis encode proteins that participate in the functioning of the diﬀerentiated osteoclast, while some of the genes identiﬁed in animal models aﬀect osteoclastogenesis. Some of the genes that are associated with osteopetrosis based on such studies are listed in Table 5.3. These genes regulate osteoclast number, activity, and/or function. Osteoclastogenesis is dependent on the RANKL system and macrophage-colony stimulating factor [82]. Osteoblasts produce RANK-ligand which activates osteoclast formation and development by binding to a receptor on the osteoclast-forming cells. The osteoblasts also secrete a soluble factor, OPG, that binds to the same receptor and blocks osteoclastogenesis; this is part of the coupling between osteoblasts and osteoclasts. Additionally, matrix proteins present in bone are important for recruiting the osteoclasts to the bone surface. Once on the bone surface, osteoclasts degrade the mineral and organic matrices of bone by secreting hydrochloric acid and proteases (cathepsins, matrix metalloproteases, etc.) [134]. The mineral is dissolved by the hydrochloric acid, and the bicarbonate-chloride channel, its aﬃliated membrane protein OSM1, and the vacuolar Hþ -adenosine triphosphatase (V-ATPase) are required for removal of the mineral. Matrix metalloproteases, along with the protease cathepsin K, remove the organic matrix. The release of these proteins is important for the recruitment and diﬀerentiation of osteoblasts at the bone surface. These osteoblasts then ﬁll in the ‘‘pits’’ made by osteoclasts,

5.3 The Cellular and Molecular Bases of Osteopetrosis and Osteoporosis Table 5.3 The molecular basis of diﬀerent variants of osteopetrosis.

Gene with mutation

Description

Model

Disease characteristics

Reference(s)

TCIRG1a

a3 -subunit of V-type Hþ -ATPase

Human

Recessive osteopetrosis

121

Atp6Ia

Vacuolar-proton pump, Hþ transporting (member I)

Mouse

Severe osteopetrosis

122

116-kDa V-ATPasea

116 kDa osteoclastspeciﬁc vacuolar proton ATPase subunit

Mouse Human

Autosomal recessive lethal

123, 124

Syk

Syk tyrosine kinase

Mouse

Osteopetrotic

125

D11S1889

? – also associated with muscle function

Human

Autosomal dominant – generalized osteosclerosis, most pronounced at the cranial vault – no fractures

126

TRAP5b

Osteoclast-derived serum tartrate-resistant acid phosphatase 5b

Human

Albers–Schonberg disease

127

ClCN7

Chloride channel 7

Human Mouse

Albers–Schonberg disease (autosomal dominant osteopetrosis type II) – large ineﬀective osteoclasts – gray lethal mouse

127, 129

OSTM1

Osteopetrosis-associated transmembrane protein 1

Human Mouse

Albers–Schonberg disease

128

CAII

Carbonic anhydrase 2

Human

Cranial thickening, distal renal tubular acidosis and increased fractures

130

CSF

Macrophage stimulating factor

Mouse

Deﬁcient macrophages and osteoclasts

131

CATK

Cathepsin K

Mouse Human

Pycnodysostosis

83–85, 132, 133

GL

Gray lethal

Mouse Human

Severe, juvenile, thickened metaphyses

78

a OC116-KDa

(refers also to ATP6i, TCIRGI, a3) subunit of the osteoclast vacuolar Hþ -ATPase (V-Hþ -ATPase) proton pump.

73

74

5 Osteoporosis and Osteopetrosis

and consequently there is a ‘‘cycle’’ of formation and resorption, with factors such as parathyroid hormone (PTH), parathyroid hormone-related peptide (PTHrP), prostaglandins, vitamin D, and a variety of growth and transcription factors released by osteoclasts from the bone matrix activating osteoblast precursors to form a new bone matrix, which must then be mineralized.

5.4 Biomineralization in Osteopetrosis and Osteoporosis

An examination of preclinical models of osteopetrosis and osteoporosis reveals that, in general, the mineral (and mechanical) properties are distinct from those in age- and gender-matched controls. The fact that osteopetrosis is predominantly a disease of defective osteoclast activity provokes the thought of why bone formation is impaired. The most likely here answer is that, because osteoblastic activity and osteoclastic activity are coupled, when the osteoclasts cannot remove bone they are unable to release factors that recruit osteoblasts to deposit new bone (collagen and mineral). Thus, there is a decreased activity of both osteoblasts and osteoclasts. The exceptions are those cases with defects in the chloride channel, where tissue culture studies have indicated that isolated osteoblasts or marrow stromal cells can deposit bone in a normal fashion. However, even when osteoblastic activity is decreased, there may be additional bone formation as bone is deposited on calciﬁed cartilage spicules rather than on newly formed osteoid. This results in the presence of smaller mineral crystals and matrices which are rich in cartilage proteins. In osteoporosis, a similar argument holds. It is diﬃcult to determine which comes ﬁrst: accelerated remodeling in the presence of osteoblasts that cannot match the rate and therefore cannot rebuild bone that has been removed; or decreased formation with osteoclasts that try to remove the new, poorly mineralized matrix before secondary mineralization is complete. In either case there is a vicious cycle, as any new mineralized tissue that forms is rapidly removed, and what is left is an older matrix with larger crystals. The underlying mechanisms for mineral deposition in osteopetrosis and osteoporosis are most likely the same as in the normal individual (as reviewed elsewhere in this volume). The osteoblasts deposit an organic matrix (predominantly collagen), while non-collagenous proteins associated with the collagen provide nucleation sites for the ﬁrst deposition of mineral crystals, and other proteins regulate the growth of these crystals. Local calcium, phosphate and hydroxide ion concentrations, each regulated by the cell, determine the rates at which mineral can form. In the normal case there is a balance of mineral formation and removal. It is only when these cellular activities become unbalanced – as in osteopetrosis and osteoporosis – that the mechanisms are impaired and the bone properties compromised.

References

Acknowledgments

These studies were supported by NIH grants DE04141, AR037661, AR043125, and AR046121. The author is grateful to Drs. Peter Bullough, Edward DiCarlo, and Joseph M. Lane for their advising and inspiring her to pursue these investigations.

References 1 N.E. Lane, Am. J. Obstet. Gynecol. 2

3 4 5 6

7

8

9 10

11

12

2006, 194, S3–S11. St. Jude’s Hospital Web Site, 2005, http://www.stjude.org/print/ 0,2510,404_2166_2944,00.html. J. Bollerslev, Clin. Genet. 1987, 31, 86–90. J. Bollerslev, P.E. Andersen, Jr., Acta Orthop. Scand. 1989, 60, 110–112. J. Carolino, J.A. Perez, A. Popa, Am. Fam. Physician. 1998, 57, 1293–1296. Y. Gong, M. Vikkula, L. Boon, J. Liu, P. Beighton, R. Ramesar, L. Peltonen, H. Somer, T. Hirose, B. Dallapiccola, A. De Paepe, W. Swoboda, B. Zabel, A. Superti-Furga, B. Steinmann, H.G. Brunner, A. Jans, R.G. Boles, W. Adkins, M.J. van den Boogaard, B.R. Olsen, M.L. Warman, Am. J. Hum. Genet. 1996, 59, 146–151. H. Hartikka, O. Makitie, M. Mannikko, A.S. Doria, A. Daneman, W.G. Cole, L. Ala-Kokko, E.B. Sochett, J. Bone Miner. Res. 2005, 20, 783–789. O. Makitie, A.S. Doria, F. Henriques, W.G. Cole, S. Compeyrot, E. Silverman, R. Laxer, A. Daneman, E.B. Sochett, J. Pediatr. 2005, 146, 395–401. J. Dwyer, Am. J. Clin. Nutr. 2006, 83, 415S–420S. F. Lazner, M. Gowen, D. Pavasovic, I. Kola, Hum. Mol. Genet. 1999, 8, 1839–1846. H.C. Blair, L.J. Robinson, M. Zaidi, Biochem. Biophys. Res. Commun. 2005, 328, 728–738. G. Russell, G. Mueller, C. Shipman, P. Croucher, Novartis Found. Symp. 2001, 232, 251–267.

13 C.G. Keith, Arch. Ophthal. 1968, 79,

234–241. 14 P. Maroteaux, M. Lamy, JAMA 1965,

191, 715–717. 15 A.L. Boskey, in: S.C. Cowin (Ed.),

16 17 18 19 20 21 22 23

24 25

26

27

28

Bone Biomechanics, 3rd edn. CRC Press, Boca Raton, FL, 2001, pp. 5.1– 5.34. M.C. van der Meulen, K.J. Jepsen, B. Mikic, Bone 2001, 29, 101–104. A. Boskey, R. Mendelsohn, J. Biomed. Opt. 2005, 10, 031102–031105. A.L. Boskey, R. Mendelsohn, Vib. Spectrosc. 2005, 38, 107–114. A. Carden, M.D. Morris, J. Biomed. Opt. 2000, 5, 259–268. L.P. Schot, A.H. Schuurs, J. Steroid Biochem. Mol. Biol. 1990, 37, 461–465. D.N. Kalu, Bone Miner. 1991, 15, 175– 191. W.S. Jee, W. Yao, J. Musculoskelet. Neuronal. Interact. 2001, 1, 193–207. C.M. Bagi, S.C. Miller, B.M. Bowman, G.L. Blomstrom, E.P. France, Bone 1992, 13, 35–40. L. Vico, J. Bone Miner. Res. 2003, 18, 561–569. S.C. Manolagas, R.S. Weinstein, J. Bone Miner. Res. 1999, 14, 1061– 1066. N.E. Lane, W. Yao, M. Balooch, R.K. Nalla, G. Balooch, S. Habelitz, J.H. Kinney, L.F. Bonewald, J. Bone Miner. Res. 2006, 21, 466–476. T. Takeda, M. Hosokawa, K. Higuchi, M. Hosono, I. Akiguchi, H. Katoh, Arch. Gerontol. Geriatr. 1994, 19, 185– 192. M.J. Silva, M.D. Brodt, Z. Fan, J.Y. Rho, J. Biomech. 2004, 37, 1639–1646.

75

76

5 Osteoporosis and Osteopetrosis 29 S. Mohan, A. Kapoor, A. Singgih, Z.

30

31

32

33

34

35

36 37

38

39

40

41

42

Zhang, T. Taylor, H. Yu, R.B. Chadwick, Y.S. Chung, L.R. Donahue, C. Rosen, G.C. Crawford, J. Wergedal, D.J. Baylink, J. Bone Miner. Res. 2005, 20, 1597–1610. B. Bar-Shira-Maymon, R. Coleman, A. Cohen, E. Steinhagen-Thiessen, M. Silbermann, Calcif. Tissue Int. 1989, 44, 36–45. M.J. Silva, M.D. Brodt, M. Ko, Y. AbuAmer, J. Bone Miner. Res. 2005, 20, 419–427. M.J. Silva, M.D. Brodt, B. Wopenka, S. Thomopoulos, D. Williams, M.H. Wassen, M. Ko, N. Kusano, R.A. Bank, J. Bone Miner. Res. 2006, 21, 78–88. A.M. Delany, M. Amling, M. Priemel, C. Howe, R. Baron, E. Canalis, J. Clin. Invest. 2000, 105, 915–923. D.T. Gilmour, G.J. Lyon, M.B. Carlton, J.R. Sanes, J.M. Cunningham, J.R. Anderson, B.L. Hogan, M.J. Evans, W.H. Colledge, EMBO J. 1998, 17, 1860–1870. A.L. Boskey, D.J. Moore, M. Amling, E. Canalis, A.M. Delany, J. Bone Miner. Res. 2003, 18, 1005–1011. D. Faibish, S.M. Ott, A.L. Boskey, Clin. Orthop. Relat. Res. 2006, 443, 28–38. T. Xu, P. Bianco, L.W. Fisher, G. Longenecker, E. Smith, S. Goldstein, J. Bonadio, A. Boskey, A.M. Heegaard, B. Sommer, K. Satomura, P. Dominguez, C. Zhao, A.B. Kulkarni, P.G. Robey, M.F. Young, Nat. Genet. 1998, 20, 78–82. A.L. Boskey, M.F. Young, T. Kilts, K. Verdelis, Cells Tissues Organs. 2005, 181, 144–153. S. Viguet-Carrin, P. Garnero, P.D. Delmas, Osteoporosis Int. 2006, 17, 319–336. K.J. Jepsen, M.B. Schaﬄer, J.L. Kuhn, R.W. Goulet, J. Bonadio, S.A. Goldstein, J. Biomech. 1997, 30, 1141–1147. N.P. Camacho, L. Hou, T.R. Toledano, W.A. Ilg, C.F. Brayton, C.L. Raggio, L. Root, A.L. Boskey, J. Bone Miner. Res. 1999, 14, 264–272. R. Pereira, J.S. Khillan, H.J. Helminen, E.L. Hume, D.J. Prockop, J. Clin. Invest. 1993, 91, 709–716.

43 K.M. Kozloﬀ, A. Carden, C. Bergwitz,

44

45

46

47 48

49 50

51

52

53

54

55

A. Forlino, T.E. Uveges, M.D. Morris, J.C. Marini, S.A. Goldstein, J. Bone Miner. Res. 2004, 19, 614–622. D.O. Sillence, H.E. Ritchie, T. Dibbayawan, D. Eteson, K. Brown, Am. J. Med. Genet. 1993, 45, 276– 283. W. Traub, T. Arad, U. Vetter, S. Weiner, Matrix Biol. 1994, 14, 337– 345. U. Vetter, E.D. Eanes, J.B. Kopp, J.D. Termine, P.G. Robey, Calcif. Tissue Int. 1991, 49, 248–250. J.Y. Reginster, N. Burlet, Bone 2006, 38, S4–S9. D.B. Lewis, H.D. Liggitt, E.L. Eﬀmann, S.T. Motley, S.L. Teitelbaum, K.J. Jepsen, S.A. Goldstein, J. Bonadio, J. Carpenter, R.M. Perlmutter, Proc. Natl. Acad. Sci. USA 1993, 90, 11618–11622. E. Abe, J. Bone Miner. Res. 2000, 15, 663–673. X.B. Wu, Y. Li, A. Schneider, W. Yu, G. Rajendren, J. Iqbal, M. Yamamoto, M. Alam, L.J. Brunet, H.C. Blair, M. Zaidi, E. Abe, J. Clin. Invest. 2003, 112, 924–934. D.G. Winkler, M.K. Sutherland, J.C. Geoghegan, C. Yu, T. Hayes, J.E. Skonier, D. Shpektor, M. Jonas, B.R. Kovacevich, K. Staehling-Hampton, M. Appleby, M.E. Brunkow, J.A. Latham, EMBO J. 2003, 22, 6267– 6276. S.L. Ferrari, S. Deutsch, S.E. Antonarakis, Curr. Opin. Lipidol. 2005, 16, 207–214. M. Bonyadi, S.D. Waldman, D. Liu, J.E. Aubin, M.D. Grynpas, W.L. Stanford, Proc. Natl. Acad. Sci. USA 2003, 100, 5840–5845. N. Bucay, I. Sarosi, C.R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H.L. Tan, W. Xu, L.D. Lacey, W.J. Boyle, W.S. Simonet, Genes Dev. 1998, 12, 1260–1268. P. Pogoda, M. Priemel, A.F. Schilling, M. Gebauer, P. Catala-Lehnen, F. Barvencik, F.T. Beil, C. Munch, M. Rupprecht, C. Muldner, J.M. Rueger, T. Schinke, M. Amling, J. Bone Miner. Metab. 2005, 23(Suppl.), 97–102.

References 56 M.C. Monier-Faugere, Z. Geng, Q.

57 58

59 60

61

62

63

64

65

66

67 68

69 70

71

72

Qi, I. Arnala, H.H. Malluche, J. Bone Miner. Res. 1996, 11, 446–455. A.S. Turner, Vet. J. 2002, 163, 232– 239. L. Mosekilde, S.E. Weisbrode, J.A. Safron, H.F. Stills, M.L. Jankowsky, D.C. Ebert, C.C. Danielsen, C.H. Sogaard, A.F. Franks, M.L. Stevens, Bone 1993, 14, 379–382. C.P. Jerome, P.E. Peterson, Bone 2001, 29, 1–6. M.D. Grynpas, B. Huckell, K.P. Pritzker, R.G. Hancock, M.J. Kessler, P. R. Health Sci. J. 1989, 8, 197–204. M.D. Grynpas, P. Patterson-Allen, D.J. Simmons, Calcif. Tissue Int. 1986, 39, 57–62. K. Lundon, M. Dumitriu, M. Grynpas, Bone Miner. 1994, 24, 135– 149. S.J. Gadeleta, A.L. Boskey, E. Paschalis, C. Carlson, F. Menschik, T. Baldini, M. Peterson, C.M. Rimnac, Bone 2000, 27, 541–550. D. Pienkowski, T.M. Doers, M.C. Monier-Faugere, Z. Geng, N.P. Camacho, A.L. Boskey, H.H. Malluche, J. Bone Miner. Res. 1997, 12, 1936–1943. L.M. Miller, J. Tibrewala, C.S. Carlson, Cell Mol. Biol. (Noisy-legrand) 2000, 46, 1035–1044. W. Balemans, L. Van Wesenbeeck, W. Van Hul, Calcif. Tissue Int. 2005, 77, 263–274. S.C. Marks, Jr., Clin. Orthop. Relat. Res. 1984, 189, 239–263. A. Wojtowicz, A. Dziedzic-Goclawska, A. Kaminski, W. Stachowicz, K. Wojtowicz, S.C. Marks, Jr., M. Yamauchi, Bone 1997, 20, 127–132. A.L. Boskey, S.C. Marks, Jr., Calcif Tissue Int. 1985, 37, 287–292. N. Sakagami, N. Amizuka, M. Li, K. Takeuchi, M. Hoshino, M. Nakamura, K. Nozawa-Inoue, N. Udagawa, T. Maeda, Micron 2005, 36, 688–695. M.E. Jackson, V. Shalhoub, J.B. Lian, G.S. Stein, S.C. Marks, Jr., J. Cell Biochem. 1994, 55, 366–372. T.J. Martin, N.A. Sims, Trends Mol. Med. 2005, 11, 76–81.

73 J. Tuukkanen, A. Koivukangas, T.

74 75

76

77

78

79

80

81

82

83

84

85

86

Jamsa, K. Sundquist, C.A. Mackay, S.C. Marks, Jr., J. Bone Miner. Res. 2000, 15, 1905–1911. A. Boskey, Crit. Rev. Eukaryot. Gene Expr. 2003, 13, 109–116. U. Kornak, D. Kasper, M.R. Bo¨sl, E. Kaiser, M. Schweizer, A. Schulz, W. Friedrich, G. Delling, T.J. Jentsch, Cell 2001, 104, 205–215. N. Chalhoub, N. Benachenhou, V. Rajapurohitam, M. Pata, M. Ferron, A. Frattini, A. Villa, J. Vacher, Nat. Med. 2003, 9, 399–406. A. Ramirez, J. Faupel, I. Goebel, A. Stiller, S. Beyer, C. Stockle, C. Hasan, U. Bode, U. Kornak, C. Kubisch, Hum. Mutat. 2004, 23, 471–476. P. Quarello, M. Forni, L. Barberis, C. Deﬁlippi, M.F. Campagnoli, L. Silvestro, A. Frattini, N. Chalhoub, J. Vacher, U. Ramenghi, J. Bone Miner. Res. 2004, 19, 1194–1199. S. Schaller, K. Henriksen, M.G. Sorensen, M.A. Karsdal, Drug News Perspect. 2005, 18, 489–495. J.C. Scimeca, A. Franchi, C. Trojani, H. Parrinello, J. Grosgeorge, C. Robert, O. Jaillon, C. Poirier, P. Gaudray, G.F. Carle, Bone 2000, 26, 207–213. K. Roundy, R. Smith, J.J. Weis, J.H. Weis, J. Bone Miner. Res. 2003, 18, 278–288. T. Wada, T. Nakashima, N. Hiroshi, J.M. Penninger, Trends Mol. Med. 2006, 12, 17–25. W.S. Hou, D. Bromme, Y. Zhao, E. Mehler, C. Dushey, H. Weinstein, C.S. Miranda, C. Fraga, F. Greig, J. Carey, D.L. Rimoin, R.J. Desnick, B.D. Gelb, J. Clin. Invest. 1999, 103, 731–738. N. Fratzl-Zelman, A. Valenta, P. Roschger, A. Nader, B.D. Gelb, P. Fratzl, K. Klaushofer, J. Clin. Endocrinol. Metab. 2004, 89, 1538–1547. M. Gowen, F. Lazner, R. Dodds, R. Kapadia, J. Feild, M. Tavaria, I. Bertoncello, F. Drake, S. Zavarselk, I. Tellis, P. Hertzog, C. Debouck, I. Kola, J. Bone Miner. Res. 1999, 14, 1654–1663. C.Y. Li, K.J. Jepsen, R.J. Majeska, J. Zhang, R. Ni, B.D. Gelb, M.B.

77

78

5 Osteoporosis and Osteopetrosis

87 88

89 90 91 92

93

94 95 96

97

98

99

100

101

Schaﬄer, J. Bone Miner. Res. 2006, 21, 865–875. H.W. Leipold, J.E. Cook, Am. J. Pathol. 1977, 86, 745–748. R.B. Ashman, W.C. Van Buskirk, S.C. Cowin, P.M. Sandborn, M.K. Wells, J.C. Rice, Calcif. Tissue Int. 1985, 37, 73–76. D.C. Wolf, W.G. Van Alstine, J. Vet. Diagn. Invest. 1989, 1, 262–264. E.V. Schmidt, R.E. Smith, Am. J. Pathol. 1982, 106, 297–299. R.E. Smith, Curr. Top Microbiol. Immunol. 1982, 101, 75–94. R. Michael McClain, E. Wolz, A. Davidovich, F. Pfannkuch, J.A. Edwards, J. Bausch, Food Chem. Toxicol. 2006, 44, 56–80. M.M. Tondravi, S.R. McKercher, K. Anderson, J.M. Erdmann, M. Quiroz, R. Maki, S.L. Teitelbaum, Nature 1997, 386, 81–84. P. Soriano, C. Montgomery, R. Geske, A. Bradley, Cell 1991, 64, 693–702. O. Jacenko, BioEssays 1995, 17, 277– 281. G.G. McGill, M. Horstmann, H.R. Widlund, J. Du, G. Motyckova, E.K. Nishimura, Y.L. Lin, S. Ramaswamy, W. Avery, H.F. Ding, S.A. Jordan, I.J. Jackson, S.J. Korsmeyer, T.R. Golub, D.E. Fisher, Cell 2002, 109, 707–718. P. Roschger, K. Matsuo, B.M. Misof, W. Tesch, W. Jochum, E.F. Wagner, P. Fratzl, K. Klaushofer, Bone 2004, 34, 776–782. A.L. Boskey, S. Gadaleta, C. Gundberg, S.B. Doty, P. Ducy, G. Karsenty, Bone 1998, 23, 187–196. Y.Y. Kong, H. Yoshida, I. Sarosi, H.L. Tan, E. Timms, C. Capparelli, S. Morony, A.J. Oliveira-dos-Santos, G. Van, A. Itie, W. Khoo, A. Wakeham, C.R. Dunstan, D.L. Lacey, T.W. Mak, W.J. Boyle, J.M. Penninger, Nature 1999, 397, 315–323. T. Yamashita, I. Sekiya, N. Kawaguchi, K. Kashimada, A. Nifuji, Y.I. Nabeshima, M. Noda, J. Endocrinol. 2001, 168, 347–351. B. Dabovic, R. Levasseur, L. Zambuto, Y. Chen, G. Karsenty, D.B. Rifkin, Bone 2005, 37, 25–31.

102 R. Schenk, W.A. Merz, R. Muhlbauer,

103

104

105

106 107

108

109

110

111

112

113

114

115

R.G. Russell, H. Fleisch, Calcif. Tissue Res. 1973, 11, 196–214. N.P. Camacho, C.L. Raggio, S.B. Doty, L. Root, V. Zraick, W.A. Ilg, T.R. Toledano, A.L. Boskey, Calcif. Tissue Int. 2001, 69, 94–101. M.P. Whyte, D. Wenkert, K.L. Clements, W.H. McAlister, S. Mumm, N. Engl. J. Med. 2003, 349, 457–463. R. Michael McClain, E. Wolz, A. Davidovich, F. Pfannkuch, J.A. Edwards, J. Bausch, Food Chem. Toxicol. 2006, 44, 56–80. M. Grynpas, Calcif. Tissue Int. 1993, Suppl. 1, S57–S64. Q.Y. Huang, R.R. Recker, H.W. Deng, Osteoporos. Int. 2003, 14, 701–715. R.F. Klein, J. Allard, Z. Avnur, T. Nikolcheva, D. Rotstein, A.S. Carlos, M. Shea, R.V. Waters, J.K. Belknap, G. Peltz, E.S. Orwoll, Science 2004, 303, 229–232. Y. Yershov, T.H. Baldini, S. Villagomez, T. Young, M.L. Martin, R.S. Bockman, M.G. Peterson, R.D. Blank, J. Bone Miner. Res. 2001, 16, 992–1003. M. Shimizu, K. Higuchi, S. Kasai, T. Tsuboyama, M. Matsushita, T. Matsumura, S. Okudaira, M. Mori, A. Koizumi, T. Nakamura, M. Hosokawa, Mamm. Genome 2002, 13, 335–340. W. Gu, X. Li, K.H. Lau, B. Edderkaoui, L.R. Donahae, C.J. Rosen, W.G. Beamer, K.L. Shultz, A. Srivastava, S. Mohan, D.J. Baylink, Funct. Integr. Genomics 2002, 1, 375– 386. F.M. Williams, T.D. Spector, J. Musculoskelet. Neuronal Interact. 2006, 6, 27–35. D.A. Glass, II, P. Bialek, J.D. Ahn, M. Starbuck, M.S. Patel, H. Clevers, M.M. Taketo, F. Long, A.P. McMahon, R.A. Lang, G. Karsenty, Dev. Cell 2005, 8, 751–764. V. Krishnan, H.U. Bryant, O.A. Macdougald, J. Clin. Invest. 2006, 116, 1202–1209. L. Van Wesenbeeck, E. Cleiren, J. Gram, R.K. Beals, O. Benichou, D.

References

116

117

118

119 120

121

122

123

Scopelliti, L. Key, T. Renton, C. Bartels, Y. Gong, M.L. Warman, M.C. De Vernejoul, J. Bollerslev, W. Van Hul, Am. J. Hum. Genet. 2003, 72, 763–771. M. Kato, M.S. Patel, R. Levasseur, I. Lobov, B.H. Chang, D.A. Glass, II, C. Hartmann, L. Li, T.H. Hwang, C.F. Brayton, R.A. Lang, G. Karsenty, L. Chan, J. Cell Biol. 2002, 157, 303–314. S.L. Holmen, T.A. Giambernardi, C.R. Zylstra, B.D. Buckner-Berghuis, J.H. Resau, J.F. Hess, V. Glatt, M.L. Bouxsein, M. Ai, M.L. Warman, B.O. Williams, J. Bone Miner. Res. 2004, 19, 2033–2040. J.B. van Meurs, F. Rivadeneira, M. Jhamai, W. Hugens, A. Hofman, J.P. van Leeuwen, H.A. Pols, U.A. Gitterlinden, J. Bone Miner. Res. 2006, 21, 141–150. D. Rowe, A. Lichtler, Endocrine 2002, 17, 67–75. C. Sobacchi, A. Frattini, P. Orchard, O. Porras, I. Tezcan, M. Andolina, R. Babul-Hirji, I. Baric, N. Canham, D. Chitayat, S. Dupuis-Girod, I. Ellis, A. Etzioni, A. Fasth, A. Fisher, B. Gerritsen, V. Gulino, E. Horwitz, V. Klamroth, E. Lanino, M. Mirolo, A. Musio, G. Matthijs, S. Nonomaya, L.D. Notarangelo, H.D. Ochs, A. Superti Furga, J. Valiaho, J.L. van Hove, M. Vihinen, D. Vujic, P. Vezzoni, A. Villa, Hum. Mol. Genet. 2001, 10, 1767–1777. A. Frattini, P.J. Orchard, C. Sobacchi, S. Giliani, M. Abinun, J.P. Mattsson, D.J. Keeling, A.K. Andersson, P. Wallbrandt, L. Zecca, L.D. Notarangelo, P. Vezzoni, A. Villa, Nat. Genet. 2000, 25, 343–346. Y.P. Li, W. Chen, Y. Liang, E. Li, P. Stashenko, Nat. Genet. 1999, 23, 447– 451. J.C. Scimeca, A. Franchi, C. Trojani, H. Parrinello, J. Grosgeorge, C. Robert, O. Jaillon, C. Poirier, P.

124 125

126

127

128

129

130

131

132 133

134

Gaudray, G.F. Carle, Bone 2000, 26, 207–213. K.U. Ogbureke, Q. Zhao, Y.P. Li, Front. Biosci. 2005, 10, 2940–2954. A. Mocsai, M.B. Humphrey, J.A. Van Ziﬄe, Y. Hu, A. Burghardt, S.C. Spusta, S. Majumdar, L.L. Lanier, C.A. Lowell, M.C. Nakamura, Proc. Natl. Acad. Sci. USA 2004, 101, 6158– 6163. E. Van Hul, J. Gram, J. Bollerslev, L. Van Wesenbeeck, D. Mathysen, P.E. Andersen, F. Vanhoenacker, W. Van Hul, J. Bone Miner. Res. 2002, 17, 1111–1117. S.L. Alatalo, K.K. Ivaska, S.G. Waguespack, M.J. Econs, H.K. Vaananen, J.M. Halleen, Clin. Chem. 2004, 50, 883–890. P.F. Lange, L. Wartosch, T.J. Jentsch, J.C. Fuhrmann, Nature 2006, 440, 220–223. U. Kornak, A. Ostertag, S. Branger, O. Benichou, M.C. de Vernejoul, J. Clin. Endocrinol. Metab. 2006, 91, 995–1000. K.J. Borthwick, N. Kandemir, R. Topaloglu, U. Kornak, A. Bakkaloglu, N. Yordam, S. Ozen, H. Mocan, G.N. Shah, W.S. Sly, F.E. Karet, J. Med. Genet. 2003, 40, 115–121. W. Wiktor-Jedrzejczak, A. Bartocci, A.W. Ferrante, Jr., A. Ahmed-Ansari, K.W. Sell, J.W. Pollard, E.R. Stanley, Proc. Natl. Acad. Sci. USA 1990, 87, 4828–4832. G. Motyckova, D.E. Fisher, Curr. Mol. Med. 2002, 2, 407–421. P. Saftig, E. Hunziker, O. Wehmeyer, S. Jones, A. Boyde, W. Rommerskirch, J.D. Moritz, P. Schu, K. von Figura, Proc. Natl. Acad. Sci. USA 1998, 95, 13453–13458. K. Henriksen, M.G. Sorensen, R.H. Nielsen, J. Gram, S. Schaller, M.H. Dziegiel, V. Everts, J. Bollerslev, M.A. Karsdal, J. Bone Miner. Res. 2006, 21, 58–66.

79

81

6 Biomimetic Bone Substitution Materials Matthias Epple

Abstract

Bone defects are ﬁlled with diﬀerent materials, with the patient’s own bone (e.g., harvested from the iliac crest) being the ‘‘gold standard’’ in surgery. However, the insuﬃcient supply of this autologous bone in the case of major defects, and the need for a second operation to harvest this bone, have triggered research into semi- or fully synthetic bone substitution materials. In general, it is desirable that a bone defect is ﬁlled by newly grown bone after some time; that is, the implant material should be biodegradable and permit or (even better) stimulate the ingrowth of bone. Thus, the regeneration depends on the body’s own restorative capability, and it may be assumed that a material which has properties close to natural bone will be advantageous. Within the current concepts for bone substitution materials, the role of biomimetic – that is, ‘‘bone-resembling’’ – implants is highlighted. Keywords: implants.

surgery, bone, calcium phosphate, polymers, metals, osteoblasts,

6.1 The Clinical Need for Bone Substitution Materials

There are a number of clinical situations where lost bone has to be replaced. Typical examples are complicated fractures, explantation sites of bone tumors, bone loss around endoprostheses, and bone loss in the jaw around lost or extracted teeth [1–5]. In all of these cases, the bone defect must be ﬁlled with a suitable material which has a suﬃcient mechanical stability, does not cause chemically adverse reactions (e.g., the release of acids or toxic metals), and does not lead to an adverse biological reaction (e.g., inﬂammation or an allergic reaction). Ideally, the material should be biodegradable and eventually be replaced by newly grown

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

82

6 Biomimetic Bone Substitution Materials

bone, leading to a restitutio ad integrum of the bone defect. Note that bone as a living tissue is undergoing a constant biomineralization, the so-called ‘‘remodeling’’; that is, osteoclasts are continuously resorbing bone and osteoblasts are simultaneously forming new bone [6]. Therefore, the process of the replacement of an implant by new bone must rely on the same processes as the biomineralization in healthy bone. Consequently, the most straightforward approach is the use of explanted native bone from a donor site of the same patient, and its re-implantation into the defect site. Usually, such a transplant is well accepted by the body and rapidly integrated into the surrounding bone tissue. Therefore, this autologous bone implant is usually referred to as the ‘‘gold standard’’ in clinical medicine. However, the fact that there is not much spare bone in the human body (typically, bone is harvested from the iliac crest), and that there is sometimes a rapid resorption of the implanted bone (faster than the growth of new bone), sets a limit to this approach. The need for an additional operation to explant the bone from the donor site, which often is accompanied by additional pain for the patient, is another restriction. Due to high clinical demand, many diﬀerent semi- or fully synthetic materials have been proposed to treat bone defects. Bone from other donors (from bone banks, so-called ‘‘allogenic transplants’’) can also be implanted, but this is limited by the need to suppress an adverse immune reaction and to exclude transmitted infections. Nevertheless, such allogenic transplants still constitute a considerable part of clinical practice, due also to the fact that there is often ‘‘spare bone’’ available, for example from the removed femoral heads after the implantation of an artiﬁcial hip joint. The next logical step is to implant bone from animals – the so-called ‘‘xenogenic transplants’’ – where there is a potentially unlimited supply. However, concerns about immune reactions and infections are even more severe in this case, and therefore xenogenic bone can only be used after extensive chemical and/or thermal treatment in order to exclude all hazardous biological material. Thus, fully synthetic biomaterials for bone substitution oﬀer great potential, provided that they fulﬁll all mechanical, chemical and biological requirements. The high economic opportunities associated with this process have resulted in many diﬀerent biomaterials for this application.

6.2 Synthetic Materials Used for Bone Substitution

Materials science is a well-developed ﬁeld of science, and in recent years many materials have been proposed as bone substitution materials, with a more or less biological relationship to the original material, living bone. The main requirements for a bone substitute include:

6.2 Synthetic Materials Used for Bone Substitution

A suﬃcient mechanical stability which, ideally, is identical to that of bone. A low mechanical stability leads to disintegration and undesired destabilization of the implantation site. Conversely, a high mechanical stability, characterized by a high stiﬀness (high module of elasticity), leads to stress-shielding of the surrounding bone and potentially to bone loss around the implant. A biodegradability which is adapted to the biological requirements; that is, it should be rapid enough to allow new bone to grow into the implantation site, but not so rapid that a mechanically weak point results. Ideally, the combined mechanical strength of the implant and of the ingrowing bone should remain constant throughout the regenerative process. A high porosity, which allows the ingrowth of bone tissue during regeneration. This requires typical pore sizes of a few hundred micrometers, which were shown to be well-suited to cell invasion [7]. For a good bone ingrowth the pores should be interconnected, as in native bone (i.e., not isolated). The absence of any chemical or biological irritation by substances which are released due to corrosion or degradation. Typical chemically adverse reactions are the release of immunogenic metal ions (e.g., nickel) and the release of lactic acid during the degradation of poly(lactic acid). An absence of the release of biologically adverse substances, such as immunogenic (e.g., proteins), infectious (e.g., viruses, bacteria), or toxic compounds. A possibility to adjust the shape of the implant during the operation in order to fulﬁll the surgeons’ requirements. Another possibility is a pre-shaping of the implant before surgery, based on a previous geometric analysis of the bone defect (e.g., by microtomography). A good sterilizability, storability, and processability. A price which is low enough to permit a clinical application.

These requirements are clearly manifold, and impossible to fulﬁll with any single material. This may explain why autologous bone is still the ‘‘gold standard’’, and why synthetic approaches try to mimic the original bone as closely as possible. Some examples of fully synthetic commercial bone substitution materials are shown in Figure 6.1.

83

84

6 Biomimetic Bone Substitution Materials

Fig. 6.1 Some commercial synthetic bone substitution materials. (A) A granulate of hydroxyapatite, Ca10 (PO4 )6 (OH)2 which is used to ﬁll bone defects, sometimes after mixing with blood or bone chips. (B) A bone cement consisting of a solid mixture of calcium phosphates in a small elastic balloon, an aqueous NaCl solution, and a syringe to mix these components (from left to right). The NaCl solution is added to the calcium phosphate powder in the balloon and temporary dissolution, followed by precipitation of calcium phosphate, occurs. This dispersion can be introduced into a bone defect where it hardens in situ.

(C) A granulate of b-tricalcium phosphate (b-Ca3 (PO4 )2 ; b-TCP) for the same application as in case (A). (D) Cylinders of nickeltitanium (NiTi), which are used as nonbiodegradable, weight-bearing implants in the spine. (E) A paste of nano-hydroxyapatite in water with high mineral content which can be injected into bone defects. Note that hardening does not occur in this case because no precipitation occurs, and because the paste will not dry in the ‘‘moist’’ biological environment. (F) Porous blocks of b-TCP which are individually machined and can be used as three-dimensional defect-ﬁlling materials.

6.3 Ceramics and Bone Cements

Bone and dentin mineral consists of nanocrystals of carbonated apatite, the formula of which may expressed simpliﬁed as Ca10-x (PO4 )6-x (CO3 )x (OH)2 , and sometimes denoted as dahllite [8, 9]. Interestingly, the same mineral in terms of composition and crystal size was found in atherosclerotic lesions [10], pointing to similar formation pathways. A whole range of diﬀerent calcium phosphates has been identiﬁed [9, 11–13], and in general all are biocompatible due to their similarity to bone and tooth mineral. Consequently, many attempts have been un-

6.3 Ceramics and Bone Cements

dertaken to use calcium phosphates as bone substitution materials, and today many diﬀerent products are available internationally, though most are based on hydroxyapatite (HAP or HA), Ca10 (PO4 )6 (OH)2 , or b-tricalcium phosphate (bTCP), or b-Ca3 (PO4 )2 [14]. Some current problems involve the inherent brittleness of ceramics, as this may lead to mechanical failure at the operation site and a sometimes inadequate biodegradation (often too slow). The ﬁrst of these problems can be solved by using sintering processes which increase the hardness of a material; this in turn often slows down the rate of biodegradation [15–18]. The sometimes slow rate of biodegradation can be understood when the biological mechanism for degradation of calcium phosphates is considered. Bone tissue is continually undergoing a permanent remodeling process [19]; that is, old bone is being resorbed by osteoclasts and new bone is being formed by osteoblasts [5, 20]. Osteoclasts are also responsible for the biodegradation of calcium phosphate implants [18, 21], and function by creating a secluded compartment between the cell and the bone, characterized by the so-called ‘‘ruﬄed border’’ of the osteoclast. An acidic pH of about 4, caused by the presence of hydrochloric acid, is created by proton and chloride pumps [22–25], and this in turn leads to a dissolution of the nanoscopic bone mineral crystals, because all calcium phosphates are soluble in acids [9, 26]. Because calcined ceramics consist of microcrystals instead of nanocrystals, they have a lower solubility and are dissolved only slowly under the conditions of osteoclastic resorption (compare the morphology of the material in Figs. 6.3 and 6.4 at high magniﬁcation). This was also shown experimentally in vitro [16, 27–32]. It is therefore important to assure a good solubility of the calcium phosphate under the conditions of osteoclastic dissolution [18], either by choosing a phase with higher solubility (e.g., b-TCP [33], octacalcium phosphate, OCP [34], or carbonated apatite [22, 35–38]), or by keeping the size of the crystals within the nanometer range [38–42]. Slowly degrading ceramic implants may cause problems if further traumatic fractures occur at the same site [43]. Bone cements consisting of carbonated apatite can be precipitated in situ by mixing powders and solutions which contain the components of carbonated apatite, and injecting the resulting paste into the bone defect [44–46]. Such bone cements have gained some clinical acceptance and show a good biodegradability [31]. Glass ceramics (‘‘Bioglasses’’), which are based on ‘‘CaOP2 O5 SiO2 ’’ are also used as bone substitution materials, with an adjustable range of properties, depending on the composition [47, 48]. In summary, it appears to be advantageous to come as close as possible to the biological example – that is, to nanosized carbonated apatite. Unfortunately, although it is possible to prepare bone mineral-like calcium phosphate nanoparticles [40, 49, 50], it has not yet been possible to mimic completely the ﬁnely structured composite of collagen-calcium phosphate by synthetic means, despite promising attempts [51–58]. After all, this hierarchical structure is responsible for the exceptional mechanical properties of bone ! [6, 59, 60]. Although calcium phosphates are by far the most important ceramics to be used as bone substitution materials, it has also been proposed to use calcium sulfate [61] or calcium carbonate [62, 63] in such a role. Unfortunately, however,

85

86

6 Biomimetic Bone Substitution Materials

both materials show a rather rapid biodegradation, due either to their high solubility either in general (calcium sulfate) or under acidic conditions (calcium carbonate).

6.4 Polymers

Polymers have a greater elasticity than ceramics, and furthermore their properties can be ﬁne-tuned over a wide range by varying the type of polymer, its chain length and crystallinity, and/or by preparing co-polymers and by blending two or more polymers [64–66]. In particular, this permits the adjustment of mechanical properties and of the rate of biodegradation [67]. If biodegradable [68] polymers are to be used, it must be ensured that the degradation products – that is, the monomers and oligomers of the polymer – do not induce any adverse reactions. Consequently, the number of polymers in clinical use is limited. For bone substitution, two main synthetic classes of materials are used: (i) polymethylmethacrylate (PMMA) and its derivatives; and (ii) various polyesters from hydroxycarboxylic acids [67]. 6.4.1 PMMA-Based Materials

Bone cements based on PMMA are applied as a paste of the liquid monomer and solid particles of oligomers, together with an initiator [69, 70]. The material polymerizes on the implantation site and integrates as a tough, very stable implant. It is often applied to ﬁx total hip endoprostheses in the femoral leg. The drawbacks of PMMA use are the heat evolution during polymerization, which may lead to necrosis of the surrounding bone, and the release of small amounts of oligomers into the surrounding tissue. PMMA is not biodegradable and does not induce the ongrowth of bone. 6.4.2 Polyester-Based Materials

Polyesters such as polyglycolide (polyglycolic acid; PGA) and polylactide (polylactic acid; PLA) and co-polyesters of these, are applied in medicine as biodegradable implant materials [67, 71–74]; they have also been studied extensively as bone substitution materials [75–80] and as scaﬀolds for tissue engineering [81–85]. They show good mechanical characteristics (they are more elastic than ceramics) and degrade to the corresponding hydroxycarboxylic acids, which are easily metabolized [71]. Occasionally, an accumulation of the acidic degradation products has been observed which led to serious inﬂammation and damage of the surrounding tissues [86, 87]. However, this can be countered by adding basic salts such as calcium

6.7 Bone Substitutes of Biological Origin

carbonate or carbonated apatite to the materials [88–91]; this has the added advantage of making the material more biocompatible.

6.5 Metals

Metals are often used in medicine as surgical materials, mainly as plates, nails or screws. For bone substitution, very few (and usually porous [92]) materials have been proposed, including porous titanium [93], porous nickel-titanium (NiTi) [94–97], porous tantalum [98], and magnesium alloys [99]. Of course, they are all not biomimetic in terms of chemical composition because there are no elemental metals in the human body, but they can show a good biocompatibility in bone contact, especially when they are coated with calcium phosphate (the bone mineral [100–106]; see also Chapter 7). Usually, these are permanent implants in which degradation or corrosion is not desirable, although during recent years a number of magnesium alloys have been proposed which are aimed to degrade in the body in order to make room for the ingrowing bone [99].

6.6 Composites

Bone is a composite material which consists mainly of collagen (‘‘an elastic polymer’’) and calcium phosphate (‘‘a tough ceramic’’). Its extraordinary mechanical properties of being simultaneously elastic and hard [6, 107] suggest that just one replacement material alone will not be able to fulﬁll all requirements. Consequently, investigations are ongoing in order to develop composite materials from polymers and ceramics which are aimed to mimic the properties of natural bone. Many diﬀerent polymers have been combined with calcium phosphate as ceramic ﬁller material; calcium phosphate is usually chosen because of its excellent biocompatibility, its biodegradability, and its good mechanical properties when used as a ﬁller material. For a review on polymer/calcium phosphate composite materials used for bone substitution, the reader is referred to Ref. [108].

6.7 Bone Substitutes of Biological Origin

Nacre is well received by the body upon implantation, and may even contain osteoinductive substances [62, 109, 110]. Other biological materials comprise corals [111–13], chemically or thermally treated bone xenografts [114], and hydrothermally treated calcareous algae [14, 115]. The latter material consists of a calcium carbonate skeleton which is converted into hydroxyapatite or b-tricalcium phosphate by hydrothermal treatment with ammonium phosphate [115]. The external

87

88

6 Biomimetic Bone Substitution Materials Fig. 6.2 A bone substitution material which was obtained by hydrothermal conversion of the calcareous skeleton of red algae into hydroxyapatite. Note the preservation of the porous skeleton of the algae in the ﬁnal product.

shape and internal porosity are well preserved (Fig. 6.2). This material demonstrated a good clinical success, especially in oral surgery, and its performance can be improved my mixing with autologous bone chips [116, 117]. Juniper wood has also been proposed as a bone substitution material [118]. It is important to note that all materials of biological origin must be suitably conditioned before any implantation due to the risk of infection, and this may involve chemical and/or thermal treatment. A bone substitution material which was derived from bovine bone by chemical and thermal treatment is shown in Figure 6.3 [14]. Of particular note is the porous structure, as well as the non-resolved microstructure at the highest magniﬁcation, which shows no individual crystals and a morphology which is close to that of original bone. The sample still consists of collagen and calcium phosphate nanocrystals. A bone substitution material obtained by the calcination of bovine bone is shown in Figure 6.4 [14]. The interconnecting pore structure of the original bone is still present, but the former nanocrystals have sintered into microcrystals (these are clearly visible at the highest magniﬁcation). The driving forces for these developments were both the chemical similarity to bone (mineral) and the morphological similarity to cancellous bone, which allows an easy ingrowth of bone. The graded nature of bone (cortical and cancellous bone) is also an important property to reproduce in bone substitution materials, not only to control the

6.7 Bone Substitutes of Biological Origin Fig. 6.3 A bone substitution material which was obtained by thermal and chemical treatment of bovine bone (to remove all infectious components). Chemically, the sample consists of collagen and nanocrystalline calcium phosphate. Morphologically, the interconnecting porosity of the original bone is preserved.

Fig. 6.4 A bone substitution material which was obtained by calcination of bovine bone. It consists exclusively of inorganic components, mainly hydroxyapatite from the original bone mineral, and preserves the interconnecting porosity of natural bone. On the microscale, however, it is clear that the calcium phosphate nanocrystals have sintered into microcrystals.

89

90

6 Biomimetic Bone Substitution Materials

mechanical properties but also to achieve spatially diﬀerent rates of biodegradation [119, 120].

6.8 Biological Functionalization of Synthetic Materials

One concept which goes beyond the pure chemical, structural or morphological variation of a synthetic biomaterial involves its functionalization with biologically active molecules. These can be either substances which enhance bone growth after being released (e.g., morphogenetic proteins [121–123]; see also Chapter 2), or substances which attract cells to the implant surface, making use of special recognition sequences such as integrin ligands ([124, 125]; see also Chapter 8). These approaches must be distinguished on a conceptual basis, however: the ﬁrst method induces bone formation due to drug release in the vicinity of the implant, whereas the second method modiﬁes the surface only – that is, it acts in a spatially deﬁned manner. Synthetic bone substitution materials can be used as carriers for biomolecules, with bone growth being induced in their vicinity during release (see, e.g., Refs. [112, 116, 126–130]).

6.9 An Example of a Synthetic Biomimetic Bone Substitution Material

The general concepts behind the preparation of biomimetic materials involve biodegradable materials, materials with a ‘‘bone-like’’ porous structure, and a mechanical performance resembling that of bone. If a biological functionalization is set aside, this must be achieved using synthetic or biologically derived materials. Such an example is described in the following section. A combined eﬀort made by chemists, engineers, and clinicians led to an individually shaped skull implant which consisted of polylactides and calcium phosphate/calcium carbonate (fully biodegradable), and was structured according to the biological requirements at the implantation site [79, 80, 131]. The implant was designed to mimic the cancellous and cortical structure of bone; that is, it is a functionally graded material like natural bone [6] (Fig. 6.5). The porous inside (‘‘cancellous bone’’) consisted of rapidly degrading poly(dl-lactide) and calcium carbonate, pointed towards the dura mater, and permitted the ingrowth of bone from that region. The compact exterior (‘‘cortical bone’’), consisted of slowly degrading poly(l-lactide) and nanoscopic carbonated apatite. An integrated micro computed tomography/computer-aided manufacturing process chain (‘‘TICC’’; [79, 80]) permitted a patient-speciﬁc shaping of the implant. Subsequent experiments conducted in animals showed the implant to have excellent biocompatibility and to be almost completely degraded and/or substituted by newly grown bone [80].

Acknowledgments

Fig. 6.5 A biomimetic bone substitution implant for the treatment of skull defects. The part pointing towards the brain (top) consists of porous poly(dl-lactide), together with calcium carbonate, the part pointing towards the skin (bottom) consists of compact poly(l-lactide), together with nano-calcium phosphate.

6.10 Conclusions and Future Developments

While the in-vitro preparation of new bone by tissue engineering remains in its infancy, despite considerable eﬀorts having been made (see Chapter 10), surgeons will continue to rely either on autologous bone grafts or on synthetic biomaterials. This will involve strategies that are known from Nature, and include materials which are biodegradable, mechanically and morphologically optimized, and possibly also nano-structured. Although this can be achieved by using synthetic materials, in future the biological functionalization of implants will attract an increasing amount of attention. This functionalization will be introduced not only onto the implant surface in order to achieve interaction with the surrounding tissue and cells, but also internally, perhaps to incorporate drugs and biomolecules that will promote healing and induce future bone growth. Clearly, these modiﬁcations will involve more costly preparations and formulations than are presently in use, and competition will no doubt evolve between the design and development of biologically optimized implants and cost-limiting procedures of the social security systems.

Acknowledgments

The biomimetic skull implant was developed in a major project funded by the Deutsche Forschungsgemeinschaft (DFG). The author wishes to acknowledge his partners and co-workers, including Thomas Annen, Harald Euﬁnger, Chris-

91

92

6 Biomimetic Bone Substitution Materials

tian Rasche, Carsten Schiller, Inge Schmitz, Michael Wehmo¨ller, and Stephan Weihe, all of whom have contributed to these results.

References 1 R.A. Kenley, K. Yim, J. Abrams, E.

2 3

4 5

6 7

8 9

10

11

12

13 14 15

16

Ron, T. Turek, L.J. Marden, J.O. Hollinger, Pharmaceut. Res. 1993, 10, 1393–1401. J.M. Rueger, Orthopa¨de 1998, 27, 72– 79. A.S. Greenwald, S.D. Boden, V.M. Goldberg, Y. Khan, C.T. Laurencin, R.N. Rosier, J. Bone Joint Surg. 2001, 83A, 98–103. S.V. Dorozhkin, M. Epple, Angew. Chem. 2002, 114, 3260–3277. M. Amling, A.F. Schilling, P. Pogoda, M. Priemel, J.M. Rueger, Eur. J. Trauma 2006, 32, 102–106. S. Weiner, H.D. Wagner, Annu. Rev. Mater. Sci. 1998, 28, 271–298. S.L. Ishaug, G.M. Crane, M.J. Miller, A.W. Yasko, M.J. Yaszemski, A.G. Mikos, J. Biomed. Mater. Res 1997, 36, 17–28. U. Akiva, H.D. Wagner, S. Weiner, J. Mater. Sci. 1998, 33, 1497–1509. S.V. Dorozhkin, M. Epple, Angew. Chem. Int. Ed. 2002, 41, 3130– 3146. A. Becker, M. Epple, K.M. Mu¨ller, I. Schmitz, J. Biol. Inorg. Chem. 2004, 98, 2032–2038. R.Z. LeGeros, Calcium phosphates in oral biology and medicine. Karger, Basel, 1991. J.C. Elliot, Structure and chemistry of the apatites and other calcium orthophosphates, Vol. 18. Elsevier, Amsterdam, 1994. R.Z. LeGeros, Clin. Orthop. Rel. Res. 2002, 395, 81–98. D. Tadic, M. Epple, Biomaterials 2004, 25, 987–994. J.M. Rueger, W. Linhart, D. Sommerfeldt, Orthopa¨de 1998, 27, 89–95. Y. Doi, H. Iwanaga, T. Shibutani, Y. Morikawa, Y. Iwayama, J. Biomed. Mater. Res. 1999, 47, 424–433.

17 D. Briem, W. Linhart, W. Lehmann,

18

19

20

21

22

23 24

25 26

27

28

29

30 31

N.M. Meenen, J.M. Rueger, Unfallchirurg 2002, 105, 128–133. A.F. Schilling, W. Linhart, S. Filke, M. Gebauer, T. Schinke, J.M. Rueger, M. Amling, Biomaterials 2004, 25, 3963–3972. R. Weinkamer, M.A. Hartmann, Y. Brechet, P. Fratzl, Phys. Rev. Lett. 2004, 93, 228102. P. Ducy, M. Amling, S. Takeda, M. Priemel, A.F. Schilling, F.T. Beil, J. Shen, C. Vinson, J.M. Rueger, G. Karsenty, Cell 2000, 100, 197–207. A.F. Schilling, S. Filke, S. Brink, H. Korbmacher, M. Amling, J.M. Rueger, Eur. J. Trauma 2006, 32, 107–113. Y. Doi, T. Shibutani, Y. Moriwaki, T. Kajimoto, Y. Iwayama, J. Biomed. Mater. Res. 1998, 39, 603–610. G. Vaes, Clin. Orthoped. Rel. Res. 1988, 231, 239–271. S.L. Teitelbaum, M.M. Tondravi, F.P. Ross, J. Leucocyte Biol. 1997, 61, 381– 388. S.L. Teitelbaum, Science 2000, 289, 1504–1508. P. Koutsoukos, Z. Amjad, M.B. Tomson, G.H. Nancollas, J. Am. Chem. Soc. 1980, 102, 1553–1557. S. Yamada, D. Heymann, J.M. Bouler, G. Daculsi, Biomaterials 1997, 18, 1037–1041. R.P. Shellis, A.R. Lee, R.M. Wilson, J. Coll. Interface Sci. 1999, 218, 351– 358. A. Leeuwenburgh, P. Layrolle, F. Barre`re, J.D. de Bruijn, J. Schoonman, C.A. van Blitterswijk, K. de Groot, J. Biomed. Mater. Res. 2001, 56, 208–215. D. Tadic, M. Epple, Eur. J. Trauma 2002, 28, 136–137. C.R. Hankermeyer, K.L. Ohashi, D.C. Delaney, J. Ross, B.R. Constantz, Biomaterials 2002, 23, 743–750.

References 32 M.T. Fulmer, I.C. Ison, C.R.

33 34

35

36 37

38 39

40 41

42

43

44

45

46

47 48

49

Hankermayer, B.R. Constantz, J. Ross, Biomaterials 2002, 23, 751– 755. F. Peters, D. Reif, Mat.-wiss. u. Werkstoﬀtech. 2004, 35, 203–207. S. Kamakura, Y. Sasano, T. Shimizu, K. Hatori, O. Suzuki, M. Kagayama, K. Motegi, J. Biomed. Mater. Res. 2002, 59, 29–34. J.C. Merry, I.R. Gibson, S.M. Best, W. Bonﬁeld, J. Mater. Sci. Mater. Med. 1998, 9, 779–783. I.R. Gibson, W. Bonﬁeld, J. Biomed. Mater. Res. 2002, 59, 697–708. W. Suchanek, P. Shuk, K. Byrappa, R.E. Riman, K.S. TenHuisen, V.F. Janas, Biomaterials 2002, 23, 699–710. D. Tadic, F. Peters, M. Epple, Biomaterials 2002, 23, 2553–2559. L.M. Rodriguez-Lorenzo, M. ValletRegi, J.M.F. Ferreira, Biomaterials 2001, 22, 583–588. D. Tadic, M. Epple, Biomaterials 2003, 24, 4565–4571. D. Tadic, A. Veresov, V.I. Putlayev, M. Epple, Mat.-wiss. u. Werkstoﬀtech. 2003, 34, 1048–1051. D. Tadic, F. Beckmann, K. Schwarz, M. Epple, Biomaterials 2004, 25, 3335–3340. W. Linhart, D. Briem, M. Amling, J.M. Rueger, J. Windolf, Unfallchirurg 2004, 107, 154–157. B.R. Constantz, I.C. Ison, M.T. Fulmer, R.D. Poser, S.T. Smith, M. VanWagoner, J. Ross, S.A. Goldstein, J.B. Jupiter, D.I. Rosenthal, Science 1995, 267, 1796–1799. E. Fernandez, F.J. Gil, M.P. Ginebra, F.C.M. Driessens, J.A. Planell, S.M. Best, J. Mater. Sci. Mater. Med. 1999, 10, 169–176. E. Fernandez, F.J. Gil, M.P. Ginebra, F.C.M. Driessens, J.A. Planell, S.M. Best, J. Mater. Sci. Mater. Med. 1999, 10, 177–183. L.L. Hench, J. Am. Ceram. Soc. 1998, 81, 1705–1728. A.A. Ignatius, C. Schmidt, D. Kaspar, L.E. Claes, J. Biomed. Mater. Res. 2001, 55, 285–294. T. Welzel, W. Meyer-Zaika, M. Epple, Chem. Commun. 2004, 1204–1205.

50 A. Tampieri, G. Celotti, E. Landi,

51

52

53 54

55

56

57

58 59

60

61

62 63

64

65 66

Anal. Bioanal. Chem. 2005, 381, 568– 576. Y. Doi, T. Horiguchi, Y. Moriwaki, H. Kitago, T. Kajimoto, Y. Iwayama, J. Biomed. Mater. Res. 1996, 31, 43–49. C. Du, F.Z. Cui, Q.L. Feng, X.D. Zhu, K. de Groot, J. Biomed. Mater. Res. 1998, 42, 540–548. J.D. Hartgerink, E. Beniash, S.I. Stupp, Science 2001, 294, 1684–1688. D. Lickorish, J.A.M. Ramshaw, J.A. Werkmeister, V. Glattauer, C.R. Howlett, J. Biomed. Mater. Res. 2004, 68A, 19–27. S. Liao, F. Watari, M. Uo, S. Ohkawa, K. Tamura, W. Wang, F.Z. Cui, J. Biomed Mater. Res. Appl. Biomater. 2005, 74B, 817–821. H. Ehrlich, T. Douglas, D. Scharnweber, T. Hanke, R. Born, S. Bierbaum, H. Worch, Z. allg. anorg. Chem. 2005, 631, 1825–1830. A. Yokoyama, M. Gelinsky, T. Kawasaki, T. Kohgo, U. Koenig, W. Pompe, F. Watari, J. Biomed. Mater. Res. Appl. Biomater. 2005, 75B, 464– 472. E. Sachlos, D. Gotora, J.T. Czernuska, Tissue Eng. 2006, 12, 2479–2487. H. Peterlik, P. Roschger, K. Klaushofer, P. Fratzl, Nature Mater. 2006, 5, 52–55. H.S. Gupta, U. Stachewicz, W. Wagermaier, P. Roschger, H.D. Wagner, P. Fratzl, J. Mater. Res. 2006, 21, 1913–1921. M.A. Rauschmann, T.A. Wichelhaus, V. Stirnal, E. Dingeldein, L. Zichner, R. Schnettler, V. Alt, Biomaterials 2005, 26, 2677–2684. P. Westbrook, F. Marin, Nature 1998, 392, 861–862. Y.W. Kim, J.J. Kim, Y.H. Kim, J.Y. Rho, Biomaterials 2001, 23, 2089– 2096. A.G. Mikos, A.J. Thorsen, L.A. Czerwonka, Y. Bao, R. Langer, D.N. Winslow, J.P. Vacanti, Polymer 1994, 35, 1068–1077. R. Langer, Acc. Chem. Res. 2000, 33, 94–101. R. Langer, N.A. Peppas, A. I. Ch. E. J. 2003, 49, 2990–3006.

93

94

6 Biomimetic Bone Substitution Materials ¨ stman, H. Pihlajama¨ki, 67 O. Bo 68

69 70

71

72 73 74 75 76

77 78

79

80

81

82 83

84

Biomaterials 2000, 21, 2615–2621. F. von Burkersroda, L. Schedl, A. Go¨pferich, Biomaterials 2002, 23, 4221–4231. K.D. Ku¨hn, Bone cements. Springer, Berlin, 2000. B.D. Ratner, A.S. Hoﬀman, F.J. Schoen, Biomaterials Science. An Introduction to Materials in Medicine. Academic Press, 2004. M. Vert, G. Schwarch, J. Coudane, J. Macromol. Sci. Pure Appl. Chem. 1995, A32, 787–796. N.A. Ashammakhi, J. Biomed. Mater. Res. 1996, 33, 297–303. O.M. Bo¨stman, Clin. Orthop. Rel. Res. 1996, 329, 233–239. E. Chiellini, R. Solaro, Adv. Mater. 1996, 8, 305–313. N. Ashammakhi, P. Rokkanen, Biomaterials 1997, 18, 3–9. S.L. Ishaug-Riley, G.M. Crane, A. Gurlek, M.J. Miller, A.W. Yasko, M.J. Yaszemski, A.G. Mikos, J. Biomed. Mater. Res. 1997, 36, 1–8. K. Schwarz, M. Epple, Macromol. Rapid Commun. 1998, 19, 613–617. L. Lu, S.J. Peter, M.D. Lyman, H.L. Lai, S.M. Leite, J.A. Tamada, J.P. Vacanti, R. Langer, A.G. Mikos, Biomaterials 2000, 21, 1595–1605. C. Schiller, C. Rasche, M. Wehmo¨ller, F. Beckmann, H. Euﬁnger, M. Epple, S. Weihe, Biomaterials 2004, 25, 1239–1247. H. Euﬁnger, C. Rasche, J. Lehmbrock, M. Wehmo¨ller, S. Weihe, I. Schmitz, C. Schiller, M. Epple, Biomaterials 2007, 28, 475–485. C.K. Breuer, T. Shin’oka, R.E. Tanel, G. Zund, D.J. Mooney, P.X. Ma, T. Miura, S. Colan, R. Langer, J.E. Mayer, J.P. Vacanti, Biotechn. Bioeng. 1996, 50, 562–567. K.C. Dee, R. Bizios, Biotechn. Bioeng. 1996, 50, 438–442. L.E. Niklason, J. Gao, W.M. Abbott, K.K. Hirschi, S. Houser, R. Marini, R. Langer, Science 1999, 284, 489– 493. Q.Q. Qiu, P. Ducheyne, P.S. Ayyaswamy, J. Biomed. Mater. Res. 2000, 52, 66–76.

85 R. Zhang, P.X. Ma, J. Biomed. Mater.

Res. 2000, 52, 430–438. 86 C. Martin, H. Winet, J.Y. Bao,

Biomaterials 1996, 17, 2373–2380. 87 H. Winet, J.Y. Bao, J. Biomater. Sci.

Polymer Edn. 1997, 8, 517–532. 88 C.M. Agrawal, K.A. Athanasiou,

89

90 91

92 93

94

95

96

97

98

99

100

J. Biomed. Mater. Res. 1997, 38, 105– 114. W. Linhart, F. Peters, W. Lehmann, A.F. Schilling, K. Schwarz, M. Amling, J.M. Rueger, M. Epple, J. Biomed. Mater. Res. 2001, 54, 162– 171. C. Schiller, M. Epple, Biomaterials 2003, 24, 2037–2043. W. Linhart, W. Lehmann, M. Siedler, F. Peters, A.F. Schilling, K. Schwarz, M. Amling, J.M. Rueger, M. Epple, J. Mater. Sci. 2006, 41, 4806–4813. G. Ryan, A. Pandit, D.P. Apatsidis, Biomaterials 2006, 27, 2651–2670. M. Bram, A. Laptev, H.P. Buchkremer, D. Sto¨ver, Mat.-wiss. u. Werkstoﬀtech. 2004, 35, 213–218. M. Assad, P. Jarzem, M.A. Leroux, C. Coillard, A.V. Chernyshov, S. Charette, C.H. Rivard, J. Biomed. Mater. Res. Appl. Biomater. 2003, 64B, 107–120. M. Assad, A.V. Chernyshov, P. Jarzem, M.A. Leroux, C. Coillard, S. Charette, C.H. Rivard, J. Biomed. Mater. Res. Appl. Biomater. 2003, 64B, 121–129. M. Assad, F. Likibi, P. Jarzem, M.A. Leroux, C. Coillard, C.H. Rivard, Mat.-wiss. u. Werkstoﬀtech. 2004, 35, 219–223. O. Prymak, D. Bogdanski, M. Ko¨ller, S.A. Esenwein, G. Muhr, F. Beckmann, T. Donath, M. Assad, M. Epple, Biomaterials 2005, 26, 5801– 5807. H. Matsuno, A. Yokoyama, F. Watari, M. Uo, T. Kawasaki, Biomaterials 2001, 22, 1253–1262. F. Witte, J. Reifenrath, P.P. Mu¨ller, H.A. Crostack, J. Nellesen, F.W. Bach, D. Bormann, M. Rudert, Mat.-wiss. u. Werkstoﬀtech. 2006, 37, 504–508. T. Kokubo, Acta Mater. 1998, 46, 2519–2527.

References 101 P. Habibovic, F. Barre`re, C.A. van

102

103

104

105

106

107 108 109

110 111 112

113

114

115

116

Blitterswijk, K. de Groot, P. Layrolle, J. Am. Ceram. Soc. 2002, 85, 517– 522. Y. Han, K. Xu, G. Montay, T. Fu, J. Lu, J. Biomed. Mater. Res. 2002, 60, 511–516. L. Sun, C.C. Berndt, K.A. Khor, H.N. Cheang, K.A. Gross, J. Biomed. Mater. Res. 2002, 62, 228–236. J. Choi, D. Bogdanski, M. Ko¨ller, S.A. Esenwein, D. Mu¨ller, G. Muhr, M. Epple, Biomaterials 2003, 24, 3689– 3696. H. Schliephake, D. Scharnweber, M. Dard, S. Ro¨ßler, A. Sewing, C. Hu¨ttmann, J. Biomed. Mater. Res. 2003, 64A, 225–234. J.W.M. Vehof, J. van den Dolder, J.E. de Ruijter, P.H.M. Spauwen, J.A. Jansen, J. Biomed. Mater. Res. 2003, 64A, 417–426. S. Weiner, L. Addadi, H.D. Wagner, Mater. Sci. Eng. 2000, C11, 1–8. M. Neumann, M. Epple, Eur. J. Trauma 2006, 32, 125–131. H. Liao, H. Mutvei, L. Hammarstro¨m, T. Wurtz, J. Li, Biomaterials 2002, 23, 2693–2701. M. Ni, B.D. Ratner, Biomaterials 2003, 24, 4323–4331. U. Ripamonti, S. Ma, A.H. Reddi, Matrix 1992, 12, 202–212. J. Vuola, T. Bo¨hling, H. Go¨ransson, P. Puolakkainen, J. Biomed. Mater. Res. 2002, 59, 152–159. C.N. Demers, M. Tabrizian, A. Petit, R.C. Hamdy, H. L’Yahia, J. Biomed. Mater. Res. 2002, 59, 403–410. Y. Acil, I.N.G. Springer, V. Broek, H. Terheyden, S. Jepsen, J. Cell. Biochem. 2002, 86, 90–98. D. Turhani, E. Watzinger, M. Weißenbo¨ck, B. Cvikl, D. Thurnher, G. Wittwer, K. Yerit, R. Ewers, J. Oral Maxillofac. Surg. 2005, 63, 673–681. R. Ewers, J. Oral Maxillofac. Surg. 2005, 63, 1712–1723.

117 F. Wanschitz, M. Figl, A. Wagner, R.

118 119

120

121 122 123

124

125

126

127

128 129

130 131

Ewers, Int. J. Oral Maxillofac. Surg. 2006, 21, 433–438. K.A. Gross, E. Ezerietis, J. Biomed. Mater. Res. 2003, 64A, 672–683. A. Tampieri, G. Celotti, S. Sprio, A. Delcogliano, S. Franzese, Biomaterials 2001, 22, 1365–1370. W. Pompe, H. Worch, M. Epple, W. Friess, M. Gelinsky, P. Greil, U. Hempel, D. Scharnweber, K. Schulte, Mater. Sci. Eng. 2003, A362, 40–60. A.H. Reddi, Nature Biotechnol. 1998, 16, 247–252. E.H.J. Groeneveld, E.H. Burger, Eur. J. Endocrinol. 2000, 142, 9–21. P.H. Warnke, I.N. Springer, J. Wiltfang, Y. Acil, H. Euﬁnger, M. Wehmo¨ller, P.A.J. Russo, H. Bolte, E. Sherry, E. Behrens, H. Terheyden, Lancet 2004, 364, 766–770. M. Kantlehner, D. Finsinger, J. Meyer, P. Schaﬀner, A. Jonczyk, B. Diefenbach, B. Nies, H. Kessler, Angew. Chem. 1999, 111, 587–590. J.A. Rowley, Z. Sun, D. Goldman, D.J. Mooney, Adv. Mater. 2002, 14, 886–889. J. Gille, B. Dorn, J. Kekow, J. Bruns, P. Behrens, Int. Orthopaed. 2002, 26, 203–206. L. Meseguer-Olmo, M.J. Ros-Nicola´s, M. Clavel-Sainz, V. Vicente-Ortega, M. Alcaraz-Banos, A. Lax-Pe´rez, D. Arcos, C.V. Ragel, M. Vallet-Regi, J. Biomed. Mater. Res. 2002, 61, 458– 465. H.W. Kim, J.C. Knowles, H.E. Kim, Biomaterials 2004, 25, 1279–1287. D. Tadic, T. Welzel, P. Seidel, E. Wu¨st, E. Dingeldein, M. Epple, Mat.wiss. u. Werkstoﬀtech. 2004, 35, 1001– 1005. M. Vallet-Regi, Dalton Trans. 2006, 5211–5220. M. Wehmo¨ller, K. Neuking, M. Epple, T. Annen, H. Euﬁnger, Mat.-wiss. u. Werkstoﬀtech. 2006, 413–415.

95

97

7 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants Tadashi Kokubo and Hiroaki Takadama

Abstract

Most bone-bonding bioactive materials form bone-like apatite on their surfaces after being implanted into the living body, and bond to neighboring bone through this apatite layer. The apatite layer can be reproduced on the surfaces of materials in an organic-substance-free acellular simulated body ﬂuid (SBF) with ion concentrations almost equal to those of human blood plasma. The bone-bonding ability of a material is often evaluated by examining the ability of apatite to form on the material in SBF. In this chapter, the validity of this method for evaluating the bone-bonding bioactivity of a material, the ion concentrations of SBF, the materials able to form apatite, the characteristics of apatite, the bone-bonding mechanisms of bioactive materials, and the mechanisms of apatite formation, are reviewed. Key words: simulated body ﬂuid (SBF), bioactive material, apatite-forming ability, bone-bonding ability, bone substitute, bone-like apatite.

7.1 Introduction

Various materials, including Bioglass [1], sintered hydroxyapatite [2], sintered beta-tricalcium phosphate [3], biphasic ceramics of hydroxyapatite and tricalcium phosphate [4], and glass–ceramic A-W [A ¼ apatite (Ca10 (PO4 )6 (O, F2 )); W ¼ wollastonite (CaOSiO2 )] [5], can bond to living bone. These are referred to as ‘‘bioactive’’ materials, and many are currently in clinical use as important bone substitutes. Most of them bond to living bone through an apatite layer that forms on their surfaces after implantation into the living body. This apatite formation has been reproduced on their surfaces in an organic-substance-free acellular simulated body ﬂuid (SBF), with ion concentrations almost equal to those of human blood plasma [6]. This indicates that the bone-bonding bioactivity of a Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

98

7 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants

material can be evaluated in preliminary fashion, before conducting animal experiments, by examining apatite formation on its surface in SBF. As a result, the number of animals required to evaluate the bone-bonding ability of a material can be reduced, and today many laboratories utilize SBF as a standard tool when testing the bioactivity of new materials. In this chapter, the correlation between the bone-bonding bioactivity of materials and their apatite-forming ability in SBF, the ion concentrations of SBF, the dependence of apatite formation on the material, the characteristics of apatite, the bone-bonding mechanisms of bioactive materials, and the mechanisms of apatite formation on these materials, are described.

7.2 Qualitative Correlation of Bone-Bonding Bioactivity of a Material with Apatite Formation on its Surface in SBF

Materials that bond to living bone through an apatite or calcium phosphate layer that forms on their surfaces after implantation into the living body include: a Bioglass 45S5 type-glass in a Na2 OaCaOaSiO2 aP2 O5 system [7]; bioactive glasses in the Na2 OaCaOaB2 O3 aAl2 O3 aP2 O5 system [8]; glasses in the CaOaSiO2 system [9]; Ceravital-type glass–ceramics containing crystalline apatite in the Na2 OaCaOaSiO2 aP2 O5 system [10]; Glass–ceramic A-W, containing crystalline apatite and wollastonite, in the MgOaCaOaSiO2 aP2 O5 system [11]; Bioverite-type glass–ceramics containing crystalline apatite and ﬂuorophlogopite in the Na2 OaMgOaCaOaAl2 O3 aSiO2 aP2 O5 aF system [12]; sintered hydroxyapatite [13]; biphasic ceramics of hydroxyapatite and beta-tricalcium phosphate; sintered calcium sulfate [14]; a composite of glass–ceramic A-W with polyethylene [15]; titanium metal subjected to NaOH and heat treatments [16]; and tantalum metal subjected to NaOH and heat treatments [17]. An example of an interface of glass–ceramic A-W to living bone is shown in Figure 7.1. All of these bone-bonding bioactive glasses, glass–ceramics, sintered crystalline ceramics, composites and metals have been conﬁrmed as forming an apatite on their surfaces in SBF within 4 weeks [4, 6–8, 10, 14, 18–23], except for the Bioverite-type glass–ceramic, which has not been investigated for apatite forma-

7.2 Qualitative Correlation of Bone-Bonding Bioactivity of a Material with Apatite Formation

Fig. 7.1 Transmission electron microscopy image of the interface of glass–ceramic A-W and rabbit tibial bone [13].

tion on its surface in SBF. An apatite layer formed on a glass–ceramic A-W in SBF is illustrated in Figure 7.2. When a small amount of Al2 O3 was added to the composition of Bioglass-type glass [24], CaOaSiO2 glass [25], and glass–ceramic A-W [26], the resultant glasses and glass–ceramics did not form an apatite or calcium phosphate layer on their surfaces in the living body, and did not bond to the neighboring bone. In addition, none of these materials with added Al2 O3 formed apatite on their surfaces within 4 weeks in SBF [14, 27, 28]. It can be concluded from these results that the essential requirement for a material to bond to living bone is the formation of an apatite or calcium phosphate layer on its surface, and that the bone-bonding bioactivity of a material can be evaluated by examining the formation of an apatite layer on its surface in SBF. However, it should be noted here that a small number of cases in which a material bonds to living bone without yielding a detectable apatite layer at their interfaces have been reported. Sintered beta-tricalcium phosphate and a natural calcite of calcium carbonate are examples [29, 30], with neither material forming an apatite layer on its surface within 4 weeks in SBF [31, 32]. In fact, the bone-bonding properties of these materials might be related to their high resorbability in the living body. One case in which a material – abalone shell – does not bond to living bone, despite forming an apatite or calcium phosphate layer on its surface in the living

99

100

7 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants

Fig. 7.2 The apatite layer formed on glass–ceramic A-W in simulated body ﬂuid [19].

body – has been also reported. Abalone shell also forms an apatite layer on its surface in SBF [32], and suppression of the bone-bonding bioactivity of this material might be attributed to foreign body reactions elicited by small amounts of protein contained in the shell. From the above ﬁndings, it can be concluded that a material which is able to form an apatite layer on its surface in SBF may bond to living bone through the apatite layer that forms on its surface, as long as the material does not release any component that induces toxic or immune responses in the surrounding tissue. Based on these ﬁndings, the examination of apatite formation on a surface of a material in SBF would be a useful tool for predicting the bone-bonding bioactivity of a material, before progressing to animal experiments. Indeed, by using this method not only the number of animals but also the duration of animal experiments required to evaluate the bone-bonding bioactivity of a material can be greatly reduced.

7.3 Quantitative Correlation of Bone-Bonding Bioactivity and Apatite-Forming Ability in SBF

Not all bioactive materials show equal bone-bonding ability; rather, the time required for a material to bond to living bone, and the amount of bone formed

7.4 Ion Concentrations of SBF

Fig. 7.3 The rate of bone formation on a cross-section of a defect of rabbit femur when ﬁlled with glass particles 6 weeks after implantation compared with time of surface apatite formation in simulated body ﬂuid [33].

around a material in a given time, will vary widely depending on the material involved. The time required for a bioactive material to become fully covered with apatite in SBF also varies, depending on the material. In order to investigate the relationship between bone formation in vivo and apatite formation in SBF, bone formation in defects in rabbit femurs ﬁlled with Na2 OaCaOaSiO2 glass particles (the SiO2 contents of which were changed from 70.0 to 50.0 mol%, with a constant Na2 O/CaO molar ratio of one) were examined [33]. The time required for the same glasses in SBF to form bone-like apatite which fully covered their surfaces was also measured [34]. The data provided in Figure 7.3 show clearly that bone formation around glass particles increases with the increasing apatiteforming ability of the glasses in SBF. This, in turn, indicates that the bonebonding bioactivity of a material can be evaluated not only qualitatively but also quantitatively, by examining the apatite-forming ability on the material’s surface in SBF.

7.4 Ion Concentrations of SBF

In all of the above-described investigations, the organic-substance-free acellular solution used as the SBF had ion concentrations as ﬁrst reported by Kokubo et al. in 1990 [6], and as corrected by the same authors in 1991 [35]. However, the ion concentrations of this SBF were not exactly equal to those of human blood plasma (see Table 7.1), as SBF is richer in Cl ions and poorer in HCO3 ions than is human blood plasma [36]. In 2003, Oyane et al. proposed a revised simulated body ﬂuid (r-SBF), in which the ion concentrations were identical to those of human blood plasma [37]. However, r-SBF had a strong tendency to produce

101

102

7 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants Table 7.1 Ion concentrations of simulated body ﬂuid (SBF) and human blood plasma.

Ion concentration [mmol]

Blood plasma SBF

NaB

KB

Mg 2B

Ca 2B

ClC

HCO3C

HPO4 2C

SO4 2C

142.0 142.0

5.0 5.0

1.5 1.5

2.5 2.5

103.0 147.8

27.0 4.2

1.0 1.0

0.5 0.5

precipitates of calcium carbonate, as it is highly supersaturated with respect to hydroxyapatite and calcite [38]. In 2004, the method for preparing conventional SBF was further reﬁned and simpliﬁed such that it could be easily prepared and subjected to round-robin testing by 10 research institutes [39]. This reﬁned SBF recipe (the details of which have been published [40]) was proposed to the International Organization for Standardization as a standard solution for in-vitro monitoring of the apatiteforming ability of implant materials. Simulated body ﬂuids with higher ion concentrations (e.g., 1.5 and 4 SBF, where ion concentrations are 1.5- or four-times those of SBF) have also been used to evaluate the bone-bonding abilities of materials, or the production of a bone-like apatite layer on materials. It should be noted, however, that no correlation has been identiﬁed between apatite formation in such solutions and bone-bonding ability, and that the apatite formed in these solutions diﬀers in composition from bone mineral [41].

7.5 Materials Able to Form Apatite

Despite human body ﬂuid being highly supersaturated with respect to apatite (even under normal conditions [42]), apatite does not usually precipitate in the living body, except at sites of bone tissue, as the energy barrier for its nucleation is high. This means that, once apatite nuclei have been formed catalytically on a material, they can grow spontaneously by consuming the calcium and phosphate ions from the surrounding body ﬂuid. The question persists, however, as to what type of material induces apatite nucleation. In an attempt to answer this question, various types of gels prepared using sol-gel methods were soaked in SBF, and their apatite-forming abilities examined. Although SiO2 [43], TiO2 [44], ZrO2 [45], Nb2 O5 [46] and Ta2 O5 [47] gels were seen to form apatite on their surfaces, Al2 O3 [44] gels did not (Fig. 7.4), which indicated that the SiaOH, TiaOH, ZraOH, NbaOH and TaaOH groups that were abundant on the surfaces of the gels were eﬀective in inducing apatite

7.6 Composition and Structure of Apatite

Fig. 7.4 Apatite formed on (left) SiO2 and (right) TiO2 gels in simulated body ﬂuid [44].

nucleation. Subsequently, Tanahashi et al., using self-assembled monolayers, showed that COOH and PO4 H2 groups were also eﬀective for apatite nucleation [48]. Based on these ﬁndings, titanium metal and its alloys were subjected to NaOH solution and heat treatment to form sodium titanate on their surfaces. These treated materials were found to form bone-like apatite on their surfaces in the living body, and to bond to living bone [49], and subsequently were applied for use in hip-joint prostheses.

7.6 Composition and Structure of Apatite

The calcium phosphate layer formed on bioactive materials after implantation into living bodies has been identiﬁed by micro X-ray diﬀraction [50] and electron [51] diﬀraction as a nanosized crystalline apatite. However, it has been diﬃcult to obtain more detailed structural information for these calcium phosphate layers formed in vivo. More detailed structural information can be obtained for apatite formed on bioactive materials in SBF. According to observations made with transmission electron microscopy (TEM), the apatite on both glass–ceramic A-W [52] and NaOH- and heat-treated titanium metal [53] in SBF takes the shape of thin needles of 10 nm thickness and 100 nm length (Fig. 7.5). This apatite has a Ca/P atomic ratio of about 1.65, which is less than the stoichiometric value of 1.67, and contains a small amount of Naþ and Mg 2þ ions beside CO3 2 ions [19, 52, 53]. As these characteristics are similar to those of bone mineral, the material may be referred to as ‘‘bone-like’’ apatite.

103

104

7 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants

Fig. 7.5 Transmission electron microscopy (left) and energy dispersive X-ray analysis (right) images of apatite formed on NaOH- and heattreated titanium metal in simulated body ﬂuid [53].

7.7 Mechanism of Bonding of Bioactive Material to Bone

As described above, most bioactive materials form apatite on their surfaces after being implanted into the living body. As this surface apatite is very similar to bone mineral in its composition, structure and morphology, the bone-producing cells (osteoblasts) could preferentially proliferate and diﬀerentiate on its surface to produce collagen and apatite, similar to their behavior on the surface of fractured bone (Fig. 7.6) [54]. As a result, the surrounded bone may come into direct

Fig. 7.6 Transmission electron microscopy image of the interface of glass–ceramic A-W and rabbit tibial bone at early stage after implantation [54].

7.8 Mechanisms of Apatite Formation

contact with the surface apatite layer on materials. When this occurs, a tight chemical bond is formed between the apatite in the bone and the surface apatite to reduce their interface energy.

7.8 Mechanisms of Apatite Formation

If apatite formation on bioactive materials implanted into the living body can be reproduced on their surfaces in SBF, then the mechanisms of apatite formation on the materials might be revealed by the examining surface structural changes of the materials as a function of soaking time in SBF. Based on TEM observations and zeta potential measurements, the mechanism of apatite formation on sintered hydroxyapatite in body environment is interpreted as follows [55]. The sintered hydroxyapatite is initially negatively charged on its surface, and combines with positively charged Ca 2þ ions in the surrounding ﬂuid. As a result, Ca-rich amorphous calcium phosphate is formed on the sintered hydroxyapatite. As the Ca 2þ ions accumulate, the sintered hydroxyapatite becomes positively charged on its surface and reacts with negatively charged phosphate ions. As a result, Ca-poor amorphous calcium phosphate is formed which is eventually transformed into the more stable, nanosized crystalline bone-like apatite. This mechanism is essentially the same in ﬂuids containing proteins [56]. Apatite formation on NaOH- and heat-treated titanium metal in a body environment is similarly interpreted as follows [53, 57]. The treated titanium metal releases Naþ ions from its surface sodium titanate layer via exchange with H3 Oþ ions in the ﬂuid, to form TiaOH groups (Fig. 7.7). As a result, its surface becomes negatively charged and reacts with positively charged Ca 2þ ions to form calcium titanate. As the calcium ions accumulate, the positively charged surface reacts with negatively charged phosphate ions, forming amorphous calcium phosphate. As this phase is metastable, it eventually transforms into nanosized, crystalline bone-like apatite.

Fig. 7.7 The mechanism of bone-like apatite formation on NaOH- and heat-treated titanium metal in vitro [57]. SBF ¼ simulated body ﬂuid.

105

106

7 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants

7.9 Summary

Simulated body ﬂuid, which is used to test the bone-bonding bioactivity of various materials, is identical to human blood plasma in terms of its ion concentrations, but does not contain organic substances such as proteins. Nevertheless, the soaking of bioactive materials in SBF can reproduce the apatite formation seen on such materials in the living body. SBF is easily prepared and relatively stable at body temperature. Moreover, it is useful for evaluating the bone-bonding bioactivity of new materials and investigating the mechanisms of apatite formation on, and bone bonding of, materials.

References 1 L.L. Hench, R.J. Splinter, W.C. Allen,

2

3 4

5

6

7

8 9

10

11

T.K. Greenlee, J. Biomed. Mater. Res. 1972, 2, 117–141. M. Jarcho, J.I. Kay, R.H. Gummaer, H.P. Drobeck, J. Bioeng. 1977, 1, 79–92. B.J. Rejda, J.G.J. Peelen, K. de Groot, J. Bioeng. 1977, 1, 93–97. R.Z. Legeros, S. Lin, R. Rohanizadeh, D. Mijares, J.P. Legeros, J. Mater. Sci. Mater. Med. 2003, 14, 201–209. T. Kokubo, M. Shigematsu, Y. Nagashima, M. Tashiro, T. Nakamura, T. Yamamuro, S. Higashi, Bull. Inst. Chem. Res. Kyoto Univ. 1982, 60, 260–268. T. Kokubo, H. Kushitani, S. Sakka, T. Kitsugi, T. Yamamuro, J. Biomed. Mater. Res. 1990, 24, 721–734. M. Ogino, F. Ohuchi, L.L. Hench, J. Biomed. Mater. Res. 1980, 14, 55–64. ¨ .H. Anderson, K.H. Karlsson, J. NonO Cryst. Solids 1991, 129, 145–151. K. Ohura, T. Nakamura, T. Yamamuro, T. Kokubo, Y. Ebisawa, Y. Kotoura, M. Oka, J. Biomed. Mater. Res. 1991, 25, 357–365. C. Ohtsuki, H. Kushitani, T. Kokubo, S. Kotani, T. Yamamuro, J. Biomed. Mater. Res. 1991, 25, 1363–1370. T. Kitsugi, T. Nakamura, T. Yamamuro, T. Kokubo, T. Shibuya, M. Takagi, J. Biomed. Mater. Res. 1987, 21, 1255–1271.

¨ land, W. Vogel, K. Naumann, 12 W. Ho

13

14

15

16

17

18 19

J. Biomed. Mater. Res. 1985, 19, 303–312. M. Neo, S. Kotani, T. Nakamura, T. Yamamuro, C. Ohtsuki, T. Kokubo, Y. Bando, J. Biomed. Mater. Res. 1992, 26, 1419–1432. H. Chan, D. Mijares, J.L. Ricci, Transactions of the Seventh World Biomaterials Congress. Australian Society for Biomaterials Inc., Brunswick Lower, Victoria, Australia, 2004, p. 627. J.A. Juhasz, S. Ishii, S.M. Best, M. Kawashita, M. Neo, T. Kokubo, T. Nakamura, W. Bonﬁeld, Transactions of the Seventh World Biomaterials Congress, Australian Society for Biomaterials Inc., Brunswick Lower, Victoria, Australia, 2004, p. 665. S. Nishiguchi, S. Fujibayashi, H.-M. Kim, T. Kokubo, T. Nakamura, J. Biomed. Mater. Res. 2003, 67A, 28–35. H. Kato, T. Nakamura, S. Nishiguchi, Y. Matsusue, M. Kobayashi, T. Miyazaki, H.-M. Kim, T. Kokubo, J. Biomed. Mater. Res. Appl. Biomater. 2000, 53, 28–35. C. Ohtsuki, T. Kokubo, T. Yamamuro, J. Non-Cryst. Solids 1992, 143, 84–92. T. Kokubo, S. Ito, T. Huang, T. Hayashi, S. Sakka, T. Kitsugi, T. Yamamuro, J. Biomed. Mater. Res. 1990, 24, 331–343.

References 20 T. Kokubo, M. Kushiyani, Y. Ebisawa,

21

22

23

24

25

26

27

28

29

30

31

32

33

34

T. Kitsugi, S. Kotani, K. Oura, T. Yamamuro, Bioceramics, Volume 1. Ishiyaku EuroAmerica, Tokyo, 1988, pp. 157–162. J.A. Judasz, S.M. Best, W. Bonﬁeld, M. Kawashita, N. Miyata, T. Kokubo, T. Nakamura, J. Mater. Sci. Mater. Med. 2003, 14, 489–495. H.-M. Kim, F. Miyaji, T. Kokubo, T. Nakamura, J. Biomed. Mater. Res. 1996, 32, 409–417. T. Miyazaki, H.-M. Kim, F. Miyaji, T. Kokubo, T. Nakamura, J. Biomed. Mater. Res. 2000, 50, 35–42. ¨ .H. Andersson, G. Liu, K.H. O Karlsson, L. Niemi, J. Miettinen, J. Juhanoja, J. Mater. Sci. Mater. Med. 1990, 1, 219–227. K. Ohura, T. Nakamura, T. Yamamuro, Y. Ebisawa, T. Kokubo, Y. Kotoura, M. Oka, J. Mater. Sci. Mater. Med. 1992, 3, 95–100. T. Kitsugi, T. Yamamuro, T. Nakamura, T. Kokubo, Int. Orthop. 1989, 13, 199–206. ¨ .H. Andersson, G. Liu, K. O Kangasniemi, J. Juhanoja, J. Mater. Sci. Mater. Med. 1992, 3, 145–150. Y. Ebisawa, T. Kokubo, K. Ohura, T. Yamamuro, J. Mater. Sci. Mater. Med. 1990, 1, 239–244. S. Kotani, Y. Fujita, T. Kitsugi, T. Nakamura, T. Yamamuro, J. Biomed. Mater. Res. 1991, 25, 1303–1315. Y. Fujita, T. Yamamuro, T. Nakamura, S. Kotani, J. Biomed. Mater. Res. 1991, 25, 991–1003. C. Ohtsuki, T. Kokubo, M. Neo, S. Kotani, T. Yamamuro, T. Nakamura, Y. Bando, Phos. Res. Bull. 1991, 1, 191–196. C. Ohtsuki, Y. Aoki, T. Kokubo, Y. Fujita, S. Kotani, T. Yamamuro, Transactions of the 11th Annual Meeting of Japanese Society for Biomaterials. Japanese Society for Biomaterials, Toshima-ku, Tokyo, Japan, 1989, p. 12. S. Fujibayashi, M. Neo, H.-M. Kim, T. Kokubo, T. Nakamura, Biomaterials 2003, 24, 1349–1356. H.-M. Kim, F. Miyaji, T. Kokubo, C. Ohtsuki, T. Nakamura, J. Am. Ceram. Soc. 1995, 78, 2405–2411.

35 T. Kokubo, Biomaterials 1991, 12,

155–163. 36 J.E. Gamble, Chemical Anatomy,

37

38

39

40 41

42

43

44

45

46

47

48 49

50

51

Physiology and Pathology of Extracellular Fluid. Harvard University Press, Cambridge, 1967, pp. 1–17. A. Oyane, H.-M. Kim, T. Furuya, T. Kokubo, T. Miyazaki, T. Nakamura, J. Biomed. Mater. Res. 2003, 65A, 188–195. A. Oyane, K. Onuma, A. Ito, H.-M. Kim, T. Kokubo, T. Nakamura, J. Biomed. Mater. Res. 2003, 64A, 339–348. H. Takadama, M. Hashimoto, M. Mizuno, T. Kokubo, Phos. Res. Bull. 2004, 17, 119–125. T. Kokubo, H. Takadama, Biomaterials 2006, 27, 2907–2915. H.-M. Kim, K. Kishimoto, F. Miyaji, T. Kokubo, T. Yao, Y. Suetsugu, J. Tanaka, T. Nakamura, J. Biomed. Mater. Res. 1999, 46, 228–235. W. Neuman, M. Neuman, The Chemical Dynamics of Bone Mineral. University of Chicago, 1958, p. 34. P. Li, C. Ohtsuki, T. Kokubo, K. Nakanishi, N. Soga, T. Nakamura, J. Am. Ceram. Soc. 1992, 75, 2094– 2097. P. Li, C. Ohtsuki, T. Kokubo, K. Nakanishi, N. Soga, K. de Groot, J. Biomed. Mater. Res. 1994, 28, 7–15. M. Uchida, H.-M. Kim, T. Kokubo, T. Nakamura, J. Am. Ceram. Soc. 2001, 84, 2041–2044. T. Miyazaki, H.-M. Kim, T. Kokubo, C. Ohtsuki, H. Kato, T. Nakamura, J. Ceram. Soc. Japan 2001, 109, 929–933. T. Miyazaki, H.-M. Kim, T. Kokubo, H. Kato, T. Nakamura, J. Sol-Gel Sci. Technol. 2001, 21, 83–88. M. Tanahashi, T. Matsuda, J. Biomed. Mater. Res. 1997, 34, 305–315. T. Kokubo, H.-M. Kim, M. Kawashita, T. Nakamura, J. Mater. Sci. Mater. Med. 2004, 15, 99–107. T. Kokubo, C. Ohtsuki, S. Kotani, T. Kitsugi, T. Yamamuro, Bioceramics, Volume 2. German Ceramic Society, Cologne, 1990, pp. 113–121. M. Neo, S. Kotani, Y. Fujita, T. Nakamura, T. Yamamuro, Y. Bando,

107

108

7 Simulated Body Fluid (SBF) as a Standard Tool to Test the Bioactivity of Implants C. Ohtsuki, T. Kokubo, J. Biomed. Mater. Res. 1992, 26, 255–267. 52 C. Ohtsuki, Y. Aoki, T. Kokubo, Y. Bando, M. Neo, T. Nakamura, J. Ceram. Soc. Japan 1995, 103, 449–454. 53 H. Takadama, H.-M. Kim, T. Kokubo, T. Nakamura, J. Biomed. Mater. Res. 2001, 57, 441–448. 54 M. Neo, T. Nakamura, C. Ohtsuki, T. Kokubo, T. Yamamuro, J. Biomed. Mater. Res. 1993, 27, 999–1006.

55 H.-M. Kim, T. Himeno, M.

Kawashita, T. Kokubo, T. Nakamura, J. R. Soc. Interface 2004, 1, 17–22. 56 T. Kokubo, T. Himeno, H.-M. Kim, M. Kawashita, T. Nakamura, Bioceramics, Volume 16, Trans. Tech. Pub., Switzerland, 2004, pp. 139–142. 57 H.-M. Kim, T. Himeno, M. Kawashita, J.H. Lee, T. Kokubo, T. Nakamura, J. Biomed. Mater. Res. 2003, 67A, 1305–1309.

109

8 Stimulation of Bone Growth on Implants by Integrin Ligands Mo´nica Lo´pez-Garcı´a and Horst Kessler

Abstract

A successful biointegration of orthopedic and craniofacial implants requires a strong mechanical interaction between the surface of the artiﬁcial material and the surrounding natural bone tissue. Osseointegration of implants is known to be a biological process that occurs by formation of new peri-implant bone in direct contact with the synthetic surface. Mimicking the physiological adhesion process of osteoblasts to the extracellular matrix (ECM), by coating of implant surfaces with speciﬁc cell-adhesive molecules, was proven to enhance osteoblast adhesion in vitro and to accelerate osseointegration of implants in vivo. Cell adhesion is mediated by integrins, a class of heterodimeric transmembrane cell receptors that bind selectively to diﬀerent proteins of the ECM and transduce information to the nucleus through cytoplasmic signaling pathways. The peptide sequence Arg-Gly-Asp (RGD), is by far the most eﬀective and extensively studied ligand to promote osteoblast adhesion and proliferation on implants through integrin stimulation. The biofunctionalization of diﬀerent surfaces with RGD peptides and mimetics has resulted in major improvements in bone implant technology. Key words: integrins, RGD peptides, cell adhesion, surface coating, implants, osseointegration.

8.1 Introduction 8.1.1 Biomimetic Materials for Implant Technology

Biomaterials are designed to restore or replace a damaged part of the body and/or its associated functions. They may be used either in a permanent way or as temHandbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

110

8 Stimulation of Bone Growth on Implants by Integrin Ligands

porary support for cells and tissues, but in both cases they should exhibit a pronounced compatibility with the biological environment [1]. Most of the commonly employed materials (polymers, ceramics, metals, etc.) are non-toxic, have suﬃcient mechanical stability and elasticity, and are stable towards enzymatic degradation. However their non-physiological character often leads to undesirable processes, such as graft rejection, inﬂammations, infections, local tissue wasting and implant encapsulation, as well as thrombosis and embolization [2]. These biological responses are associated with a poor integration after implantation into the living body; that is, an inadequate interaction between the artiﬁcial material and the biological tissue. Controlled cell–biomaterial interaction is of utmost importance to avoid graft rejection and to favor a successful implantation. In particular, a strong mechanical contact between the implant surface and the surrounding tissue is required for the osseointegration of bone implants. Biofunctionalization of the implant material for rapid and speciﬁc cell colonization of their surfaces is of growing interest in implant technology and tissue engineering [3]. Biomimetic surface modiﬁcation takes advantage of the power of speciﬁc biomolecular recognition events to control implant–tissue interactions without compromising the desirable bulk characteristics of an implant material. During the past few decades, those working in the ﬁelds of material science, surface engineering, chemistry, physics, biology, biochemistry and medicine have attempted to functionalize the surfaces of implant materials with bioactive molecules in order to enable signaling to adjacent cells and to obtain a desired cellular response. Cell–biomaterial interactions can be either speciﬁc or unspeciﬁc. Unspeciﬁc interactions are diﬃcult to control because they are based on properties common to multiple cell types. The design of biomimetic materials takes proﬁt of the speciﬁc interactions, related to deﬁned chemical structures, such as ligands that interact with their corresponding cell surface receptor [4]. The most often-employed procedure to enhance cell adhesion and proliferation on synthetic surfaces targets the integrin receptors. 8.1.2 Integrins and RGD Sequence

The integrin family represents the most numerous and versatile group of cell adhesion receptors, which regulate the cell–cell and cell–extracellular matrix (ECM) interactions in multicellular organisms [5]. These interactions inﬂuence many fundamental cellular functions such as motility, proliferation, diﬀerentiation, and apoptosis. Therefore, integrins not only play a major role as anchoring molecules but they are also involved in many biological processes, such as embryogenesis, blood coagulation, immune response, hemostasis and regulation of the balance between cellular proliferation and apoptosis [6]. Integrins are heterodimeric transmembrane proteins composed of two noncovalently associated subunits (a and b). The 18 a and eight b known subunits combine to form 24 diﬀerent heterodimers (Fig. 8.1) which diﬀer in their ligand speciﬁcity [7].

8.1 Introduction

Fig. 8.1 The integrin family: the 24 known heterodimers.

The tripeptide sequence Arg-Gly-Asp (RGD) was identiﬁed as a minimal essential cell adhesion peptide sequence in ﬁbronectin [8], since when cell-adhesive RGD motifs have been identiﬁed in many other ECM proteins, including vitronectin, ﬁbrinogen, collagen, laminin, and osteopontin [9]. About half of the 24 integrins have been shown to bind to ECM molecules in a RGD-dependent manner [9]. The RGD sequence is the most eﬀective and often-employed sequence to stimulate cell adhesion on synthetic surfaces. This is based on its widespread distribution and use in the organism, its ability to address more than one cell adhesion receptor, and its biological impact on cell anchoring, behavior, and survival. 8.1.3 Natural Proteins or Synthetic Peptides as Cell-Adhesive Molecules?

In the early studies, the surface of implant materials was coated with celladhesive ECM natural proteins which contain the RGD sequence in its structure [10]. However, the use of these proteins bears some disadvantages (Table 8.1) which prevent their practical use for medical applications. Most of these problems can be overcome when these macromolecular ligands are reduced to small recognition sequences, as small synthetic RGD peptides [11]. The RGD peptide sequence, structure and conformation play a crucial function in the ligand–receptor interaction and/or in the stability of the interaction. Sometimes, linear peptides experience slow enzymatic degradation [11a, 12], but small cyclic peptides are known to exhibit excellent long-term stability [13] as well as higher selectivity. Cyclic derivatives can interact with integrins more eﬀectively than linear RGD peptides because the cyclization induces conformational stability as well as enhancing the preferred three-dimensional (3-D) structure for receptor interactions. However, cyclization usually results in low (or a lack of ) activity, and

111

112

8 Stimulation of Bone Growth on Implants by Integrin Ligands Table 8.1 Evolution, advantages and disadvantages of the diﬀerent coating methods.

Integrin ligand

Advantages

Disadvantages

Proteins (1st generation)

Speciﬁcity, analogy to natural adhesion molecules

Enzymatic instability, immunogenicity, infection and inﬂammation risks, diﬃculties in anchoring, high costs

Peptides (2nd generation)

No risk of contamination, not immunogenic, higher temperature and pH stability, can be packed with higher density on surfaces, cost eﬀectiveness

Enzymatic instability, lack of selectivity (e.g., osteoblast vs. platelets)

Cyclic peptides (3rd generation)

High selectivity, enzymatic stability

Moderate costs

Peptidomimetics (4th generation)

Very low costs, may be highly speciﬁc, high stability in the sterilization process as well as in vivo

–

only when the bioactive conformation is matched can super-activity and receptorsubtype selectivity be found. In the case of RGD peptides, a ‘‘spatial screening’’ procedure was applied to optimize the structure–activity relationship [14]. Modiﬁcation of the amino acid sequences ﬂanking the RGD motif or changing its 3-D structure has been shown to modify the ligand selectivity [15]. For example, the D-phenylalanine residue (f ¼ D-Phe) following the RGD binding sequence in peptides of the type cyclo(-RGDfK-) is essential to enhance the av -selectivity versus the platelet receptor aIIb b 3 (to induce the preferred adhesion of osteoblasts rather than of platelets). Cyclization also causes proteolytic stability. Both factors are of great importance for achieving a correct implant osseointegration in vivo. More recently, research into the design of speciﬁc integrin antagonists has focused on the development of non-peptidic integrin-selective mimetics, due to their specially desirable properties [16]. 8.1.4 Integrin-Mediated Cell Adhesion

The process of integrin-mediated cell adhesion involves four diﬀerent, partly overlapped events [17]: cell attachment; cell spreading; organization of actin cytoskeleton; and the formation of focal adhesions. In the ﬁrst step – the ‘‘initial’’ attachment – the cell contacts the surface and some ligand binding occurs; this allows the cell to resist gentle shear forces. Afterwards, the cell begins to ﬂatten

8.1 Introduction

and its plasma membrane spreads over the substratum. The third eﬀect noticed is the actin organization into microﬁlament bundles forming an actin cytoskeleton, sometimes referred to as ‘‘stress ﬁbers’’. In the fourth step, the formation of focal adhesions occurs, which link molecules of the ECM to components of the cell’s actin cytoskeleton. During these four steps, integrins mediate physical anchoring and transmembrane signaling processes [5, 7]. The structure of the integrins allows them to function as bidirectional cellular signal transducers [18]. Conformational changes induced by the binding of ligands to integrins invoke signaling cascades inside the cell that regulate gene expression, activate kinases, and direct cytoskeletal organization (outside-in signaling) [5]. Alternatively, internal cellular activation can produce both conformational changes and multimeric clustering of the integrins, which results in non-consecutive binding to ligands, ECM components, as well as other cells (inside-out signaling) [19]. The type and degree of the signaling event is determined by the conformation and nature of the ligand, and is regulated by divalent cations bound to metal ion-dependent adhesion sites (MIDAS) on the integrin receptors. A great number of signaling events following the formation of focal adhesions are known. These include activation of focal adhesion kinase (FAK), extracellular signal-regulated kinase (ERK), small Rho GTPases and phosphatidyl inositol 4phosphate 5-kinase (PIP 5-kinase), and some elements of the mitogen-activated protein kinase (MAPK) pathway [5, 20], although many details are not yet clear. Nevertheless, it is well established that integrin-mediated cell spreading and focal adhesion formation induces survival and the proliferation of anchoragedependent cells [21]. The biofunctionalization of implant surfaces in order to mimic the biological environment and to stimulate a speciﬁc cell colonization is based on this eﬀect (Fig. 8.2A). In contrast, loss of attachment causes apoptosis in many cell types, a situation referred to as ‘‘anoikis’’ [22]. Anoikis can even be induced in the presence of immobilized ECM molecules when non-immobilized soluble ligands such as RGD peptides are added (Fig. 8.2B). Based on this princi-

Fig. 8.2 The opposing eﬀects of integrin ligands. (A) Immobilized ligands: these act as agonists of the extracellular matrix (ECM), leading to cell adhesion and survival. (B) Non-immobilized ligands: these act as antagonists of the ECM, leading to cell detachment and apoptosis.

113

114

8 Stimulation of Bone Growth on Implants by Integrin Ligands

Fig. 8.3 Scanning electron micrographs of adherent cells on substrates containing covalently grafted peptide (GRGDY). (A) Spheroid cells with no ﬁlapodial extensions. (B) Spheroid cells with one to two ﬁlapodial extensions. (C) Spheroid cells with more than two ﬁlapodial extensions. (D) Flattened morphology representative of well-spread cells. (Figure in according to Ref. [26])

ple, inhibition assays as well as some therapeutic applications of RGD peptides and peptidomimetics in the ﬁeld of osteoporosis, renal failure, cancer and angiogenesis have been developed [23]. Recently, the role of the anoikis model was questioned concerning RGD peptide-induced apoptosis and the inhibition of angiogenesis [22b, 24]. Cell proliferation and apoptosis are two contrasting biological processes, both of which are integrin-dependent. Round cells are related to apoptosis, whereas cell spreading is related to its survival, focal adhesion formation, and proliferation [25]. Four types of adherent cell morphologies on surfaces have been described [26] (Fig. 8.3). With increasing concentrations of the integrin receptors on the surface (0.1, 1, 10 and 100 fmol cm2 ), the predominant morphology type shifts from A to D. Therefore, cell adhesion on biomaterials promoted by integrin ligands clearly depends on its surface density. Cell attachment as a function of RGD concentration shows a sigmoidal increase [27], indicating that there is a critical minimum density for cell response, which depends on the surface material and the type of cells. For comparable cell adhesion studies, it is important to know the density of the RGD peptides on the surface. The surface-bound RGD peptide may be determined using integrin-speciﬁc antibodies, although on the other hand an eﬃcient method by radiolabeling with 125 I of a tyrosine-containing RGD peptide was recently reported [28]. The amount of bound peptide was seen to depend on the material and the concentration of the peptide in the coating solution used, although this was in the pmol cm2 range for all surfaces examined [poly(methyl methacrylate) (PMMA), titanium and silicon]. Nanoscale patterned surfaces were recently designed to address the precise molecular topology of focal adhesions by using block-copolymer micelle nanolithography [29]. These systems consist of a hexagonally packed rigid template of celladhesive gold nanodots coated with cyclo(-RGDfK-)-thiol peptide and separated by non-adhesive regions. The gold dots are small enough (<8 nm) to allow the bind-

8.2 Improvements in Implant-Osseointegration by Surface Modiﬁcation with Integrin Ligands

ing of one single integrin per dot, and are positioned with high precision at 28-, 58-, 73-, and 85-nm spacing. Spacing between the dots must be smaller than 65 nm in order to achieve stable adhesion of diﬀerent cells. The distribution of the surface is also crucial. When a lower total number of dots is used they still induce cell adhesion if they are clustered. Hence, it is clear that stable adhesion requires a spatial clustering of integrins to form focal adhesion from which the stress ﬁbers within the cells can be formed. These properties are shared by a variety of cultured cells, including MC3T3-osteoblasts, B16-melanocytes, REF52-ﬁbroblasts, and 3T3-ﬁbroblasts.

8.2 Improvements in Implant-Osseointegration by Surface Modiﬁcation with Integrin Ligands 8.2.1 Mechanisms of Bone Grafting

Bone formation is the result of a series of sequential events that begin with recruitment and proliferation of osteoprogenitor cells from surrounding tissues, followed by osteoblastic diﬀerentiation, matrix formation, and mineralization. In bone grafting there are three basic biological mechanisms to favor bone regeneration: osteogenesis; osteoconduction; and osteoinduction [30]: In osteogenesis, viable osteoblasts or osteogenic cells (boneforming cells) are transplanted from one part of the body to the site where new bone is needed. Cancellous bone or marrow grafts provide such cells. Osteoconduction refers to the ability of some materials to serve as a scaﬀold on which bone cells can attach, migrate, grow, and divide. In this way, the bone-healing response is ‘‘conducted’’ through the graft site. Osteogenic cells function much better when they have an osteoconductive matrix to which they can attach. As only living cells can form new bone, the success of any bone-grafting procedure is dependent upon having suﬃcient osteogenic cells in the area. In some circumstances, the number of osteogenic cells in the surrounding tissues may be limited (areas of scarring, previous surgery or infection, bone gaps, areas treated with radiation therapy, etc.), in which cases osteogenesis might be applied in combination with the osteoconductive stimulation. The induction of bone formation refers to the capacity of some natural substances in the body to stimulate primitive ‘‘stem cells’’ or immature bone cells to grow and mature, forming healthy bone tissue. Most – but not all – of these agents are

115

116

8 Stimulation of Bone Growth on Implants by Integrin Ligands

proteins known as ‘‘peptide growth factors’’ or ‘‘cytokines’’. Several methods have been developed to prepare bone matrix which retains the natural growth factors. 8.2.2 Modiﬁcations on Implant Surfaces to Improve its Osseointegration

The osseointegration of biomimetic materials into bone tissue takes place when new bone forms in direct contact with the implant surface. This biological process is determined by initial cell–surface interactions and favored by osteoconductive scaﬀolds [31]. The ability of the biomaterial to allow osteoprogenitor cell adhesion and migration in the early stages of wound healing is crucial for later steps of the bone formation cascade. Numerous experimental studies have shown that bone–implant contact can be enhanced, in terms of both speed and intensity of bone formation, by modiﬁcations on the implant surface. These modiﬁcations might focus on optimizing the mechanical properties of the scaﬀold by mimicking the structure of the natural tissue. The morphology of an implant surface, including its microtopography and roughness, has been shown to be related to successful bone ﬁxation [32]. Osteoblasts exhibit roughness-dependent phenotypic characteristics; they tend to attach more easily to surfaces with a rougher microtopography, whereas ﬁbrous connective tissue is found more frequently on smooth implant surfaces [33]. Accelerated and increased bone contact with the implant surface was, for example, achieved by coating the implant with hydroxyapatite [34]. However, better results may be achieved when the implant surface is modiﬁed with bioligands such as peptides, adhesion proteins, enzymes, or growth factors. In this area, the most often employed approach targets integrin-speciﬁc ligands (mainly RGD peptides and RGD mimetics) to enable signaling to adjacent cells [17, 35]. This approach, which is depicted schematically in Figure 8.4, has resulted in advanced improvements in bone implant technology.

Fig. 8.4 Enhanced implant–bone interaction by integrin-speciﬁc ligands.

8.2 Improvements in Implant-Osseointegration by Surface Modiﬁcation with Integrin Ligands

Integrin-mediated cell adhesion to interfaces modiﬁed with speciﬁc ligands depends on many factors, including the aﬃnity and speciﬁcity of the ligands to the particular integrin, the mechanical strength of the ligand support, spacer length, overall ligand concentration, surface topography, and ligand density. The strategies that target integrin ligands to improve the osseointegration and longevity of these implants focus on promoting bone growth directly at the implant surface through control of all these parameters. In the absence of a direct bone-toimplant interface, anchorage of the implant occurs via an intermediate layer (up to 10 mm in thickness) of ﬁbrous connective tissue. This ﬁbrous tissue is mechanically unstable and is therefore detrimental to implant ﬁxation and the long-term stability of an endosseous implant. 8.2.3 Structure of the Coating Molecules

The surface-coating systems consist, in general, of four major parts (Fig. 8.5) that should be optimized for the design of successful biomimetic materials: Implant material: non-toxic, osteoconductive, proteolytic and mechanical stable. Anchor: this allows binding to the implant surface and depends directly on the implant material. In Figure 8.5 are included the most frequently employed implant–anchor systems. Biomaterial surfaces can be chemically modiﬁed using covalent coupling or self-assembly techniques. Some concrete examples will be described in the next section.

Fig. 8.5 The structure of coating molecules.

117

118

8 Stimulation of Bone Growth on Implants by Integrin Ligands

Spacer: this provides the required minimum distance between the RGD binding sequence and the anchoring moiety for eﬀective integrin-mediated cell adhesion on the implant surface. A change of its hydrophilicity resulted in no signiﬁcant diﬀerences on cell attachment [36]. Aminohexanoic acid, derivates of polyethyleneglycol (e.g., HEGAS: 20-amino-3,6,9,12,15,18-hexaoxaeicosanoic acid), as well as photoswitchable and photolabile units, were used. Speciﬁc integrin ligand (see Section 8.1.3). 8.2.4 Stimulation of Osteoblasts Adhesion and Proliferation on Implants Promoted by Integrin Ligands

Here, some examples – both in vivo and in vitro – are described to illustrate the enhancement of cell adhesion and proliferation induced by speciﬁc RGD peptides and peptidomimetics, using diﬀerent coating systems. 8.2.4.1 Poly(methyl methacrylate) The adhesion and proliferation of osteoblasts on poly(methyl methacrylate) (PMMA) discs in vitro was achieved by coating with cyclo(-RGDfK-) peptide using an acrylamide anchor [36]. This cyclic pentapeptide is selective for the av b 3 and the av b 5 integrin receptors [13a, 14, 23e, 37], which are expressed on osteoblasts. The lysine residue (or, if required, a glutamic acid residue) allows coupling of the RGD peptide to diﬀerent spacer-anchor systems which are developed for the improvement of implant integration. The distance between the RGD binding sequence and the anchoring moiety is a crucial parameter for cell adhesion. In this case, a minimum spacer length of about 3.5 nm was required for eﬀective integrin-mediated cell adhesion on the PMMA surface. The RGD peptide with an optimized spacer was linked covalently to the surface through an acrylamide anchor, which was radically polymerized onto the PMMA graft used as the bone implant. The coated surfaces bind eﬀectively to murine as well as to human osteoblasts, with the cell-adhesion rate rising as the ligand density on the surface increases. Even with the relatively high numbers of suspended cells (50 000 per cm 2 of surface), adhesion rates of up to 100% were observed on RGDfunctionalized PMMA (Fig. 8.6). Cells attached to coated surfaces formed focal adhesions and stretched on the surface, whereas the shape of cells attached to non-coated surfaces was spherical. Observing the osteoblasts bound to RGDcoated PMMA over a period of 22 days showed them to have proliferated (increasing in number by a factor of 10), and to have formed a homogeneous cell layer at the polymer surface. However, the osteoblasts attached to untreated PMMA died after a few days. Eﬀects in vivo [using the same cyclo(-RGDfK-) peptide] were investigated by the implantation of functionalized PMMA granulate cylinders into the patellar groove

8.2 Improvements in Implant-Osseointegration by Surface Modiﬁcation with Integrin Ligands

Fig. 8.6 Optical microscopy images of attached MC3T3H1 mouse osteoblasts (dark) on uncoated poly(methyl methacrylate) (PMMA) surfaces (A) and on PMMA bone cement coated with cyclic(-RGDfK-) peptide (the peptide concentration in the coating solution was 100 mM). Image (B) is

representative of the whole coated surface. Image (A) shows the only area of the uncoated PMMA where cell adhesion was observed. (Figure reproduced from Ref. [36] by copyright permission of Wiley-VCH Verlag GmbH & Co. KGaA.)

of rabbits [27a, 38]. PMMA RGD-coated pellets are integrated more quickly and more strongly into regenerating bone tissue of rabbits than are uncoated pellets (Fig. 8.7). The newly formed bone contacts the modiﬁed implant surface directly (no ﬁbrous layer was observed between the implant and bone tissue), and an ingrowth of bone tissue towards the center of the porous implant was clearly visible (Fig. 8.7A). In contrast, uncoated implants were separated from newly formed bone by a ﬁbrous tissue layer (Fig. 8.7B), which prevented any direct implant– bone interaction. Hence, coating the implants with av integrin-speciﬁc RGD peptides accelerated their osseointegration compared to uncoated, granulate cylinders.

Fig. 8.7 Cross-section of implanted poly(methyl methacrylate) (PMMA) implants (original magniﬁcation, 16). (A) RGDcoated PMMA implant; (B) uncoated PMMA implant. Color index: white ¼ PMMA beads; green-blue ¼ already existing bone and newly

formed bone [visible in image (A) around the PMMA beads]; light brown ¼ newly formed osteoid (bone precursor); dark brown ¼ ﬁbrous tissue. (Figure reproduced from Ref. [27a] by copyright permission of Wiley-VCH Verlag GmbH & Co. KGaA.)

119

120

8 Stimulation of Bone Growth on Implants by Integrin Ligands

8.2.4.2 Silks The unique mechanical properties of silks suggest also exciting opportunities for biomaterial scaﬀolds formed by these proteins and their variants, based on their unique mechanical properties (high strength, resistance to compressive forces, ﬂexibility, etc.), opportunities for genetic tailoring of structure and thus function and biocompatibility. Silk ﬁlms were covalently decorated with integrin recognition sequences (RGD) as well as other bioactive ligands involved in the induction of bone formation [39]. The osteoblast responses to the silk ﬁlms support the concept that the proteins serve as suitable bone-inducing matrices. Stimulated osteoblast-based mineralization in vitro (calcium deposition and nodule formation) was particularly signiﬁcant for RGD-modiﬁed silk. 8.2.4.3 Titanium Titanium and its alloys are some of the most-often employed scaﬀold materials for bone reconstructive purposes, due to their excellent biocompatibility and good mechanical properties [40]. Nevertheless, the osteoconductive properties of the material are insuﬃcient to allow complete osseointegration [41]. Many divergent strategies have been implemented in order to improve the clinical applications of titanium, with the best results having been obtained when the surface of the material was functionalized with RGD peptides. The evaluation (both, quality and quantity) of new bone formed in response to modiﬁed titanium rods was carried out in vivo (in the rat femur). The titanium surfaces were coated with the linear peptide RGDC (through the thiol group of the cysteine) [42]. In spite of a lack of speciﬁcity of this peptide towards osteoblasts, histomorphometric analysis of cross-sections showed a signiﬁcantly thicker shell of new bone formed around RGD-modiﬁed implants compared to plain implants after 2 weeks (26:2 G 1:9 versus 20:5 G 2:9 mm; P < 0:01). A noteworthy increase in bone thickness for RGD-coated implants was also observed after 4 weeks, while bone surrounding controls showed almost no change (32:7 G 4:6 versus 22:6 G 4:0 mm; P < 0:02). However, mechanical pull-out testing conducted after 4 weeks revealed that the average interfacial shear strength of peptide-modiﬁed rods was not statistically signiﬁcant, perhaps due to a smooth implant model, instability of the ﬁxation of thiol groups, or proteolytic degradation of the linear peptide. Similar in vivo studies were conducted using titanium implants (Ti6 Al4 V) coated with collagen and also with covalently bound cyclo(-RGDfK[3-mercaptopropionyl]-) onto the collagen. At 3 months after implantation into the alveolar crest of beagle dogs, the RGD-functionalized implants displayed a bone contact rate that was twice that observed with implants coated only with collagen [43]. The same peptide with thiol anchor was used to coat press-ﬁt porous-Ti6 Al4 V implants inserted in the cancellous bone region of a canine tibia [44]. Histomorphometry and mechanical push-out testing showed encouraging results, with a signiﬁcant twofold increase in bone ongrowth and reduced ﬁbrous tissue ﬁxation being observed for the RGD-coated implants. Bone volume was signiﬁcantly enhanced in a 0 to 100 mm zone around the implant. A moderate

8.2 Improvements in Implant-Osseointegration by Surface Modiﬁcation with Integrin Ligands

increase in mechanical ﬁxation, median ultimate shear strength and energy to failure were also described. Positive eﬀects were also observed when a RGDcoated porous titanium ﬁber mesh implant was inserted into the cranium of a rabbit [45]. The cell adhesion of human osteoprogenitor cells on titanium surfaces (Ti6 Al4 V), coated with a range of linear and cyclic RGD peptides, was examined along with the integrin expression on these cells [46]. Cyclo(-RGDfK-) was observed not only to promote cell adhesion of human bone marrow stromal cells (HBMSC) on biomaterials but also to increase cell diﬀerentiation and mineralization through the activation of tyrosine kinases, focal adhesion kinase (p125FAK) and mitogen-activated protein (MAP) kinases. In this study, RGD linear peptides did not enhance osteoblast diﬀerentiation. These data have significant clinical potential for bone tissue engineering in humans, as they suggest that cyclo(-RGDfK-) constitutes a good candidate to promote not only cell adhesion but also the ingrowth and the production of a bone-speciﬁc matrix once grafted onto a functionalized material. To summarize, the results of all of these studies have shown that cyclic-RGD coating increases the osseointegration of different types of titanium implant. 8.2.4.4 RGD Mimetics Another promising strategy to control the interactions between surfaces and adherent cells focuses on the covalent grafting of speciﬁc peptides to polymers on which unspeciﬁc cell adhesion (protein adsorption) is essentially suppressed. Thus, diblock copolymers based on monoamine poly(ethylene glycol)-blockpoly(d,l-lactic acid) (H2 N-PEG-PLA) were functionalized with av b 3 =av b 5 integrinspeciﬁc cyclic RGD peptides to promote the adhesion of human osteoblasts. These adhesion experiments revealed a signiﬁcant increase in the number of cells and spreading on the RGD-modiﬁed surfaces [47]. The coating with RGD mimetics constitutes another attractive alternative to promote osteoblast adhesion on implant surfaces. A number of non-peptidic av -selective RGD mimetics have been developed [16, 48] as potential drugs to treat cancer, osteoporosis, acute renal failure, restenosis, arthritis, and retinopathy. However, their use to enhance cell adhesion has been only recently described [49]. The suitable positions for anchor groups, necessary to attach the RGD mimetics to implant surfaces, without interfering with integrin binding, were identiﬁed by taking into account modeling studies on the non-peptidic av b 3 ligands [50] (using the X-ray structure of the av b 3 head group [51] containing the cyclic peptide cilengitide [52]). As the guanidine and carboxy groups of the ligand are essential for binding to the integrin subunits a and b, respectively [18b], the aromatic rings of the highly av b 3 -selective diacylhydrazine scaﬀold [16] were chosen to position the anchor groups (R 1 and R 2 ; Fig. 8.8) [49b]. Substances 1 to 3 were irreversibly linked to titanium (Ti6 Al4 V) surfaces through a thiol anchor, in order to evaluate their cell-adhesion properties (Fig. 8.8). The plating eﬃciency was increased to 42.9% (100 mm compound 2 in coating solution) compared to 9.4% on the unmodiﬁed titanium. Compound 2 was shown to stimulates cell

121

122

8 Stimulation of Bone Growth on Implants by Integrin Ligands

Fig. 8.8 Adhesion of MC3T3E1 mouse osteoblasts on uncoated and RGD mimetic-coated titanium surfaces.

adhesion as eﬃciently as the cyclo(-RGDfK[3-mercaptopropionyl]-) peptide [27b]. Compound 1 was slightly less potent, most likely due to the signiﬁcantly shorter linker, which makes the integrin ligand less accessible to the integrin. Compound 3, despite having comparable activity in the binding assay of isolated av b 3 , provided no stimulation of osteoblast adhesion when bound to the titanium surface in repeated testing, though this may have been due to an unfavored orientation for integrin binding. In conclusion, RGD mimetics 1 and 2 are av -selective integrin ligands for surface coating which exhibit a potency for stimulated osteoblast adhesion similar to cyclic RGD peptides when immobilized on titanium. The advantages of mimetics compared to RGD peptides include greater stability to enzymatic degeneration, pH variations and heat, as well as a much cheaper synthesis. New anchor systems have also been studied to allow for an easy and practical coating of implants with av -speciﬁc integrin ligands. Thus, a simple – but eﬃcient – method for biofunctionalization of titanium surfaces with the av -speciﬁc cyclo(-RGDfK-) peptide, through branched phosphonic acid anchors, has been recently reported [53]. The anchor system consisted of four phosphonopropionic acids linked together by a branching unit composed of three Lys residues, which improved binding to the Ti surface by the multimeric eﬀect. These provided extremely tight binding and, when compared to a thiol anchor, the amount of immobilized peptide was greater [44]. It is clear that the same coating compound can also be used to stimulate cell attachment to apatite or other potassium phosphates. The photochemical control of cell adhesion by changing the distance and orientation of the integrin ligands to the surface through light-induced E,Z isomerization was also recently described [54]. In a recent in vitro study of osseointegration through integrin ligands, the surface coating of freshly resected and cleaned human bone discs with cyclic RGD-peptides (using phosphonate anchors) was examined [55]. Human osteo-

References

blasts were seeded onto the RGD-coated bone discs, after which the adherence and growth of the cells was analyzed. The coating of bone discs with RGDpeptides did not improve the adhesion rate of osteoblast-like cells to the discs, but signiﬁcantly accelerated (by up to 40%) the growth of these cells within 8 days of attachment. It is intriguing to speculate, therefore, that the coating of bone implants with RGD-peptides prior to reimplantation might serve as a novel approach to accelerate the ingrowth and healing of a bone implant.

8.3 Conclusions

The biofunctionalization of diﬀerent materials with av -speciﬁc integrin ligands has led to promising results in the ﬁeld of orthopedic and craniofacial implant technology. This methodology not only results in an enhanced cell adhesion but also increases cell diﬀerentiation and mineralization. As a consequence, ﬁbrous tissue formation around the implant is signiﬁcantly reduced, thereby improving osseointegration of the implant in vivo. These results impact greatly on the development of clinically eﬀective biomaterials for bone grafting.

References 1 (a) R. Langer, L.G. Cima, J.A.

2

3 4

5 6

Tamada, E. Wintermantel, Biomaterials 1990, 11, 738–745; (b) H. Shin, S. Jo, A.G. Mikos, Biomaterials 2003, 24, 4353–4364. (a) M. Epple, J.M. Rueger, Chem. Tech. Lab. 1999, 47, 1405–1410; (b) R. Thull, Materialwiss. Werkst. 2001, 32, 949–952. L.G. Griﬃth, G. Naughton, Science 2002, 295, 1009–1014. (a) T. Blunk, A. Gopferich, J. Tessmar, Biomaterials 2003, 24, 4335–4335; (b) H. Sinn, S. Jo, A.G. Mikos, Biomaterials 2003, 24, 4335–4335; (c) S. Drotleﬀ, U. Lungwitz, M. Breunig, A. Dennis, T. Blunk, J. Tessmar, A. Go¨pferich, Eur. J. Pharm. Biopharm. 2004, 58, 385–407. R.O. Hynes, Cell 1992, 69, 11–25. (a) E. Ruoslahti, M.D. Pierschbacher, Science 1987, 238, 491–497; (b) S.M. Albelda, C.A. Buck, FASEB J. 1990, 4, 2868–2880; (c) J. Travis, Science 1993, 260, 906–908.

7 A. van der Flier, A. Sonnenberg,

Cell Tissue Res. 2001, 305, 285–298. 8 M.D. Pierschbacher, E. Ruoslahti,

Nature 1984, 309, 30–33. 9 M. Pfaﬀ, Recognition sites of RGD-

dependent integrins, in: J.A. Eble (Ed.), Integrin-Ligand Interaction. Springer-Verlag, Heidelberg, 1997, pp. 101–121. 10 (a) J.M. Li, M.J. Menconi, H.B. Wheeler, M.J. Rohrer, V.A. Klassen, J.E. Ansell, M.C. Appel, J. Vasc. Surg. 1992, 15, 1010–1017; (b) T. Miyata, M.S. Conte, L.A. Trudell, D. Mason, A.D. Whittemore, L.K. Birinyi, J. Surg. Res. 1991, 50, 485–493; (c) R. Vohra, G.J. Thomson, H.M. Carr, H. Sharma, M.G. Walker, Br. J. Surg. 1991, 78, 417–420; (d) G.J. Thomson, R.K. Vohra, M.H. Carr, M.G. Walker, Surgery 1991, 109, 20–27; (e) J. Kaehler, P. Zilla, R. Fasol, M. Deutsch, M. Kadletz, J. Vasc. Surg. 1989, 9, 535–541; (f ) J.M. Sentissi, K. Ramberg, Jr., T.F. O’Donnell, R.J.

123

124

8 Stimulation of Bone Growth on Implants by Integrin Ligands

11

12

13

14

15

16

17 18

19 20

21

Connolly, A.D. Callow, Surgery 1986, 99, 337–343; (g) J.M. Seeger, N. Klingman, J. Surg. Res. 1985, 38, 641–647. (a) N. Weix, D. Klee, H. Hocker, Biomaterials 2001, 2, 81–86; (b) Y. Ito, M. Kajihara, Y. Imanishi, J. Biomed. Mater. Res. 1991, 25, 1325–1337; (c) T. Boxus, R. Touillaux, G. Dive, J. Marchand-Brynaert, Bioorg. Med. Chem. 1998, 6, 1577–1595; (d) J.A. Neﬀ, K.D. Caldwell, P.A. Tresco, J. Biomed. Mater. Res. 1998, 40, 511–519. (a) G.B. Fields, J.L. Lauer, Y. Dori, P. Forns, Y.-C. Yu, M. Tirrell, Biopolymers 1998, 47, 143–151. (a) M. Aumailley, M. Gurrath, G. Muller, J. Calvete, R. Timpl, H. Kessler, FEBS Lett. 1991, 291, 50–54; (b) M. Gurrath, G. Mu¨ller, H. Kessler, M. Aumailley, R. Timpl, Eur. J. Biochem. 1992, 210, 911–921; (c) B. Ivanov, W. Grzesik, F.A. Robey, Bioconjug. Chem. 1995, 6, 269–277. R. Haubner, D. Finsinger, H. Kessler, Angew. Chem. Int. Ed. 1997, 36, 1374– 1389. (a) U. Hersel, C. Dahmen, H. Kessler, Biomaterials 2003, 24, 4385–4415; (b) P. Shaﬀner, D. Dard, Cell Mol. Life Sci. 2003, 60, 119–132. G.A.G. Sulyok, C. Gibson, S.L. Goodman, G. Ho¨lzemann, M. Wiesner, H. Kessler, J. Med. Chem. 2001, 44, 1938–1950. R.G. Lebaron, K.A. Athanasiou, Tissue Eng. 2000, 6, 85–103. (a) M.A. Arnaout, Immunol. Rev. 2002, 186, 141–163; (b) K.-E. Gottschalk, H. Kessler, Angew. Chem. Int. Ed. 2002, 41, 3767–3774. A.R. Horwitz, J.T. Parsons, Science 1999, 286, 1102–1103. (a) A. van der Flier, A. Sonnenberg, Cell Tissue Res. 2001, 305, 285–298; (b) B. Geiger, A. Bershadsky, Curr. Opin. Cell Biol. 2001, 13, 584–592; (c) N.J. Boudreau, P.L. Jones, Biochem. J. 1999, 339, 481–488; (d) M.A. Schwartz, Trends Cell Biol. 2001, 11, 466–470. (a) F.G. Giancotti, Nat. Cell Biol. 2000, 2, E13–E14; (b) C.S. Chen, M.

22

23

24

25

26 27

28

29

Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Science 1997, 276, 1425– 1428. (a) E. Ruoslahti, M.D. Pierschbacher, V.L. Woods, Bull. Inst. Pasteur 1994, 92, 242–247; (b) D.G. Stupack, X.S. Puente, S. Boutsaboualoy, C.M. Storgard, D.A. Cheresh, J. Cell Biol. 2001, 155, 459–470. (a) M. Grano, P. Zigrino, S. Colucci, G. Zambonin, L. Trusolino, M. Serra, N. Baldini, A. Teti, P.C. Marchisio, A.Z. Zallone, Exp. Cell Res. 1994, 212, 209–218; (b) E. Noiri, J. Gailit, D. Sheth, H. Magazine, M. Gurrath, G. Muller, H. Kessler, M.S. Goligorsky, Kidney Int. 1994, 46, 1050–1058; (c) S. Stromblad, J.C. Becker, M. Yebra, P.C. Brooks, D.A. Cheresh, J. Clin. Invest. 1996, 98, 426–433; (d) S. Cheng, W.S. Craig, D. Mullen, J.F. Tschopp, D. Dixon, M.D. Pierschbacher, J. Med. Chem. 1994, 37, 1–8; (e) R. Haubner, R. Gratias, B. Diefenbach, S.L. Goodman, A. Jonczyk, H. Kessler, J. Am. Chem. Soc. 1996, 118, 7461–7472. (a) C.D. Buckley, D. Pilling, N.V. Henriquez, G. Parsonage, K. Threlfall, D. Scheel-Toellner, D.L. Simmons, A.N. Albar, J.M. Lord, M. Salmon, Nature 1999, 397, 534–539; (b) R.O. Hynes, Nat. Med. 2002, 8, 918–921. C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Science 1997, 276, 1425–1428. S.P. Massia, J.A. Hubbell, J. Cell. Biol. 1991, 114, 1089–1100. (a) M. Kantlehner, P. Schaﬀner, D. Finsinger, J. Meyer, A. Jonczyk, B. Diefenbach, B. Nies, G. Ho¨lzemann, S.L. Goodman, H. Kessler, ChemBioChem. 2000, 1, 107–114; (b) B. Jeschke, J. Meyer, A. Jonczyk, H. Kessler, P. Adamietz, N.M. Meenen, M. Kantlehner, C. Goepfert, B. Nies, Biomaterials 2002, 23, 3455– 3463. J. Auernheimer, R. Haubner, M. Schottelius, H.-J. Wester, H. Kessler, Helv. Chim. Acta 2006, 89, 833–840. M. Arnold, E.A. Cavalcanti-Adam, R. Glass, J. Blu¨mmel, W. Eck, M.

References

30 31

32 33

34

35

36

37

38

39

40

Kantlehner, H. Kessler, J.P. Spatz, ChemPhysChem. 2004, 5, 383–388. J. Glowacki, J.B. Mulliken, Clin. Plast. Surg. 1985, 12, 233–241. (a) D.A. Puleo, A. Nanci, Biomaterials 1999, 20, 2311–2321; (b) K. Anselme, Biomaterials 2000, 21, 667–681. K. Thomas, S.D. Cook, J. Biomed. Mater. Res. 1985, 19, 875–901. J. Lincks, B.D. Boyan, C.R. Blanchard, C.H. Lohmann, Y. Liu, D.L. Cochran, D.D. Dean, Z. Schwartz, Biomaterials 1998, 19, 2219–2232. (a) M. Rocca, M. Fini, G. Giavaresi, N. Nicoli Aldini, R. Giardin, Int. J. Artif. Organs 2001, 24, 649–654; (b) G. Balasundaram, M. Sato, T.J. Webster, Biomaterials 2006, 27, 2798–2805. (a) M.D. Pierschbacher, E. Ruoslahti, J. Biol. Chem. 1987, 262, 17294– 17298; (b) A. Rezania, K.E. Healy, Biotechnol. Prog. 1999, 15, 19–32; (c) T. Matsuura, R. Hosokawa, K. Okamoto, T. Kimoto, Y. Akagawa, Biomaterials 2000, 21, 1121–1127. M. Kantlehner, D. Finsinger, J. Meyer, P. Schaﬀner, A. Jonczyk, B. Diefenbach, B. Nies, H. Kessler, Angew. Chem. Int. Ed. 1999, 38, 560–562. (a) M. Pfaﬀ, K. Tangemann, B. Mu¨ller, M. Gurrath, G. Mu¨ller, H. Kessler, R. Timpl, J. Engel, J. Biol. Chem. 1994, 269, 20233–20238; (b) M. Friedlander, P.C. Brooks, R.W. Shaﬀer, C.M. Kincaid, J.A. Varner, D.A. Cheresh, Science 1995, 270, 1500–1502. P. Schaﬀner, J. Meyer, M. Dard, R. Wenz, B. Nies, S. Verrier, H. Kessler, M. Kantlehner, J. Mater. Sci. 1999, 10, 837–839. (a) S. Soﬁa, M.B. McCarthy, G. Gronowicz, D.L. Kaplan, J. Biomed. Mater. Res. 2001, 54, 139–148; (b) J. Chen, G.H. Altman, V. Karageorgiou, R. Horan, A. Collette, V. Volloch, T. Colabro, D.L. Kaplan, J. Biomed. Mater. Res. 2003, 67A, 559–570. (a) D.M. Brunette, P. Tengvall, M. Textor, P. Thomsen (Eds.), Titanium in medicine. Springer-Verlag, Berlin,

41

42

43

44

45

46

47

48

49

2001; (b) O.E. Pohler, Injury 2000, S4(31), D7–D13; (c) J.A. Jansen, A.F. von Recum, J.P. van der Waerden, K. de Groot, Biomaterials 1992, 13, 959– 968. J. van den Dolder, E. Farber, P.H.M. Spauwen, J.A. Jansen, Biomaterials 2003, 24, 1745–1750. D.M. Ferris, G.D. Moodie, P.M. Dimond, C.W.D. Gioranni, M.G. Ehrlich, R.F. Valentini, Biomaterials 1999, 20, 2323–2331. H. Schliephake, D. Scharnweber, M. Dard, S. Ro¨ßler, A. Sewing, J. Meyer, D. Hoogestraat, Clin. Oral Impl. Res. 2002, 13, 312–319. B. Elmengaard, J.E. Bechtold, K. Søballe, Biomaterials 2005, 26, 3521–3526. H.C. Kroese-Deutman, J. Van den Dolder, P.H.M. Spauwen, J.A. Jansen, Tissue Eng. 2005, 11, 1867–1875. (a) S. Verrier, S. Pallu, R. Bareille, A. Jonczyk, J. Meyerc, M. Dard, J. Ame´de´e, Biomaterials 2002, 23, 585– 596; (b) S. Pallu, R. Bareille, M. Dard, H. Kessler, A. Jonczyk, M. Vernizeau, J. Ame´de´e-Vilamitjana, Peptides 2003, 24, 1349–1357; (c) M.C. Porte´Durrieu, F. Guillemot, S. Pallu, C. Labruge`re, B. Brouillaud, R. Bareille, J. Ame´de´e, N. Barthe, M. Dard, Ch. Baquey, Biomaterials 2004, 25, 4837– 4846. E. Lieb, M. Hacker, J. Tessmar, L.A. Kunz-Schughart, J. Fiedler, C. Dahmen, U. Hersel, H. Kessler, M.B. Schulza, A. Go¨pferich, Biomaterials 2005, 26, 2333–2341. (a) C. Gibson, G.A.G. Sulyok, D. Hahn, S.L. Goodman, G. Ho¨lzemann, H. Kessler, Angew. Chem. Int. Ed. 2001, 40, 165–169; (b) G. Ho¨lzemann, Drugs 2001, 4, 72–81; (c) J.S. Kerr, A.M. Slee, S.A. Mousa, Expert Opin. Invest. Drugs 2000, 9, 1271–1279; (d) R. Dayam, F. Aiello, J. Deng, Y. Wu, A. Garofalo, X. Chen, N. Neamati, J. Med. Chem. 2006, 49, 4526–4534. (a) S. Biltresse, M. Attolini, G. Dive, A. Cordi, G.C. Tucker, J. MarchandBrynaert, Bioorg. Med. Chem. 2004, 12, 5379–5393; (b) C. Damen, J.

125

126

8 Stimulation of Bone Growth on Implants by Integrin Ligands Auernheimer, A. Meyer, A. Enderle, S.L. Goodman, H. Kessler, Angew. Chem. Int. Ed. 2004, 43, 6649–6652. 50 L. Marinelli, A. Lavecchia, K.-E. Gottschalk, E. Novellino, H. Kessler, J. Med. Chem. 2003, 46, 4393–4404. 51 J.-P. Xiong, T. Stehle, R. Zhang, A. Joachimiak, M. Frech, S.L. Goodman, M.A. Arnaout, Science 2002, 296, 151–155. 52 M.A. Dechantsreiter, E. Planker, B. MathI, E. Lohof, G. Ho¨lzemann, A. Jonczyk, S.L. Goodman, H. Kessler, J. Med. Chem. 1999, 42, 3033–3040.

53 (a) J. Auernheimer, D. Zukowski, C.

Dahmen, M. Kantlehner, A. Enderle, S.L. Goodman, H. Kessler, ChemBioChem. 2005, 6, 2034–2040; (b) J. Auernheimer, H. Kessler, Bioorg. Med. Chem. Lett. 2006, 16, 271–273. 54 J. Auernheimer, C. Dahmen, U. Hersel, A. Bausch, H. Kessler, J. Am. Chem. Soc. 2005, 127, 16107–16110. 55 U. Magdolen, J. Auernheimer, C. Dahmen, J. Schauwecker, H. Gollwitzer, J. Tu¨bel, R. Gradinger, H. Kessler, M. Schmitt, P. Diehl, Int. J. Mol. Med. 2006, 17, 1017–1021.

127

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles from Titanium and other Materials Fumio Watari, Kazuchika Tamura, Atsruro Yokoyama, Kenichiro Shibata, Tsukasa Akasaka, Bunshi Fugetsu, Kiyotaka Asakura, Motohiro Uo, Yasunori Totsuka, Yoshinori Sato, and Kazuyuki Tohji

Abstract

Biocompatible titanium (Ti) causes inﬂammation when in the form of abraded ﬁne particles, whereas asbestos, which is a type of clay mineral, induces mesothelioma after long-term, high-level exposure. In addition to these materials’ properties, such as toxicity or biocompatibility, these phenomena may be seen to occur as a result of the ‘‘particle eﬀect’’. The cytotoxicity of ﬁne particles of Ti, Fe, Ni and TiO2 was investigated in vitro using human neutrophils as probe cells, and also with the tissue reaction in-vivo implantation test. Biochemical functional analyses of cell survival rate, lactate dehydrogenase (LDH) activities, superoxide anion, cytokines and the microscopic observation of cellular morphology showed that the stimulatory eﬀects on neutrophils and inﬂammation in soft tissue became more prominent as the particle size became smaller (<100 mm). Moreover, such eﬀects were especially pronounced for particles <10 mm (about cell size), when phagocytosis was induced. Inductively coupled plasma elemental analysis showed dissolution from Ti particles to be negligible. Results with Fe were quantitatively similar to those with Ti, despite Fe being soluble. Taken together, these results indicated that the stimulus produced is based, non-speciﬁcally, on the physical size and shape eﬀect of particles, and is more pronounced on the micro/nano scale. This is diﬀerent from the material-dependent, chemical toxicity eﬀects that are caused by ionic dissolution and normally dominant in bulk materials. Key words: nanotoxicology, biocompatibility, titanium, superoxide, cytokine, neutrophil, macrophage, phagocytosis.

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

128

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

9.1 Introduction

Metallic titanium (Ti) is highly corrosion-resistant due to the presence of a thin and stable protective oxide layer which forms on its surface. Moreover, it is one of the most biocompatible metals known to mankind [1, 2], and hence is regarded as the near-ideal material for implantation, being used widely in orthopedics and dentistry. Unfortunately, Ti has a weak point in that it has a low abrasion resistance [3, 4], whereby abraded ﬁne particles produced from the sliding parts of artiﬁcial joints often cause inﬂammation in the surrounding tissues. The reason why this material behaves in diﬀerent ways depending on its particle size is unclear. By comparison, asbestos – a silicaceous clay-type mineral – induces mesothelioma after long-term, high-level exposure. These phenomena may originate not only from the material’s biological properties, such as toxicity or biocompatibility, but also from particulate eﬀects, including physical size and morphology. Recently, signiﬁcant advances have been made in the development of a drug delivery system (DDS) to administer anticancer agents and to conduct gene transfections. In this regard, it is essential to establish an understanding of the principles involved, and of the biological reactivity of micro/nanoparticles when developing these biomedical applications. In this chapter we present details of studies aimed at determining the eﬀects of ﬁne particles, in terms of their size, on cytotoxicity in vitro and biocompatibility in vivo. These investigations included biochemical functional analyses with human neutrophils as probe cells, as well as histological observations in animal implantation tests [5–7]. In this way, the eﬀects of Ti, Fe, Ni and TiO2 particles could be compared directly. Human neutrophils, which play a central, non-speciﬁc role in the initial stages of inﬂammation resulting from contact with a foreign body(ies), were used as probes. Particles that were either smaller (0.5 to 3 mm) or larger (10, 50, and 150 mm) than neutrophils were used to determine the relationship between cell or particle size and cytotoxicity.

9.2 Materials and Methods 9.2.1 Specimens

Ti, Fe, Ni and TiO2 particles of 99.9% purity and of various sizes (from 300 nm to 150 mm) were used in these experiments. In order to equate experimental conditions between materials of the particle group, particles of 0.5, 3 and 10 mm were extracted by sedimentation, while those <300 nm in size were extracted by ultra-

9.2 Materials and Methods

ﬁltration. Fullerene (C60 ), multiwalled carbon nanotubes (CNT) and hat-stacked carbon nanoﬁbers (CNF) – a derivative of CNTs – were used for some additional experiments. 9.2.2 Dissolution Testing of Ti Particles

After immersion of Ti particles in Hank’s balanced salt solution (HBSS) at 37 C for one month, the suspension was ﬁltered through a 0.45-mm pore membrane to remove Ti particles. The supernatant was then analyzed using inductively coupled plasma-atomic emission spectrometry (ICP-AES) elemental analysis (ICPS-8100; Shimazu, Tokyo, Japan). 9.2.3 Probe Cells

Neutrophils were separated from human peripheral blood obtained from healthy volunteers, using 6% isotonic sodium chloride containing hydroxyethyl starch and lymphocyte isolation solution (Ficoll-Hypaque TM ; Amersham Pharmacia Biotech AB, Sweden). After mixing with HBSS, the particles were maintained at 37 C, the neutrophils added, and the whole mixture was used for a variety of cell toxicity tests. A human acute monocytic leukemia cell line (THP-1) was also used for additional experiments. 9.2.4 Biochemical Analyses of Cellular Reactions to Materials

Cell survival rates, LDH activities and superoxide anion (O 2 ) production per 10 6 neutrophils were measured. Cytokines of tumor necrosis factor alpha (TNF-a) and interleukin 1b (IL-1b) were measured using ELISA kits (Endogen, Inc., USA). Morphological changes of neutrophils mixed with HBSS and containing various particles were observed using optical microscopy (OM) (Zeiss Axioskop; Germany) and scanning electron microscopy (SEM) (Hitachi S-4300; Tokyo, Japan). 9.2.5 Animal Experiments

Particles were inserted into the subcutaneous connective tissue of the abdominal region of Wistar rats (aged 11–12 weeks; body weight 350–380 g). Specimens were prepared through the usual process of ﬁxation, embedding, sectioning, and staining with hematoxylin and eosin, and then observed histopathologically. A compulsory exposure test was also performed using 30-nm TiO2 particles. The observation of internal diﬀusion of nanoparticles was conducted by elemental mapping in air using X-ray scanning analytical microscopy (XSAM) (Horiba

129

130

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

XGT-2000V; Tokyo, Japan) without the pretreatments of ﬁxation, dehydration and staining after sectioning.

9.3 Results 9.3.1 Dependence of Tissue Reaction In Vivo on Material Macroscopic Size

A comparison of tissue reaction with macroscopic size (1 mm 10 mm) of Ni, Ti, Fe and Ag after one week of implantation in the dorsal thoracic region of rats is shown in Figure 9.1. In each case, the implant was originally situated in the upper space. With Ni, there was an expansion of capillary vessels, and the tissues in the distant regions became necrotic and degenerated. Fibrous connective tissue surrounding the implant was already formed for Ti and Fe from the earlier stage, and was still in the course of formation for Ag at this stage. Such a comparison highlights the diﬀerent sensitivities of biocompatibility of these metals [5].

Fig. 9.1 Histological image of rat soft tissue after one-week implantation with (a) Ni, (b) Ti, (c) Fe and (d) Ag particles of macroscopic size (1 mm 10 mm).

9.3.2 Eﬀect of Particle Size on Biocompatibility 9.3.2.1 Size Distribution of the Abraded Particles A typical example of size distribution of abraded particles from Ti, produced by grinding with a dental air turbine, is shown in Figure 9.2. The largest proportion of particles were approximately 5 mm in size, followed by 0.8 mm, 0.2 mm, and

9.3 Results

Fig. 9.2 Typical size distribution of abraded particles of Ti produced by grinding with a dental air turbine.

0.07 mm. These more sensitive particle sizes, based on biological reaction, were further investigated in the following sections. 9.3.2.2 Particle Size Dependence In Vitro The survival rate of human neutrophils in HBSS containing Ti particles is shown graphically in Figure 9.3, with HBSS solution used as the control. The average

Fig. 9.3 Dependence of cell survival rate on Ti particle size [6].

131

132

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

Fig. 9.4 Dependence of interleukin (IL)-1b release from neutrophils on Ti particle size [6].

survival rate was seen to decrease in line with decreases in particle size, with signiﬁcant diﬀerences from control being observed for the 0.5 and 3 mm particles. ICP elemental analysis showed that the dissolution from Ti particles was negligible (i.e., below detection limits) [6]. LDH activities in each particle solution were related inversely to the cell survival rate, with signiﬁcant diﬀerences from control being observed for the 0.5 and 3 mm particles. Superoxide production also tended to increase with decreasing particle size, and became prominent as the particle size reached <10 mm. Interleukin-1b release from neutrophils in HBSS containing Ti particles is shown in Figure 9.4, and tended to increase as the particle sizes decreased. The most pronounced increases were for the 0.5 and 3 mm particles. The release of cytokine TNF-a showed a similar behavior to that for IL-1b. Scanning electron microscopy and OM images of human neutrophils exposed to Ti, TiO2 and Ni particles in HBSS solution are shown in Figure 9.5. The morphology of the neutrophil in HBSS as a control (see Fig. 9.5a) changes, depending on the degree of stimulus. When stimulated by 0.5 or 3 mm Ti particles, the neutrophil surface was changed to become either knobbly or smooth, by transformation of the cell membrane. In addition, a neutrophil may extend its pseudopodia in order to phagocytose a TiO2 or Ti particle (Fig. 9.5c,d); the particles already phagocytosed can be observed inside the cell (Fig. 9.5d). The morphology of neutrophils exposed to Ni particles was generally either transformed or destroyed (Fig. 9.5b). In the case of Ti particles larger than 10 mm, phagocytosis was not

9.3 Results

Fig. 9.5 (a,b,d) Scanning electron microscopy and (c) optical microscopy images of human neutrophils. (a) Control neutrophil in HBSS; (b) after exposure to particles of Ni (500 nm); (c) after exposure to TiO2 particles (300 nm); (d) after exposure to Ti particles (500 nm) [6].

observed and the form of the neutrophils was changed only minimally. The pronounced phenomena of biochemical cell reactions observed for particle sizes below 10 mm (see Figs. 9.3 and 9.4) were closely related to the phagocytosis (Fig. 9.5). 9.3.2.3 Particle Size Dependence In Vivo A series of in-vivo tests showed that Ti particles larger than 100 mm were surrounded by a ﬁbrous connective tissue layer, which is the usual reaction for biocompatible materials such as the bulk size of a Ti implant. As the particle sizes became smaller, inﬂammation was seen to occur; however, the presence of particles < 10 mm (about cell size) caused phagocytosis by macrophages to be induced. Numerous inﬂammatory cells were observed in the surroundings of the Ti particles, and both macrophages and neutrophils showed degenerative changes in their morphology [7]. The histological images of rat soft tissues into which 3 mm and 10 mm Ti particles had been inserted for 5 days are shown in Figure 9.6. In the case of the 3 mm particles, macrophage-based phagocytosis had occurred and the Ti particles were observed inside the cells, such that the cytoplasm of an inﬂammatory cell contained numerous small black particles. In contrast, the 10 mm Ti particles were seen to locate outside the cells, with phagocytosis rarely being observed and the tissue showing a much lesser degree of inﬂammation.

133

134

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

Fig. 9.6 Tissue reaction to (a) 3 mm and (b) 0.5 mm Ti particles after 5-day implantation.

Fig. 9.7 Dependence of TNF-a release from neutrophils on particle size for Ti, Fe, and Ni.

9.3 Results

9.3.2.4 Material Dependence of the Particle Size Eﬀect In Vitro The comparative release of TNF-a by Ti, Fe and Ni particles is illustrated graphically in Figure 9.7. Both, Ti and Fe particles, showed a tendency towards increased TNF-a release as the particle size decreased (and to similar degrees), whereas Ni demonstrated a particle size-dependent release, with relatively lower values. Of note, the behavior of Fe was almost equal to that of Ti in a quantitative sense, despite clear diﬀerences in the chemical properties of these two materials. Interleukin-1b, superoxide anion and LDH production showed, likewise, a similar tendency to increase compared to a decrease in particle size for both Ti and Fe. Ni also demonstrated a particle size-dependency in the cell functional tests, although the quantitative values obtained diﬀered from those obtained with Ti and Fe. Typically, the cell survival rate was signiﬁcantly lower than in controls, but LDH activity was higher. However, the degree of superoxide anion and IL-1b production was less with Ni than with Ti and Fe. 9.3.2.5 Material Dependency of Tissue Reaction to Particles In Vivo Comparative histopathological observations of the soft tissues of rats into which 3 mm particles of Ti, Fe and Ni were inserted for 5 days are shown in Figure 9.8.

Fig. 9.8 Histopathological observation of in-vivo reaction to 3 mm particles. (a) Ti; (b) Fe; (c) Ni.

135

136

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

Macrophage-based phagocytosis was eﬀected for both Ti and Fe particles, while necrosis occurred in the case of Ni particles. 9.3.3 Shape Eﬀect

Scanning electron microscopy images of a set of massive and acicular TiO2 particles, of equivalent diametric size and longitudinal length (ca. 10 mm), are shown in Figure 9.9. The subsequent particle size-dependence of IL-1b release from neutrophils for these TiO2 particles is illustrated graphically in Figure 9.10. TNF-a

Fig. 9.9 Scanning electron microscopy images of massive and acicular TiO2 particles with equivalent diametric size and longitudinal length, respectively.

Fig. 9.10 Particle size dependence of IL-1b release from neutrophils for massive and acicular TiO2 particles.

9.3 Results

release, which is related to phagocytosis, showed very similar size-dependencies for both massive and acicular particles, but for IL-1b the massive particle shape showed a dependency similar to that for TNF-a, where cytokine production was enhanced below a particle size of 10 mm, but reduced for particles > 10 mm (Fig. 9.10). In contrast, production levels of IL-1b remained high for 10 mm acicular particles, indicating that this type of particle has a more stimulatory role than do massive particles. Of note, IL-1b represents a diﬀerent type of stimulus and inﬂammation compared to TNF-a, despite both cytokines being indicators of inﬂammation. 9.3.4 The Origin of the Particle Size Eﬀect

For the same material, it is clear that cell and tissue reactions diﬀer between the bulk state and ﬁne particles. The results of both in-vitro biochemical cell functional tests and in-vivo animal tests were in accordance with the ﬁnding that, when in bulk, even biocompatible materials such as Ti and TiO2 become stimulatory as their particle size is decreased, and that this is especially pronounced below a size of 10 mm, when phagocytosis is induced. The relationship between cell/tissue and particle size is shown schematically in Figure 9.11. The cellstimulatory and tissue-inﬂammatory natures of particles, as described above, originate from the relative size relationship between the particle and the cell/tissue. Previously, these phenomena were characterized as non-speciﬁc cytotoxicity aris-

Fig. 9.11 The relationship between cell/tissue and particle size.

137

138

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

ing from physical size eﬀects; these diﬀer from chemical toxicity eﬀects, which are based on ionic dissolution and usually are dominant in bulk materials. 9.3.5 Toxicity Level of Particle Size Eﬀect for Bioactive and Bioinert Materials

The comparative production of TNF-a in THP-1 cells (a human acute monocytic leukemia cell line) following treatment with fullerene (C60 ), carbon nanoﬁbers (CNFs) and lipopeptide, is shown graphically in Figure 9.12. Clear diﬀerences were apparent in TNF-a production among CNFs, depending on the treatment conditions utilized (whether acid- or CHAPS-treated) [8]. CNFs, which are derivatives of CNTs, have a crystal structure in which a graphene sheet forms a circular cone, with several cones being stacked towards the needle-axis. C60 , CNFs and CNTs [9–17] are bioinert materials that show similar behavior to Ti or TiO2 in terms of both cell functional tests and size-dependence, leading to inﬂammation in the tissues in vivo [18–22]. The diacylated lipopeptide FSL-1 was used as a positive control, as it is known to induce macrophages to produce TNF-a. The concentration of C60 and CNFs was 0.1 to 10 mg mL 1 , and that of the lipopeptide 10 ng mL 1 . Although the dose of lipopeptide was much lower, the level of activation of THP-1 cells was far beyond that of C60 and CNFs. This implied that the stimulatory level by bioactive and bioinert materials was far lower (perhaps 1/1000th to 1/10 000th) compared to the toxicity level of microbial lipopeptide, which is a type of bacterial endotoxin. It should be recognized, however, that micro/nanoparticles still induce phagocytosis and inﬂammation.

Fig. 9.12 Comparison of TNF-a release for carbon nanoﬁbers and endotoxin (lipopeptide) [8].

9.3 Results

Fig. 9.13 Dependence of TNF-a release from neutrophils on particle size down to nanometer size.

9.3.6 Nanotoxicology 9.3.6.1 Size-Dependent Stimulus Down to Nanometer Size The dependence on particle size (down to nanometer) of TNF-a release from neutrophils is shown graphically in Figure 9.13. Below a particle size of 0.5 mm, metallic Ti is diﬃcult to treat as it is easily oxidized in air, and consequently TiO2 was used to create a smaller particle size. Metallic Ti and TiO2 showed similar dependencies, in both qualitative and quantitative manner, which reﬂected the non-speciﬁc properties of a stimulus induced by a particle size eﬀect. The stimulus – in this case the amount of TNF-a released – was pronounced below a particle size of 10 mm, exhibited a maximum from a size of about 1 mm down to 0.5 mm, and decreased with sizes < 0.2 mm. Although the level of TNF-a release was low at the 50 nm particle size, this might be preferred for the biological use of nanoparticles, as the stimulus was decreased. In this situation, however, the biophylactic system would no longer function adequately to combat any invasion by nanoparticles. 9.3.6.2 Internal Diﬀusion of Nanoparticles Nanoparticles (especially those <50 nm in size) may invade the internal body through either the respiratory or digestive systems. Figure 9.14 illustrates internal whole-body Ti mapping of rats, using X-ray analytical microscopy (XSAM), to show the distribution of 30-nm TiO2 particles inside the body, following compulsory inhalational exposure. Condensation of the particles occurred from the respiratory system to the urinary bladder by diﬀusion in the whole body through the cardiovascular system, following direct uptake into blood vessels from the lungs. It is likely that the existence of nanoparticles is not recognized by the body’s defense system, and consequently an understanding of their behavior within the body is essential if they are to be used as eﬀective drug delivery systems. Thus, it is important that the internal dynamics of particles be determined if nanoparticles are to be used in this way.

139

140

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

Fig. 9.14 X-ray scanning analytical microscopy Ti mapping of internal distribution of 30 nm TiO2 particles after compulsory exposure test.

9.3.6.3 Toxicity-Enhancing Eﬀects of Biostimulatory Materials by Nanosizing Following a one-year implantation of 0.5 mm Ni particles, tumors were seen to occur in the subcutaneous tissues of rats (Fig. 9.15). Although Ni is known to be toxic when in macroscopic form (see Fig. 9.1), its toxicity is enhanced remarkably when in a ﬁne-particle state. Such a speciﬁc surface eﬀect increases reciprocally with particle size, and may lead to the enhancement of adverse chemical and toxic eﬀects.

Fig. 9.15 Tumor induced after one-year implantation of 0.5 mm Ni particles.

9.4 Discussion 9.4.1 Particle Size-Dependence of Cytotoxicity

The results of these biochemical functional analyses and microscopic morphological observations showed cell survival to have decreased (see Fig. 9.3) and, corre-

9.4 Discussion

spondingly, the activity of LDH (an indicator of cell disruption) to have been increased, as the Ti particle size was reduced from 150 mm to 0.5 mm. Both, superoxide anion and cytokine (IL-1b, TNF-a) production also tended to increase as particle sizes decreased, notably for those of 0.5 and 3 mm. Whilst superoxide anions are released from intracellular organs and cell membranes when stimulated, IL1b is released when neutrophils are stimulated by foreign bodies, leading to the inﬂammation reaction cascade. TNF-a release is closely related to phagocytosis. Both, OM and SEM observations (see Fig. 9.2) indicated that only Ti particles of 0.5 and 3 mm were phagocytosed by neutrophils (which themselves were 5–10 mm in diameter), whereas phagocytosis proved to be diﬃcult for the 10-, 50-, 150-mm Ti particles. The results of cell survival (Fig. 9.3), LDH activity increase, superoxide production and release of IL-1b (Fig. 9.4) and TNF-a (Fig. 9.5) were shown to be particle size-dependent, with eﬀects becoming more apparent as the particle size was reduced. This was especially pronounced when the size was less than that of the cell. ICP elemental analysis conﬁrmed that dissolution from Ti particles was below the limits of detection, and eﬀectively negligible; thus, the biofunctional changes demonstrated in the present studies were not caused by any chemical effects of the Ti ions but rather were due to physical size eﬀects of the Ti particles. 9.4.2 Particle Size-Dependence in Soft Tissues

The in-vivo implantation of a variety of Ti particle sizes in the soft tissue of rats showed that particles > 100 mm were surrounded by a ﬁbrous connective tissue layer, this being the normal reaction for biocompatible materials such as a Ti implant. Particles < 100 mm in size tended to cause inﬂammation, but this was even more pronounced for smaller particles. Particles of <10 mm – about cell size – caused numerous inﬂammatory cells to appear around the particle agglomerates and for macrophage-based phagocytosis to be induced. Long-term inﬂammatory reactions were stimulated by 0.5- and 3-mm Ti particles. As dissolution from particles was undetected in vitro, the main eﬀects caused by Ti in rats appeared to be caused by the particles themselves rather than by ions. These results in vivo can be explained by assuming an increased release of superoxide, TNF-a and IL-1b, and subsequent cytotoxic stimulation in the soft tissue around the inserted Ti particles. 9.4.3 Comparison of Ti, Fe, and Ni Particles

In order to verify whether the cytotoxicity of ﬁne particles originated from dissolved ions or from the particles per se, ICP elemental analyses were carried with HBSS containing Ti particles. Dissolution from Ti particles was below the limits of detection and negligible, which inferred that Ti was chemically stable and insoluble, whereas Fe was easily dissolved [23–26]. By contrast, Ni was seen to dis-

141

142

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles

solve to some extent and hence to be highly toxic. Data acquired with regards to TNF-a release (see Fig. 9.7) indicated that Ti and Fe – both of which are biocompatible or are minimally cytotoxic when in macroscopic form (see Fig. 9.1) – demonstrated quantitatively similar size dependencies, despite Ti being insoluble and Fe soluble in HBSS. This lack of cytotoxicity by Ti and Fe, coupled with a lack of Ti dissolution, strongly suggested that any cytotoxicity related to the Ti and Fe particles was due to their physical size and not to the presence of dissolved ions. Ni showed a similar size-dependency, although the values of cell survival rate, superoxide anion and cytokine release were lowered. Signiﬁcantly lower cell survival rates and higher LDH activities suggested that cell destruction occurred with a higher probability following exposure to Ni particles, and this might surpass the stimulatory eﬀect of Ni and lower superoxide anion and IL-1b production. Previous results obtained with macroscopic Ni inserted into the soft tissue showed that necrosis and inﬂammation occurred which was dependent upon the distance from the Ni surface, though is also the function of concentration of dissolved Ni ions, as analyzed by XSAM [23–26]. The extreme toxicity of Ni was shown to originate from its dissolved ions. 9.4.4 The Eﬀect of Micro-/Nanosizing on Biological Reactions

In general, the biocompatibility of a material depends on the chemical solubility of its ions, followed by its absorption into the cells and tissues. Corrosion resistance [27] is, therefore, a prerequisite for biomaterials, and this holds true for most materials of macroscopic size. The eﬀects of micro-/nanosizing are often explained in terms of their speciﬁc surface area, which increases in reciprocally proportional manner to particle size. The enhancement of chemical reactivity based on ionic dissolution can be more easily understood by this eﬀect, an example of which is the accelerated toxicity of Ni (see Fig. 9.15), with tumor being generated after the long-term implantation of 0.5-mm particles, compared to necrosis in the short term. Other eﬀects of changes in material properties when nanosizing include the increasing eﬀect of the ratio of surface constituent atoms, and the quantum eﬀect. All of these eﬀects may be classiﬁed as the eﬀects of the material itself. The eﬀects of micro/nanosizing also depend upon the size relationship between cells/tissues and particles. Physical particle size and shape eﬀects occur non-speciﬁcally in many materials, and are unaﬀected by surface area eﬀects. Such eﬀects also seem to occur more easily in bioactive and bioinert materials, which are minimally inﬂuenced by chemical dissolution and show good biocompatibility on the macroscopic scale (as is the case for Ti and Fe; see Fig. 9.1). Thus, toxicity levels may be 10 3 - to 10 4 -fold lower than with endotoxins. Although this may not cause serious problems in the short term and in small quantities, it still causes inﬂammation in vivo, despite in-vitro toxicity being low. This situation would be serious when numbers of ﬁne particles increase in quantity with time, and remain present over the long term; an example would be an artiﬁcial joint,

References

the lifetime of which is limited by the inﬂammation induced by abraded particles. A similar eﬀect is seen with asbestos, where the long-term phagocytosis of acicular particles leads to chronic inﬂammation, together with superoxide production, defect formation in DNA, and carcinogenesis. 9.4.5 Terminology on ‘‘Nanotoxicology’’

The term ‘‘Nanotoxicology’’ does not necessarily accurately reﬂect the eﬀects caused by the particles described above. In biology, the size of the fundamental unit – the cell – is approximately 10 mm, and particle eﬀects are pronounced in the range of 3 mm to 0.5 mm. However, particles < 50 nm in size may invade the body directly, via either the respiratory or digestive system, without stimulating the biophylactic system. Changes in function occurring over this size range are seen with hydroxyapatite, which is not only osteoconductive but also nonabsorbable on a macroscopic scale, and is thus suited to biostructural implantation. Consequently, as apatite nanoparticles can form composites with collagen, apatite may function as a bone substitute, inducing absorption of the composite by osteoclasts and new bone formation by osteoblasts [28, 29]. Thus, particle effects on biological reactions may indeed be termed ‘‘micro/nanotoxicology’’.

Acknowledgments

These studies were conducted with ﬁnancial support from Health and Labour Sciences Research Grants of the Research on Chemical Substance Assessment from the Ministry of Health, Labour and Welfare of Japan (H18-ChemistryGeneral-006).

References 1 H. Matsuno, A. Yokoyama, F. Watari,

2

3

4

5

M. Uo, T. Kawasaki, Biomaterials 2001, 22, 1253–1262. M. Uo, K. Asakura, A. Yokoyama, K. Tamura, Y. Totsuka, T. Akasaka, F. Watari, Chem. Lett. 2005, 34, 776– 777. Y. Tamura, A. Yokoyama, F. Watari, M. Uo, T. Kawasaki, Mater. Trans. 2002, 43, 3043–3051. Y. Tamura, A. Yokoyama, F. Watari, T. Kawasaki, Dent. Mater. J. 2002, 21, 355–372. F. Watari, A. Yokoyama, M. Omori, T. Hirai, H. Kondo, M. Uo, T. Kawasaki,

6

7

8

9

Composites Sci. Tech. 2004, 64, 893– 908. K. Tamura, N. Takashi, R. Kumazawa, F. Watari, Y. Totsuka, Mater. Trans. 2002, 43, 3052–3057. R. Kumazawa, F. Watari, N. Takashi, Y. Tanimura, M. Uo, Y. Totsuka, Biomaterials 2002, 23, 3757–3764. K. Kiura, Y. Sato, M. Yasuda, B. Fugetsu, F. Watari, K. Tohji, K. Shibata, J. Biomed. Nanotech. 2005, 1, 359–364. I.D. Rosca, F. Watari, M. Uo, T. Akasaka, Carbon 2005, 43, 3124– 3131.

143

144

9 Biochemical and Pathological Responses of Cells and Tissue to Micro- and Nanoparticles 10 M. Ushiro, K. Uno, T. Fujikawa, Y.

11

12

13

14 15

16

17

18

19

Sato, K. Tohji, F. Watari, W.J. Chun, Y. Koike, K. Asakura, Phys. Rev. B 2006, 73, 144103/1–11. K. Asakura, W.J. Chun, K. Tohji, Y. Sato, F. Watari, Chem. Lett. 2005, 34, 382–383. B. Fugetsu, S. Satoh, T. Shiba, T. Mizutani, Y. Lin, N. Terui, Y. Nodasaka, K. Sasa, K. Shimizu, T. Akasaka, M. Shindoh, K. Shibata, A. Yokoyama, M. Mori, K. Tanaka, Y. Sato, K. Tohji, S. Tanaka, N. Nishi, F. Watari, Environ. Sci. Technol. 2004, 38, 6890–6896. B. Fugetsu, S. Satoh, A. Iles, K. Tanaka, N. Nishi, F. Watari, The Analyst (London) 2004, 129, 565– 566. T. Akasaka, F. Watari, Chem. Lett. 2005, 34, 826–827. T. Akasaka, F. Watari, Y. Sato, K. Tohji, Mater. Sci. Eng. C 2005, 26, 675–678. N. Aoki, A. Yokoyama, Y. Nodasaka, T. Akasaka, M. Uo, Y. Sato, K. Tohji, F. Watari, Chem. Lett. 2006, 35, 508– 509. N. Aoki, A. Yokoyama, Y. Nodasaka, T. Akasaka, M. Uo, Y. Sato, K. Tohji, F. Watari, J. Biomed. Nanotechnol. 2005, 1, 402–405. A. Yokoyama, Y. Sato, Y. Nodasaka, S. Yamamoto, T. Kawasaki, M. Shindoh, T. Kohgo, T. Akasaka, M. Uo, F. Watari, K. Tohji, Nano Lett. 2005, 5, 157–161. Y. Sato, A. Yokoyama, K. Shibata, Y. Akimoto, S. Ogino, Y. Nodasaka,

20

21

22

23

24

25

26 27 28

29

T. Kohgo, K. Tamura, T. Akasaka, M. Uo, K. Motomiya, B. Jeyadevan, M. Ishiguro, R. Hatakeyama, F. Watari, K. Tohj, Mol. BioSystems 2005, 1, 176–182. Y. Sato, T. Shibata, H. Kataoka, S. Ogino, B. Fugetsu, A. Yokoyama, K. Tamura, T. Akasaka, M. Uo, K. Motomiya, B. Jeyadevan, R. Hatakeyama, F. Watari, K. Tohji, Mol. BioSystems 2005, 1, 142–145. W. Wang, M. Omori, F. Watari, A. Yokoyama, Dent. Mater. J. 2005, 24, 478–486. M. Uo, K. Tamura, Y. Sato, A. Yokoyama, F. Watari, Y. Totsuka, K. Tohji, Small 2005, 1, 816–819. M. Uo, F. Watari, A. Yokoyama, H. Matsuno, T. Kawasaki, Biomaterials 2001, 21, 677–685. M. Uo, F. Watari, A. Yokoyama, H. Matsuno, T. Kawasaki, Biomaterials 2001, 22, 1787–1794. M. Uo, F. Watari, A. Yokoyama, H. Matsuno, T. Kawasaki, Biomaterials 1999, 20, 747–755. M. Uo, K. Asakura, T. Kohgo, F. Watari, Chem. Lett. 2006, 35, 66–67. F. Watari, J. Electron Microsc. 2005, 54, 299–308. A. Yokoyama, M. Gelinsky, T. Kawasaki, T. Kohgo, U. Konig, W. Pompe, F. Watari, J. Biomed. Mater. Res. B: Appl. Biomater. 2005, 75B, 464–472. S. Liao, W. Wang, M. Uo, S. Ohkawa, T. Akasaka, K. Tamura, F. Cui, F. Watari, Biomaterials 2005, 26, 7564– 7571.

145

10 Tissue Engineering of Bone Hans-Peter Wiesmann, Beate Lu¨ttenberg, and Ulrich Meyer

Abstract

Bone tissue engineering is one of the most developing ﬁelds of research, especially with regards to clinical applications. It will oﬀer the opportunity to enable the regeneration of bone damaged by disease or trauma and perhaps, in some cases, the ‘‘creation’’ of new bone. In future years bone loss will become more important, especially with the increasing age of patients. One main precondition for the development of successful bone tissue engineering is an understanding of the details of the biomineralization process. In this chapter, a brief overview is provided of the biomineralization steps that take place in extracellular matrix formation in vivo and in vitro. A state-of-the-art summary is also provided of the existing methods in bone tissue engineering, including scaﬀold design, the diﬀerent cell types in use, and methods of stimulation. In future, improved combinations of scaﬀolds, cells and cell-culturing methods, together with new technical possibilities, will help us to achieve bone tissue engineering processes that can provide individual bone substitutes. Keywords: bone tissue engineering, biomineralization, extracellular matrix, osteoblasts, cell stimulation, bioreactors, scaﬀold design.

10.1 Tissue Engineering: What Does it Mean?

Tissue engineering is not a well-deﬁned term, rather, it is more a technical concept which combines various disciplines and ﬁelds of research. Two more-or-less accepted deﬁnitions of tissue engineering are in common use: US National Science Foundation: The application of the principles and methods of engineering and the life sciences towards the fundamental understanding of structure/ function relationship in normal and pathological mammalian Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

146

10 Tissue Engineering of Bone

tissues and the development of biological substitutes to restore, maintain or improve function. European Commission (Scientiﬁc Committee on Medical Products and Medical Devices; SCMPMD): Tissue engineering is the regeneration of biological tissue through the use of cells, with the aid of supporting structures and/or biomolecules. The US National Science Foundation deﬁnition of tissue engineering includes in vivo engineering techniques as well as the use of cells from diﬀerent sources (autogenic, allogenic, xenogenic), whereas the European Commission deﬁnition narrows it to the extracorporeal (in vitro) aspect of this technique. In this chapter, bone tissue engineering means the construction, repair or replacement of damaged or missing bone in humans and animals. The aim is to focus on the biological and biophysical principles in extracorporeal bone tissue engineering which are implied in the European Commission deﬁnition (see above). Tissue engineering is a multidisciplinary subject that brings together a variety of ﬁelds, including materials sciences, biological sciences and clinical disciplines [1]. Although the clinical aspects are currently in their early stages, commercial bone tissue engineering in clinical practice does already exist in certain cases [2], with a few companies engaged in this ﬁeld. However, many of their products do not comply with the deﬁnition of tissue engineering used in this chapter, in that they are simply combinations of cells and biomaterials [3]. One of the ﬁrst investigators to describe early attempts in the ﬁeld of bone regeneration and replacement was Lexer who, in 1908, used freshly amputated and cadaver allografts for joint reconstruction [4]. During the mid-1960s, Urist demonstrated that matrix implantation was followed immediately by autoinduction involving the ingrowth of cells from the host bed [5]. Since then, the induction of bone cell diﬀerentiation has been identiﬁed, with the development of new strategies for bone tissue engineering, and polypeptides, demineralized powder or biophysical stimuli having been used to stimulate mesenchymal tissue. More recently, cell transplantation strategies have evolved which attempted to deliver living cells onto alloplastic implants in order to produce bone. This new approach combines the biological properties of living cells with the physical properties of specially designed materials [6]. For this, one must take into account the importance of any environmental support to ensure survival of the grafted cells after implantation. Reconstructive surgery uses biomaterials, autografts or allografts, although restrictions such as donor site morbidity or immunologic barriers for allografts and the risk of transmitting infectious diseases remain. Bone tissue engineering should oﬀer the possibility of helping to regenerate damaged tissue (resulting from disease or trauma), replacing malfunctioning tissue, or even creating new ‘‘bone’’. The critical feature of this multistep process of new bone formation is that of biomineralization.

10.2 Components of Bone Tissue Engineering

10.2 Components of Bone Tissue Engineering

The three main parties involved in extracorporeal tissue engineering are bone cells, extracellular scaﬀolds, and (sometimes) growth factors or other stimulatory strategies. The cells used for this purpose must have the ability to be proliferated or even diﬀerentiated in vitro. Bone devices can contain allogeneic, xenogeneic or immunogenic autologous cells and, when assembled with an appropriate scaﬀold and the eventual application of growth factors, the result is a three-dimensional (3-D) complex tissue that is ready for implantation. A principal pathway of cell mediated tissue engineering is shown in Figure 10.1. To date, it has proved impossible to construct a large bone-mineral complex outside the living body.

Fig. 10.1 The principal pathways of bone tissue engineering. BPS ¼ biophysical stimulation.

10.2.1 Osteoblasts

Osteoblasts, which are cuboidal cells that secrete the bone matrix, possess four maturation stages: preosteoblasts (precursor cells); lining cells; osteoblasts; and osteocytes. Osteoblasts originate from mesenchymal progenitor cells and undergo a gradual process of diﬀerentiation which is accompanied by a temporal sequence

147

148

10 Tissue Engineering of Bone

of the expression of genes encoding typical cell-dependent phenotype marker proteins. Fortunately, it is possible to maintain and proliferate osteoblasts or osteoblastlike cells in vitro, and such cells can be derived from several anatomical sites, using a variety of explant procedures. The various osteoblast-like cell types are stable extracorporeally over long periods of time. In current clinical practice, autologous osteoblast-like cells are the most common cell type, because they carry no legal problems, nor any major risk of immune rejection. However, one possible limitation with osteoblast-like cells with regards to tissue engineering might be that they are less eﬀective at rebuilding damaged bone within a reasonable time. 10.2.2 Bone Marrow Cells

The bone marrow of a healthy human adult contains approximately 0.001% osteoprogenitor cells among the fraction of nucleated cells. The successful in vitro use of bone marrow explants depends very heavily on the number of progenitor cells, although unfortunately the population of these cells decreases not only with age but also especially with disease – the precise situation when tissue engineering is normally required. In addition to these clinical and applicational problems, it is known that human bone marrow osteoprogenitors can be isolated and enriched by using selective markers [7, 8]. Several preclinical investigations and clinical studies have shown that bone marrow has the potential to lead to bone regeneration [9–11], but in order to enhance such promise it will be essential to develop new techniques for selecting, expanding and controlling the progenitor cell fraction. 10.2.3 Marrow-Derived Stem Cells

The isolation of mesenchymal stem cells (MSC) is commonly based on density gradient centrifugation and cell culture techniques following bone marrow aspiration. It has been shown that MSC can be isolated and cultured without undergoing diﬀerentiation and without losing their osteogenic, chondrogenic or adipogenic potential over a period of time [12, 13]. There are two critical points concerning the use of MSC. The ﬁrst point is the technical challenge that, until now, no speciﬁc marker has been determined to ensure reproducible cell isolation. The second and main problem with in vitro cultured and diﬀerentiated MSC is that they seem unable to mineralize in a bone-like manner [14, 15], and this drastically limits their applications, notably in relation to extracorporeal bone tissue engineering. Apart from this, the potential of MSC is indisputable, and with new technical developments combined with basic research these cells may represent the most promising approach to extracorporeal tissue engineering.

10.2 Components of Bone Tissue Engineering

10.2.4 Vascular Cells

Endothelial progenitor cells are good candidates for delivery together with bone cells inside the implant. They are ubiquitous within the body and can promote vascularization at sites where new vessels are needed after implantation. Muyarama et al. [16] hypothesized that endothelial progenitor cells might not only participate directly in neovascularization as a substrate for new endothelial cells, but can also be capable of delivering growth factors and/or cytokines that stimulate the vascularization process in general. Further investigations with endothelial progenitor cells might help to improve the extracorporeal formation of mature bone constructs. Bearing in mind that vasculogenesis is an essential primary step of bone mineralization, the importance of endothelial cells in bone tissue engineering becomes apparent [17]. 10.2.5 Scaﬀold Design and Cell Compatibility

Bone scaﬀolds used in tissue engineering should resemble the morphology of the bone at the site of implantation, and are also intended to degrade slowly as they are replaced piece by piece with new bone. In order to achieve a good seeding of cells, surface parameters such as wettability, conductibility, topography, element composition and release are critical. These points are important for cells assembled with the scaﬀold in vitro as well as for those cells surrounding the implantation bed in vivo. The pore size may also be shown as crucial for the ingrowth of osteoblasts and small vessels, since it seems that small pores (ca. 150 mm diameter) hinder bone formation and vascular invasion [2, 18, 19]. With regards to the clinical applications of scaﬀolds, the main point is that their degradation should be completed after an adequate period of time, though other points to be taken into account include a disadvantageous pH-lowering when synthetic polymers are degraded, an ability to be surgically ﬁxed, and that they themselves should be neither toxic nor immunogenic. Initially, implanted scaﬀolds will inﬂuence the bony implantation bed, although the body may also react against the scaﬀolds in various cellular and non-cellular ways. In order to evaluate such possible reactions, it is necessary to guarantee that the scaﬀolds have reproducible properties, and consequently computer-based fabrication technologies (e.g., solid free form; SFF [20]) are increasingly replacing conventional procedures. To date, four diﬀerent types of material are in common use for the fabrication of scaﬀolds: organic materials of natural origin (e.g., collagen or ﬁbrin); inorganic materials of natural origin (e.g., coralline hydroxyapatite); synthetic organic materials (e.g., polymers); and synthetic inorganic materials (e.g., glass ceramics).

149

150

10 Tissue Engineering of Bone

10.2.6 Bioreactors

Culture dishes and ﬂasks, as the most simple forms of bioreactors, do not provide a convenient solution for the fabrication of 3-D bone tissue in vitro. In order to avoid the limited nutrient supply that occurs within a culture dish or ﬂask, spinner ﬂasks can be used, as these ensure a better and continuous nutrient supply, even when cells are seeded into 3-D scaﬀolds. Another type of bioreactor is the ‘‘rotating wall vessel’’, although recent investigations have shown that bone cell/ polymer constructs grown in this type of reactor did not demonstrate any convenient cell diﬀerentiation towards the osteoblastic phenotype. Another possible approach, ‘‘ﬂow chambers’’, combines nutrient supply with physiological strains. For the manufacture of scaﬀolds and cells in vitro, closed bioreactor systems offer major advantages over open systems, mainly because closed bioreactors can modulate each of the diﬀerent parameters (e.g., temperature, culture medium, ﬂuid ﬂow); this in turn inﬂuences the cell/scaﬀold product in a controlled and reproducible manner. 10.2.7 In-Vitro Cell Stimulation 10.2.7.1 Biophysical Stimulation Within their natural environment, bones are continuously aﬀected by mechanical forces caused by muscular contractions and/or body movements that, perhaps unsurprisingly, inﬂuence osteoblast proliferation, orientation, gene activity, and other features of cell activity. Electrical ﬁelds are also thought to inﬂuence osteoblast physiology; for example, it has been shown that the long-term electrical stimulation of osteoblasts resulted in an altered gene expression pattern to enhance the synthesis of extracellular matrix (ECM) compounds and biomineral formation [21, 22]. In vivo, bone defect regeneration can also be enhanced by pulsed electrical ﬁelds (Fig. 10.2). Thus, for bone tissue engineering, biophysical stimu-

Fig. 10.2 Histology of critical size defect in rabbit mandibular bone after 3 weeks. Left: without stimulation; right: with electrical stimulation. Staining was carried out with a combination of Alizarin Red-S and toluidine blue. Scale bars ¼ 1 mm.

10.3 Bone Biomineralization in Tissue Engineering Ex Vivo and In Vivo

lation might represent a valuable means of producing a mature product for subsequent implantation [22]. 10.2.7.2 Biochemical Stimulation Cytokines and other bioactive proteins may serve as useful supplements to enhance bone biomineralization in vitro [23]. Typical candidates among these substances include transforming growth factor b (TGF-b), bone morphogenic proteins (BMP), ﬁbroblast growth factor (FGF), platelet-derived growth factor (PDGF), and insulin-like growth factor (IGF). One crucial aspect with regards to tissue engineering is that these factors require systems which ensure the delivery not only of a single, correct dose, but also – to produce an optimal eﬀect – the delivery of several well-deﬁned doses over a longer period of time. Only such an approach would provide an adequate and enhanced biological response at the site of implantation. Today, several growth factors are available commercially as recombinant proteins, and consequently the future availability of cytokines will undoubtedly increase, thereby facilitating their application. One delivery system that should be mentioned in this context is that of gene therapy, whereby cells are transfected with a DNA-vector containing the DNAsequence of the required cytokine. If the implant carries the encoding DNAvector, and the cells of the environment grow into the implant, then at best the DNA-vector will be taken up and the therapeutic agent synthesized. Among this method utilizing a passive approach to gene therapy, extracorporeal tissue engineering can also use active transfection of the bone cells in vitro, and these are subsequently implanted together with the scaﬀold. Unfortunately, however, such gene therapy raises technical and legal problems that must be considered before it is used.

10.3 Bone Biomineralization in Tissue Engineering Ex Vivo and In Vivo 10.3.1 Principles of ECM Biomineralization

During the multistep process of bone mineralization, Ca-phosphate crystal nucleation and growth are seen as the key steps. In the early stages of crystal formation the unstable crystal nuclei grow to reach a size at which they are stable, whereafter collagen mineralization begins at the surface of the collagen ﬁber. Initially, at the ﬁber surface a larger quantity of mineral substance per unit volume exists in comparison to the interior of the collagen ﬁber [24]. The crystallites in the surface region are shorter and more densely packed (cf. peritubular collagen-free dentine and intertubular dentine) [25]. The complex interactions that occur between collagen and the non-collagenous proteins [26] seem to facilitate primary crystal formation and subsequent crystal growth. Apatite collagen mineralization is induced by the non-collagenous proteins bound in parallel strands along the collagen ﬁber surface [27].

151

152

10 Tissue Engineering of Bone

The primary apatite formations are strands composed of nanometer-sized particles arranged in parallel to the crystallographic bipolar c-axis of apatite [15, 28]. The center-to-center distances between the nanometer-sized apatite particles in the mineral strands may reﬂect the distances between the crystal nucleating sites along the non-collagenous matrix proteins [29]. The strands composed of nanometer-sized particles grow rapidly in length and then fuse with neighbors to form needles. Such growth in thickness causes the lateral bonds within the ﬁbrils to break, whereupon the particles fuse with their neighbors to form plate-like crystallites. This crystal arrangement would produce the primary structure for biomechanical qualities such as high tensile strength. 10.3.2 Principles of Bone Formation

Before starting to utilize bone tissue engineering techniques, it is ﬁrst essential to investigate and understand the structure, function, and formation process of bone itself. Bones constitute the structural framework of the body, and contain a variety of diﬀerent cell types, including vascular cells, marrow cells, preosteoblasts, osteocytes, chondroblasts, and osteoclasts. The other major component of bone is the organic matrix, which consists of 95% collagen, the remaining 5% being proteoglycans and other non-collagenous proteins. Bone is a highly dynamic tissue which has distinct processes of assembly, degradation, and remodeling. Although all of the above-mentioned cells are necessary for the construction of ‘‘real bone’’, limited cell sources may suﬃce for the in-vitro synthesis of ‘‘bone-like’’ constructs. Two distinct developmental processes have been identiﬁed for the ‘‘natural’’ formation of bone. The ﬁrst process – endochondral ossiﬁcation – yields the so-called ‘‘long bones’’ that comprise the facial bones, vertebrae, apendicular skeleton and the lateral clavicles. The cranium and medial clavicles, on the other hand, consist of ﬂat bone that is the product of intramembranous ossiﬁcation. Whilst both types of ossiﬁcation begin with an initial condensation of mesenchyme, the endochondral process involves an additional regulation step mediated by a cartilaginous template, though both processes result in the eventual formation of calciﬁed bone. An additional macroscopic distinction concerning bone structure can be made with regards to morphology and function: Cortical (or compact) bone comprises approximately 80% of the mature skeletal bone; this has mechanical and/or protective functions. Cancellous (or trabecular) bone comprises approximately 20% of the skeletal bone, and is more involved in metabolic functions. Characteristic of cortical bone are the densely packed collagen ﬁbrils which form concentric lamellae. In contrast, cancellous bone has a more loosely orga-

10.4 Clinical Demands

nized porous structure, comprised of an array of plates and rods of bone tissue which form an open cell foam. In cortical bone the majority of cells are osteocytes and mature osteoblasts, surrounded by a mineralized matrix. The osteoblasts, which originate from MSC, are responsible for building up the organic matrix (osteoid) but, after fulﬁlling their function, they mainly evolve to osteocytes that are located in small lacunae within the mineralized bone; others become lining cells on the surface of the bone. Communication between the osteocytes is achieved via a gap junction-network interconnecting the cells, and hence it is assumed that the osteocytes play a very important role in adaptational processes to external stimuli. The trabeculae of the cancellous bone are covered with osteoblasts and bone-lining cells. The latter are in an inactive state, and the osteoblasts actively build up bone tissue by secreting components of the extracellular organic matrix. 10.3.3 Particular Features of Extracorporeal Biomineralization

For extracorporeal bone tissue engineering it is essential to diﬀerentiate between the precipitation of calcium phosphate and the formation of bone-like apatite structures. For this, the mineralization process and resulting crystal structures must be studied in a detailed manner [22]. Exact bone-like mineral formation on extracorporeal surfaces has not yet been clearly demonstrated, and this remains a long-standing problem with regards to bone tissue engineering in vitro. De novo biomineralization of the matrix in bone and in cell culture is connected with matrix vesicles [30, 31], and it has been shown that such membrane-invested particles serve as the initial sites of calciﬁcation. Ex vivo, the matrix secreted by osteoblasts undergoes mineralization independently of the bone formation process. Osteoblast cell cultures only show osteoid-like matrices when they are grown in a multilayered structure, and the mineralization nodules are formed after several weeks [32]. However, matrix vesicle mineralization is globular, but not structured, over longer distances, and therefore in vivo collagen biomineralization is necessary to adapt the bony tissue to the mechanical and hydrodynamic requirements. The extra- and intra-collagenous mineralization process is highly complex, and is accompanied by non-collagenous proteins both outside and on the collagen surface. Following nucleation and mineralization, a well-formed and structured composite consisting mainly of collagen and hydroxyapatite is generated by the bone cells. Ex-vivo, mechanical and hydrodynamic impulses are missing and hence collagen mineralization is frequently not clear [22].

10.4 Clinical Demands

Clinical applications of extracorporeal bone tissue engineering require model systems that can exactly mimic the clinical situation in the patient. Animal models

153

154

10 Tissue Engineering of Bone

(including those in mature animals) allow analyses to be made of the implanted material in a functional environment that is comparable with the human situation. However, since clinical studies in humans are restricted to investigations of outcome parameters, appropriate animal studies should also include similar outcome measures. Before clinical tissue engineering can be planned, the location and the anatomic nature of the bone defect must be investigated. Additional aspects should also be considered, for example if there are possible advantages of cellular bone constructs over non-living materials, and if clinically proven applied models for bone tissue regeneration already exist. Biomechanical aspects of the skeletal implantation site and immunological reactions towards the cell/scaﬀold construct should also be taken into account. Although extracorporeal new bone tissue engineering strategies have been introduced clinically, especially in craniofacial surgery [33], they often result from studies by orthopedic surgeons seeking to reﬁne techniques for bone healing in fracture repair and defect reconstruction [34–38].

10.5 Future Aspects

Currently, techniques of extracorporeal bone tissue engineering are showing great promise, though new developments in the ﬁelds of customized tissue reconstruction, nutritional support of the transplant, stem cell research and genetic engineering will undoubtedly improve these applications during the following years. In the area of customized tissue reconstruction, both computed tomography (CT) and magnetic resonance imaging (MRI) provide valuable detailed anatomic information in designing the scaﬀold. The production and accurate insertion of individualized scaﬀolds are supported by rapid prototyping (RP) and computeraided design/computer-aided manufacturing (CAD/CAM). These results in the emerging ﬁeld of computer-aided tissue engineering (CATE) [39] which, one day, will surely allow the individualized fabrication of bone substitutes in common clinical practice. In vascularization and nerve regeneration, neovascularization at the site of implantation is regarded as a limiting factor for the survival of tissue-engineered constructs. Although the in-vitro creation of an exact vascular structure to implant may be unrealizable in the near future, it might be possible either to inﬂuence the resident cells to assist in this process (e.g., with growth factors such as VEGF [40]) or even to generate scaﬀolds with properties similar to those of the native vessels [41]. The repair of long peripheral nerve gaps also remains a clinical challenge which, at present, cannot be solved. For this reason, methods to develop synthetic alternatives to autografts are of special interest [42]. With regards to genetic engineering, gene therapy involves the introduction of encoding genes into speciﬁc cells such that expression of the desired protein can

References

be controlled in vivo. Alternatives to gene therapy, such as conventional proteinbased therapies, have certain disadvantages, most notably the limited amounts of protein produced at the target site, the need for extensive recombinant expression technologies to obtain suﬃcient protein, and an inability to regulate protein levels in vivo. Successful gene therapy implicates eﬀective transfection, followed by controlled transcription and translation of the recombinant gene. Genetic engineering strategies can be distinguished into ex vivo-transfection of cells and local or systemic delivery systems although, as the latter approach has major limitations, local gene therapy seems to show more promise for bone healing. Transfection can be achieved either in vivo or in vitro: in vivo methods include introduction of the DNA-vector directly into the host tissue or via a DNA-vectorcontaining matrix (gene-activated matrix, GAM), whereas transfection in vitro requires a cell-harvesting step. The main diﬀerence between these two approaches is that the in vitro method generates a higher rate of transfected cells compared to the more-or-less fortuitous cell transfection in vivo. It should be noted that gene therapy implies not only advantages but also severe risks, however, including those of mutagenesis and oncogenesis due to integration of the viral vectors into the host cell genome. Hence, special model systems to evaluate possible disadvantages must be developed before gene therapy can enter clinical practice.

References 1 E. Lavik, R. Langer, Appl. Microbiol. 2 3

4 5 6 7 8

9

10

11

Biotechnol. 2004, 65, 1. U. Ripamonti, J.R. Tasker, Curr. Pharm. Biotechnol. 2000, 1, 47. J.W. Calvert, L.E. Weiss, M.J. Sundine, Clin. Plast. Surg. 2003, 30, 641. E. Lexer, Arch. Klin. Chir. 1908, 86, 939. M.R. Urist, Science 1965, 150, 893. J.M. Orban, K.G. Marra, J.O. Hollinger, Tissue Eng. 2002, 8, 529. P.J. Simmons, B. Torok-Storb, Blood 1991, 78, 55. K. Stewart, S. Walsh, J. Screen, C.M. Jeﬀeriss, J. Chainey, G.R. Jordan, J.N. Beresford, J. Bone Miner. Res. 1999, 14, 1345. I.T. Jackson, L.R. Scheker, J.G. Vandervord, J.G. McLennan, Br. J. Plast. Surg. 1981, 34, 422. H. Ohgushi, V.M. Goldberg, A.I. Caplan, Acta Orthop. Scand. 1989, 60, 334. H. Ohgushi, V.M. Goldberg, A.I. Caplan, J. Orthop. Res. 1989, 7, 568.

12 S.P. Bruder, N. Jaiswal, S.E. Haynes-

worth, J. Cell. Biochem. 1997, 64, 278. 13 M.F. Pittenger, A.M. Mackay, S.C.

14

15

16

17

18

19

Beck, R.K. Jaiswal, R. Douglas, J.D. Mosca, M.A. Moorman, D.W. Simonetti, S. Craig, D.R. Marshak, Science 1999, 284, 143. N. Jaiswal, S.E. Haynesworth, A.I. Caplan, S.P. Bruder, J. Cell. Biochem. 1997, 64, 295. U. Plate, S. Arnold, U. Stratmann, H.P. Wiesmann, H.J. Ho¨hling, Connect. Tissue Res. 1998, 38, 149. T. Murayama, O.M. Tepper, M. Silver, H. Ma, D.W. Losordo, J.M. Isner, T. Asahara, C. Kalka, Exp. Hematol. 2002, 30, 967. J. Kleinheinz, H.P. Wiesmann, U. Stratmann, U. Joos, Mund Kiefer Gesichtschir. 2002, 6, 175. J.H. Kuhne, R. Bartl, B. Frisch, C. Hammer, V. Jansson, M. Zimmer, Acta Orthop. Scand. 1994, 65, 246. O. Gauthier, J.M. Bouler, E. Aguado, P. Pilet, G. Daculsi, Biomaterials 1998, 19, 133.

155

156

10 Tissue Engineering of Bone 20 P.S. D’Urso, W.J. Earwaker, T.M.

21

22

23 24

25

26

27

28 29

Barker, M.J. Redmond, R.G. Thompson, D.J. Eﬀeney, F.H. Tomlinson, Br. J. Plast. Surg. 2000, 53, 200. H. Wiesmann, M. Hartig, U. Stratmann, U. Meyer, U. Joos, Biochim. Biophys. Acta 2001, 1538, 28. H.P. Wiesmann, U. Joos, U. Meyer, Int. J. Oral Maxillofac. Surg. 2004, 33, 523. H. Schliephake, Int. J. Oral Maxillofac. Surg. 2002, 31, 469. H.J. Ho¨hling, R.H. Barckhaus, E.R. Krefting, J. Althoﬀ, P. Quint, in: Ultrastructure of Skeletal Tissues. Academic Press, 1990, p. 41. H.J. Ho¨hling, in: Handbook of Microscopic Anatomy, Volume V/6. Springer-Verlag, Berlin, 1989, p. 474. U. Meyer, T. Meyer, J. Vosshans, U. Joos, J. Cranio-Maxillofac. Surg. 1999, 27, 222. H.P. Wiesmann, U. Meyer, U. Plate, H.J. Ho¨hling, Int. Rev. Cytol. 2005, 242, 121. H.J. Ho¨hling, T. Hake, R. Katterbach, Adv. Fluorine Res. 1966, 4, 201. S. Arnold, U. Plate, H.P. Wiesmann, U. Stratmann, H. Kohl, H.J. Ho¨hling, J. Microsc. 2001, 202, 488.

30 H.C. Anderson, J. Cell Biol. 1967, 35,

81. 31 E. Bonucci, J. Ultrastruct. Res. 1967,

20, 33. 32 H.P. Wiesmann, N. Nazer, C. Klatt, T.

33 34 35

36

37

38 39 40

41 42

Szuwart, U. Meyer, J. Oral Maxillofac. Surg. 2003, 61, 1455. E. Alsberg, E.E. Hill, D.J. Mooney, Crit. Rev. Oral Biol. Med. 2001, 12, 64. G.M. Crane, S.L. Ishaug, A.G. Mikos, Nat. Med. 1995, 1, 1322. H.J. Mankin, M.C. Gebhardt, W.W. Tomford, Orthop. Clin. North Am. 1987, 18, 275. C. Perka, O. Schultz, R.S. Spitzer, K. Lindenhayn, G.R. Burmester, M. Sittinger, Biomaterials 2000, 21, 1145. W.C. Puelacher, J.P. Vacanti, N.F. Ferraro, B. Schloo, C.A. Vacanti, Int. J. Oral Maxillofac. Surg. 1996, 25, 223. C.A. Vacanti, L.J. Bonassar, Clin. Orthop. Relat. Res. 1999, 367, 375. W. Sun, P. Lal, Comput. Methods Programs Biomed. 2002, 67, 85. M. Nomi, A. Atala, P.D. Coppi, S. Soker, Mol. Aspects Med. 2002, 23, 463. A. Ratcliﬀe, Matrix Biol. 2000, 19, 353. R.V. Bellamkonda, Biomaterials 2006, 27, 3515.

Part II Teeth

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

159

11 Formation of Teeth Katharina Reichenmiller and Christian Klein

Abstract

During tooth development, several diﬀerent mineralization processes take place. The tooth is a functional system of dental hard and soft tissues, and the formation of teeth underlies genetic and subsequent cellular control. Mineralization arises by the occurrence of matrix vesicles, followed subsequently by extracellular matrix mineralization processes. Cells are able to secrete a calcifying matrix, and their metabolism is responsible for initiating apatite formation at selected, preﬁxed extracellular sites. The balance between cell proliferation, diﬀerentiation and mineralization is regulated by involving the pathways of known and unknown hormones, cytokines and morphogenic factors. Dentin is composed of ca. 70 wt% inorganic compounds, 10 wt% water, and 20 wt% organic matrix, and consists predominantly of collagen ﬁbers. Enamel is composed of 95 wt% inorganic compounds, 4 wt% water, and 1 wt% organic matrix, and is the hardest tissue in the vertebrate body. Enamel contains long, thin crystallites of substituted hydroxyapatite. Cementum, which forms a basic part of the tooth and is also a component of the periodontium, consists of 61 wt% inorganic compounds, 12 wt% water, and 27 wt% organic matrix. Key words: ameloblast, amelogenesis, cementoblast, cementogenesis, collagen, dentinogenesis, odontoblast, pulp, tooth development, vesicle.

11.1 Introduction

Teeth consist of a unique composition of the three dental hard tissues, namely dentin, enamel, and cementum. The three compounds form a complex structure based on hydroxyapatite. Innervated soft tissue, otherwise referred to as the dental pulp, is covered by dentin which, in turn, forms the core of a tooth. In the crown area of a tooth the dentin is covered by enamel, the hardest and most-mineralized Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

160

11 Formation of Teeth

Fig. 11.1 Histological cross-section through a human molar (incomplete root development). Due to decalciﬁcation of the preparation, the highly mineralized enamel is dissolved and is no longer present; the space of the former enamel is shown in yellow. Hematoxylin eosin staining; illustration courtesy of Claus Loest, University of Tubingen/Germany.

of the three tissues. In the root area the outer layer of the dentin is coved by cementum (Fig. 11.1), which anchors the tooth in its socket by inserting desmodontal ﬁbers. Only the combination of these three dental hard tissues ensures that the tooth as a system can function for a time span of about seven decades. Furthermore, this combination makes it possible that teeth can withstand mechanical forces such as chewing pressures up to 800 N, wear (abrasion), chemical or physical interactions such as changes in pH-value or variations in temperature, and biological eﬀects such as bacteria and toxins. Although the role that dental pulp tissue plays is not completely understood, in animal experiments it has been shown that neuronal signals trigger diﬀerentiation of odontoblast-like cells, and in the process hard substance neogenesis is induced [1, 2]. The development of hard tissue can take place in the pulp throughout the life of a tooth, and thus a functioning (albeit limited) cellular reaction may occur to external inﬂuences such as trauma, attrition, abrasion, or caries. This is an important prerequisite for applying dentinogenesis as model for biomineralization, in order to gain insight into the biomineralization processes of hydroxyapatite [3].

11.1 Introduction

Fig. 11.2 Schematic presentation of the association of enamel knot signaling with morphogenesis and odontoblast diﬀerentiation. (A) During the bud stage, the condensed odontogenic mesenchyme induces the formation of the primary enamel knot at the tip of the epithelium. (B) During the cap stage, the enamel knot expresses signaling molecules, which regulate formation of the dental papilla and growth of the cervical loops of the epithelium. (C) In the bell stage, signals from the secondary enamel knots regulate the formation of cusps and may induce the initiation of terminal diﬀerentiation of odontoblasts. Diﬀerentiation proceeds toward the intercuspal areas and cervical loops. (D) An enlarged view of (B) shows the induction of dental papilla cells in the dental mesenchyme underlying the primary enamel knot. (E) An enlarged view of the cusp tip at

the time of secondary enamel knot formation (stage slightly preceding that in C) shows the induction of odontoblast diﬀerentiation in the dental papilla cells underlying the secondary enamel knot. (F) An enlarged view of the odontoblast diﬀerentiation proceeding at the slope of the cusp (in the region of the vertical arrow in the left cusp in C). After initiation odontoblast diﬀerentiation at the cusp tip, the diﬀerentiation signals may come from the epithelium in which the expression of several enamel knot signals are spreading (arrows from left to right) and/or signals may be relayed by diﬀerentiating odontoblasts (vertical arrow). The odontoblasts secrete dentin and induce the terminal diﬀerentiation of ameloblasts (arrows from right to left). (Figure reproduced from Ref. [84], with permission.)

161

162

11 Formation of Teeth

11.2 Odontogenesis

11.2 Odontogenesis

The determination – and therefore the positioning – of dental hard substanceforming tissue precedes the actual formation and development of dental hard substances in the framework of embryologic development (Figs. 11.2A–C and 11.3A–E). Enamel is derived from the oral epithelium, whereas dentin and cementum are of ectomesenchymal origin. In humans, this process of tooth development starts 28 to 40 days after ovulation, with a thickening of the oral epithelium, which later evolves into the dental lamina, caused by the subepithelial ectomesenchyme in the area of the oral cavity space. The mechanisms for the epithelium to start depth growth, thus forming the dental lamina, are not precisely known. The subepithelial ectomesenchyme is recruited from cells that have migrated into the oral cavity and originally stem from the neural crest ectomesenchyme of the maxillary and mandibular processes. This is an interactive process between the oral epithelium and the subjacent extomesenchyme [4]. Starting from this dental ridge, compact epithelial buds are formed (bud stage) between the 7th and 10th weeks. By proliferation of the bud margins, as well as by diﬀerentiation of cells, the so-called ‘‘cap stage’’ is reached (8th to 10th weeks after ovulation). During the cap stage the enamel organ separates into four areas: the outer enamel epithelium; the stellate reticulum; the stratum intermedium; and the inner enamel epithelium. Later, the tooth germ enters into the bell stage, which marks the completion of this development and thus the end of the proliferation phase (Figs. 11.2A–C, 11.3A–E and 11.4). The germs of the anterior deciduous teeth reach this stage between the 12th and 16th weeks, and the posterior teeth normally between the 15th and 21st weeks, after ovulation. During the bell stage, the dental organ – which comprises the dental papilla, surrounded by the ________________________________________________________________________________ H Fig. 11.3 Tooth development. Histologic sections through human prenatal jaws at diﬀerent stages of odontogenesis. In order to keep track of the tooth, the germs have been turned around in some illustrations, so that tooth germs from the mandibula and maxilla are facing in the same direction. (A) Bud stage: epithelial invagination from the dental lamina (DL) into the underlying ectomesenchyme; ec ¼ epithelial cells; mc ¼ mesenchymal cells; arrows indicate the basement membrane. 8th gestational week (gw), maxillary process. (B) Cap stage: epithelial cap (enamel organ) with enamel knot (EK) and dental lamina (DL); dental papilla (DP) as ectomesenchymal condensation; alveolar bone (AB). 9th gw,

mandibula. (C) Late cap stage: enamel knot (EK) as signaling center of the enamel organ dental lamina (DL), dental papilla (DP), alveolar bone (AB). The arrow indicates a cross-section through a blood vessel. 10th gw, maxilla. (D) Bell stage: dental lamina (DL); dental papilla (DP); dental sac (DS); enamel organ (EO); stellate reticulum (SR). 22nd gw, mandibula, permanent dentition germ. (E) Early formation of dentin and enamel. Ameloblasts (A); enamel (E); odontoblasts (O); dentin (D). The dental papilla becomes the future pulp (P); CEJ ¼ cemento-enamel junction. 22nd gw, mandibula, deciduous molar germ. Hematoxylin and eosin staining; illustration courtesy of Werner Goetz, University of Bonn/Germany.

163

164

11 Formation of Teeth

bell-shaped dental organ – is made up from ectomesenchymal tissue. The outer enamel epithelium is the outer boundary and devolves into the inner enamel epithelium at the margin of the ‘‘bell’’ (the so-called cervical loop). The inner enamel epithelium forms the inner surface of the bell and is divided from the dental papilla by a basal membrane. On top of the single-layer of cells from the inner enamel epithelium, from which ameloblasts are diﬀerentiating, rests the three- to four-layered stellate reticulum. The internal space of the dental organ is taken up by the stellate reticulum, the star-shaped cells of which are imbedded in a mucopolysaccharide-enriched intercellular area. The basal membrane of the inner enamel epithelium is the mold of the initially built dentin, and represents the future enamel–cemental junction. Along that line it comes to additional induction for diﬀerentiation of odontoblasts (¼ dentin-forming cells) and ameloblasts (¼ enamel-forming cells) (Fig. 11.4). Thus, the dental papilla causes diﬀerentiation of the inner enamel epithelium, which again induces the forming and alignment of odontoblasts on the basal membrane of the inner enamel epithelium and the formation of primary dentin. As soon as the ﬁrst layer of primary dentin becomes the basis for forming enamel, diﬀerentiation of the inner enamel epithelium from pre-ameloblasts to ameloblasts is induced (Figs. 11.2E–G, 11.3A–E and 11.4). The above-described diﬀerentiation is a chronological as well as spatial proceeding progress. This progress starts with a relatively small bell shape in the small area where mineralization of the future cusps is due to start. The small bell shape becomes increasingly larger as a result of proliferation in the area of the cervical loop – while the enamel and dentin are already deposited – until the form has reached its ﬁnal size of the future dentinal core [4–7].

Fig. 11.4 Schematic representation of the processes during tooth development in the late bell stage.

11.3 Dentinogenesis

On average, human tooth eruption of the ﬁrst primary tooth takes place at the age of 6 to 12 months, although at this stage the root is not completed. Human permanent maxillary anterior teeth begin to mineralize around the date of birth, they erupt approximately at the age of six years, and root development is normally completed about three years later. 11.2.1 Genes Involved in Odontogenesis

Some genes are known to be involved in the development of teeth, for example msh homeobox homolog 1 (Msx1), bone morphogenetic protein 4 (Bmp4), paired box gene 9 (Pax9), and lymphoid enhancer binding factor 1 (Lef1). An important signaling center during tooth development until the cusps mineralize represents the enamel knot (Figs. 11.2 and 11.3C). In a study investigating the developing cranial tissues of mice, it was shown that in the signaling interplay of the epithelium and mesenchyme during the transition from bud stage to cap stage, epithelial-driven ﬁbroblast growth factor (FGF), which is induced by Lef1, in turn induces runt-related transcription factor 2 (Runx2) expression in the dental mesenchyme through Msx1. Runx2 regulates mesenchymal FGF 3 expression. This and other downstream targets of Runx2 induce sonic hedgehog homolog (Shh) expression in the enamel knot. Runx2 expression controls activin expression, which reciprocally acts on the dental epithelium and activates the ectodysplasin A receptor (Edar) expression in the enamel knot [8]. The interaction of Pax9 with Msx1 and the subsequent controlling of Bmp4 expression seems also to be involved in the transition from the bud stage to the cap stage during tooth development [9]. 11.2.2 Stem Cells

Dental pulp stem cells represent another chapter in tooth formation. Pulp tissue is of mesenchymal origin, and contains pluripotent cells with stem cell character. Several studies have been conducted on both permanent third molars as well as on erupted primary teeth analogous to bone marrow stem cells [10, 11]. It has become clear that a great deal of work has still to be done before dental therapy is able to beneﬁt from these insights into the conditions required for the regeneration of dental tissues.

11.3 Dentinogenesis

Dentin is formed by odontoblasts developing from ectomesenchymal cells of the ‘‘dental papilla’’ in contact with the basal membrane of the inner enamel epithe-

165

166

11 Formation of Teeth

lium. It has long been considered that this contact with the inner enamel epithelium and/or with other associated extracellular tissue of epithelial origin served as trigger for the initial diﬀerentiation of odontoblasts. In the meantime, it has been established that a substrate enriched by ﬁbronectin is the crucial precondition for diﬀerentiation of odontoblasts [15]. Several diﬀerent types of dentin have been identiﬁed, with the classiﬁcation being made according to stages of development and recorded as primary, secondary, and tertiary dentin. Primary dentin is present in the largest quantities, and is formed until the growth of the root has been completed. Subsequently, secondary dentin is formed in physiological terms, albeit in much lesser quantities. Certain stimuli cause the growth of tertiary dentin lifelong [12–15]. The classiﬁcation of dentin can also be based on the mechanism of its formation, as described in the following sections. 11.3.1 Mantle and Circumpulpal Dentin

The inner enamel epithelium inﬂuences the diﬀerentiation of ectomesenchymal cells in the dental papilla. The cells closest to the inner enamel membrane align themselves on the tooth bell and polarize, lose their ability to divide, and form a junctional complex. These pre-odontoblasts begin to excrete an extracellular dentin matrix pre-form consisting of type I and type III collagen, ﬁbronectin, and glycosaminoglycans. The dentin layer that these pre-odontoblasts secrete directly at the dentinoenamel junction has a thickness of about 10 to 30 mm, and is termed ‘‘mantle dentin’’ once mineralization is completed. It can be distinguished from the circumpulpal dentin, which is formed by diﬀerentiated odontoblasts [5–7, 13, 16–18]. The mineralization of mantle dentin proceeds by budding vesicles from the tips of the odontoblast processes. The course of budding is polarized, with only a particular part of the outer cell membrane generating matrix vesicles [19]. As the concentrations of calcium and phosphate ions increases, hydroxyapatite crystallizes in the area of the inner membrane of the vesicles. An elemental analysis of freeze-dried matrix vesicles of rats points to early mineral deposition occurring as dot-like nuclei inside the vesicles [20]. First, calcium ions are transfected to matrix vesicles by calcium-binding molecules that are concentrated in the matrix vesicle structure. The local intra- and perivascular phosphate concentration is increased by the enhanced activity of phosphohydrolases. Their incidence in the matrix membrane is very high, so the processes take place predominantly in the matrix membrane as, for example, the activity of alkaline phosphatase. When the solubility product of calcium and phosphate ions is reached, precipitation of the ﬁrst mineral deposits occurs near the inner surface of the matrix vesicle membrane, where calcium- and phosphate-concentrating molecules coincidence occurs [19]. The continuing growth of crystallites causes the vesicles to dehisce and thus to release crystallites into the extravesicular matrix, a process which might involve phospholipases and proteases. The released crystals function in the extravesicular ﬂuid as seeds for mineralization. When mineralization is initiated in such a way,

11.3 Dentinogenesis

apparently no further reason exists for the presence of matrix vesicles. Independently of matrix vesicles, phosphophoryn initiates the mineralization of circumpulpal dentin. Phosphophoryn, a macromolecule arising in the extracellular matrix (ECM), binds calcium ions with high aﬃnity and at the same time aggregates collagen ﬁbrils on the mineralization front. The interspace between the carboxyl and phosphate groups, as positioned in the phosphophoryn molecule, might cause speciﬁc crystal orientation relative to the collagen substrate for the future mantle dentin and circumpulpal dentin. The phospophoryn might function as a link between collagen ﬁbrils and crystals [21]. With proceeding deposition of the dental matrix, the odontoblast process elongates and becomes embedded in mineralized tissue, such that the tubular structure of dentin comes into existence. These dentinal tubules act as guide trail for external stimuli such as bacterial toxins, cavity preparation, or the ingredients of dental restorative materials, thus inducing a further apposition of hard tissue in the pulp cavity, the so-called ‘‘tertiary’’ dentin formation. Initially, studies conducted in animals and various analyses showed that neuronal signals cause a diﬀerentiation of odontoblasts and subsequent hard tissue synthesis [1, 2], despite the nerves penetrating only 50 to 70 mm into predentin/dentin [22]. Therefore, cilia – which have been identiﬁed in the odontoblast membrane in a supranuclear position – were considered to serve as a link between ﬂuid movement in the tubules and odontoblast/nerve response [23]. However, details of the exact relationship between neuronal impulses and hard tissue synthesis function is not entirely known. Diﬀerentiated odontoblasts are polar cells some 40–50 mm in length, and with a diameter of about 7 mm. The odontoblast’s characteristic form is a strong, distal cytoplasmic process showing a distinguished microtubules and microﬁlament network. This process remains intact in length, even during the retraction of odontoblasts from the cementoenamel junction, and is characteristic for dentin. The predentin secreted from odontoblasts is built up from ﬁbers of collagen type I, glycoproteins, and glycosaminoglycans. As the predentin still has to mature to a dentin matrix, there is a space of 5 to 20 mm between the frontlines of odontoblasts and mineralization (Fig. 11.5). It is assumed that mineralization occurs by

Fig. 11.5 Odontoblasts at the interface pulp-dentin; the odontoblasts are not detachable due to their anchorage in the dentinal tubules.

167

168

11 Formation of Teeth

matrix vesicles and that, at a later time the centers of mineralization become conﬂuent, although the particular sequences and pathways involved have not yet been determined. 11.3.2 Intertubular Dentin

So-called ‘‘intertubular dentin’’, which forms the largest part of mineralized circumpulpal dentin, is formed by the mineralization of predentin. The matrix of intertubular dentin is rich in collagen I ﬁbrils, while hydroxyapatite crystals of about 40 nm length are formed in and around the collagen ﬁbrils of intertubular dentin [14, 16, 17, 24–32]. 11.3.3 Peritubular Dentin

Intertubular dentin is distinguished from a second and minor component of circumpulpal dentin, the peritubular dentin. The latter is formed by the secretion of a matrix consisting of mucopolysaccharides in the space resulting from a reduction of the odontoblast process diameter, although almost no collagen ﬁbers are involved. Peritubular dentin is mineralized to a greater degree than intertubular dentin, and contains 50% ﬁbers. Bone sialoprotein and osteonectin have been localized in the peritubular dentin [4]. The formation of peritubular dentin is a lifelong process which may be increased by external stimuli; this can lead to a complete macroscopic obturation of the dentinal tubuli [14, 28, 29, 32]. About 90% of the ECM of dentin is type I collagen. As collagen I cannot initiate the mineralization of hydroxyapatite [33], a special group of macromolecules takes on the initiation and control of matrix mineralization; those such materials that have been identiﬁed are mostly proteins and are referred to as non-collagenous matrix proteins (NCPs). Initially, only two NCPs were found exclusively in dentin in the framework of dentinogenesis, namely dentin sialoprotein (DSP) and dentin phosphoprotein (DPP). Both are cleavage products of a precursor protein, the dentin sialophosphoprotein (DSPP) [34, 35]. In histological sections through human prenatal jaws and in cultured human pulp-derived cells, DSPP and other non-collagenous proteins were localized immunohistochemically (Fig. 11.6A and B), and were also relatively quantiﬁed using RT-PCR [36]. In rats it could be shown that proteolysis is carried out by the bone morphogenetic protein 1 (BMP1), belonging to a group of metal proteinases which are, among other things, responsible for the synthesis of collagen ﬁbrils [37]. DSPP mRNA was meanwhile detected in the bone of mice, albeit in clearly smaller quantities than in teeth [38, 39]. These data were veriﬁed by unpublished data from the present authors’ studies with bovine and human odontoblasts. At which stage cleavage into DSP and DPP occurs – and what function this has during the biomineralization of dentin – is not known. DPP has been proposed to promote and inhibit mineral deposition during denti-

11.3 Dentinogenesis

Fig. 11.6 Cap stage: immunohistochemical staining of paraﬃnembedded histological sections through human prenatal jaw in the 22nd gestational week (gw). (A) Anti-Ca-ATPase visualized with Alexa 488 (green), anti-osteopontin with Alexa 594 (red) and nuclei with DAPI (blue). (B) Anti-DSPP visualized with Alexa 488. (Anti-DSPP kindly provided by Larry Fisher, NIH.)

nogenesis. Indeed, experiments on bovine DSP show, after dephosphorylation of DPP by alkaline phosphatase, a reduced aﬃnity for apatite compared with phosphorylated DPP when bound to collagen ﬁbrils; phosphorylated DPP signiﬁcantly promotes the rate per time of hydroxyapatite crystal growth compared to the dephosphorylated DPP [40]. The most commonly found protein in dentin is the above-mentioned phosphophoryne. This was detected using immunolocalization, the protein being identiﬁed between the gap and the overlapped zone along the ﬁbrils closest to the mineralization front. It seems as if the structure of the collagen matrix would change shortly prior to mineralization. These processes seem to be controlled by the addition of highly anionic non-collagenous proteins and their interaction associated with collagen. These changes may be responsible for the formation of collagenous netting capable of mineralization [41]. A number of other non-collagenous proteins have been identiﬁed, including dentin matrix acidic phosphoprotein (DMP1), which is a member of the sibling family (small integrin binding ligand N-linked glycoproteins). DMP1 is highly expressed in mineralizing tooth and bone, and is critical for mineralization and tooth morphogenesis [42, 43]. In a rat model, DMP1 was shown to act as a morphogene on undiﬀerentiated mesenchymal cells of the dentin–pulp complex. In this case, cells diﬀerentiate into odontoblast-like cells and regenerate a dentin-like tissue [44]. Another group of non-collagenous proteins are the insulin-like growth factors (IGFs), their binding proteins (IGFBPs), and receptors. IGF I is known to be ex-

169

170

11 Formation of Teeth

Fig. 11.7 Immunohistochemical staining of expression of dentin sialophosphoprotein. Cultured human pulp-derived cells after 3 weeks of cultivation on ThermanoxTM (Nunc GmbH & Co. KG, Wiesbaden, Germany).

pressed during tooth development and to be on a constant level during the diﬀerentiation of human pulp-derived cells, while IGF II is expressed on a low level. IGFBP-3 levels increase during the diﬀerentiation of human pulp-derived cells, while IGFBP-2 levels remain constant during the course of cultivation. Most strikingly, there is evidence for direct cellular actions by the up-regulation of the number of IGF I receptors (IGF-IR), and in this way controlling the activity of IGF I [45].

11.4 Amelogenesis

Those cells which form enamel are termed ‘‘ameloblasts’’. With regards to the formation of dental enamel, higher vertebrates have evolved a process of biomineralization that yields the strongest material in the biological world. The timing and position of the crystal initiation, as well as crystal structure, shape and orientation have all been brought under genetic control [46, 47]. Following its formation, enamel is not subjected to any further remodeling. In contrast to bone or dentin, crystallite growth during amelogenesis continues throughout the entire period of formation, so that these crystallites are the largest of the hard tissues. In order to make such crystallite growth possible, the primary secretory protein matrix must be resorbed. In amelogenesis, it is possible to distinguish between the following developmental phases: Formation and secretion of the enamel matrix; Initial mineralization of the enamel matrix; Crystal nucleation and crystal elongation; Re-resorption of the enamel matrix; and Secondary mineralization of enamel (crystal maturation).

11.4 Amelogenesis

The enamel matrix is secreted from ameloblasts, which retract during secretion, thus forming the enamel–cemental junction. This newly secreted enamel matrix consists mostly of non-collagenous proteins (@90% amelogenin) and of 1–2% carbohydrates and lipids. In the case of amelogenin – the main non-collagenous structural protein in amelogenesis – the two domains which have been identiﬁed as being required for its proper selfassembly into supramolecular structures are referred to as nanospheres, and provide control of the hydroxyapatite crystal formation [48]. The remaining part of the non-collagenous proteins mostly comprise proline, leucine, histidine and glutamic acid, which function as a control system during germ formation and regulate germ growth (e.g., ameloblastin, enamelin). Although these proteins are always subjected to genetic control, the exact process of mineral germ formation and the details of mineralization remain unknown. The crystals have a thin, needle-like structure of about 15 A˚ thickness, 300 A˚ width and 800 to 1200 A˚ length. On completion of the initial mineralization, the mineral content is about 30–60% by volume. In that state, the developing enamel has a relatively soft consistency, is jelly-like, glassy, and contains ﬂuid [46, 49–61]; this softness is due to hydration of the glycosaminoglycans of the stellate reticulum. This composition may be responsible for equalizing the pressure generated by proliferation and matrix secretion in the dental papilla. Maturation of the enamel starts when the enamel layer has reached its predesigned thickness. In a ﬁrst step, secretory proteases cleave the organic matrix, and the fragments are resorbed by ameloblasts. The resulting pores are ﬁlled with ﬂuid. Complex re-resorption of the organic matrix is needed to justify such a high occurrence of mineral content (ca. 98%). The continuous growth is enabled by the re-resorption, and not by formation of additional crystallites. The conditions of crystallite growth are controlled by the layer of ameloblasts – that is, by inﬂuencing parameters such as calcium ion availability, pH-value, and the reresorption of ﬂuid and peptides [46, 55, 57–59, 62–64]. The Vicker’s hardness reached by matured enamel is between 300 and 430; this complies with the hardness of steel, but the enamel is much more brittle. The layer thickness of enamel can be up to 2.5 mm in the area of the cup tips. The layer of ameloblasts turns into junctional epithelium (part of the gingiva) during dentition, and therefore – unlike bone – enamel cannot regenerate. Thus, enamel damage can be corrected only by dental intervention [7, 58]. Enamel is arranged cross-sectionally in prisms which provide information about the prior course of the ameloblasts. Each arcade-like prism is surrounded by interprismatic enamel, and the morphology of the ameloblasts is described as follows. The term ‘‘proximal’’ is used to refer to the end of the cell nearest the stratum intermedium, to which the prisms are connected by GAP junctions. The term ‘‘distal’’ (apical) term is used to identify the secretory pole of the cell, next to the enamel. The cytoplasm is strongly polarized [4]. The proximal cytoplasm contains many mitochondria, and a ‘‘terminal web’’ of cytoplasmic ﬁlaments associated with a zonula adherens. The distal cytoplasm, which amounts to slightly more than half of the total cell volume, is equipped

171

172

11 Formation of Teeth

with an extensive rough endoplasmic reticulum (RER), cisternae, and a welldeveloped Golgi complex [65].

11.5 Cementogenesis

Cementum, a heterogeneous hard tissue, is a thin, calciﬁed layer of a bone-like tissue that covers the dentin of the tooth root. The formation of cementum occurs during root development and throughout the life of the tooth, and functions as an area of attachment for the periodontal ligament ﬁbers. The area cementum, which joins the enamel at the cervix of the tooth, is known as the ‘‘cementoenamel junction’’. Root cementum exists in several histologic types: acellular extrinsic ﬁber cementum (AEFC); acellular intrinsic ﬁber cementum (AIFC); cellular intrinsic ﬁber cementum (CIFC); and cellular mixed stratiﬁed cementum (CMSC) [66–69]. The primary function of the root cementum is to anchor the tooth in its socket by inserting desmodontal ﬁbers onto the root surface; in this way an elastic suspension is made possible for the tooth in its socket (periodontal ligament). Cementum also serves as a reservoir for cementocytes which, under special conditions, have the capability of diﬀerentiating into cementoblasts. The latter are able to restore continuity when resorptions or fractures occur on the roots of teeth, or to pad lacunes. Protective factors within the cementum prevent the resorption of exposed dentin by osteoclasts. Mature cementoblasts are relatively large cells with a highly basophilic cytoplasm; during CIFC formation they secrete cementum relatively rapidly and subsequently become entrapped in the matrix as cementocytes [66, 68, 70]. Immunohistochemical analyses of IGF system components in human deciduous teeth have revealed that odontoclasts do not express IGFs or the IGF-IR, but do contain IGFBPs and the IGF-IIR [71]. These results suggest that odontoclasts, in contrast to osteoclasts, may not respond to IGFs, but are most likely involved in the release and sequestration of IGFs from cementum during the resorption process. In contrast to odontoclasts, cementoblasts and periodontal ligament ﬁbroblasts express the IGF-IR. 11.5.1 Acellular Extrinsic Fiber Cementum (AEFC)

One of the ﬁrst relevant steps during cementogenesis is disruption of continuity of the so-called Hertwig’s epithelial root sheath (HERS) by adjacent cells (see Fig. 11.4). In mice, the above-mentioned sonic hedgehog (shh) seems to be involved in this process [72]. HERS is a double layer of epithelial cells that is continuous with and extends apically from the apical rim of the enamel organ. The deposition of root cementum begins just apical of the cervical enamel, and during human tooth development HERS loses contact with the root surface after the diﬀerentiation of odontoblasts. At the cap stage of human tooth development, the apical rim

11.5 Cementogenesis

of the enamel organ forms the edge of the apical foramen. After detachment and disintegration of the HERS, the AEFC forms on the growing root, when ﬁbroblasts of the dental follicle come into contact with the unmineralized matrix of the dentin [4]. These ﬁbroblasts secrete in a unipolar direction and deposit collagen ﬁbrils on the dentin surface, in order to form a thin layer of perpendicularly directed ‘‘fringe ﬁbers’’ [73]. When the dentin mineralization front slowly reaches the outer area of the mantle dentin it touches the ‘‘fringe ﬁbers’’, which in turn disappear by slow mineralization, in order to complete the process of AEFC formation. Within the life expectancy of a tooth the growth of thickness of the AEFC continues annually by 1.5 to 3.0 mm [4]. During the initial formation of cementum, two processes occur simultaneously: the ﬁrst process is loss of the external basal lamina adjacent to the cells of the dental follicle proper; the second process is the start of the diﬀerentiation of pre-cementoblasts. Whichever passive or active metabolic functions are assigned to epithelial cells, their function is considered to be far more complex than those of the inner enamel epithelium cells or inner root sheath cells in odontoblast differentiation [74, 75]. Acellular extrinsic ﬁber cementum lacks cells, and is composed of densely packed striated collagen ﬁber bundles embedded in a glycosamine-rich matrix. These ﬁbers termed ‘‘Sharpey’s ﬁbers’’ – are orientated perpendicular to the root surface. The acellular extrinsic ﬁber cementum is located in the coronal part of the root, and covers 40–70% of the root surface. Its function is to attach the root to the periodontal ligament, which means anchoring the tooth in its socket by ﬁbers. The desmodontal surface of cementum is characterized by inserting Sharpey’s ﬁbers and a thin layer of non-calciﬁed cementoid [76]. Cementoid is rich in proteoglycans, which are known to function as IGF carriers. In the cementum, several IGF components were found, indicating roles in tissue homeostasis or attachment [71]. Any additional growth of the AEFC is very slow (<0.1 mm per day), and it forms parallel to the root development of a tooth. As soon as the cervical loop has advanced into the area, the root becomes a double cell layer, the HERS, the cells of which cease to diﬀerentiate but then cause an induction of osteoblasts on the inside of the root sheath. Later, root dentin is formed by these cells. As soon as the formation of pre-dentin begins, the cell layer of the Hertwig’s epithelial root sheath dissolves. A special population of ﬁbroblasts secrete a lawn-like structure made from collagenous ﬁbrils on the exposed pre-dentin; this structure is arranged vertical to the surface. These ﬁbrils are then mineralized under a steady extension of the dentin–cement-junction [77–81]. 11.5.2 Cellular Intrinsic Fiber Cementum (CIFC)

The CIFC is formed in connection with adaptation and repair processes, and does not have the capability of anchoring a tooth. It contains cementocytes embedded in a collagenous matrix, and can be found in lacunes of resorption and in root fracture planes. CIFC comes into existence both pre- and post-eruptively,

173

174

11 Formation of Teeth

and is predominantly formed apically in the area of furcations. This type of cementum is formed by cementoblasts secreting collagenous ﬁber bundles, which run parallel to the root, into the collagenous matrix surface (¼ intrinsic ﬁbers). The matrix is secreted both against the pre-dentin surface of a root and around cementoblasts, whereas the latter wall themselves in and become cementocytes. This is analogous to bone formation by osteoblasts/osteocytes [4, 77–80]. 11.5.3 Cellular Mixed Stratiﬁed Cementum (CMSC)

In addition to the above-described types of cementum there is a hybrid of both types, the so-called ‘‘cellular mixed stratiﬁed cementum’’. This consists of alternating layers and irregular sequence-deposited portions of AEFC and CIFC/ AIFC. CMSC covers the apex, the apical third of the roots, and the furcational aspects of human teeth. The dynamics of this layering, and the fact that, temporarily, the root surfaces may remain unsupported by Sharpey’s ﬁbers as a consequence of this layering, are discussed. Cellular mixed cementum serves to reshape root surfaces to accommodate for physiologic drift and non-physiologic shifting of teeth in the tooth socket, as well as for the repair of resorption sites [66, 69, 82]. It is layered by AEFC for attachment to the periodontal ligament [80]. 11.5.4 Acellular Intrinsic Fiber Cementum (AIFC)

This type of cementum is secreted in unipolar mode by cementocytes [79], and forms shortly before and during dentition, covering the dentin in forms such as tongues or isles. By secreting matrix slowly from only one surface, the cementoblasts avoid their subsequent entrapment in matrix as cementocytes [4]. 11.6 Acknowledgments

These studies were supported by the Deutsche Forschungsgemeinschaft (DFG) SPP 1117, RE-1562-3. This chapter is dedicated to Prof. Dr. H. E. Reichenmiller, to mark the occasion of his 70th birthday. References 1 M.R. Byers, M.V. Narhi, Crit. Rev.

Oral. Biol. Med. 1999, 10, 4. 2 H.D. Rodd, F.M. Boissonade, J. Dent. Res. 2001, 80, 389. 3 E. Ba¨uerlein, Biomineralization: Progress in Biology, Molecular Biology

and Application. Completely revised and extended edition, 2nd edn. WileyVCH, Weinheim, 2004. 4 P.R. Garant, Oral Cells and Tissues, 1st edn. Quintessence Publishing Co., Inc., Illinois, 2003.

References 5 J.V. Ruch, Biochem. Cell Biol. 1998, 76, 6 7

8 9

10

11

12 13 14

15

16 17 18

19 20

21 22

23

923. J.V. Ruch, H. Lesot, C. Begue-Kirn, Int. J. Dev. Biol. 1995, 39, 51. G.-H. Schuhmacher, H. Schmidt, H. Bo¨rning, W. Richter, in: G.-H. Schuhmacher, H. Schmidt, H. Bo¨rning, W. Richter (Eds.), Anatomie und Biochemie der Za¨hne. Gustav Fischer Verlag, Stuttgart, New York, 1990, p. 55. H.M. Ryoo, X.P. Wang, Crit. Rev. Eukaryot. Gene Expr. 2006, 16, 143. T. Ogawa, H. Kapadia, J.Q. Feng, R. Raghow, H. Peters, R.N. D’Souza, J. Biol. Chem. 2006, 281, 18363. S. Gronthos, M. Mankani, J. Brahim, P.G. Robey, S. Shi, Proc. Natl. Acad. Sci. USA 2000, 97, 13625. S. Shi, P.M. Bartold, M. Miura, B.M. Seo, P.G. Robey, S. Gronthos, Orthod. Craniofac. Res. 2005, 8, 191. D. Tziafas, Int. J. Dev. Biol. 1995, 39, 281. D. Tziafas, Int. Endod. J. 1994, 27, 61. A.J. Smith, N. Cassidy, H. Perry, C. Begue-Kirn, J.V. Ruch, H. Lesot, Int. J. Dev. Biol. 1995, 39, 273. H. Lesot, S. Lisi, R. Peterkova, M. Peterka, V. Mitolo, J.V. Ruch, Adv. Dent. Res. 2001, 15, 8. A. Linde, T. Lundgren, Int. J. Dev. Biol. 1995, 39, 213. T. Sasaki, P.R. Garant, Anat. Rec. 1996, 245, 235. H.E. Schroeder, in: H.E. Schroeder (Ed.), Orale Strukturbiologie. Georg Thieme Verlag, Stuttgart, New York, 2000, p. 85. H.C. Anderson, Curr. Rheumatol. Rep. 2003, 5, 222. U. Stratmann, K. Schaarschmidt, H.P. Wiesmann, U. Plate, H.J. Hohling, Cell Tissue Res. 1996, 284, 223. A. George, J. Hao, Cells Tissues Organs 2005, 181, 232. T. Ibuki, M.A. Kido, T. Kiyoshima, Y. Terada, T. Tanaka, J. Dent. Res. 1996, 75, 1963. H. Magloire, M.L. Couble, A. Romeas, F. Bleicher, Cell. Biol. Int. 2004, 28, 93.

24 M. Goldberg, M. Takagi, Histochem. J.

1993, 25, 781. 25 W.T. Butler, Connect. Tissue Res. 1995,

33, 59. 26 A. Linde, Connect. Tissue Res. 1995,

33, 163. 27 W.T. Butler, H. Ritchie, Int. J. Dev.

Biol. 1995, 39, 169. 28 M. Goldberg, J.J. Lasfargues, J. Dent.

1995, 23, 15. 29 M. Goldberg, D. Septier, S. Lecolle,

30

31 32 33 34

35

36

37

38

39

40

41

42

H. Chardin, M.A. Quintana, A.C. Acevedo, G. Gafni, D. Dillouya, L. Vermelin, B. Thonemann, Int. J. Dev. Biol. 1995, 39, 93. M. Goldberg, D. Septier, S. Lecolle, L. Vermelin, P. Bissila-Mapahou, J.P. Carreau, A. Gritli, A. BlochZupan, Connect. Tissue Res. 1995, 33, 105. W.T. Butler, Eur. J. Oral Sci. 1998, 106 Suppl. 1, 204. I.A. Mjor, O.B. Sveen, K.J. Heyeraas, Quintessence Int. 2001, 32, 427. Y. Chen, B.S. Bal, J.P. Gorski, J. Biol. Chem. 1992, 267, 24871. M. MacDougall, D. Simmons, X. Luan, J. Nydegger, J. Feng, T.T. Gu, J. Biol. Chem. 1997, 272, 835. C. Qin, R.G. Cook, R.S. Orkiszewski, W.T. Butler, J. Biol. Chem. 2001, 276, 904. K.M. Reichenmiller, C. Klein, Int. Endod. J. 2005, 38, 941, abstract 67. B.M. Steiglitz, M. Ayala, K. Narayanan, A. George, D.S. Greenspan, J. Biol. Chem. 2004, 279, 980. C. Qin, J.C. Brunn, E. Cadena, A. Ridall, H. Tsujigiwa, H. Nagatsuka, N. Nagai, W.T. Butler, J. Dent. Res. 2002, 81, 392. C. Qin, J.C. Brunn, E. Cadena, A. Ridall, W.T. Butler, Connect. Tissue Res. 2003, 44 Suppl 1, 179. A.M. Milan, R.V. Sugars, G. Embery, R.J. Waddington, Eur. J. Oral Sci. 2006, 114, 223. E. Beniash, W. Traub, A. Veis, S. Weiner, J. Struct. Biol. 2000, 132, 212. Y. Lu, S. Zhang, Y. Xie, Y. Pi, J.Q. Feng, Cells Tissues Organs 2005, 181, 241.

175

176

11 Formation of Teeth 43 K.D. Moses, W.T. Butler, C. Qin, Eur. 44

45

46 47

48 49

50 51

52 53 54

55

56

57 58

59

60

61

J. Oral Sci. 2006, 114, 216. A. Almushayt, K. Narayanan, A.E. Zaki, A. George, Gene Ther. 2006, 13, 611. K.M. Reichenmiller, C. Mattern, M.B. Ranke, M.W. Elmlinger, Horm. Res. 2004, 62, 33. J.P. Simmer, A.G. Fincham, Crit. Rev. Oral Biol. Med. 1995, 6, 84. J.C. Hu, Y. Yamakoshi, F. Yamakoshi, P.H. Krebsbach, J.P. Simmer, Cells Tissues Organs 2005, 181, 219. M.L. Paine, M.L. Snead, Orthod. Craniofac. Res. 2005, 8, 239. D. Deutsch, J. Catalano-Sherman, L. Dafni, S. David, A. Palmon, Connect. Tissue Res. 1995, 32, 97. A.G. Fincham, J. Moradian-Oldak, Connect. Tissue Res. 1995, 32, 119. C. Robinson, J. Kirkham, S.J. Brookes, W.A. Bonass, R.C. Shore, Int. J. Dev. Biol. 1995, 39, 145. A.G. Fincham, J.P. Simmer, Ciba Found. Symp. 1997, 205, 118. T. Sasaki, M. Takagi, T. Yanagisawa, Ciba Found. Symp. 1997, 205, 32. T. Aoba, H. Komatsu, Y. Shimazu, H. Yagishita, Y. Taya, Connect. Tissue Res. 1998, 38, 129. C. Robinson, S.J. Brookes, R.C. Shore, J. Kirkham, Eur. J. Oral Sci. 1998, 106 Suppl 1, 282. A.G. Fincham, J. Moradian-Oldak, J.P. Simmer, J. Struct. Biol. 1999, 126, 270. J.D. Bartlett, J.P. Simmer, Crit. Rev. Oral Biol. Med. 1999, 10, 425. H.E. Schroeder, in: H.E. Schroeder (Ed.), Orale Strukturbiologie. Georg Thieme Verlag, Stuttgart, New York, 2000, p. 37. J. Moradian-Oldak, I. Jimenez, D. Maltby, A.G. Fincham, Biopolymers 2001, 58, 606. M.L. Paine, S.N. White, W. Luo, H. Fong, M. Sarikaya, M.L. Snead, Matrix Biol. 2001, 20, 273. S. Oida, T. Nagano, Y. Yamakoshi, H. Ando, M. Yamada, M. Fukae, J. Dent. Res. 2002, 81, 103.

62 A.R. do Espirito Santo, P.D. Novaes,

63 64 65 66

67 68 69

70

71

72 73 74 75 76

77 78 79 80

81 82

S.R. Line, Eur. J. Oral Sci. 2006, 114 Suppl 1, 333. C.E. Smith, Crit. Rev. Oral Biol. Med. 1998, 9, 128. J. Moradian-Oldak, Matrix Biol. 2001, 20, 293. C.E. Smith, A. Nanci, Int. J. Dev. Biol. 1995, 39, 153. H.E. Schroeder, in: A. Oschke (Ed.), Handbook of Microscopic Anatomy. Springer-Verlag, Berlin, Heidelberg, 1989, p. 23. D.D. Bosshardt, K.A. Selvig, Periodontology 2000 1997, 13, 41. H.E. Schroeder, Int. Rev. Cytol. 1992, 142, 1. T. Yamamoto, T. Domon, S. Takahashi, M. Wakita, Anat. Embryol. (Berl) 1996, 193, 495. T. Yamamoto, T. Domon, S. Takahashi, M. Wakita, Anat. Embryol. (Berl) 1996, 193, 495. W. Go¨tz, U. Kruger, S. Ragotzki, S. Lossdorfer, A. Jager, Connect. Tissue Res. 2001, 42, 291. M. Nakatomi, I. Morita, K. Eto, M.S. Ota, J. Dent. Res. 2006, 85, 427. D.D. Bosshardt, H.E. Schroeder, Cell Tissue Res. 1992, 267, 321. M.I. Cho, P.R. Garant, J. Periodontal. Res. 1988, 23, 268. T.G. Diekwisch, Connect. Tissue Res. 2002, 43, 245. B.K.B. Berkovitz, R. Holland, N. Nagai, Oral Anatomy, Histology & Embryology, 3rd edn. Mosby Ltd., Edinburgh, 2002. R.L. MacNeil, M.J. Somerman, J. Periodontal. Res. 1993, 28, 550. H.F. Thomas, Int. J. Dev. Biol. 1995, 39, 231. D.D. Bosshardt, H.E. Schroeder, Anat. Rec. 1996, 245, 267. H.E. Schroeder, in: H.E. Schroeder (Ed.), Orale Strukturbiologie. Georg Thieme Verlag, Stuttgart, New York, 2000, p. 144. T.G. Diekwisch, Int. J. Dev. Biol. 2001, 45, 695. H.E. Schroeder, Schweiz Monatsschr. Zahnmed. 1993, 103, 550.

177

12 The Structure of Teeth: Human Enamel Crystal Structure Fre´de´ric Cuisinier and Colin Robinson

Abstract

In this chapter we review and compare data on enamel crystals obtained with high-resolution transmission electron microscopy (HRTEM) and atomic force microscopy (AFM). The combination of AFM and HRTEM images is useful when elucidating enamel crystal growth mechanisms, since AFM provides information about surface properties (morphology, chemistry, physics) while HRTEM permits the determination of crystallographic structure. From a comparison of these data it was possible to validate a crystal growth mechanism involving the fusion of smaller mineral subunits. It appears that such fusion occurs during the secretory stage of enamel formation, with the fusion lines being still visible in the crystals at the maturation stage. Key words: calcium phosphate, enamel crystal, crystal growth, atomic force microscopy (AFM), high-resolution electron microscopy (HREM).

12.1 Introduction

The detailed structure and development of enamel crystals – and indeed of crystals from all mineralized tissues – have been examined largely using highresolution electron microscopy. Such studies have produced a wealth of detailed information concerning not only overall crystal morphology but also structural information at the crystallographic level. Until recently, no other technologies have been capable of delivering information at this high degree of resolution, but with the development of scanning ﬁeld microscopies – and especially atomic force microscopy (AFM) – it has become possible to examine crystal surfaces at this level, with the added advantage of investigating surface chemistry. In this chapter we review and compare current data on enamel crystals obtained with high-resolution transmission electron microscopy (HRTEM) and Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

178

12 The Structure of Teeth: Human Enamel Crystal Structure

AFM. Moreover, by combining these data we have attempted to develop a proposal for a mechanism of crystal growth. The data have been drawn from investigations of both human and rodent (rat and mouse) tissues. Human enamel crystals were collected from the secretory stage of enamel formation and observed using HRTEM, while some mature human enamel has been observed using AFM. The rodent crystals were collected at both secretory and maturation stage and studied with AFM and chemical force microscopy (CFM).

12.2 HRTEM Observations

During the secretory stage of human enamel, two mineral forms have been observed using HRTEM, namely ribbon-like crystals [1–3] and nanometer-sized, roughly spherical, particles [4]. Judging from lattice fringes, the ribbon-like crystals have a structure which is close to that of hydroxyapatite (Fig. 12.1). They are, however, far from being stoichiometric, and present many structural defects, mainly as dislocations and – perhaps important for this study – grain boundaries [2, 3]. The ribbon-like morphologies were variable, and often exhibited considerable curvature [3], with surfaces characterized by the presence of nanoscale steps and irregularities similar to those demonstrated using AFM [5]. Their elongated shape was observed along many crystallographic zone axes, and was observable at both low and high magniﬁcation.

Fig. 12.1 Human enamel crystal observed along [103] zone axis. (A) General view. (B) Enlarged image with a calculated image of hydroxyapatite. (C) Enlargement of the boundary.

12.3 AFM Observations

Fig. 12.2 High-resolution electron microscopy image of two crystals fused by a large angle boundary (indicated by the black arrowheads).

Figure 12.1 illustrates the presence of elongated crystals along [103] zone axis. This crystal is composed of two grains with a boundary (indicated by the arrowhead), and each grain showing a regular and perfect crystallographic structure. The lengths of the two grains along the (100) planes are 25 and 35 nm respectively, with thicknesses of around 10 nm and 20 nm. A crystal observed along the [001] zone axis is shown in Figure 12.2. This crystal is also formed by two grains with a junction between. Each grain is observed along the [001] zone axis, and although each is extremely small they show the characteristic hexagonal shape of enamel crystals. The structure of the area between both grains (indicated by the two arrowheads) is complex, and is almost certainly due to the rather high disorientation of the two grains. In both cases it appears that the two grains fuse to form longer (Fig. 12.1) or wider (Fig. 12.2) particles. This is similar to the fusing of particles that is also observed during bone and dentine crystal growth [6–8]. In addition to the ribbon-like structures described above, HRTEM also revealed the presence of approximately 2 nm-sized particles close to the ribbon-like crystals in developing human enamel [3]. Such a mean diameter of 2 nm would not be detected using AFM, and consequently the fate and role of these particles remain unclear.

12.3 AFM Observations

Using both AFM and CFM, it was possible to show the classical crystal morphology of enamel. In addition, particles of 20 to 30 nm diameter were revealed at low pH within rat and human enamel crystals (Fig. 12.3). These particles were pre-

179

180

12 The Structure of Teeth: Human Enamel Crystal Structure

Fig. 12.3 (A) Height image of a crystal from the maturing stage of a rat incisor. A smooth featureless surface can be seen with some suggestion of banding. (B) Height image of a crystal from the maturing stage of a rat incisor, subjected to pH 4. The banded appearance can be seen with bands @30–50 nm in width; each band appears to comprise

spherical domains @30–40 nm in diameter. Arrows indicate bands and spherical domains. (C) Lateral force/friction image of the crystal in (B). Spherical domains exhibit a light color; that is, high friction indicating high positive polarity. The arrows indicate spherical positive domains.

Fig. 12.4 Suggested arrangement of spherical domains, within an enamel crystal. Spherical domains are either as stacked polygons (hexagons) or occupy a shallow spiral.

12.4 Discussion

sumably revealed by selective etching or polishing at grain boundaries [9], implying structural (if not chemical) discontinuity between them. The particles appeared to form six-membered hexagonal clusters or a shallow spiral, and it was suggested that these may have fused to form developing enamel crystals, as shown in Figure 12.4 [10]. It seems likely that these particles correspond to those identiﬁed in the HRTEM studies, their somewhat larger size most likely resulting from convolution artifacts that occur in lateral measurements made with AFM.

12.4 Discussion

Both HRTEM and AFM/CFM data support the view that enamel crystals form by the fusion of smaller subunits. These may represent initial mineral nuclei and may give rise to speciﬁc binding sites for modulating protein [9]. Whilst the particles observed using HRTEM during the secretory stage did not have the roughly spherical shape obtained using AFM, this diﬀerence may have been due to tip convolution and the fact that only the outer surfaces of the particles could be imaged. In this case, it would be more useful to compare the equidistance of line separation between spherical domains (ca. 25 nm) and the larger dimensions of the particles measured on HRTEM images. In support of a larger size, however, is the fact that mouse enamel crystals observed with AFM and CFM showed larger sizes, with a mean width of 50 nm. Changes in morphology may also be eﬀected by interaction with modulating proteins. For example, amelogenin can inﬂuence calcium phosphate crystals grown on ﬂuoroapatite (FAP) to have a spherical morphology and a diameter of 50 nm [11]. From a strict crystallographic point of view, growth mechanisms involving spherical particles (as observed with AFM) would be unlikely, as hydroxyapatite crystallizes in the hexagonal system to form prismatic crystals. The plate-like or ribbon-like crystal morphologies seen when using HRTEM may result from ionic substitutions, while extracellular matrix organic components by, for example adsorption or complexation, can also modify the morphology of the crystals. However, spherical structures have not been observed using HRTEM. It seems likely therefore that the true morphology of initial mineral particles might more closely resemble that observed using HRTEM (see Fig. 12.2). This would not preclude pre-crystalline particles with only a short range order and a roughly spherical morphology. The proposed mechanism involving the fusion of smaller crystallites could also explain the paradox of enamel crystal formation, in that the earliest transmission electron microscopy studies showed the ﬁrst crystals observed to be ribbon-like in nature. However it is possible that, due to the low resolution of the microscope used, small crystals (20–30 nm) were not observed; rather, these were observed only after their fusion.

181

182

12 The Structure of Teeth: Human Enamel Crystal Structure

12.5 Conclusions

Both HRTEM and AFM/CFM have revealed the presence of regular, similarly sized, nanoscale subunit structures in enamel crystals, not only in mature tissue but also during development. Crystal formation appears to involve the fusion of these structures to generate the long ribbon-like crystals characteristic of enamel. They may represent original initiation sites for mineral precipitation. In fundamental terms, the development of biological crystals by the fusion of extremely small subunits provides a degree of control over ﬁnal size, morphology and disposition/location which is not easily available for crystal development by the acquisition of ions from solution.

References 1 F.J.G. Cuisinier, P. Steuer, J.C.

2

3

4

5

6

Voegel, R.M. Frank, J. Biol. Buccale 1990, 8, 149–154. F.J.G. Cuisinier, J.C. Voegel, J. Yacaman, R.M. Frank, J. Cryst. Growth 1992, 116, 314–318. F.J.G. Cuisinier, P. Steuer, B. Senger, J.C. Voegel, R.M. Frank, Calcif. Tissue Int. 1992, 51, 259–268. F.J.G. Cuisinier, P. Steuer, B. Senger, J.C. Voegel, R.M. Frank, Cell Tissue Res. 1993, 273, 175–182. J. Kirkham, S.J. Brookes, R.C. Shore, W.A. Bonass, D.A. Smith, M.L. Wallwork, C. Robinson, Connect. Tissue Res. 1998, 38, 89–100. P. Houlle, J.C. Voegel, P. Sschultz, F.J.G. Cuisinier, J. Dental Res. 1997, 76, 895–904.

7 F.J.G. Cuisinier, P. Steuer, A. Brisson,

8 9

10

11

J.C. Voegel, J. Cryst. Growth 1995, 156, 443–453. F.J.G. Cuisinier, Curr. Opinion Sol. State Mater. Sci. 1996, 1, 436–439. C. Robinson, K. Yamamoto, S.D. Connell, J. Kirkham, H. Nakagaki, A.D. Smith, Eur. J. Oral Sci. 2005, 114, 99–104. C. Robinson, S. Connell, J. Kirkham, R.C. Shore, A. Smith, J. Mater. Chem. 2004, 14, 2242–2248. S. Habelitz, A. Kullar, S.J. Marshall, P.K. DenBesten, M. Balooch, G.W. Marshall, W. Li, J. Dent. Res. 2004, 83, 698–702.

183

13 Design Strategies of Human Teeth: Biomechanical Adaptations Paul Zaslansky and Steve Weiner

Abstract

Human teeth are exquisitely designed to fulﬁll their essential functions in mastication. They are composed of complex hierarchical and graded structures that function together in an integrative manner. Although tooth materials, speciﬁcally enamel and dentin, have been well investigated, surprisingly little is known about the details of how whole teeth function mechanically when compressive loads are applied. Studies using strain gauges, photoelasticity and simulation methods show that the enamel cap acts as a stiﬀ, but deformable, body. More comprehensive mapping of the crowns of human premolars by electronic speckle patterncorrelation interferometry (ESPI) conﬁrm these observations and highlight details of the design strategies of the enamel cap. The crown not only deforms but also rotates under load. Much of the load is transferred onto a crucial interphase in dentin just below the dentino-enamel junction. This relatively soft zone compresses asymmetrically when loaded. The root also has asymmetric mechanical properties, that presumably reﬂect the manner in which it functions under stress. Many key questions remain to be addressed before a more complete understanding of the design strategies of whole teeth is obtained. Key words: human tooth function, whole-crown deformation, enamel cap, dentino-enamel junction (DEJ) interphase.

13.1 Introduction

Many animals produce teeth with working surfaces that are hardened in one way or another and a bulk that provides the essential tough support [1]. Although not all teeth are mineralized, this basic tooth design strategy is found in many vertebrates and invertebrates. In humans, the complexity of tooth structures [2] suggests that additional design features contribute to function. Such features are the Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

184

13 Design Strategies of Human Teeth: Biomechanical Adaptations

results of structural adaptations that allow human teeth to repeatedly and reliably fulﬁll their mechanical functions for many years. The adaptation is in both materials properties and three-dimensional (3-D) design, although the signiﬁcance of these details is largely unknown. Indeed, much remains to be discovered about the ways in which natural teeth respond to daily physiological loading. The major objective of this chapter is to consider some principles of the biomechanical adaptations of whole human teeth. These teeth have hierarchical and graded structures [3, 4] that have evolved into eﬃcient tools for mastication [5]. They outperform all currently available man-made dental replacements, be-it restorations, crowns, or implants. Teeth diﬀer in shape and size according to their position in the jaw, where they normally function as a group of cutting and grinding devices [6]. Primarily, they fulﬁll a function that is essential for survival, namely food procurement and processing [7]; consequently, evolutionary pressure has compelled teeth to be ﬁnely tuned to their function. An understanding of the structure–function relationship of teeth [8] is, therefore, both fascinating and important. The fact that enamel and dentin possess graded variations in mechanical properties was ﬁrst reported during the late 1950s. Microhardness proﬁles from the outer surface of enamel into dentin revealed an increase and then a decrease in enamel hardness, followed by a signiﬁcant decrease in hardness of dentin just beneath the dentino-enamel junction (DEJ) [9]. Hardness, as well as other elastic and plastic characteristics (e.g., stiﬀness, strength and fracture toughness), have been researched extensively during the past century [8, 10]. Earlier studies sought single accurate estimates of properties such as modulus, proportion limit, and so forth. However, the materials in teeth are graded, and the results of recent studies have suggested that a range of measurements more accurately reﬂects the properties of both enamel [4] and dentin [11]. Consequently, estimates of modulus in the range of 50@120 GPa for enamel or 5@30 GPa for dentin could all be correct. There are, therefore, no unique elastic constants such as the modulus of enamel or dentin. The large variation in properties is related to the gradual but substantial variations in microstructure found within distances of several hundreds of micrometers. Thus, structural complexity and the challenge of analyzing small sample sizes make human teeth inherently problematic to study. The overall approach to studying the function of teeth has, for the most part, followed the traditional materials research approach, namely measuring the materials properties and then using mathematical models and simulations to understand the behavior of the whole tooth. With the development of numerical simulation methods for dental research [12], the elastic properties (moduli and Poisson’s ratios) have often been used for modeling by the ﬁnite element and similar methods. While these studies provide important insights, the approach of modeling teeth in function is inherently limited, because only simpliﬁed models of the complex structure are created. Hence, it is necessary to make direct measurements of tooth deformation under load in order to increase our understanding about the exquisite design of the whole tooth. The following section describes the relatively few data that exist on measurements of whole human teeth.

13.2 Deformation of Whole Teeth under Load

13.2 Deformation of Whole Teeth under Load

Teeth deform due to forces that develop during food mastication or contact with other teeth. The forces vary, depending upon the nature of the food being chewed [7], although axial compression predominates. Most studies of function of whole teeth report strain or stress due to load that is applied in a direction perpendicular to the cutting surface. Anderson [13] used strain gauges built into restorations of patients that were given food to consume, and succeeded in correlating in-vivo loads to various food types. He emphasized the calibration procedure for the accurate measurement of bite forces, and found low stresses (up to only 150 N) in the tooth, in contrast to earlier measurements. Neumann and DiSalvo [14] also tested the eﬀects of applying loads to teeth by measuring the shortening of whole teeth under compression. Their measurements were performed in vitro on teeth machined to have upper and lower parallel ﬂat surfaces. An increased compressibility of teeth was found to occur at low loads, and from the stress–strain curves these authors derived reasonable tooth crown modulus values of 80@110 GPa. Remarkably, these results fall within the range reported for isolated enamel. Neumann and DiSalvo [14] reasoned that the enamel cap deformed markedly at low loads, claiming that something in the enamel organization undergoes change during physiological loading conditions. Thus, they suggested that whole-tooth behavior was directly related to the interactions of the mineral/organic components, albeit in an unspeciﬁed way. Haines, Berry and Poole [15] recognized the importance of the interplay between dentin and enamel when whole teeth are compressed. Using extensometers, they studied the deformation of the enamel cap along and across whole crowns. Their results reproduced the tooth-shortening ﬁndings of Neuman and DiSalvo [14], while proving that the enamel cap becomes deformed in a nonlinear manner. Haines et al. [15] emphasized that the cap appeared to be substantially more compressible under small loads, and suggested that this occurs because of changes to the environment surrounding the mineralized crystals. They also hypothesized that volume shrinkage in the cap appears as deformation both along and across the crown. Using the method of photoelastic stress visualization developed by Mahler and Peyton [16], Hood [17] placed photoelastically active dental ﬁllings into class V cavities. Restorations of these cavities are commonly made near the lower buccal (gingival) margins of tooth crowns, far from the chewing edge of teeth (see Fig. 13.1 for tooth nomenclature deﬁnitions). Hood showed that there is substantial deformation of crowns of teeth loaded at the cusp tip, with stress developing at the outer lower edge of the enamel cap. During these experiments, teeth were subjected to extreme desiccation and heating (>100 C). Non-compressed teeth were used during control experiments, to prove that the ﬁndings were not due to experimental artifacts. Hood later documented crown deformation by additional methods, where teeth with and without restorations were observed [18], and noted that premolars expanded bucco-lingually when compressed on axis.

185

186

13 Design Strategies of Human Teeth: Biomechanical Adaptations

Fig. 13.1 Tooth nomenclature conventions. Schematic representations of (a) teeth relations in the human mouth and (b) tooth areas on and within any single tooth. The upper and lower jaws of humans have identical numbers and types of teeth: four incisors, two canines, four premolars, and six molar teeth (including wisdom teeth). Cheek/ lip sides of all teeth (black arrows) are termed facial, labial or buccal (B), and they reach down to the gums/gingiva (arrowheads, G). Contact areas between adjacent teeth closer to the midline are the

mesial aspects of teeth (M). Sides of teeth facing away from the midline are the distal aspects of teeth (D). The internal areas near the tongue are the lingual sides (lower jaw, L) or palatal sides of teeth (in the upper jaw). Dentin, which is known to house tubules radiating out from the pulp chamber/canal, forms the bulk of the tooth. It is lined with cementum in the root, and is covered by an enamel cap in the tooth crown. The dentinoenamel junction (DEJ) forms the interface between dentin and enamel in the crown.

Hood also noted a shortening of whole-tooth crowns subjected to load [19], as has been reported by earlier investigators [14]. Sakaguchi and co-workers [20] showed that the extent of deformation of buccal and lingual cusps varied signiﬁcantly, depending on where load was applied on the working surface. These authors concluded that the cusps deform independently from each other, by noting that deformations were not detected on the outer crown surfaces just millimeters away from the loaded cusps. This concept probably holds only to a limited extent, because intact cusps seem to support each other. Signiﬁcant deformation is found when whole-crown integrity is compromised, as seen for example in Hood’s ﬁndings of increased cusp ﬂexure when he produced cavities within intact teeth, gradually enlarging them [18]. The importance of structural integrity was also highlighted by Popowics et al. [21], who studied human and pig teeth loaded in compression. These authors considered fracture experiments important for the characterization of deformation response, and used their results to identify diﬀerences between tooth cusps. They suggested that vertical cracks in enamel appeared as a result of the reinforced microstruc-

13.2 Deformation of Whole Teeth under Load

ture that sustains higher loads in human teeth. Popowics et al. also emphasized the importance of the complete structure for normal tooth function. Water is clearly a key component in the functioning of whole teeth [8]. Fox [22] reproduced the whole-tooth deformation experiments of Haines et al. [15], and addressed the role that water plays in the enamel cap. Fox demonstrated the existence of hysteresis loops when employing cyclic measurements that involved loading and unloading of teeth, and hypothesized that the water content changed during compression of the cap – contributing to the toughness. Fox concluded that the tooth crown acts like a stiﬀ sponge. Paphangkorakit and Osborn [23] showed that water ﬂows out of the crown when low loads (<120 N) are applied, and determined this by directly measuring ﬂow out of the root canal. It was concluded that a volume change of the pulp chamber occurred due to compression and shortening of the entire crown under load. This may well be the case, considering that the main component of teeth is dentin that appears to deform viscoelastically. The viscoelastic properties of dentin [11, 24] may play an important role in the extensive deformation and recovery of the crown, but as yet the role water plays is not well understood [8]. In order to obtain a more comprehensive perspective of tooth crown deformation under compressive loads, we have measured the deformation of the natural surfaces of whole human premolars that were mechanically loaded so as to be directly comparable to physiologic mastication [25]. The displacements of many points on the surface of the crowns were monitored simultaneously by using an optical mapping technique termed electronic phase shifting speckle interferometry (ESPI) (Fig. 13.2a). Speckles of light appear due to interference between laser beams reﬂected from many points on the tooth surface [26] (Fig. 13.2b). All teeth, immersed in water, were illuminated repeatedly by three interferometers, each of which was composed of two beams split from a single laser source, aligned symmetrically at an angle to the surface along the X, Y, or Z axis [27]. The surface speckle patterns were recorded by a camera facing the tooth (see additional details in [28]). Surface displacements resulted in local shortening or lengthening of the optical paths, seen as variations in intensity. Optical path changes produce shifts of the speckle phases. The phase-shifts were numerically quantiﬁed for all points on the image of the crown [28], and displacement maps were obtained (see typical example in Fig. 13.2c). In this way, speckles were used to track sub-micron surface deformations in three orthogonal directions. An example of a vector map combining X and Y displacement ﬁelds is shown in Figure 13.2d. Stresses were applied in two ways, in order to measure the response of premolar crowns to loads commonly produced during everyday mastication. They were either localized around the tip of the highest (main) cusp for simulating the mastication of hard food, or applied over most of the upper portion of the main cusp for simulating the mastication of soft food. Small increments of stress were applied in a cumulative manner, and maps of the orthogonal (3-D) displacements were obtained after each increment (a typical example of tooth crown deformation with increasing load is shown in Fig. 13.4). Micrometer-scale relative displacement magnitudes (RDMs) were calculated: RDMs for each point represent

187

188

13 Design Strategies of Human Teeth: Biomechanical Adaptations

Fig. 13.2 Speckle interferometry deformation measurements. (a) Schematic diagram of a speckle interferometer used to determine surface displacements along the X, Y, and Z axes of a mechanically compressed tooth. F is the applied force. (b) Speckle ﬁeld formed by bright spots that are used to track displacements over areas of several micrometers squared on the tooth surface. (c) Displacement maps are obtained for each axis, as seen by the typical sub-micron values depicted in the gray-level image. The numerical estimates of displacement obtained by electronic speckle patterncorrelation interferometry (ESPI) are calculated relative to a reference point on the deforming surface, thus circumventing any global, rigid-body movement. (d) Vector map representation of the combined surface displacements along the X and Y directions. Although informative, such vector maps are problematic for whole-tooth deformation interpretation, because the directions and

sizes of the vectors are a function of the location of the reference point. This is shown schematically in Figure 13.3, where vector maps of identical displacement ﬁelds produced by an ideally compressed rectangle are shown using two diﬀerent reference points. To circumvent this limitation, the X, Y, and Z directional displacement values were reduced for every point to single estimates of the magnitude of displacement in space. This was done by adding the squares of the average of each component of displacement (three orthogonal directions) and calculating the square root. The new values represent the relative displacement magnitudes (RDMs) of each point, and they are no longer vectors [25]. Rather, they form new maps of RDMs of the diﬀerent areas on the surface revealing deformation patterns of the tooth, independent of a reference point. RDMs are scalars that merely indicate how much each point has moved in space, relative to the average of all other points on the tooth crown.

13.2 Deformation of Whole Teeth under Load

Fig. 13.3 Relative displacements in vector representation are aﬀected by the reference point location. (a) Surface displacements of the compressed rectangle represented as vectors relative to point near the top of the object, marked by a cross. (b) Same displacements as in (a), but drawn with reference to a point found on the lower side of the object. Note that the vector ﬁeld appears very diﬀerent, although the displacement ﬁeld is identical (as noted in legend to Fig. 13.2).

Fig. 13.4 Relative displacement magnitude (RDM) maps showing deformation of a premolar during incremental loading. When analyzing displacements at the various load levels (indicated by the force value adjacent to each ﬁgure), the relative movement of the crown surface in X, Y, and Z directions is

determined. These are combined into RDM maps [25]. RDM variations reveal characteristic deformation patterns of the whole crown in micrometers. The ﬁnal load level (105 N in this case) shows the greatest level of detail.

the average displacement derived from the three axes (see Figs. 13.4 and 13.5; see legend of Fig. 13.2, [25] and Fig. 13.3 for additional details). For comparison between the contributions of morphology versus structure, exact acrylic (polymethylmethacrylate; PMMA) replicas were tested in a similar manner [25]. Figure 13.5 represents a typical RDM map of a lower ﬁrst premolar, loaded at the tip of the main cusp. It shows, for example, that the cusp on which the load is applied moves relative to the main bulk of the tooth. The RDM values vary in two diﬀerent ways. The ﬁrst variation is an overall gradual decrease in values from the tip of the cusp down to the center of the crown. This shows that the enamel cap has rotated (down and to the left of the image shown in Fig. 13.5) relative to the central bulk of the crown. The center is found approximately midway between the buccal and lingual outermost points (points B and L in Fig. 13.5). High RDM levels are found at the lower edges of the crown just beneath the enamel margins.

189

190

13 Design Strategies of Human Teeth: Biomechanical Adaptations

Fig. 13.5 Typical RDM representation of deformation in a lower premolar under load. Load was applied to the tip of the tooth (arrow), which is seen to deform relative to the center of the crown [midway between the buccal (B) and lingual (L) aspects of the tooth]. The tooth cap is thus deforming; there

is also some rotation of the crown. This is seen as higher RDM levels at points away from the center of the crown (mainly on upper and lower margins of the RDM map). Note displacements are measured in micrometers, whereas tooth dimensions are millimeters.

It has been shown [25] that this indicates a certain degree of rotation relative to the root. Thus, the cap redistributes the load applied at the tip, which results in some rotation of the crown relative to the main tooth axis. The tooth cap must be bending with respect to the root. The second type of variation observed is in the relative distribution of RDM values within the bulk and the buccal side of the crown. Substantial deformation of the enamel cap is observed when noting that the upper part of the main cusp moves towards the buccal side, which hardly moves. RDM values are minimal (near B in Fig. 13.5), resembling those found at the center of the tooth crown and indicating little relative motion. Deformation is therefore seen to be distributed 3@4 mm away from the loaded tip. The RDM data thus show that the enamel cap both rotates and deforms under moderate load. Deformation of this nature might reveal details of the generic phenomenon that Haines et al. [15] and others had noted. These ﬁndings are consistent with the observations of the buccal and lingual cusps moving and deforming to diﬀerent extents when load was applied at various points on the working surfaces of intact teeth [20]. It should be noted that there is an area on the dentin surface just below the enamel cap that is deformed signiﬁcantly in the buccal area, but not as much on the lingual side. The enamel cap thus transfers some of the applied stress down to this region in an asymmetric manner. Dentists are well aware that known pathologies (such as class V lesions) often occur in this region of premolars [29]. The form associated with stress is termed non-carious cervical lesions or ‘‘abfractions’’ [30]. The present observations may provide a means of visualizing and quantifying the dynamics of formation of this pathology.

13.3 Mechanical Behavior of the Enamel Cap

RDM maps readily identify locations on the tooth surface where the smallest displacements occur. RDM map minima are invariably found in the region where the premolars are usually in contact with adjacent teeth. This is in contrast to acrylic replicas, in which the locations of the minimum RDM values are found to be very close to the geometric center of the crown. Thus, the location of minimum RDM displacement is shifted by the structure of the tooth. If premolars and other natural teeth undergo minimal deformation at their mesial or distal contact points with neighboring teeth (areas M and D in Fig. 13.1), deformation is likely to be channeled along a lingual–buccal axis. Teeth may therefore be designed to become least deformed in the direction of their neighbors so as to minimize the forces exerted against adjacent teeth. This design has the advantage of reducing potential damage such as cracking or fracture when chewing, because the teeth themselves are by far the hardest objects encountered during mastication [7]. The extent to which these ﬁndings are representative of the manner in which teeth other than premolars respond to loading must be considered. Preliminary measurements of upper 3rd molars show marked rotation of the crown, depending on the point of application of load. Little or no deformation appears to occur in the crowns under physiological loads. This response is diﬀerent from that observed for premolars, and it raises fascinating questions concerning the possibility of microstructural diﬀerences between these two teeth types. Whole-tooth studies of human premolars using ESPI highlight two paradigms: (i) the enamel cap is primarily a stiﬀ, but deformable body; and (ii) the overall load is transferred via the enamel cap in an asymmetric manner onto the underlying dentin. Additional questions also need to be considered regarding the overall performance of teeth: What is the contribution of the root to load distribution? How is the periodontal ligament that holds the tooth in the jaw involved? In the following sections we will address several aspects of these paradigms in more detail, and discuss broader notions regarding tooth performance.

13.3 Mechanical Behavior of the Enamel Cap

Enamel is stiﬀ and therefore undergoes only minimal deformation while transferring load to the underlying dentin. The key to the unusual properties of enamel lies in its unique structure, which consists of very long crystals of carbonated apatite (Dahllite) arranged in bundles. These bundles are progressively interwoven at higher hierarchical levels [31]. Craig and Peyton [32] demonstrated that hardness varies across the cap, while Heines et al. [15] noted that the enamel cap could not act as a stiﬀ rigid unit, and that it must become deformed during function. During the four decades that followed these studies, many investigators had studied enamel properties. Some results clearly demonstrated marked deformation of the enamel cap, as do our own measurements [25]. Most researchers however, have not reported or discussed the

191

192

13 Design Strategies of Human Teeth: Biomechanical Adaptations

signiﬁcance of the gradual change in properties. The application of new, highly resolved indentation techniques to the study of enamel [33–35] has provided reﬁned elastic and plastic measurements, but all seem to produce single or average values for enamel stiﬀness, hardness and toughness. Enamel structure varies in diﬀerent directions, and thus possesses structural anisotropy. Xu et al. [35] investigated mechanical anisotropy in enamel, relating it to the microstructure by indentation. They report toughness values approximately threefold lower along the enamel rod orientation than across the rods. Habelitz et al. [33], using nano-indentation with much higher resolution, found enamel to be only moderately anisotropic with moduli of 87:5 G 2:2 GPa and 72:2 G 4:5 GPa along and across the rods, respectively. White et al. [36] found a similar ratio of 1.4 for toughness across the rods as compared to the orthogonal direction, and emphasized the importance of protein and water to plasticize the enamel and toughen it. These authors were of the opinion that anisotropy was less important than previously believed [37]. The mechanical anisotropy, as well as the measured gradients in hardness, all point to the fact that there is no single ‘‘correct’’ value for any mechanical property of enamel. On the contrary, it could well be that the variations themselves are key properties that allow the whole structure to function for many years. In computer simulations of whole-tooth deformation, the use of single values enables simple models to be constructed, and these have provided some important insights into enamel cap function. Yettram et al. [38] and Goel et al. [39] used numerical models assuming diﬀerent constraints and boundary conditions, yet both groups concluded that high stresses occur at the margin of the tooth crown due to stresses ﬂowing around the cap and into dentin. The form of enamel around dentin and the enamel elastic properties were considered to be the crucial factors in these studies. Rees and Jacobsen [40] used ﬁnite element modeling to reﬁne the published values of modulus of enamel and dentin by numerically reproducing measurements of cusp deformation reported by Hood [18]. Thus, numerical simulations were in fact used to identify modulus values that simulated buccal and lingual cusp displacements similar to those determined by experimental measurements of loaded natural teeth [41]. The simulations reproduced wholetooth response to load by varying density and property parameters in the model, and subsequent numerical studies were successful in accounting for the structural anisotropy of enamel [41, 42]. Spears et al., for example [41], modeled anisotropy and showed that orientation of the enamel could explain the variations in modulus results reported in the literature. However, these authors noted that density variations (seen as the percentage volume of organic matrix in the mineral) had a greater inﬂuence on the outcome of the predicted stiﬀness. Thus, anisotropy alone – even when using much higher values than those reported by White et al. [36] and Habelitz et al. [33] – is not suﬃcient to explain property differences determined by diﬀerent experiments. The results obtained by simulations are clearly aﬀected by the level of microstructure detail that is incorporated into the models. To date, none of the simulations has fully reproduced the complex structure of the enamel cap, although an increased level of detail is being

13.3 Mechanical Behavior of the Enamel Cap

achieved [43, 44]. Modeling has produced clinically interesting predictions of deformation patterns in the cap, suggesting that high levels of stress must occur at the cervical regions [45]. Modeling was also able to account for tooth orientation studies: upper jaw molar teeth appear to be aligned so as to coincide with minimization of tensile stresses in enamel. Consequently, the distal molars in the maxilla are best oriented mesially and buccally as compared with anterior upper molars [46]. A signiﬁcant advance in understanding the enamel cap has been made by Cuy et al. [4], who measured structure and materials properties in the context of a graded design. These authors demonstrated localized variations of both hardness and stiﬀness of the enamel, and correlated these values with elemental analyses of the enamel constituents. They showed that a structural asymmetry exists between the palatal and buccal sides of upper molars, and that the enamel was found to be stiﬀer on the palatal surfaces of the enamel cap. Cuy et al. reasoned that this must be an adaptation of the cap microstructure speciﬁc to this tooth type, because upper jaw molars are routinely loaded on the innermost cusp [46]. They argued that such structural adaptation enables the teeth to withstand greater stress on the palatal cusps, known to be the supporting cusps during routine mastication [6]. The same group [4] also suggested that this is the result of higher concentrations of calcium and phosphate on the palatal side. These results are consistent with the measurements of whole-teeth deformation performed by ESPI, where greater deformation is observed on the buccal side of lower premolars. The lingual sides show even distributions of RDM values (see Fig. 13.5) indicative of uniform displacements and consequently little strain. Thus, the enamel cap appears to be asymmetrically structurally ﬁne-tuned, with stiﬀness varying from area to area. This might be one reason for the superior performance of teeth as longlasting cutting and grinding tools. The ESPI observations predict that the enamel cap should be stiﬀ enough to transfer loads applied to one cusp over the whole dentin support, while being tough enough not to incur irreversible damage when the cap becomes deformed. An experiment by Wood et al. [47] showed that, indeed, the enamel is capable of restricting and conﬁning the dentin so that it is attached to enamel in this way. Slices along and across tooth crowns were produced. Slices across the crown had enamel which formed a closed ring around a core of dentin. The slices were hydrated and dehydrated and, using Moire´ interferometry, the dentin was shown to be totally conﬁned by the enamel ring. If the ring was not intact, then the dentin could freely expand and contract. This shows that the enamel cap is capable of ‘‘straight-jacketing’’ dentin, suggesting that a tight transfer of load occurs when the enamel is compressed during mastication. Lucas [7] considers that teeth function as tools that are ﬁnely tuned to handle food which must be torn or fractured during mastication. He reasons that teeth are optimized for compression so that they generate internal tensile stresses and microdamage in the foods. Thus, teeth are not particularly sharp but rather they are designed to support the cusps during compression, so that the crushing-offood task can be carried out throughout the tooth/animal lifetime, without irre-

193

194

13 Design Strategies of Human Teeth: Biomechanical Adaptations

versible damage. Load must therefore be distributed from enamel into dentin, and the former must undergo some deformation that redistributes loads in the tooth bulk under elastic or viscoelastic conditions. The enamel cap can be very stiﬀ, but is able to deform without failing, provided that it is supported by dentin. The transfer of load into dentin is the subject of the next section.

13.4 The Role of Crown Dentin During Load Bearing

The enamel cap rests on dentin that forms the bulk of the tooth. In crown dentin there are at least three diﬀerent types of structural material (Fig. 13.6): (i) Highly mineralized and stiﬀ peritubular dentin [48] composed of mainly carbonatedapatite crystals [49]. (ii) Intertubular dentin composed of about 20 wt% collagen, ca. 65 wt% carbonated apatite crystals, and water. (iii) Other constituents that have no direct known load-bearing function, such as tubule lumens, various tissue components and water-ﬁlled voids. In root dentin, the peritubular dentin is absent. Enamel is attached to the crown dentin via the DEJ, and it is now well established that this junction is more than a simple interface between two tooth materials [50]. The DEJ zone is a complex interphase where a variety of structural features are found in both enamel and dentin [51]. The structure of dentin in this zone diﬀers from the remaining crown dentin in terms of crystal organization, collagen ﬁber arrangement and orientation, variations in the amount and type of non-collagenous proteins, and the absence of peritubular dentin. Hard-

Fig. 13.6 Crown dentin microstructure. A typical scanning electron microscopy (SEM) image of a fracture surface (dehydrated) of crown dentin. While all dentin is structurally anisotropic, only crown dentin has an extra phase of highly mineralized peritubular linings surrounding the tubules (black

arrow). Functionally, however, anisotropy is minimal, appearing to be at most 10% [11, 56]. The extent to which this mechanical anisotropy is important for the longevity of natural and restored teeth is yet to be explored.

13.4 The Role of Crown Dentin During Load Bearing

ness measurements show a sharp minimum adjacent to the interface, with a gradual increase in the overall values further into the dentin [32]. This decrease corresponds to a decrease in intensity of backscattered electron microscopy signals [52]. This same pattern is evident from quantitative scanning microradiography of healthy teeth, where a minimum of mineral mass density is observed just beneath the DEJ [53]. The fact that this interphase zone fulﬁlls an important role during load bearing was ﬁrst directly demonstrated by Wang and Weiner [52]. Using Moire´ interferometry, these authors showed an increase in strain in a zone that was 200@ 300 mm thick beneath the enamel. This zone corresponds to the zone of softer dentin, originally reported by Craig and Peyton [32]. Although fringes of high strain in this region were observed by Hood [17, 18], they were discarded as being experimental artifacts, unrelated to deformation diﬀerences between enamel and the underlying dentin. Moire´ interferometry was used to map strains on slices of dentin [52], where Wang and Weiner showed that an asymmetry exists in the extent of sub-DEJ strain, comparing the buccal and lingual sides of the tooth. Wood et al. [47] reproduced these ﬁndings by measuring the strains produced during dehydration using high-resolution projected Moire´ interferometry. They found high levels of strain just beneath the enamel corresponding to the transition DEJ zone. Again, an asymmetry between the lingual and buccal sides was observed. Zaslansky et al. [54] used ESPI to directly measure the stiﬀness of dentin in the sub-DEJ soft zone, and showed that dentin on the buccal side of upper premolars beneath the enamel layer was signiﬁcantly softer than bulk dentin to a depth of approximately 300 mm, with a modulus averaging about 4 GPa, whereas on the lingual side the modulus was ca. 10 GPa. Thus, asymmetry is clearly an important aspect of the structural design of human premolar dentin. The bulk of the crown dentin has been more thoroughly investigated. It is characterized by peritubular lined tubules running from a few hundred microns beneath the DEJ all the way to the dental pulp. The peritubular dentin modulus is substantially higher than that of the intertubular dentin, and various values of stiﬀness ranging from 15 to 30 GPa have been reported [11]. Peritubular dentin is mostly absent in the DEJ zone, and this is probably one reason why it can act as a cushion or spring under load. The manner in which the bulk of the crown dentin contributes to strategies of load distribution is diﬃcult to measure experimentally, and to date mainly models have provided information about stress distribution. Invariably, crown dentin is modeled as an isotropic material, supporting the enamel cap and providing a tough, functionally homogeneous and isotropic backing substrate. It seems reasonable to predict, however, that an experimental detailed analysis of crown dentin might show that it is matched to the overlying soft sub-DEJ dentin and surrounding enamel cap, and that crown dentin too has an asymmetric microstructure and materials properties when comparing lingual to buccal sides. The tubules and associated peritubular dentin certainly make the material itself structurally anisotropic. Collagen ﬁbers appear to run approximately orthogonal to the tubule directions, bridging the intertubular dentin. They are thought to

195

196

13 Design Strategies of Human Teeth: Biomechanical Adaptations

assist in cross-stabilizing the highly directional tubules that radiate outwards form the pulp [55]. Early measurements reported crown dentin to be isotropic [8], although conﬂicting reports have emerged. Palamara et al. [56] measured deformation of grids that were sputtered onto the surfaces of tooth samples. These authors found that both elastic and fracture properties of crown dentin diﬀered by about 10% when observed along the direction of the tubules as compared with the orthogonal orientation; the modulus values they report however seem to be very low, with 10.7 GPa along the tubules and a higher 11.9 GPa across the tubules. Kinney et al. [11] report much greater values of modulus, and reasoned that previously performed indentation measurements could not detect the small extent of anisotropy. They subsequently used resonant ultrasound spectroscopy [57] and reported a small degree of anisotropy in the modulus values (23.2 GPa along the tubules versus 25.0 GPa across the tubules). Interestingly, such anisotropy seems elegantly to match and counteract enamel anisotropy: lower fracture toughness along the enamel rods and orthogonal to the DEJ and dentin [35] appear to be reinforced by dentin being stiﬀer in the plane parallel to the DEJ and orthogonal to the tubule orientation [11]. Nevertheless, the signiﬁcance of crown dentin anisotropy to whole tooth function is, as yet, poorly understood. When the structural variations of crown dentin are incorporated into numerical computer-simulation models, an improved match is obtained between the experimental Moire´ results [52] and the calculated stress–strain ﬁelds [58]. This has been shown for two-dimensional (2-D) models, and with newer and faster computers and advanced simulation software, increasing detail can be added to ﬁnite element simulations showing an improved ﬁt with experimental data [43]. With higher resolution of photoelastic methods [59], it is to be expected that an even better agreement between photoelastic simulation and real measurement results will be obtained. Crown dentin is physically continuous with root dentin. The latter is surrounded by cementum and attached by ﬁbers of the periodontal ligament to the surrounding bone. The following section discusses aspects of these important features.

13.5 The Role of the Root and Supporting Structures

Root dentin, although structurally continuous with crown dentin, appears to have signiﬁcantly diﬀerent characteristics. Root dentin has no peritubular dentin stiﬀening the structure internally, and has no stiﬀ enamel supporting it externally. Root dentin is composed only of intertubular dentin and tubules. Intertubular dentin in the root has the same structure as intertubular dentin in the crown, namely mineralized collagen ﬁbers aligned orthogonal to the tubule orientation [60]. The ﬁbers have a preferred orientation along the main axis of the tooth in the plane orthogonal to the tubules. Thus, whether the load is compressive or

13.5 The Role of the Root and Supporting Structures

tensile, the root is designed so that the stiﬀest structural orientation matches the load direction, which is along the long root axis. Notably, root dentin is surrounded by cementum that forms a soft attachment through the ﬁbers of the periodontal ligament (PDL) to the jaw (alveolar) bone. Experiments and simulations repeatedly ﬁnd that stress ﬂows around the cap and down into the root [12, 17, 38, 39, 52, 58, 59]. This leads to the conclusion that stress is concentrated below the cervical margins of the crown. Such stress concentration might arguably be an experimentally recurring artifact not necessarily occurring in teeth under physiological conditions, yet it clearly highlights the crucial role the supporting structures have for stress re-distribution in the root. Atmaram and Mohammed [61] state that stresses in the tooth are not affected by PDL properties, yet it appears that simulations of tooth function need to be performed with the supporting PDL and bone properly modeled [62] if they are to be of functional signiﬁcance. It has been shown [63] that current concepts of modeling the PDL generate rigid body displacements between the tooth and the bone, a movement that in reality must be highly non-linear. New forms of modeling are thus needed to account for the PDL performance and its signiﬁcance to whole-tooth function [63]. The roots of whole teeth loaded in isolation from the surrounding environment undergo signiﬁcant strain. However, in the oral cavity and due to the low stiﬀness of the PDL (0.05 GPa) [64], high levels of strain probably rarely occur. Rather, when low loads are applied (<20 N) [65] the tooth is pushed into the socket and compressed against the bone. Presumably, higher loads are then distributed from the root dentin directly into the surrounding bone. The issue of root dentin anisotropy has been addressed by several authors. By carefully cutting specimens of dentin parallel to the tooth growth incremental plane as well as along the other two orthogonal planes, Wang and Weiner [60] obtained the surprising result that in essence root (intertubular) dentin is isotropic, at least with respect to the manner in which it responds to indentation. This observation is consistent with other studies, such as those of Kinney and co-workers [66], who used nano-indentation to study both intertubular and peritubular dentin in the crown, and found an isotropy of the intertubular dentin. It should be noted that the dentin is not isotropic with regard to fracture [55]. The isotropic elastic nature of intertubular dentin might have major beneﬁts for stress redistribution. Multiple loading directions are possible, and any applied stresses can be redistributed in diﬀerent directions without compromising the tooth integrity. Although a preferred orientation of the ﬁbrils was found approximately along the root axis [60], this had almost no eﬀect on Vickers indentation values. The key structural features that account for the intertubular dentin isotropy are the more or less random arrangement of the ﬁbril bundles within the plane perpendicular to the tubules and no preferred orientation around the ﬁber axes [60]. Thus, it is likely that the root dentin can redistribute stress along the entire length of the root. The root is surrounded by an intermediate cementum layer that appears to be attached by a graded interphase [67] which might act as a stress-breaker or a spring.

197

198

13 Design Strategies of Human Teeth: Biomechanical Adaptations

An asymmetric distribution of mineral content around the axis of the root was reported [68], indicating that the root has more mineral on the bucco-lingual sides as compared with the mesial or distal sides of the root. The extent to which this characterizes the whole length of the root should be investigated further, because dentin has been found to be softer on the buccal and lingual sides of the root as compared with the mesial and distal sides [28]. High-sensitivity measurements were performed by ESPI where strain was optically determined on compressed samples taken from the buccal, lingual, mesial, and distal sides of the root. It appears, therefore, that the root is designed such that bending of the tooth is directed along a bucco-lingual direction, orthogonal to the dental arch line. Here, neighboring teeth are normally found. This bending appears to be asymmetric and buccally inclined. In premolars, the outer (buccal) side tends to be softer [28], but the buccal was found to be stiﬀer than the lingual in lower incisors [59]. Whether similar property diﬀerences occur in the surrounding PDL and bone tissues is unknown; however, the asymmetry of the root corresponds well with the characteristics of an asymmetrically deforming whole tooth, matched by the enamel cap and supporting DEJ regions. A great deal of what is known about the function of the PDL relies on tooth mobility studies related to orthodontic positioning of teeth in the mouth [65, 69]. Simulations have been used to reproduce tooth movement results [64] from which the PDL properties have been determined. The signiﬁcance of modeling PDL for accurate ﬁnite element studies of whole teeth is well known [62, 70]. It remains to be seen if indeed the PDL is really only signiﬁcant when small loads are applied [23], and to what extent this is important for tooth function. It is interesting to note that a PDL does not exist for tooth implants. Moreover, how important is the PDL? Could the inclusion of a soft, well-designed and properly matched interphase layer near bone be the key to long-term use of artiﬁcial implants serving as replacement parts? These are pertinent clinical questions aﬀecting the oral health of many current and future patients.

13.6 Broader Implications and Conclusions

Much remains to be determined with respect to the manner in which whole teeth deform during function. Here, we have described some key features that highlight the exquisite adaptation of the whole intact tooth. The mechanisms discussed as possible tooth responses to load are summarized in Figure 13.7 where, in addition to the well-documented intrusion (Fig. 13.7a), both deformation (Fig. 13.7b) and crown outward-rotation (Fig. 13.7c) are illustrated. Clearly, asymmetry is a key design strategy, and subtle but important diﬀerences are found in the microstructure of both dentin and enamel on the buccal and lingual tooth sides. The arrangement of materials at all length scales, ranging from the nanometer scale of mineral and organic components, through the micron scale of crystal prism orientations, peritubular and intertubular dentin and up to the meso scale of sub-

13.6 Broader Implications and Conclusions

Fig. 13.7 Possible deformation responses of teeth. The whole-tooth response is probably a combination of whole body movement (a), compression (b), and bending/rotation of the cap/crown (c). Thus, a combination of responses to load probably exists.

millimeter dimensions such as the soft sub-DEJ dentin, all appear to be ﬁnely matched with the challenges of tooth response to load. The focus of almost all studies is on the hard material components of the tooth, and the importance of the softer components in teeth are often overlooked. The fact that the pulp cavity is considered not to be important for load distribution [71] may need to be reconsidered. What are the eﬀects of variations in the size and location of this soft-tissue-ﬁlled void on the performance of the whole structure? How do changes known to occur in the pulp, such as the formation of reparatory dentin, aﬀect tooth function? What is the importance of pulp-cavity agerelated reductions in height and volume ratios [72]? Should these aﬀect treatment of young versus aged teeth?

199

200

13 Design Strategies of Human Teeth: Biomechanical Adaptations

The degree to which response to load in teeth is related to pathology is an issue under much debate [73]. The etiology and conditions of the formation of stressinduced cervical lesions for example, is still unknown. Decisions of when and how to restore teeth aﬀected by abfractions have only partial answers with both ﬁnancial and ethical consequences in terms of the obligations that dentists have to patients who wish to retain their teeth for as long as possible. Teeth thus appear to function as complex units, with all components contributing to the distribution of stress. The response to load appears to be non-linear and asymmetric, with parts of the crown deforming while the structure rotates. Several components found within teeth appear to function quite diﬀerently at low and high levels of load: the enamel cap, DEJ soft zone, root dentin and PDL are examples of complex tooth structures that exhibit the intricate interplay between soft tissue and mineral found in human teeth.

Acknowledgments

The authors wish to thank Prof. John Currey, Prof. Peter Fratzl, Prof. Asher A. Friesem, Dr. Ron Shahar, Dr. Meir Barak, and Dr. Netta Lev-Tov Chattah for helpful suggestions and discussions. S.W. is the incumbent of the Dr. Walter and Dr. Trude Burchardt Professorial Chair of Structural Biology. Support for this research was provided from grant RO1 DE006954 from the National Institute of Dental and Craniofacial Research to Dr. S. Weiner, Weizmann Institute of Science.

References 1 S. Weiner, P. Zaslansky, in: R.L. Reis,

S. Weiner (Eds.), Learning From Nature How to Design New Implantable Biomaterials: From Biomineralization Fundamentals to Biomimetic Materials and Processing Routes, Vol. 171. Kluwer Academic, Dodrecht, 2004, pp. 3–13. 2 A.R. Ten Cate, Oral Histology: Development, Structure and Function. Mosby-Year Book, Inc., St. Louis, 1994. 3 W. Tesch, N. Eidelman, P. Roschger, F. Goldenberg, K. Klaushofer, P. Fratzl, Calcif. Tissue Int. 2001, 69, 147–157. 4 J.L. Cuy, A.B. Mann, K.J. Livi, M.F. Teaford, T.P. Weihs, Arch. Oral Biol. 2002, 47, 281–291.

5 P. Dechow, D.S. Carlson, in: C.

6

7

8

9

McNeill (Ed.), Science and Practice of Occlusion. Quintessence Publishing, Chicago, 1997, pp. 3–22. G.D. Douglas, R.T. DeVreugd, in: C. McNeill (Ed.), Science and Practice of Occlusion. Quintessence Publishing, Chicago, 1997, pp. 69–78. P.W. Lucas, Dental Functional Morphology: How Teeth Work. Cambridge University Press, Cambridge, 2004. N. Waters, in: Symposia of the Society for Experimental Biology. Mechanical Properties of Biological Materials, Vol. 34. Cambridge University Press, London, 1980, pp. 99–135. R. Craig, P. Gehring, F. Peyton, J. Dent. Res. 1959, 38, 624–630.

References 10 M. Braden, in: Y. Kawamura (Ed.),

11

12 13 14 15 16 17 18 19 20

21

22 23 24 25

26

27

28

Physiology of Oral Tissues, Vol. 2. S. Karger, Basel, New York, 1976, pp. 1–37. J. Kinney, S. Marshall, G. Marshall, Crit. Rev. Oral Biol. Med. 2003, 14, 13–29. R.W. Thresher, G.E. Saito, J. Biomech. 1973, 6, 443–449. D.J. Anderson, J. Dent. Res. 1956, 35, 671–673. H.H. Neumann, N.A. DiSalvo, J. Dent. Res. 1957, 36, 286–290. D.J. Haines, D.C. Berry, D.F.G. Poole, J. Dent. Res. 1963, 42, 885–888. D.B. Mahler, F.A. Peyton, J. Dent. Res. 1955, 34, 831–838. J.A. Hood, N. Z. Dent. J. 1972, 68, 116–131. J.A. Hood, Int. Dent. J. 1991, 41, 25– 32. J.A. Hood, J. Dent. Res. 1968, 47, 1030. R.L. Sakaguchi, E.W. Brust, M. Cross, R. Delong, W.H. Douglas, Dent. Mater. 1991, 7, 186–190. (a) T.E. Popowics, J.M. Rensberger, S.W. Herring, Arch. Oral Biol. 2004, 49, 595–605; (b) T.E. Popowics, J.M. Rensberger, S.W. Herring, Arch. Oral Biol. 2001, 46, 1–12. P.G. Fox, J. Mater. Sci. 1980, 15, 3113–3121. J. Paphangkorakit, J.W. Osborn, Arch. Oral Biol. 2000, 45, 1033–1041. M.G. Duncanson, E. Korostoﬀ, J. Dent. Res. 1975, 54, 1207–1212. P. Zaslansky, R. Shahar, A.A. Friesem, S. Weiner, Adv. Funct. Mater. 2006, 16, 1925. R. Jones, C. Wykes, in: P.L. Knight, W.J. Firth, S.D. Smith (Eds.), Electronic Speckle Pattern Correlation Interferometry, 2nd edn. Cambridge University Press, Cambridge, 1989, pp. 1–63. R. Jones, C. Wykes, in: P.L. Knight, W.J. Firth, S.D. Smith (Eds.), Electronic Speckle Pattern Correlation Interferometry, 2nd edn. Cambridge University Press, Cambridge, 1989, pp. 165–196. P. Zaslansky, J.D. Currey, A.A. Friesem, S. Weiner, J. Biomed. Opt. 2005, 10, 024020.

29 W.C. Lee, W.S. Eakle, J. Prosthet. Dent.

1996, 75, 487–494. 30 (a) J.O. Grippo, J. Esthet. Dent. 1991,

31

32 33

34

35

36

37 38

39 40 41

42 43

44 45

46 47

48

3, 14–19; (b) J.O. Grippo, M. Siming, S. Schreiner, J. Am. Dent. Assoc. 2004, 135, 1109–1118. A. Boyde, in: M. Stack, R. Fearnhead (Eds.), Tooth Enamel. J. Wright & Sons, Bristol, 1965, pp. 163–167. R.G. Craig, F.A. Peyton, J. Dent. Res. 1958, 37, 661–668. S. Habelitz, S.J. Marshall, G.W. Marshall, M. Balooch, Arch. Oral Biol. 2001, 46, 173–183. R. Hassan, A.A. Caputo, R.F. Bunshah, J. Dent. Res. 1981, 60, 820– 827. H.H.H. Xu, D.T. Smith, S. Jahanmir, E. Romberg, J.R. Kelly, V.P. Thompson, E.D. Rekow, J. Dent. Res. 1998, 77, 472–480. S.N. White, W. Luo, M.L. Paine, H. Fong, M. Sarikaya, M.L. Snead, J. Dent. Res. 2001, 80, 321–326. S. Lees, F.R. Rollins, J. Biomech. 1972, 15, 557–566. A.L. Yettram, K.W.J. Wright, H.M. Pickard, J. Dent. Res. 1976, 55, 1004– 1011. V.K. Goel, S.C. Khera, K. Singh, J. Prosthet. Dent. 1990, 64, 446–454. J.S. Rees, P.H. Jacobsen, Clin. Mater. 1993, 14, 35–39. I.R. Spears, R. Vannoort, R.H. Crompton, G.E. Cardew, I.C. Howard, J. Dent. Res. 1993, 72, 1526–1531. J.S. Rees, P.H. Jacobsen, J. Oral Rehabil. 1995, 22, 451–454. B. Dejak, A. Mlotkowski, M. Romanowicz, J. Prosthet. Dent. 2005, 94, 520–529. G.A. Macho, Y. Jiang, I.R. Spears, J. Hum. Evol. 2003, 45, 81–90. D. Palamara, J.E.A. Palamara, M.J. Tyas, H.H. Messer, Dent. Mater. 2000, 16, 412–419. I.R. Spears, G.A. Macho, Am. J. Phys. Anthropol. 1998, 106, 467–482. J.D. Wood, R.Z. Wang, S. Weiner, D.H. Pashley, Dent. Mater. 2003, 19, 159–166. J. Kinney, M. Balooch, S. Marshall, G. Marshall, T. Weihs, Arch. Oral Biol. 1996, 41, 9–13.

201

202

13 Design Strategies of Human Teeth: Biomechanical Adaptations 49 S.J. Jones, A. Boyde, in: A. Linde

50

51

52 53

54

55 56

57

58 59 60

(Ed.), Dentin and Dentinogenesis, Vol. 1. CRC Press, Boca Raton, 1984, pp. 81–134. S.N. White, M.L. Paine, W. Lou, M. Sarikaya, H. Fong, Z. Yu, Z.C. Li, M.L. Snead, J. Am. Chem. Soc. 2000, 83, 238–240. S. Weiner, F. Nudelman, E. Sone, P. Zaslansky, L. Addadi, Biointerphases 2006, 1, 12–14. R. Wang, S. Weiner, J. Biomech. 1998, 31, 135–141. J.C. Elliott, P. Anderson, X.J. Gao, F.S.L. Wong, G.R. Davis, S.E.P. Dowker, J. X-Ray Sci. Technol. 1994, 4, 102–117. P. Zaslansky, A.A. Friesem, S. Weiner, J. Struct. Biol. 2006, 153, 188–199. R.Z. Wang, Dent. Mater. 2005, 21, 429–436. J.E.A. Palamara, P.R. Wilson, C.D.L. Thomas, H.H. Messer, J. Dent. 2000, 28, 141–146. J.H. Kinney, J.R. Gladden, G.W. Marshall, S.J. Marshall, J.H. So, J.D. Maynard, J. Biomech. 2004, 37, 437– 441. B. Huo, J. Biomech. 2005, 38, 587– 594. A. Kishen, A. Asundi, J. Biomed. Opt. 2005, 10, 034010. R.Z. Wang, S. Weiner, Connect. Tissue Res. 1998, 39, 269–279.

61 G.H. Atmaram, H. Mohammed,

J. Dent. Res. 1981, 60, 873–877. 62 T. Kuroe, H. Itoh, A.A. Caputo,

J. Dent. Res. 1998, 77, 498. 63 E.B. DeLaCasas, T.P.M. Cornacchia,

64 65 66

67

68

69 70 71

72 73

P.H. Guouvea, C.A. Cimin, Comput. Methods Biomech. Biomed. Eng. 2003, 6, 65–73. J.S. Rees, P.H. Jacobsen, Biomaterials 1997, 18, 995–999. D.C.A. Picton, Arch. Oral Biol. 1963, 8, 109–118. J. Kinney, M. Balooch, G.W. Marshall, S.J. Marshall, Arch. Oral Biol. 1999, 44, 813–822. S.P. Ho, M. Balooch, S.J. Marshall, G.W. Marshall, J. Biomed. Mater. Res. Part A 2004, 70A, 480–489. A. Kishen, U. Ramamurty, A. Asundi, J. Biomed. Mater. Res. 2000, 51, 650– 659. G.J. Parﬁtt, J. Dent. Res. 1960, 39, 608–618. J.S. Rees, J. Oral. Rehabil. 2001, 28, 425–432. C. Rubin, N. Krishnamurthy, E. Capilouto, H. Yi, J. Dent. Res. 1983, 62, 82–86. T. Oi, H. Saka, Y. Ide, Int. Endod. J. 2004, 37, 46–51. L.A. Litonjua, S. Andreana, A.K. Patra, R.E. Cohen, Biomed. Mater. Eng. 2004, 14, 311–321.

203

14 Clinical Aspects of Tooth Diseases and their Treatment Peter Ga¨ngler and Wolfgang H. Arnold

Abstract

From an evolutionary point of view, teeth are one of the earliest biomineralization structures of vertebrates. Teeth are unique and diﬀerent from any other biomineralization outcome of vertebrates. This is reﬂected in the specialized human odontogenesis and periodontogenesis and in developmental anomalies, in the clinical appearance and susceptibility to infectious diseases such as dental caries and periodontitis, and ﬁnally in the regeneration pattern after traumatic injuries. All pathological features developing during the natural history of dental caries and periodontal diseases, and also the regenerative patterns, are perturbations of the intrinsically conserved pattern of mineralized tissue reactivity. Clinical aspects of tooth diseases and their treatment are characterized by: biomineralization, demineralization, remineralization, resorption, and apposition. Hydroxyapatite biomineralization was a major step forward in evolution. The high remodeling potential of hydroxyapatite plays a major role in disease initiation and progression and in the treatment strategies. Key words: teeth, evolution, demineralization, remineralization, hard-tissue resorption, hard-tissue apposition, dental caries, periodontal diseases, developmental anomalies, dental trauma.

14.1 Introduction

According to newer data emerging from fossil records, teeth have either evolved from conodont elements or from denticles of the odontode type in the earliest agnathic chordates, before the dermal armor and tooth-like placoid scales appeared during the evolution of ﬁshes [1, 2]. This projects the evolutionary origin of teeth back to a point some 50 million years earlier than previously thought [3]. Therefore, a new developmental model has been proposed on the basis of the separaHandbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

204

14 Clinical Aspects of Tooth Diseases and their Treatment

tion of odontogenic from osteogenic and chondrogenic neural crest potential [4]. According to this modular development, each interactive morphogenic unit may be committed to form either cartilage, bone, odontodes, or teeth. From this evolutionary point of view, teeth and later periodontal tooth attachments are one of the earliest biomineralization structures of vertebrates. The highly complex mineralization process, and the diﬀerent structures of teeth in terms of enamel, enameloid, dentin, predentin and periodontal tooth-to-bone attachment, vary widely throughout evolution (Fig. 14.1). Teeth are indeed unique, and diﬀer from any other biomineralization outcome of vertebrates – a point which is reﬂected in the specialized human odontogenesis and periodontogenesis and in developmental anomalies, in the clinical appearance and susceptibility to infectious diseases such as dental caries and periodontitis, and ﬁnally in the regeneration pattern after traumatic injuries [5]. Because of the ‘‘phylogenetic memory’’, and of genetic information remaining quiescent in the genome, perturbations of mineralized tissue interactions can alter gene expression. Therefore, all pathological features which have developed during the natural history of dental caries and periodontal diseases, as well as the regenerative patterns, are perturbations of the intrinsically conserved pattern of mineralized tissue reactivity. Pathologic host responses of the connective tissue (mineralized as well as soft tissues) to infection and trauma do not represent a new quality; rather, they are simply ‘‘repeating’’ earlier phylogenetic reaction patterns which are normal biomineralization features of our ancestors from ﬁshes and reptiles to mammals. Consequently, clinical aspects of tooth diseases and their treatment are characterized by: Biomineralization, as the normal or disturbed developmental process of tooth formation and attachment (odontogenesis and periodontogenesis) of the human semi-diphyodont dentition. Ectopic biomineralization, represented by dental calculus formation as a common host response to accumulating bioﬁlms (plaque) on tooth surfaces, and by the formation of displaced crown cementum. Demineralization, as the reversible or irreversible mineral loss due to bacterial metabolism in bioﬁlms. Remineralization, represented by the uptake of calcium and phosphate ions in demineralized subsurface lesions of mainly enamel and, to lesser extent, of dentin and cementum. Resorption of mineralized structures (enamel, dentin, cementum, alveolar bone) by clastic cell activity (dentinoclasts, cementoclasts, osteoclasts, external granuloma). Apposition of hard tissues exceeding the normal life-long formation of secondary dentin, cementum, and the remodeling of alveolar bone due to the slow, but continuously

14.1 Introduction

Fig. 14.1 Phylogenetic development from homodont to heterodont teeth. (a) Conodont tooth of reptile (crocodile); (b) heterodont molar of herbivorous mammal (horse, functional wear) with exposed dentin, deep enamel ﬁssures and crown cementum; (c) heterodont molar of carnivorous mammal (wolf, no functional wear) with no exposed

dentin and very small pulp chamber; (d) permanent growing incisors and slow but continuous erupting molars of rodent mammal (beaver); (e) heterodont molars of omnivorous mammal (adolescent human, with no wear); (f ) heterodont premolar and molars of omnivorous mammal (adult human).

205

206

14 Clinical Aspects of Tooth Diseases and their Treatment

erupting teeth. According to this terminology, apposition takes place during the formation of irritation dentin and denticles, the de novo formation of cementum and alveolar bone, hypercementosis, and ankylosis of teeth [5, 6]. The unique nature of mineralized dental and periodontal hard tissues and structures is further attributed to the cellular origin: Enamel produced by ameloblasts disappearing during tooth eruption and, therefore, with no cellular regeneration. Dentin produced by odontoblasts, the only cell which plays a life-long role in biomineralization, unipolar with centripedal matrix secretion. Irritation dentin produced by either (much) fewer numbers of ‘‘eternal’’ primary odontoblasts, or by transformed stem cells into secondary odontoblasts to mineralize the pulp chamber in response to wear and to slow, but continuous, eruption. Cementum, divided into acellular-aﬁbrillar cementum (crown cementum), acellular-ﬁbrillar cementum, and cellular-ﬁbrillar cementum. Alveolar bone with a normal remodeling rate according to functional occlusion and mastication. Because of the abundance of genetic information (some quiescent, some active), and of the composite structure of diﬀerent mineralized tissues of very diﬀerent cellular ectodermal, mesectodermal and mesodermal origins, and, ﬁnally, because of the innate and acquired host responses to the bacterial infection of dental caries and periodontal inﬂammation, the pathogenesis of this mainly life-long disease exhibits diﬀerent clinical features during longlasting stagnating phases and bursts of activity. This is characterized by the concept of progression and stagnation [7]. The lack of regeneration of cavitated enamel and dentin is the reason for restoration with alloplastic biomaterials. To date, all of these materials are far from being optimal replacement biomaterial substitutes. An ideal restoration material would be a composite matrix catcher of calcium and phosphate ions from supersaturated saliva, mimicking the mineralization of hydroxyapatite.

14.2 Tooth Development and Developmental Anomalies

Tooth development is the most complex biomineralization process, and diﬀers from any other mineral formation in vertebrates. Teeth can undergo cellmediated resorption or external demineralization, and are either stable until they are shed (deciduous teeth) or they function for a complete life-time (permanent teeth).

14.2 Tooth Development and Developmental Anomalies

14.2.1 Developmental Features and Elemental Analysis of Early Mineralization

Tooth development starts with an infolding of the oral epithelium into the underlying mesectodermal mesenchyme, forming the dental lamina with tooth buds. A strictly genetically determined cascade of diﬀerent induction mechanisms between the epithelial and mesenchymal cells leads to condensation of the mesenchyme with subsequent invagination of the epithelium of the tooth buds and formation of the inner and outer enamel epithelium (Fig. 14.2a–c) [8]. Dentin is the product of odontoblasts which derive from the mesenchymal cells adjacent to the inner enamel epithelium. These diﬀerentiate into pre-odontoblasts, which begin to secrete the predentin matrix. Two diﬀerent types of matrix proteins have been distinguished: (i) collagenous proteins of which collagen type I is the most prominent in mineralized tissues; and (ii) non-collagenous proteins such as proteoglycans and phosphoproteins, which are believed to regulate the biomineralization process [9]. The collagenous proteins, which are secreted near the cell body of the diﬀerentiating odontoblasts, build up the scaﬀold for the subsequent crystallization of hydroxyapatite. While further secreting collagen, the apical cell process of the odontoblast elongates and forms the odontoblast process, with its diﬀerent functions. Adjacent to the cell body, collagen secretion persists, whilst in the middle part it is mainly proteoglycans that are produced; dentin-phosphoproteins and matrix vesicles are secreted in the apical part of the odontoblast process [10]. The secretion of collagen near to the cell body results in an unstructured meshwork of collagen ﬁbers. Proteoglycans convert the collagen ﬁbers into a structured threedimensional (3-D) meshwork to which dentin-phosphoproteins are bound. Because of their electronegativity, dentin-phosphoproteins have a high aﬃnity for calcium ions, which leads either to direct nucleation of hydroxyapatite or to the binding of nanocrystals of hydoxyapatite to the collagen ﬁbers, which have been formed in matrix vesicles [11]. The phospholipid membrane of the matrix vesicles, which are secreted in the apical part of the odontoblast process, contains annexin V [12, 13], a calcium channel which increases the Ca 2þ concentration within the matrix vesicles and which, together with the Na-dependent phosphate transporter, leads to nucleation of hydroxyapatite nanocrystals in the matrix vesicles [14]. Further growth of the nanocrystals leads to destruction of the phospholipids membrane, whereupon the hydroxyapatite nanocrystals are liberated from the matrix vesicles. The hydroxyapatite crystals are attached to the collagen ﬁbers, and this results in the typical composite structure of dentin containing organic ﬁbers and inorganic hydroxyapatite crystals. These molecular biological mechanisms explain the relatively low calcium and phosphorus contents in predentin close to the cell body of the odontoblasts, and the rapid increase in the calcium and phosphorus content with a sharp mineralization zone towards the predentin-dentin border (Fig. 14.2d,e) [15]. Enamel is the product of ameloblasts that are secreting matrix proteins directly, without producing collagen and matrix vesicles. The main matrix proteins responsible for hydroxyapatite crystallization of enamel are amelogenin and enamelin [16]. However, these matrix proteins are cleaved by proteinases which are also

207

208

14 Clinical Aspects of Tooth Diseases and their Treatment

Fig. 14.2 Three-dimensional (3-D) features of human tooth development and element analysis of early mineralization. (a) Histological section of a developing tooth in the bell stage, depicting the outer and inner enamel epithelium from which enamel will originate, and condensed mesenchyme from which the pulp tissue and odontoblasts will develop. (b) 3D-reconstruction of the tooth anlage of a frontal tooth showing already mineralized enamel (blue), mineralized

dentin (pink) underneath enamel and the extension of the enamel organ (gray transparent). (c) Overlay picture of scanning electron micrograph of mineralized dentin and predentin and element mapping of Ca. The rapid increase in the content of Ca in predentin towards mineralizing dentin is documented. (d) 3D-reconstruction of four tooth anlage in a human mandible with the enamel organ (gray transparent) and mineralizing enamel (blue) and dentin (pink).

14.2 Tooth Development and Developmental Anomalies

Fig. 14.2 (e) Energy-dispersive X-ray (EDX) line scan of predentin and mineralizing dentin of a human tooth bud. The relative carbon content is decreasing towards mineralizing dentin due to the increase of the Ca and P content. In the mineralizing zone there is a steep increase in the Ca and P contents.

secreted by ameloblasts and resorbed by the ameloblast processes [16, 17]. Thus, two diﬀerent stages of enamel production can de diﬀerentiated: (i) the secretory period, where amelogenin is processed by metalloproteinases and the crystallization of enamel prisms is initiated; and (ii) the maturation period, where the activity of serine proteinases rapidly degrades the matrix to very small peptides and amino acids, which are eventually resorbed by ameloblasts [18]; the enamel prisms subsequently grow by further crystallization of hydroxyapatite. Because of the rapid degradation of matrix proteins once formed, enamel crystals cannot be remodeled, and the orientation of the crystals is determined by the orientation of the ameloblasts [19]. Following the initiation of hydroxyapatite crystallization, and cleavage and resorption of the matrix proteins, the enamel prisms undergo maturation with a constant increase in mineralization. Enamel is free of collagen ﬁbers and is composed of 95% hydroxyapatite crystals, 4% water, and only 1% organic material.

209

210

14 Clinical Aspects of Tooth Diseases and their Treatment

14.2.2 Developmental Anomalies

According to the diﬀerent stages of tooth development, four classes of various aberrations determine the clinical picture: Class 1 anomalies of the number of teeth develop very early, and are strictly of hereditary origin. Class 2 anomalies develop later as a consequence of disturbed folding during the bell stage (Fig. 14.3), according to the invagination theory [20]; these malformed teeth are also of hereditary origin. Class 3 anomalies represent some very rare (of genetic origin) and some other very common anomalies of mineralization (with metabolic or traumatic etiology) (Fig. 14.4).

Fig. 14.3 Class 2 anomaly of premolarization of the human incisor 12 due to excessive invagination during the bell stage.

Fig. 14.4 Class 3 anomaly of enamel hypoplasia with aplastic areas and exposed dentin due to neonatal hypoxia during the mineralization of permanent incisors.

Dentin dysplasia Odontogenesis hereditaria imperfecta

Molarization (of premolars) Gemination or confusion (of tooth buds) Internal invagination or (rare) evagination (Dens invaginatus, sive evaginatus) Root malformations

Hypoplasias (very common due to disturbances in mineral metabolism, trauma, inﬂammation, speciﬁc infections, ﬂuoride intoxication)

Dentinogenesis hereditaria imperfecta

Premolarization (of canines and incisors)

Hyperdontia (supernumerous teeth, mostly malformed, emboliform-like, up to 3.5%)

Hypodontia (less teeth, mainly wisdom teeth up to 35%, premolars and incisors up to 5%)

Crowding

Amelogenesis hereditaria imperfecta (hypoplastic, hypomaturated, and hypomineralized enamel)

Invagination irregularities (wrong folding during bell stage)

Anodontia (no teeth, very rare)

Functional irregularities (dry mouth, bruxism, temporomandibular joint disorders)

Face malformations

Mastication disturbances

Non-occlusion

Ectopic positions

Eruption Class 4 anomalies of position of teeth

Stage of mineralization Class 3 anomalies of mineralization of teeth

Bell stage Class 2 anomalies of form of teeth

Bud stage Class 1 anomalies of number of teeth

Table 14.1 Developmental tooth malformations according to the formation stages, and their clinical appearance.

14.2 Tooth Development and Developmental Anomalies 211

212

14 Clinical Aspects of Tooth Diseases and their Treatment

Class 4 anomalies develop during the latest stage of tooth eruption; these irregular positions of teeth are of genetic origin and/or due to local factors.

Except for class 4 malformations, all other anomalies develop long before tooth eruption, and will never disappear (see Table 14.1).

14.3 Dental Caries

Dental caries is an alimentary modiﬁed, polybacterial non-speciﬁc infectious disease with multifactorial etiology. It is caused by many diﬀerent strains of bioﬁlm bacteria. However, the initiation and further development in longlasting periods of stagnation and bursts of progression (Fig. 14.5) depend on interactions of microorganisms, as well as on host responses of the macroorganism. The disease progression seen from the enamel and pulpal side is mainly characterized by alternating processes of demineralization and remineralization (Fig. 14.6). Under

Fig. 14.5 3D-reconstruction of approximal caries lesions in deciduous molars. (a) Three small undulated initial caries lesions, which are separated. (b) Large approximal caries lesion covering most of the surface with a small translucent zone. Color coding: blue ¼ enamel; pink ¼ dentin; red ¼ body of the caries lesion; yellow transparent ¼ translucent zone.

14.3 Dental Caries

physiological conditions, there is an equilibrium between the loss of calcium and phosphate ions from enamel (dentin or cementum) and the uptake of these ions from the supersaturated saliva via the bioﬁlm ﬂuid. If the pH falls unbuﬀered below 5.5 due to the complex metabolic activity of plaque bacteria, it is likely that the demineralization prevails. This clinically invisible mineral loss is called ‘‘precaries’’, and this is followed by subsurface demineralization creating the porous body of the lesion with a pore volume of ca. 25% in contrast to the normal volume of 0.1%. The remineralization of these initial caries lesions is clinically and experimentally well established [21, 22]. Non-invasive treatment with the outcome of caries reversals or stagnating subsurface lesions with no cavitation is the main aim of contemporary clinical dentistry [23]. The carious infection starts with bacterial colonization of the porous body of the enamel lesion, and cell-mediated mineral formation by pulpal odontoblasts are the consequences of early dentin lesions to that infection. This means, that stagnating, still non-cavitated enamel lesions lead to intertubular, peritubular and intratubular dentin formation, resulting in a hypermineralized translucent zone (Figs. 14.7 and 14.8). This host response mimics dentin reactions to wear or the durodentin formation of enamel-free teeth in many animals. Further progression of the disease is characterized by the superﬁcial caries with cavitation. The consequence from the pulpal side is a manifest dentin lesion with outer zones of destruction by demineralization and proteolysis of the collagen matrix and inner zones of cell-mediated mineralization of the translucent zone and later of the irritation dentin (Fig. 14.7b). Deep caries lesions exhibit softened dentin close to the pulp chamber, and chronic and acute inﬂammatory reactions are obligatory host responses. The alternating process of regeneration and, therefore, further dentin mineralization and of degeneration ending up in pulp necrosis can take years or even decades. From a biomineralization point of view the treatment strategy and the successful outcome is aimed at ‘‘simple blocking’’ of the bioﬁlm accumulation at all different levels to avoid further infection: Plaque control to prevent precaries and to treat initial caries (mainly due to the permanent bioavailability of ﬂuorides in saliva). Alloplastic tooth restorations to block the bioﬁlm maturation on dentin surfaces and to restore masticatory and cosmetic functions. Root canal ﬁlling treatment to avoid endodontal bioﬁlms, to stop periapical inﬂammatory reactions, to prevent aggressive bone and tooth resorption, and to support bone remodeling and cementogenesis. In conclusion, dental caries is an excellent example of the strength of biomineralization as an important part of host responses to cope with the environment. Physico-chemical active remineralization of enamel and outer zones of dentin

213

214

14 Clinical Aspects of Tooth Diseases and their Treatment

Fig. 14.6 Biomineralization consequences of the pathogenesis of dental caries diseases according to the concept of progression and stagnation [7]. Comparison of factors contributing to either bursts of disease progression or to long-lasting periods of stagnation. SC ¼ stem cell; T ¼ T lymphocyte; B ¼ B lymphocyte; P ¼ plasma cell; L ¼ lymphocyte.

14.3 Dental Caries

Fig. 14.7 Polarized light microscopy of caries lesions. (a) Initial enamel caries subsurface lesion with typical zones. (b) Root caries lesion with hypermineralized translucent dentin and dead tracts due to empty dentin tubules.

Fig. 14.8 Root caries lesion. Scanning electron microscopy image of the translucent zone and dentin tubules which contain intratubular dentin produced by the odontoblast processes.

215

216

14 Clinical Aspects of Tooth Diseases and their Treatment

and cementum (‘‘dental caries seen from the oral side’’) is quite successfully supported by cell-mediated ongoing mineralization of dentin, cementum and new bone (‘‘dental caries seen from the pulpal and periodontal side’’). This is what makes prevention and treatment of caries on an individual basis of risk assessment clinically safe and functionally successful.

14.4 Periodontal Diseases

The most common periodontal diseases are gingivitis and periodontitis; these are infectious diseases caused by supragingival and subgingival bioﬁlms, whereby mainly Gram-negative anaerobic bacteria shift towards pathogenetic conditions and overcome the host responses. All of the various gingivitis forms have no consequences for biomineralization, because they represent a pure (and usually reversible) inﬂammation of the soft tissue only. In contrast, periodontitis involves loss of the tooth support by bone loss and the resorption of cementum (Figs. 14.9 and 14.10) until the inﬂammation is stopped or the loose tooth is exfoliated or extracted. Except for early bone loss, periodontitis – like manifest caries lesions – is a life-long disease, although the alveolar bone loss takes place during shortterm bursts of cell-mediated resorption by osteoclasts followed by long-term stable conditions with spontaneous regeneration of bone and cementum defects and de novo formation of the periodontal ligament between the two mineralized tissues (Fig. 14.11). This is explained by the concept of progression (within hours,

Fig. 14.9 Pathohistology of marginal periodontitis. Considerable bone loss with past resorbing activity and lacunae formation is visible; the intact periodontal ligament and apposition of cementum represent the reduced tooth support.

14.4 Periodontal Diseases

Fig. 14.10 Pathohistology of marginal periodontitis. Scanning electron microscopy image of root resorption lacunae. (a) Irregular scattered resorption lacunae on the root surface. (b) As (a), but at higher magniﬁcation (see scale bar), showing deep bowl-like resorption pattern.

days and weeks) and stagnation (within months, years or even decades) (Fig. 14.12) [7, 24]. The average loss of bone support in untreated periodontitis teeth is 0.1 mm year1, while inter-individual and intra-individual diﬀerences range from 0.08 to more than 2 mm year1. Therefore, untreated aggressive periodontitis cases result in tooth loss within a few years. A very rare disease entity is the external granuloma which rapidly resorbs enamel, dentin, cementum, and alveolar bone by the action of clastic cells. The transformation of these cells is not well understood. In contrast, the activity of cementoclasts and osteoclasts during bursts of progression of periodontitis is stimulated by inﬂammatory cytokines. This rather well-known cascade of inﬂammation is inﬂuenced by the pathogenicity of diﬀerent bioﬁlms, as well as by the systemic and local immunity. From a phylogenetic point of view, the loss of mineralized tooth support mimics the natural exfoliation pattern of deciduous teeth, in permanent teeth aimed at the exfoliation of infected tissues. Moreover, the common ankylosis (Fig. 14.11a) between bone and cementum ‘‘repeats’’ the acrodontal tooth-tobone attachment in most teeth of ﬁsh, and in some reptiles. The main aim of dental treatment is to halt disease progression by removing subgingival bioﬁlms and their niches. In cases of deep vertical bony defects, guided tissue regeneration by blocking the downgrowth of gingival epithelium with diﬀerent membranes shows promise [25]. The procedure stimulates the cellmediated de-novo formation of specialized connective tissue structures: cementum, alveolar bone and periodontal ligament. Eﬀorts have been made to stimulate the osteoneogenesis, osteoconduction or osteoinduction by bone substitution materials. Autogenic, allogenic and xenogenic transplants such as intraoral or extraoral living bone, or human and bovine demineralized freeze-dried bone allografts, seem to have some future potential, whereas to date alloplastic biomaterials

217

218

14 Clinical Aspects of Tooth Diseases and their Treatment

Fig. 14.11 Pathohistology of marginal periodontitis and dental caries. Resorption and apposition. (a) Apposition of bone to the root surface, which results in a tight connection between the tooth and alveolar bone (ankylosis) and the development of a secondary periodontal ligament. (b) Alveolar bone resorption by osteoclasts within the periodontal ligament. (c) Resorption of dentin by odontoclasts.

(hydroxyapatite, b-tricalcium phosphate, bioactive glasses, polymers) have not demonstrated any reproducible regeneration patterns under experimental conditions. The same is true for other growth factors [platelet-derived growth factor (PDGF), insulin-like growth factor (IGF), bone morphogenic protein (BMP), erythroblast macrophage protein (EMP)] which have been investigated in animal models with spontaneous, periodontitis teeth. At this point, the biological rationale behind periodontal treatment should perhaps be taken into account. Clearly, the aim is to stop the shedding process of the infected tooth tissues within an environment which is exposed to the open oral cavity and all bioﬁlms. With regards to the regenerative potential of bone substitution materials and various growth factors, periodontitis teeth exhibit completely

14.4 Periodontal Diseases

Fig. 14.12 Biomineralization consequences of the pathogenesis of periodontal diseases, according to the concept of progression and stagnation [7]. Comparison of factors contributing to either bursts of rapid disease progression with periodontal pocket formation or to longlasting periods of stagnation. T ¼ T lymphocyte; B ¼ B lymphocyte.

219

220

14 Clinical Aspects of Tooth Diseases and their Treatment

diﬀerent biomineralization structures compared to any other bone structure in the body.

14.5 Dental Trauma

Dental trauma may be allocated to two classes, namely acute injuries and chronic disturbances of the teeth and the bone support. 14.5.1 Acute Dental Trauma

In vertebrates, the teeth are the only biomineralized structures exposed to the environment, and therefore acute trauma is rather common. Infractions and fractures of enamel and the concussion of teeth have consequences for survival of the pulp, with the risk of degeneration and necrosis after years, or more than a decade after trauma. There are no reactions of mineralized tissues, however, and in contrast dentin and cementum fractures may demonstrate a rather high regenerative potential that may be exploited for treatment strategies. In most cases, a certain cycle of events is apparent, resulting in repair. The cycle begins with an acute inﬂammatory response to the trauma, followed by the stimulation of clastic cells resorbing dentin, cementum and/or bone, the replacement of resorbed hard tissues by granulomatous cells, attracting later blastic cells to repair the dentin, cementum, and bone. It is likely that new odontoblasts are derived from pulpal stem cells, while new cementoblasts and osteoblasts may derive from the stem cells of bone marrow. This potential for the cell-mediated de-novo formation of mineralized tissues is rather high compared to that of the chronic infectious diseases of dental caries and periodontitis. Consequently, the superimposition of infection after trauma leads to a considerable reduction in repair capacity. 14.5.2 Chronic Dental Trauma

Omnivorous teeth show attrition (tribological wear) and abrasion (wear due to abrasives), even at a young age, whereby the formation of secondary dentin and the slow, but continuous, eruption of human teeth are suﬃcient responses. The pulp chamber represents a type of depository for the formation of secondary dentin, when it is needed. The less the wear, and the less the functional eruption, the greater the chance that dental caries and periodontal inﬂammatory destruction can be maintained [26]. In contrast to the natural phenomenon of life-long wear and eruption, extraordinary challenges such as bruxism or acid erosion may have clinical consequences. Pathologic wear leads to the development of areas of fracture-prone

References

enamel, and may even cause exposure of the pulp chamber. In general, eroded enamel surfaces will be remineralized within hours or days. Likewise, following an acid challenge, the same mechanism of remineralization of the opened tubule entrances in root dentin will be enforced, supported by the bioavailability of ﬂuoride ions, within a few days, such that the hypersensitivity of the tooth is minimized or blocked. Teeth truly tell tales, and hydroxyapatite biomineralization was truly the major step forward in evolution. Under the physico-chemical conditions of the maritime and terrestrial environments inhabited by vertebrates, hydroxyapatite forms acellular mineral structures (dental enamel), unipolar cellular mineralized tissues with centripedal growth (dentin) and, ﬁnally, diﬀerent forms of bone and cementum. Hydroxyapatite may be regarded, therefore, as the optimal biomineral in fetal, infantile and adult formation, as well as for resorption when needed. This high remodeling potential may play a major role not only in the initiation and progression of diseases, but also in strategies for their treatment.

References 1 R.J. Krejsa, H.C. Slavkin, J. Dent. Res. 2 3

4 5

6

7 8

1987, 66, 144. M.M. Smith, B.K. Hall, Biol. Rev. 1990, 65, 277. M.M. Smith, M.I. Coates, in: M.F. Teaford, M.M. Smith, M.W.J. Ferguson (Eds.), Development, Function and Evolution of Teeth. Cambridge University Press, Cambridge, 2000, p. 133. M.M. Smith, B.K. Hall, Evolut. Biol. 1993, 27, 387. P. Ga¨ngler, W.H. Arnold, in: P. Ga¨ngler, et al. (Eds.), Konservierende Zahnheilkunde und Parodontologie. Thieme, Stuttgart, New York, 2005, p. 21. P. Ga¨ngler, in: M.F. Teaford, M.M. Smith, M.W.J. Ferguson (Eds.), Development, Function and Evolution of Teeth. Cambridge University Press, Cambridge, 2000, p. 173. P. Ga¨ngler, Zahn-Mund-Kieferheilkd. 1985, 73, 477. B.K.B. Berkovitz, G.R. Holland, B.J. Moxham, in: B.K.B. Berkovitz, G.R. Holland, B.J. Moxham (Eds.), Oral Anatomy, Histology and Embryology. Mosby, Edinburgh, London, New York, 2002, p. 290.

9 A.L. Boskey, Connect. Tissue Res. 1996,

35, 357. 10 A. Linde, M. Goldberg, Crit. Rev. Oral

Biol. Med. 1993, 4, 679. 11 W.T. Butler, Connect. Tissue Res. 1995,

33, 59. 12 B.R. Genge, L.N. Wu, R.E. Wuthier,

J. Biol. Chem. 1990, 265, 4703. 13 T. Kirsch, H.D. Nah, D.R. Demuth,

14

15

16 17

18

19 20

G. Harrison, E.E. Golub, S.L. Adams, M. Paciﬁci, Biochemistry 1997, 36, 3359. C. Montessuit, J.P. Bonjour, J. Caverzasio, J. Bone Miner. Res. 1995, 10, 625. F. Neues, W.H. Arnold, F. Fischer, F. Beckmann, P. Ga¨ngler, M. Epple, Mat. Wiss. Werkstoﬀtech. 2006, 37, 426–431. J.D. Bartlett, J.P. Simmer, Crit. Rev. Oral Biol. Med. 1999, 10, 425. C.E. Smith, J.R. Pompura, S. Borenstein, A. Fazel, A. Nanci, Anat. Rec. 1989, 224, 292. S.J. Brookes, J. Kirkham, R.C. Shore, W.A. Bonass, C. Robinson, Connect. Tissue Res. 1998, 39, 89. T. Aoba, Anat. Rec. 1996, 245, 208–18. P. Ga¨ngler, Nova Acta Leopoldina NF262 1986, 85, 525.

221

222

14 Clinical Aspects of Tooth Diseases and their Treatment 21 O. Backer Dirks, J. Dent. Res. 1966,

24 S.S. Socransky, A.D. Haﬀajee, J.M.

45, 503. 22 T. Koulourides, H. Cueto, W. Pigman, Nature 1961, 189, 226. 23 O. Fejerskov, E. Kidd (Eds.), Dental Caries: The Disease and its Clinical Management. Blackwell Munksgaard, Oxford, 2003.

Goodson, J. Linde, J. Clin. Periodontol. 1984, 11, 21. 25 T. Karring, S. Nyman, J. Gottlow, L. Laurell, Periodontology 1993, 1, 26. 26 Z. Ugur, P. Ga¨ngler, A.O. Karababa, Dtsch. Zahna¨rztl. Z. 2001, 56, 172.

223

15 Dental Caries: Quantifying Mineral Changes Susan M. Higham and Philip W. Smith

Abstract

Dental caries is a disease characterized by the net loss of mineral from the dental hard tissues. Despite being largely preventable, it accounts for signiﬁcant morbidity and consumes signiﬁcant resources in developed and developing countries. Tooth decay is preferably managed by early detection and institution of proven methods of caries prevention which promote remineralization of the carious lesion. Thus, the ability to monitor mineral changes in dental caries forms a crucial part of the evaluation of potential preventive and clinical strategies aimed at managing tooth decay. A number of in vitro and in vivo studies have been conducted with the aim of elucidating the factors which may be involved in the development, prevention, and remineralization of the dental hard tissues. This chapter concentrates on two key techniques used to evaluate the early events in dental caries; namely transverse microradiography and quantitative light-induced ﬂuorescence, and their application in a variety of in vitro and in vivo studies involving the demineralization and remineralization of caries lesions. Keywords: tooth decay, dental caries, caries prevention, in vitro, in vivo, demineralization, remineralization, caries detection, microradiography, quantitative lightinduced ﬂuorescence.

15.1 Introduction

Dental caries is an infectious transmissible disease which is characterized by a net loss of mineral from the dental hard tissues. It has a multifactorial etiology, and the factors involved in the development of dental caries have been known for many years. Ultimately, the clinical presentation is the result of the complex interplay of factors which relate to the presence of cariogenic dental plaque, quantitative and qualitative aspects involving salivary function, diet, behavior and lifeHandbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

224

15 Dental Caries: Quantifying Mineral Changes

style, together with suﬃcient time for the various factors to act. Although caries may aﬀect any of the mineralized dental tissues, the majority of research investigations have been directed at elucidating the processes involved in the development of enamel caries, along with the development of strategies aimed at its prevention, for the pragmatic reason of reducing the burden of dental decay in childhood. As ageing populations retain natural teeth for longer, there are strong reasons for detailed study of the carious processes aﬀecting the other mineralized dental tissues, namely cementum and dentine.

15.2 Enamel Caries

Human enamel is a highly mineralized tissue in which calcium phosphate crystals, mostly as hydroxyapatite, comprise approximately 99% of the dry weight. Hydroxyapatite crystals found in enamel contain impurities such as carbonate and ﬂuoride ions which can inﬂuence its solubility. The tooth surface is typically covered by pellicle, derived from salivary proteins, and at times dental plaque, and in turn these interact with salivary ﬂuid. Saliva is normally supersaturated with respect to calcium and phosphate ions, and in turn this steady state favors the integrity of the crystals of hydroxyapatite. When considering the carious process at the tooth level, bacterial fermentation of carbohydrates leads to the production of a variety of organic acids by dental plaque. When the pH is lowered as a result of acid production, oral ﬂuids become undersaturated with respect to hydroxyapatite and, given suitable conditions, this may then lead to dissolution of the enamel hydroxyapatite crystals. This process of demineralization can be reversed, however, and remineralization may occur if the oral ﬂuids once again become supersaturated with respect to apatite. Caries, therefore, is a dynamic process and has been likened to an ionic see-saw; whether net gain or loss of ions occurs is dependent upon the interaction of the factors mentioned earlier. Caries-preventive strategies aim to redress the loss of mineral by promoting conditions which favor the remineralization of depleted enamel crystals. As the carious process is fundamentally a biologically based physico-chemical reaction involving mineral changes, many investigations have aimed at studying the processes of enamel demineralization and remineralization, and this in turn has involved the examination of mineral changes that take place in enamel. Classically, these studies have involved the use of in vitro and in situ model systems designed to allow the quantiﬁcation of mineral change, thus permitting the evaluation of cariogenic challenges to the teeth and caries-preventive strategies.

15.3 Dentine Caries

Dentine, like bone, can be regarded as a mineralized connective tissue, and whilst both tissues share several similarities they also demonstrate signiﬁcant diﬀer-

15.4 Analyzing Mineral Changes in Dental Caries

ences. Dentine is a highly organized composite vital tissue consisting of mineral in the form of hydroxyapatite crystals, which in turn are arranged in an organic matrix, principally in the form of type I collagen. The organic phase also contains smaller amounts of substances such as phosphoproteins, proteoglycans, and glycoproteins. This complex hydrated composite is arranged as curved orientated tubules surrounded by a highly mineralized peritubular organic matrix, embedded in a less-mineralized intertubular matrix. If dentine is analyzed by weight, mineral accounts for approximately 70% and the organic phase for 20%, the remainder being water. In contrast to enamel, the hydroxyapatite crystals in dentine are much smaller and are arranged with their c-axes parallel with the collagen ﬁbers in the organic matrix. Dentine vitality results from its close association with viable pulpal connective tissue; this allows the pulp–dentine complex to react to the carious process through further deposition of mineral in an attempt to protect the vulnerable pulp from assault. In the crown of the tooth, dentine is contiguous with the overlying enamel at the enamel–dentine junction, and therefore may be aﬀected by the carious processes in enamel caries. If net demineralization continues in enamel, the enamel eventually will break down and cavity formation results. This in turn may lead to further demineralization of dentine crystals and subsequently cariogenic bacteria may invade dentine; protein degradation may then occur. It is important to note that, like enamel caries, the early process of caries in dentine involves mineral changes in hydroxyapatite crystals. As mentioned previously, the retention of teeth later in life is becoming more prevalent, and this – when coupled with gingival recession – leads to exposure of the roots of the teeth. When surrounded by periodontium root dentine is covered by a thin (30–50 mm) layer of another mineralized tissue, cementum, which forms part of the attachment apparatus of the tooth to its socket in the jaw. Collagen ﬁbers traverse the periodontal ligament and run from the alveolar bone and become embedded into the cementum. When the gingival tissues recede and the roots become exposed to the mouth, the thin layer of cementum is often lost, leaving an exposed dentine surface; this gives rise to the possibility of caries aﬀecting the dentine of the roots of such teeth. The early stages of root caries have been likened to those in enamel caries, with a similar sequence of physicochemical events taking place involving subsurface demineralization.

15.4 Analyzing Mineral Changes in Dental Caries

Methods used both in vitro and in situ require the creation of artiﬁcially produced caries lesions in the enamel of extracted teeth. These lesions may then be subjected to a range of diﬀerent conditions either in vitro, or alternatively placed in the mouths of experimental subjects for evaluation in situ. Following the desired experimental interventions, the samples of enamel are then prepared in the laboratory and undergo analysis which aims to quantify the amount of mineral

225

226

15 Dental Caries: Quantifying Mineral Changes

change in the enamel sample. Over the years, a variety of methods have been used to evaluate the mineral changes seen in the carious process; these include microradiography [1], microhardness [2], polarized light [3]; light scattering [4]; iodine absorptiometry [5]; confocal laser scanning [6]; microtomography [7]; and quantitative laser- and light-induced ﬂuorescence [8]. Microradiography has generally been recognized as the most practical and reliable technique for the direct measurement of mineral loss and gain in the dental hard tissues. Using this technique, mineral changes can be quantiﬁed, mineral distribution can be analyzed, and the method is applicable to enamel, dentine or cementum [9]. Three techniques using microradiography have been described in the literature: (i) transverse microradiography (TMR) [10]; (ii) longitudinal microradiography (LMR) [11]; and (iii) wavelength-independent microradiography [12]. Among these microradiography techniques, TMR is considered to be the most practical, and to yield the most reliable information. It is generally held to be the ‘‘gold standard’’ for quantifying mineral changes, and has been used by us for a number of years. When considering alternative techniques it is important to assess their abilities with reference to TMR. Until recently, the nature of the preparation and testing involved in the analysis of enamel samples has precluded quantiﬁcation of mineral change in the early stages of enamel dissolution in vivo, and in addition has not permitted longitudinal quantiﬁcation of mineral changes in enamel caries. Studies conducted over several years within our research group have been involved with the development of in situ caries models, and more recently with the technique of quantitative light-induced ﬂuorescence (QLF), which provides the exciting prospect of being able to quantify longitudinal mineral changes seen in enamel caries in vivo. The remainder of the chapter aims to demonstrate to the reader the ways in which these two techniques have been utilized to quantify mineral changes in dental caries. This type of investigation is usually conducted in order to provide an insight not only into the mechanisms involved in the caries process but also, importantly, to allow the evaluation of potential preventive and restorative treatments aimed at managing dental caries. The nature of TMR usually means that it has been used in in vitro investigations, and also in studies which use in situ caries models, whilst QLF has been used directly in vivo as well as in the other types of investigation, several of which have been used to validate this technique against the TMR gold standard. 15.4.1 Transverse Microradiography

The basis of TMR is the measurement of X-ray absorption by a section of tooth, and simultaneously comparing this to the absorption of X-rays by a known standard. Plano-parallel sections of tooth (typically 80 mm for enamel) are prepared from the sample to be analyzed. These are placed on high-resolution photographic ﬁlm along with the standard aluminum stepwedge, and irradiated with

15.4 Analyzing Mineral Changes in Dental Caries Fig. 15.1 A microradiograph showing a subsurface lesion (ssl) represented by the hypomineralized dark zone.

Fig. 15.2 Screen output from the transverse microradiography (TMR) software. (Illustration courtesy of Inspektor Research Systems BV, The Netherlands.)

monochromatic X-rays. Absorption of X-rays by the tooth sample and stepwedge are directly reﬂected in the optical density of the developed ﬁlm (the microradiograph) (Fig. 15.1). The mineral content and analysis is calculated by means of Angmar’s formula [1], which is utilized by dedicated software to produce mineral content proﬁles that relate to mineral loss DZ (vol% mm), depth of the carious lesion Ld (mm), and lesion width Lw (mm). DZ is the integrated diﬀerence between the microradiograph of the sample exhibiting mineral loss and that of the sound sample, whilst Ld and Lw are determined from mineral distribution in the sample being analyzed (Fig. 15.2). 15.4.2 TMR Studies

Quantiﬁcation of mineral loss by TMR has produced much valuable information regarding the degree of mineral loss, and the depth and width of lesions. The

227

228

15 Dental Caries: Quantifying Mineral Changes

technique does have the major disadvantage that it is destructive, and this limits its use to in vitro or in situ experimental designs. In vitro studies, whilst providing much valuable information and allowing a high degree of control, are far removed from the natural situation occurring in the mouth. In situ model systems have the advantage that, although enamel lesions can be prepared, handled and analyzed outside the mouth, they are worn in the mouths of human volunteers and subjected to the varying conditions of the oral environment. In situ investigations therefore oﬀer conditions for the studies that are closer to the in vivo situation than experiments performed entirely in vitro. The enamel slabs resulting from both in vitro and in situ studies are examined in the same way by TMR. Many studies have employed TMR to quantify the degree of mineral loss or gain in tooth tissues following the consumption of various foods and drinks, and the use of dentifrices and mouthrinses of diﬀering formulations. Remineralization of human enamel has been shown to be enhanced following the chewing of chewing gum sweetened with sorbitol after meals and snacks; the eﬀect is thought to be due to an increased stimulation of salivary ﬂow during the chewing period [13]. A remineralizing eﬀect was also found with sucrosesweetened gum, although this had to be chewed for longer periods of time than sugar-free gums [14]. The consumption of cheese after sucrose challenges was found signiﬁcantly to reduce the demineralizing eﬀect of the challenge alone [15]. Not only does cheese stimulate salivary ﬂow, but the constituents of dairy products also have the potential to remineralize carious lesions by replacing lost calcium and phosphate. A sucrose-containing snack food that also contains milk products was found to lead to a large increase in the mineral gain of caries lesions when consumed as an in-between meal snack [16]. In situ models have been used to investigate the interactions between ﬂuoride and the frequency of sucrose consumption [17]. Important information regarding the number of carbohydrate intakes that can be consumed before a net demineralization occurs has been gained from this type of study. Although most of the information from TMR studies has added to our knowledge of the carious process in enamel, recent attempts have been made to investigate the potential for TMR to elucidate mineral changes in early dentine caries. This has involved both in vitro and in situ experiments [18, 19] which examined the inﬂuence of various parameters on the development of artiﬁcial caries lesions created in human dentine, and their subsequent behavior with respect to mineral changes under remineralizing conditions. It is possible to create artiﬁcial carieslike lesions in root dentine, although careful attention needs to be paid to the experimental parameters so that lesions appropriate to the type of in situ investigation are created [18]. A persistent drawback of all potential in situ dentine caries models is that it is not possible to elicit a response form the vital pulp–dentine complex. The ability to investigate the behavior of dentine in vivo under caries attack would represent a major step forwards. However, at present it is not possible to envisage using an in situ dentine caries model to study mineral changes in early dentine caries [19].

15.5 Quantitative Light-Induced Fluorescence

15.5 Quantitative Light-Induced Fluorescence

QLF is an optical technique which utilizes the principle that tooth enamel autoﬂuoresces under certain lighting conditions (Fig. 15.3). As teeth demineralize, there is a loss of ﬂuorescence which can be detected, quantiﬁed, and longitudinally monitored. This loss of ﬂuorescence is thought to be due to an increase in porosity that occurs as carious lesions develop. Uptake of water into the lesions occurs, and this leads to a decrease in the refractive index of the lesion. As a consequence, there is an increase in the scattering of light and a decreased light-path length and absorption, which in turn decreases auto-ﬂuorescence [20].

Fig. 15.3 White-light image of a tooth (left) with corresponding quantitative light-induced ﬂuorescence (QLF) image (right), showing an early demineralized area; this is visible due to the loss in ﬂuorescence resulting in a dark spot. (Image courtesy of Inspektor Research Systems BV, The Netherlands.)

Fig. 15.4 The quantitative light-induced ﬂuorescence (QLF) handpiece housing the light-guide, camera and mirror. (Illustration courtesy of Inspektor Research Systems BV, The Netherlands.)

229

230

15 Dental Caries: Quantifying Mineral Changes

The development of software to quantify changes in enamel ﬂuorescence into mineral content [21] was a major advance, and has allowed QLF to be developed for use in vivo. This provides major advantages for dental researchers and clinicians involved in the study and management of dental caries. The teeth are illuminated from an arc lamp with a liquid light guide with a peak intensity of 370 nm. In addition, a yellow ﬁlter with a wavelength of 520 nm is placed in front of a microcamera (charged couple device; CCD) which is used to capture images of the teeth (Fig. 15.4). The light guide and camera are housed in a hand-held device similar in size to a dental handpiece (Fig. 15.5) that incorporates a small mirror.

Fig. 15.5 Schematic diagram showing the QLF system. (Image courtesy of Inspektor Research Systems BV, The Netherlands.)

Fig. 15.6 Example of output from QLF software, showing longitudinal demineralization measurements. (Image courtesy of Inspektor Research Systems BV, The Netherlands.)

15.5 Quantitative Light-Induced Fluorescence

This design makes it suitable for intraoral use and the capture of images of all tooth surfaces. The images are stored electronically and analyzed using customized software (Inspektor Research Systems BV, The Netherlands). The program detects the dark areas of the image where ﬂuorescence loss resulting from mineral loss has occurred, and simulates the ﬂuorescence radiance of sound enamel adjacent to the lesion. The absolute decrease in ﬂuorescence is expressed by DF, whereas the parameter used to express ﬂuorescence radiance loss integrated over the lesion area (in mm 2 ) is DQ (Fig. 15.6), which is comparable to DZ, the total mineral loss measured by TMR [22]. Many studies have been performed showing mineral changes in teeth using QLF, and these fall broadly into those conducted either in vitro or in vivo. 15.5.1 In Vitro QLF Studies

With all new technology it is important to evaluate whether the system is reliable, and that it measures what it claims to measure. To this end, a number of validation studies have been performed, and their results reviewed [23, 24]. These studies have compared the quantiﬁcation of mineral in teeth measured by QLF with measurements made by TMR, the accepted gold standard. A series of in vitro studies [25–27] has shown a good correlation between QLF and TMR (r ¼ 0:84), indicating that QLF is a reliable and valid technique. In our laboratories, we have performed many in vitro studies to identify the optimum conditions which should be adhered to when using QLF. These have included investigating the eﬀect of dehydration (Fig. 15.7), the presence of stain and plaque, lighting conditions, and focal distance [31–33]. Treatment has been suggested by studies performed in vitro [24–26]. Subsurface caries-like lesions were produced in areas of exposed enamel during a demineralization phase, and quantiﬁed using QLF. The teeth were then subjected to a remineralizing regime with the levels of mineral quantiﬁed (Figs. 15.8 and 15.9). The results of these studies indicated that QLF had the ability to monitor the demineralization and remineralization of enamel adjacent to orthodontic brackets on extracted human premolars.

Fig. 15.7 (a) Appearance a tooth with a caries lesion after bench drying for 24 h. (b) The same tooth following immersion in saliva.

231

232

15 Dental Caries: Quantifying Mineral Changes

Fig. 15.8 (a) Image showing a two-color adhesive tape applied to the tooth surface to ensure that the varnish and etch are retained in speciﬁc zones. (b) Image showing a premolar with bracket attached. Zone v is the lingual bracket with cleat; zone w is the varnished area; zones x and y are exposed areas.

Fig. 15.9 The extent of demineralization and remineralization with time around orthodontic brackets.

The present authors’ research group has also reported the detection in vitro of smooth-surface caries in deciduous teeth [5]. In this respect, QLF was found to be capable of quantifying mineral changes as accurately as in the permanent dentition, indicating signiﬁcant possibilities for the use of this non-invasive technique in children in a clinical setting (Fig. 15.10). Diﬀerences in the optical properties of dentine and enamel have made the QLF-based evaluation of mineral change in dentine diﬃcult; this in turn has led to the use of dyes [34], which is not ideal in vivo. The study of root caries is problematic using model systems (despite being used routinely with much success for enamel caries), as extracted dentine behaves very diﬀerently from vital dentine.

15.5 Quantitative Light-Induced Fluorescence

Fig. 15.10 Example of a deciduous tooth showing increasing levels of demineralization after longitudinal monitoring by QLF with lesion reconstruction (right-hand side).

The clinical diagnosis of root caries can also be problematic, due to confounding factors such as plaque, the proximity of gingival tissue, and the poor resolution of radiographs. Given these diﬃculties, further advances will be needed before techniques such as QLF can be used to monitor the progression of root caries in vivo. Despite the fact that the prevalence of caries is generally declining in young people, it remains a suﬃciently large problem to be a public health concern. When teeth do suﬀer from dental caries, decay aﬀecting the occlusal surfaces accounts for the majority of lesions [35]. These lesions are often only clearly visible on radiographs at a relatively advanced stage, and there would appear to be a need to develop a technique capable of detecting lesions on occlusal surfaces at an earlier stage. An interpretive index has been developed for occlusal caries that may be used by clinicians to extrapolate QLF readings and make informed decisions with regards to the clinical management of occlusal caries. This was performed in vitro using natural lesions in extracted teeth, ranging in severity from sound to grossly cavitated [36]. In addition to being imaged using QLF, each tooth was examined by visual clinical inspection, radiographically, and histologically (Fig. 15.11) using widely used and accepted criteria [37]. Data obtained from the QLF images showed good correlation with histological data, and the proposed occlusal caries index could provide the potential to guide the clinician with regards to the extent of a lesion aﬀecting a particular tooth. Although much of our attention has focused on the diagnosis of early primary caries, it is also very important to be able to detect caries occurring around dental

233

234

15 Dental Caries: Quantifying Mineral Changes

Fig. 15.11 Examples (from left to right) of white-light images, histology, QLF images and analysis reconstructions, showing lesion extent.

restorations, and our studies have also been directed towards this. Thus, in an in vitro study, QLF was able to detect demineralization longitudinally adjacent to restorative materials [38]. 15.5.2 In Vivo QLF Studies

Many of above-described in vitro studies were conducted to validate and optimize the conditions in which to use QLF, with the intention of developing the technique for use in vivo. Over the past few years, several studies have reported ﬁndings relating to the use of QLF to monitor the progression or regression of caries in vivo. Indeed, QLF was found to be a sensitive method for the longitudinal quantiﬁcation of mineral content in incipient caries lesions on the smooth surfaces of children’s teeth [39], as well as being more sensitive than clinical visual inspection, thereby allowing caries to be detected at an earlier stage [40]. The technique has also been found suﬃciently sensitive to detect mineral changes of only 20% from baseline, and with relatively small numbers of experimental subjects (n < 40) per group [41]. In the clinical situation, QLF has been found to be particularly useful in monitoring whitespot lesions that occur naturally around ﬁxed orthodontic brackets [40], where caries regressed when preventive regimens proved eﬀective. To date, our experiences (unpublished results) have suggested that QLF holds great promise for use in child patients, in adults suﬀering with xerostomia (Fig. 15.12), and in situations where secondary caries has developed (Fig. 15.13).

15.5 Quantitative Light-Induced Fluorescence

Fig. 15.12 White-light image of the dentition of a patient with xenostomia, and corresponding QLF images of the lower right (LR), lower left (LL), lower right (LR), upper right (UR), upper left (UL). Note the dark areas showing demineralization not seen on the white-light image.

235

236

15 Dental Caries: Quantifying Mineral Changes

Fig. 15.13 Examples of secondary caries. White-light image and (b) QLF image of a failing restoration in tooth 43. (c) White-light image and (d) QLF image of a failing restoration in tooth 22.

Clearly, QLF has great potential for providing sensitive and reliable detection of caries at a very early stage in vivo, before it is detectable by clinical inspection or conventional radiological techniques. QLF also oﬀers powerful diagnostic imaging capabilities for the detection of caries in a variety of clinical presentations.

References ¨ m, J.-E. Glas, 1 B. Angmar, D. Carlstro 2 3 4

5 6

7

8

9

10

11

J. Ultrastruct. Res. 1963, 8, 12. J. Arends, J. Schuthof, W.L. Jongebloed, Caries Res. 1980, 14, 190. L.M. Silverstone, Oral Sci. Rev. 1973, 3, 100. J.J. ten Bosch, H.C. van der Mei, P.C.F. Borsboom, Caries Res. 1984, 18, 540. H. Almqvist, F. Lagerlof, Caries Res. 1993, 27, 100. M. Fontana, Y. Li, A.J. Dunipace, T.W. Noblitt, G. Fischer, B.P. Katz, G.K. Stookey, Caries Res. 1996, 30, 317. J.C. Elliot, F.S.L. Won, P. Anderson, G.R. Davis, S.E.P. Dowker, Connect. Tiss. Res. 1998, 38, 61. M. van der Veen, E. de Josselin de Jong, Monograph in Oral Sci. 2000, 17, 144. J.J.M. Damen, R.A.M. Extercate, J.M. ten Cate, Adv. Dent. Res. 1997, 11, 415. E. de Josselin de Jong, J.J. ten Bosch, J. Noordmans, Phys. Med. Biol. 1987, 32, 887. E. de Josselin de Jong, A.H.I.M. van der Linden, P.C.F. Borsboom, J.J. ten Bosch, Caries Res. 1988, 22, 153.

¨ ter, J. Noordsmans, J.J. 12 F.M. Herkstro ten Bosch, Caries Res. 1990, 24, 399. 13 S.A. Leach, G.T.R. Lee, W.M. Edgar,

J. Dent. Res. 1986, 68, 1644. 14 R.H. Manning, W.M. Edgar, J. Clin.

Dent. 1992, 3, 71. 15 M.F. da Silva, G.N. Jenkins, R.C.

16 17

18 19 20 21

22

23

Burgess, H.J. Sandham, Caries Res. 1986, 20, 263. S.M. Higham, P.W. Smith, P. Walsh, W.M. Edgar, Caries Res. 1991, 25, 233. M.S. Duggal, K.J. Toumba, B.T. Amaechi, M.B. Kowash, S.M. Higham, J. Dent. Res. 2001, 80, 1721. P.W. Smith, K.P. Preston, S.M. Higham, J. Dent. 2005, 33, 253. P.W. Smith, K.P. Preston, S.M. Higham, J. Dent. 2005, 33, 269. B. Angmar-Mansson, J.J. ten Bosch, Adv. Dent. Res. 1987, 1, 14. E. de Josselin de Jong, F. Sundstrom, H. Westerling, S. Tranaeus, J.J. ten Bosch, B. Angmar-Mansson, Caries Res. 1995, 29, 2. I.A. Pretty, W.M. Edgar, S.M. Higham, Int. J. Paed. Dent. 2002, 12, 158–167. J.J. ten Bosch, in: G.K. Stookey (Ed.), Early detection of dental caries II.

References

24

25

26

27

28 29 30 31 32

Indiana University School of Dentistry 1999, p. 261. I.A. Pretty, W.M. Edgar, S.M. Higham, in: G.K. Stookey (Ed.), Early detection of dental caries III. Indiana University School of Dentistry 2003, p. 253. S. Al-Khateeb, J.M. ten Cate, B. Angmar-Mansson, E. de Josselin de Jong, R.A.M. Exterkate, G. Sundstrom, A. Oliveby, Adv. Dent. Res. 1997, 11, 502. S. Al-Khateeb, A. Oliveby, E. de Josselin de Jong, B. AngmarMansson, Caries Res. 1997, 31, 132. S. Al-Khateeb, C.-M. Forsberg, E. de Josselin de Jong, B. AngmarMansson, G. Sundstrom, R.A.M. Exterkate, Am. J. Orthod. Dentofacial Orthop. 1998, 113, 595. B.T. Amaechi, S.M. Higham, J. Biomed. Optics 2002, 7, 7. I.A. Pretty, W.M. Edgar, S.M. Higham, J. Oral Rehab. 2002, 29, 369. I.A. Pretty, W.M. Edgar, S.M. Higham, J. Oral Rehab. 2004, 31, 179. P.E. Benson, N. Pender, S.M. Higham, Eur. J. Orthod. 2003, 25, 149. P.E. Benson, N. Pender, S.M. Higham, Eur. J. Orthod. 2003, 25, 159.

33 I.A. Pretty, N. Pender, W.M. Edgar,

34

35 36

37 38

39

40

41

S.M. Higham, Eur. J. Orthod. 2003, 25, 217. I.A. Pretty, G.S. Ingram, E.A. Agalamanyi, W.M. Edgar, S.M. Higham, J. Oral Rehab. 2003, 30, 1151. L.W. Ripa, G.S. Leske, A.O. Varma, J. Public Health Dent. 1988, 48, 8. S.M. Higham, P.W. Smith, I.A. Pretty, in: G.K. Stookey (Ed.), Early detection of dental caries III. Indiana University School of Dentistry 2003, p. 195. K.R. Ekstrand, D.N. Ricketts, E.A.M. Kidd, Caries Res. 1997, 31, 224. I.A. Pretty, P.W. Smith, W.M. Edgar, S.M. Higham, Dent. Mater. 2003, 19, 368. S. Tranaeus, S. Al-Khateeb, S. Bjorkman, S. Twetman, B. AngmarMansson, Eur. J. Oral Sci. 2001, 109, 71. J.G. Boersma, M.H. van der Veen, M.D. Lagerweij, B. Bokhort, B. PrahlAnderson, Caries Res. 2005, 39, 41. G.J. Eckert, B.P. Katz, S. Ofner, A.G. Ferreria Zandona, H. Eggertson, T. Doi, G.K. Stookey, J.S. Wefel, Caries Res. 2005, 39, 294.

237

239

16 Periodontal Regeneration Hom-Lay Wang and Lakshmi Boyapati

Abstract

An ultimate goal of periodontal therapy is not only to halt inﬂammatory disease but also to promote cell growth and new attachment. Traditional periodontal treatments aim to arrest disease progression by providing access to debride root surfaces and to eliminate/reduce periodontal pockets. Regenerative procedures, in addition, aim to restore/regenerate lost tissues. Periodontal regeneration, however, is not always a predictable or realistic goal of therapy. In this chapter we describe the biology of periodontal wound-healing and clinical applicability and outcomes of guided tissue regeneration (GTR) utilizing diﬀerent barrier membranes, osseous grafting, and newly developed biologic modiﬁers. Key words: periodontal disease, wound healing, regeneration, bone grafting, membranes, guided tissue regeneration, biologic modiﬁers.

16.1 Deﬁnitions

In this chapter, the following deﬁnitions are used: Regeneration refers to the reproduction or reconstitution of a lost or injured tissue [1]. Periodontal regeneration is deﬁned as the restoration of lost periodontium or supporting tissues, and includes the formation of new alveolar bone, new cementum, and new periodontal ligament. Repair describes the healing of a wound by tissue that does not fully restore the architecture or the function of the part [1]. New attachment is deﬁned as the union of connective tissue or epithelium with a root surface that has been deprived of Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

240

16 Periodontal Regeneration

its original attachment apparatus. The new attachment may be epithelial adhesion or connective tissue adaptation or attachment, and may include new cementum. Reattachment describes the reunion of epithelium and connective tissue with a root surface [1]. Guided tissue regeneration (GTR) describes procedures which attempt to regenerate lost periodontal structures through diﬀerential tissue responses, and typically refers to the regeneration of periodontal attachment [1]. Barrier techniques are used for excluding connective tissue and gingiva from the root, in the belief that they will interfere with regeneration [1]. Bone ﬁll is deﬁned as the clinical restoration of bone tissue in a treated periodontal defect. Bone ﬁll does not address the presence or absence of histologic evidence of new connective tissue attachment or the formation of new periodontal ligament [1]. Open probing clinical attachment is used to describe the tissue seen at re-entry surgery after regeneration procedures [2]. This term is not commonly used because the clinical attachment cannot be probed in an open environment.

16.2 Periodontal Wound Healing 16.2.1 Wound-Healing Principles

Surgical debridement and resective procedures are traditional surgical treatments employed to improve clinical parameters and to arrest disease progression [3–6]. Regeneration of the attachment apparatus is unlikely following these therapeutic treatments [7]. Healing is typically by repair, with the formation of a long junctional epithelium [8, 9]. The healing of periodontal wounds parallels that of incisional wounds in other parts of the body, with an inﬂammatory phase followed by the formation of nascent granulation tissue, which is later remodeled [10, 11]. Healing is initiated with the formation of a ﬁbrin clot between the ﬂap margin and the root surface, and replacement of this ﬁbrin clot by connective tissue attached to the root surface [11]. It has been proposed that the ‘‘ﬁbrin linkage’’ is critical to connective tissue attachment to the root surface. When the tensile strength of the linkage is exceeded, resulting in a tear, then healing with a long junctional epithelium will result [12]. Factors that can complicate the process of periodontal regeneration include the presence of multiple specialized cell types, stromal–cellular interactions, diverse microbial ﬂora, a transepithelial environment, and the presence of an avascular

16.3 Techniques Used for Regeneration

root surface [13]. More predictable regenerative outcomes might be anticipated as the periodontal wound healing process is better understood. 16.2.2 Compartmentalization

The concept of compartmentalization was ﬁrst introduced by Melcher in 1976 [14]. The connective tissues of the periodontium are divided into four compartments: the lamina propria of the gingiva; the periodontal ligament; the cementum; and the alveolar bone. Guided tissue regeneration is based on the exclusion of gingival connective tissue cells and the prevention of epithelial downgrowth into the wound. GTR procedures and membranes are used to accomplish the objectives of controlled cell/tissue repopulation, space maintenance, and clot stabilization [15–17]. 16.2.3 Evaluating Regeneration

Guided tissue regeneration is more eﬀective than open-ﬂap debridement (OFD) in improving clinical attachment levels, probing depth reduction, and bone ﬁll in intrabony and furcation defects [18–22]. However, the true nature of attachment to root surfaces can only be evaluated via histologic examination. Improved probing attachment levels following treatment can be misleading as they may only reﬂect an increased resistance to probing [23, 24]. Furthermore, radiographic analysis and clinical observations made at re-entry will not reveal either new connective tissue attachment or whether a long junctional epithelium has formed between the root surface and the bone [15, 25]. A study with non-human primates examined the method of evaluating osseous repair suggested by previous investigators. The study results showed that presence of new attachment should not be based on re-entry, clinical measurements and radiographs, but rather should rely on more deﬁnitive observations such as histological assessment [9]. Another consideration is that, even with histologic analysis, the attachment level prior to therapy must also be determined. A notch created at the most apical level of calculus is an appropriate marker [26], but clearly this depends on the presence of calculus.

16.3 Techniques Used for Regeneration 16.3.1 Root Surface Biomodiﬁcation 16.3.1.1 Root Surface Conditioning Root surface conditioning with tetracycline, citric acid, and ethylenediaminetetraacetic acid (EDTA) has been used as part of regenerative procedures. Surface de-

241

242

16 Periodontal Regeneration

mineralization has been claimed to enhance new attachment via the following eﬀects: elimination of bacterial endotoxin [27, 28]; exposure of collagen ﬁbrils [12, 29]; prevention of migration of epithelial cells [30]; removal of the smear layer [31]; and new cementum formation and increased ﬁbroblastic attachment to the treated root surfaces [32]. Although both, in-vitro and animal studies, have demonstrated enhanced connective tissue attachment following acid demineralization [12, 33, 34], reports of human histology demonstrating regeneration following demineralization are almost non-existent [35, 36]. A systematic review of results from clinical trials using any type of root conditioning agent, concluded that there was a lack of clinically signiﬁcant beneﬁt of regeneration in patients with chronic periodontitis [37]. Adverse eﬀects such as delayed healing and inﬂammatory pulpal damage have also been reported [38, 39]. 16.3.2 Bone Replacement Grafts

Grafting procedures attempt to stimulate the regeneration of lost periodontal structures through either conductive or inductive processes: Osteoconduction involves the graft material acting as a scaﬀold to support new tissue growth, where the graft material is eventually replaced with host tissue. Osteoinduction processes involve the graft material directing the host tissues to regenerate the lost structures. Osteogenesis refers to the formation or development of new bone by cells contained in the graft. The clinical objectives of bone grafting for periodontal regeneration include: (i) probing depth reduction; (ii) clinical attachment gain; (iii) bone ﬁll of the osseous defect; and (iv) the regeneration of new bone, cementum and periodontal ligament, as determined by histological analysis [40]. All four criteria must be met in order to prove the occurrence of periodontal regeneration. Graft material selection should be determined based on biologic acceptability, predictability, clinical feasibility, minimal operative hazards, minimal postoperative sequelae, and patient acceptance [40]. For the treatment of intrabony defects, bone grafts increase bone ﬁll, minimize crestal bone loss, gain clinical attachment level, and reduce periodontal probing depths (Table 16.1) compared to OFD [18]. For the treatment of furcations, the combination of bone graft and GTR results in greater clinical outcomes than GTR alone [22].

16.3 Techniques Used for Regeneration Table 16.1 Eﬀects of various osseous grafts on average defect ﬁll and probing depth reduction.

Graft

Defect ﬁll [%]

Probing depth reduction [mm]

Autograft Allograft Alloplasts Open ﬂap debridement

75–80 60–70 <50 <50

2.5–3.0 1.7–2.0 1.0–1.5 1.0

16.3.2.1 Autografts Bone autografts of intraoral origin may be harvested from a healing extraction site, edentulous ridge, exostoses, tori, the maxillary tuberosity, ramus, or chin. Extraoral sites include grafts from the iliac crest, tibia, and cranium. Autogenous iliac cancellous bone and marrow are graft materials of high osteogenic potential [40], and can be used either fresh or frozen. Case reports demonstrating successful bone ﬁll after their use in furcations, dehiscences, and intraosseous defects of various morphologies have been reported [41–44]. Additional case reports have shown mean bone ﬁlls of 3.3 to 3.6 mm in a large number of intraosseous defects, and a 2.5 mm increase in crestal bone height [45]. The histological evaluation of treated sites showed strong evidence of periodontal regeneration [44], although diﬃculties in obtaining the graft material, the high cost of hospitalization for the patient, the possibility of root resorption with fresh grafts, the potential of tooth loss following its use, and the likelihood of morbidity associated with collection of the material have greatly limited the use of iliac crest grafts in periodontal therapy [46]. Several types of autogenous bone graft can be used for periodontal regeneration, including cortical bone chips, osseous coagulum, or a blend of cortical and cancellous bone [47–49]. Cortical bone chips are still occasionally used today, but these have largely been replaced by osseous coagulum and bone blend because the cortical chips are generally much larger particles (1559:6 183 mm), and thus have a higher potential for sequestration [50]. A mean bone ﬁll range of 1.2 to 3.4 mm (>50% of the initial defect) has been reported following intraoral grafts [51, 52]. Healing following these procedures may be via regeneration or the formation of long junctional epithelium, or a combination of these [25]. 16.3.2.2 Allografts Allografts involve bone taken from one human for transplantation to another. The four main types of allograft are frozen cancellous iliac bone and marrow cryopreserved bone from the head of a femur, freeze-dried bone allograft (FDBA), and demineralized freeze-dried bone allograft (DFDBA). Although the clinical results after using frozen iliac allografts are favorable [53], the need for exten-

243

244

16 Periodontal Regeneration Table 16.2 Comparison of freeze-dried bone allograft (FDBA) and demineralized freeze-dried bone allograft (DFDBA).

FDBA

DFDBA

Demineralized More bone morphogenetic protein

Not demineralized Better space maintenance Slower resorption rate compared to DFDBA Osteoconductive More radio-opaque Breakdown by way of foreign body reaction Primary indication bone augmentation associated with implant treatment (e.g., guided bone regeneration, sinus grafting, ridge augmentation)

expression potential Possible osteoinduction Osteoconductive More radiolucent Rapid resorption Primary indication periodontal disease associated with natural teeth

sive cross-matching to decrease the likelihood of both graft rejection and disease transmission has precluded its widespread utilization in periodontal therapy. The diﬀerences noted between freeze-dried bone allograft (FDBA) and demineralized freeze-dried bone allograft (DFDBA) are listed in Table 16.2. 16.3.2.2.1 Mineralized Cortical FDBA This approach was introduced to periodontal therapy in 1976 [54]. In general, controlled clinical trials indicate bone ﬁll ranging from 1.3 to 2.6 mm using FDBA, with approximately 63% of defects showing b 50% bone ﬁll [55–57]. Histological evidence of regeneration with FDBA has yet to be demonstrated in humans, however. 16.3.2.2.2 Demineralized DFDBA This approach is based on experiments indicating that the demineralization of cortical bone grafts causes the release of bone morphogenic proteins (BMPs), thus enhancing osteogenic potential [58]. However, commercial bone banks do not verify the speciﬁc amount of BMPs, nor the level of inductive capacity in any graft material that they sell [59, 60]. A high osteoinductive potential and more rapid turnover than FDBA suggest superior properties for periodontal defects [61]. The particle size of DFDBA used in studies is generally in the range of 250 to 1000 mm [26, 62, 63]. Human histologic data have shown that sites grafted with DFDBA had more new cementum, new bone, and new periodontal ligament formation when compared to non-grafted sites [62, 64]. Bone ﬁll following the grafting of intrabony defects generally ranges from 1.7 to 2.9 mm (60–70% of the defect depth) [19, 55, 65]. The use of bone grafts from other humans has raised concerns of disease transmission, such as HIV/AIDS, bovine spongiform encephalopathy (BSE), and hepatitis, although strict procurement and processing techniques signiﬁcantly reduce

16.3 Techniques Used for Regeneration

this risk [66, 67]. The risk of HIV transmission has been calculated as approximately 1 in 8 million [68]. 16.3.2.2.3 Human Mineralized Bone Human mineralized bone (Puros; Zimmer Dental, Carlsbad, CA, USA) is an allograft of cancellous bone that undergoes the unique Tutoplast process. This consists of the following stages: (i) delipidation with acetone and ultrasound; (ii) osmotic treatment; (iii) oxidation with hydrogen peroxide to destroy unwanted proteins; (iv) solvent dehydration with acetone to preserve the collagen ﬁber structure; and (v) low-dose gamma irradiation. The manufacturers claim that Tutoplast preserves the trabecular pattern and mineral structure better than freeze-drying, and thus provides a more osteoconductive material. Recently, an initial human case series evaluated the use of Puros to treat periodontal osseous defects. Single osseous defects in nine patients were grafted and subsequently re-entered at 6 months. On average, the probing depth was improved by 3 mm, and there was 2.5 mm bone ﬁll [69]. In addition, Puros has been used to treat class II mandibular defects in a randomized controlled clinical study, whereby 30 patients were assigned randomly to OFD, Puros, or Puros with a bioabsorbable collagen membrane. The results showed that the mineralized human cancellous allograft, with or without a collagen membrane, can signiﬁcantly improve bone ﬁll in mandibular class II furcation defects [70, 71]. 16.3.2.3 Xenografts A xenograft is a graft taken from a donor of another species. Two currently available types of xenograft are the coralline calcium carbonate and bovine-derived hydroxyapatite (OsteoGraf N-300/N700 and BioOss, respectively). The former xenograft is derived from the exoskeleton of marine coral species, and most reports with this show bone ﬁll, though human histology demonstrating true regeneration is limited [72, 73]. Bovine-derived hydroxyapatite is more popular than coralline calcium carbonate. BioOss and OsteoGraf have similar chemical compositions, but the latter has a crystalline structure rather than the microporous structure of BioOss. The results of a clinical trial showed that BioOss implantation resulted in pocket depth reduction, a gain in clinical attachment, and bone ﬁll in periodontal defects to the same extent as DFDBA [74]. Both, human histology [72, 75] and animal experiments [76], have suggested a potential beneﬁt of placing BioOss in periodontal bony defects. However, the majority of studies suggest that these materials are osteoconductive and are resorbed very slowly [77]; thus, their use is more appropriately conﬁned to bone ﬁll or to implant-related applications such as guided bone regeneration around implants, sinus lift procedures, and ridge augmentation [78, 79]. The risk of disease transmission, of which BSE is a particular concern, is unlikely [80]. The World Health Organization has labeled bone as type IV (no transmission) for prion diseases [81]. 16.3.2.4 Alloplasts Alloplastic materials are synthetic, inorganic, biocompatible, or bioactive bone graft substitutes. Six types of alloplastic material are currently commercially avail-

245

246

16 Periodontal Regeneration

able, namely: (i) hydroxyapatite cement; (ii) non-porous hydroxyapatite; (iii) porous hydroxyapatite; (iv) beta tricalcium phosphate; (v) polymethylmethacrylate/ hydroxyethylmethacrylate (PMMA/HEMA) calcium layered polymer; and (vi) bioactive glass. The ideal properties of an alloplastic graft material are as follows [82]: biocompatibility, minimal ﬁbrotic reaction, an ability to undergo remodeling and support new bone formation, comparable strength to cortical/cancellous bone, and a similar modulus of elasticity to bone to prevent fatigue fracture under cycling load. Tricalcium phosphate and bioactive glass are absorbable, whereas porous, non-porous hydroxyapatite materials and PMMA/HEMA polymer are non-absorbable. In general, synthetic bone grafts – when used to treat periodontal osseous defects – have demonstrated improved clinical parameters when compared to non-grafted controls [83, 84]. Similar clinical results have also been found when allografts and alloplasts were compared [85, 86]. However, particles are frequently encapsulated by ﬁbrous connective tissue during re-entry surgery or histologic analysis [85, 87, 88]. 16.3.3 Guided Tissue Regeneration 16.3.3.1 Non-Absorbable Membranes Guided tissue regeneration was ﬁrst performed in animal and human studies using Millipore ﬁlters [15, 89]. The ﬁrst commercially available non-absorbable membrane was made from expanded polytetraﬂuoroethylene (ePTFE), and composed of two parts. The coronal collar comprises an open microstructure to facilitate ingrowth of connective tissue but prevent apical migration of the epithelium. The remaining occlusive portion prevents the gingival tissue from interfering with the healing process at the root surface. Non-absorbable membranes made from ePTFE have been considered the ‘‘gold standard’’ for GTR procedures due to a long history of positive clinical results [90–93], although absorbable membranes have demonstrated similar clinical and histologic outcomes when compared to ePTFE [94–99]. The most signiﬁcant problem with non-absorbable membranes is early membrane exposure; this occurs especially in thin tissues, and often results in bacterial contamination leading to poor clinical outcomes [100–102]. 16.3.3.2 Absorbable Membranes The major advantage of absorbable membranes is that they do not require a second surgical procedure. The beneﬁts of absorbable membranes compared to non-absorbable membranes are listed in Table 16.3. In addition, a lower incidence of early membrane exposure has been reported when compared to ePTFE [103]. Three major types of absorbable membrane are currently available: collagen; polyglycolic/polylactic polymer; and acellular dermal matrix. Degradable polymers of polylactic acid, polyglycolic acid, or mixtures of both, have provided similar clinical results compared with other membranes [98, 104–107]. One study

16.3 Techniques Used for Regeneration Table 16.3 Advantages of bioabsorbable membranes.

No need for re-entry surgery to retrieve barrier membrane:

Biologically absorbable materials hold potential to:

Be more tissue-compatible Promote wound coverage Resist/prevent microbial colonization

Reduce surgery time and reduce cost Reduce overall treatment morbidity Increase patient acceptance Reduce risk of loss of regenerated attachment due to re-entry

Table 16.4 Current commonly used collagen membranes.

Product

Company

Source

Components

BioMend

Zimmer (USA)

Bovine tendon

Type I collagen

Bioguide

Geistlich (Switzerland)

Porcine dermis

Type I and III collagen

Ossix

3i (USA)

Bovine tendon

Type I collagen

Regaurd

Ace Surgical (USA)

Bovine tendon

Type I collagen

Neomen membrane

Citagenix Inc. (Canada)

Bovine tendon

Type I collagen

Alloderm

Biohorizons (USA)

Human dermis

Type I collagen, elastin, ECM components

Pericardium

Zimmer (USA)

Human pericardium

Type I collagen

ECM ¼ extracellular matrix.

in which a co-polymer of polylactic acid and polyglycolic acid was compared with a type I collagen membrane in the treatment of infrabony defects, reported similar clinical improvements for both membranes [108]. Comparable regenerative outcomes using various absorbable membranes have also recently been conﬁrmed [109, 110]. Collagen membranes can be crosslinked to extend absorption time and reduce antigenicity (Table 16.4). They are well-tolerated, malleable, support cell proliferation, facilitate early wound stabilization and maturation, are chemotactic for ﬁbroblasts, and are absorbed naturally [111–113]. Collagen membranes are also effective in inhibiting epithelial migration and promoting new connective tissue attachment [99, 114–117]. They also have certain hemostatic advantages, notably their ability to induce platelet aggregation (which facilitates early clot formation)

247

248

16 Periodontal Regeneration

and to stabilize wounds, both of which are considered essential for successful regeneration [118]. Other materials such as human dermal matrix and pericardium have not yet been fully studied in periodontics, and further studies in this ﬁeld will surely be forthcoming.

16.3.4 Biologic Modiﬁers 16.3.4.1 Growth Factors/Cytokines Growth factors are a class of naturally occurring biologic mediators that regulate the proliferation, migration and/or extracellular matrix synthesis of cells, including those of the periodontium [119]. Growth factors that might have therapeutic potential in periodontal regeneration include platelet-derived growth factor (PDGF), insulin-like growth factors (IGFs), ﬁbroblast growth factors (FGFs), and transforming growth factor-b [120]. In one study it was shown that the use of puriﬁed human PDGF BB mixed with bone allograft resulted in signiﬁcant periodontal regeneration in class II furcations and intrabony defects [121]. In another study, PDGF and IGF were used in conjunction to treat periodontal osseous defects and furcations in humans, with favorable results [122]. Recently, a large-scale randomized controlled clinical trial evaluated the safety and eﬃcacy of puriﬁed recombinant human PDGF-BB mixed with a synthetic beta-tricalcium phosphate (beta-TCP) matrix for the treatment of advanced periodontal osseous defects. The low-dose PDGF group showed greater clinical attachment level (CAL) gain and percentage bone ﬁll at 6 months compared to controls (beta-TCP alone) [123]. However, additional studies are required to elucidate fully the potential of growth factors for enhancing periodontal regeneration. 16.3.4.2 Bone Morphogenetic Proteins (BMPs) BMPs include a large number of proteins that belong to the transforming growth factor (TGF)-beta superfamily. These are characterized by their ability to induce bone and cartilage formation de novo [58, 124, 125]. An initial human study indicated that osteogenin combined with DFDBA signiﬁcantly enhanced the regeneration of a new attachment apparatus [126], although since then conﬂicting results have been reported [127]. Factors that may contribute to variable results include: the inﬂuence of root conditioning, occlusal loading, BMP dose, release characteristics of the carrier, and the periodontal disease model being utilized [128]. Recombinant BMP-2, -3, and -7 have been most extensively studied in animals, and have demonstrated enhanced cementogenesis and osteogenesis [129, 130]. The ability of BMP-2 to stimulate the formation of a functionally oriented PDL has been questioned, and has been associated with ankylosis [127, 129]. More recently, recombinant human BMP-12 demonstrated an ability to regenerate alveolar bone and periodontal attachment in a dog model [131]. Further studies are required to verify the potential applications of BMPs in periodontal regeneration.

16.4 Factors Inﬂuencing GTR Success

16.3.4.3 Pep-Gen p-15 Pep-Gen p-15 is an anorganic bovine graft which is enhanced with a synthetic 15-amino acid sequence (p-15). P-15 is a collagen-derived cell binding peptide that is reported to attract and bind ﬁbroblasts and osteoblasts, and promote binding to the anorganic bovine hydroxyapatite carrier [132]. Successful reports of clinical and human histologic results have been relatively few in number, however [21, 133, 134]. A recent study evaluated the healing of critical-sized fenestration defects in a canine model compared to putty or particulate forms of this graft. More new bone formation was noted for putty compared to particulate, both of which were superior to the non-grafted control [135]. Additional clinical and histologic data are needed to establish true periodontal regeneration when this material is used. 16.3.4.4 Enamel Matrix Derivative (EMD) EMD is a preparation of low molecular-weight enamel matrix proteins that is harvested from developing porcine tooth buds using acetic acid [136–138]. Amelogenins constitute around 90% of the preparation [139], which is solubilized in a propylene glycol alginate carrier solution. Amelogenins are extremely hydrophobic, which causes them to aggregate together and to stimulate crystallization and cementogenesis. EMD is available commercially as Emdogain, and is FDAapproved for the treatment of angular periodontal defects [140]. Several human case series have demonstrated improved clinical attachment levels, probing depth reduction and radiographic bone ﬁll [138, 140, 141], although reports of histologic evidence of regeneration in humans have been inconsistent and mostly limited to case reports [142–144]. A recent multicenter, randomized trial compared EMD with barrier membranes for the treatment of class II mandibular furcations [145]. The reduction in open horizontal furcation depth (the primary outcome variable) was signiﬁcantly greater in the EMD group, although secondary outcome variables such as probing depth, bleeding on probing and attachment level did not diﬀer statistically between the groups [146]. A systematic review of human studies supported the use of EMD in the treatment of periodontal bony defects to improve clinical attachment levels and reduce probing depths, although long-term stability remains to be proven [147]. When EMD and GTR were used to treat intrabony defects and compared after 4 years, no signiﬁcant diﬀerence was noted between groups [148]. Hence, further studies are required to determine the long-term beneﬁts of EMD.

16.4 Factors Inﬂuencing GTR Success

Factors that may inﬂuence the outcome of GTR treatment are listed in Figure 16.1. Indications and contraindications for GTR are listed in Table 16.5.

249

250

16 Periodontal Regeneration

Fig. 16.1 Factors inﬂuencing guided tissue regeneration (GTR) success.

Table 16.5 Indications and contraindications for guided tissue regeneration.

Indications Narrow two- or three-wall infrabony defects Circumferential defects Class II molar furcations Recession defects

Contraindications Any medical condition contraindicating surgery Infection at defect site Poor oral hygiene Heavy smoking Tooth mobility > 1 mm Defect depth < 4 mm Width of attached gingiva at defect site a 1 mm Thickness of attached gingiva at defect site a 0:5 mm Furcations with short root trunks Generalized horizontal bone loss Advanced lesions with little remaining bone support Multiple defects

16.4 Factors Inﬂuencing GTR Success

16.4.1 Patient Factors

Patient factors that may impair healing following GTR include cigarette smoking, diabetes mellitus, immune-mediated, and uncontrolled periodontal disease in other areas of the mouth [149]. The type of periodontal disease does not seem to inﬂuence the outcome of regenerative therapy [150]. Cigarette smoking is associated with poorer clinical outcome after regenerative therapy [151–153], presumably as it impairs neutrophil function, alters ﬁbroblast functions (e.g., collagen synthesis and attachment), causes a shift to a more pathogenic microbial ﬂora, and reduces the vascularity of tissues. Heavy smokers in particular demonstrate less attachment gain and probing depth reductions [154]. The best results have been observed in healthy, non-smoking patients demonstrating good plaque control and compliance with recommended oral hygiene measures [150, 155].

16.4.2 Defect/Local Factors

Case selection is critical for ensuring successful clinical outcomes, and defect access, root anatomy, anatomical considerations, and morphology of the defect all are critical factors in the decision-making process. Narrow furcations are diﬃcult to debride thoroughly with curettes, and the use of ultrasonics with an open approach is often needed for adequate root surface preparation [156, 157]. Generally, three-walled and two-walled defects respond much better to therapy than one-walled defects [158]. Defect depth – both horizontal and vertical – will inﬂuence bone ﬁll where deeper defects can expect a greater amount of bone ﬁll [92]. In intrabony defects, greater success is shown in defects that are >3 mm deep and <2 mm in width [152, 153, 159]. Mobility at baseline may also compromise clinical attachment level gain after regenerative therapy [160]. Thin tissues have been found to show signiﬁcantly less clinical improvement and percentage root coverage [161–164]. Class II furcation sites are most indicated for osseous grafting. The therapeutic end-point of success is closure demonstrated on re-entry, or clinical improvement of the extent of involvement as assessed by probing. Generally, buccal mandibular class II furcation involvements respond best to regenerative approaches [165]. Mandibular class III and interproximal maxillary class II furcation involvements have demonstrated less evidence to support the use of regenerative techniques, while maxillary class III involvements had the least support [165–167]. One critical factor to consider is that of interproximal bone height; when interproximal bone is at a more coronal level than the furcation entrance, coverage and stabilization of the membrane will be easier and complete furcation closure is more likely [168].

251

252

16 Periodontal Regeneration

16.4.3 Treatment Factors

The type of membrane used may inﬂuence clinical success. In the past, nonabsorbable ePTFE has been considered the ‘‘gold standard’’ in regeneration, but more recent clinical trials have shown similar outcomes for ePTFE, absorbable collagen and PLA/PGA membranes [94, 98, 115, 169]. One major advantage of absorbable membranes is the lack of need for second-stage surgery, although a recent study showed that cross-linked bovine-derived membranes resulted in prolonged biodegradation, foreign body reaction, and decreased tissue integration and vascularization [170]. Systematic reviews comparing the use of membranes with bone grafts in furcation lesions have demonstrated enhanced bone ﬁll, probing depth reduction and clinical attachment gain when compared to membrane alone [22]. In contrast, the use of combination therapy was not shown to be superior to GTR alone for the treatment of intrabony defects [18, 22]. Other important aspects include the preservation of ﬂap tissue to ensure graft coverage (sulcular incision), extending the incision as far interdentally as possible to preserve the papilla, and extending the ﬂap at least one tooth mesial and distal to the graft site, prevention of ﬂap perforations, and removing all granulation tissue to maximize space available for graft material. Tension-free primary ﬂap closure is essential for graft success. 16.4.4 Postoperative Care

Membrane exposure during wound healing resulted in less clinical attachment gain [100, 103, 150, 171]. Membrane exposure has been reported more commonly after the use of non-absorbable membranes [103, 172]. When the membranes have been exposed to the oral environment they are rapidly colonized with periodontal pathogens, though the use of chlorhexidine rinses has been advocated by some authors to limit contamination [173]. The use of antibiotics has been reported widely, but is largely empirical. The rationale for their use is that infection has been considered a major cause of incomplete healing after GTR therapy, and that antibiotics may enhance graft success [149, 174]. Plaque control is a fundamental prerequisite to successful regeneration. Good oral hygiene has been shown to improve the regeneration of intrabony defects after various forms of periodontal surgery [175]. In addition to self-performed plaque control, compliance with a maintenance schedule is key [150], and for this reason a recall schedule of 3 months to maintain low plaque and gingival indices has been advocated [176].

16.5 Surgical Principles

16.5 Surgical Principles

Common clinical uses for periodontal regeneration include the treatment of furcation, intrabony, and recession defects. 16.5.1 Furcation Defects

When compared to OFD, in the treatment of furcation defects, GTR procedures show more favorable gains in vertical probing attachment level, reductions in vertical probing depth, and improvements in horizontal probing depth [22]. The best results are found using a combination of GTR and bone replacement grafts. Mandibular buccal and lingual furcations show the most favorable results after GTR therapy [22, 93, 114]. First and second mandibular molars have demonstrated a similar response to regeneration [150]. Maxillary buccal furcations may show positive results, but interproximal furcations are not predictable [166]. Molars with longer root trunks tend to respond more favorably to regenerative treatment than those with short root trunks [155]. The least favorable results are found in mandibular and maxillary class III furcations [2, 167]. 16.5.2 Intrabony Defects

GTR is an eﬀective treatment modality for the treatment of intrabony defects, with deep, narrow defects with a signiﬁcant three-wall component being the most amenable to treatment [149, 177]. No advantage has been found with the use of a combination technique compared to membrane barrier alone in the treatment of conﬁned intrabony defects [22]. Therefore, the additional use of bone graft in GTR for the treatment of intrabony defects is often unnecessary. 16.5.3 Root Coverage

Root coverage can be accomplished with GTR or GTR combined with bone grafting materials. An average root coverage of 76.4 G 11.3% and occurrence of complete root coverage of 33.1 G 20.4% have been reported [178]. The advantages of GTR-based root coverage are excellent color match and an elimination of the need to harvest donor tissue. However, the predictability of root coverage is inferior to that obtained with connective tissue grafting, and success is highly dependent on tissue thickness [178]. Root coverage in thin tissues is signiﬁcantly less than in thick tissues [161–163], though the exact deﬁnition of thin versus thick tissue has yet to be resolved (it is nominally 1 mm).

253

254

16 Periodontal Regeneration

16.5 Surgical Principles

16.5.4 Surgical Techniques [113, 179, 180]

The surgical techniques used by the authors for periodontal regeneration, specifically when GTR was employed, are illustrated in Figures 16.2–16.4. The initial incision should be made away from the defect so that closure is not directly over the defect. A full-thickness mucoperiosteal ﬂap should be reﬂected 2–3 mm beyond the defect. Apical to the mucogingival junction, a partial-thickness ﬂap is continued by blunt dissection to free the ﬂap from tension. Granulation tissue is removed and curettes or burs are used to plane and contour the exposed root surface. Where appropriate, interdental papilla are deepithelialized with a blade or diamond bur to provide a bleeding tissue bed. Epithelium should also be removed from the inner surface of the ﬂap with a sharp curette or diamond bur. The membrane should be trimmed so that it extends 2–3 mm beyond the margins of the defect in all directions. A trial membrane can serve as a template for the ﬁnal membrane. The ﬂap should be trimmed/released if needed to achieve primary tension-free closure. Cortical perforations with a 12-round bur are made to create bleeding at the defect site to facilitate bloody supply and the migration of progenitor cells. The graft material or biologic modiﬁer is placed at the defect site to create space or support the membrane. The membrane is added to the site and, if stable, ﬁxation is not necessary. If ﬁxation is needed, then pins, sutures, bone screws, or tacks can be used to achieve membrane stability. The surgical site should be closed with passive tension. Dressings should be used with caution because they may displace the graft material and collapse the membrane at the defect site. Postoperative care should comprise the following: Antibiotic (amoxicillin) for a minimum of 10 days. Warm salt water rinses for the ﬁrst 2 to 3 weeks.

________________________________________________________________________________ H Fig. 16.2 (a) Preoperative radiograph showing the infrabony defect on the distal of right lower mandibular 2nd molar. (b) Initial probing showed a 9-mm probing pocket depth on the right lower mandibular 2nd molar. (c) The initial intra-sulcular incision. (d) The extent of the osseous defect (8 mm) is shown after ﬂap elevation. (e) The bone graft (human mineralized allograft; Puros TM ; Zimmer Dental Inc., Carlsbad, CA, USA) was mixed with the patient’s own blood and placed on the defect. (f ) The collagen

membrane (BioMend4; Zimmer Dental Inc., Carlsbad, CA, USA) was correctly trimmed and placed to cover the defect. (g) The ﬂap was repositioned coronally and sutured with 4-0 Vicryl4 suture (Ethicon, Somerville, NJ, USA). (h) Uneventful healing was demonstrated at 2 weeks postoperatively. (i) Postoperative radiography at 1 year showed bone regeneration. (j) Clinical probing at 1 year showed 4 mm of probing pocket depth; hence, a 5-mm probing pocket depth reduction had been achieved.

255

256

16 Periodontal Regeneration

16.5 Surgical Principles

Fig. 16.4 (a) An initial buccal view showed 4 mm of Miller’s class I recession defect on the upper right 1st premolar. (b) Two diverging vertical releasing incisions were performed on the mesial and distal sides of upper right 1st premolar, with ﬂap refection. (c) The recession defect was properly prepared with curets and burs (low- and high-speed) to create a biologically acceptable concave

root surface. The collagen membrane (BioMend4) was properly trimmed to cover the recession defect (3 mm beyond), and then secured with 5-0 gut suture. (d) The ﬂap was coronally repositioned and sutured with 4-0 Vicryl4 suture. (e) Uneventful healing was demonstrated at 2 weeks postoperatively. (f ) Healing at 6 months postoperatively showed 100% root coverage.

________________________________________________________________________________ H Fig. 16.3 (a) The preoperative radiograph suggested furcation involvement on the left lower mandibular 1st molar. (b) Surgical view of class II furcation with 5-mm horizontal probing pocket depth. (c) A demineralized, freeze-dried bone allograft was placed to slightly overﬁll after de-cortication. (d) The collagen membrane (BioMend4) was trimmed and placed 3 mm beyond the furcation defect. (e) The ﬂap was coronally repositioned and sutured with 4-0 Vicryl4

suture. (f ) Uneventful healing was demonstrated at 1 week postoperatively. (g) Postoperative radiography at 1 year showed furcation bone ﬁll. (h) One-year postoperative clinical probing showed a 3-mm probing pocket depth. (i) Two-year postoperative clinical probing showed a 3-mm probing pocket depth. (j) Three-year postoperative clinical probing showed the stability of a 3-mm probing pocket depth.

257

258

16 Periodontal Regeneration

Chlorhexidine gluconate 0.12% mouthrinse for the next 3 weeks. Sutures are removed at 10–14 days. Gentle brushing with a soft brush can resume after 3 weeks, and ﬂossing after 1 month. The surgical site is checked every 2 weeks over a 2-month period.

16.6 Conclusions

The ultimate goal of periodontal therapy is not only to arrest the disease process, but also to regenerate lost tissues. Guided tissue regeneration therapy has been shown to be a predictable treatment modality for select periodontal defects. Although much speculation on actual new histological clinical attachment has surfaced, clinical parameters and long-term studies have shown positive outcomes. Several options are available for GTR and grafting materials to achieve this goal. Critical factors involved in achieving optimal results include case selection, ﬂap management, patient management, technique, and graft selection. Clinicians should be able correctly to select cases that are amenable to regenerative therapy, and to use the appropriate grafting materials when necessary. As new materials such as BMPs, growth factors and EMDs are developed, each must be evaluated in order to ensure that materials are used only when properly indicated.

References 1 M.A. Cohen, M. Rethman, et al.,

2

3

4

5

6

Glossary of Periodontal Terms, 4th edn. American Academy of Periodontology, Chicago, 2001. W. Becker, B.E. Becker, L. Berg, J. Prichard, R. Caﬀesse, E. Rosenberg, Int. J. Periodontics Restorative Dent. 1988, 8, 8. W. Becker, B.E. Becker, C. Ochsenbein, G. Kerry, R. Caﬀesse, E.C. Morrison, J. Prichard, J. Periodontol. 1988, 59, 351. W.B. Kaldahl, K.L. Kalkwarf, K.D. Patil, M.P. Molvar, J.K. Dyer, J. Periodontol. 1996, 67, 93. B.L. Pihlstrom, R.B. McHugh, T.H. Oliphant, C. Ortiz-Campos, J. Clin. Periodontol. 1983, 10, 524. S.P. Ramfjord, R.G. Caﬀesse, E.C. Morrison, R.W. Hill, G.J. Kerry, E.A.

7 8 9 10 11

12 13 14

Appleberry, R.R. Nissle, D.L. Stults, J. Clin. Periodontol. 1987, 14, 445. A.M. Polson, L.C. Heijl, J. Clin. Periodontol. 1978, 5, 13. J. Caton, S. Nyman, J. Clin. Periodontol. 1980, 7, 212. J. Caton, S. Nyman, H. Zander, J. Clin. Periodontol. 1980, 7, 224. U.M. Wikesjo, R.E. Nilveus, K.A. Selvig, J. Periodontol. 1992, 63, 158. U.M. Wikesjo, M. Crigger, R. Nilveus, K.A. Selvig, J. Periodontol. 1991, 62, 5. A.M. Polson, M.P. Proye, J. Periodontol. 1983, 54, 141. C.A. McCulloch, Periodontol. 2000 1993, 1, 16. A.H. Melcher, J. Periodontol. 1976, 47, 256.

References 15 S. Nyman, J. Lindhe, T. Karring, H.

16

17

18

19 20 21

22 23 24

25 26

27 28 29 30 31 32 33

34

Rylander, J. Clin. Periodontol. 1982, 9, 290. J.G. Caton, E.L. DeFuria, A.M. Polson, S. Nyman, J. Periodontol. 1987, 58, 546. S. Nyman, J. Gottlow, J. Lindhe, T. Karring, J. Wennstrom, J. Periodont. Res. 1987, 22, 252. M.A. Reynolds, M.E. AichelmannReidy, G.L. Branch-Mays, J.C. Gunsolley, Ann. Periodontol. 2003, 8, 227. J.T. Mellonig, Int. J. Periodontics Restorative Dent. 1984, 4, 40. E.S. Rosenberg, G.K. Fox, C. Cohen, J. Esthet. Dent. 2000, 12, 248. R.A. Yukna, D.P. Callan, J.T. Krauser, G.H. Evans, M.E. Aichelmann-Reidy, K. Moore, R. Cruz, J.B. Scott, J. Periodontol. 1998, 69, 655. K.G. Murphy, J.C. Gunsolley, Ann. Periodontol. 2003, 8, 266. G.C. Armitage, G.K. Svanberg, H. Loe, J. Clin. Periodontol. 1977, 4, 173. M.A. Listgarten, R. Mao, P.J. Robinson, J. Periodontol. 1976, 47, 511. M.A. Listgarten, M.M. Rosenberg, J. Periodontol. 1979, 50, 333. G.M. Bowers, B. Chadroﬀ, R. Carnevale, J. Mellonig, R. Corio, J. Emerson, M. Stevens, E. Romberg, J. Periodontol. 1989, 60, 683. J.J. Aleo, F.A. De Renzis, P.A. Farber, J. Periodontol. 1975, 46, 639. C.G. Daly, J. Clin. Periodontol. 1982, 9, 386. A.M. Polson, M.P. Proye, J. Clin. Periodontol. 1982, 9, 441. A.M. Polson, P.J. Hanes, J. Clin. Periodontol. 1987, 14, 357. J.S. Garrett, M. Crigger, J. Egelberg, J. Periodontal Res. 1978, 13, 155. A.A. Register, F.A. Burdick, J. Periodontol. 1976, 47, 497. R.G. Caﬀesse, M.J. Holden, S. Kon, C.E. Nasjleti, J. Clin. Periodontol. 1985, 12, 578. V.P. Terranova, L.C. Franzetti, S. Hic, R.M. DiFlorio, R.M. Lyall, U.M. Wikesjo, P.J. Baker, L.A. Christersson, R.J. Genco, J. Periodontal Res. 1986, 21, 330.

35 W.B. Albair, C.M. Cobb, W.J. Killoy,

J. Periodontol. 1982, 53, 515. 36 R.M. Frank, G. Fiore-Donno, G.

Cimasoni, J. Periodontol. 1983, 54, 389. 37 A. Mariotti, Ann. Periodontol. 2003, 8,

205. 38 E.C. Pettersson, I. Aukhil,

J. Periodontal Res. 1986, 21, 543. 39 P.C. Ryan, G.M. Newcomb, G.J.

40 41 42 43 44 45 46 47

48 49 50 51 52 53

54

55

56 57 58 59

Seymour, R.N. Powell, J. Clin. Periodontol. 1984, 11, 633. R.G. Schallhorn, J. Periodontol. 1977, 48, 570. R.G. Schallhorn, Periodontal Abstr. 1967, 15, 101. R.G. Schallhorn, J. Periodontol. 1968, 39, 145. R.G. Schallhorn, Int. Dent. J. 1980, 30, 101. M.R. Dragoo, H.C. Sullivan, J. Periodontol. 1973, 44, 599. R.G. Schallhorn, W.H. Hiatt, W. Boyce, J. Periodontol. 1970, 41, 566. R.G. Schallhorn, W.H. Hiatt, J. Periodontol. 1972, 43, 67. S.J. Froum, M. Ortiz, R.T. Witkin, R. Thaler, I.W. Scopp, S.S. Stahl, J. Periodontol. 1976, 47, 287. C.L. Nabers, T.J. O’Leary, J. Periodontol. 1965, 36, 5. D.J. Zaner, R.A. Yukna, J. Periodontol. 1984, 55, 406. J.T. Mellonig, Crit. Rev. Oral Biol. Med. 1992, 3, 333. W.H. Hiatt, R.G. Schallhorn, J. Periodontol. 1973, 44, 194. M.M. Rosenberg, J. Periodontol. 1971, 42, 195. W.H. Hiatt, R.G. Schallhorn, A.J. Aaronian, J. Periodontol. 1978, 49, 495. J.T. Mellonig, G.M. Bowers, R.W. Bright, J.J. Lawrence, J. Periodontol. 1976, 47, 125. J.M. Rummelhart, J.T. Mellonig, J.L. Gray, H.J. Towle, J. Periodontol. 1989, 60, 655. N. Blumenthal, J. Steinberg, J. Periodontol. 1990, 61, 319. J.T. Mellonig, Dent. Clin. North Am. 1991, 35, 505. M.R. Urist, Science 1965, 150, 893. Z. Schwartz, J.T. Mellonig, D.L. Carnes, Jr., J. de la Fontaine, D.L.

259

260

16 Periodontal Regeneration

60

61 62

63

64

65

66 67 68 69 70

71

72 73

74

75

76

Cochran, D.D. Dean, B.D. Boyan, J. Periodontol. 1996, 67, 918. Z. Schwartz, A. Somers, J.T. Mellonig, D.L. Carnes, Jr., D.D. Dean, D.L. Cochran, B.D. Boyan, J. Periodontol. 1998, 69, 470. J.T. Mellonig, Clin. Orthop. Relat. Res. 1996, 116. G.M. Bowers, B. Chadroﬀ, R. Carnevale, J. Mellonig, R. Corio, J. Emerson, M. Stevens, E. Romberg, J. Periodontol. 1989, 60, 675. G.E. Pearson, S. Rosen, D.A. Deporter, J. Periodontol. 1981, 52, 55. G.M. Bowers, B. Chadroﬀ, R. Carnevale, J. Mellonig, R. Corio, J. Emerson, M. Stevens, E. Romberg, J. Periodontol. 1989, 60, 664. G. Quintero, J.T. Mellonig, V.M. Gambill, G.B. Pelleu, Jr., J. Periodontol. 1982, 53, 726. L.S. Martin, J.S. McDougal, S.L. Loskoski, J. Infect. Dis. 1985, 152, 400. J.T. Mellonig, A.B. Prewett, M.P. Moyer, J. Periodontol. 1992, 63, 979. B.E. Buck, T.I. Malinin, M.D. Brown, Clin. Orthop. Relat. Res. 1989, 129. S. Vastardis, R.A. Yukna, Compend. Contin. Educ. Dent. 2006, 27, 38. Y.P. Tsao, R. Neiva, K. Al-Shammari, T.J. Oh, H.L. Wang, J. Periodontol. 2006, 77, 641. Y.P. Tsao, R. Neiva, K. Al-Shammari, T.J. Oh, H.L. Wang, J. Periodontol. 2006, 77, 416. J.T. Mellonig, Int. J. Periodontics Restorative Dent. 2000, 20, 19. M.L. Nevins, M. Camelo, S.E. Lynch, R.K. Schenk, M. Nevins, Int. J. Periodontics Restorative Dent. 2003, 23, 9. C.R. Richardson, J.T. Mellonig, M.A. Brunsvold, H.T. McDonnell, D.L. Cochran, J. Clin. Periodontol. 1999, 26, 421. M. Camelo, M.L. Nevins, R.K. Schenk, M. Simion, G. Rasperini, S.E. Lynch, M. Nevins, Int. J. Periodontics Restorative Dent. 1998, 18, 321. L.P. Clergeau, M. Danan, S. Clergeau-Guerithault, M. Brion, J. Periodontol. 1996, 67, 140.

77 M. Spector, Clin. Plast. Surg. 1994, 21,

437. 78 M. Hallman, S. Lundgren, L.

79

80 81 82

83 84 85

86

87 88 89

90 91

92

93

94

95

96

Sennerby, Clin. Implant. Dent. Relat. Res. 2001, 3, 87. M.B. Hurzeler, C.R. Quinones, A. Kirsch, C. Gloker, P. Schupbach, J.R. Strub, R.G. Caﬀesse, Clin. Oral Implants Res. 1997, 8, 476. A. Sogal, A.J. Tofe, J. Periodontol. 1999, 70, 1053. D.M. Asher, A.M. Padilla, M. Pocchiari, Biologicals 1999, 27, 265. W.R. Moore, S.E. Graves, G.I. Bain, Aust. N. Z. J. Surg. 2001, 71, 354. R.A. Yukna, E.T. Mayer, S.M. Amos, J. Periodontol. 1989, 60, 544. R.A. Yukna, J. Periodontol. 1994, 65, 342. J.D. Barnett, J.T. Mellonig, J.L. Gray, H.J. Towle, J. Periodontol. 1989, 60, 231. J.A. Bowen, J.T. Mellonig, J.L. Gray, H.T. Towle, J. Periodontol. 1989, 60, 647. S. Froum, S.S. Stahl, J. Periodontol. 1987, 58, 103. S.S. Stahl, S. Froum, J. Periodontol. 1986, 57, 211. S. Nyman, J. Gottlow, T. Karring, J. Lindhe, J. Clin. Periodontol. 1982, 9, 257. J. Gottlow, S. Nyman, T. Karring, J. Clin. Periodontol. 1992, 19, 315. W. Becker, B.E. Becker, M. Handelsman, C. Ochsenbein, T. Albrektsson, J. Periodontol. 1991, 62, 703. M. Handelsman, M. Davarpanah, R. Celletti, Int. J. Periodontics Restorative Dent. 1991, 11, 350. R. Pontoriero, J. Lindhe, S. Nyman, T. Karring, E. Rosenberg, F. Sanavi, J. Clin. Periodontol. 1988, 15, 247. P. Eickholz, T.S. Kim, R. Holle, E. Hausmann, J. Periodontol. 2001, 72, 35. R.G. Caﬀesse, C.E. Nasjleti, E.C. Morrison, R. Sanchez, J. Periodontol. 1994, 65, 583. R.G. Caﬀesse, L.F. Mota, C.R. Quinones, E.C. Morrison, J. Clin. Periodontol. 1997, 24, 747.

References 97 B.S. Black, M.E. Gher, J.B. Sandifer,

98 99 100 101

102 103

104 105

106

107

108 109

110

111

112 113 114

115

S.E. Fucini, A.C. Richardson, J. Periodontol. 1994, 65, 598. P. Eickholz, B. Pretzl, R. Holle, T.S. Kim, J. Periodontol. 2006, 77, 88. P. Bunyaratavej, H.L. Wang, J. Periodontol. 2001, 72, 215. E.E. Machtei, J. Periodontol. 2001, 72, 512. M. De Sanctis, G. Zucchelli, C. Clauser, J. Clin. Periodontol. 1996, 23, 1039. H. Nowzari, J. Slots, J. Clin. Periodontol. 1994, 21, 203. P. Bouchard, J.L. Giovannoli, C. Mattout, M. Davarpanah, D. Etienne, J. Clin. Periodontol. 1997, 24, 511. J. Caton, G. Greenstein, U. Zappa, J. Periodontol. 1994, 65, 1037. M. Christgau, G. Schmalz, E. Reich, A. Wenzel, J. Clin. Periodontol. 1995, 22, 306. A.R. Vernino, T.A. Ringeisen, H.L. Wang, M. Derhalli, J. Rapley, S.J. Nechamkin, J. Brekke, Int. J. Periodontics Restorative Dent. 1998, 18, 572. A.R. Vernino, H.L. Wang, J. Rapley, S.J. Nechamkin, T.A. Ringeisen, M. Derhalli, J. Brekke, Int. J. Periodontics Restorative Dent. 1999, 19, 56. J.S. Mattson, S.J. Gallagher, M.H. Jabro, J. Periodontol. 1999, 70, 510. A. Stavropoulos, A. Sculean, T. Karring, Clin. Oral Investig. 2004, 8, 226. L. Gineste, M. Gineste, L. Bluche, A. Guilhem, A. Elefterion, P. Frayssinet, D. Duran, H.L. Wang, Int. J. Periodontics Restorative Dent. 2005, 25, 61. P. Locci, M. Calvitti, S. Belcastro, M. Pugliese, M. Guerra, L. Marinucci, N. Staﬀolani, E. Becchetti, J. Periodontol. 1997, 68, 857. T. Takata, H.L. Wang, M. Miyauchi, Clin. Oral Implants Res. 2001, 12, 332. H.L. Wang, M. Miyauchi, T. Takata, J. Periodontal Res. 2002, 37, 340. H.L. Wang, R.B. O’Neal, C.L. Thomas, Y. Shyr, R.L. MacNeil, J. Periodontol. 1994, 65, 1029. H.L. Wang, R.L. MacNeil, Dent. Clin. North Am. 1998, 42, 505.

116 C.C. Chen, H.L. Wang, F. Smith,

117

118

119 120 121

122

123

124

125

126

127

128 129

130

131

G.N. Glickman, Y. Shyr, R.B. O’Neal, J. Periodontol. 1995, 66, 838. H.L. Wang, R.B. O’Neal, L.M. MacNeil, Pract. Periodontics Aesthet. Dent. 1995, 7, 59. A.D. Steinberg, G. LeBreton, R. Willey, S. Mukherjee, J. Lipowski, J. Periodontol. 1986, 57, 516. S. Rosenkranz, A. Kazlauskas, Growth Factors 1999, 16, 201. O. Anusaksathien, W.V. Giannobile, Curr. Pharm. Biotechnol. 2002, 3, 129. M. Nevins, M. Camelo, M.L. Nevins, R.K. Schenk, S.E. Lynch, J. Periodontol. 2003, 74, 1282. T.H. Howell, J.P. Fiorellini, D.W. Paquette, S. Oﬀenbacher, W.V. Giannobile, S.E. Lynch, J. Periodontol. 1997, 68, 1186. M. Nevins, W.V. Giannobile, M.K. McGuire, R.T. Kao, J.T. Mellonig, J.E. Hinrichs, B.S. McAllister, K.S. Murphy, P.K. McClain, M.L. Nevins, D.W. Paquette, T.J. Han, M.S. Reddy, P.T. Lavin, R.J. Genco, S.E. Lynch, J. Periodontol. 2005, 76, 2205. A.J. Celeste, J.A. Iannazzi, R.C. Taylor, R.M. Hewick, V. Rosen, E.A. Wang, J.M. Wozney, Proc. Natl. Acad. Sci. USA 1990, 87, 9843. E. Ozkaynak, P.N. Schnegelsberg, D.F. Jin, G.M. Cliﬀord, F.D. Warren, E.A. Drier, H. Oppermann, J. Biol. Chem. 1992, 267, 25220. G. Bowers, F. Felton, C. Middleton, D. Glynn, S. Sharp, J. Mellonig, R. Corio, J. Emerson, S. Park, J. Suzuki, et al., J. Periodontol. 1991, 62, 690. U.M. Wikesjo, P. Guglielmoni, A. Promsudthi, K.S. Cho, L. Trombelli, K.A. Selvig, L. Jin, J.M. Wozney, J. Clin. Periodontol. 1999, 26, 392. G.N. King, D.L. Cochran, J. Periodontol. 2002, 73, 925. T.J. Sigurdsson, M.B. Lee, K. Kubota, T.J. Turek, J.M. Wozney, U.M. Wikesjo, J. Periodontol. 1995, 66, 131. W.V. Giannobile, S. Ryan, M.S. Shih, D.L. Su, P.L. Kaplan, T.C. Chan, J. Periodontol. 1998, 69, 129. U.M. Wikesjo, R.G. Sorensen, A. Kinoshita, X. Jian Li, J.M. Wozney, J. Clin. Periodontol. 2004, 31, 662.

261

262

16 Periodontal Regeneration 132 R.S. Bhatnagar, J.J. Qian, A.

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

Wedrychowska, M. Sadeghi, Y.M. Wu, N. Smith, Tissue Eng. 1999, 5, 53. R.A. Yukna, J.T. Krauser, D.P. Callan, G.H. Evans, R. Cruz, M. Martin, J. Periodontol. 2000, 71, 1671. R.A. Yukna, J.T. Krauser, D.P. Callan, G.H. Evans, R. Cruz, M. Martin, J. Periodontol. 2002, 73, 123. S. Vastardis, R.A. Yukna, E.T. Mayer, B.L. Atkinson, J. Periodontol. 2005, 76, 1690. J.C. Rincon, H.R. Haase, P.M. Bartold, J. Periodontal Res. 2003, 38, 290. V. Cattaneo, C. Rota, M. Silvestri, C. Piacentini, A. Forlino, A. Gallanti, G. Rasperini, G. Cetta, J. Periodontal Res. 2003, 38, 568. N. Donos, L. Glavind, T. Karring, A. Sculean, Int. J. Periodontics Restorative Dent. 2003, 23, 507. S.J. Brookes, C. Robinson, J. Kirkham, W.A. Bonass, Arch. Oral Biol. 1995, 40, 1. A. Sculean, G.C. Chiantella, A. Miliauskaite, M. Brecx, N.B. Arweiler, Int. J. Periodontics Restorative Dent. 2003, 23, 345. A. Sculean, E. Reich, G.C. Chiantella, M. Brecx, Int. J. Periodontics Restorative Dent. 1999, 19, 157. R. Parodi, G. Liuzzo, P. Patrucco, G. Brunel, G.A. Santarelli, V. Birardi, B. Gasparetto, Int. J. Periodontics Restorative Dent. 2000, 20, 584. R. Parodi, G.A. Santarelli, B. Gasparetto, Int. J. Periodontics Restorative Dent. 2004, 24, 57. L. Francetti, M. Del Fabbro, M. Basso, T. Testori, R. Weinstein, J. Clin. Periodontol. 2004, 31, 52. S. Jepsen, B. Heinz, K. Jepsen, M. Arjomand, T. Hoﬀmann, S. Richter, E. Reich, A. Sculean, J.R. Gonzales, R.H. Bodeker, J. Meyle, J. Periodontol. 2004, 75, 1150. J. Meyle, J.R. Gonzales, R.H. Bodeker, T. Hoﬀmann, S. Richter, B. Heinz, M. Arjomand, E. Reich, A. Sculean, K. Jepsen, S. Jepsen, J. Periodontol. 2004, 75, 1188. W.V. Giannobile, M.J. Somerman, Ann. Periodontol. 2003, 8, 193.

148 A. Sculean, P. Windisch, G.C.

149

150

151

152

153

154 155

156 157

158 159

160

161

162 163 164 165

Chiantella, N. Donos, M. Brecx, E. Reich, J. Clin. Periodontol. 2001, 28, 397. A.B. Novaes, Jr., D.B. Palioto, P.F. de Andrade, J.T. Marchesan, Braz. Dent. J. 2005, 16, 87. E.E. Machtei, M.I. Cho, R. Dunford, J. Norderyd, J.J. Zambon, R.J. Genco, J. Periodontol. 1994, 65, 154. M.S. Tonetti, G. Pini-Prato, P. Cortellini, J. Clin. Periodontol. 1995, 22, 229. L. Trombelli, C.K. Kim, G.J. Zimmerman, U.M. Wikesjo, J. Clin. Periodontol. 1997, 24, 366. F. Klein, T.S. Kim, S. Hassfeld, H.J. Staehle, P. Reitmeir, R. Holle, P. Eickholz, J. Periodontol. 2001, 72, 1639. H. Preber, L. Linder, J. Bergstrom, J. Clin. Periodontol. 1995, 22, 946. G.M. Bowers, R.G. Schallhorn, P.K. McClain, G.M. Morrison, R. Morgan, M.A. Reynolds, J. Periodontol. 2003, 74, 1255. R.C. Bower, J. Periodontol. 1979, 50, 23. J.I. Matia, N.F. Bissada, J.E. Maybury, P. Ricchetti, Int. J. Periodontics Restorative Dent. 1986, 6, 24. K.A. Selvig, B.G. Kersten, U.M. Wikesjo, J. Periodontol. 1993, 64, 730. P. Cortellini, G.M. Bowers, Int. J. Periodontics Restorative Dent. 1995, 15, 128. P. Cortellini, M.S. Tonetti, N.P. Lang, J.E. Suvan, G. Zucchelli, T. Vangsted, M. Silvestri, R. Rossi, P. McClain, A. Fonzar, D. Dubravec, P. Adriaens, J. Periodontol. 2001, 72, 1702. C. Baldi, G. Pini-Prato, U. Pagliaro, M. Nieri, D. Saletta, L. Muzzi, P. Cortellini, J. Periodontol. 1999, 70, 1077. R.J. Harris, J. Periodontol. 1998, 69, 1426. L.H. Huang, R.E. Neiva, H.L. Wang, J. Periodontol. 2005, 76, 1729. C.R. Anderegg, D.G. Metzler, B.K. Nicoll, J. Periodontol. 1995, 66, 397. E.E. Machtei, R.G. Schallhorn, Int. J. Periodontics Restorative Dent. 1995, 15, 146.

References 166 R. Pontoriero, J. Lindhe, J. Clin. 167 168

169

170

171

172

Periodontol. 1995, 22, 756. R. Pontoriero, J. Lindhe, J. Clin. Periodontol. 1995, 22, 810. J. Horwitz, E.E. Machtei, P. Reitmeir, R. Holle, T.S. Kim, P. Eickholz, J. Clin. Periodontol. 2004, 31, 105. A.R. Vernino, F.L. Jones, R.A. Holt, R.E. Nordquist, J.W. Brand, Int. J. Periodontics Restorative Dent. 1995, 15, 84. D. Rothamel, F. Schwarz, M. Sager, M. Herten, A. Sculean, J. Becker, Clin. Oral Implants Res. 2005, 16, 369. K.A. Selvig, B.G. Kersten, A.D. Chamberlain, U.M. Wikesjo, R.E. Nilveus, J. Periodontol. 1992, 63, 974. A. Friedmann, F.P. Strietzel, B. Maretzki, S. Pitaru, J.P. Bernimoulin, Clin. Oral Implants Res. 2002, 13, 587.

173 A.B. Novaes, Jr., F.G. Gutierrez, I.F.

174

175 176 177 178

179 180

Francischetto, A.B. Novaes, J. Periodontol. 1995, 66, 864. J.J. Sanders, W.W. Sepe, G.M. Bowers, R.W. Koch, J.E. Williams, J.S. Lekas, J.T. Mellonig, G.B. Pelleu, Jr., V. Gambill, J. Periodontol. 1983, 54, 1. B. Rosling, S. Nyman, J. Lindhe, J. Clin. Periodontol. 1976, 3, 38. M.G. Newman, J. Periodontol. 1993, 64, 1166. P. Cortellini, G. Pini Prato, M.S. Tonetti, J. Periodontol. 1993, 64, 261. T.W. Oates, M. Robinson, J.C. Gunsolley, Ann. Periodontol. 2003, 8, 303. H.L. Wang, M.J. Carroll, Quintessence Int. 2001, 32, 504. H.L. Wang, K.F. Al-Shammari, Quintessence Int. 2002, 33, 715.

263

265

17 Tissue Engineering of Teeth Misako Nakashima

Abstract

The ultimate goal of endodontics and operative dentistry is to retain metabolic function and to reconstitute the occlusal and aesthetic function of teeth. Dental pulp is critical in maintaining homeostasis of the teeth. Deep caries and pulpal exposure have been treated by pulp capping or pulp amputation to preserve pulp tissue, but with limited success. Recent advances in stem cell biology and gene therapy technology have provided the potential for pulp stem cells as treatment in the regeneration of dentin-pulp complexes after injury such as caries. A rational approach to the tissue engineering of dentin is based on the triad of: responding stem/progenitor cells; morphogenetic signals for tissue induction; and the extracellular matrix scaﬀold to optimize the cellular microenvironment. In this chapter we summarize a number of issues related to this triad, aimed at the tissue engineering of dentin–pulp complexes, as well as details of preclinical studies of gene therapy to harness pulp stem cells with a morphogenetic signal, and bone morphogenetic protein (BMP) for dentin/pulp regeneration. Recent advances in the bioengineering of whole teeth are also discussed. Key words: tissue engineering, dentin regeneration, pulp regeneration, dental pulp stem/progenitor cells, odontoblasts, pulp exposure, gene therapy, cell therapy, bone morphogenetic proteins, scaﬀold.

17.1 Introduction

Dental caries, which causes an early loss of dental pulp and resultant tooth loss, is one of the most prevalent health problems worldwide. The teeth have signiﬁcant roles in mastication, speech and the aesthetics of the face, and all are critical components of well-being and a good quality of life. Dysfunction of the chewing apparatus may cause pain in the head, face, and jaw [1], while the condition of Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

266

17 Tissue Engineering of Teeth

dental occlusion is closely related to lower-extremity dynamic strength, agility, and balance function [2]. Masticatory function is also associated with cognitive status and vascular dementia in the elderly population [3]. The pulp tissue has functions in maintaining the homeostasis of teeth in terms of their vascular supply and innervation, and this is essential for the longevity of teeth. Reactionary/reparative dentin formation after noxious stimuli such as caries and immunological defense reactions against bacterial inﬁltration and external stimuli, is well known. Under deep caries and pulpal exposure, however, traditional direct pulp capping and pulpal amputation with calcium hydroxide often lead to unsatisfactory results due to bacterial inﬁltration and inﬂammation in the pulp tissue. Even if the reparative dentin is formed, its orientation is towards the apical region, and the remaining pulp becomes reduced, which in turn might cause diﬃculties in root canal access in case of further endodontic treatment. Thus, eﬀective vital pulp therapy with dentin and pulp regeneration should revolutionize the clinical management of deep caries and pulpal inﬂammation. In this respect, an enhancement of the natural healing potential of pulp tissue, together with the bioengineering of odontoblasts from stem/progenitor cells in vivo and ex vivo, might be promising. It is possible that tissue engineering and regenerative medicine for dentin and pulp might provide the ultimate goal of endodontics and operative dentistry, to retain metabolic function and reconstitute the occlusal and aesthetic functions of teeth.

17.2 The Triad

The triad for dentin and pulp regeneration is based on three basic components of biologic tissues: (i) responding cells; (ii) inductive morphogenetic signals such as bone morphogenetic proteins (BMPs); and (iii) an extracellular matrix (ECM) scaﬀold (Fig. 17.1). In recent years this triad has formed the focal point of several intense research programs, some details of which are described in the following sections. 17.2.1 Pulp Stem/Progenitor Cells 17.2.1.1 Isolation Pulp tissue contains a heterogeneous population of cells, including pulp cells, vascular endothelial cells, undiﬀerentiated mesenchymal cells, ﬁbroblasts, pericytes and the odontoblast diﬀerentiation hierarchy, as well as immune cells and neurocytes. Highly proliferative pulp cells have been isolated from adult pulp tissues of many diﬀerent species. Human clonogenic dental pulp stem cells (DPSCs)/progenitor cells, which have self-renewal capability and multi-lineage diﬀerentiation potential, have also been isolated using enzymatic disaggregation [4, 5]. Human DPSCs were further isolated by magnetic-activated cell sorting

17.2 The Triad

Fig. 17.1 The key elements of tissue engineering and dentin regeneration. The triad comprises stem cells, a scaﬀold of the extracellular matrix, and signals of morphogens.

using antibodies reactive to STRO-1, CD146, which is suggestive of a potential perivascular origin [6]. A distinct class of tissue-speciﬁc multipotent stem cells, known as the ‘‘Side Population’’ (SP) cells have been isolated on the basis of their ability to pump out Hoechst 33342, a DNA-binding ﬂuorescent dye [7]. SP cells up-regulate the genes implicated in the quiescent status, pluripotency and asymmetric division [8]. Pulp SP cells have also been isolated from many diﬀerent species [9] (Fig. 17.2). CD150 mRNA, which was shown recently to be a useful marker for distinguishing stem cells from other progenitors [10], was highly expressed in porcine pulp SP cells compared to non-SP cells [9]. The expression of smooth muscle cell a-actin, NG2 and desmin – all of which are markers for microvascular pericytes – was much lower in SP cells. CD31, Vegfr2 and CD105 endothelial markers were highly expressed in SP cells compared to non-SP cells. Bcrp1, which was attributed to the SP phenotype [11], was localized in the perivascular region [9], overlapping with CD31. However, 50% of the SP cells were CD31-positive, as assessed by ﬂow cytometry in porcine SP cells, which suggested a possible alternative niche other than the perivascular region in pulp tissue. The vast majority of SP cells from bone marrow possess cell-surface markers characteristic of hematopoietic stem cells (HSCs), c-Kit (CD117)þ , Sca-1þ , Thy1(CD90) low , CD31þ , CD135 and lineage . However, only a minority of the antigen-deﬁned subsets are SP cells, which suggests that the latter are not distinct from established stem cell phenotypes but rather form a small subset of

267

268

17 Tissue Engineering of Teeth

Fig. 17.2 Pulp side population (SP) cells from porcine adult pulp tissue. (A) Primary pulp cells contained 0.2% population with relatively low Hoechst 33342 ﬂuorescence, SP cells. (B) Colony formation derived from a single SP cell in 8 days. (C) Adipogenic conversion on day 34 (Oil Red O staining). (D) Chondrogenic conversion on day 28 (Alcian blue staining). (E) Neurogenic conversion on day 30 (immunostaining of neuromodulin).

them [12]. The SPþ subset of KTLSþ (c-Kit(CD117)þ , Thy-1(CD90) low , lineage , and Sca-1þ ) cells is more enriched in long-term repopulating cell content than the SP subset of KTLSþ [12]. Mouse pulp cells were also found to represent a similar SPþ subset of KTLSþ to bone marrow (unpublished data). Speciﬁc markers are further required to identify the diﬀerent hierarchy of pulp cells, which might resolve the question of whether all of the colony-forming cells are derived from a single pluripotent stem cell, or from committed progenitors belonging to several distinct lineages. 17.2.1.2 Self-Renewal Stem cells have self-renewal capability, with each cell division giving rise to one replacement stem cell and one transit amplifying cell by asymmetric cell division

17.2 The Triad

[13]. Alternatively, a stem cell gives rise to two daughter transit amplifying cells each time, while another stem cell might give rise to two daughter stem cells by symmetrical division [13]. The normal localization site of stem cells, the niche, provides a specialized environment which is composed of mesenchymal cells and ECM. Interactions within niches are crucial to the self-renewal process [13]. Oct4/POU3-4, which is implicated in the maintenance of a proliferative lifespan, was highly expressed in pulp SP cells compared with non-SP cells; the expression of Bmi1 and Stat3 was also high in pulp SP cells. The expression of telomerase reverse transcriptase (Tert), which maintains telomere length, and also preserves genomic stability and the long-term viability of highly proliferative organs [14], was higher in pulp SP cells [9]. The latter cells also had higher cumulative cell number and a longer proliferative life-span compared to non-SP cells, which suggests that they have a self-renewal, replicative capacity. The results of further, comprehensive study are awaited to detect the signaling components, Wnt, Notch and BMP in the niches of pulp tissue, as have been identiﬁed in other organs [15–17]. 17.2.1.3 Multipotential Diﬀerentiation Tissue stem cells have the potential to diﬀerentiate into other unrelated organs. For example, the multipotency of bone marrow stem cells to engraft into cardiac muscle, vascular endothelium, liver, and skeletal muscle has been described [18– 21]. Mesenchymal stem cells from tissues such as bone marrow and adipose tissue have multi-diﬀerentiation potential in vitro [22–24]; indeed, bone marrow cells even have the potential to give rise to dental mesenchymal-like cells [25]. The pulp stem cells are also multipotent (Fig. 17.2), representing adipogenic, chondrogenic, osteogenic, neurogenic and angiogenic as well as dentinogenic potential in vitro [4, 5, 9, 26, 27]. CD146-positive human pulp stem cells were shown to induce ectopic dentin formation after transplantation with hydroxyapatite/ tricalcium phosphate scaﬀold [6], while porcine SP cells induced dentin regeneration on the amputated pulp in vivo after implantation of the three-dimensional pellet treated with BMP2 [9]. 17.2.2 Morphogenetic Signals, BMPs

The key – and most promising – regulators in dentin and pulp regeneration are the family of BMPs [28, 29]. BMP family members are involved sequentially and repeatedly throughout embryonic tooth development, initiation, morphogenesis, cytodiﬀerentiation and matrix secretion [30]. Recently, we have cloned ten BMP family members [Bmp2, Bmp4, Bmp6, Bmp7, Bmp8, Growth/diﬀerentiation factor (Gdf )1, Gdf5, Gdf6, Gdf7, Gdf11 and glial cell line-derived neurotrophic factor (GDNF)] from rat incisor pulp tissue [31, 32]. BMP4 from the epithelium induces the mesenchyme to be odontogenic [33], while Bmp2, Bmp4 and Bmp7 signals are expressed in the enamel knot, which inﬂuences both epithelial and mesenchymal cells, and is responsible for maintenance of the enamel knot and subsequent morphogenesis of the epithelium [34]. These signals also regulate

269

270

17 Tissue Engineering of Teeth

the patterning of the tooth crown by inﬂuencing the initiation of secondary knots, together with mesenchymal signals such as BMP4 [35–37]. Bmp2, Bmp4, Bmp6, Bmp7, and Gdf11 are expressed during odontoblast diﬀerentiation [32, 38–40], and Bmp4 and Bmp5 during ameloblast diﬀerentiation. The BMP signaling networks are complex during morphogenesis. The BMP signaling pathway is regulated at three levels: intracellular domains; membrane sites; and extracellular sites [41]. The BMP antagonists, noggin, follistatin and ectodin, modulate the bioavailability of BMPs in tooth development [30, 42–44]. Two transmembrane receptors – type I (ActR-IA, BMPR-IA and BMPR-IB) and type II (ActR-IIA, ActR-IIB, and BMPR-II) – each with serine-threonine kinase activity, are expressed during tooth development [30, 45]. BMP signals are transduced from the plasma membrane to the nucleus through a limited number of Smad proteins, receptor-activated Smads (R-Smads), common mediator Smads (co-Smads) and inhibitory Smads (I-Smads) during tooth morphogenesis [46]. BMP2 stimulates diﬀerentiation of the pulp cells into odontoblasts in monolayer cultures, in 3-D cultures, and in organ culture [47–50]. Recombinant human BMP7 elicited similar eﬀects in cultured tooth slices [51], while recombinant human BMP2, BMP4 and GDF11, when soaked in beads, also stimulated odontoblast diﬀerentiation in organ cultures of mouse dental papillae cells [33, 52]. It is still unclear how the abrupt transition of stem/progenitor cells from the quiescent to the active state is regulated, in terms of proliferation, migration, differentiation, and matrix secretion following pulp tissue injury. Clearly, the molecular control mechanisms underlying morphogen release need to be elucidated for therapeutic use in regenerative endodontics. 17.2.3 Scaﬀold

Interactions between the inductive signaling by morphogens and responding cells are modulated by the ECM, the components of which tether active morphogens to confer an optimal conformation and perhaps protect them from proteolysis [53]. BMPs bind to collagen type I and type IV, procollagen type II, heparan sulfate, heparin, ﬁbrillins, proteoglycans, as well as to BMP antagonists such as noggin, chordin and DAN [54]. The critical functions of the scaﬀold are to: provide a physico-chemical and biological 3-D microenvironment for cell growth and diﬀerentiation, promoting cell adhesion and migration; serve as a carrier for morphogens such as BMPs in protein therapy, and for cells in cell therapy; be eﬀective for the transport of nutrients, oxygen, and waste; be gradually degraded and replaced by regenerative tissue, retaining the feature of the ﬁnal tissue structure; and possess biocompatibility, non-toxicity, and proper physical and mechanical strength.

17.3 Dentin Regeneration

Some strategies for reparative and regenerative dentin formation involve the integration of prefabricated scaﬀolds suited for dentinogenesis. Scaﬀolds are constructed from either natural or synthetic polymers. The natural polymers (e.g., collagen and ﬁbronectin) have the advantages of good biocompatibility and bioactivity, whereas the synthetic polymers enable precise control over the physicochemical properties such as degradation rate, porosity, microstructure, and mechanical properties [55]. The diﬀerentiation of odontoblasts is seen in contact with a collagenous matrix of osteodentin formed by a response to calcium hydroxide pulp capping [56]. A tubular matrix lined with odontoblasts is also formed when demineralized or native dentin matrix is implanted in the pulp tissue [57– 60]. Demineralized dentin matrix is inactivated by 4 M guanidine hydrochloride. BMP2, together with this scaﬀold, induces tubular dentin formation in the cavity on the amputated pulp. However, a tendon collagen scaﬀold does not induce tubular formation [61, 62]. Therefore, the presence of an insoluble dentin matrix seems to be important in order for pulp cells to attach and diﬀerentiate into odontoblasts during reparative dentinogenesis. The precise physico-chemical surface requirement, however, has not yet been elucidated. Fibronectin might mediate interactions between these substrates and pulp cells and reorganize the cytoskeleton during odontoblast polarization [63]. Inter-odontoblastic collagen ﬁbers might help to guide the pre-odontoblasts in their migration, adhesion, and arrangement and reparative dentin formation, providing a biological/mechanical tie that binds the ﬁbers of the central pulp and newly formed odontoblasts [64]. Bone sialoprotein is capable of stimulating the diﬀerentiation of cells which secrete an organized ECM and a thick reparative dentin matrix formation [65]. Certain synthetic scaﬀolds may have application in dentin/pulp regeneration [66, 67]; here, the most frequently used synthetic materials are poly(lactic acid) (PLA), poly(glycolic acid) (PGA), and their co-polymer, poly(lactic-co-glycolic acid) (PLGA). Synthetic hydrogels include poly(ethylene glycol) (PEG)-based polymers, and those modiﬁed with cell-surface adhesion peptides, such as arginine, glycine, and aspartic acid (RGD), can improve cell adhesion and matrix synthesis within the 3-D network [68]. Scaﬀolds containing inorganic compounds such as hydroxyapatite and calcium phosphate are used to enhance mineralized tissue conductivity [69]. In view of the role of the 3-D cellular microenvironment, scaﬀolds will need to be improved in the quest to regulate the regeneration of dentin.

17.3 Dentin Regeneration 17.3.1 Protein Therapy

Protein therapy with some BMPs, such as recombinant human BMP2, BMP4, and BMP7, has been established to induce reparative/regenerative dentin formation [28, 29, 70]. However, a requirement of large doses of recombinant BMPs,

271

272

17 Tissue Engineering of Teeth

coupled with their short half-lives, have suggested that the delivery system needs to be optimized, and this has not yet been achieved. 17.3.2 Gene Therapy 17.3.2.1 In-Vivo BMP Gene Therapy Gene therapy represents an alternative strategy, and it is clinically possible to apply BMPs in dentin regeneration. The main components for developing an eﬃcient and safety gene delivery system in dentin regeneration are: (i) higher transduction rates into targeted cells; (ii) suﬃcient long-term, time-controlled and localized expression of transgene; and (iii) safety ensured, minimal attendant immune responses and/or toxicity. The use of both viral and non-viral vectors has been reported to express transgene and produce endogenous BMPs in dental pulp tissue. In-vivo direct infection (Fig. 17.3, left) with adenovirus containing a full-length Bmp7 gene induced only small amounts of poorly organized mineralizing masses in dental pulp tissue; reversible pulp is produced experimentally by the injection of bacterial lipopolysaccharide [71]. Initially, a non-viral gene delivery approach was used, by electroporation, to induce dentin regeneration. This method has certain advantages over the viral gene delivery, namely that: it permits and stable manufacture/manipulation at low cost and with a high level of purity, and minimal risk of replication or incorporation;

Fig. 17.3 The two main strategies for pulp therapy to regenerate dentin. Left: For the in-vivo method, the natural healing potential of pulp tissue is enhanced by direct local application of BMP proteins or BMP genes into the exposed dental pulp. Right: For the ex-vivo method, stem/ progenitor cells are ﬁrst isolated, treated with BMP proteins or BMP genes to diﬀerentiate into odontoblasts on the surface of scaﬀold, and then transplanted onto the exposed pulp.

17.3 Dentin Regeneration

the inserted size of transgene is unlimited; there is weak immunogenicity; it allows episomal, transient insertion into the host cells; repeated administration is possible; and there is no limited dose of the vector for administration [72].

The eﬃciency of gene transduction is relatively low, although electroporation using pulsed electric ﬁelds eﬃciently transduced a Gdf11 cDNA plasmid into pulp cells in an organ culture of mouse dental papilla, with a resultant induced diﬀerentiation of odontoblasts in vitro [52]. Gdf11 gene transfer in the amputated dental pulp, however, induced only a small amount of reparative dentin, mainly due to the electrode impairing the pulp tissue by virtue of its thermal and physical invasiveness, and this resulted in incomplete and non-homogeneous osteodentin formation [52]. The next procedure to be assessed was ultrasoundmediated gene delivery, using microbubbles. The non-invasiveness, non-toxicity, safety (as conﬁrmed by diagnostic ultrasound equipment), extreme localized transduction and simple operation proved advantageous for gene therapy with sonoporation in dentin regeneration. For the transfer of plasmids by sonoporation (with microbubbles) into amputated/exposed pulp tissue, Optison (at a concentration of 5–10%) provided the most eﬃcient method. Unlike electroporation, ultrasound-mediated Gdf11 plasmid gene transfer induced the diﬀerentiation of pulp stem cells into odontoblasts in vitro, and subsequent homogeneous and complete reparation of dentin in vivo, without causing any tissue damage or cell necrosis [73]. 17.3.2.2 Ex-Vivo BMP Cell Therapy and Gene Therapy From a clinical viewpoint, when dental pulp tissue is accidentally exposed during the removal of caries, and/or if inﬂammation is localized in the coronal region, ultrasound-mediated in-vivo gene therapy with BMP might be eﬀective for dentin regeneration. In case of severe, irreversible pulpitis however, few stem/progenitor cells are left in the dental pulp and the eﬃcacy of in-vivo gene therapy might be questioned [71]. Hence, an alternative approach must be attempted, namely exvivo cell therapy or gene therapy (Fig. 17.3, right). In this case, pulp stem/progenitor cells are transduced with BMP protein or BMP gene to diﬀerentiate into odontoblasts in vitro, and are then transplanted into the amputated/exposed pulp tissue (Fig. 17.3, right). Rutherford [71] has shown that cultured dermal ﬁbroblasts, when transduced with adenovirus-mediated Bmp7, induced reparative dentin formation following autogenous transplantation onto the exposed pulp with pulpitis. The present authors’ studies have also shown that cell therapy and ex-vivo gene therapy could initially induce osteodentin, followed by longlasting tubular dentin formation on experimentally exposed pulp, either by transplanting BMP2-treated pulp stem/progenitor cells [49] or Gdf11-electrotransfected pulp stem/progenitor cells [74] in a 3-D pellet, respectively (Fig. 17.4). Those cells treated with BMP proteins or transduced with Bmp genes directly contributed to reparative/regenerative dentin matrix formation after transplantation. At the

273

274

17 Tissue Engineering of Teeth

Fig. 17.4 Reparative/regenerative dentin formation by cell therapy and ex-vivo gene therapy. Pulp stem/progenitor cells were treated with recombinant BMP2 (A) or electrotransfected with Gdf11 gene (B) to diﬀerentiate into odontoblasts in the 3-D pellet culture, and autogenously transplanted on the amputated pulp. Large amounts of osteodentin (OD) and tubular dentin (TD) were formed after 1 month (1M) and 3 months (3M).

same time, it is likely that various growth/diﬀerentiation factors were secreted by the cells and retained in their matrix. These were gradually released from the transplantation pellet, and resulted in a further induction of host pulp migrating cells into odontoblast diﬀerentiation [74]. The autogenously transplanted cells had the potential to integrate seamlessly into the exposed pulp tissue, without disrupting normal function. Moreover, a regenerative response of host pulp cells was observed following the transplantation of BMP-transduced cells even under pulpitis, demonstrating the eﬃcacy of this ex-vivo approach to dentin regeneration for the clinical treatment of endodontic teeth. Whilst tissue-engineered dentin/pulp grafts might serve as promising substitutes for artiﬁcial dental materials (such as resin and metal), the main challenge for the clinical application of ex-vivo gene therapy is that of the cell source. The most reliable cell source is that of autologous pulp stem/progenitor cells isolated from teeth which have been extracted for the purpose of orthodontic treatment. Such sources, which include the third molar and deciduous teeth, have certain limitations, mainly that the stem cells are rarely available from middle-aged and elderly patients, only limited numbers are available, and that they cannot be completely puriﬁed, expanded and maintained for long periods without altering the phenotype capable of diﬀerentiation in odontoblasts after BMP transduction. However, a recent study demonstrated long-term cryopreserved properties of pulp stem cells [27], and suggested the introduction of a ‘‘stem cell bank’’ where ther-

17.3 Dentin Regeneration

apeutic individual pulp stem cells isolated from useless teeth at a younger age could be preserved until their clinical use later in life. In this respect, the immunosuppressive activity of pulp stem cells [75] suggested a potential use for allogenic pulp stem cells. This improved and reﬁned safety, coupled with eﬃcient isolation and expansion of pulp stem cells requires a better understanding of stem cell biology. Hence, in time it may be possible to identify pulp stem cell markers of diﬀerent diﬀerentiation hierarchies, and to harvest and transduce the most preferable mesenchymal stem cells from sources such as fat, bone marrow and skeletal muscle, in a minimally invasive manner, for subsequent odontoblast diﬀerentiation and dentin regeneration. The preclinical data reported by the present authors have demonstrated the safety and eﬃcacy of ex-vivo cell therapy and gene therapy of BMPs (Fig. 17.3). However, the pharmacokinetics of the transplanted pulp cells must be examined in greater detail prior to clinical use, and the integrity and permeability of the reparative dentin must be evaluated systematically. It should be noted that neither inﬂammatory cell inﬁltration nor leakage were observed in the pulp tissue, despite complete exposure of the reparative dentin for 3 months following the exvivo gene therapy, which indicates the formation of a functional physical barrier

Fig. 17.5 Tubular dentin formation with optimal orientation for clinical application. After the removal of caries, a deﬁned scaﬀold is made from the mold. The pulp stem cells treated with BMP protein or transfected with BMP gene attach to the scaﬀold and diﬀerentiate into odontoblasts, which leads to the formation of a functional tubular dentin–pulp complex in 3-D culture. This is then transplanted onto the exposed pulp.

275

276

17 Tissue Engineering of Teeth

to protect the pulp. Tubular dentin formation with optimal orientation might provide a better permeability barrier, although the orientation of tubular dentin depends upon the scaﬀold. Recent advances in novel biomaterials and nanomaterials may lead to further progress in this area, however. In time, the use of stem cells on a deﬁned scaﬀold may yield functional dentin–pulp complexes that can be transplanted onto the exposed pulp (Fig. 17.5).

17.4 Pulp Regeneration 17.4.1 Vasculogenesis

The critical importance of vasculature in tissue repair and regeneration is well known, and vascular inﬁltration to the engineered graft contributes greatly to regeneration of the dentin–pulp complex. Vascular endothelial growth factor (VEGF), which is a prominent regulator of angiogenesis, induces chemotaxis, proliferation and diﬀerentiation of human dental pulp cells [76, 77]. Thus, the presence of VEGF in dentin matrix [78] and the response of dental pulp cells to VEGF raises the possibility endothelial progenitor cells being present in the dental pulp, alongside progenitors for odontoblasts and neuronal cells [9, 79]. In view of the role of endothelial progenitor cells in vascularization during tissue regeneration, it is likely that VEGF and vascular endothelial cells are critical for dentin and pulp regeneration. The results of a recent study showed that pulp stem/ progenitor cells migrate to the injury site in response to injury to endothelial cells [80]. VEGF, the secretion of which is stimulated by both BMP4 [81] and BMP7 [82], up-regulates BMP2 expression in capillary endothelial cells [83], and interacts synergistically with BMP2 and BMP4 to stimulate the recruitment of stem cells [84]. BMPs stimulate angiogenesis through the production of VEGF by osteoblasts [85]. It is possible that cross-talk occurs between endothelial cells and odontoblasts in dentin–pulp regeneration, as has been suggested in bone formation and fracture healing. The therapeutic treatment of neovascularization opens unprecedented opportunities for life-threatening diseases as, well as bone disease [86, 87]. The use of gene therapy to stimulate vascular growth [88] permits the local stimulation of vascularization during regeneration. In this way, an enhanced regeneration of dentin–pulp complex by in-vivo gene therapy of VEGF, together with an appropriate ratio of BMPs or ex-vivo gene therapy using pulp stem cells or endothelial progenitor cells transduced from these genes, should demonstrate great clinical potential in endodontic treatments. 17.4.2 Neurogenesis

The dental pulp is richly innervated, the nerve supply entering through the apical foramen along with the vascular elements. The nerves then proceed to the coro-

17.5 Whole-Teeth Regeneration

nal area, where they form a plexus in proximity to the odontoblasts, and ﬁnally enter the dentinal tubules. Both, sensory and sympathetic nerves are involved [79]. The pulpal nerves play a key role in the regulation of blood ﬂow, dentinal ﬂuid ﬂow and pressure [89–91], as well as contributing to angiogenesis, the extravasation of immune cells and the regulation of inﬂammation to minimize initial damage, maintaining pulp tissue and strengthening the pulpal defense mechanisms [92]. Thus, neuro-pulpal interactions and nerve regeneration are essential for successful gene therapy in endodontic treatment. Nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF) and glial cell line-derived neurotrophic factor (GDNF) are all expressed in dental pulp [93], with GDNF also being transported retrogradely to trigeminal neuron cell bodies from dental pulp [94]. The pronounced eﬀects of BMP members on neurogenesis [95–102] support their demonstrated eﬃcacy on nerve regeneration in gene therapy for dentin–pulp complex regeneration. Recent developments into the mechanism(s) of the neurotrophism of pulp stem cells [93] should, in time, ensure further advances into the regeneration of nerves, based on neuro-pulpal interactions.

17.5 Whole-Teeth Regeneration

Tissue engineering techniques have been utilized in the management of lost teeth, tooth autotransplantation and dental implants. Although the development of whole tooth crowns from ectopically transplanted embryonic tooth buds has long been recognized [103–106], it was only recently that bioengineered teeth were successfully generated from dissociated cells of pig tooth bud or rat cultured tooth bud cells seeded onto biodegradable polymer scaﬀolds and grown in the omenta [107–109]. The entire tooth structure, including the enamel-covered dentin and cementum-covered dentin, was formed. This process was mediated by an epithelial–mesenchymal interaction similar to that occurring in natural tooth development, and derived from the epithelial and mesenchymal stem/progenitor cells. Computer-generated models of the bioengineered 21-week tooth crowns revealed a normal spatial organization of enamel, dentin and pulp tissues, although the enamel failed to cover the entire crown surface and the dentin did not enclose the entire pulp center [110]. The transplantation of dissociated tooth bud cells into the socket of an extracted tooth bud in dogs, albeit with a poorer blood supply compared with the omentum, led to an induced tubular dentin regeneration [111]. One possible problem when transplanting dissociated tooth bud cells with a polymer scaﬀold is that the bioengineered teeth are generally small, and the shape and coordinated formation of enamel and cementum cannot be regulated [109]. The predictable formation of a precisely shaped molar or incisor remains a daunting challenge, although in this respect physical and mechanical forces might play a role, in addition to the epithelial–mesenchymal interaction. More recently, Honda et al. [112] showed that exposure to shear stress for 2 h facilitates odontogenic cell diﬀerentiation in vitro, leading subsequently to early regeneration of the bioengineered tooth in vivo.

277

278

17 Tissue Engineering of Teeth

A critical function of the cap-stage mesenchyme in controlling tooth morphogenesis and inducing epithelial cell plasticity (i.e., an ability to undergo conversion between diﬀerent epithelial cell types) was demonstrated by performing dissociation–re-association experiments. Dissociated cap-stage mouse tooth epithelia were re-associated with a dental mesenchyme or dissociated mesenchymal cells, and then cultured in vitro; this resulted in complete tooth morphogenesis, despite a loss of positional information [113, 114]. Further experiments have shown that the number of dissociated mesenchymal cells reassociated with the epithelial compartment might modulate the shape of the crown [115]. A recent exciting report also claimed that the embryonic oral epithelium was able to direct odontogenesis when recombined with non-dental ectomesenchymal stem cells in adult renal capsules [25]. On this basis, it was suggested that odontogenic signals might be able to instruct tooth crown formation in embryonic or adult stem cells, without the presence of a scaﬀold [116]. However, a major – as yet unresolved – problem here is to deﬁne accurately the stage and source of oral epithelium and/or ectomesenchymal stem cells for potential clinical use.

17.6 Conclusions and Future Perspectives

Despite extensive progress having been made in tissue engineering during recent years, there are currently no clear mechanical and biological means of replicating the structure and function of small-sized teeth, including dentin, enamel, cementum and dental pulp, in the routine clinical treatment of caries and/or missing teeth. Advances in stem/progenitor cell sourcing, together with a better understanding of the ECM scaﬀold, will be crucial to the success of such clinical applications. However, it seems likely that, through the development of bioreactor design and highly eﬀective scaﬀolds, a bioengineered prosthetic tooth will, in time, be developed, thereby providing an ‘‘oﬀ-the-shelf ’’ cell product. Recent spectacular advances in morphogenetic signals, responding stem cells and biomimetic scaﬀolds augur well for the future of tooth tissue engineering. References 1 M.M. Siccoli, C.L. Bassetti, P.S.

4 S. Gronthos, M. Mankani, J.

Sandor, Lancet Neurol. 2006, 5, 257– 267. 2 T. Yamaga, A. Yoshihara, Y. Ando, Y. Yoshitake, Y. Kimura, M. Shimada, M. Nishimuta, H. Miyazaki, J. Gerontol A. Biol. Sci. Med. Sci. 2002, 57, M616–M620. 3 H. Miura, K. Yamasaki, M. Kariyasu, K. Miura, Y. Sumi, J. Oral Rehabil. 2003, 30, 808–811.

Brahim, P.G. Robey, S. Shi, Proc. Natl. Acad. Sci. USA 2000, 97, 13625–13630. 5 S. Gronthos, J. Brahim, W. Li, L.W. Fisher, N. Cherman, A. Boyde, P. DenBesten, P.G. Robey, S.J. Shi, Dent. Res. 2002, 81, 531– 535. 6 S. Shi, S. Gronthos. J. Bone Miner. Res. 2003, 18, 696–704.

References 7 M.A. Goodell, K. Brose, G. Paradis,

8

9

10

11

12

13

14 15 16 17 18

19

20

21 22

23 24

A.S. Conner, R.C. Mulligan, J. Exp. Med. 1996, 183, 1797–1806. C. Rochon, V. Frouin, S. Bortoli, K. Giraud-Triboult, V. Duverger, P. Vaigot, C. Petat, P. Fouchet, B. Lassalle, O. Alibert, X. Gidrol, G. Pietu, Exp. Cell Res. 2006, 312, 2074– 2082. K. Iohara, L. Zheng, M. Ito, A. Tomokiyo, K. Matsushita, M. Nakashima, Stem Cells 2006, 11, 2493–2503. M.J. Kiel, O.H. Yilmaz, T. Iwashita, O.H. Yilmaz, C. Terhorst, S.J. Morrison, Cell 2005, 121, 1109–1121. S. Zhou, J.J. Morris, Y. Barnes, L. Lan, J.D. Schuetz, B.P. Sorrentino, Proc. Natl. Acad. Sci. USA 2002, 99, 12339–12344. D.J. Pearce, C.M. Ridler, C. Simpson, D. Bonnet, Blood 2004, 103, 2541– 2546. M.R. Alison, M. Brittan, M.J. Lovell, N.A. Wright, Handbook Exp. Pharmacol. 2006, 174, 185–227. C. Autexier, N.F. Lue, Annu. Rev. Biochem. 2006, 75, 493–517. L. Li, T. Xie, Annu. Rev. Cell Dev. Biol. 2005, 21, 605–631. K.A. Moore, I.R. Lemischka, Science 2006, 311, 1880–1885. A. Wilson, A. Trumpp, Nat. Rev. Immunol. 2006, 6, 93–106. K.A. Jackson, S.A. Majka, H. Wang, J. Pocius, C.J. Hartley, M.W. Majesky, M.L. Entman, L.H. Michael, K.K. Hirschi, M.A. Goodell, J. Clin. Invest. 2001, 107, 1395–1402. F.D. Camargo, R. Green, Y. Capetanaki, K.A. Jackson, M.A. Goodell, Nat. Med. 2003, 9, 1520–1527. G.G. Wulf, K.L. Luo, K.A. Jackson, M.K. Brenner, M.A. Goodell, Haematologica 2003, 88, 368–378. D.N. Kotton, R. Summer, A. Fine, Exp. Hematol. 2004, 32, 340–343. M. Reyes, T. Lund, T. Lenvik, D. Aguiar, L. Koodie, C.M. Verfaillie, Blood 2001, 98, 2615–2625. A.I. Caplan, Tissue Eng. 2005, 11, 1198–1211. B.M. Strem, K.C. Hicok, M. Zhu, I. Wulur, Z. Alfonso, R.E. Schreiber,

25

26

27

28 29 30

31

32

33 34 35

36

37 38 39 40

41 42

J.K. Fraser, M.H. Hedrick, Keio J. Med. 2005, 54, 132–141. A. Ohazama, S.A.C. Modino, I. Miletich, P.T. Sharpe, J. Dent. Res. 2004, 83, 518–522. G. Laino, A. Graziano, R. d’Aquino, G. Pirozzi, V. Lanza, S. Valiante, A. De Rosa, F. Naro, E. Vivarelli, G. Papaccio, J. Cell Physiol. 2006, 206, 693–701. G. Papaccio, A. Graziano, R. d’Aquino, M.F. Graziano, G. Pirozzi, D. Menditti, A. De Rosa, F. Carinci, G. Laino, J. Cell. Physiol. 2006, 208, 319–325. M. Nakashima, A.H. Reddi, Nat. Biotech. 2003, 21, 1025–1032. M. Nakashima, Cytokine Growth Factor Rev. 2005, 16, 369–376. K. Heikinheimo, C. Begue-Kirn, O. Ritvos, T. Tuuri, J.V. Ruch, Eur. J. Oral Sci. 1998, 106, 167–173. M. Nakashima, T. Toyono, T. Murakami, A. Akamine, Arch. Oral Biol. 1998, 43, 745–751. M. Nakashima, T. Toyono, A. Akamine, A. Joyner, Mech. Dev. 1999, 80, 185–189. S. Vainio, I. Karavanova, A. Jowett, I. Thesleﬀ, Cell 1993, 75, 45–58. I. Thesleﬀ, J. Cell Sci. 2003, 116, 1647–1648. I. Thesleﬀ, J. Jernvall, Cold Spring Harbor Symp. Quant. Biol. 1997, 62, 257–267. J. Jernvall, T. A˚berg, P. Kettunen, S. Keranen, I. Thesleﬀ, Development 1998, 125, 161–169. I. Thesleﬀ, M. Mikkola, Int. Rev. Cytol. 2002, 217, 93–135. K. Heikinheimo, J. Dent. Res. 1994, 73, 590–597. T. A˚berg, J. Wozney, I. Thesleﬀ, Dev. Dynam. 1997, 210, 383–396. M.N. Helder, H. Karg, T.J. Bervoets, S. Vukicevic, E.H. Burger, R.N. D’Souza, J.H. Woltgens, G. Karsenty, A.L. Bronckers, J. Dent. Res. 1998, 77, 545–554. W. Balemans. W. Van Hul, Dev. Biol. 2002, 250, 231–250. A.S. Tucker, K.L. Matthews, P.T. Sharpe, Science 1998, 282, 1136– 1138.

279

280

17 Tissue Engineering of Teeth 43 Y. Kassai, P. Munne, Y. Hotta, E.

44

45

46

47

48

49

50 51 52

53

54 55 56 57 58 59 60 61 62 63

Penttila, K. Kavanagh, N. Ohbayashi, S. Takada, I. Thesleﬀ, J. Jernvall, N. Itoh, Science 2005, 309, 2067–2070. M.V. Plikus, M. Zeichner-David, J.A. Mayer, J. Reyna, P. Bringas, J.G. Thewissen, M.L. Snead, Y. Chai, C.M. Chuong, Evol. Dev. 2005, 7, 440–457. A. Nadiri, S. Kuchler-Bopp, F. PerrinSchmitt, H. Lesot, Cell Tissue Res. 2006, 324, 33–40. X. Xu, L. Jeong, J. Han, Y. Ito, P. Bringas, Y. Chai, Int. J. Dev. Biol. 2003, 47, 31–39. C. Be´gue-Kirn, A.J. Smith, J.V. Ruch, J.M. Wozney, A. Purchio, D. Hartmann, H. Lesot, Int. J. Dev. Biol. 1992, 36, 491–503. M. Nakashima, H. Nagasawa, Y. Yamada, A.H. Reddi, Dev. Biol. 1994, 162, 18–28. K. Iohara, M. Nakashima, M. Ito, K. Ishikawa, A. Nakasima, A. Akamine, J. Dent. Res. 2004, 83, 590–595. T. Saito, M. Ogawa, Y. Hata, K. Bessho, J. Endod. 2004, 30, 205–208. A.J. Sloan, R.B. Rutherford, A.J. Smith, Arch. Oral Biol. 2000, 45, 173–177. M. Nakashima, K. Mizunuma, T. Murakami, A. Akamine, Gene Ther. 2002, 9, 814–818. V.M. Paralkar, A.K. Nandedkar, R.H. Pointer, H.K. Kleinman, A.H. Reddi, J. Biol Chem. 1990, 265, 17281– 172814. A.H. Reddi, Arthritis Res. 2001, 3, 1–5. B. Sharma, J.H. Elisseeﬀ, Ann. Biomed. Eng. 2004, 32, 148–59. U. Schro¨der, J. Dent. Res. 1985, 64, 541–548. G. Anneroth, G. Bang, Odontologisk. Revy. 1972, 23, 315–328. M. Nakashima, Endod. Dent. Traumatol. 1989, 5, 279–286. M. Nakashima, Arch. Oral Biol. 1990, 35, 277–281. D. Tziafas, T. Lambrianidis, P. Beltes, J. Endod. 1993, 19, 116–122. M. Nakashima, J. Dent. Res. 1994, 73, 1515–22. M. Nakashima, Arch. Oral Biol. 1994, 39, 1085–1089. D. Tziafas, A.J. Smith, H. Lesot, J. Dentistry 2000, 28, 77–92.

64 Y. Kitasako, S. Shibata, C.F. Cox, J.

65

66

67

68 69

70

71 72 73

74

75

76

77

78 79

Tagami, Int. Endo. J. 2002, 35, 996– 1004. F. Decup, N. Six, B. Palmier, D. Buch, J.J. Lasfargues, E. Salih, M. Goldberg, Clin. Oral Invest. 2000, 4, 110–119. K.S. Bohl, J. Shon, B. Rutherford, D.J. Mooney, J. Bio Science, Polymer Edition 1998, 9, 749–764. M. Goldberg, N. Six, F. Decup, D. Buch, E. Soheili Majd, J.J. Lasfargues, E. Salih, L. Stanislawski, Adv. Dent. Res. 2001, 15, 91–95. J.A. Burdick, K.S. Anseth, Biomaterials 2002, 23, 4315–4323. J.A. Jadlowiec, A.B. Celil, J.O. Hollinger, Expert Opin. Biol. Ther. 2003, 3, 409–423. R.B. Rutherford, in: R.P. Lanza, R. Langer, L. Vacanti (Eds.), Principles of Tissue Engineering. 2nd edn. Academic Press, San Diego, 2000, pp. 847–853. R.B. Rutherford, Eur. J. Oral Sci. 2001, 109, 422–424. K.Y. Ng, Y. Liu, Medicinal Res. Rev. 2002, 22, 204–223. M. Nakashima, K. Tachibana, K. Iohara, M. Ito, M. Ishikawa, A. Akamine, Human Gene Ther. 2003, 14, 591–597. M. Nakashima, K. Iohara, M. Ishikawa, M. Ito, A. Tomokiyo, T. Tanaka, A. Akamine, Hum. Gene Ther. 2004, 15, 1045–1053. L. Pierdomenico, L. Bonsi, M. Calvitti, D. Rondelli, M. Arpinati, G. Chirumbolo, E. Becchetti, C. Marchionni, F. Alviano, V. Fossati, N. Staﬀolani, M. Franchina, A. Grossi, G.P. Bagnara, Transplantation 2005, 80, 836–842. K. Matsushita, R. Motani, T. Sakuta, N. Yamaguchi, T. Koga, K. Matsuo, S. Nagaoka, K. Abeyama, I. Maruyama, M. Torii, J. Dent. Res. 2000, 79, 1596– 1603. L. Artese, C. Rubini, G. Ferrero, M. Fioroni, A. Santinelli, A. Piattelli, J. Endod. 2002, 28, 20–23. D.J. Roberts-Clark, A.J. Smith, Arch. Oral Biol. 2000, 45, 1013–1016. M.R. Byers, H. Suzuki, T. Maeda, Microsc. Res. Tech. 2003, 60, 503–515.

References 80 S. Mathieu, A. El-Battari, J. Dejou, I.

81

82 83

84

85

86

87 88 89 90

91 92 93

94

95

96 97

98

About, Arch. Oral Biol. 2005, 50, 109–113. H. Tokuda, D. Hatakeyama, S. Akamatsu, K. Tanabe, M. Yoshida, T. Shibata, O. Kozawa, Arch. Biochem. Biophys. 2003, 415, 117–125. L.C. Yeh, J.C. Lee, Mol. Cell. Endocrinol. 1999, 153, 113–124. P.J. Bouletreau, S.M. Warren, J.A. Spector, Z.M. Peled, R.P. Gerrets, J.A. Greenwald, M.T. Longaker, Plast. Reconstr. Surg. 2002, 109, 2384–2397. H. Peng, A. Usas, A. Olshanski, A.M. Ho, B. Gearhart, G.M. Cooper, J. Huard, J. Bone Miner. Res. 2005, 20, 2017–2027. M.M. Deckers, R.L. van Bezooijen, G. van der Horst, J. Hoogendam, C. van Der Bent, S.E. Papapoulos, C.W. Lowik, Endocrinology 2002, 143, 1545–1553. Y. Cao, A. Hong, H. Schulten, M.J. Post, Cardiovasc. Res. 2005, 65, 639– 648. P. Madeddu, Exp. Physiol. 2005, 90, 315–326. S. Yla¨-Herttuala, K. Alitalo, Nat. Med. 2003, 9, 694–701. L.M. Olgart, J. Dent. Res. 1985, 64, 572–578. S. Kim, M. Liu, S. Simchon, J.E. Dorscher-Kim, Proc. Finn. Dent. Soc. 1992, 88, 387–392. N. Vongsavan, B. Matthews, Proc. Finn. Dent. Soc. 1992, 88, 491–497. I. Fristad, Acta Odontol. Scand. 1997, 55, 236–254. I.V. Nosrat, C.A. Smith, P. Mullally, L. Olson, C.A. Nosrat, Eur. J. Neurosci. 2004, 19, 2388–2398. I.H. Kvinnsland, K. Luukko, I. Fristad, P. Kettunen, D.L. Jackson, K. Fjeld, C.S. von Bartheld, M.R. Byers, Eur. J. Neurosci. 2004, 19, 2089–2098. P. Lein, X. Guo, A.M. Hedges, D. Rueger, M. Johnson, D. Higgins, Int. J. Dev. Neurosci. 1996, 14, 203–215. P.C. Mabie, M.F. Mehler, J.A. Kessler, J. Neurosci. 1999, 19, 7077–7088. C. Schneider, H. Wicht, J. Enderich, M. Wegner, H. Rohrer, Neuron 1999, 24, 861–870. K. Nakashima, T. Taga, Mol. Neurobiol. 2002, 25, 233–244.

99 H.H. Wu, S. Ivkovic, R.C. Murray, S.

100

101

102

103 104

105 106 107

108

109

110

111

112

113

114

115

116

Jaramillo, K.M. Lyons, J.E. Johnson, A.L. Calof, Neuron 2003, 37, 197–207. K. Tsuchida, Curr. Drug Targets Immune Endocr. Metabol. Disord. 2004, 4, 157–166. G. Ge, D.R. Hopkins, W.B. Ho, D.S. Greenspan, Mol. Cell. Biol. 2005, 25, 5846–5858. J. Kim, H.H. Wu, A.D. Lander, K.M. Lyons, K.M. Matzuk, A.L. Calof, Science 2005, 308, 1927–1930. H.C. Slavkin, L.A. Bavetta, Arch. Oral Biol. 1981, 26, 303–307. M. Yamada, P. Bringas, Jr., M. Grodin, M. MacDougall, E. Cummings, J. Grimmett, B. Weliky, H.C. Slavkin, J. Biol. Buccale 1980, 8, 127–139. M. Mina, E.J. Kollar, Arch. Oral Biol. 1987, 32, 123–127. H.F. Thomas, E.J. Kollar, Arch. Oral Biol. 1989, 34, 27–35. C.S. Young, S. Terada, J.P. Vacanti, M. Honda, J.D. Bartlett, P.C. Yelick, J. Dent. Res. 2002, 81, 695–700. M.T. Duailibi, S.E. Duailibi, C.S. Young, J.D. Bartlett, J.P. Vacanti, P.C. Yelick, J. Dent. Res. 2004, 83, 523– 528. M.J. Honda, Y. Sumita, H. Kagami, M. Ueda, Arch. Histol. Cytol. 2005, 68, 89–101. C.S. Young, S.W. Kim, C. Qin, O. Baba, W.T. Butler, R.R. Taylor, J.D. Bartlett, J.P. Vacanti, P.C. Yelick, Arch. Oral Biol. 2005, 50, 259–265. M.J. Honda, T. Ohara, Y. Sumita, T. Ogaeri, H. Kagami, M. Ueda, J. Oral Maxillofac. Surg. 2006, 64, 283–289. M.J. Honda, Y. Shinohara, Y. Sumita, A. Tonomura, H. Kagami, M. Ueda, Bone 2006, 39, 125–133. B. Hu, A. Nadiri, S. Bopp-Kuchler, F. Perrin-Schmitt, H. Lesot, J. Dent. Res. 2005, 84, 521–525. B. Hu, A. Nadiri, S. Bopp-Kuchler, F. Perrin-Schmitt, S. Wang, H. Lesot, Arch. Oral Biol. 2005, 50, 131–136. B. Hu, A. Nadiri, S. Kuchler-Bopp, F. Perrin-Schmitt, H. Peters, H. Lesot, Tissue Eng. 2006, 12, 1–7. A.H. Yen, P.T. Sharpe, Expert Opin. Biol. Ther. 2006, 6, 9–16.

281

Part III Pathological Calciﬁcations

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

285

18 Aspects of Pathological Calciﬁcations Inge Schmitz

Abstract

Unwanted, pathological calciﬁcations can be categorized based on the initiating stimulus: trauma, tumor, disturbances of mineral metabolism, inﬂammation, or idiopathic (unknown) causes, and can be found in almost all tissues. For example, pulmonary calciﬁcations are common asymptomatic ﬁndings, usually discovered on routine chest X-ray or at autopsy, and can occur in a wide variety of disorders, while vascular calciﬁcation (arteriosclerosis) is a common ﬁnding in elderly people, and causes heart disease and cerebral disorders. Several cellular components are involved in pathological calciﬁcations, and there is some evidence suggesting pathological mineralizations and ossiﬁcations are regulated processes. In terms of structure and elemental analysis, there are no signiﬁcant differences between pathological mineralization and bone formation. Calciﬁcations may be present in biological tissue, as well as in synthetic graft material, for example in vascular prostheses. Key words: heterotopic ossiﬁcation, pulmonary calciﬁcation, metastatic calciﬁcation, arteriosclerosis, AV-shunt, ossiﬁcation of synthetic grafts.

18.1 Introduction

Intact calciﬁcations – for example the formation of bone and skeleton and bone repair during fracture healing – are fundamental processes in vertebrates, where bone-forming osteoblasts and bone-resorbing osteoclasts and several other components are involved. Also involved are several cellular components, including connective tissue, matrix vesicles, mineral contents (calcium and phosphorus), hormones, vitamins and several proteins. In medicine, the formation of new bone is fundamental to bone repair and to orthopedic and dental surgery, for example in promoting the ﬁxation of implants. Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

286

18 Aspects of Pathological Calciﬁcations

Pathological calciﬁcations, which are characterized by the deposition of calcium phosphate-rich structures in regions where bone or calcium phosphate-rich structures are not normally present, are deemed to be unwanted reactions. Generally, pathological calciﬁcation and bone formation can be categorized into groups based on the initiating stimulus: trauma, tumors, disturbances of mineral metabolism, inﬂammation, or idiopathic (unknown) causes. Among others, ﬁbroblasts, endothelial cells and pericytes are involved in pathological bone formation and calciﬁcation. As early as 1924, Leriche and Policard discovered that, under certain conditions, ﬁbroblasts may induce new bone at almost every site of the body [1]. Later, in 1976, Friedenstein showed that ﬁbroblasts derived from bone marrow can form bone; these were referred to either as: (i) determined osteogenic precursor cells (DOPC); or (ii) ﬁbroblasts derived from other tissues, which can generate new bone under conditions of stimulation, or inducible osteogenic precursor cells (IOPC) [2]. Recently acquired evidence has suggested that vascular calciﬁcation is an active cell-mediated process with similarities to bone formation, and is not merely a simple precipitation of calcium and phosphate. On the ultrastructural level, there are no major diﬀerences between the steps of calciﬁcation seen in bone and arteries (Fig. 18.1) or other biological structures.

Fig. 18.1 Pathological bone-like calciﬁcation of an artery. Transmission electron microscopy image showing detail of the calciﬁcation area. The calciﬁcation is comparable to bone formation, with calcium phosphate crystallization.

18.1.1 Examples of Pathological Calciﬁcation

Calciﬁcations are frequently found in traumatic or necrotic tissues, for example after burns [3], in patients having hip prostheses [4], in injured blood vessels

18.1 Introduction

after coronary stenting [5], in lungs (caused by inﬂammations or infections such as tuberculosis), and in almost any other tissue. Cases have been reported where calciﬁcations have been identiﬁed even years or decades after injury [6]. Burke et al. studied 108 cases of sudden coronary death at autopsy and found calciﬁcations in the hearts of all male patients aged b50 years, and in all female patients aged b60 years [7], with micro fractures being found in 92% of the calciﬁed heart valves [8]. A paradox commonly seen in patients with severe aortic calciﬁcation is the simultaneous presence of demineralized bone in osteoporosis [9]. One rare pathological ﬁnding is calciﬁcation of the optic nerve caused by trauma [10]. Chemical burning of the eye with calcium-containing corrosives, as well as irrigation with eye drops containing phosphate buﬀer solutions following eye burns, also bear the risk of corneal calciﬁcation [11, 12]. The incidence of coronary artery calciﬁcation in patients with chronic kidney disease (CKD) and undergoing renal dialysis is up to ﬁvefold higher than in agematched individuals with angiographically proven coronary artery disease [13]. Here, two contributing factors inﬂuence calciﬁcation, namely hyperphosphatemia (plus hypercalcemia) and hyperparathyroidism. In a large variety of tumors, calciﬁcations are present either in the primary lesion or in metastases. Many tumor metastases, for example in the lung, are calciﬁed [14, 15]. Micro calciﬁcations play an important role in the detection of breast cancer, especially early-stage breast cancer. Tumoral calcinosis is a rare disease in its own right, and is characterized by the massive deposition of calcium phosphate in the soft tissue of large joints [16–19]. The arteriopathia calciﬁcans neonatum, another rare disease, is characterized by calciﬁcations of mostly medium-sized arteries and an accumulation of calciﬁed deposits along elastic ﬁbers, leading to destruction of the normal vessel structure. The etiology of this condition is unclear [20]. 18.1.2 Regulation of Calciﬁcations

Pathological and natural calciﬁcations are regulated by many factors; hence, osteopontin, a highly phosphorylated glycoprotein, is a potent inhibitor of vascular calciﬁcation in vitro and in vivo, and is also a regulator of biological calciﬁcation. Osteopontin expression has been identiﬁed in bone, in osteoblast-like cells in cultures, in tumors, in vascular smooth muscle cells (SMCs), in macrophages and myoepithelial cells, and is also expressed extracellularly around calciﬁed foci in a variety of tissues, for example in calciﬁed areas of breast cancer [21]. The osteoblasts themselves regulate the uptake and exchange of ions and small molecules via gap junctions (Fig. 18.2). Electric coupling of adjacent osteoblasts was demonstrated using microelectrodes, whereby changes in membrane potential amplitudes evoked by intracellular current injections in one cell were directly related to those in the coupled cell [22].

287

288

18 Aspects of Pathological Calciﬁcations

Fig. 18.2 Coupling between osteoblasts via gap junctions. Transmission electron microscopy image of two osteoblast-like cells gap junctions (arrows). The cells are derived from calvarian fragments of newborn rats.

18.2 Heterotopic Ossiﬁcation 18.2.1 Calciﬁcation in Ulcera of Patients with Paraplegia

Heterotopic ossiﬁcation (HO) is the abnormal development of bone tissue within soft tissue or muscle. The etiology and pathology of the condition are uncertain, but HO of the locomotor system may result from burns, tetanus, paraplegia or intensive sporting practice, such as horse riding. Symptoms include localized swelling, fever, pain, and a loss of joint motility. Indeed, surgical resection of HO is often needed to preserve joint motility. The heterotopic formation of bone in ulceras (Fig. 18.3) may be caused by increased pressure and tractive forces, and is reported to have an incidence of 20 to 30%. Histological and electron microscopic examinations of specimens taken

Fig. 18.3 Heterotopic ossiﬁcation in an ulcus of a patient with paraplegia. A histological section through the ulcus, showing the ossiﬁcation zones (arrows).

18.2 Heterotopic Ossiﬁcation

Fig. 18.4 Steps of heterotopic ossiﬁcation present in an ulcus of a patient with paraplegia. Transmission electron microscopy images. (a) Starting mineralization at collagen ﬁbers. (b) Mature calciﬁed crystals comparable to calcium phosphate crystals seen during bone formation or mineralization in the arteries (cf. Fig. 18.1).

from patients with calciﬁed ulceras showed a subdivision in an outer ulcus zone, a zone with granulation tissue, a blood vessel-rich proliferation zone, and a zone of ossiﬁcation. Diﬀerent degrees of mineralization and calciﬁcation up to the formation of mature bone may be found. Immunohistochemical and electron microscopic examinations have suggested that pericytes might be involved in the process of ossiﬁcation. As early as 1992, Brighton and colleagues showed that pericytes in culture can mineralize and secrete alkaline phosphatase and osteocalcin. Based on these ﬁndings, this group re-named pericytes as the ‘‘waiting-sleeping’’ stem cells of osteoneogenesis. Ultrastructurally, there are no diﬀerences between the stages of mineralization in bone or in osteoblast-like cell cultures, or in heterotopic ossiﬁcations (compare Figs. 18.1 and 18.4); immunohistochemically positive reactions to osteocalcin, bone morphogenetic protein (BMP), collagen type IV and desmin were identiﬁed. Moreover, a transformation of pericytes to pre-osteoblasts and, ultimately, to osteoblasts was demonstrated [23]. 18.2.2 Calciﬁcations of the Lung

Pulmonary calciﬁcation is a common asymptomatic ﬁnding which is usually discovered on routine chest X-radiography or at autopsy, and can occur in a wide variety of disorders. Calciﬁcations may be present in the lung parenchyma, the pleura, in lymph nodes, and the chest wall. Often, these calciﬁcations are manifestations of previous inﬂammatory processes (Fig. 18.5), but they may also be correlated to tumors, to medical therapy, to metabolic disorders, and to occupational exposure. Often, the possible degree of respiratory restriction does not correlate with the degree of mineralization or ossiﬁcation; for example, patients with

289

290

18 Aspects of Pathological Calciﬁcations

Fig. 18.5 Ossiﬁcation of human lung parenchyma due to the deposition of foreign material. Light microscopy image showing the lung parenchyma, with ossiﬁcation (*), foreign bodies (**), and a small blood vessel (***).

massive ossiﬁcations may be asymptomatic. Although calciﬁcations of the lung are usually harmless, they can oﬀer important information contributing to the correct diagnosis. Often, pulmonary mineralization/ossiﬁcation can be related to tuberculosis, and computed tomography (CT) ﬁndings in patients with inactive tuberculosis frequently include calciﬁed nodules or consolidations. Calciﬁcation is a common reaction to the inﬂammation caused by tuberculosis, and is also seen in animals, where calcosperite-like bodies have been identiﬁed in the tracheal mucus of a dog with tuberculosis [24]. Diﬀuse pulmonary calciﬁcation can sometimes be associated with the formation of mature bone in the pulmonary parenchyma. For diagnostic purposes, the lung tissue can be digested using sodium hypochlorite (10–12%), and any foreign materials or bony structures isolated and investigated separately (Fig. 18.6).

Fig. 18.6 Mature bone isolated from lung parenchyma (same case as Fig. 18.5). The bone was isolated after digestion of the lung parenchyma with sodium hypochlorite. Ossiﬁcations up to the formation of mature bone and initiated by inﬂammatory processes were caused by foreign material inhaled during an occupational exposure.

18.3 Vascular Calciﬁcations: Arteriosclerosis

Among 326 lung specimens taken randomly during autopsy, calciﬁcations of both lungs were found in 21 cases, while 14 patients showed calciﬁcations only in the right lung, and 12 only in the left lung. Pleural mineralizations were seen in 47 cases. These calciﬁcations were most likely caused by inﬂammatory processes. X-radiography and micrography of the lungs (taken at random in autopsy cases) showed a variety of conditions to exist, ranging from small, branch-like ossiﬁcations up to large ossiﬁed areas with the formation of mature bone. In most cases, the patient history leading to these changes were unclear, however. 18.2.2.1 Metastatic Pulmonary Calciﬁcations Ossiﬁcation of the lung tissue is rare in carcinoid tumors of the lung or pulmonary blastomas, and it is hardly ever found in lung metastases of extrathoracic epithelial tumors. According to ﬁndings based on the examination of metastatic pulmonary calciﬁcations, primary colorectal adenocarcinoma is the most probable primary lesion (Fig. 18.7). Among the investigated cases, lung metastases and primary tumor tissue were immunopositive for BMP2/4 and osteonectin antibodies, and energy-dispersive X-ray (EDS) microanalysis of the lesions demonstrated calcium and phosphorus contents typical of bone. Again, with regards to diagnosis, when ossiﬁcations are found in lung metastases there is a strong possibility of primary colorectal adenocarcinoma being present.

Fig. 18.7 Calciﬁed lung metastases of a primary colorectal adenocarcinoma. Light microscopy image; hematoxylin and eosin staining. Note the metastases with typical structure of the colon (*), but no typical lung structure. ** indicates ossiﬁcation.

18.3 Vascular Calciﬁcations: Arteriosclerosis 18.3.1 Calciﬁcations of Arteries

Arteries are among the most predisposed sites for the calciﬁcations that lead to a multitude of cardiac and cerebral complications. Vascular calciﬁcations can be correlated with aging, diabetes mellitus and renal dialysis, with the causes of

291

292

18 Aspects of Pathological Calciﬁcations

arterial calciﬁcations including changes in mineral metabolism, high-density lipoprotein (HDL) levels and in the expression of mineral-regulating proteins. The activity of vascular smooth muscle cells (VMCs), vesicle release, mediators of inﬂammation such as oxidation, carbonyl stress, C-reactive protein, TGF-b, and cytokines released by endothelial cells, along with several other components, are also involved. Vascular calciﬁcation is common in uremic hyperphosphatemia, where serum phosphate levels well above normal [25]; in fact, serum phosphate was found to be reliable predictor of morbidity and mortality. Vascular mesenchymal cells may respond to elevated serum phosphate concentrations by raising the levels of pro-mineralizing molecules. Hence, whilst SMCs were seen to undergo matrix mineralization in the presence of 2 mM inorganic phosphate, SMCs treated with 1.4 mM inorganic phosphate failed to mineralize [26]. The human aorta and the carotis are privileged sites for the development of vascular calciﬁcation, which may in turn be related to the hemodynamics of blood ﬂow. Evidence exists that vascular calciﬁcation is a highly regulated process which involves both inductive and inhibitory mechanisms, and in recent years several hypotheses attempting to explain vascular calciﬁcation have been proposed. Epple and Lanzer have discussed four mechanisms, each of which they considered might be responsible for the calciﬁcation of blood vessels [27]: A loss of inhibitory action on the crystallization of biological macromolecules. Nucleation of calcium phosphate precipitation by dead cells and/or their membranes. Autocatalytic nucleation of cholesterol by antibodies against cholesterol crystals. Formation of bone-like structures in advanced atherosclerotic lesions. Vascular calciﬁcations show many similarities to bone formation (see Fig. 18.1), and matrix-vesicle-like structures have been identiﬁed in calciﬁed blood vessels. Diphosphonates, citrates, aluminum and ferric ions are all potential inhibitors of vascular calciﬁcation; in synthetic systems these materials inhibit the crystal growth of apatite and other biologically relevant calcium phosphate phases [28]. 18.3.1.1 Calciﬁcation of the Tunica Media (Mo¨nckeberg’s Arteriosclerosis) Arteries consist of three distinct concentric layers; the tunica externa or adventitia; the tunica media; and the tunica intima (which bears the endothelial cells). Calciﬁcation can occur in two sites, namely the tunica intima and tunica media. Mo¨nckebergs arteriosclerosis is a degenerative and apparently noninﬂammatory disease in which the tunica media of small and medium-sized muscular arteries calciﬁes, with or without concurrent arteriosclerosis. The vessel lumina remains open. Two diﬀerent pathological processes may be involved in the formation of Mo¨nckeberg’s sclerosis: (i) a degenerative process leading to apoptosis or necrosis of medial SMCs; and (ii) an osteogenic process leading to calciﬁcation and the formation of bone-like structures [29].

18.3 Vascular Calciﬁcations: Arteriosclerosis

18.3.1.2 Calciﬁcation of the Tunica Intima (Arteriosclerosis) Calciﬁcations are prominent features of arteriosclerotic plaques, and are frequently found in the arteries of elderly patients, although even in adolescence small calcium-rich granules may be present. Arteriosclerotic calciﬁcations occur in the intimal layer of the arteries, and can result in severe stenosis of the vascular lumen, leading ﬁnally to clinical complications such as heart disease or stroke. Complicated lesions are characterized by the grade of calciﬁcation and the possible additional presence of ossiﬁcations (type Vb/VII, calciﬁed ﬁbroatheroma lesion; type VI, complicated lesion) [30–32]. According to Mu¨ller and colleagues [33], classiﬁcation of the calciﬁcation of arteries can be divided into diﬀerent histological degrees: grade 0 ¼ no/minimal amounts of mineral deposits; grade 1 ¼ small spots of calciﬁed deposits; grade 2 ¼ intermediate, non-conﬂuent regions of calcium deposits; and grade 3 ¼ large, solid calciﬁed deposits (Fig. 18.8). Among randomly taken specimens of arteries (carotis), only 50% of the samples could be classiﬁed as grade 0 (no/minimal deposits of calcium and phosphorus) based on the Mu¨ller classiﬁcation. EDS analysis has been found to be a very sensitive method for detecting small accumulations of calcium phosphate.

Fig. 18.8 Diﬀerent degrees of calciﬁcation present in human carotis. Scanning electron microscopy images; EDS-mapping for calcium, or calcium and phosphorus. (a) Detection of small, calcium-rich deposits in a so-called lipid-rich soft plaque (grade 1 M€ uller classiﬁcation). (b) Grade 2, with non-conﬂuent calcium-rich deposits. (c,d) Grade 3, large solid calciﬁed deposits leading to severe stenosis of the lumen.

293

294

18 Aspects of Pathological Calciﬁcations

Fig. 18.9 Ossiﬁcation of human arteries. (a) Scheme of an ossiﬁed artery; three distinct zones can be distinguished. 1 ¼ tunica intima; 2 ¼ tunica media; 3 ¼ tunica externa. Plaque formation (p) and ossiﬁcation (o) can be seen in the tunica media. (b) Light microscopy image, hematoxylin and eosin staining, showing ossiﬁcation (*).

18.3.2 Ossiﬁcations of Arteries

Arteries may show variable degrees of calciﬁcation to a point where, infrequently, there is the presence of bone (Fig. 18.9). In general, ossiﬁcations are present in 5% of investigated cases [34], but they occur predominantly at the external regions of arteriosclerotic plaques. Here, three diﬀerent zones may be distinguished: (i) a central zone, with mature lamellar bone; (ii) a zone, with osteoblast-like cells (immunopositive for markers for osteoblasts) and giant cells frequently found at the periphery of newly formed bone; and (iii) a zone consisting of an accumulation of foam cells, ﬁbroblasts, and connective tissue. 18.3.3 Characterization of Atherosclerotic Plaques of the Human Aorta

Element analytical studies of plaque material taken from patients with macroscopically diﬀerent degrees of calciﬁcation according to the Stary classiﬁcation [30–32] revealed that plaques contained about 60–70% biological carbonated apatite (in a dry state) in a nanocrystalline form, with particle sizes of approximately 20 nm. Ultrastructurally, there were many similarities to normal bone, with proof of typical calcospherites, and mineralization processes starting at collagen ﬁbrils (Fig. 18.10). Histological, ultrastructural and chemical investigations revealed no signiﬁcant diﬀerences between ossiﬁcations in plaques of the aorta compared to bone. Heterogeneously located deposits of calcium and phosphorus were detected using EDS-analysis, whilst chemically the calciﬁed deposits consisted of carbonated hydroxyapatite, usually with a calcium deﬁciency induced by the incorporation of hydrogen phosphate groups. The mean molar ratio of Ca:P was 1.4:1 according to EDX-analysis, which pointed to a calcium-deﬁcient hydroxyapatite (note that stoichiometric hydroxyapatite, Ca10 (PO4 )6 (OH)2 , has a Ca:P ratio of

18.3 Vascular Calciﬁcations: Arteriosclerosis

Fig. 18.10 Calciﬁcation of the human aorta. Transmission electron microscopy images of plaque material. (a) Collagen ﬁbrils with small electron-dense calciﬁcation sites. Original magniﬁcation, 20 000; scale bar ¼ 0:05 mm. (b) Radially grown aggregate of calcium phosphate, bound to a collagen ﬁber. Original magniﬁcation, 30 000; scale

bar ¼ 0:03 mm. (c) Concentrically grown aggregate of calcium phosphate crystals, showing diﬀerent ‘‘layers’’ of radially grown nanocrystals. Original magniﬁcation, 48 000; scale bar ¼ 0:02 mm. (d) Typical calcospherites in dense arrangement. Original magniﬁcation, 48 000; scale bar ¼ 0:02 mm.

Fig. 18.11 Chemical analysis of calciﬁed deposits isolated from human aorta. High-resolution X-ray powder diﬀractometry of atherosclerotic plaques in comparison to callus bone.

295

296

18 Aspects of Pathological Calciﬁcations

Fig. 18.12 Chemical analysis of calciﬁed deposits isolated from human aorta. Infrared spectroscopy of calciﬁed aorta samples in comparison to callus bone.

1.67:1). No diﬀerences were seen in the calciﬁcation process and the mineral composition (using infrared spectroscopy) between individual patients in diﬀerent stages of arteriosclerosis; neither were any diﬀerences identiﬁed compared to bone formation (Figs. 18.11 and 18.12) [35].

18.4 Calciﬁcation of Synthetic Vascular Grafts 18.4.1 Chronic Kidney Disease-Dialysis and Vascular Calciﬁcation of Arteries and Arteriovenous Shunts

Both, vascular calciﬁcation and inﬂammation are common in patients with chronic kidney disease (CKD) who, compared to healthy subjects, exhibit increased (two- to ﬁvefold) coronary arterial and peripheral arterial calciﬁcations. The risk factors for vascular calciﬁcation include the duration of renal dialysis, diabetes mellitus, aging, hyperphosphatemia, hyperparathyroidism, and calcium or vitamin D supplementation [36]. Cardiovascular disease is the leading cause of death in patients with CKD, and high levels of serum phosphate and parathyroid hormone play a critical role in the pathogenesis of cardiovascular problems and trigger pathological calciﬁcations. When examining the synthetic arteriovenous (AV) shunts of patients undergoing renal dialysis, mineralization up to various degrees of ossiﬁcation was identiﬁed in all cases of long-term (>4 years) dialysis patients, both in the arteries and the AV shunts.

18.4 Calciﬁcation of Synthetic Vascular Grafts

Fig. 18.13 Calciﬁcation of a vascular synthetic PTFE graft. (a) Scanning electron microscopy (SEM) image of synthetic PTFE graft, with typical architecture consisting of nodes (n) and ﬁbrils (f ). (b) SEM image of the explanted graft. (c) EDS-mapping for phosphorus (P) and calcium (Ca); calciﬁcation is closely related to the graft structure.

The calciﬁcation of synthetic material may be one cause of graft failure, and polytetraﬂuoroethylene (PTFE) (Fig. 18.13), Dacron, polyurethane (PU) and silicon are all biocompatible polymers commonly used as materials for vascular grafts. The chemical composition of grafts, together with the geometry and architecture of ﬁbrils, are all components that serve as targets and inﬂuence the process of calciﬁcation. ‘‘Dead’’ grafts are comparable with dead cells in blood vessels, and represent crystallization sites. Park et al. [37], when comparing the rate and extent of calciﬁcation of PTFE, PU and silicon grafts (in-vitro testing), found PTFE-based materials to calcify more quickly than either PU or silicon. Gorna and Gogolewski [38] investigated the in-vitro degradation and calciﬁcation of aliphatic PUs with various degrees of hydrophilicity, and found all to be calciﬁed in vitro, though susceptibility to mineralization increased with the material’s hydrophilicity. The Ca:P atomic ratio of the crystals growing was found to depend on the chemical composition of the tested materials, and ranged from 0.94 to 1.55 [38]. By using intravascular ultrasound and angiography, Castagna et al. [39] showed the calcium to be located primarily in the graft wall in 40% of cases, which indicated that the graft itself could

Fig. 18.14 A macroscopic view of an explanted AV-shunt. The biological tissue surrounding the explanted graft was digested using sodium hypochlorite. Note the numerous defects caused by puncturing during dialysis.

297

298

18 Aspects of Pathological Calciﬁcations

calcify. Defects in the grafts, which may for example have been caused by puncturing during dialysis, led to disturbances of the inﬂux and exchange of molecules, and favored an imbalance of phosphorus and calcium that promoted calciﬁcation (Fig. 18.14). 18.4.2 Ossiﬁcation of Synthetic Grafts

In cases where ossiﬁcation occurred adjacent to the graft or directly on the graft ﬁbrils (Fig. 18.15), EDS-analysis and mapping for phosphorus and calcium clearly identiﬁed those calciﬁed/ossiﬁed areas that could be distinguished from the graft. Immunohistochemical studies, using antibodies against BMP 2/4 (markers against osteoblasts) showed a positive reaction of small cells associated with the graft periphery and to cells on the graft surface (Fig. 18.16).

Fig. 18.15 Ossiﬁcation of a synthetic vascular graft (an explanted AV-shunt). The ossiﬁcation (*) is located mainly at the periphery of the synthetic material (**).

Fig. 18.16 Ossiﬁcation of synthetic grafts. Light microscopy image of immunohistochemical testing. * indicates the PTFE graft; ** indicates the immunopositive osteoblast-like cells (BMP2/4-positive) seen at the periphery of the graft; a few such cells were located on the graft.

References

18.5 Conclusions

As pathological calciﬁcations/ossiﬁcations may occur in almost every biological tissue, this chapter has focused on only certain aspects of clinical/pathological calciﬁcation. Pathological calciﬁcations have many parallels with the natural formation of bone, and appear to be regulated, although the exact manner in which this is achieved remains unclear. Future investigations, notably of an interdisciplinary nature, are required to resolve the situation.

References 1 R. Leriche, A. Policard, Surg. Gynecol. 2 3

4 5

6 7

8

9 10

11 12

13 14

Obstet. 1926, 43, 308. A.J. Friedenstein, Int. Rev. Cytol. 1976, 47, 327–360. S.L. Peterson, M.M. Mani, C.M. Crawford, J.R. Neﬀ, J.M. Hiebert, J. Trauma 1989, 29, 365–369. M.A. Ritter, R.B. Vaughan, J. Bone Joint Surg. 1977, 59A, 345–351. M. Mosseri, L.F. Satler, A.D. Pichard, R. Waksman, Cardiovasc. Revasc. Med. 2005, 6, 147–153. K.B. Poon, S.H. Chien, K.B. Tsai, G.T. Lin, Acta Orthop. 2005, 76, 728–729. A.P. Burke, A. Taylor, A. Farb, G.T. Malcom, R. Virmani, Z. Kardiol. 2000, 89 (Suppl. 2), 49–53. E.R. Mohler, F. Gannon, C. Reynolds, R. Zimmermann, M.G. Keane, F.S. Kaplan, Circulation 2001, 103, 1522– 1528. K.I. Bostro¨m, Z. Kardiol. 2000, 89 (Suppl. 2), II69–II74. J.L. Crompton, J. O’Day, A. Hassan, J. Neuroophthalmol. 2004, 24, 293– 294. M. Daly, S.J. Tuft, P.M. Munro, Cornea 2005, 24, 761–765. N.F. Scharge, S. Compa, B. Ballmann, M. Reim, S. Langefeld, Graefes Arch. Clin. Exp. Ophthalmol. 2005, 243, 780–784. S.M. Moe, N.X. Chen, Circ. Res. 2004, 95, 560–567. C.W. Maile, B.A. Rodan, J.D. Godwin, J.T. Chen, C.E. Ravin, Br. J. Radiol. 1982, 650, 108–113.

¨ ller, 15 J. Birzele, I. Schmitz, K.M. Mu Pathologie 2003, 24, 66–69. 16 S. Bittmann, H. Ulus, Schweiz.

17

18 19

20

21

22

23 24

25 26

27 28

Rundsch. Med. Prax. 2005, 94, 1309– 1311. L. Pilloni, D. Fanni, O. Nardello, M. Cagetti, G. Faa, Pathologica 2004, 96, 470–474. O. Nardello, M. Muggiamu, M. Cagetti, Chir. Ital. 2005, 57, 103–107. S. Mahadevan, B. Adhisivam, C.N. Kumar, Indian J. Pediatr. 2005, 72, 889–890. K.M. Mu¨ller, M. Blaschke, R.v. Basewitz, Beitr. Path. 1976, 157, 84–92. A.M. Tokes, J. Krausz, J. Kulka, M. Jackel, A. Kadar, Orv. Hetil. 2002, 143, 1841–1846. K. Schirrmacher, I. Schmitz, E. Winterhager, O. Traub, F. Bru¨mmer, D. Jones, D. Bingmann, Calcif. Tissue Int. 1992, 51, 285–290. A. Bosse, Veroﬀ. Pathol. 1997, 146, 1–168. N.B. Bauer, E. O’Neill, B.J. Sheahan, J. Cassidy, H. McAllister, Vet. Clin. Pathol. 2004, 33, 168–172. C.M. Giachelli, Orthod. Craniofac. Res. 2005, 8, 229–231. S. Jono, A. Shioi, Y. Ikari, Y. Nishizawa, J. Bone Miner. Metab. 2006, 24, 176–181. M. Epple, P. Lanzer, Z. Kardiol. 2001, 90 (Suppl. 3), 2–5. R.Z. LeGeros, Z. Kardiol. 2001, 90 (Suppl. 3), 116–124.

299

300

18 Aspects of Pathological Calciﬁcations 29 A. Shioi, H. Taniwaki, S. Jono, Y.

30

31

32 33

Okuno, H. Koyama, K. Mori, Y. Nishozawa, Am. J. Kidney Dis. 2001, 38 (4 Suppl. 1), S47–S49. H.C. Stary, A.B. Chandler, S. Glagov, J.R. Guyton, W. Insull, M.E. Rosenfeld, S.A. Schaﬀer, C.J. Schwartz, W.D. Wagner, R.W. Wissler, Circulation 1994, 89, 2462– 2478. H.C. Stary, A.B. Chandler, R.E. Dinsmore, V. Fuster, S. Glagov, W. Insull, M.E. Rosenfeld, C.J. Schwartz, W.D. Wagner, R.W. Wissler, Circulation 1995, 92, 1355–1374. H.C. Stary, Z. Kardiol. 2000, 89 (Suppl. 2), 28–35. Th. Deneke, P.H. Grewe, S. Ruppert, K. Balzer, K.M. Mu¨ller, Z. Kardiol. 2000, 89 (Suppl. 2), II36–II48.

34 Th. Deneke, K. Langner, P.H. Grewe,

35

36

37 38 39

E. Harrer, K.M. Mu¨ller, Z. Kardiol. 2001, 90 (Suppl. 3), 106–115. A. Becker, M. Epple, K.M. Mu¨ller, I. Schmitz, J. Inorg. Biochem. 2004, 98, 2032–2038. H. Al Humoud, N. Al-Hilali, A.A. Ahmad, V.T. Ninan, M.R. Nampoory, A.M. Rizk, J.H. Ali, K.V. Johny, Transplant. Proc. 2005, 37, 4183–4186. J.C. Park, M.J. Song, Y.S. Hwang, H. Suh, Yonsei Med. J. 2001, 42, 304–310. K. Gorna, S. Gogolewski, J. Biomed. Mater Res. A. 2003, 67, 813–827. M.T. Castagna, G.S. Mintz, P. Ohlmann, J. Kotani, A. Maechara, N. Gevorkian, E. Cheneau, E. Stabile, A.E. Ajani, W.O. Suddath, K.M. Kent, L.F. Satler, A.D. Pichard, N.J. Weissman, Circulation 2005, 11, 1148–1152.

301

19 Atherosclerosis: Cellular Aspects Diane Proudfoot and Catherine M. Shanahan

Abstract

Calciﬁcation is an extremely common pathology in atherosclerosis, and correlates with the severity of coronary artery disease. Calcium crystal development occurs at very early stages of the disease, and involves both physico-chemical interactions as well as complex cellular interactions within the arterial intima. Vascular smooth muscle cells (VSMCs) and cells associated with inﬂammation, such as the macrophage, are the two main cell types associated with atherosclerosis. Both have a role in the development of calciﬁcation via release of various factors including matrix vesicles, bone-related proteins and extracellular matrix components which create an environment that will support calcium crystal growth. Cell death within the atherosclerotic plaque is also an important mechanism in generating calciﬁed deposits. In this chapter, we discuss the roles of the diﬀerent types of cell within atherosclerotic plaques, and how they might contribute to or limit calciﬁcation. Key words: atherosclerosis, cells, calciﬁcation, calcium, vascular smooth muscle, macrophages, apoptosis, matrix vesicles.

19.1 Introduction

The normal human artery wall consists of an endothelial cell layer at the luminal side, under which lies a small intima of vascular smooth muscle cells (VSMCs), followed by the internal elastic lamina (Fig. 19.1). Between the internal elastic lamina and the external lamina lie several layers of VSMCs and their associated matrix. These VSMCs have a ‘‘contractile’’ phenotype and are essential in maintaining vascular tone. To the outside of the external elastic lamina is the adventitia, made up of matrix, ﬁbroblasts, microvessels, and nerves. Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

302

19 Atherosclerosis: Cellular Aspects

Fig. 19.1 Cellular components of the normal vessel wall (A) and the atherosclerotic intima (B). IEL ¼ internal elastic lamina; OEL ¼ external elastic lamina.

Atherosclerosis is a disease of the intima, initiated by damage to the endothelium and resulting in an inﬁltrate of monocytes and lymphocytes from the circulation into the vessel wall. Upon entering the intima and exposure to modiﬁed lipids, monocytes diﬀerentiate into macrophages and can accumulate large amounts of lipid, becoming ‘‘foam cells’’. Macrophages/foam cells release proinﬂammatory mediators, and these factors promote the migration of contractile VSMCs from the media into the intima. VSMCs in atherosclerotic plaques have a diﬀerent phenotype from those in the medial layer of the vessel wall. They are known as a ‘‘synthetic’’ or ‘‘repair’’ phenotype, which indicates that they become non-contractile and synthesize matrix components, including collagen. Although increased numbers of VSMCs accumulating in the intima may cause atherosclerotic plaques to become larger in size, VSMCs are thought to be beneﬁcial as the ﬁbrous cap matrix they produce creates a stronger plaque that is less likely to rupture – a main cause of myocardial infarction or stroke. Macrophages, on the other hand, are capable of killing neighboring VSMCs, and are therefore thought to be harmful in terms of progression of the disease as they reduce the number of VSMCs capable of mediating repair. Vascular calciﬁcation occurs at two distinct areas of the artery wall: the intima and the media. Intimal calciﬁcation occurs within atherosclerotic plaques, has a diﬀuse, punctate morphology, and appears as aggregates of calcium crystals. Medial calciﬁcation occurs along the elastic lamellae, and is commonly found in aging, diabetes, and uremia. These deposits can also coalesce to produce larger, solid crystals. Occasionally, calciﬁcation is found that has the characteristics of true bone. The triggers for the induction of calcium crystal formation, regulation of crystal growth and whether cells within the vessel wall orchestrate or limit calciﬁcation, are not fully understood but much progress has been made over the past decade. In this chapter we discuss the contribution made by diﬀerent cell types to the development of calciﬁcation, and examine the complex interactions that occur between these cell types in the context of atherosclerosis.

19.2 Role of VSMCs in Vascular Calciﬁcation

19.2 Role of VSMCs in Vascular Calciﬁcation 19.2.1 Release of Apoptotic Bodies and Vesicles

VSMCs contribute to the development of calciﬁcation in atherosclerosis in several diﬀerent ways (Table 19.1). One of these is the generation of vesicles that can act as a nidus for the initiation of calcium crystal growth. Indeed, the vesicles generated by VSMCs have some similarities with mineralizing matrix vesicles (MVs) found in cartilage and bone. MVs are extracellular membrane-derived particles released by budding from the surface of chondrocytes and osteoblasts that range from 30 nm to 1 mm in diameter [1]. The conditions that induce MV budding and release are not known, but they may be related to the cell cycle, and possibly also to apoptosis. Mineralization initiates inside the vesicle from a pre-existing nucleation core complex and then progresses around the vesicle membrane, which eventually ruptures. Calciﬁcation then proceeds along the associated matrix. MV production in cartilage occurs in normal non-mineralized articular cartilage as well as the mineralizing growth plate cartilage. This indicates that not all released vesicles undergo calciﬁcation. The release of calciﬁcation-competent vesicles is dependent on the diﬀerentiation state of the chondrocyte. Calcifying

Table 19.1 Summary of the role of diﬀerent cells types in atherosclerotic calciﬁcation.

Cell type

Role in calciﬁcation

Reference(s)

VSMC

Positive and negative Production of inhibitors of calciﬁcation Production of inducers of calciﬁcation Phagocytosis of MV

82 36, 37 16, 31

Positive and negative Production of inhibitors of calciﬁcation Production of inducers of calciﬁcation

83 61, 84

CVCs and pericytes

Macrophages

Osteoclasts

Positive and negative Phagocytosis of MV Phagocytosis of crystals Pro-inﬂammatory releasing TNFa (inducer of CVC calciﬁcation) Negative Removal of crystals

CVC ¼ calcifying vascular cell; MV ¼ matrix vesicle; VSMC ¼ vascular smooth muscle cell.

33, 34 67 66, 67

77

303

304

19 Atherosclerosis: Cellular Aspects

MVs are selectively enriched in speciﬁc proteins such as alkaline phosphatase and ATP-hydrolyzing enzymes for the generation of local phosphate ions. They also contain annexins II, V, and VI [2]. Annexins V and VI normally reside within the cytoplasm, and are selectively relocated to the plasma membrane and released in MV, where they are thought to act as calcium channels in the vesicle membrane. This inﬂux of Ca 2þ is essential for mineral growth. In addition to acting as calcium channels, annexin V binds to types II and X collagen that anchor vesicles to the extracellular matrix (ECM). Certain factors can stimulate cartilage vesicle calciﬁcation, such as retinoic acid, calcium and phosphate ions, b-glycerophosphate, 1,25-OH vitamin D3 , and dexamethasone [1, 3]. Vesicular structures with similar morphology to cartilage MVs have been identiﬁed in human arteries [4, 5]. VSMC-derived vesicles have particularly been observed in association with atherosclerosis and hypertension [4–9]. When isolated from atherosclerotic plaques, vesicles have been shown to calcify in vitro if incubated in synthetic cartilage lymph, a buﬀer that induces MV calciﬁcation [10]. Many vesicles may originate from apoptotic VSMCs, as some have been shown to express the pro-apoptotic protein BAX [4]; however, vesicles are also produced from viable VSMCs [5]. Vesicles isolated from apparently normal areas of human arteries have the potential to undergo calciﬁcation in vitro, but to a lesser extent than those from plaques. This may be due to several factors, including cholesterol, which has been shown to increase plaque vesicle calciﬁcation [11]. MVs appear to be distinct from apoptotic bodies, as they are released from viable cells and do not contain organelle remnants. In cartilage growth plate chondrocytes, the mineralization of MV requires alkaline phosphatase and annexins [12]. Apoptotic bodies derived from the same cells also undergo mineralization, but via mechanisms that do not require alkaline phosphatase or annexins. The exact mechanisms involved in apoptotic body calciﬁcation are not known, but a lipid component exposed on apoptotic cells, phosphatidyl serine (PS), can provide a calcium-binding site [13] as well as a membrane surface suitable for calcium crystal deposition [14]. Apoptotic bodies are therefore thought to have a ‘‘default’’ mechanism which allows the precipitation of calcium and phosphate ions on their outer membrane surface. Our studies in cultured human VSMCs have shown that apoptotic bodies and MV derived from human VSMCs can undergo calciﬁcation when incubated in synthetic cartilage lymph buﬀer [15, 16]. In addition, the stimulation of apoptosis led to increased calciﬁcation in VSMC multicellular nodule cultures, while inhibition of apoptosis reduced the level of calciﬁcation. We have therefore speculated that apoptosis and/or vesicle generation is a crucial initiating event in vascular calciﬁcation. Apoptosis of VSMCs may be triggered in several diﬀerent ways in the atherosclerotic plaque: (i) VSMC interaction with inﬂammatory cells that express cell-surface death ligands and activation of death receptors such as Fas [17]; (ii) secretion of pro-apoptotic cytokines such as tumor necrosis factor (TNF) by macrophages and interaction with TNF receptor 2 on the surface of VSMCs; and (iii) oxidized low-density lipoprotein (LDL), mechanical stress and reactive oxygen species (ROS) cause VSMC apoptosis via release of cytochrome c from the mito-

19.2 Role of VSMCs in Vascular Calciﬁcation

chondria [18]. Therefore, several factors known to be expressed in plaques – as well as interactions with inﬂammatory cells – can lead to VSMC apoptosis and would be expected to encourage apoptotic body (nidus) production. On the other hand, anti-apoptotic factors are also found in plaques such as insulin-like growth factor (IGF-1), which is a potent cell-survival factor [19]. Both IGF and its receptor have been detected in atherosclerotic lesions, although VSMCs in atherosclerotic plaques have a reduced capacity to respond to IGF, compared with medial VSMCs [19, 20]. IGF-1 has recently been shown to inhibit calciﬁcation in vascular cells in vitro, via ERK and PI3 kinase pathways, but it may also block mineralization through eﬀects on inhibiting apoptosis [21]. Plateletderived growth factor (PDGF) has been reported to inhibit apoptosis, mediated by survivin, a member of the ‘‘inhibition of apoptosis’’ family [22]. In addition, the ECM has a major role in protecting cells from apoptosis [23]. Thus, the determination of whether a VSMC undergoes apoptosis in an atherosclerotic plaque may depend on its location, phenotype, and exposure to pro- or anti-apoptotic factors. Much less is known about the factors that determine vesicle generation and release from viable cells. However, stress, hypertension, and increased extracellular calcium and phosphate ion concentrations have been shown to increase vesicle release [16, 24, 25]. VSMCs exposed to elevated calcium levels, equivalent to those found in uremic patients or at sites of inﬂammation in atherosclerotic plaques, release vesicles containing preformed calcium crystals within the vesicle sap [16]. This shows that environmental conditions are important in determining whether calciﬁcation is initiated in vesicles. Aside from VSMCs, macrophages, dendritic cells, T lymphocytes, mast cells and endothelial cells have also been reported to cause vesicle release [26–29], though whether such release from these or other cell types in atherosclerotic plaques is a mechanism for initiating calciﬁcation is not yet known. 19.2.2 Phagocytosis

As described above, apoptotic bodies and MV can act as a nidus for the initiation of calcium crystal formation. Under physiological conditions, however, apoptotic bodies are normally rapidly phagocytosed to avoid secondary necrosis and stimulation of inﬂammation. Although macrophages are thought of as the ‘‘professional’’ phagocytic cell in atherosclerotic plaques, VSMCs have also been shown to phagocytose apoptotic bodies [30, 31]. Therefore, eﬃcient phagocytosis removes a potential nidus for calciﬁcation. It is not yet known whether the smaller vesicles released by viable VSMCs are also phagocytosed. In the atherosclerotic plaque, certain factors such as modiﬁed lipids interfere with phagocytosis and can compete with phagocyte binding to apoptotic bodies [32]. In cultured VSMCs, apoptotic bodies competed with acetylated LDL for the scavenger receptor class A1 (SRA1) on the surface of VSMCs [31]. Thus, failure to clear apoptotic bodies would allow calcium crystal growth to proceed, and/or may be an important in-

305

306

19 Atherosclerosis: Cellular Aspects

Fig. 19.2 The balance between various factors inﬂuencing atherosclerotic calciﬁcation. BMP ¼ bone morphogenetic protein; BSP ¼ bone sialoprotein; MGP ¼ matrix Gla protein; MV ¼ matrix vesicle.

ﬂammatory stimulus. However, opsonins (factors that aid phagocytosis) such as osteopontin and fetuin are also found in atherosclerotic plaques, and these would be expected to encourage phagocytosis [16, 33, 34], thus highlighting the importance of a balance between factors with opposing actions in the complex atherosclerotic environment (Fig. 19.2). 19.2.3 VSMC Osteo/Chondrocytic Conversion

Another way in which VSMCs contribute to calciﬁcation is via the induction of expression of various calciﬁcation-regulatory proteins normally associated with bone and cartilage. Indeed, experiments in cultured cells have shown that VSMCs isolated from normal areas of human vessels spontaneously take on an osteo/ chondrocytic phenotype [35]. The triggers for this conversion in culture are not known, but are likely to involve dissociation from their three-dimensional (3-D) matrix and cell–cell contacts, as well as exposure to culture medium components such as serum and, potentially, bone morphogenetic proteins (BMPs). Calciﬁcation-regulatory proteins such as bone sialoprotein (BSP), osteopontin, osteocalcin, matrix Gla protein (MGP), osteonectin, collagen I and II, alkaline phosphatase and BMPs have been detected in atherosclerotic plaques [36, 37]. Many of these proteins are expressed by VSMCs that have undergone osteo/ chondrocytic conversion mediated by expression of obligate bone and cartilage transcription factors such as Cbfa1, Msx2, and Sox9 [37]. Some of these proteins are inducers of calciﬁcation (e.g., BSP [38] and BMP-2 [39]), but most are known to be inhibitors of calciﬁcation (e.g., osteopontin, osteocalcin, and MGP) [40, 41]. Interestingly, MGP is constitutively expressed by contractile VSMCs in the normal media, and the expression of its mRNA is down-regulated in calciﬁed vessels [35]. MGP contains ﬁve residues of an uncommon amino acid, g-carboxyglutamic acid, which is formed by a vitamin K-dependent modiﬁcation of speciﬁc glutamic acid residues. The Gla residues confer calcium-binding properties to MGP and this is thought to inhibit calciﬁcation via calcium ion and calcium crystal binding.

19.2 Role of VSMCs in Vascular Calciﬁcation

MGP null mice display extensive vascular calciﬁcation, although their arteries contain chondrocytes rather than VSMCs. MGP has since been shown to play a role in cell diﬀerentiation via binding and antagonizing BMPs [42–44]. Thus, loss of MGP expression – allowing BMPs to act unopposed – may be a key event leading to VSMC osteo/chondrocytic conversion. Indeed, BMP-2 – the moststudied BMP in the vasculature – has been shown to decrease expression of VSMC contractile markers [45]. BMP-2 is synthesized by VSMCs, periadventitial myoﬁbroblasts and endothelial cells, and it is thought to be induced in atherosclerotic plaques by oxidative stress, inﬂammation, and hyperglycemia [39, 46]. BMP-2 has also been shown to induce apoptosis in pulmonary VSMCs [47], another potential mechanism for the calciﬁcation-inducing action of BMP-2. Interestingly, not all BMPs have stimulatory eﬀects on vascular calciﬁcation; BMP-7 has anti-calciﬁcation properties, partly via reducing phosphate levels and partly via eﬀects on VSMCs [48]. Other potential triggers for osteo/chondrocytic conversion of VSMCs in plaques include lipids, oxidative stress, and calcium and phosphate ions. Early deposits of calciﬁcation have been demonstrated in and around isolated VSMCs within the lipid core [49], and we have also detected calciﬁed, osteo/chondrocytic, lipid-ﬁlled VSMCs in human carotid atherosclerotic plaques [31]. Interestingly, in cultures of human VSMCs, spontaneous lipid accumulation occurred in multicellular nodules prior to calciﬁcation, and the addition of modiﬁed lipoproteins stimulated calciﬁcation as well as changing the time-course of bone-associated protein gene expression [31]. Oxidized lipids have also been shown to induce osteoblastic diﬀerentiation and calciﬁcation of bovine calcifying vascular cells [50]. Giachelli’s group has shown that phosphate ions, when in excess of the physiological extracellular concentration, can induce osteogenic diﬀerentiation of human VSMCs via membrane phosphate transporter proteins, and this results in the deposition of diﬀuse calciﬁcation [17]. Phosphate treatment of VSMCs increased the expression of Cbfa1 and osteocalcin, while simultaneously downregulating smooth muscle contractile marker proteins [51]. Extracellular phosphate is taken up by cells via a sodium-dependent phosphate transporter, Pit-1, and subsequent increases in intracellular phosphate ion concentration induce the expression of mRNA for mineralization-regulating genes [17]. By reducing Pit-1 levels using RNA interference, it was recently reported that the Pit-1mediated VSMC calciﬁcation was not associated with apoptosis or cell-derived vesicles; however, Pit-1 was necessary for the phenotypic modulation in response to elevated phosphate [52]. Extracellular calcium ions have also been shown to alter VSMC gene expression and to induce calciﬁcation [16, 53, 54]. Levels of extracellular ionic calcium can be elevated, particularly at sites of cell death or inﬂammation in atherosclerotic plaques [55, 56]. Although supraphysiological extracellular calcium ion levels induced Pit-1 expression in VSMCs [54], however, VSMCs also responded to increases in extracellular calcium by increasing expression of MGP mRNA via a calcium-sensing signaling mechanism [53]. As MGP is a potent inhibitor of calciﬁcation, the response to increased MGP mRNA by extracellular calcium ions may

307

308

19 Atherosclerosis: Cellular Aspects

be a feedback mechanism to prevent calciﬁcation. In contrast, experiments in cultured human VSMCs showed that calcium ions could induce calciﬁcation of VSMCs, due to increased release and calciﬁcation of VSMC-derived MV [16], indicating that the inhibitory actions of MGP can be overcome. Thus, a number of factors – including the loss of MGP together with exposure to BMPs, lipids and calcium and phosphate ions – may be responsible for the osteogenic phenotype of VSMCs found in atherosclerotic lesions (see Table 19.2). It should also be noted that some of the components of atherosclerotic plaques that stimulate osteogenesis and calciﬁcation, such as ROS, have the opposite eﬀect on

Table 19.2 Summary of the components of atherosclerotic plaques that may inﬂuence osteogenic diﬀerentiation and calciﬁcation.

Component of atherosclerotic plaque

Eﬀect on osteogenic diﬀerentiation

Eﬀect on calciﬁcation

Reference(s)

Lipids Cholesterol HDL Oxidized LDL Acetylated LDL

? þ þ þ

þ þ þ þ

11, 85 86 86 31

Bone proteins MGP BMP-2 Osteopontin Osteocalcin Bone sialoprotein Alkaline phosphatase

þ ? ? ? ?

þ þ þ

41, 83 83 40 87 38 1

Elevated extracellular phosphate ions

þ

þ

16, 17

Elevated extracellular calcium ions

þ

þ

16, 54

Oxidative stress

þ

þ

57

Cytokines TNFa TGFb IL-1, IL-6

þ þ þ

þ þ þ

68 85, 88 86

Cells Endothelial cells Macrophages

þ þ?

þ þ

89 70

19.2 Role of VSMCs in Vascular Calciﬁcation

bone cell calciﬁcation [57]. These observations may help to explain the parallel build-up of calciﬁcation in arteries and the loss of calciﬁcation in osteoporosis.

19.2.4 Role of Calcifying Vascular Cells and Pericytes

Calcifying vascular cells (CVCs) is the term given to a subpopulation of VSMCs isolated from the media of the bovine vessel wall by dilutional cloning. CVCs spontaneously form nodules, express bone proteins and calcify in culture, and have been extremely useful in determining the eﬀects of various atherosclerotic plaque components on the calciﬁcation process (Table 19.2). The presence of these cells in the vessel wall suggests the possibility that a subset of cells in the normal aorta may be responsible for driving calciﬁcation in atherosclerosis and in medial calciﬁcation. These cells have a distinct pattern of bone-like gene expression, as they calcify in culture and have also been shown to retain the ability to diﬀerentiate into other mesenchymal lineages besides osteoblasts [58]. However, there are no markers available to distinguish them from VSMCs, which also show mesenchymal diﬀerentiation potential [59]. The existence of these cells raises the possibility that clonal ‘‘stem’’ cells may also be present in atherosclerotic plaques. Potential sources of these stem cells, which may be responsible for initiating true bone formation in the vessel wall, are microvascular pericytes and mesenchymal stem cells. Microvascular pericytes were the ﬁrst cell types shown to link vascular cells with osteogenesis [60], and have some similarities with VSMCs of larger vessels and with CVCs. The main diﬀerences between CVCs and pericytes are that the latter are present in microvessels; in vivo they are closely associated with endothelial cells, and they have been reported to express a unique marker protein, 3G5. This pericyte marker has been detected in cells within human atherosclerotic plaques [36]. The origin of the pericyte-like cells in plaques is not clear, although many advanced plaques contain neovascular microvessels that will be surrounded by pericytes. The adventitia also contains microvessels, and pericytes may migrate from the adventitia to the intima. Importantly, cultured vascular pericytes were shown to have osteoblastic and chondrocytic potential when implanted subcutaneously in diﬀusion chambers [61]. They also form nodules spontaneously in culture and have a distinct pattern of bone gene expression as they calcify. The VSMCs, CVCs and pericytes described above are all capable of diﬀerentiating to an osteogenic phenotype, where they express several proteins normally associated with bone and cartilage formation. Another potential source of osteo/ chondrocytic cells in atherosclerotic plaques is that of mesenchymal stem cells, which could enter the plaque from the blood vessel lumen or from neovascular microvessels within the plaques. However, whether VSMCs, CVCs, pericytes or other mesenchymal cells orchestrate or serve to inhibit calciﬁcation in atherosclerosis is not yet known.

309

310

19 Atherosclerosis: Cellular Aspects

19.3 Role of Inﬂammatory Cells 19.3.1 Macrophages

As described above, intimal calciﬁcation occurs in the context of apoptotic cells in a lipid-rich environment. Hence, as atherosclerosis is predominantly an inﬂammatory process, it is not surprising that the cells most commonly associated with intimal calciﬁcation are inﬂammatory cells such as macrophages. The exact role of inﬂammatory cells in initiating or maintaining atherosclerotic calciﬁcation is not known, but macrophages could act in several diﬀerent ways to inﬂuence the calciﬁcation process. First, it has been well established that inﬂammatory cells are associated with the deposition of calcium crystals in degenerative arthritis. Macrophages have been reported to respond to crystals by phagocytosing and solubilizing them [62, 63]. The engulfment of crystals in this way may be of some beneﬁt, but this also leads to an increased cell survival time in macrophages and prolonged inﬂammation. The recruitment of macrophages to joints has been related to high concentrations of calcium ions or crystals at these sites. Macrophages have been shown to migrate towards a high concentration of extracellular calcium, and this property was absent in macrophages isolated from mice that lacked the calcium-sensing receptor [64]. Given that calcium levels may be elevated at sites of cell death and inﬂammation, macrophages are likely to be recruited to these sites as part of the innate immune response. Because of the damaging eﬀects of crystals observed in arthritis, it was hypothesized that calcium crystals might also have pro-inﬂammatory eﬀects on macrophages in human atherosclerotic plaques [65]. It was recently demonstrated that calcium crystals do indeed stimulate production of inﬂammatory cytokines in human macrophages [66, 67]. In the study conducted by Nadra and colleagues, it was shown that phagocytosis of the crystals was essential in causing a stimulation of TNFa release from macrophages. A crystal diameter of 1 mm or less was noted as being the optimal size for phagocytosis, with larger-sized crystals having no eﬀect on TNFa release. TNFa has been shown in other studies to stimulate calciﬁcation of vascular cells in vitro, either via the cAMP signaling pathway or potentially via the induction of apoptosis [68, 69]. It was therefore concluded that crystals small enough to be phagocytosed by macrophages in atherosclerotic plaques could have the eﬀect of increasing calcium deposition via TNFa release, leading to a vicious cycle of calcium crystal uptake and deposition [67]. In the same study, it was also suggested that small crystals – particularly in the early stages of atherosclerosis – could have damaging eﬀects, whereas larger aggregates of calcium crystals may be inert. Another way in which macrophages could potentially inﬂuence intimal calciﬁcation is via the engulfment of apoptotic bodies. Macrophages are ‘‘professional’’ phagocytes that are well equipped rapidly to engulf apoptotic cells, with the ap-

19.3 Role of Inﬂammatory Cells

parent aim of avoiding an inﬂammatory response such as that stimulated by necrotic cells. However, apoptotic cells and bodies are detected in atherosclerotic plaques, indicating a failure of phagocytosis. As described earlier, the acellular lipid core of the atherosclerotic plaque contains oxidized lipids that can compete with phagocyte binding to apoptotic bodies [32]. Thus, failure to clear apoptotic bodies would allow calcium crystal growth to proceed, and/or may be an important inﬂammatory stimulus. Lastly, macrophages may inﬂuence atherosclerotic calciﬁcation via the release of numerous secretory products. Factors within atherosclerotic plaques such as oxidized LDL and lipopolysaccharide (LPS) can activate macrophages, leading to the release of inﬂammatory cytokines, including interleukins, TNFa, and macrophage chemotactic factor (MCP-1), as well as anti-inﬂammatory cytokines such as transforming growth factor b (TGFb). They can also release osteopontin, ROS, prostaglandins, matrix-degrading enzymes and several other factors that may aﬀect calciﬁcation. One study addressed whether macrophages could inﬂuence calciﬁcation by using human monocyte-derived macrophages and bovine CVCs in co-culture [70]. Macrophages were found to stimulate calciﬁcation when the cells were in direct contact with each other, but when the two cell types were separated by a membrane only LPS-treated macrophages stimulated calciﬁcation – an eﬀect which was shown to occur via the release of TNFa. Therefore, under certain circumstances, macrophages can stimulate the calciﬁcation of vascular cells. Of particular interest is the secretion of osteopontin by macrophages. Osteopontin is a multifunctional protein that can act as a potent inhibitor of calciﬁcation by binding to the mineral and preventing crystal growth. In addition, it can act as an opsonin, encouraging phagocytosis, and it also has roles in cell adhesion to the mineral surface and in inﬂammation. It was also recently shown that osteopontin could stimulate the synthesis of carbonic anhydrase, an enzyme that regulates pH, and mice lacking this enzyme were seen to develop vascular calciﬁcation [71]. Osteopontin can therefore not only inhibit calciﬁcation but also can potentially reverse the calciﬁcation process by inducing an acidic pH, via its stimulation of carbonic anhydrase production/activity. Furthermore, in atherosclerotic plaques, S100A8 and A9 mRNA and protein have been found in subsets of macrophages, foam cells and endothelial cells of microvessels [28]. S100A8 and A9 are members of the S100 family of calcium ion-binding proteins that are highly expressed in various inﬂammatory conditions; they are also thought to inﬂuence calciﬁcation by modulating phospholipid-Ca 2þ binding within MV. S100A9 in particular was detected in MV extracts isolated from human atherosclerotic plaques. The fact that S100A9 was not detected in VSMCs indicated that MV derived from other cell types, such as macrophages and endothelial cells, are likely to be the source of S100A9-rich MV in plaques. Thus, as for VSMCs, macrophages may have either pro- or anticalciﬁcation eﬀects. Much less is known about how other cell types associated with inﬂammation may inﬂuence atherosclerotic calciﬁcation.

311

312

19 Atherosclerosis: Cellular Aspects

19.3.2 Dendritic Cells, Mast Cells and T Lymphocytes

Cells with the morphological appearance of dendritic cells that also express CD1d and S-100 proteins have been detected in atherosclerotic plaques [6, 72]. Importantly, they have been found in close proximity to calciﬁed deposits in both early fatty streaks and advanced atherosclerotic plaques. They are also found in the non-diseased artery wall in the subendothelial space, and are thought to form networks that are involved in the screening of potentially harmful antigens [73]. In atherosclerotic plaques, they are thought to contribute to calciﬁcation when they undergo necrosis, releasing calcium-binding S100 proteins [72]. Mast cells have been detected in close association with early stages of diﬀuse calciﬁcation, and also with larger calciﬁed deposits in human atherosclerotic plaques [74]. Mast cells produce tryptase, which is a protease with numerous properties; it can degrade ﬁbronectin and collagen VI, and it can activate matrix metalloproteinases while also stimulating collagen synthesis; it also acts as a chemoattractant for neutrophils and eosinophils. Mast cells also release histamine, heparin, TNFa and TGFb, which can aﬀect the calciﬁcation of VSMCs. Other inﬂammatory cells such as T lymphocytes may have potential eﬀects on atherosclerotic calciﬁcation, but this has not yet been investigated. T lymphocytes secrete pro- and anti-inﬂammatory cytokines, and have recently been shown to kill VSMCs in plaques via surface expression of TRAIL [75]. As described in Section 19.3.1, T lymphocytes, dendritic cells and mast cells produce extracellular vesicles, but whether they are capable of initiating calciﬁcation is not yet known.

19.4 The Role of Osteoclasts: Is there a Possibility for Calciﬁcation-Regression?

The calcium crystals in atherosclerotic plaques can appear as: (i) nanoparticulate, diﬀuse deposits; (ii), coalesced crystals that can be several millimeters in length; or (iii) a bone-like material that has many features of true bone, such as osteocytes within the calciﬁed material, bone marrow and osteoblasts lining the bone surface. Osteoclasts are deﬁned as large, multinucleated cells with a ruﬄed border derived from the monocytic lineage of hematopoietic cells that resorb bone, thereby creating characteristic ‘‘pits’’ on the bone surface. The osteoclasts achieve this function by forming an actin ring where the cell tightly apposes the mineralized matrix, thereby allowing a protease-rich, acidic environment necessary for controlled resorption [76]. Osteoclasts express tartrate-resistant acid phosphatase (TRAP), cathepsin K, calcitonin receptors, Hþ ATPase and carbonic anhydrase. In histological and electron microscopy studies, TRAP-positive cells have been detected at the edges of bone-like structures in atherosclerotic lesions [77]. This raises the possibility that some of the larger, bone-like calcium crystals may be dissolved to some extent, thus lowering the amount of calciﬁcation in the

References

plaques. It is also interesting that macrophages exposed to calcium crystals have been reported to undergo osteoclastic diﬀerentiation [78]. In order to mature into osteoclasts, pre-osteoclasts require macrophage colony-stimulating factor (MCSF) and the ligand for receptor activator of NF-kappaB (RANKL) [79], both of which have been detected in atherosclerotic plaques [80, 81]. Loss of osteoprotegerin expression by VSMCs in atherosclerotic plaques may also promote osteoclast diﬀerentiation. Osteoclast-like cells, derived from monocytic cells, may therefore be beneﬁcial in reducing the plaque calciﬁcation burden. However, it should be noted that high local levels of calcium and phosphate ions released by focal acidiﬁcation could potentially further contribute to calcium deposition.

19.5 Conclusions

In the atherosclerotic plaque a growing number of proteins and substances are now known either positively or negatively to regulate the initiation and growth of calcium crystals. No doubt the cellular composition of the plaque and the balance between the cellular production of these factors will determine the resulting level of calciﬁcation. We have yet to determine: (i) whether vesicles produced from diﬀerent cell types in plaques have calciﬁcation potential, and what regulates their release and calciﬁcation; (ii) whether the induction of bone-gene expression by VSMCs in plaques initiates or limits calciﬁcation; and (iii) whether osteoclast-like cells are functional in plaques, causing a reduction in calciﬁcation burden in atherosclerosis.

References 1 H.C. Anderson, Clin. Orthop. Relat. 2

3 4

5 6 7 8 9

Res. 1995, 314, 266. T. Kirsch, G. Harrison, E.E. Golub, H.D. Nah, J. Biol. Chem. 2000, 275, 35577. W. Wang, T. Kirsch, J. Cell Biol. 2002, 157, 1061. M.M. Kockx, G.R. De Meyer, J. Muhring, W. Jacob, H. Bult, A.G. Herman, Circulation 1998, 97, 2307. K.M. Kim, Fed. Proc. 1976, 35, 156. Y.V. Bobryshev, R.S. Lord, B.A. Warren, Atherosclerosis 1995, 118, 9. K.M. Rogers, W.E. Stehbens, Pathology 1986, 18, 64. J.A. Esterly, S. Glagov, D.J. Ferguson, Am. J. Pathol. 1968, 52, 325. M. Aikawa, S. Koletsky, Am. J. Pathol. 1970, 61, 293.

10 H.H. Hsu, N.P. Camacho,

Atherosclerosis 1999, 143, 353. 11 H.H. Hsu, Biochim. Biophys. Acta

2003, 1638, 235. 12 T. Kirsch, W. Wang, D. Pfander,

J. Bone Miner. Res. 2003, 18, 1872. 13 J.M. Cotmore, G. Nichols, Jr., R.E.

Wuthier, Science 1971, 172, 1339. 14 D. Skrtic, E.D. Eanes, Bone Miner.

1992, 16, 109. 15 D. Proudfoot, J.N. Skepper, L.

Hegyi, M.R. Bennett, C.M. Shanahan, P.L. Weissberg, Circ. Res. 2000, 87, 1055. 16 J.L. Reynolds, A.J. Joannides, J.N. Skepper, R. McNair, L.J. Schurgers, D. Proudfoot, W. Jahnen-Dechent, P.L. Weissberg, C.M. Shanahan, J. Am. Soc. Nephrol. 2004, 15, 2857.

313

314

19 Atherosclerosis: Cellular Aspects 17 S. Jono, M.D. McKee, C.E. Murry, A.

18 19

20

21

22

23

24

25 26

27

28

29

30

31

32

Shioi, Y. Nishizawa, K. Mori, H. Morii, C.M. Giachelli, Circ. Res. 2000, 87, E10. T.D. Littlewood, M.R. Bennett, Curr. Opin. Lipidol. 2003, 14, 469. P. Delafontaine, Y.H. Song, Y. Li, Arterioscler. Thromb. Vasc. Biol. 2004, 24, 435. V.A. Patel, Q.J. Zhang, K. Siddle, M.A. Soos, M. Goddard, P.L. Weissberg, M.R. Bennett, Circ. Res. 2001, 88, 895. K. Radcliﬀ, T.B. Tang, J. Lim, Z. Zhang, M. Abedin, L.L. Demer, Y. Tintut, Circ. Res. 2005, 96, 398. O.P. Blanc-Brude, J. Yu, H. Simosa, M.S. Conte, W.C. Sessa, D.C. Altieri, Nat. Med. 2002, 8, 987. J.E. Meredith, Jr., B. Fazeli, M.A. Schwartz, Mol. Biol. Cell 1993, 4, 953. G. Fleckenstein-Grun, F. Thimm, A. Czirfuzs, S. Matyas, M. Frey, J. Cardiovasc. Pharmacol. 1994, 24 (Suppl. 2), S75. C.M. Shanahan, Clin. Nephrol. 2005, 63, 146. Y.V. Bobryshev, R.S. Lord, T. Watanabe, J. Submicrosc. Cytol. Pathol. 1997, 29, 553. M.C. Martinez, A. Tesse, F. Zobairi, R. Andriantsitohaina, Am. J. Physiol. Heart Circ. Physiol. 2005, 288, H1004. M.M. McCormick, F. Rahimi, Y.V. Bobryshev, K. Gaus, H. Zreiqat, H. Cai, R.S. Lord, C.L. Geczy, J. Biol. Chem. 2005, 280, 41521. K. Laulagnier, C. Motta, S. Hamdi, S. Roy, F. Fauvelle, J.F. Pageaux, T. Kobayashi, J.P. Salles, B. Perret, C. Bonnerot, M. Record, Biochem. J. 2004, 380, 161. M.R. Bennett, D.F. Gibson, S.M. Schwartz, J.F. Tait, Circ. Res. 1995, 77, 1136. D. Proudfoot, J.D. Davies, J.N. Skepper, P.L. Weissberg, C.M. Shanahan, Circulation 2002, 106, 3044. D.A. Bird, K.L. Gillotte, S. Horkko, P. Friedman, E.A. Dennis, J.L. Witztum, D. Steinberg, Proc. Natl. Acad. Sci. USA 1999, 96, 6347.

33 M.D. McKee, A. Nanci, Connect.

Tissue Res. 1996, 35, 197. 34 H. Wang, M. Zhang, M. Bianchi,

35

36

37

38 39 40 41

42

43

44

45

46

47

48 49 50

B. Sherry, A. Sama, K.J. Tracey, Proc. Natl. Acad. Sci. USA 1998, 95, 14429. C.M. Shanahan, N.R. Cary, J.R. Salisbury, D. Proudfoot, P.L. Weissberg, M.E. Edmonds, Circulation 1999, 100, 2168. K. Bostrom, K.E. Watson, S. Horn, C. Wortham, I.M. Herman, L.L. Demer, J. Clin. Invest. 1993, 91, 1800. K.L. Tyson, J.L. Reynolds, R. McNair, Q. Zhang, P.L. Weissberg, C.M. Shanahan, Arterioscler. Thromb. Vasc. Biol. 2003, 23, 489. G.K. Hunter, H.A. Goldberg, Proc. Natl. Acad. Sci. USA 1993, 90, 8562. K.A. Hruska, S. Mathew, G. Saab, Circ. Res. 2005, 97, 105. T. Wada, M.D. McKee, S. Steitz, C.M. Giachelli, Circ. Res. 1999, 84, 166. G. Luo, P. Ducy, M.D. McKee, G.J. Pinero, E. Loyer, R.R. Behringer, G. Karsenty, Nature 1997, 386, 78. A.F. Zebboudj, M. Imura, K. Bostrom, J. Biol. Chem. 2002, 277, 4388. K. Bostrom, D. Tsao, S. Shen, Y. Wang, L.L. Demer, J. Biol. Chem. 2001, 276, 14044. R. Wallin, D. Cain, S.M. Hutson, D.C. Sane, R. Loeser, Thromb. Haemost. 2000, 84, 1039. K.E. King, V.P. Iyemere, P.L. Weissberg, C.M. Shanahan, J. Biol. Chem. 2003, 278, 11661. R.N. Willette, J.L. Gu, P.G. Lysko, K.M. Anderson, H. Minehart, T. Yue, J. Vasc. Res. 1999, 36, 120. S. Zhang, I. Fantozzi, D.D. Tigno, E.S. Yi, O. Platoshyn, P.A. Thistlethwaite, J.M. Kriett, G. Yung, L.J. Rubin, J.X. Yuan, Am. J. Physiol. Lung Cell. Mol. Physiol. 2003, 285, L740. M.R. Davies, R.J. Lund, K.A. Hruska, J. Am. Soc. Nephrol. 2003, 14, 1559. H.C. Stary, Z. Kardiol. 2000, 89 (Suppl. 2), 28. F. Parhami, A.D. Morrow, J. Balucan, N. Leitinger, A.D. Watson, Y. Tintut, J.A. Berliner, L.L. Demer, Arterioscler. Thromb. Vasc. Biol. 1997, 17, 680.

References 51 S.A. Steitz, M.Y. Speer, G. Curinga,

52 53

54 55

56 57

58

59

60

61

62

63 64

65 66

67

H.Y. Yang, P. Haynes, R. Aebersold, T. Schinke, G. Karsenty, C.M. Giachelli, Circ. Res. 2001, 89, 1147. X. Li, H.Y. Yang, C.M. Giachelli, Circ. Res. 2006, 98, 905. A. Farzaneh-Far, D. Proudfoot, P.L. Weissberg, C.M. Shanahan, Biochem. Biophys. Res. Commun. 2000, 277, 736. H. Yang, G. Curinga, C.M. Giachelli, Kidney Int. 2004, 66, 2293. R.S. Kaslick, A.I. Chasens, I.D. Mandel, D. Weinstein, R. Waldman, T. Pluhar, R. Lazzara, J. Periodontol. 1970, 41, 93. B. Zimmermann, Cell Tissue Res. 1994, 275, 345. N. Mody, F. Parhami, T.A. Saraﬁan, L.L. Demer, Free Radic. Biol. Med. 2001, 31, 509. Y. Tintut, Z. Alfonso, T. Saini, K. Radcliﬀ, K. Watson, K. Bostrom, L.L. Demer, Circulation 2003, 108, 2505. J.D. Davies, K.L. Carpenter, I.R. Challis, N.L. Figg, R. McNair, D. Proudfoot, P.L. Weissberg, C.M. Shanahan, J. Biol. Chem. 2005, 280, 3911. A.M. Schor, T.D. Allen, A.E. Canﬁeld, P. Sloan, S.L. Schor, J. Cell Sci. 1990, 97 (Pt. 3), 449. M.J. Doherty, B.A. Ashton, S. Walsh, J.N. Beresford, M.E. Grant, A.E. Canﬁeld, J. Bone Miner. Res. 1998, 13, 828. R.W. Evans, H.S. Cheung, D.J. McCarty, Calcif. Tissue Int. 1984, 36, 645. J.A. Hamilton, G. McCarthy, G. Whitty, Arthritis Res. 2001, 3, 242. I.T. Olszak, M.C. Poznansky, R.H. Evans, D. Olson, C. Kos, M.R. Pollak, E.M. Brown, D.T. Scadden, J. Clin. Invest. 2000, 105, 1299. D. Proudfoot, C.M. Shanahan, Herz 2001, 26, 245. P. Laquerriere, A. GrandjeanLaquerriere, E. Jallot, G. Balossier, P. Frayssinet, M. Guenounou, Biomaterials 2003, 24, 2739. I. Nadra, J.C. Mason, P. Philippidis, O. Florey, C.D. Smythe, G.M. McCarthy, R.C. Landis, D.O. Haskard, Circ. Res. 2005, 96, 1248.

68 Y. Tintut, J. Patel, F. Parhami,

69

70

71

72 73

74 75

76 77

78

79

80

81

82

83

L.L. Demer, Circulation 2000, 102, 2636. J.J. Boyle, P.L. Weissberg, M.R. Bennett, Arterioscler. Thromb. Vasc. Biol. 2003, 23, 1553. Y. Tintut, J. Patel, M. Territo, T. Saini, F. Parhami, L.L. Demer, Circulation 2002, 105, 650. S.A. Steitz, M.Y. Speer, M.D. McKee, L. Liaw, M. Almeida, H. Yang, C.M. Giachelli, Am. J. Pathol. 2002, 161, 2035. Y.V. Bobryshev, R.S. Lord, Cardiovasc. Res. 1995, 29, 689. G. Millonig, H. Niederegger, W. Rabl, B.W. Hochleitner, D. Hoefer, N. Romani, G. Wick, Arterioscler. Thromb. Vasc. Biol. 2001, 21, 503. M. Jeziorska, C. McCollum, D.E. Woolley, J. Pathol. 1998, 185, 10. K. Sato, A. Niessner, S.L. Kopecky, R.L. Frye, J.J. Goronzy, C.M. Weyand, J. Exp. Med. 2006, 203, 239. R. Baron, L. Neﬀ, D. Louvard, P.J. Courtoy, J. Cell Biol. 1985, 101, 2210. M. Jeziorska, C. McCollum, D.E. Wooley, Virchows Arch. 1998, 433, 559. K.D. Merkel, J.M. Erdmann, K.P. McHugh, Y. Abu-Amer, F.P. Ross, S.L. Teitelbaum, Am. J. Pathol. 1999, 154, 203. K. Matsuzaki, N. Udagawa, N. Takahashi, K. Yamaguchi, H. Yasuda, N. Shima, T. Morinaga, Y. Toyama, Y. Yabe, K. Higashio, T. Suda, Biochem. Biophys. Res. Commun. 1998, 246, 199. M.E. Rosenfeld, S. Yla-Herttuala, B.A. Lipton, V.A. Ord, J.L. Witztum, D. Steinberg, Am. J. Pathol. 1992, 140, 291. C.R. Dhore, J.P. Cleutjens, E. Lutgens, K.B. Cleutjens, P.P. Geusens, P.J. Kitslaar, J.H. Tordoir, H.M. Spronk, C. Vermeer, M.J. Daemen, Arterioscler. Thromb. Vasc. Biol. 2001, 21, 1998. C.M. Shanahan, N.R. Cary, J.C. Metcalfe, P.L. Weissberg, J. Clin. Invest. 1994, 93, 2393. A.F. Zebboudj, V. Shin, K. Bostrom, J. Cell Biochem. 2003, 90, 756.

315

316

19 Atherosclerosis: Cellular Aspects 84 C. Farrington, I.S. Roberts, A.M.

87 C.M. Shanahan, D. Proudfoot, A.

Heagerty, A.E. Canﬁeld, Biochem. Soc. Trans. 1998, 26, S3. 85 K.E. Watson, K. Bostrom, R. Ravindranath, T. Lam, B. Norton, L.L. Demer, J. Clin. Invest. 1994, 93, 2106. 86 F. Parhami, B. Basseri, J. Hwang, Y. Tintut, L.L. Demer, Circ. Res. 2002, 91, 570.

Farzaneh-Far, P.L. Weissberg, Crit. Rev. Eukaryot. Gene Expr. 1998, 8, 357. 88 A. Simionescu, K. Philips, N. Vyavahare, Biochem. Biophys. Res. Commun. 2005, 334, 524. 89 V. Shin, A.F. Zebboudj, K. Bostrom, J. Vasc. Res. 2004, 41, 193.

317

20 The Biological and Cellular Role of Fetuin Family Proteins in Biomineralization Cora Scha¨fer and Willi Jahnen-Dechent

Abstract

The ﬁnal step of biomineralization is a chemical precipitation reaction that occurs spontaneously in supersaturated or metastable salt solutions. In physiological bone formation, ‘‘osteogenesis’’, and also in pathological mineralization, ‘‘ectopic mineralization or calciﬁcation’’, genetic programs direct precursor cells into a mineralization-competent state. Therefore, all tissues not meant to mineralize must be actively protected against the chance precipitation of mineral. Fetuin-A is a blood protein that acts as a potent inhibitor of ectopic mineralization. Fetuin-A-deﬁcient mice develop severe soft tissue calciﬁcation. Fetuin-A combines with calcium and phosphate into transiently soluble colloidal particles termed calciprotein particles. Thus, fetuin-A is a systemic inhibitor of pathological mineralization complementing local inhibitors acting in a cell- or tissuerestricted fashion. Key words: fetuin, alpha2-HS glycoprotein, calciprotein particles, pathological mineralization, ectopic calciﬁcation, knockout mice.

20.1 Osteogenesis and Bone Mineralization versus Calciﬁcation

In vertebrates, although mineralization is usually restricted to the bones and teeth, it may also occur outside the skeleton ectopically (out of place) when precursor cells inappropriately receive signals to develop into bone cells. The ectopic activation of osteogenesis also contributes to the calciﬁcation of blood vessels, including medial arterial calciﬁcation, atherosclerotic aortic calciﬁcation (intima calciﬁcation), and aortic valve calciﬁcation [1]. Calciﬁcation is a known major shortcoming of vessel prostheses and bioartiﬁcial heart valves [2]. Ectopic calciﬁcation may also be considered a primitive response to chronic infection, whereby pathogens that cannot be cleared are encased. Calciﬁcation sites in soft tissue organs Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

318

20 The Biological and Cellular Role of Fetuin Family Proteins in Biomineralization

20.1 Osteogenesis and Bone Mineralization versus Calciﬁcation

are in fact diagnostic of parasitic infection, for example tuberculoma and breast cancer. The innate immunity response to Mycobacterium tuberculosis involves vitamin D-like compounds, which enhance local calciﬁcation at the infection sites [3]. Calciﬁcation is common in tissue remodeling [4], and most cells are proﬁcient in the shedding of ‘‘matrical lipidic debris’’, which readily calciﬁes. Following tissue insult and the ensuing remodeling activity, large amounts of cell remnants break oﬀ and drift into the extracellular matrix (ECM). Extreme cases of vascular calciﬁcation are associated with typical bone structures, including trabeculae, lacunae, and areas resembling marrow [5, 6]. Several chapters in this book have detailed clear parallels of bone formation and ectopic calciﬁcation (see Chapters 1, 21, and 24). The roles of mineralization regulators and bone-related genes in the context of smooth muscle cell mineralization are illustrated schematically in Figure 20.1. Both osteogenesis and ectopic calciﬁcation invariably terminate in matrix mineralization as the ﬁnal step. The process of matrix mineralization requires a calciﬁable substrate such as collagen, the activity of alkaline phosphatase to provide high local phosphate concentrations, and an absence of soluble inhibitors (e.g., pyrophosphate) or tissue-bound inhibitors (e.g., matrix Gla protein) [7]. Most of the calcium available for precipitation circulates in the blood. Historically, the classic investigations of Blumenthal and colleagues demonstrated that the blood serum contains potent inhibitors of spontaneous calcium salt precipitation [8], preventing mineralization of the blood itself. In the absence of circulat________________________________________________________________________________ H Fig. 20.1 Calciﬁcation-related genes ‘‘at work’’ in vascular smooth muscle cells (VSMCs) undergoing metaplasia. Following ‘‘transdiﬀerentiation’’ into mineralizing VSMCs, the cells elaborate markers of the osteo/chondrogenic lineage. The Na/ phosphate co-transporter Pit-1 mediates phosphate transport into the cell. Elevated phosphate levels in the cytoplasm upregulate expression of Runx2/Cbfa-1, an osteogenic transcription factor. In addition, hyperphosphatemia enhances the production of apoptotic bodies and matrix vesicles that nucleate vascular mineral deposition. The TGF-b-like cytokine bone morphogenetic protein-7 (BMP-7) maintains the contractile phenotype (via similar to mothers against decapentaplegic 6 (Smad 6) and Smad 7 signaling), BMP-2 and TGF-b 1 enhance the osteogenic phenotype. Extracellular calcium is transported into matrix vesicles (MV) by Ca 2þ -channel-forming annexins II, V, and VI. Calcium enhances the phosphate-dependent osteogenic diﬀerentiation by up-regulation of

Pit-1 expression. Pyrophosphate (PP) acts as an inhibitor of basic calcium phosphate (BCP) crystal growth. The concentration of PP is controlled by nucleotide pyrophosphatase/phosphotransferase1 (ENPP1) which generates PP, the PP transporter ANK, and tissue non-speciﬁc alkaline phosphatase (TNAP), which cleaves PP. Unlike MV-mediated mineralization, apoptotic body (AB)-mediated mineralization does not require alkaline phosphatase and annexins. In addition, phosphatidylserine (PS) is localized on opposite sides of the plasma membrane of MV (inside) and AB (outside). PS is externalized to the outer membrane leaﬂet during apoptosis. Fetuin-A prevents intravesicular BCP growth in MV, and thus reduces calcium-induced apoptosis in VSMCs. BMPR-I ¼ BMP receptor-I; MGP ¼ matrix GLA protein; NTP ¼ NMP, nucleotide tri (mono) phosphate; OPN ¼ osteopontin; SM-MHC ¼ smooth muscle myosin heavy chain; TNF-a ¼ tumor necrosis factor-a.

319

320

20 The Biological and Cellular Role of Fetuin Family Proteins in Biomineralization

ing ‘‘systemic’’ inhibitors, an organism would run a high risk of mineralizing the extracellular ﬂuid, which itself is typically a metastable solution with regards to calcium and phosphate solubility. This phenomenon was aptly termed ‘‘Lot’s wife’s problem’’, and has been addressed previously [9]. Candidate systemic inhibitor proteins included bulk serum proteins such as albumin [10, 11]. Apart from their aﬃnity towards calcium apatite, these proteins also bind several more ligands, including lipids, proteases, growth factors and ECM. It was, therefore, difﬁcult to decide if the inhibition of calcium salt precipitation in vitro was fortuitous and due to bulk binding, or whether it represented a true physiological function of a given protein. With the advent of gene targeting technology however, this function could be tested in vivo, in mutant mice. Subsequently, studies conducted in our laboratory have shown that a2 -HS glycoprotein/fetuin-A (genetic symbol Ahsg or Fetua), a serum protein, is a bona ﬁde systemic inhibitor of calciﬁcation.

20.2 a2 -HS Glycoprotein/Fetuin-A is a Systemic Inhibitor of Ectopic Calciﬁcation

The name a2 -HS glycoprotein refers to the fact that this protein migrates with the alpha-2 fraction of serum proteins upon traditional cellulose acetate paper-based electrophoresis. Furthermore, it is reminiscent of the two co-discoverers of this protein in humans [12], namely Heremans [13] and Schmid [14]. Bovine fetuinA was described in 1944 by Pedersen as fetuin (from the Latin, fetus), the most abundant globular serum protein in fetal calf serum [15]. Following the discovery of a second fetuin, fetuin-B [16, 17], the protein originally named fetuin was renamed fetuin-A [16]. Fetuins belong to the cystatin superfamily of cysteine protease inhibitors, which encompass a series of closely related liver-derived serum proteins. Further members of this superfamily sharing cystatin-like domains are the kininogens and histidine-rich glycoproteins [18, 19]. Fetuin-A has been implicated in several diverse functions, including osteogenesis and bone resorption [20], regulation of insulin activity [21], hepatocyte growth factor activity [22], response to systemic inﬂammation [23], and inhibition of unwanted mineralization [24–26]. These seemingly diverse functions show that fetuins are multi-ligand binding proteins that potentially interfere with any biochemical pathway the components of which they can bind and sequester. Using gene knockout technology in mice, the inhibition of ectopic calciﬁcation was shown to be the major biological function of fetuin-A [24, 27]. Fetuin-A-deﬁcient mice on a mixed C57BL/6-129 genetic background displayed only a mild calciﬁcation phenotype [27]. The lack of generalized ectopic mineralization in fetuin-A-deﬁcient mice was somewhat anticipated, because fetuin-A accounted for only a fraction of the inhibition of apatite precipitation observed with total serum of normal mice [25]. Against the calciﬁcation-prone genetic background DBA/2 [28], severe, systemic calciﬁcation occurred (see Fig. 20.2) [24].

20.2 a2 -HS Glycoprotein/Fetuin-A is a Systemic Inhibitor of Ectopic Calciﬁcation

Fig. 20.2 Extensive soft tissue calciﬁcation in Ahsg/fetuin-A-deﬁcient DBA/2 mice. (A) Microcomputed tomography (mCT) of a fetuin-A knockout mouse. Usually, only the bones and teeth should be visible, but the bright spots in the mouse are calciﬁed lesions in the skin, heart, omentum, and kidneys. (B) Histological sections show scale-

like calciﬁed lesions in almost all organs, including tongue, lung, heart, gut, pancreas, and urogenital tract. The liver is mostly spared from calciﬁcation. This phenotype is associated with a reduced life-span and fertility; these mice may live well into adulthood, but stop breeding at an age of about 6 months.

Mice suﬀered calciﬁcation aﬀecting the kidney, myocardium, lungs, and skin. The phenotype of these animals closely resembled uremia-associated arteriolopathy/calciphylaxis, with its clinical hallmarks [24]. Secondary hyperparathyroidism was observed in older mice (aged >5 months) due to kidney damage. A recent study of the cardiovascular system in these mice showed extensive cardiac calciﬁcation and ﬁbrosis with impaired cardiac function [29], reminiscent of models of dystrophic cardiac calciﬁcation [30, 31]. Interestingly, the large arteries were spared from calciﬁcation. Taken together, it was demonstrated by using reverse genetics that the serum protein a2 -HS glycoprotein/fetuin-A is a systemic inhibitor of ectopic calciﬁcation. The question remained, however, as to whether Ahsg deﬁciency is also important in human pathology. To this end, a clinical study was conducted in uremic and healthy subjects, the results of which showed that a lack of Ahsg correlated with the severity of calciﬁcation and, indeed, was (statistically) a highly signiﬁcant

321

322

20 The Biological and Cellular Role of Fetuin Family Proteins in Biomineralization

predictor of short-term morbidity and mortality in uremic patients [32]. This ﬁnding has subsequently been conﬁrmed by two other groups [33, 34].

20.3 The Mechanism of Fetuin-A Inhibition of Calciﬁcation

Ahsg/fetuin-A is easily puriﬁed and can be obtained in large quantities for structure–function analyses. Important parts of the three-dimensional (3-D) structure can be modeled after the known structure of chicken egg white cystatin [35]. Taken together, these data provided an excellent opportunity to study the mechanism of calciﬁcation inhibition by a mammalian protein. By using dynamic light scattering and transmission electron microscopy, it was shown that Ahsg solubilizes apatite as a colloid [26]. This was reminiscent of how apolipoproteins ensheath and thereby solubilize insoluble lipids such as cholesterol. In analogy to the lipoprotein particles of varying buoyant density (HDL, LDL, VLDL, etc.) comprising apolipoproteins and lipids, the calcium- and phosphate-containing Ahsg colloid was referred to as the ‘‘calciprotein particle’’ (CPP), the molecular structure of which is described in detail in Volume I, Chapter 24. Importantly, although the inhibitory eﬀect is transient for up to 36 h at body temperature, within 24 h the CPPs undergo a marked morphological transformation from non-diﬀractive nanospheres with a diameter of @50 nm to larger, more crystalline irregular spheres of up to several hundred nanometers in size. It is important to remember that Ahsg binds calcium phosphate, though bovine fetuin-A calcium binding is rather poor (K d 0:95 104 M) [36]. Even if three calciumbinding sites were to exist, Ahsg/fetuin-A (10 mM serum concentration) would cause only a minute change in the serum calcium concentration of 2.5 mM. Therefore, albumin (1 mM serum concentration) should be considered the major binding protein for ionized calcium, whilst Ahsg/fetuin-A is a highly eﬀective scavenger of basic calcium phosphate (BCP), which precipitates in the absence of this protein [24].

20.4 The Fate of Calciprotein Particles

The route and mechanism of CPP elimination, and the clearance of calciﬁed lipidic debris from the body by endothelial cells and by macrophages is shown schematically in Figure 20.3. There are three equally important mechanisms which prevent extracellular calciﬁcation: (i) hormone-regulated calcium homeostasis to prevent excessive ﬂuctuations in extracellular calcium; (ii) the stabilization of calcium phosphate, which will form spontaneously as a soluble colloid (CPP) to prevent mineral from pre-

20.4 The Fate of Calciprotein Particles

Fig. 20.3 Hypothetical pathway of removal of calciprotein particles (CPP) from circulation by endothelial cells and tissue-resident phagocytes. In healthy individuals CPP may form spontaneously in small numbers, or may occur as a spill-over into the blood of bone catabolism. When mineral homeostasis is severely disturbed (e.g., in renal dialysis patients), bouts of hypercalcemia and hyperphosphatemia will result in the formation of high numbers of CPP. Fetuin-A will be consumed in the process. CPP in the interstitial space will be phagocytosed by macrophages (as depicted), or by any other phagocytosing tissue-resident cell type. In hyperlipidemia, excess low-density lipoprotein (LDL) will concur and be cleared through

similar pathways. Too much CPP remnants, in combination with lipid droplets, may overwhelm the clearing capacity of the reticuloendothelial system phagocytes, and perhaps also the lymphatic system; this may result in apoptosis and the deposition of calciﬁed apoptotic cell remnants also rich in lipid. Fetuin-A appears to stabilize CPP in circulation and mediates their eﬃcient uptake by phagocytes (this latter function remains to be veriﬁed experimentally). Note that many ‘‘inhibitors of calciﬁcation’’ activate monocytes/macrophages and stimulate phagocytosis, and therefore the stimulation of calciﬁed remnant removal may be equally important as the inhibition of precipitation. Figure modiﬁed after Ref. [62].

cipitating and clogging the small vessels; and (iii) an eﬃcient clearing of CPPs by phagocytic cells such that excessive CPP – and hence precipitable mineral – is cleared from the circulation. Whilst the main established mechanisms for the removal of calcium phosphate crystals are phagocytosis and acidiﬁcation [37–39], the reticuloendothelial system (RES) is also capable of removing particulate matter (e.g., cell remnants, molecular aggregates, and ‘‘mineral dirt’’) from the circulation. This network of phagocytic cells encompasses endothelial cells and macrophages in the liver, spleen, and bone marrow, and it is likely that CPPs are phagocytosed and recycled in the RES. Annexin binding may mediate fetuin uptake by pinocytosis, as both annexin-II and -VI have been identiﬁed as putative cell-surface receptors for fetuin-A in the presence of Ca 2þ ions [40]. Cell culture

323

324

20 The Biological and Cellular Role of Fetuin Family Proteins in Biomineralization

assays with ‘‘professional’’ phagocytes have shown that fetuin-A and the related histidine-rich glycoprotein (Hrg) generally act as ‘‘opsonins’’. Hrg binds components of the humoral immune system, and is an opsonizing agent for apoptotic and necrotic cells [41–43]. Fetuins augmented phagocytosis in monocytes, macrophages, osteoclasts and dendritic cells [44–50], in addition to the shuﬄing of calciﬁed vesicles in smooth muscle cells [51] (see Chapter 21). Fetuins also inhibited the inﬂammatory response of neutrophils towards BCP crystals [11]. It is striking that the removal of calciﬁed remnants from bone turnover is usually non-inﬂammatory, and even the large calcareous deposits of Ahsg knockout mice showed no signs of inﬂammatory cell inﬁltrates [24, 29]. Fetuin-A coating may well render phagocytosed material non-inﬂammatory by carrying along antiinﬂammatory polyanions such as spermine [23, 52] and the immunosuppressive cytokine transforming growth factor beta (TGF-b) [20, 53]. In conclusion, the plasma protein fetuin-A protects the body from unwanted calciﬁcation in various ways: Chemically, by binding to calcium phosphate nuclei and inhibiting further mineral growth [25]. Biochemically, by stabilizing and opsonizing CPPs to be cleared from the circulation before the growing crystals reach a critical size at which they begin to precipitate [26]. On the cellular level, by alleviating the detrimental eﬀects of calcium overload during the shuﬄing of calcifying vesicles, thus indirectly inhibiting apoptosis [54] (see also Chapter 21). On a systemic level, by binding and antagonizing TGF-b and BMP, thereby regulating their osteogenic activity [20, 55]. In summary, these roles suggest that fetuin-like proteins have a more general function in the prevention of calciﬁcation and innate immunity, and it is likely that anti-calciﬁcation mechanisms and general tissue remodeling may become merged [56, 57]. Today, a dual role is emerging for fetuin-A and osteopontin (see Chapter 24) as mineral-binding proteins in the inhibition of calciﬁcation and immunomodulation, especially as both agents are present in calciﬁed atherosclerotic plaques [58, 59]. Whilst fetuin-A is derived from serum, osteopontin may be expressed by macrophages resident in the plaques. It might well be that these key components of the anti-calcifying machinery, which originally evolved to fend oﬀ excessive mineral deposition, are now utilized in a more elaborate manner to discriminate self from non-self. Therefore, we should perhaps begin to consider ectopic calciﬁcation not only as a lack of inhibitors of precipitation (the chemical viewpoint) but also as the evolution of molecules that promote opsonization and the timely removal of calciﬁed remnants (the immunological viewpoint). Investigations into apoptotic cell clearing [60], plaque deposition diseases and atherosclerosis [61] have already led to immunology being embraced as the main focus of research in the harnessing of potential therapies. Likewise, the reversal of calciﬁcation by immunological means will undoubtedly soon become a ‘‘hot topic’’ of research.

References

References 1 J.S. Shao, J. Cai, D.A. Towler,

2 3

4 5

6

7

8 9

10 11

12 13

14 15 16

Arterioscler. Thromb. Vasc. Biol. 2006, 26, 1423–1430. F.J. Schoen, R.J. Levy, Ann. Thorac. Surg. 2005, 79, 1072–1080. P.T. Liu, S. Stenger, H. Li, L. Wenzel, B.H. Tan, S.R. Krutzik, M.T. Ochoa, J. Schauber, K. Wu, C. Meinken, D.L. Kamen, M. Wagner, R. Bals, A. Steinmeyer, U. Zugel, R.L. Gallo, D. Eisenberg, M. Hewison, B.W. Hollis, J.S. Adams, B.R. Bloom, R.L. Modlin, Science 2006, 311, 1770–1773. F.N. Ghadially, Ultrastruct. Pathol. 2001, 25, 243–267. E.R. Mohler, III, F. Gannon, C. Reynolds, R. Zimmerman, M.G. Keane, F.S. Kaplan, Circulation 2001, 103, 1522–1528. J.L. Hunt, R. Fairman, M.E. Mitchell, J.P. Carpenter, M. Golden, T. Khalapyan, M. Wolfe, D. Neschis, R. Milner, B. Scoll, A. Cusack, E.R. Mohler, III, Stroke 2002, 33, 1214– 1219. M. Murshed, D. Harmey, J.L. Millan, M.D. McKee, G. Karsenty, Genes Dev. 2005, 19, 1093–1104. N.C. Blumenthal, Clin. Orthop. 1989, 247, 279–289. W. Jahnen-Dechent, in: E. Ba¨uerlein (Ed.), Biomineralization Progress in Biology, Molecular Biology and Application. 2nd edn. Wiley-VCH, Weinheim, 2004. J. Garnett, P. Dieppe, Biochem. J. 1990, 266, 863–868. R.A. Terkeltaub, D.A. Santoro, G. Mandel, N. Mandel, Arthritis Rheum. 1988, 31, 1081–1089. H.E. Schultze, K. Heide, H. Haupt, Naturwissenschaften 1962, 49, 15–17. J.F. Heremans, Les Globulines Se´riques ` me Gamma. Arscia, Brussels, du Syste 1960. K. Schmid, W. Bu¨rgi, Biochim. Biophys. Acta 1961, 47, 440–453. K.O. Pedersen, Nature 1944, 154, 575. E. Olivier, E. Soury, P. Ruminy, A. Husson, F. Parmentier, M. Daveau, J.P. Salier, Biochem. J. 2000, 350, 589–597.

17 B. Denecke, S. Gra¨ber, C. Scha¨fer, A.

18

19

20

21

22

23

24

25

26

27

28

Heiss, M. Wo¨ltje, W. Jahnen-Dechent, Biochem. J. 2003, 376, 135–145. P.P. Cheung, L.A. Cannizzaro, R.W. Colman, Cytogenet. Cell Genet. 1992, 59, 24–26. B.C. Hennis, R.R. Frants, E. Bakker, R.H. Vossen, E.W. van der Poort, L.A. Blonden, S. Cox, P.M. Khan, N.K. Spurr, C. Kluft, Genomics 1994, 19, 195–197. M. Szweras, D. Liu, E.A. Partridge, J. Pawling, B. Sukhu, C. Clokie, W. Jahnen-Dechent, H.C. Tenenbaum, C.J. Swallow, M.D. Grynpas, J.W. Dennis, J. Biol. Chem. 2002, 277, 19991–19997. S.T. Mathews, G.P. Singh, M. Ranalletta, V.J. Cintron, X. Qiang, A.S. Goustin, K.L. Jen, M.J. Charron, W. Jahnen-Dechent, G. Grunberger, Diabetes 2002, 51, 2450–2458. T. Ohnishi, O. Nakamura, N. Arakaki, Y. Daikuhara, Eur. J. Biochem. 1997, 243, 753–761. M. Ombrellino, H. Wang, H. Yang, M. Zhang, J. Vishnubhakat, A. Frazier, L.A. Scher, S.G. Friedman, K.J. Tracey, Shock 2001, 15, 181– 185. C. Scha¨fer, A. Heiss, A. Schwarz, R. Westenfeld, M. Ketteler, J. Floege, W. Mu¨ller-Esterl, T. Schinke, W. JahnenDechent, J. Clin. Invest. 2003, 112, 357–366. T. Schinke, C. Amendt, A. Trindl, O. Po¨schke, W. Mu¨ller-Esterl, W. JahnenDechent, J. Biol. Chem. 1996, 271, 20789–20796. A. Heiss, A. DuChesne, B. Denecke, J. Gro¨tzinger, K. Yamamoto, T. Renne´, W. Jahnen-Dechent, J. Biol. Chem. 2003, 278, 13333–13341. W. Jahnen-Dechent, T. Schinke, A. Trindl, W. Mu¨ller-Esterl, F. Sablitzky, S. Kaiser, M. Blessing, J. Biol. Chem. 1997, 272, 31496–31503. F.A. van den Broek, R. Bakker, M. den Bieman, A.X. FielmichBouwman, A.G. Lemmens, H.A. van Lith, I. Nissen, J.M. Ritskes-Hoitinga, G. van Tintelen, L.F. van Zutphen,

325

326

20 The Biological and Cellular Role of Fetuin Family Proteins in Biomineralization

29

30

31

32

33

34

35

36

37 38 39

40

41

42

Biochem. Biophys. Res. Commun. 1998, 253, 204–208. M.W. Merx, C. Scha¨fer, R. Westenfeld, V. Brandenburg, S. Hidajat, C. Weber, M. Ketteler, W. JahnenDechent, J. Am. Soc. Nephrol. 2005, 16, 3357–3364. S. Korﬀ, N. Riechert, F. Schoensiegel, D. Weichenhan, F. Autschbach, H.A. Katus, B.T. Ivandic, Virchows Arch. 2005, 1–9. S. Korﬀ, F. Schoensiegel, N. Riechert, D. Weichenhan, H.A. Katus, B.T. Ivandic, Physiol. Genomics 2006, 25, 387–392. M. Ketteler, P. Bongartz, R. Westenfeld, J.E. Wildberger, A.H. Mahnken, R. Bohm, T. Metzger, C. Wanner, W. Jahnen-Dechent, J. Floege, Lancet 2003, 361, 827–833. P. Stenvinkel, K. Wang, A.R. Qureshi, J. Axelsson, R. Pecoits-Filho, P. Gao, P. Barany, B. Lindholm, T. Jogestrand, O. Heimburger, C. Holmes, M. Schalling, L. Nordfors, Kidney Int. 2005, 67, 2383–2392. A.Y. Wang, J. Woo, C.W. Lam, M. Wang, I.H. Chan, P. Gao, S.F. Lui, P.K. Li, J.E. Sanderson, Nephrol. Dial. Transplant. 2005, 20, 1676–1685. W. Bode, R. Engh, D. Musil, U. Thiele, R. Huber, A. Karshikov, J. Brzin, J. Kos, V. Turk, EMBO J. 1988, 7, 2593–2599. M. Suzuki, H. Shimokawa, Y. Takagi, S. Sasaki, J. Exp. Zool. 1994, 270, 501–507. M.D. Fallon, S.L. Teitelbaum, A.J. Kahn, Lab. Invest. 1983, 49, 159–164. S.A. Nesbitt, M.A. Horton, Science 1997, 276, 266–269. J. Salo, P. Lehenkari, M. Mulari, K. Metsikko, H.K. Vaananen, Science 1997, 276, 270–273. M.N. Kundranda, S. Ray, M. Saria, D. Friedman, L.M. Matrisian, P. Lukyanov, J. Ochieng, Biochim. Biophys. Acta 2004, 1693, 111–123. N.N. Gorgani, C.R. Parish, S.B. Easterbrook Smith, J.G. Altin, Biochemistry 1997, 36, 6653–6662. N.N. Gorgani, B.A. Smith, D.H. Kono, A.N. Theoﬁlopoulos, J. Immunol. 2002, 169, 4745–4751.

43 A.L. Jones, I.K. Poon, M.D. Hulett,

44

45

46 47 48

49

50

51

52

53

54

55

56 57 58

C.R. Parish, J. Biol. Chem. 2005, 280, 35733–35741. S.P. Hart, C. Jackson, L.M. Kremmel, M.S. McNeill, H. Jersmann, K.M. Alexander, J.A. Ross, I. Dransﬁeld, Am. J. Pathol. 2003, 162, 1011–1018. H.P. Jersmann, K.A. Ross, S. Vivers, S.B. Brown, C. Haslett, I. Dransﬁeld, Cytometry 2003, 51A, 7–15. J.G. Lewis, C.M. Andre, Immunol. Commun. 1981, 10, 541–547. J.G. Lewis, C.M. Andre, Immunology 1981, 42, 481–487. M.S. Lamkin, C. Colclasure, W.S. Lloyd, J.M. Doherty, W. Gonnerman, K. Schmid, R.B. Nimberg, Cancer Res. 1986, 46, 4650–4655. G.C. Colclasure, W.S. Lloyd, M. Lamkin, W. Gonnerman, R.F. Troxler, G.D. Oﬀner, W. Bu¨rgi, K. Schmid, R.B. Nimberg, J. Clin. Endocrinol. Metab. 1988, 66, 187–192. L. Thiele, J.E. Diederichs, R. Reszka, H.P. Merkle, E. Walter, Biomaterials 2003, 24, 1409–1418. J.L. Reynolds, J.N. Skepper, R. McNair, T. Kasama, K. Gupta, P.L. Weissberg, W. Jahnen-Dechent, C.M. Shanahan, J. Am. Soc. Nephrol. 2005, 16, 2920–2930. H. Wang, M. Zhang, M. Bianchi, B. Sherry, A. Sama, K.J. Tracey, Proc. Natl. Acad. Sci. USA 1998, 95, 14429–14434. M. Demetriou, C. Binkert, B. Sukhu, H.C. Tenenbaum, J.W. Dennis, J. Biol. Chem. 1996, 271, 12755–12761. J.L. Reynolds, A.J. Joannides, J.N. Skepper, R. McNair, L.J. Schurgers, D. Proudfoot, W. Jahnen-Dechent, P.L. Weissberg, C.M. Shanahan, J. Am. Soc. Nephrol. 2004, 15, 2857– 2867. B. Rittenberg, E. Partridge, G. Baker, C. Clokie, R. Zohar, J.W. Dennis, H.C. Tenenbaum, J. Orthop. Res. 2005, 23, 653–662. J. Savill, V. Fadok, Nature 2000, 407, 784–788. P.M. Henson, D.L. Bratton, V.A. Fadok, Curr. Biol. 2001, 11, R795–R805. F.W. Keeley, E.E. Sitarz, Atherosclerosis 1985, 55, 63–69.

References 59 Y. Matsui, S.R. Rittling, H. Okamoto,

M. Inobe, N. Jia, T. Shimizu, M. Akino, T. Sugawara, J. Morimoto, C. Kimura, S. Kon, D. Denhardt, A. Kitabatake, T. Uede, Arterioscler. Thromb. Vasc. Biol. 2003, 23, 1029– 1034.

60 V.A. Fadok, G. Chimini, Semin.

Immunol. 2001, 13, 365–372. 61 Y.I. Miller, M.K. Chang, C.J. Binder,

P.X. Shaw, J.L. Witztum, Curr. Opin. Lipidol. 2003, 14, 437–445. 62 P. Libby, Nature 2002, 420, 868– 874.

327

329

21 Stone Formation Pierfrancesco Bassi

Abstract

The formation of stones in humans stems from a wide range of underlying disorders. That clinicians seek the underlying causes of lithiasis is imperative in order to direct the management of the condition. During recent years, many advances have been made in the genetics, pathophysiology, diagnostic imaging, medical treatment, medical prevention, and surgical intervention of lithiasis. In this chapter we provide a brief general background, and focus mainly on the pathophysiology of stones. Although important advances have been made in understanding lithiasis from the basis of single gene defects, our understanding of polygenetic causes of stones remains largely elusive. A substantial proportion of data that have resulted in new methods of treatment and prevention, and which may be either empirical or deﬁnitive, has focused on the chemical composition of the precipitating solute(s). Advances in the management of lithiasis depend on the combined eﬀorts of clinicians and scientists to understand the pathophysiology of stone formation. Key words: urinary stones, urolithiasis, testicular microlithiasis, biliary stones, gallbladder stones, sialolithiasis, supragingival stones, pancreatic stones, broncholithiasis, pulmonary alveolar microlithiasis.

21.1 Urinary Stones

Although the formation of stones in the urinary tract aﬀects 5 to 10% of the population in Europe and North America [1], even higher frequencies have been reported from other parts of world. Indeed, there are only a few geographical areas in which stone disease is rare, for example in Greenland and in the coastal areas of Japan [2]. The annual incidence of stone formation in the industrialized world is generally considered to be 1500–2000 cases per one million population [2]. AlHandbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

330

21 Stone Formation

though the chemical composition of stones varies widely, a common denominator is the high risk of recurrent stone formation when the ﬁrst stone is removed, although there is considerable variation among individuals. 21.1.1 Pathogenesis

Urinary stone formation, or urolithiasis, is a consequence of complex physical processes [3]. Whilst the pathogenesis of urinary stone formation is a multifactorial process, the main prerequisite is that of urinary crystal formation, and for this the urine must be supersaturated [4]. The crystallization potential of urine is related not only to the concentration of the salt(s) in question, but also to the presence or absence of other compounds, such as inhibitors, complexors, or promotors [5]. Other factors are represented by anatomic abnormalities. The sequence of events leading to urinary stone formation is: saturation ! supersaturation ! nucleation ! crystal growth or aggregation ! crystal retention ! stone formation. Some or all of these processes contribute to the development of a clinically signiﬁcant stone. The supersaturation of urine is considered the most important driving force behind stone formation. This is based on the binding of salts, which occurs after a certain concentration is reached. The thermodynamic solubility product (KSP) of a compound deﬁnes the saturation of that compound in solution [6], and is equal to the product of a pure chemical in equilibrium between a solid and solvent in solution. If the salt concentration is less than the KSP, the compound remains in solution, but if the salt concentration exceeds the KSP then the compound will precipitate. This process is called homogeneous nucleation. Nuclei are formed from the ﬁrst crystals that do not dissolve, and have a characteristic lattice pattern. In urine, nuclei usually form on existing surfaces, a process known as heterogeneous nucleation. Epithelial cells, urinary casts, red blood cells and other crystals can act as nucleating foci in urine. The degree of saturation required for heterogeneous nucleation to occur is much less than for homogeneous nucleation. However, once a nucleus is formed – and particularly if it becomes anchored – then crystallization can occur at a lower saturation than is required for the formation of the initial nucleus. The initial nucleus can grow by the precipitation of additional salt on the lattice framework (crystal growth). In humans, the earliest site of stone formation is the papillary duct or the collecting duct tubule, where the diameter is 50 to 200 mm. The time needed for a crystal to grow to a diameter of 200 mm depends on the state of supersaturation of the urine. When the nuclei have formed they are able to separate one from another, ﬂoat freely, and become kinetically active. Under certain circumstances, these nuclei come into close contact and, due to chemical or electrical forces, can bind to each other, a process known as crystal aggregation. Although it is impossible for crystal growth alone to give rise to a crystal which is large enough to occlude the lumen of the collecting duct, aggregates of crystals easily can attain such a size

21.1 Urinary Stones

[7]. This combination of crystal growth and crystal aggregation provides the explanation for the genesis of urinary stones. Although crystals may form, they generally do not become very large and tend to ‘‘wash out’’ before they become clinically signiﬁcant, due mainly to their short transit time within the urinary tract. However, an anatomic or functional abnormality can cause an obstruction to the ﬂow of urine and the retention of urinary crystals (crystal retention). In general, crystal aggregates are too fragile to occlude a collecting duct for long enough to give rise to a stone, but if a crystal is retained in the kidney then growth can occur over long periods of time whenever the urine is supersaturated or there is aggregation of new crystals [8]. Anatomical abnormalities in the kidney, such as medullary sponge kidney or ureteropelvic junction, or even an increase in crystal epithelial adherence, can lead to crystal retention. 21.1.1.1 Inhibitors of Stone Formation In normal urine, the concentration of calcium oxalate (CaOx) may be many-fold greater than its solubility. This is made possible due to the presence in urine of inhibitors; these substances modify or alter crystal growth, thus preventing stone formation, and they also allow for higher concentrations of calcium phosphate to be held in solution than in pure solvents. Although urine may be supersaturated with a salt, the inhibitors can prevent stone formation [9] by forming complexes with active surface compounds, thereby reducing the binding of calcium to oxalate. In fact, many individuals who excrete increased amounts of calcium and oxalate do not form stones. The list of compounds that can retard stone crystallization is extensive, and includes metal ions such as magnesium, simple compounds such as citrate, and macromolecules [10–12]. 21.1.1.1.1 Citrate Citrate is the most important urinary stone inhibitor [13], and the most abundant organic anion in human urine. By chelating calcium ions, citrate eﬃciently lowers supersaturation, which is the driving force of crystallization [14], and also reduces induction time and the rate of nucleation [15]. The presence of urinary citrate also permits the excretion of basic compounds, without raising the urinary pH, and this in turn provides a defense against alkali loads, without precipitating calcium phosphate. Citrate is able to complex calcium in a soluble form, which in turn prevents the crystal growth of calcium phosphate and CaOx in urine [16, 17]. 21.1.1.1.2 Proteins A number of other important stone formation inhibitors have been identiﬁed, the absence or reduction of which can lead to increased stone formation [18]. For example, normal urine contains small amounts of protein, much of which is secreted by the tubular epithelial cells of the kidney. These proteinaceous urinary inhibitors include Tamm–Horsfall mucoprotein (THP) [19], one of the main components of the urinary proteins. THP is a glycoprotein that is produced and se-

331

332

21 Stone Formation

creted by the thick ascendant limb of the loop of Henle. In normal subjects, THP inhibits crystal aggregation but has little eﬀect on the nucleation and growth of CaOx crystals [20]. However, THP activity is inﬂuenced by its own concentration, urinary pH and ionic strength, and it therefore plays a dual role in crystal formation, depending on the environmental conditions. THP isolated from the urine of recurrent stone-formers sometimes promotes CaOx aggregation; this is due to its tendency to self-aggregate, which prevents it from interacting eﬀectively with CaOx monohydrate crystals [19]. Another inhibitor, nephrocalcin, is an acidic glycoprotein which contains gcarboxyglutamic acid and acts by impairing crystal nucleation, growth, and aggregation. Osteopontin (OPN) is a phosphorylated protein which is widely distributed among the tissues and is found in association with dystrophic calciﬁcation, including the organic matrix of kidney stones [21–23]. OPN is a powerful inhibitor of crystal formation and growth in vitro, and there is also some evidence that it is implicated in stone disease, where the primary emphasis is on its interaction with CaOx, the major constituent of calcium-containing stones [24]. Fibronectin is a multifunctional a2-glycoprotein that is distributed throughout the extracellular matrix and body ﬂuids [25, 26]. Fibronectin is oversecreted from the renal tubular cells as a result of CaOx crystal stimulation, and inhibits the aggregation of CaOx crystals and their adhesion to renal tubular cells [27]. Substances which form soluble complexes with the lattice ions for speciﬁc crystals, such as CaOx, are termed complexing agents. These reduce free ionic activity and thus lower the level of saturation of the stone-forming substance. Pure promotors of stones are rare; glycosaminoglycans promote crystal nucleation but inhibit crystal aggregation and growth [28]. THP may act either as a promotor or an inhibitor of crystallization. The presence of a non-crystalline organic matrix in urinary stone has been recognized since the 17th century [29], with chemical analyses of the matrix having revealed the presence of 65% hexosamine and 10% bound water. The matrix also contains substances similar to the uromucoids found in urine, except that the latter also contains 3.5% sialic acid. Some authors have suggested that the matrix is nothing more than a co-precipitate with the crystals that form stones [30], but taking into consideration the microscopic ﬁndings that show the presence of a concentric, lamellated structure, the matrix may be a ground substance for stone formation. Du Toit [31] has suggested that one factor in stone formation might be an alteration in the excretion of the enzymes urokinase and sialidare, as decreased urokinase and increased sialidare levels in urine lead to the formation of mineralizable stone matrix. It is known that Proteus mirabilis and Escherichia coli each decrease urokinase and increase sialidase activity. According to these ﬁndings, E. coli may cause urolithiasis by producing matrix substances that in turn increase crystal adherence to the epithelium. Matrix calculi are found predominantly in individuals with recurrent infections by urease-producing organisms; they are radiolucent and may be mistaken for a uric acid stone, although the association with alkaline urine and infection helps to maintain this distinction. In-

21.1 Urinary Stones

vestigations in whole animals and in cells have indicated that exposure to high levels of oxalate and/or CaOx crystals produces cellular injury at the level of the proximal kidney tubular cells, in turn inducing changes that may range from cellular adaptation to cell death [32]. It has also been shown that injury produced by CaOx is a predisposing factor to subsequent CaOx stone formation [33]. In fact, oxalate increases the availability of free radicals by inhibiting the enzymes responsible for their degradation; thus, reactive oxygen species (ROS) can damage the mitochondrial membranes and produce a reduction in mitochondrial transmembrane potential [34], events which are recognized as early processes in the apoptotic pathways. It has also been shown that the exposure of renal epithelial cells to oxalate produces increased DNA synthesis, altered gene expression, and apoptosis. The role of free radicals in lithogenesis is also supported by the fact that vitamin E and selenium – both known anti-oxidants – can prevent in vitro the lipid peroxidation of renal proximal tubular cells [35]. Additionally, oxalate injury can result in shedding of the microvillous brush border. Changes in the membrane lipids may manifest as an increase in lipid urinary content [36]. As a consequence, a signiﬁcant increase in the urinary excretion of proteins and lipids in stone-formers may be indicative of cellular damage resulting from prolonged exposure to oxalate and/or the deposition of CaOx crystals in the tubules [37]. CaOx, struvite and uric acid stones contain two- to fourfold more lipids than proteins [38]. The lipid matrix is a good nucleator of CaOx crystals from a metastable solution, and the formation of a complex between calcium and acidic phospholipids is considered to be one of the initial steps in stone formation [39]. The results of several studies have shown that lipids participating in the crystallization of calcium phosphate form complexes with calcium, and bind tightly to the crystals [37]. Interestingly, some cells do not seem to respond to oxalate injury, which suggests that changes in gene expression might protect against apoptosis and consequently against lithiasis. 21.1.2 Classiﬁcation of Urinary Stones

Although kidney stones are composed of diﬀerent types of crystal, most of the crystals present are CaOx or calcium phosphate (or a combination thereof ), uric acid, magnesium ammonium phosphate (also known as struvite or infection stones), cystine, and miscellaneous types, as might occur with high urinary levels of drug metabolites. 21.1.2.1 Calcium Stones Calcium stones (Fig. 21.1) are heterogeneous in both composition and pathophysiology. CaOx and calcium phosphate are the most frequently encountered compounds, but CaOx stones occur much more frequently than their calcium phosphate counterparts. Some 70 years ago, Randall described plaque-like lesions in the renal papillae, which were invariably present in patients with CaOx stones, although they sometimes were also present in individuals who did not form

333

334

21 Stone Formation

Fig. 21.1 Plain X-radiograph (left) and urogram (right) showing a calcium oxalate (CaOx) stone (circles) in the proximal ureter.

stones [4]. Now referred to as Randall’s plaques, these lesions were believed to be the nidus upon which CaOx stones arose and grew. Microscopically, these plaques seem to arise from the basement membrane of the thin limbs of the loops of Henle, expand through the interstitium (sometimes encasing the renal tubules and vas recta), and eventually protrude into the uroepithelium in the renal papillae. Randall’s plaques are composed of calcium phosphate, and seem to provide the platform on which CaOx crystals may form initially through heterogeneous nucleation, and then to grow to a nephrolith. Hypercalciuria, hypercalcemia and hyperoxaluria are the three most common metabolic disorders that lead to calcium stone formation. These and other causes of calcium-containing stones are listed in Table 21.1. 21.1.2.1.1 Hypercalciuria This common metabolic abnormality is encountered in 60% of stone-formers. There are several metabolic causes of hypercalciuria that lead to stone formation [40]. Absorptive hypercalciuria is caused by an increased intestinal calcium absorption; this leads to an increase in renal ﬁltered calcium and in turn to hypercalciuria. Another problem is that a high renal calcium level can produce a hypocalcemic state, which then causes a secondary hyperparathyroidism. Parathyroid

21.1 Urinary Stones Table 21.1 Common underlying causes of calcium-containing calculi.

Hypercalciuria (50%)

Idiopathic hypercalciuria (90%) Primary hyperparathyroidism High-sodium diet High-protein diet Medications (loop diuretics, calcium supplements)

Hyperoxaluria (10–20%)

Crohn’s disease Chronic pancreatitis Celiac sprue High vitamin C intake (in some patients) Calcium restriction Primary hyperoxaluria

Hypocitraturia (10–40%)

Idiopathic (90%) Metabolic acidosis (diarrhea) Distal renal tubular acidosis Potassium depletion

Hyperuricosuria (10–20%)

Excessive dietary purine intake

Other

Genitourinary abnormalities (medullary sponge kidney)

hormone causes intestinal absorption and bone resorption of calcium, which leads to further calcium wasting and hypercalciuria. Resorptive hypercalciuria is caused by excessive bone resorption of calcium, primarily as a result of hyperparathyroidism. A renal phosphate leak creates hyperphosphaturia, which can produce the excess 1,25-(OH)2 -vitamin D that causes hypercalciuria. 21.1.2.1.2 Hypercalcemia Hypercalcemia is caused by hyperparathyroidism, sarcoidosis, steroid therapies, malignancy, idiopathic causes, and immobilization. Immobilization, which is the second most common cause of stone disease, causes an increased bone resorption of calcium which, in turn, increases renal ﬁltration. This leads to hypercalciuria and the precipitation of urinary CaOx and calcium phosphate stones. 21.1.2.1.3 Hyperoxaluria Primary hyperoxaluria is caused by an autosomal recessive disorder in oxalate biosynthesis that causes an increase in the hepatic production of oxalate. Secondary hyperoxaluria, enteric oxaluria, is caused by the intestinal hyperabsorption of oxalate. Enteric hyperoxaluria is seen in inﬂammatory bowel disease, bowel resection, and small bowel bypass procedures. With these conditions there is an increase in bile salt and fatty acids that combine with calcium, leading to increased oxalate available for absorption. This eﬀect of oxalate is dependent on the so-

335

336

21 Stone Formation

called calcium-oxalate interaction, whereby oxalate absorption from the bowel is aﬀected by formation of poorly absorbed CaOx, and the amount of free oxalate in urine is aﬀected by formation of a soluble complex of calcium and oxalate. In fact, with increased intestinal absorption of oxalate, there is an increase in urinary oxalate leading to formation of CaOx stones, because oxalate complexes with calcium within the renal tubules, precipitates, and aggregates to form stones. The bile acids and fats within the lumen of the bowel bind and reduce the amount of free calcium, leaving an increased amount of oxalate to be absorbed from the intestine and excreted in the urine. These patients also have low urinary citrate and magnesium levels as a result of chronic metabolic acidosis due to chronic diarrhea. All of these factors lead to CaOx stone formation. 21.1.2.2 Uric Acid Stones Uric acid is an end-product of purine metabolism, and is the same crystalline material that causes gout, an arthritic condition. Hence, purine-rich foods (e.g., red meat, ﬁsh, chicken) should be avoided by people who are prone to uric acid stone formation. A high body-mass index, glucose intolerance, and overt type 2 diabetes are common ﬁndings in uric acid stone-formers. Conditions associated with uric acid nephrolithiasis are listed in Table 21.2.

Table 21.2 Conditions associated with uric acid nephrolithiasis.

Congenital conditions

Disorders causing uric acid overproduction

Hypoxanthine guanine phosphoribosyl-

transferase deﬁciency Phosphoribosylpyrophosphate synthetase

over-activity Urate transporter defects Acquired conditions

Volume depletion Increased purine states

Uricosuric drugs

Idiopathic

Gouty diathesis Primary gout

Glucose-6-phosphatase deﬁciency Congenital hypouricemia with

hyperuricosuria

Excessive dehydration Chronic diarrhea Myelo/lymphoproliferative disorders Other malignancies with or without chemotherapy Hemolytic disorders High animal protein intake Probenecid High-dose salicylates Radiocontrast agents

21.1 Urinary Stones

Fig. 21.2 Urogram showing a renal radiolucent uric acid stone (arrow) in the pelvis of the kidney.

The solubility of uric acid depends on the acidity or alkalinity of the urine. In acidic urine (pH < 5.5), uric acid crystals are precipitated and this leads to stone formation. If urine is alkaline, however, the uric acid remains soluble and does not precipitate. Knowledge of these facts forms the basis of the medical treatment of uric acid stones. Gout, myeloproliferative disorders and chemotherapy are common causes of uric acid stones, which are radiolucent on plain X-ray, but can be visualized with urography (Fig. 21.2) or computed tomography scanning. 21.1.2.3 Magnesium Ammonium Phosphate Stones, Struvite or Infection Stones Depending on their composition, these stones are known as triple-phosphate, struvite, or infection stones. Infection stones are formed as a result of persistent infections caused by urease-producing bacteria, and urinary tract obstructions are frequently involved [41]. Infection stones, which are characterized by their exceptionally rapid growth rate, consist of monoammonium urate, struvite (magnesium ammonium phosphate; MgNH4 PO4 6H2 O) and/or carbonate apatite [42]. Carbonate apatite begins to crystallize at a urinary pH of 6.8, and struvite precipitates only above pH 7.2. The pathogenesis of struvite stones mainly relates to urinary tract infection by urease-producing bacteria (this often correlates with urinary ﬂow disturbances). In the urine, the urea is converted to ammonia and CO2 by the catalytic action of the bacterial enzyme, urease; this causes the urinary pH to rise and, after a

337

338

21 Stone Formation

short time, to be maintained at pH 7.2 to 8.0. The ammonia continues to be hydrolyzed, forming ammonium ions, but in the presence of urease-producing bacteria, Mg 2þ – which is a normal component of urine – is rendered insoluble at alkaline pH and precipitates as struvite due to the presence of NH4 þ and PO4 3 . Bacteria that produce urease act on the urea present in urine to form ammonia and bicarbonate/carbonate ions. The most common bacterium associated with struvite stones is Proteus [41], although other bacteria such as Klebsiella, Pseudomonas and Staphylococcus species have also been implicated. 21.1.2.4 Cystine Stones Cystinuria is an autosomal recessive disorder caused by the defective transport of cystine and dibasic amino acids (lysine, ornithine, arginine) in the brush border of the proximal renal tubules and intestinal tract. This leads to an increased urinary excretion of these compounds, although the only compound to form stones is cystine, due to its low solubility inducing precipitation in the urinary tract [43]. Factors that should raise suspicion are a young age at presentation, mildly radioopaque stones (on urography), a family history of the condition, and characteristic hexagonal cystine crystals in the urine.

21.1.3 Risk Factors

Risk factors associated with the formation of urinary stones may be distinguish between non-genetic and genetic categories. 21.1.3.1 Non-Genetic Factors 21.1.3.1.1 Diet Whilst no kidney stone disorder can be explained by nutrition alone, diet does play a crucial role in the pathogenesis of the most widespread forms of nephrolithiasis, such as CaOx and calcium phosphate and uric acid stones, and triggering the formation of stones in people who are suitably predisposed. The ﬁrst crucial point concerns water intake, the importance of which cannot be underestimated in the prevention of stone diseases. The most eﬀective way to increase the urine volume (which in stone-forming adults should always be in excess of 2 L per day) is to drink suﬃcient quantities of water [44]. With regards to CaOx stones, considerable controversy persists surrounding the various factors that determine urinary oxalate levels. The majority (55–70%) of urinary oxalate is derived from the metabolism of glyoxalate and ascorbic acid, and the remainder is from dietary sources. Nevertheless, diet and absorption may provide the critical quantity of additional oxalate necessary to trigger the formation of CaOx stones. Some foods are particularly high in oxalate content; these include rhubarb, spinach, beetroot, parsley, okra, soya beans and many soya prod-

21.1 Urinary Stones

ucts, yams, wheat bran, nuts, peanut butter, sesame seeds, black pepper, chocolate, chocolate drinks, and tea. Most fruits, cereals and vegetables contain only small quantities of oxalate, but if these are extracted and concentrated (e.g., in cranberry juice tablets) then the resulting oxalate content can be very high. The percentage of oxalate absorbed may vary widely depending on dietary intake and composition. It is also dependent on body size, presumably as a function of the greater intestinal area available for absorption [11, 44]. Whilst several risk factors have been implicated in the pathogenesis of uric acid nephrolithiasis, the underlying mechanisms are not completely understood. In fact, stones can develop as a result of congenital or acquired causes, or they may be idiopathic. In the case of acquired causes, a low-carbohydrate high-protein diet (e.g., the Atkins’ diet), which is commonly used for weight reduction, delivers an exaggerated acid load to the kidney, reducing the urinary pH and increasing the propensity for uric acid stone formation. Several recent studies on the metabolic aspects of idiopathic uric acid nephrolithiasis have suggested a link between insulin resistance and low urinary pH in gouty diathesis. Individuals with this condition share many features of the metabolic syndrome, including hyperglycemia and hyperinsulinemia, hypertension, a high body mass index, and hypertriglyceridemia. Through a variety of mechanisms – but ﬁrst and foremost through an increase in the body’s acid load – an excess intake of animal proteins induces multiple metabolic urinary alterations such as hypercalciuria, hyperoxaluria, hyperuricuria, hypocitraturia and excessive urinary acidiﬁcation, thus exposing the subject to the risk of forming both calcium and uric acid stones. Diets high in animal protein have been shown to raise the risk for kidney stone disease, and indeed epidemiological studies have disclosed a strong association between stone disease and the more aﬄuent members of industrialized societies who consume large amounts of animal protein. Several mechanisms have been suggested to explain protein-induced hypercalciuria. One theory suggests that excessive animal protein intake provides an excess of sulfur-containing amino acids (e.g., methionine) that can be metabolized to sulfuric acid. In an attempt to help buﬀer this excess acid load, bone is resorbed to provide phosphate and carbonate for buﬀering purposes. The other product of this bone dissolution, namely calcium, is cleared by the kidney, resulting in increased urinary calcium excretion. According to an alternative mechanism for explaining the deleterious eﬀects of animal protein intake on stone disease, the animal protein diet – when its electrolyte composition and protein content were kept the same as for a vegetarian diet – conferred an increased risk for uric acid stones but, because of opposing factors, not for CaOx or calcium phosphate stones [45]. An excessive intake of carbohydrates, especially of the simple form, can generate a state of hyperinsulinism, leading to reduced renal tubular reabsorption of calcium, and hence to hypercalciuria. In contrast, there are carbohydrates among the precursors of the endogenous synthesis of oxalate, so that hyperoxaluria can

339

340

21 Stone Formation

also be generated by this route. This is the reason why a limited intake of simple sugars is recommended in the diet. Excessive fat intake, except for foods containing high quantities of omega-3 fatty acids, has also been proven to increase oxaluria, possibly due to an increase in the intestinal absorption of oxalate [46]. The traditional advice given to stoneformers to reduce foods containing calcium by eliminating milk and milk derivatives is no longer accepted as valid. In fact, a low-calcium diet can reduce calciuria in hyperabsorbers of calcium, but can trigger a negative calcium balance and a loss of bone calcium. Unfortunately, very few trials have evaluated the impact of diet change on stone recurrences in a controlled manner and for a suﬃciently long period of time (at least 3 years). Indeed, the famous low-calcium diet that has been applied for years throughout the world to reduce calciuria in hypercalciuric subjects and to attempt to prevent stone recurrences, has never been tested for prolonged periods under controlled conditions [47]. For this reason, in 1993 Borghi et al. [48] commenced a randomized trial in hypercalciuric stone-forming males, aimed at comparing the eﬀects of the ‘‘traditional’’ low-calcium diet (400 mg day1 ) with an ‘‘antistone-forming’’ diet consisting of a normal calcium intake, a low animal protein and salt intake, and a high potassium intake. The results of this trial showed that calcium excretion fell signiﬁcantly with both diets, to a mean value of approximately 170 mg per day. In contrast, urinary oxalate increased in patients following the ‘‘traditional’’ low-calcium diet, and decreased in those treated with the ‘‘anti-stone-forming’’ diet. After 5 years, 12 of the 60 (20%) hypercalciuric subjects following the ‘‘anti-stone-forming’’ diet had recurrences, compared to 23 of the 60 (38.3%) patients following the ‘‘traditional’’ low-calcium diet [49]. Vitamin E, an anti-oxidant, has been shown to prevent the lipid peroxidation of renal proximal tubular cells in vitro. Selenium, acting as an anti-oxidant is also involved. Vegetables and fruits increase the urinary excretion of the stone-inhibiting citrate, and consequently the consumption of foods with a high oxalate content (spinach, rhubarb, beetroot, chard and nuts) should always be kept to a minimum, or at least combined at the same time with foods providing a plentiful supply of calcium (e.g., spinach with a cheese gratin). This prevents the absorption of large quantities of oxalate from the intestine, which would lead to an increase in its excretion via the urine [46]. Some authors [50] have reported that green tea has anti-oxidant eﬀects, with such treatment having been shown to increase superoxide dismutase (SOD) activity compared to the stone group. In addition, the degree of apoptosis in the urinary stone group was signiﬁcantly increased compared to that in the group treated with green tea. 21.1.3.1.2 Body Size As body weight increases, the excretion of calcium, oxalate and uric acid also increases, both in normal subjects and in stone formers [51]. In contrast, even a

21.1 Urinary Stones

small drop in body weight in subjects with calcium stone disease is associated with a considerable reduction of lithogenous salts in their urine, and vice versa. In conclusion, obesity and weight gain increase the risk of kidney stone formation, with the magnitude of the increased risk perhaps being greater in women than in men [52]. 21.1.3.1.3 Environment There is a growing body of evidence from NASA and the Russian space program showing that humans exposed to the microgravity environment of space have a greater risk for developing renal stones [53]. Increased bone resorption and the attendant hypercalciuria and hyperphosphaturia contribute signiﬁcantly to raising the urinary state of saturation with respect to the calcium salts, namely CaOx and calcium phosphate. However, other environmental and dietary factors may also adversely aﬀect urine composition and increase stone formation risk. Reductions in urinary volume have been consistently observed in short-duration ﬂights [54–56]. 21.1.3.2 Genetic Factors As a family history is a known risk factor of urolithiasis [57, 58], genetic factors have been postulated to play an important role in the risk of urolithiasis. Stone formation seems to be inherited with a polygenic mechanism (Table 21.3), and in the majority of children with urolithiasis a metabolic cause of stone formation can be identiﬁed. In fact, although several metabolic disorders (e.g., cystinuria or primary hyperoxaluria) have been elucidated in children, metabolic causes for urolithiasis are rarely found in adults. Approximately 50% of patients labeled as having idiopathic hypercalciuria have a positive family history of kidney stones [59]. In the past, familial idiopathic hypercalciuria has been described as an autosomal dominant trait, but this is clearly an oversimpliﬁcation of the genetics of familial hypercalciuria. In the case of uric acid stones, distinct geographical and ethnic variations in the incidence of this type of stone development have been identiﬁed that are compatible with environmental or genetic susceptibility in some populations. For example, although the incidence of uric acid stones in the United States is reported as between 5 and 9.7% of all kidney stones analyzed, it is 25% in Germany, and as high as 39.5% in Israel [60]. Recently, Ombra et al. identiﬁed a possible genetic basis for this increased risk by studying a homogeneous population in Sardinia with a high incidence of uric acid stones [61]. The majority of these patients had a low urinary pH with high titratable acidity. In addition, one-third of the individuals were hyperuricosuric, with a daily uric acid excretion in excess of 700 mg. The major abnormality in this population was, therefore, a high urine acidity. The investigators used multi-step linkage and allele-sharing analysis to identify a locus on chromosome 10q21-22 which was associated with increased susceptibility to uric acid stone formation. Subsequent investigations further identiﬁed a candidate gene, named ZNF365 [62].

341

342

21 Stone Formation

Table 21.3 Gene defects connected to stone formation.

Increased excretion

Disease/Syndrome

Calcium

Familial idiopathic hypercalciuria

Cystine

Defect

Gene locus

Familial hypocalcemia with hypercalciuria

Calcium-sensing receptor

3q13.3-21

Dent’s disease

Mutation of CLCN5-gene

Xp 11.22

X-linked recessive nephrolithiasis type I

Mutation of CLCN5-gene

Xp 11.22

X-linked-recessive hypophosphatemic rickets (XLRH)

Mutation of CLCN5-gene

Xp 11.22

Distal renal tubular acidosis

Mutation of RTA-1 gene?

Bartter’s syndrome

Na-K-2Cl co-transporter

7q11.23

William’s syndrome

Deletion of the elastin gene (ELN) calcitonin receptor gene (?)

7q21.3

Wilson’s disease

Copper-transporting protein

13p14.1-21.1

Cystinuria type I

rBAT/D2H (SLC3A1)

2p21

Cystinuria type III (type II) Oxalate

19q13.1

Primary hyperoxaluria type 1

Alanine:glyoxylate-aminotransferase

Primary hyperoxaluria type 2

Glyoxylate-reductase/D-Glycerate dehydrogenase

Lesch–Nyhan syndrome

Hypoxanthine-guanine phosphoribosyltransferase

Xq26-27.2

Phosphoribosyl-pyrophosphatesynthetase superactivity

Phosphoribosyl-pyrophosphatesynthetase

Xq22-24

Glycogen-storage disease type 1

Glucose-6-phosphatase

17q21

2,8 Dihydroxyadenine

Dihydroxyadeninuria

Adenine-phosphoribosyltransferase

16q22.2-22.3

Xanthine

Xanthinuria

Xanthine-oxidase

2p23-22

Uric acid

2q27.3

21.3 Biliary and Gallbladder Stones

21.2 Other Urological Stones: Testicular Microlithiasis

Testicular microlithiasis is an uncommon condition characterized by calciﬁcations within the seminiferous tubules. Although the true prevalence in a normal population has not been deﬁned, the condition was ﬁrst described radiologically by Priebe and Garret in 1970 [63]. Morphologically, the microliths consist of degenerated intratubular cells which form a calciﬁed core. In a histopathological classiﬁcation, two diﬀerent forms of testicular microlithiasis have been described by Renshaw [64]. The most frequent type is represented by laminated calciﬁcations that occurred in association with testicular malignancies, in cryptorchid testes, and also in normal testes. The second type consists of hematoxylin bodies that are exceptionally encountered in connection with testicular malignancies [65]. A prognostic value of testicular microlithiasis concerning the development of testicular cancer cannot be derived from the published data [66], and the clinical significance of testicular microlithiasis remains unclear [67, 68].

21.3 Biliary and Gallbladder Stones

Biliary stone disease is a common disorder, usually associated with stones in the gallbladder, and can cause signiﬁcant complications. The formation of stone usually occurs in the presence of the following factors: abnormalities of the bile constituents; bile stasis; and the presence of a nidus for stone formation [69]. The cause of gallbladder stone is very complicated, and a variety of factors are involved [70]. Three conditions must be met to permit the formation of cholesterol gallstones: (i) the bile must be supersaturated with cholesterol; (ii) nucleation must be kinetically favorable; and (iii) the cholesterol crystals must remain in the gallbladder long enough to agglomerate into stones. Most patients have cholesterol gallstones, the initial and essential ﬁrst stage of which is considered to be the nucleation of cholesterol crystals [71]. Cholesterol is insoluble in water, but can dissolve in bile, where the concentration may reach 10 to 20 mmol L1 , which is about 106-fold greater than the solubility in water [72]. From a crystallogenic viewpoint, the precondition for a solute to be separated out from a solution and to form crystals is that it must be in a state of supersaturation, which is an unstable state in thermodynamic terms [73]. However, it is only when in this state that the solute can be crystallized [74]. This also causes an imbalance between nucleation-leading factors and antinucleation factors, and abnormal function of the gallbladder, and so on [75]. Excessive cholesterol may be stored in vesicles (i.e., spherical bilayers of cholesterol and phospholipids, without bile salts), provided that suﬃcient phospholipid is available. When relatively low amounts of phospholipids are present, then cholesterol crystal formation occurs in supersaturated bile, and this is the start of gallstone formation. Primary bile salts (i.e., cholate and chenodeoxycholate) are

343

344

21 Stone Formation

synthesized from cholesterol in the liver, while secondary bile salts (mainly deoxycholate) are formed from primary bile salts in the intestine by bacterial transformation [76]. Cholesterol crystallization is promoted by hydrophobic bile salts (chenodeoxycholate, deoxycholate), and by phospholipids with unsaturated acyl chains. Iron deﬁciency has been shown to alter the activity of several hepatic enzymes, leading to increased bile cholesterol saturation in the gallbladder and the promotion of cholesterol crystal formation [77].

21.4 Miscellaneous Stones 21.4.1 Sialolithiasis

Sialolithiasis, or the formation of sialoliths or salivary stones, typically occurs in the ducts of the submandibular and parotid glands of middle-aged adults. Sialolithiasis is the most common cause of salivary gland obstruction [78], and it can be complete or partial and may show recurrence. Sialolithiasis is the most common disease of salivary glands, aﬀecting 12 in every 1000 of the adult population [79], and males being aﬀected twice as much as females [80]. Although children are rarely aﬀected, a review of the literature has revealed 100 cases of submandibular calculi in children aged between 3 weeks and 15 years [81]. The retained saliva applies retrograde pressure on the salivary gland, the chyma, and the ductal system. Salivary calculi develop due to pathologic formations of calcareous deposits in the salivary ducts or glands, whereby minerals form around an organic matrix. Many theories have been proposed to explain salivary calculi formation, such as calciﬁcation around foreign bodies, desquamated epithelial cells, and microorganisms in the duct. Although the exact cause of these stones is unknown, some stones may be related to dehydration, which thickens the saliva; to decreased food intake, which lowers the demand for saliva; or to medications that decrease saliva production, including certain antihistamines, anti-hypertensive drugs and psychiatric medications. 21.4.2 Dental Stones

Dental calculus is calciﬁed dental plaque that is composed primarily of mineral as well as inorganic and organic components. Supragingival and subgingival calculus contain 37% and 58% mineral content by volume, respectively [82]. Dental calculus contains both total phospholipids and acidic phospholipids in much higher concentrations than parotid saliva [83]. In addition, the concentration of

21.4 Miscellaneous Stones

phospholipids in the saliva of heavy calculus formers is signiﬁcantly higher than that of light calculus formers. These ﬁndings suggest that phospholipids play an important role in calculus formation [84]. In theory, the supersaturation of saliva, especially plaque ﬂuid, with respect to calcium phosphate salts is the driving force for dental plaque mineralization. Although calculus can be induced in germ-free animals, human calculus development invariably involves plaque bacterial calciﬁcation. 21.4.3 Pancreatic Stones

Stone formation in the pancreatic duct system is common in chronic pancreatitis [85]. Plugs formed by precipitation of the protein within the interlobular and intralobular ducts are one of the earliest ﬁndings in chronic pancreatitis, and the protein plugs subsequently perpetuate inﬂammation of the gland through repeated obstruction of the pancreatic duct system. The role(s) of the diﬀerent protein components of pancreatic secretions is (are) controversial in pancreatic stone formation. Pancreatic stone protein is a 16-kDa acidic protein which is believed to inhibit calcium carbonate precipitation in pancreatic juice. Additional candidate proteins and mechanisms will be required in stone formation, however. Lactoferrin, for example, may play a role in the formation of the protein plugs frequently seen in chronic pancreatitis because of its ability to cause the aggregation of a large acidophilic protein, such as albumin. 21.4.4 Broncholithiasis and Pulmonary Alveolar Microlithiasis

Broncholithiasis is deﬁned as a condition in which calciﬁed or ossiﬁed material is present within the bronchial lumen [86]. A broncholith is usually formed by the erosion by, and extrusion of, a calciﬁed adjacent lymph node into the bronchial lumen. Other causes of broncholithiasis include: (i) the aspiration of bone tissue or in-situ calciﬁcation of aspirated foreign material; (ii) the erosion by, and extrusion of, calciﬁed or ossiﬁed bronchial cartilage plates; and (iii) migration to a bronchus of calciﬁed material from a distant site, such as a pleural plaque or the kidney (via a nephrobronchial ﬁstula) [87]. Pulmonary alveolar microlithiasis is a rare idiopathic disease characterized by the diﬀuse presence in the alveoli of innumerable minute calculi called microliths [88]. This disease can also present in a single area of the lung as a secondary localized disease in the presence of adenocarcinoma, tubercular remnants, or pleural mesothelioma [89–91]. The etiology and pathogenesis of the condition are still unknown, however [92, 93].

345

346

21 Stone Formation

References 1 S. Ljunghall, Scand. J. Urol. Nephrol. 2 3

4 5 6

7 8 9 10 11

12 13 14 15 16 17

18 19 20

21

22

23

1977, 41, 1. A. Serio, A. Fraioli, Nephron 1999, 81 (Suppl. 1), 26. R.J. McLean, J. Downey, L. Clapham, J.C. Nickel, Urol. Res. 1990, 18, 39. A. Randall, Ann. Surg. 1937, 105, 1009. J.S. Elliot, J. Urol. 1973, 109, 82. C.Y. Pak, O. Waters, L. Arnold, K. Holt, C. Cox, D. Barilla, J. Clin. Invest. 1977, 59, 426. D.J. Kok, Scanning Microsc. 1996, 10, 471. R.S. Malek, W.H. Boyce, J. Urol. 1977, 117, 336. R.S. Howard, M.P. Walzak, J. Urol. 1967, 98, 639. W.G. Robertson, M. Peacock, Contrib. Nephrol. 1984, 37, 1. W.G. Robertson, M. Peacock, P.J. Heyburn, D.H. Marshall, P.B. Clark, Br. J. Urol. 1978, 50, 449. W.G. Robertson, M. Peacock, Clin. Sci. 1972, 43, 499. B. Finlayson, Kidney Int. 1978, 13, 344. H.G. Tiselius, A.M. Fornander, M.A. Nilsson, Urol. Res. 1993, 21, 363. R.L. Ryall, World J. Urol. 1997, 15, 155. J.S. Elliot, M.E. Ribeiro, Invest. Urol. 1972, 10, 102. P.O. Schwille, D. Scholz, K. Schwille, R. Leutschaft, I. Goldberg, A. Sigel, Nephron 1982, 31, 194. S.R. Khan, D.J. Kok, Front. Biosci. 2004, 9, 1450. B. Hess, Miner. Electrolyte Metab. 1994, 20, 393. E.M. Worcester, Y. Nakagawa, C.L. Wabner, S. Kumar, F.L. Coe, Am. J. Physiol. 1988, 255, F1197. T. Yamate, K. Kohri, T. Umekawa, M. Iguchi, T. Kurita, J. Urol. 1998, 160, 1506. T. Yamate, T. Umekawa, M. Iguchi, T. Kurita, K. Kohri, Hinyokika Kiyo 1997, 43, 623. T. Yamate, K. Kohri, T. Umekawa, N. Amasaki, N. Amasaki, Y. Isikawa, M.

24

25 26

27

28 29 30

31

32 33

34

35

36

37

38

39 40

41

Iguchi, T. Kurita, Eur. Urol. 1996, 30, 388. G. Gambaro, A. D’Angelo, A. Fabris, E. Tosetto, F. Anglani, A. Lupo, J. Nephrol. 2004, 17, 774. J.C. Lieske, S. Deganello, J. Am. Soc. Nephrol. 1999, 10 (Suppl. 14), S422. Y. Kohjimoto, S. Ebisuno, M. Tamura, T. Ohkawa, Scanning Microsc. 1996, 10, 459. S. Ebisuno, M. Umehara, Y. Kohjimoto, T. Ohkawa, Br. J. Urol. Int. 1999, 84, 118. P.K. Grover, R.L. Ryall, V.R. Marshall, Kidney Int. 1992, 41, 149. E.L. Prien, E.L. Prien, Jr., Am. J. Med. 1968, 45, 654. B. Dussol, S. Geider, A. Lilova, F. Leonetti, P. Dupuy, M. Daudon, Y. Berland, J.C. Dagorn, J.M. Verdier, Urol. Res. 1995, 23, 45. P.J. du Toit, C.H. van Aswegen, P.L. Steyn, A. Pols, D.J. du Plessis, Urol Res. 1992, 20, 393. J.H. Wiessner, L.Y. Hung, N.S. Mandel, Kidney Int. 2003, 63, 1313. J.A. Jonassen, L.C. Cao, T. Honeyman, C.R. Scheid, Crit. Rev. Eukaryot. Gene Expr. 2003, 13, 55. L.C. Cao, J. Jonassen, T.W. Honeyman, C. Scheid, Am. J. Nephrol. 2001, 21, 69. S. Lahme, G. Feil, W.L. Strohmaier, K.H. Bichler, A. Stenzl, Urol. Int. 2004, 72, 244. A.L. Boskey, B.D. Boyan-Salyers, L.S. Burstein, I.D. Mandel, Arch. Oral Biol. 1981, 26, 779. S.R. Khan, P.A. Glenton, R. Backov, D.R. Talham, Kidney Int. 2002, 62, 2062. S.R. Khan, F. Atmani, P. Glenton, Z. Hou, D.R. Talham, M. Khurshid, Calcif. Tissue Int. 1996, 59, 357. A. Muthukumar, R. Selvam, Ren. Fail. 1997, 19, 401. G. Gambaro, G. Vezzoli, G. Casari, L. Rampoldi, A. D’Angelo, L. Borghi, Am. J. Kidney Dis. 2004, 44, 963. R.J. McLean, J.C. Nickel, K.J. Cheng, J.W. Costerton, Crit. Rev. Microbiol. 1988, 16, 37.

References 42 H. Hedelin, J.E. Brorson, L. Grenabo,

43

44

45

46 47

48

49 50 51

52 53 54 55 56 57 58

59 60 61

62 63 64

S. Pettersson, Br. J. Urol. 1984, 56, 244. P. Goodyer, I. Saadi, P. Ong, G. Elkas, R. Rozen, Kidney Int. 1998, 54, 56. W.G. Robertson, M. Peacock, P.J. Heyburn, F.A. Hanes, Scand. J. Urol. Nephrol. Suppl. 1980, 53, 15. N.A. Breslau, L. Brinkley, K.D. Hill, C.Y. Pak, J. Clin. Endocrinol. Metab. 1988, 66, 140. W.G. Robertson, M. Peacock, D.H. Marshall, Eur. Urol. 1982, 8, 334. G.C. Curhan, W.C. Willett, F.E. Speizer, D. Spiegelman, M.J. Stampfer, Ann. Intern. Med. 1997, 126, 497. L. Borghi, T. Meschi, F. Amato, A. Novarini, A. Romanelli, F. Cigala, J. Urol. 1993, 150, 1757. B. Hess, Scanning Microsc. 1996, 10, 547. Y. Itoh, T. Yasui, A. Okada, et al., J. Urol. 2005, 173, 271. G.C. Curhan, W.C. Willett, E.B. Rimm, F.E. Speizer, M.J. Stampfer, J. Am. Soc. Nephrol. 1998, 9, 1645. E.N. Taylor, M.J. Stampfer, G.C. Curhan, JAMA 2005, 293, 455. J.E. Zerwekh, Nutrition 2002, 18, 857. P.A. Whitson, R.A. Pietrzyk, C.F. Sams, J. Gravit. Physiol. 1999, 6, P87. P.A. Whitson, R.A. Pietrzyk, C.Y. Pak, J. Urol. 1997, 158, 2305. P.A. Whitson, R.A. Pietrzyk, C.Y. Pak, N.M. Cintron, J. Urol. 1993, 150, 803. H.O. Goodman, R.P. Holmes, D.G. Assimos, J. Urol. 1995, 153, 301. J.M. Kinder, C.D. Clark, B.J. Coe, J.R. Asplin, J.H. Parks, F.L. Coe, J. Urol. 2002, 167, 1965. M.J. Favus, J. Urol. 1989, 141, 719. N.S. Mandel, G.S. Mandel, J. Urol. 1989, 142, 1516. M.N. Ombra, S. Casula, G. Biino, G. Maestrale, F. Cardia, P. Melis, M. Pirastu, Urology 2003, 62, 416. F. Gianfrancesco, T. Esposito, Minerva Med. 2005, 96, 409. C.J. Priebe, R. Garret, Pediatrics 1970, 46, 785. A.A. Renshaw, J. Urol. 1998, 160, 1625.

65 A. Berger, K. Brabrand, Acta Radiol.

1998, 39, 583. 66 N. Lawrentschuk, S.J.S. Brough, R.J.

67

68 69

70 71

72

73

74

75

76

77 78 79

80

81

82

De Ryke, Aust. N. Z. J. Surg. 2003, 73, 364. M. Holm, C.E. Hoei-Hansen, E. Rajpert-De Meyts, N.E. Skakkebaek, J. Urol. 2003, 170, 1163. J.P. Ganem, Curr. Opin. Urol. 2000, 10, 99. H.M. Yang, J. Wu, J.Y. Li, L. Gu, M.F. Zhou, World J. Gastroenterol. 2003, 9, 1791. M. Rubin, R. Pakula, F.M. Konikoﬀ, Histol. Histopathol. 2000, 15, 761. P. Portincasa, A. Moschetta, G. Calamita, A. Margari, G. Palasciano, Curr. Drug Targets Immune Endocr. Metabol. Disord. 2003, 3, 67. N.G. Venneman, M. van Kammen, W. Renooij, G.P. VanbergeHenegouwen, K.J. van Erpecum, Biochim. Biophys. Acta 2005, 1686 3, 209. J.X. Zhang, E. Lundin, C.O. Reuterving, G. Hallmans, R. Stenling, E. Westerlund, P. Aman, Br. J. Nutr. 1994, 71, 861. P. Portincasa, A. Moschetta, G. Palasciano, Ann. Hepatol. 2002, 1, 121. J.J. Roslyn, R.L. Conter, L. DenBesten, Dig. Dis. Sci. 1987, 32, 609. J. Shoda, N. Tanaka, B.F. He, Y. Matsuzaki, T. Osuga, H. Miyazaki, Dig. Dis. Sci. 1993, 38, 2130. M. Kumar, B.B. Goyal, M. Mahajan, S. Singh, Indian J. Surg. 2006, 68, 80. J. Lustmann, A. Shteyer, J. Dent. Res. 1981, 60, 1386. A.K. Leung, M.C. Choi, G.A. Wagner, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 1999, 87, 331. R.A. Cawson, E.W. Odell, Essentials of Oral Pathology and Oral Medicine. Churchill Livingstone, Edinburgh, 1998. M.G. Steiner, A.R. Gould, G.M. Kushner, R. Weber, A. Pesto, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 1997, 83, 188. J. Friskopp, G. Isacsson, Scand. J. Dent. Res. 1984, 92, 25.

347

348

21 Stone Formation 83 I.D. Mandel, A. Eisenstein, Arch. Oral 84 85

86 87 88

Biol. 1969, 14, 231. A. Slomiany, B.L. Slomiany, I.D. Mandel, Arch. Oral Biol. 1981, 26, 151. M. Kitagawa, T. Hayakawa, T. Kondo, T. Shibata, Y. Sakai, H. Sobajima, H. Ishiguro, M. Tanikawa, Y. Nakae, Y. Nimura, et al., Gastroenterology 1992, 102, 2151. V.A. Vix, Radiology 1978, 128, 295. L.A. Weed, H.A. Andersen, Dis. Chest 1960, 37, 270. A. Aliperta, G. Mauro, G. Saviano, U. Bellissimo, A. Perna, S. Martuﬁ, Arch. Monaldi 1977, 31, 87.

89 M.C. Sosman, G.D. Dodd, W.D.

90

91 92

93

Jones, G.U. Pillmore, Am. J. Roentgenol. Radium Ther. Nucl. Med. 1957, 77, 947. G. Castellana, M. Gentile, R. Castellana, P. Fiorente, V. Lamorgese, Am. J. Med. Genet. 2002, 111, 220. B. Mariani, A. Bassi, Minerva Med. 1967, 58, 3761. G. Kanra, E. Tanyol, A. Gocmen, M. Yurdakok, G. Secmeer, O. Oran, O. Soylemezoglu, S. Cakir, C. Dizdarer, Turk. J. Pediatr. 1988, 30, 61. N. Chinachoti, P. Tangchai, Dis. Chest 1957, 32, 687.

349

22 Ectopic Mineralization: New Concepts in Etiology and Regulation Cecilia M. Giachelli

Abstract

Ectopic mineralization is a common response to tissue injury and systemic mineral imbalance, and can lead to severe clinical consequences when present in joints, tendons, heart valves, and blood vessels. It has been hypothesized that the mineralization of matrices in any tissue is normally controlled by the ratio of procalciﬁc and anti-calciﬁc regulatory molecules, such that abnormal deposition of mineral is avoided. Alterations in this ratio induced by injury, disease, or genetic deﬁciency are postulated to induce ectopic mineral deposition. Over the past few years we have developed in-vitro and in-vivo models of ectopic calciﬁcation in vascular tissues to investigate potential inducers and inhibitors of this process. In this chapter we will highlight recent ﬁndings in this area, focusing on the roles of phosphate and osteopontin in vascular calciﬁcation. Key words: ectopic mineralization, vascular calciﬁcation, osteopontin, phosphate, hyperphosphatemia, sodium-dependent phosphate co-transporter, Pit-1.

22.1 Introduction

Under normal conditions, most tissues other than bone remain non-calciﬁed despite circulating calcium and phosphate levels that are essentially at the solubility product for bioapatite formation. However, under certain pathological conditions, such as atherosclerosis, diabetes and uremia, blood vessels and other tissues are prone to calciﬁcation. Ectopic calciﬁcation of the blood vessels leads to stiﬀening and altered hemodynamics, may contribute to atherosclerotic plaque rupture, and has been correlated with cardiovascular mortality in both general and high-risk populations (for a review, see [1]). Ectopic calciﬁcation of cartilage and joints can lead to arthritis and ankylosis [2].

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

350

22 Ectopic Mineralization: New Concepts in Etiology and Regulation

Growing evidence suggests that vascular calciﬁcation, like bone formation, is a highly regulated process, involving both inductive and inhibitory processes [3]. For example, apatite, bone-related non-collagenous proteins, and matrix vesicles have all been observed in calciﬁed vascular lesions. Furthermore, outright cartilage and bone formation have also been documented. Finally, cells derived from the arterial media can undergo matrix mineralization under the appropriate conditions in vitro [4, 5]. These studies suggest that cell-mediated processes maintain a balance between pro-calciﬁc and anti-calciﬁc processes in the artery such that ectopic calciﬁcation is normally avoided. Under pathological conditions, this balance is upset and leads to ectopic mineralization.

22.2 Regulators of Ectopic Mineralization

Over the past ten years, our understanding of molecules and processes that regulate ectopic calciﬁcation has grown exponentially. Much of our understanding comes from the identiﬁcation of genes through linkage or targeted deletion studies that cause human and/or mouse ectopic calciﬁcation disorders, respectively. In addition, a number of in-vitro and in-vivo model systems have been developed that mimic important aspects of ectopic calciﬁcation. A number of the molecules that have been identiﬁed as playing likely roles in regulating ectopic mineralization of the vasculature and other tissues are listed in Table 22.1. These factors fall broadly into four categories: (i) circulating factors; (ii) ion transporters and homeostatic enzymes; (iii) extracellular matrix (ECM) molecules; and (iv) cell signaling molecules. 22.2.1 Circulating Factors that Regulate Ectopic Mineralization

Circulating factors thus far identiﬁed include minerals involved in apatite crystal formation such as calcium and phosphate. Hyperphosphatemia and increased calcium burden have been implicated in ectopic calciﬁcation, especially in endstage renal disease patients [6, 7]. In addition to increasing the Ca P product, these ions exert direct pro-mineralizing eﬀects on ectopically mineralizing cells, as described below (see Section 22.2.2.1). Circulating hormones that regulate serum calcium and phosphate have logically emerged as key regulators of ectopic mineralization. Vitamin D overload is routinely used to promote ectopic calciﬁcation in rodents [8], and may act predominantly by causing hypercalcemia. Likewise, FGF23 and PTH are major regulators of phosphate levels in the blood. Mice with a deﬁciency in FGF23 have severe hyperphosphatemia in addition to elevated circulating levels of 1,25dihydroxyvitamin D, and show soft tissue mineralization, growth retardation, and abnormal bones [9]. The role of PTH in vascular calciﬁcation is not conclusive, as

22.2 Regulators of Ectopic Mineralization Table 22.1 Ectopic mineralization regulatory molecules.

Ectopic mineralization regulatory molecule

Eﬀect*

Circulating factor Phosphate Calcium Pyrophosphate Fetuin Osteoprotegerin FGF23 Vitamin D overload Parathyroid hormone Magnesium High-density lipoprotein Low-density lipoprotein

P P I I I I P I I I P

Ion transporters/Homeostasis enzymes (substrate/product) Ank (pyrophosphate) Nucleotide pyrophosphatase PC-1/NPP-1 (organic phosphates/pyrophosphate) Pit-1 (phosphate) NaPi2 (phosphate) Tissue non-speciﬁc alkaline phosphatase (pyrophosphate/phosphate) Carbonic anhydrase II (CO2/protons) b-glucosidase (b-d-galactosides/b-d-glucose) Abcc6 (substrate unknown)

I I P I P I I I

Matrix molecules Matrix Gla Protein Osteopontin Fibrillin Elastin Collagen Decorin

I I I P P P

Signaling molecules/Pathways BMP2/4 Msx2/Wnt/Osx Smad6 BMP7 Runx2 (Cbfa1) IGF-1 ERK/PI3K P38MapK/PPARgamma Gas6/Axl

P P I I P P P I I

* P ¼ promotional; I ¼ inhibitory.

351

352

22 Ectopic Mineralization: New Concepts in Etiology and Regulation

some studies have linked elevated PTH [10, 11] or low PTH [12] with higher rates of vascular calciﬁcation, while others have not [7, 13–15]. On the other hand, the administration of teriperatide (human PTH 1-24) inhibited arterial calciﬁcation in diabetic LDLR-null mice [16]. Likewise, PTH and PTHrp inhibited the calciﬁcation of vascular smooth muscle cells in vitro [17], although these ﬁndings have not been conﬁrmed. Thus, it is possible that an optimal level of PTH is required for the maintenance of vessel health, and that too much or too little PTH may exacerbate vascular calciﬁcation. Finally, circulating inhibitors of apatite crystal nucleation and growth have also been identiﬁed. Pyrophosphate has gained acceptance as a major inhibitor of this type, and the human genetic disease, generalized infantile arterial calciﬁcation, is characterized by a deﬁciency in an enzyme that is responsible for pyrophosphate production (see Section 22.2.2). Likewise, mice deﬁcient in the liver-derived serum protein, fetuin, have increased spontaneous and vitamin D-induced soft tissue mineralizations that lead to myocardial stiﬀness, cardiac remodeling, and diastolic dysfunction [18, 19]. While the ability to inhibit apatite crystal growth no doubt is a major mechanism by which these molecules function to block ectopic calciﬁcation, other activities that may also be involved have been described for both [20, 21]. 22.2.2 Ion Transporters and Homeostatic Enzymes that Regulate Ectopic Mineralization

Again, human and mouse genetic mutations have identiﬁed a growing class of molecules that have common functions as ion transporters and/or enzymes involved in small charged molecule production as being very important in ectopic mineralization. In some of these cases, the mechanisms by which these factors regulate ectopic calciﬁcation are not yet apparent. For example, knockout of the b-glucosidase gene leads to an aging phenotype in mice that is associated with ectopic calciﬁcation, but the underlying mechanism for this eﬀect is unknown [22]. Likewise, mutation in the orphan ABC transporter gene, Abcc6, leads to pseudoxanthoma elasticum, a calcifying connective tissue disorder in people [23]. However, the substrate for Abcc6 has not yet been identiﬁed, nor do we understand if this is a direct or indirect eﬀect of Abcc6 deﬁciency. Finally, a deﬁciency in carbonic anhydrase II (CAR2) in mice leads to age-dependent medial calciﬁcation in small arteries in a number of organs [24]. Whilst the mechanism by which CAR2 deﬁciency predisposes to ectopic mineralization is not completely clear, it may be related to the ability of cells to generate protons and resorb mineral, and is discussed in greater detail below (see Section 22.2.3.1). In contrast, as phosphate and pyrophosphate have been strongly implicated in vascular calciﬁcation, it is not surprising that transporters and enzymes which regulate the synthesis, degradation or cellular traﬃcking of these ions have also been found to regulate ectopic mineralization. Indeed, major advances in our understanding of ectopic mineralization have come from recent investigations in this area.

22.2 Regulators of Ectopic Mineralization

In addition to dietary and hormonal inﬂuences, extracellular phosphate levels are regulated locally by the action of phosphatases, such as tissue non-speciﬁc alkaline phosphatase (TNAP), on inorganic and organic phosphate-containing molecules. TNAP is absolutely required for bone formation, and accumulating evidence indicates that it may play a similar role in mediating ectopic calciﬁcation [25]. Alkaline phosphatase is strongly induced in smooth muscle cells by calciﬁcation inducers, such as elevated phosphate, TNFa and oxidized lipids [26–28]. One key substrate for TNAP is pyrophosphate, a potent inhibitor of apatite crystal formation. In addition to generating phosphate, TNAP activity decreases pyrophosphate levels, thereby potentially promoting ectopic calciﬁcation by two mechanisms: increased phosphate levels, and decreased pyrophosphate levels. That a loss of pyrophosphate predisposes to vascular calciﬁcation is clearly demonstrated by the human genetic disease infantile arterial calciﬁcation, in which deﬁciency of the ectonucleotide pyrophosphatase phosphodiesterase (eNPP-1/PC-1), the enzyme that generates pyrophosphate from organic phosphate such as nucleoside triphosphates, leads to pyrophosphate deﬁciency and lethal arterial calciﬁcation [29]. Likewise, mice with a targeted deletion of eNPP1 show a vascular calciﬁcation phenotype, in addition to cartilage defects [30]. A similar phenotype is observed in mice deﬁcient in a major cellular pyrophosphate transporter, ANK [31]. That these molecules are part of an elegant, coordinated regulation system for biomineralization is exempliﬁed by the ﬁnding that mineralization disorders in eNPP-1 or ank-null mice were partially reversed by crossing to TNAP-null mice [32]. Finally, phosphate transporters have been implicated in the regulation of ectopic mineralization. In particular, sodium-dependent phosphate co-transporters, a large family of transporters that utilize the sodium gradient to drive phosphate uptake into cells, have been studied. The type II family member, NaPi-IIa, is expressed in the kidney and functions to facilitate phosphate resorption in kidney tubules in response to PTH, FGF23, and other hormonal stimuli. A deﬁciency in NaPi-IIa leads to hyperphosphaturia and hypercalciuria, and predisposes mice to kidney stones [33]. Our group has investigated the type III family members, Pit-1 and Pit-2, which are expressed in cardiovascular tissues and bone. These studies, the details of which are described in Section 22.2.2.1, suggest that Pit-1 is a major regulator of smooth muscle cell mineralization in vitro. 22.2.2.1 Role of Sodium-Dependent Phosphate Co-Transporters in Ectopic Mineralization As mentioned previously, elevated circulating phosphate and calcium levels may stimulate mineral precipitation not only by raising the Ca P product, but also by direct actions on calcifying cells. Culture of smooth muscle cells in elevated phosphate levels similar to those observed in hyperphosphatemic patients (>2 mM) leads to matrix mineralization. In addition, the cells undergo a dramatic phenotypic transition that is characterized by a loss of smooth muscle cell lineage marker expression and up-regulation of genes related to osteoblast/ chondrocyte diﬀerentiation, including osteopontin, osteocalcin, and Runx2 [34].

353

354

22 Ectopic Mineralization: New Concepts in Etiology and Regulation

We have determined that the eﬀects of inorganic phosphate on smooth muscle cell diﬀerentiation and mineralization are in part regulated by a sodiumdependent phosphate co-transporter [5]. Phosphonoformic acid (an inhibitor of such co-transporters) blocked smooth muscle cell mineralization and phenotype transition in response to elevated phosphate treatment [5]. Subsequent studies showed two members of the type III sodium-dependent phosphate co-transporter gene families, Pit-1 and Pit-2, were expressed in human smooth muscle cells. Furthermore, by using RNA interference, the level of Pit-1 on smooth muscle cells was found to be a major regulator of the susceptibility of these cells to phosphate-induced mineralization [35]. Furthermore, elevated calcium levels (to millimolar level) were shown to increase levels of Pit-1 mRNA in human smooth muscle cells, and to induce both phenotype change and mineralization [36]. Thus, elevated inorganic phosphate and calcium levels not only contribute to elevating the Ca P product, but also signal smooth muscle cells to undergo phenotypic transition that may promote matrix mineralization. Further studies on the signaling mechanisms involved in this inorganic phosphate-induced signaling pathway in smooth muscle cells are currently under way. 22.2.3 Extracellular Matrix Molecules that Regulate Ectopic Mineralization

A number of ECM molecules appear to play critical roles in either facilitating or inhibiting ectopic mineralization. Collagens and elastin, which are major structural proteins present in bones and blood vessels, respectively, serve as important scaﬀolding structures for apatite crystal formation in physiological bone mineralization as well as in ectopic mineralization. Apatite crystals appear to grow preferentially along and within these repeating ﬁbrillary structures in both physiological and ectopic mineralization. In addition, decorin, a collagen-binding protein, has recently been implicated as a promoter of calciﬁcation in cultured smooth muscle cells [37]. Proteins that share a common function of binding to elastin, including matrix gla protein (MGP) and ﬁbrillin, have also been shown to be regulators of ectopic mineralization. MGP-deﬁcient mice display severe cartilage and vascular calciﬁcation by the age of 2 weeks, and this progresses to osteopenia and death due to vascular rupture by 6 to 8 weeks of age [38]. Overexpression of MGP selectively in the blood vessels, but not in the liver, can rescue the vascular calciﬁcation phenotype in MGP-null mice, indicating that MGP is a local inhibitor of ectopic mineralization [39]. Interestingly, the vascular calciﬁcation observed in the MGP-null mice initiates along the elastic lamellae, suggesting a protective function of MGP on elastin integrity. In addition, MGP binds and inactivates BMP2 [40], and is a component of fetuin-containing serum complexes that may aid in clearing mineral salts from the circulation [41], indicating that it may function in multiple ways to block ectopic mineralization. Likewise, a deﬁciency in ﬁbrillin – another elastin-binding protein – leads to aortic aneurysm associated with elastin calciﬁcation in mice [42]. Finally, our

22.2 Regulators of Ectopic Mineralization

group has identiﬁed osteopontin, a secreted phosphorylated matricellular protein, as a major, inducible regulator of ectopic mineralization (see Section 22.2.3.1). 22.2.3.1 Role of Osteopontin in Ectopic Mineralization Osteopontin (OPN) is highly phosphorylated and glycosylated secreted protein that originally was discovered in bone, but was more recently identiﬁed in calciﬁed vascular lesions [43]. The role of OPN in vascular calciﬁcation has been investigated using both in-vitro and in-vivo approaches. Smooth muscle cells can be induced to mineralize in response to elevated phosphate in culture [4]. The addition of exogenous OPN potently inhibits calciﬁcation of SMC in a dosedependent fashion. OPN was intimately associated with growing apatite crystals, suggesting a physical inhibition of crystal growth as one mechanism for inhibition. The phosphorylation of OPN was required for mineral inhibitory aﬀects [44]. Finally, preliminary experiments have shown that smooth muscle cells derived from OPN-null arteries are more susceptible to mineralization compared to wild-type cells [45]. In vivo, two separate approaches were used to determine the role of OPN in ectopic mineralization. In the ﬁrst approach, OPN-null mice were crossed with MGP-mutant mice. MGP-null mice undergo spontaneous vascular calciﬁcation by the age of 2 weeks, and the mineralization of vessels was associated with highly induced OPN levels [34]. In MGP/OPN-double-null mice, the mineralization of vessels occurred earlier and was more severe, leading to earlier death compared to mice that were deﬁcient in MGP alone [46]. Finally, glutaraldehyde-ﬁxed porcine aortic valves, when implanted into OPN-null mice, calciﬁed to a much greater extent than those implanted into wild-type mice [47], and OPN not only inhibited but also caused the regression of ectopic mineral. Of major importance, the regression of calciﬁcation induced by OPN was correlated with CAR2 expression in macrophages surroundings the implants, and led to acidiﬁcation of the implants. The calciﬁcation of implanted valves could be inhibited by coating with puriﬁed, phosphorylated OPN, and this was correlated with enhanced CAR2 levels [48]. Finally, CAR2-deﬁcient mice calciﬁed subcutaneously implanted valve materials to a greater extent than wild-type mice, which was consistent with an important inhibitory activity in ectopic mineralization (R. Rajachar and C.M. Giachelli, unpublished observations). Thus, OPN may function not only as a physical inhibitor of apatite crystal growth, but may also promote mineral regression by controlling cellular gene expression patterns that favor mineral resorption. 22.2.4 Cell Signaling Pathways that Regulate Ectopic Mineralization

Recent studies have pointed to several cytokines and signaling pathways that may be particularly important in regulating ectopic mineralization, especially in vascular tissues. Several of the molecules and pathways that have come to the forefront

355

356

22 Ectopic Mineralization: New Concepts in Etiology and Regulation

recently are listed in Table 22.1. Components of the bone morphogenetic protein (BMP) signaling pathway seem to be of particular interest, as there is evidence that they may be involved in both the induction and inhibition of ectopic calciﬁcation. BMP2 has been identiﬁed in human atherosclerotic and calciﬁed aortic valve lesions, and is a known inducer of osteoblastic diﬀerentiation and mineralization [49]. In osteoblast precursor cells, BMP2 induces Msx2 signaling, and this pathway was found to be associated with calciﬁed vessels in fat-fed LDLR= mice (a model of diabetes) [50]. Furthermore, CMV-Msx2-expressing transgenic mice exhibited marked Msx2 expression in their aortas as well as aortic calciﬁcation after being fed a high-fat diet. Importantly, the Wnt signaling pathway appeared to be downstream of Msx2 signaling, since crossing CMV-Msx2 transgenic mice with the Wnt reporting TOPGAL transgenic mouse led to striking reporter activation in the aortic media as measured by b-galactosidase expression [51]. Finally, a preliminary report showed that TOPGAL mice injected three times weekly with BMP2 had elevated Msx2 and calcium levels in their aortas compared to controls, although BMP2 was not suﬃcient to induce aortic calciﬁcation but rather also required high-fat feeding [52]. A potential role for Wnt signaling in aortic valve calciﬁcation has also been recently reported [53]. Finally, Smad6, an inhibitory member of the Smad family of signaling molecules that are downstream mediators of BMP function, has been implicated in vascular calciﬁcation as Smad6-null mice develop aortic stenosis and valve calciﬁcation [54]. In contrast, Hruska’s group has presented compelling evidence that BMP7 can protect against uremia-induced vascular calciﬁcation. High-fat feeding of uremic LDLR= mice caused hyperphosphatemia, renal osteodystrophy and vascular calciﬁcation. However, treatment of these mice with BMP-7 corrected the hyperphosphatemia and ameliorated both the osteodystrophy and vascular calciﬁcation. These eﬀects of BMP-7 appeared to be due in part to a normalization of serum phosphate levels, and to direct eﬀects on vascular smooth muscle cells to inhibit osteochondrogenic phenotypic transition [55]. Runx2, a transcription factor that is absolutely required for bone and cartilage formation, has also been implicated as an important signaling molecule in vascular calciﬁcation. In vivo, Runx2 is normally not expressed in blood vessels, but is up-regulated at sites of vascular calciﬁcation [27, 34]. The calciﬁcation of vascular medial cells in vitro in response to a number of inducing factors [BMP2, elevated phosphate, interleukin (IL)-4, TNFa, etc.] leads to an up-regulation of Runx 2 and its downstream targets such as OPN and osteocalcin [3]. In a recent study, knockdown of Runx2 with small interfering RNA led to decreased SMC mineralization in response to IL-4, suggesting a required role for Runx2 [56]. In addition to stimulatory pathways, several calciﬁcation inhibitory pathways have also been described. In a spontaneously calcifying subset of arterial wall cells termed calcifying vascular cells (CVC), IGF-1 was shown to block both diﬀerentiation and mineralization, though these eﬀects were reversed by inhibitors of either the ERK or PI3K pathways [57]. In a separate study, the same investigators

22.2 Regulators of Ectopic Mineralization

found that N-3 fatty acids inhibited CVC diﬀerentiation and mineralization that was dependent on p38-MAPK and PPARgamma pathways [58]. Finally, Canﬁeld’s group identiﬁed a novel inhibitor of pericyte mineralization, the receptor tyrosine kinase, Axl [59]. More recently, statins were found to protect human aortic smooth muscle cells from elevated-phosphate-induced calciﬁcation by blocking down-regulation of the ligand for Axl, Gas6 [60]. The eﬀects of statins and Gas6 appeared to be mediated by preventing or inducing apoptosis, respectively, in this model system. 22.2.5 Roles of Cell Death and Bone Remodeling in Ectopic Mineralization

Tissue necrosis is sometimes observed at sites of ectopic mineralization. Cell death may stimulate ectopic mineralization by releasing apoptotic bodies or cell remnants that resemble matrix vesicles – small membranous structures that act as initiation sites for apatite crystallization in bone and teeth [61]. Such structures have been observed in the necrotic cores of calciﬁed atherosclerotic lesions, and may be particularly important in initiating calciﬁcation in this setting [62]. In support of this, the in-vitro culture of vascular smooth muscle cells under conditions that induce apoptosis (serum deprivation in the presence of elevated calcium or phosphate) induced apoptosis and calciﬁcation [63]. Of interest, the replenishment of serum factors (including MGP and fetuin) blocks apoptotic body calciﬁcation [64], further supporting the role of these molecules in preventing inappropriate calciﬁcation under normal conditions. A link between bone remodeling and ectopic calciﬁcation has also been proposed, and osteoporosis and cardiovascular calciﬁcation have been found to be correlated in postmenopausal women and end-stage renal disease patients [65]. Interestingly, mice deﬁcient in osteoprotegerin (OPG), a soluble member of the TNF receptor family, were osteoporotic and showed vascular calciﬁcation; this suggested that OPG and its regulators might be important in explaining the link between cardiovascular disease and osteoporosis [66]. In addition, osteoprotegerin and the bisphosphonates, alendronate and ibandronate, inhibited arterial calciﬁcation in warfarin- and/or vitamin D-treated rats at doses comparable to those used to inhibit bone resorption [67, 68]. In a subsequent study, the same investigators used a speciﬁc inhibitor of osteoclastic V-Hþ -ATPase, SB 242784, and were able to block both vascular calciﬁcation and osteoclastic resorption in rats treated with toxic doses of vitamin D [69]. These ﬁndings have led to the hypothesis that vascular calciﬁcation is linked to osteoclastic resorption. Price et al. have suggested that soft tissue calciﬁcation is promoted by circulating nucleational complexes composed of calcium phosphate mineral and the proteins fetuin and MGP that are released from remodeling bone [41]. Alternatively, alterations in systemic mineral balance that may occur during active bone remodeling (increased calcium and phosphate release to blood) are likely to predispose to vascular calciﬁcation, as previously discussed.

357

358

22 Ectopic Mineralization: New Concepts in Etiology and Regulation

22.3 Conclusions and Implications for Therapeutic Control of Ectopic Mineralization

Vascular calciﬁcation leads to a loss of compliance, elevated pulse pressure and left ventricular hypertrophy, and is strongly associated with mortality in patients with cardiovascular disease. Ectopic mineralization may also occur in bones and joints, and present new problems in the emerging ﬁelds of bone and craniofacial tissue engineering. A better understanding of cell-mediated mechanisms regulating matrix mineralization will aid in the development of novel therapeutics and strategies to address these problems. In addition, these mechanisms may also be useful in developing new strategies to promote healthy mineralization, where required.

Acknowledgments

The author thanks the current members of the laboratory who contributed to the studies described in this chapter, including Mohga El-Abbadi, James Li, Mei Speer, Rupak Rajachar, Hsueh Yang, and Amy Look. These studies were funded by NIH grants HL62329, HL081785, AR48798, and NSF grant EEC9529161.

References 1 M. El-Abbadi, C.M. Giachelli, Curr.

2 3 4 5

6 7

8

Opin. Nephrol. Hypertens. 2005, 14, 519–524. F. Rutsch, R. Terkeltaub, Curr. Opin. Rheumatol. 2003, 15, 302–310. M.Y. Speer, C.M. Giachelli, Cardiovasc. Pathol. 2004, 13, 63–70. T. Wada, M.D. McKee, S. Stietz, C.M. Giachelli, Circ. Res. 1999, 84, 1–6. S. Jono, M.D. McKee, C.E. Murry, A. Shioi, Y. Nishizawa, K. Mori, H. Morii, C.M. Giachelli, Circ. Res. 2000, 87, E10–E17. G.A. Block, T.E. Hulbert-Shearon, Am. J. Kidney Dis. 1998, 31, 607–617. W.G. Goodman, J. Goldin, B.D. Kuizon, C. Yoon, B. Gales, D. Sider, Y. Wang, J. Chung, A. Emerick, L. Greaser, R.M. Elashoﬀ, I.B. Salusky, N. Engl. J. Med. 2000, 342, 1478– 1483. N. Niederhoﬀer, Y.V. Bobryshev, I. Lartaud-Idjouadiene, P. Giummelly, J. Atkinson, J. Vasc. Res. 1997, 34, 386– 398.

9 T. Shimada, M. Kakitani, Y. Yamazaki,

10

11

12

13

14

15

H. Hasegawa, Y. Takeuchi, T. Fujita, S. Fukumoto, K. Tomizuka, T. Yamashita, J. Clin. Invest. 2004, 113, 561–568. J. Oh, R. Wunsch, M. Turzer, M. Bahner, P. Raggi, U. Querfeld, O. Mehls, F. Schaefer, Circulation 2002, 106, 100–1005. A.Y. Wang, M. Wang, J. Woo, C.W. Lam, P.K. Li, S.F. Lui, J.E. Sanderson, J. Am. Soc. Nephrol. 2003, 14, 159– 168. S. Ahmed, K.D. O’Neill, A.F. Hood, A.P. Evan, S.M. Moe, Am. J. Kidney Dis. 2001, 37, 1267–1276. J. Braun, M. Oldendorf, W. Moshage, R. Heidler, E. Zeitler, F.C. Luft, Am. J. Kidney Dis. 1996, 27, 394–401. A.P. Guerin, G.M. London, S.J. Marchais, F. Metivier, Nephrol. Dial. Transplant. 2000, 15, 1014– 1021. G.M. London, A.P. Guerin, S.J. Marchais, F. Metivier, B. Pannier, H.

References

16

17

18

19

20

21

22

23

24

25 26

27

Adda, Nephrol. Dial. Transplant. 2003, 18, 1731–1740. J.S. Shao, S.L. Cheng, N. CharltonKachigian, A.P. Loewy, D.A. Towler, J. Biol. Chem. 2003, 278, 50195– 50202. S. Jono, Y. Nishizawa, A. Shioi, H. Morii, Circulation 1998, 98, 1302– 1306. C. Schafer, A. Heiss, A. Schwarz, R. Westenfeld, M. Ketteler, J. Floege, W. Muller-Esterl, T. Schinke, W. JahnenDechent, J. Clin. Invest. 2003, 112, 357–366. M.W. Merx, C. Schafer, R. Westenfeld, V. Brandenburg, S. Hidajat, C. Weber, M. Ketteler, W. JahnenDechent, J. Am. Soc. Nephrol. 2005, 16, 3357–3364. K. Johnson, J. Goding, D. Van Etten, A. Sali, S.I. Hu, D. Farley, H. Krug, L. Hessle, J.L. Millan, R. Terkeltaub, J. Bone Miner. Res. 2003, 18, 994– 1004. M. Szweras, D. Liu, E.A. Partridge, J. Pawling, B. Sukhu, C. Clokie, W. Jahnen-Dechent, H.C. Tenenbaum, C.J. Swallow, M.D. Grynpas, J.W. Dennis, J. Biol. Chem. 2002, 277, 19991–19997. M. Kuroo, Y. Matsamura, H. Aizawa, H. Kawaguchi, T. Suga, T. Utsugi, Y. Ohyama, M. Kurabayashi, T. Kaname, E. Kume, H. Iwasaki, A. Iida, T. Shiraki-Iida, S. Nishikawa, R. Nagai, Y. Nabeshima, Nature 1997, 390, 45– 51. A.A. Bergen, A.S. Plomp, E.J. Schuurman, S. Terry, M. Breuning, H. Dauwerse, J. Swart, M. Kool, S. van Soest, F. Baas, J.B. ten Brink, P.T. de Jong, Nat. Genet. 2000, 25, 228– 231. S.S. Spicer, S.E. Lewis, R.E. Tashian, B.A. Schulte, Am. J. Pathol. 1989, 134, 947–954. M. Hui, H.C. Tenenbaum, Anat. Rec. 1998, 253, 91–94. A. Shioi, M. Katagi, Y. Okuno, K. Mori, S. Jono, H. Koyama, Y. Nishizawa, Circ. Res. 2002, 91, 9–16. N.X. Chen, K.D. O’Neill, D. Duan, S.M. Moe, Kidney Int. 2002, 62, 1724– 1731.

28 Y. Tintut, J. Patel, F. Parhami, L.L.

29

30

31 32

33

34

35 36 37

38

39

40 41

42

43

Demer, Circulation 2000, 102, 2636– 2642. F. Rutsch, N. Ruf, S. Vaingankar, M.R. Toliat, A. Suk, W. Hohne, G. Schauer, M. Lehmann, T. Roscioli, D. Schnabel, J.T. Epplen, A. Knisely, A. Superti-Furga, J. McGill, M. Filippone, A.R. Sinaiko, H. Vallance, B. Hinrichs, W. Smith, M. Ferre, R. Terkeltaub, P. Nurnberg, Nat. Genet. 2003, 34, 379–381. K. Johnson, M. Polewski, D. van Etten, R. Terkeltaub, Arterioscler. Thromb. Vasc. Biol. 2005, 25, 686–691. A.M. Ho, M.D. Johnson, D.M. Kingsley, Science 2000, 289, 265–270. D. Harmey, L. Hessle, S. Narisawa, K.A. Johnson, R. Terkeltaub, J.L. Millan, Am. J. Pathol. 2004, 164, 1199–1209. H. Chau, S. El-Maadawy, M.D. McKee, H.S. Tenenhouse, J. Bone Miner. Res. 2003, 18, 644–657. S.A. Steitz, M.Y. Speer, G. Curinga, H.Y. Yang, P. Haynes, R. Aebersold, T. Schinke, G. Karsenty, C.M. Giachelli, Circ. Res. 2001, 89, 1147– 1154. X. Li, H.Y. Yang, C.M. Giachelli, Circ. Res. 2006, 98, 905–912. H. Yang, G. Curinga, C.M. Giachelli, Kidney Int. 2004, 66, 2293–2295. J.W. Fischer, S.A. Steitz, P.Y. Johnson, A. Burke, F. Kolodgie, R. Virmani, C. Giachelli, T.N. Wight, Arterioscler. Thromb. Vasc. Biol. 2004, 24, 2391– 2399. G. Luo, P. Ducy, M.D. McKee, G.J. Pinero, E. Loyer, R.R. Behringer, G. Karsenty, Nature 1997, 386, 78–81. M. Murshed, T. Schinke, M.D. McKee, G. Karsenty, J. Cell Biol. 2004, 165, 625–630. A.F. Zebboudj, V. Shin, K. Bostrom, J. Cell. Biochem. 2003, 90, 756–765. P.A. Price, J.M. Caputo, M.K. Williamson, J. Bone Miner. Res. 2002, 17, 1171–1179. T.E. Bunton, N.J. Biery, L. Myers, B. Gayraud, F. Ramirez, H.C. Dietz, Circ. Res. 2001, 88, 37–43. C.M. Giachelli, N. Bae, M. Almeida, D.T. Denhardt, C.E. Alpers, S.M.

359

360

22 Ectopic Mineralization: New Concepts in Etiology and Regulation

44

45

46

47

48

49

50

51

52

53

54

55

56

Schwartz, J. Clin. Invest. 1993, 92, 1686–1696. S. Jono, C. Peinado, C.M. Giachelli, J. Biol. Chem. 2000, 275, 20197– 20203. M.Y. Speer, Y.C. Chien, M. Quan, H.Y. Yang, H. Vali, M.D. McKee, C.M. Giachelli, Cardiovasc. Res. 2005, 66, 324–333. M.Y. Speer, M.D. McKee, R.E. Guldberg, L. Liaw, H.Y. Yang, E. Tung, G. Karsenty, C.M. Giachelli, J. Exp. Med. 2002, 196, 1047–1055. S.A. Steitz, M.Y. Speer, M.D. McKee, L. Liaw, M. Almeida, H. Yang, C.M. Giachelli, Am. J. Pathol. 2002, 161, 2035–2046. R. Ohri, E. Tung, R. Rajachar, C.M. Giachelli, Calcif. Tissue Int. 2005, 76, 307–315. K. Bostrom, K.E. Watson, S. Horn, C. Wortham, I.M. Herman, L.L. Demer, J. Clin. Invest. 1993, 91, 1800–1809. D.A. Towler, M. Bidder, T. Latiﬁ, T. Coleman, C.F. Semenkovich, J. Biol. Chem. 1998, 273, 30427–30434. S.L. Cheng, J.S. Shao, N. CharltonKachigian, A.P. Loewy, D.A. Towler, J. Biol. Chem. 2003, 278, 45969– 45977. J.S. Shao, J. Cai, D.A. Towler, Arterioscler. Thromb. Vasc. Biol. 2006, 26, 1423–1430. N.M. Rajamannan, M. Subramaniam, F. Caira, S.R. Stock, T.C. Spelsberg, Circulation 2005, 112, I229–I134. K.M. Galvin, M.J. Donovan, C.A. Lynch, R.I. Meyer, R.J. Paul, J.N. Lorenz, V. Fairchild-Huntress, K.L. Dixon, J.H. Dunmore, M.A. Gimbrone, Jr., D. Falb, D. Huszar, Nat. Genet. 2000, 24, 171–174. M.R. Davies, R.J. Lund, S. Mathew, K.A. Hruska, J. Am. Soc. Nephrol. 2005, 16, 917–928. L.C. Hofbauer, J. Schrader, U. Niebergall, V. Viereck, A. Burchert, D. Horsch, K.T. Preissner, M. Schoppet, Thromb. Haemost. 2006, 95, 708–714.

57 K. Radcliﬀ, T.B. Tang, J. Lim,

58

59

60

61 62 63

64

65

66

67

68

69

Z. Zhang, M. Abedin, L.L. Demer, Y. Tintut, Circ. Res. 2005, 96, 398– 400. M. Abedin, J. Lim, T.B. Tang, D. Park, L.L. Demer, Y. Tintut, Circ. Res. 2006, 98, 727–729. G. Collett, A. Wood, M.Y. Alexander, B.C. Varnum, R.P. Boot-Handford, V. Ohanian, J. Ohanian, Y.W. Fridell, A.E. Canﬁeld, Circ. Res. 2003, 92, 1123–1129. B.K. Son, K. Kozaki, K. Iijima, M. Eto, T. Kojima, H. Ota, Y. Senda, K. Maemura, T. Nakano, M. Akishita, Y. Ouchi, Circ. Res. 2006, 98, 1024– 1031. C.M. Shanahan, Nephrol. Dial. Transplant. 2006, 21, 1166–1169. K.M. Kim, Fed. Proc. 1976, 35, 156– 162. J.L. Reynolds, A.J. Joannides, J.N. Skepper, R. McNair, L.J. Schurgers, D. Proudfoot, W. Jahnen-Dechent, P.L. Weissberg, C.M. Shanahan, J. Am. Soc. Nephrol. 2004, 15, 2857– 2867. J.L. Reynolds, J.N. Skepper, R. McNair, T. Kasama, K. Gupta, P.L. Weissberg, W. Jahnen-Dechent, C.M. Shanahan, J. Am. Soc. Nephrol. 2005, 16, 2920–2930. L.M. Banks, B. Lees, J.E. MacSweeney, J.C. Stevenson, Eur. J. Clin. Invest. 1994, 24, 813–817. N. Bucay, I. Sarosi, C.R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H.L. Tan, W. Xu, D.L. Lacey, W.J. Boyle, W.S. Simonet, Genes Dev. 1998, 12, 1260–1268. P.A. Price, S.A. Faus, M.K. Williamson, Arterioscler. Thromb. Vasc. Biol. 2001, 21, 817–824. P.A. Price, H.H. June, J.R. Buckley, M.K. Williamson, Arterioscler. Thromb. Vasc. Biol. 2001, 21, 1610–1616. P.A. Price, H.H. June, J.R. Buckley, M.K. Williamson, Circ. Res. 2002, 91, 547–552.

361

23 Pathological Calciﬁcation of Heart Valve Bioprostheses Birgit Glasmacher and Martin Krings

Abstract

Improvements in healthcare yields a rise in numbers of elderly people, which in turn increases the demand for healthcare among this population. Today, some 300 000 heart valve replacements are carried out worldwide each year, with about 40% of the prostheses being tissue-based. In the aortic position, these bioprostheses are the valves of choice in patients aged more than 65–70 years. Although these valves are advantageous in terms of their hemodynamics, thrombogenicity, low risk of bleeding and minimal need for anticoagulants, they suﬀer from limited and unpredictable durability. Valve failure is mainly due to tissue calciﬁcation, caused by multiple factors [1]. In this chapter we demonstrate that it is possible to mimic, induce, and investigate this pathological process in vitro by the choice of an appropriate model, whilst neglecting possible host factors such as unphysiological calcium and phosphate levels or missing inhibitory proteins. The local sites of calciﬁcation can be predicted in advance by using a nondestructive, holographic method. Multiple parameters such as the valve tissue origin, valve design, glutaraldehyde ﬁxation, and alternative chemical and irradiation treatments, have been identiﬁed as inﬂuencing calciﬁcation. Non-destructive evaluation of the calciﬁc deposits is possible using microradiography (m-X-ray), clinical and industrial computed tomography (CT) and m-CT (Synchrotron). In this way, the degree of calciﬁcation can be determined with computer image analysis. Key words: biological heart valve prostheses, xenografts, ectopic and intrinsic calciﬁcation, in-vitro calciﬁcation model, imaging.

23.1 Introduction

Since the ﬁrst valve replacement by Starr and Edwards, and Harken and colleagues, in 1960, none of the currently available prosthetic heart valves approaches Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

362

23 Pathological Calciﬁcation of Heart Valve Bioprostheses

the normal human valve in terms of hemodynamic function or freedom from valve-related complications. Whilst mechanical valves are thrombogenic, and the requisite long-term anticoagulation is associated with a high risk of bleeding, the use of xenograft valves is limited by their structural degeneration and consequent poor durability [2]. Today, the German Society for Thoracic and Cardiovascular Surgery estimates a demand for 160 operations per one million inhabitants in the treatment of acquired heart valve failure. ‘‘Bioprosthetic heart valve replacement began with the use of aortic homografts and then stentless porcine valves. At this time, the added diﬃculty of freehand sewn implant techniques was an important source of morbidity through the damaging eﬀects of cardiopulmonary bypass and poorly developed myocardial protection. Stent mounting of both human and xenograft valves greatly facilitated the implant technique and allowed widespread adoption of bioprostheses . . . . But stent mounting applies excessive stress on the biological tissue. Renewed interest in stentless xenografts was thus stimulated’’ (S. Westaby in [1]). In general, biological heart valve prostheses (bioprostheses, xenografts) are either original aortic valves or aortic valve leaﬂets from pigs (PA), or valves made from bovine pericardium (BP). All of the tissues are chemically treated, mostly with glutaraldehyde (GA; ﬁxation with either 0.5 or 0.6 wt% GA and subsequent storage either in 0.2 wt% GA or 4 wt% formaldehyde). The valves may be either stented (S) or stentless (SL), as depicted in Figure 23.1. The lifetime of these valves is unpredictably limited by calciﬁcation – that is, the ectopic, intrinsic deposition of hydroxyapatite within the valve tissue (Fig. 23.2). In fact, most valves explanted due to primary tissue failure exhibit valve stenosis due to excessive ‘‘eggshell’’ leaﬂet calciﬁcation.

Fig. 23.1 Stented bovine (left) and stentless porcine (center), and bovine stentless (right) heart valve prostheses.

23.1 Introduction

Fig. 23.2 Macroscopic view of an in-vitro-calciﬁed commercial porcine heart valve.

Thus, 40 years since their ﬁrst implantation in 1966, bioprosthetic heart valves still suﬀer from two major dysfunctions in long-term postoperative life, namely tissue rupture and calciﬁcation. Calciﬁcation is also one of the major causes of failure of native connective tissues, such as heart valves and blood vessels, as well as for many other implantable biomaterials such as polyurethane vascular grafts [3, 4]. A number of host and implant factors, either alone or in synergy, may contribute to the initiation and further development of calciﬁcation. Dead cell remnants, lipids, inadequate crosslinking and mechanical stresses are just some of the alternative explanations for the initiation of calciﬁcation [5–10]. In designing new methods of production for new biomaterials, a large number of anti-calciﬁcation treatments have been proposed, although until now no method has proven its eﬀectiveness in long-term applications in vivo [11]. Attempts to develop novel implantable biomaterials have led to the development of a number of diﬀerent in-vitro and in-vivo models to evaluate the eﬃciency of anti-calcifying treatments [12]. Typical examples of screening or preclinical investigations include subcutaneous implantation in rats, and orthotopic or heterotopic implantations in the blood circulatory system of sheep. Today, these in-vivo models have been accepted as predictive methods for evaluating the potential calciﬁcation of biomaterials. However, the need to develop reliable tests for the many anti-calciﬁcation methods that have been proposed renders the use of such animal models both economically ineﬀective and time-consuming, at least in the initial phases of testing.

363

364

23 Pathological Calciﬁcation of Heart Valve Bioprostheses

23.2 In-Vitro Calciﬁcation Models

In-vitro models allow for important alternatives in investigations of the calciﬁcation process. Although a signiﬁcant number of in-vitro models have been proposed, before they can be accepted as reliable in the screening of valve calciﬁcation or anti-calciﬁcation methods, their relevance to animal and clinical investigations should be proven [4]. During the past few years, a reliable in-vitro method for testing valve and biomaterial calciﬁcation has been developed by the authors [8–10, 13–20] in accordance with ISO 5840 as well as for testing biomaterial calciﬁcation [13, 14]. The details of the test protocol are described in Section 23.4. 23.3 Heart Valve Bioprostheses

Porcine and bovine valves from diﬀerent manufacturers, with diﬀerent ﬁxation chemicals (GA, carbodiimide-based method [10, 21]) and g-irradiation treatment) were included in the studies, as well as porcine aortic and pulmonary valves collected from local slaughterhouses and processed in our laboratories. Both, the porcine and bovine valves, were either stentless (SL) or stented (S) (see Fig. 23.1). 23.4 Calciﬁcation Hypotheses and Study Design

The pathological calciﬁcation of valves is a multifactorial process, and in order to gain insight into the development of the process, an in-vitro model might be helpful. This model should allow for the separate investigation of diﬀerent inﬂuencing parameters to verify the following hypotheses: Calciﬁcation occurs at pronounced ‘‘high-stress’’ areas of the valves (as reported by surgeons from human explants). Aortic valve leaﬂet tissue calciﬁes diﬀerently from pericardial tissue. Porcine valves calcify diﬀerently from bovine valves. Pericardial valves are produced in a more reproducible way than naturally grown aortic valves (possible impact on calciﬁcation). Stentless valves calcify less than stented valves because of reduced stress levels during the opening and closing phases (‘‘cycles’’). Stentless valves, with their aortic wall tissue, part calcify diﬀerently to stented valves. The chemical ﬁxation method has an impact on calciﬁcation (glutaraldehyde versus alternatives). The degree of calciﬁcation increases with time.

23.5 Calciﬁcation Imaging Methods

Fig. 23.3 Holographic interferograms of diﬀerent valve types, showing irregularities at the outermost surface of the leaﬂet tissue (note the regular pattern in the center valve). (Figure produced in cooperation with M. Deiwick and G. von Bally, Muenster.)

Calciﬁcation is studied in vitro with a specially designed calciﬁcation test device [8–10], and following a test protocol has been designed to meet ISO and FDA standards for aortic and mitral valves. The tests are performed at 37 C and pH 7.4 under quasi-physiological conditions, with reductions in physiological pressure across the valves being monitored according FDA and ISO standards, and with a synthetic calcifying solution with a solubility product (Ca P) of 130 (mg dL1 ) 2 . In order to compensate for consumed calciﬁcation ions, the solution is renewed weekly. The tests are performed at an increased frequency of 300 cycles min1 , such that test periods of 4 to 6 weeks result in a total of 12 to 19 million cardiac cycles. Selected valves also underwent non-destructive holographic measurements prior to calciﬁcation testing (Fig. 23.3) [8, 9].

23.5 Calciﬁcation Imaging Methods

The main problem when imaging the valves is to detect and localize the calciﬁc deposits within the tissue, but without destroying the valve. For this, clinical radiography and computed tomography (CT) have been found suitable. It has been shown that radiographically depicted calciﬁcation correlates with the mass amount of tissue calcium phosphate as detected with atomic absorption spectroscopy. In addition, von Kossa staining of histological cross-sections conﬁrmed the presence of calcium phosphates (Fig. 23.4) [15]. Furthermore, energy-dispersive X-ray (EDX) analyses of scanning electron microscopy (SEM) images conﬁrmed that the deposits contained both of calcium and phosphorus (Fig. 23.4) [15]. The extent of calciﬁcation is described as the degree of calciﬁcation, expressed as a percentage. Calciﬁcation of the valve leaﬂets may be determined weekly using mXradiography (with a conventional mammography X-ray machine operating at 7 mAs and 22 kV), and this resulted in two-dimensional (2-D) radiographs. The

365

366

23 Pathological Calciﬁcation of Heart Valve Bioprostheses

Fig. 23.4 Lower: Energy-dispersive X-ray analyses of the scanning electron microscopy images (upper; EDX analysis of indicated area), showing calcium phosphate deposition. Top right: von Kossa staining of a tissue cross-section conﬁrmed the presence of calcium phosphates. (Figure produced in cooperation with P. Koutsoukos, Patras.)

23.6 Calciﬁcation Patterns

degree of leaﬂet calciﬁcation was determined by computer image analysis, with a spatial resolution of <50 mm being achieved by magnifying and scanning the radiographs. In addition to the leaﬂet analysis, the whole valve was mX-rayed laterally before (native) and after 12 million cycles to provide an overview of the entire calciﬁcation progress [17]. In order to depict the degree of calciﬁcation three-dimensionally in the valve and wall tissue of stentless valves, computed tomography was performed. Scans were taken using clinical, industrial, and m-CT, before and after the calciﬁcation test, followed by 2-D and 3-D image analyses respectively, and m-tomography [16]. The volumetric total degrees of calciﬁcation determined via CT and m-CT measurements were comparable [16]. CT-recording was performed using a clinical spiral CT with: (i) Siemens Somatosom Sensation 16; and (ii) Tomoscan AV Philips Medical Systems (120 kV, 125 mAs mm1 , spiral mode). CT and m-CT data analysis was performed with a software developed in Aachen (DISOS2). An individual Houndsﬁeld unit (HU)related, volume-based total 3-D degree of calciﬁcation was calculated. An industrial X-ray tomograph (420 kV; BIR Inc., USA, with spatial resolution < 0.1 mm) was also used. The m-CT data were obtained using a TomoScope 30s (in cooperation with the University of Erlangen, Germany). Some valve leaﬂets have been inspected with a FOX-nanofocus m-CT-system (resolution < 1 mm; Feinfocus GmbH, Comet AG, Garbsen, Germany, in cooperation with LEBAO, Hannover Medical School) and m-CT 80 (Scanco Medical; 40–70 kV, resolution 10 mm, in cooperation with IW, Leibniz Universita¨t, Hannover). Two-dimensional analyses of the m-CT-data were performed with a demo software version of ImpactView2. In addition, hard X-ray m-tomography (Synchrotron) [22] was conducted at ESRF in Grenoble (Beamline ID-22; phase-contrast CT with a small source size of 0:7 0:5 mm, small divergence, beam size 60 m distant from the source, 1:2 2:2 mm) with a resolution of 1 mm, and also at DESY in Hamburg (both in cooperation with C. Schroer) [16].

23.6 Calciﬁcation Patterns

The in-vitro calciﬁcation ﬁndings (see Figs. 23.2 and 23.5) are generally conﬁrmed by in-vivo human explant analysis [15, 17]. Further analysis of the m-CT data with ImpactView2 showed intrinsic calciﬁcation. Calciﬁed areas were found at a HU range of þ400 to þ3000, whereas the whole valve tissue was present between 220 and þ3000 in the reconstructed m-CT data [17]. The tissue origin might have an impact on calciﬁcation, as bovine stented valves calciﬁed less than porcine stented valves [degree of calciﬁcation 18.3% (n ¼ 13) versus 39.9% (n ¼ 32)] [18]. Once initiated, calciﬁcation progresses with time, as shown at the depicted time intervals equivalent to 3, 6, 9, and 12 million cycles in Figure 23.6 [17]. Calciﬁcation was seen to occur at pronounced ‘‘high-stress’’ valve

367

368

23 Pathological Calciﬁcation of Heart Valve Bioprostheses

Fig. 23.5 Valve after 12 million-cycle, pH-controlled calciﬁcation testing. (Figure reproduced from Ref. [17].)

Fig. 23.6 Progression of leaﬂet calciﬁcation in a valve, as shown by mX-radiography. (Figure reproduced from Ref. [17].)

areas that could be detected by holographic interferography as irregular areas (e.g., as shown in Fig. 23.3), with a linear relationship (r ¼ 0.72, p ¼ 0.001) at 19 million cycles (n ¼ 34) [20]. Since mechanical stress – when visualized by holographic interferometry, as depicted in Figure 23.3 [20] – has been shown to inﬂuence calciﬁcation, the stentless-valve leaﬂets might calcify less than stented-valve leaﬂets due to lower leaﬂet tissue stresses during valve function. This was conﬁrmed by the holographic measurements. In contrast, the additional wall tissue of the stentless valves might be prone to calciﬁcation. These results conﬁrmed the hypothesis, that stented porcine and bovine valves (S-BP, S-PA) calciﬁed more than stentless valves, as shown in Figure 23.7 [22.5% (n ¼ 29) S-BP versus 32.8% (n ¼ 72) S-PA and 7.8% SL-BP (n ¼ 6) versus 19.4% SL-PA (n ¼ 20)]. In an initial study, g-irradiation treatment led to a more physiological, unimpaired valve function and less calciﬁcation (17.7% versus 25.4%) among a small number of tested porcine valves (n ¼ 4 for each group). In general, the anti-calciﬁcation treatments investigated did not improve the calciﬁcation behavior of bioprosthe-

23.7 Description of Findings

Fig. 23.7 Mean degree of leaﬂet calciﬁcation of bovine and porcine stentless versus stented bioprostheses.

ses, and to date carbodiimide treatment has been proved the most successful in the in-vitro tests [10, 21].

23.7 Description of Findings

Currently, a new procedure for the in-vitro calciﬁcation testing of bioprosthetic heart valves under dynamic accelerated conditions is available. However, one question that may be raised is whether in-vitro calciﬁcation testing reliably provides results that are similar to the mechanisms and topography of the calciﬁc deposits deposited on valves in vivo. Intrinsic calciﬁcation in tissue areas similar to that observed in vivo, together with similar crystal phases, could provide evidence of such reliability. Although these comparisons have been presented previously [14, 15], SEM investigations (Leo Supra 35VP) of the valves calciﬁed in vitro were performed in order to examine the morphological characteristics of the calciﬁc deposits (cf. Fig. 23.4). The investigations into mineralized tissues included surface- and deep-tissue examinations of cross-sectional areas of the valve leaﬂet and aortic wall tissue. The ﬁndings revealed that extensive calciﬁcation had developed inside and at the surface of the tissue, and was associated with thick or thinner collagen ﬁbers in the interior. EDX analyses of the crystal phases observed showed the presence of high calcium and phosphorus peaks. The crystal phases included nanoscale to submicron-scale calcium phosphate crystals (hydroxyapatite), mixed with a larger-sized (microscale), typically octacalcium phosphate (OCP) plate-like conﬁguration, and very large (20–100 mm) canonical geometric crystals (typically dicalcium phosphate dehydrate, CaHPO4 2H2 O or DCPD) crys-

369

370

23 Pathological Calciﬁcation of Heart Valve Bioprostheses

tals. At most crystal surfaces – and especially for the larger crystals – hydrolytic defects were visible and associated with the formation of spongy, smaller-sized crystal aggregates. Similar crystal conﬁguration and topography were also observed in bioprosthetic heart valves that were required to be explanted from humans several years after implantation following valve failure caused by calciﬁcation [25, 26].

23.8 Conclusions and Future Research

Some 40 years after their ﬁrst implantation, bioprosthetic heart valves still suﬀer from calciﬁcation, although major improvements have been made in inhibiting calciﬁcation with the development of alternative ﬁxation methods and/or new valve designs. A suitable in-vitro calciﬁcation methodology has been developed which includes appropriate imaging, and so far this has been used to identify the pathological calciﬁcation of more than 160 bioprostheses, demonstrating degrees of leaﬂet calciﬁcation ranging from 2% to 70%. The calciﬁc deposits formed were similar in composition, morphology and topography to those present in vivo. The lower calciﬁcation rates, which would be expected following the introduction of these new-generation bioprostheses, and the testing of viable tissue-engineered valves, demands for an improved, accelerated in-vitro calciﬁcation-testing technique. The high risk of microbiological contamination of bioprostheses during test periods exceeding 10 weeks further emphasizes the need for short-term invitro testing. By implementing a pH-controlled, constant solution supersaturation (CSS) methodology developed at the University of Patras [13, 14, 24, 25, 27], the test period might indeed be shortened in future investigations.

Acknowledgments

These studies were partly funded by grants from the IKYDA 2003 bilateral cooperation program DAAD (DE)-IKY (GR) (D/03/40393), the Interdisciplinary Centre for Clinical Research IZKF ‘‘BIOMAT’’, within the faculty of Medicine at the RWTH Aachen University (IZKF TVB108), by a fellowship of the International Atomic Energy Agency, by a grant from the European Synchrotron Research Facility, and by Medtronic, Santa Ana, USA. The authors are also very grateful to the valve manufacturers (Braile, Prima Plus, BioSud, Medtronic), for providing the valves used in these studies. The authors are also grateful for the technical support of W. Kalender (University of Erlangen), R. Gu¨nther and A. Mahnken (Aachen University Hospital), S. Barlach (Hannover Medical School), and to F. Bach, M. Behling and S. Mahr (Leibniz Universita¨t, Hannover). Their thanks for data reconstruction are also extended to T. Wu, D. Mavrilas, D. Kannepolous, J. Patommel and C. Schroer, and to E. Kesseylitz, T. Herzog, S. Wienecke, K. Stahley and S. Huﬀ for conducting the experimental investigations.

References

References 1 H.A. Huysmans, T.E. David, S.

2

3

4 5

6 7 8

9

10

11

12 13

14

15

Westaby (Eds.), Stentless Bioprostheses. ISIS Medical Media, Oxford, 1999. J.E. Rubay, G. El Khoury, M. Buche, in: H.A. Huysmans, T.E. David, S. Westaby (Eds.), Stentless Bioprostheses. ISIS Medical Media, Oxford, 1999, p. 47. D.J. Wheatley, L. Raco, G.M. Bernacca, I. Sim, P.R. Belcher, J.S. Boyd, Eur. J. Cardiothorac. Surg. 2000, 17, 440. J.M. Gross, J. Thorac. Cardiovasc. Surg. 2001, 121(3), 428. M.J. Thubrikar, D.J. Deck, J. Aouad, S.P. Nolan, J. Thorac. Cardiovasc. Surg. 1983, 86, 115. R.J. Levy, J. Heart Valve Dis. 1994, 3, 101. F.J. Schoen, R.J. Levy, J. Biomed. Mater. Res. 1999, 47(4), 439. B. Glasmacher, H. Reul, S. Schneppershoﬀ, S. Schreck, G. Rau, J. Heart Valve Dis. 1998, 7(4), 415. M. Deiwick, B. Glasmacher, E. Pettenazzo, D. Hammel, W. Castellon, G. Thiene, H. Reul, E. Berendes, H.H. Scheld, Thorac. Cardiovasc. Surg. 2001, 49(2), 78. M. Krings, F. Everaets, M. Torriani, M. Hendriks, B. Glasmacher, Artif. Organs 2005, 29(9), 716. N.R. Vyavahare, W. Chen, R.R. Joshi, C.-H. Lee, D. Hirsch, J. Levy, F.J. Schoen, R.J. Levy, Cardiovasc. Pathol. 1997, 6, 219. W.J. Mako, I. Vesely, J. Heart Valve Dis. 1997, 6, 316. J. Kapolos, D. Mavrilas, Y.F. Missirlis, P.G. Koutsoukos, J. Biomed. Mater. Res. (Appl. Biomater.) 1997, 38, 183. D. Mavrilas, J. Kapolos, P.G. Koutsoukos, D. Dougenis, J. Mater. Sci. Mater. Med. 2004, 15, 699. E. Pettenazzo, M. Deiwick, G. Thiene, G. Molin, B. Glasmacher, F. Martignago, T. Bottio, H. Reul, M.

16

17

18

19

20

21

22

23

24

25

26 27

28

Valente, J. Thorac. Cardiovasc. Surg. 2001, 121(3), 500. M. Krings, A. Mahnken, C. Schroer, J. Patommel, W. Kalender, B. Glasmacher, in: H. Handels, et al. (Eds.), Bildverarbeitung in der Medizin. Springer-Verlag, Berlin, 2006, p. 444. M. Krings, D. Kanellopoulou, D. Mavrilas, B. Glasmacher, Mat.-wiss. u. Werkstoﬀtech. 2006, 37, 432. B. Glasmacher, M. Krings, H. Kesper, C. Schroer, Int. J. Artif. Organs 2001, 24(8), 548. M. Krings, D. Kanellopoulou, D. Mavrilas, B. Glasmacher, Biomedizinische Technik 2005, 50, 330. B. Glasmacher, in: E. Wintermantel, S.-W. Ha (Eds.), Medizintechnik mit biokompatiblen Werkstoﬀen und Verfahren. 3rd edn. Springer, Berlin, 2002, p. 571. F. Everaerts, M. Gillissen, M. Torrianni, P. Zilla, P. Human, M. Hendriks, J. Feijen, J. Heart Valve Dis. 2006, 15, 269. P. Weiss, L. Obadia, D. Magne, X. Bourges, C. Rau, T. Weitkamp, I. Khairoun, J.M. Boulder, D. Chappard, O. Gauthier, G. Daculsi, Biomaterials 2003, 24, 4591. M. Hendriks, F. Everaerts, M. Verhoeven, J. Long-Term Eﬀ. Med. Implants 2001, 11, 163. D. Mavrilas, A. Apostolaki, J. Kapolos, P.G. Koutsoukos, M. Melachrinou, V. Zolota, D. Dougenis, J. Crystal Growth 1999, 205, 554. D. Mikroulis, D. Mavrilas, J. Kapolos, P.G. Koutsoukos, C. Lolas, J. Mater. Sci. Mater. Med. 2002, 13, 885. B.B. Tomazic, W.E. Brown, F.J. Schoen, J. Biomed. Mater. Res. 1994, 28, 35. P. Koutsoukos, Z. Amjad, M.B. Tomson, G.H. Nancollas, J. Am. Chem. Soc. 1980, 102(5), 1553. The World Medical Market Fact File. MDIS Publications Limited, 2001.

371

373

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials, Tissue Engineering and Biomineralization Pedro L. Granja, Jose´ Paulo Pereira, and Ma´rio A. Barbosa

Abstract

Medical devices and biomaterials have been acquiring increasing importance in the treatment of a large number of diseases and trauma. The internet has prompted a revolution in healthcare by vastly expanding public access to medical information, thus becoming an essential infrastructure in the healthcare environment. The Biomaterials Network (Biomat.net) is a collection of some selected internet links related to biomaterials and tissue engineering, and also some relevant links to biomedical engineering, biology, medicine and health sciences in general. It is among the most relevant information resources in the ﬁled, available online. Biomineralization is a key subject in this area of research, with special relevance on materials for orthopedic and cardiovascular applications, and the pages of Biomat.net have also been reﬂecting the interest of this ﬁeld to the biomaterials community. An overview on Biomat.net is provided. Key words: biomaterials network, internet, healthcare, resource, biomaterials, tissue engineering, biomineralization.

24.1 The Internet as a Major Healthcare Resource

The internet has prompted a revolution in healthcare by vastly expanding public access to medical information, thus becoming an essential infrastructure in the healthcare environment. The internet is not only an endless source of information, providing a wide number of useful resources, but it also constitutes a vast uncontrolled assembly of unorganized data. Everyone familiar with it is already aware of its unmatched advantages, as well as its many drawbacks. In the case of healthcare, the anarchic ﬂow of information it provides has been disturbing governments and regulatory institutions, as well as patients.

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

374

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

The widespread availability of the internet has allowed patients to be better informed about their medical conditions. It has even driven patients to elaborate their own diagnosis – designated as the ‘‘empowered patient syndrome’’ – which constitutes a profound modiﬁcation of the patient–physician relationship [1–6]. For instance, after a survey Hesse and colleagues recently reported that 48.6% of patients confessed to going ‘‘online’’ ﬁrst, before going to the physician [4]. This can become a very dangerous situation, since information contained in websites can lead to decisions based on misleading information. There are several examples of commercial healthcare companies that have exaggerated the properties of their products, inducing consumers to make erroneous judgments. The uncontrolled nature of the internet makes it diﬃcult to address responsibilities, and highlights the importance of reliable and responsible portals. The latter are websites specialized in a given subject, aimed at providing sorted, selected, and organized information. Governmental organizations such as the American Food and Drug Administration (FDA, www.fda.gov), and non-proﬁt private initiatives, such as the European Confederation of Medical Device Associations (www .eucomed.org), or the Internet Healthcare Coalition (www.ihealthcarecoalition .org), have been paying increased attention to this problem. Rather than restricting the speaker, their main goal is to empower the recipient of information that, in this case, includes consumers, healthcare professionals, academics, government oﬃcials, and representatives from companies. The healthcare industry, service providers and hospitals, and their interrelationships, have also considerably changed due to the advent of new technologies, such as electronic commerce [7–9]. Electronic data interchange and the internet are becoming increasingly useful tools, by providing communication services, such as e-mail and discussion groups, as well as direct access to scientiﬁc and health-related information resources, including libraries and databases. Improvements in design and manufacturing are considerable, through useful technical data and services, and several new tools such as remote monitoring and maintenance of manufacturing processes. A number of online resources are now focused on clinical trials, oﬀering information to study participants, clinicians, and researchers, as well as to the general public. Internet access to images and data facilitates technologies such as teleradiology with increased functionalities, thus allowing patient monitoring for diagnostic and therapeutic applications. In order to fully implement the advancements provided by these new technologies, several problems must also be eﬃciently addressed, such as the lack of eﬃcient regulation, in order to assure that information is accurate and truthful. Keeping clinical and business data private is also of major concern, and tools to keep intruders out of the systems are in constant development. The internet can also play a key role in informing the public about the impact of new technologies, therefore allowing an informed consent on the industrialization and application of emerging technologies. The public opinion can play a major role in obstructing the approval of new technologies based on their perception of risk, rather than on scientiﬁc evidence. Such was the case of transgenic foods,

24.2 Impact of Biomaterials Science in Modern Society

for instance. In the case of medical devices, the forthcoming decades are likely to be the ones of biological solutions in this ﬁeld, with a considerable increase of incorporation of biological components (cells, proteins) in medical devices, making this subject of even more pronounced importance.

24.2 Impact of Biomaterials Science in Modern Society

Medical devices and biomaterials have been acquiring increasing importance in the treatment of a large number of diseases and trauma. Artiﬁcial hearts, heart valves, pacemakers, contact lenses, mammary implants, hip and knee prostheses, and dental implants are some of the devices best known to the general public. Biomaterials science considerably contributes to improve the quality of life of an increasing amount of people. The growing life span in modern western countries is a consequence of the recent achievements in the treatment of diseases which, in part, were due to the developments in biomaterials science. In addition, the enhanced life span associated with the decrease in birth rate, has raised several additional problems, namely the increase in the proportion of elderly people in the growing population, the bodies of whom naturally do not function as well as before and show consequent increased needs for healthcare, and mainly tissue and organ replacement. Many new devices and biomaterials, less known to the public, are already in the market or likely to appear in the next few years, with increased contribution from the biological sciences, such as drug delivery systems and tissue-engineered implants [10–12]. During the past decade, we have witnessed a real revolution not only in the minds of biomaterials scientists but also with regards to the general public’s interest in this ﬁeld. Myths and prejudices are sometimes associated with the use of some of these devices and materials, sometimes hindering the development of eﬀective and safer alternatives and preventing informed consent of prospective recipients. The threat of having living cells implanted, or re-implanted, in one’s body, whether or not genetically modiﬁed, has raised unmatched public attention, along with its consequent legal and ethical issues. Mediatic legal issues related, for example, to the eventual toxicity of silicone in breast implants [13–17], of PVC in blood bags [18, 19], and of mercury in dental amalgams [20–23] have frightened companies to such an extent that silicone is no longer commercialized in some countries, PVC is to be prohibited in many others, and there is growing pressure to replace mercury in dental amalgams, even though no conclusive scientiﬁc evidence for those facts has been provided, and these examples represent huge markets [13]. Tissue engineering, which can be considered a natural evolution of biomaterials science, has become popular as today’s best way to deliver regenerative medicine, which basically consists of helping the body to heal itself by promoting regeneration through the delivery of materials, cells, and bioactive molecules. Tissue engineering was selected by

375

376

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

Time.com (from Time Magazine) as the number one of the ‘‘10 hottest jobs’’ for the 21st century [24]. Fear from the general public, health authorities and practitioners to new, not fully tested solutions raises ethical issues that immensely prolong the period between successful clinical trials and approval by health authorities. Hence, it seems wise to develop strategies to enhance public knowledge of the new developments and their real consequences, or otherwise the application of scientiﬁc progresses and many people’s lives may be at risk. It is urgent to translate from the laboratory to the clinic, and to enhance the general public’s awareness of new developments. The scientiﬁc community not only has the responsibility to fully address the scientiﬁc issues involved, but is also has the obligation to promote public awareness of new developments, as well as their associated ethical, social and economic issues. Information concerning medical devices – and namely the risks and beneﬁts associated with their use – is now available to the general public, although when Biomat.net was created only scarce, widely spread and imprecise information was available. Biomat.net has not only contributed to the enhancement of communication between researchers, but also makes available reliable information on new developments for the non-specialists.

24.3 Biomat.net as a Biomineralization Resource

Biomineralization is a key subject in biomaterials science, with special relevance on materials for orthopedic and cardiovascular applications (in the latter case, as an undesirable process). The pages of Biomat.net have been reﬂecting interest in the ﬁeld of biomineralization to the biomaterials community. Several related links are provided, notably in the terms of new developments, organizations, resources, research institutions, scientiﬁc events and books (Fig. 24.1). Biomineralization events have been endorsed by Biomat.net, namely ‘‘Biomaterials – The Next Frontiers: Biomedical, Bioelectronic, Biomineralization, Bioanalytical’’ (March 12–13, 2002, USA), ‘‘Learning From Nature How to Design New Implantable Biomaterials: From Biomineralization Fundamentals to Biomimetic Materials and Processing Routes’’ (NATO Advanced Study Institute, October 13–24, 2003, Portugal), ‘‘3rd Marie Curie Cutting Edge Conference – Biomineralisation of polymeric materials, bioactive biomaterials and biomimetic methodologies’’ (June 4–8, 2007, Portugal). Biomat.net has also provided reviews on books related to Biomineralization, namely Chitin: Fulﬁlling a Biomaterials Promise (edited by E. Khor, Elsevier), and Octacalcium Phosphate (edited by L.C. Chow and E.D. Eanes, Karger). In addition, the subject of biomineralization is a constant presence in the monthly news, which is provided through free-of-charge newsletters, highlighting new developments in the ﬁeld.

24.3 Biomat.net as a Biomineralization Resource

Fig. 24.1A Fig. 24.1 (A–F) Examples of past issues of Biomat.net, with a focus on biomineralization-related subjects on the front page, namely news and events. Events featured in the right-hand column were endorsed by Biomat.net. The ﬁgures also illustrate the evolution of Biomat.net in terms of web design.

377

378

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

Fig. 24.1B

24.3 Biomat.net as a Biomineralization Resource

Fig. 24.1C

379

380

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

Fig. 24.1D

24.3 Biomat.net as a Biomineralization Resource

Fig. 24.1E

381

382

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

Fig. 24.1F

24.4 The Biomaterials Network (Biomat.net)

24.4 The Biomaterials Network (Biomat.net) 24.4.1 An Overview

The ‘‘Biomaterials Resources on the Internet WebPage’’ was created in May 1998, and has changed considerably since its ﬁrst year of existence. It was initially aimed at gathering relevant information for biomaterials researchers, notably academic sites, societies, scientiﬁc meetings, and books. The success of this initiative encouraged its development, and in February 2000 the Biomaterials Network (Biomat.net) was published online, changing from the original collection of WWW links to an interactive source of information, where users can actively participate and communicate with each other. Biomat.net was launched as an improvement of the previous site – that is, a collection of information made available by others, but also providing networking capabilities. Biomat.net is a collection of some selected internet links related to biomaterials and tissue engineering, and also some relevant links to biomedical engineering, biology, medicine and health sciences in general. Its functionalities include a monthly newsletter, a job database, a tool designed to allow the exchange of information between researchers (with direct access to researchers’ e-mails, searchable by keywords), an updated list of relevant scientiﬁc meetings and courses, as well as updated lists of relevant internet sites organized in the following sections: Resources, Organizations, Research, Education, Journals, Books, Articles, Funding, Industry, Market, and Science Links. The site is maintained by an international Editorial Board and supervised by an international Scientiﬁc Advisory Board. Biomat.net has become a major tool in linking the biomaterials community worldwide, and has been consecutively considered the ‘‘number-one’’ online resource in this ﬁeld by distinct sources [25–33]. In August 2002, the Institute for Scientiﬁc Information (ISI) selected Biomat.net for Current Web Contents, which means that anything published on Biomat.net is now available for a worldwide audience through this major international scientiﬁc database. Membership to Biomat.net is free, and members can search all of Biomat.net without charge. Biomat.net is mainly aimed at specialists, although some information for the general public is available, such as the articles published in general-interest magazines. Future functionalities are being developed, providing information about biomaterials to the general public (see Fig. 24.2).

383

384

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

Fig. 24.2 Example of a recent issue of Biomat.net, showing its main features.

24.4.2 Objectives

Biomat.net is a non-proﬁt public service, the major goals of which include: Providing an organized and meaningful biomaterials communication resource for scientists, researchers, members of the business community, government, academia, and the general public. Acting as a resource center to disclose resources, organizations, research activity, educational initiatives, scientiﬁc events, journals, books, articles, funding opportunities, industrial developments, market analyses, jobs and every other initiative related to biomaterials science and associated ﬁelds, such as tissue engineering.

24.4 The Biomaterials Network (Biomat.net)

385

24.4.3 Team

The international team of Biomat.net, which contributes to its dissemination across the planet, is mainly constituted by biomaterials researchers, organized in an Editorial Board and a Scientiﬁc Advisory Board. Some of these groups are also involved with other websites dedicated to biomaterials, thus allowing the establishment of useful synergies and exchange of functionalities. The current team members are as follows:

Editor-in-Chief: Pedro L. Granja

(INEB, Portugal) Editor: Jose´ Paulo Pereira (INEB, Portugal) Co-editors: Ana Queiroz (INEB, Portugal) and Marcelo H. Prado da Silva (Centro Brasileiro de Pesquisas Fı´sicas – CBPF, Brazil) Site Host: INEB – Instituto de Engenharia Biome´dica, University of Porto Web Designer: Jorge Carneiro (Portugal)

Scientiﬁc Advisory Board: Arthur Brandwood (Australia) Buddy Ratner (USA) Charles Baquey (France) Clemens van Blitterswijk (The Netherlands) Ma´rio A. Barbosa (Portugal) Michael Sittinger (Germany) Michael V. Sefton (Canada) R. Geoﬀ Richards (Switzerland) Ulrich Gross (Germany) Vasif Hasirci (Turkey)

Editorial Board: News: Jose´ Paulo Pereira and

Pedro L. Granja (Portugal) Resources: Diego Mantovani

(Canada) Organizations: Dirk W. Grijpma

(The Netherlands) Research: Yunzhi Yang (China) Education: Julie Trudel (USA) Meetings: Ana Paula Filipe

(Portugal) Endorsement of events: Eduardo

A. Silva (USA) and Jose´ Paulo Pereira (Portugal) Journals: Conrado Aparicio (Spain) Books: Kanji Tsuru (Japan) Book Reviews: Ana Queiroz (INEB, Portugal) Articles: Didier Letourneur (France) ˜ a (Mexico) Funding: Cristina Pin Jobs: Rachel Williams (UK) Industry (USA): Andy Branca (USA) Industry (Europe): Joost de Bruijn (The Netherlands) Science Links: Fabio Palumbo (Italy)

386

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

24.4.4 Functionalities

Biomat.net’s functionalities include a monthly newsletter with up-to-date and relevant news, a job database, a tool allowing information exchange between researchers, as well as updated lists of biomaterials-related scientiﬁc meetings, resources, organizations, research and educational institutions, journals, books, industry, market, funding and scientiﬁc links. It also provides a monthly selection of a Top Site (Fig. 24.3).

Fig. 24.3 Monthly selection of a Top Site by Biomat.net.

24.4 The Biomaterials Network (Biomat.net)

24.4.4.1 Site Map

Biomat.net About – Mission statement What’s New – New adds to the site Search site – Search tool inside the site Feedback – Provides feedback from pageviewers Site Map – A guide through the whole site Team – Team members Networking services Join in – Membership application Researchers – Directory of members Jobs – Research and industrial job opportunities Newsletter – Monthly Newsletter

WWW links Resources – WWW biomaterials resources Organizations – Societies and relevant

institutions Research – Research institutions Education – Educational institutions Meetings – Conferences, courses, symposia Journals – Scientiﬁc journals Books – Relevant literature Articles – Related articles of general interest Funding – Funding opportunities Industry – Industrial organizations Market – Market analyzes Science links – Relevant non-biomaterials websites Top Site – Selection of top site

24.4.4.2 Membership Biomat.net is a non-proﬁt initiative. Therefore, joining Biomat.net is absolutely free of charge. However, only members are entitled to: receive updated information directly to their mailbox; browse the Directory of Researchers; and browse Job Oﬀers or Jobs Wanted.

Biomat.net presently accounts for around 5500 members from over 100 countries, and almost 1000 institutions. Pageviewers come from over 110 countries. Over 74% of them access Biomat.net from Europe and North America. Some 60% of the members are from academic, 18% from industrial, 6% from governmental institutions, 4% from hospitals, 1% from publishers, and 11% are from other non-speciﬁed activities. 24.4.4.3 Links Lists New relevant websites are added monthly to all Biomat.net’s links sections. Almost 7000 links with relevance to biomaterials science are presently listed in Biomat.net’s pages (see Fig. 24.4). 24.4.4.4 Directory of Researchers Biomat.net’s Directory of Researchers is a database of scientists and researchers, in addition to members of the business community, government and academia having an activity relevant to the biomaterials ﬁeld. It contains useful authorized information about each member, including contact, research interest and technical skills.

387

388

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials

Fig. 24.4 Example of a section of Biomat.net. The books section also includes reviews of recent books related to biomaterials and biomineralization. Books are organized by year and by title.

24.4.4.5 Jobs Job Oﬀers and Jobs Wanted are permanently being updated by the pageviewers themselves, in order to allow daily update of information. Feedback from members that have already used this functionality showed that several candidates and institutions worldwide found their jobs or recruited people through Biomat.net. Over 3500 jobs have already been announced through it pages. 24.4.4.6 Newsletter Seventy issues of Biomat.net’s monthly newsletter have been published between February 2000 and June 2006. Although its structure has been evolving over time, the newsletter’s main features remain unchanged, namely: Highlights, by the

References

Editor-in-Chief; New Developments on Biomat.net; New Meetings Endorsed by Biomat.net; Developments on Endorsed Meetings; The current month’s TOP Site; Special announcements; New Articles on European Cells and Materials e-journal (partner of Biomat.net); Biomaterials World News (tens of news every month on biomaterials, medicine, materials and science in general); and Biomaterials Articles, containing biomaterials-related articles, usually published in general-interest magazines. This newsletter is published in many other related sites and is oﬃcially distributed by some international organizations, making it impossible to eﬀectively assess its real impact. 24.4.4.7 Endorsement of Scientiﬁc Meetings Biomat.net has two ways of advertising scientiﬁc meetings: (i) they can be included in the meetings list; or (ii) they can be endorsed by Biomat.net, which means that they will be disclosed every month (until the date of the meeting) in the newsletter and front page. In order to be endorsed, an agreement is established between Biomat.net and the meeting’s organization. Biomat.net is a non-proﬁt initiative, so meetings are not sponsored ﬁnancially. The agreement is based in a simple exchange, which has proved to be fruitful for both parties: the organization recognizes the endorsement and Biomat.net discloses the event. Biomat.net also provides a website for the event, in case it is necessary. Biomat.net has already endorsed more than 315 international meetings, mainly through the dissemination of information.

References 1 N. Sparrow, The wild wild web, Eur.

2

3

4

5

Med. Device Manufacturer, November 2000. www.devicelink.com/emdm. L. Baker, T.H. Wagner, S. Singer, M.K. Bundorf, Use of the internet and e-mail for health care information: Results from a national survey, JAMA 2003, 289, 2400. A.T. McCray, Promoting health literacy, J. Am. Med. Inform. Assoc. 2005, 12, 152. B.W. Hesse, D.E. Nelson, G.L. Kreps, R.T. Croyle, N.K. Arora, B.K. Rimer, K. Viswanath, Trust and sources of health information – The impact of the internet and its implications for health care providers: Findings from the ﬁrst health information national trends survey, Arch. Intern. Med. 2005, 165, 2618. T. Lewis, Seeking health information on the internet: Lifestyle choice or

6

7

8

9

10

bad attack of cyberchondria?, Media, Cult. Soc. 2006, 28, 521. Health care information on the internet, The British Library, www.bl.uk/collections/health/ weblink.html. G. Freiherr, Electronic commerce: New Technologies are transforming business, Med. Device Diagn. Ind., May 1998. www.devicelink.com. G. Nighswonger, Exploring the internet: New tools and resources for healthcare, Med. Device Diagn. Ind., June 2000. www.devicelink.com. T. Frank, What the internet means for the medical device industry, Med. Device Technol., December 2000. www.medicaldevicesonline.com. L.L. Hench, Bioactive materials: The potential for tissue regeneration, J. Biomed. Mater. Res. 1998, 41, 511.

389

390

24 The Biomaterials Network (Biomat.net) as a Major Internet Resource for Biomaterials 11 C.W. Patrick, Jr., A.G. Mikos, L.V.

12

13

14

15

16

17

18

19

20

McIntire (Eds.) in: Frontiers in Tissue Engineering. Elsevier, Oxford, 1998, p. 3. L.G. Griﬃth, G. Naughton, Tissue engineering – current challenges and expanding opportunities, Science 2002, 295, 1009. S. Cooper, The responsibilities of a biomaterials scientist and citizen, Biomater. Forum 1997, 19, 4. J. Kohn, Current trends in the development of synthetic materials for medical applications, Med. Device Technol., November 1990. www.medicaldevicesonline.com. D.J. Lyman, in: A.F. von Recum (Ed.), Handbook of Biomaterials Evaluation: Scientiﬁc, Technical and Clinical Testing of Implant Materials, 2nd edn. Edwards Brothers, Ann Arbor, 1999, p. 37. R.G. Meeks, The Dow Corning siloxane research program: An overview and update, Med. Device Diagn. Ind., May 1999. www.devicelink.com. S.C. Gad, Safety Evaluation of Medical Devices. Marcel Dekker, New York, USA, 1997. B. Lipsitt, Metallocene polyethylene ﬁlms as alternatives to ﬂexible PVC ﬁlm for medical device fabrication, Proceedings of the Society for Plastics Engineers (SPE) Annual Technical Conference, Brookﬁeld, USA, 1997, p. 2854. S. Shang, L. Woo, Selecting materials for medical products: From PVC to metallocene polyoleﬁns, Med. Device Diagn. Ind., October 1996. www.devicelink.com. G. Balfry, A mouthful of trouble, Br. Dent. J. 2005, 198, 321.

21 R.T. Kao, S. Dault, T. Pichay,

22

23

24 25

26

27

28 29

30

31 32

33

Understanding the mercury reduction issue: the impact of mercury on the environment and human health, J. Calif. Dent. Assoc. 2004, 32, 574. P. Horsted-Bindslev, Amalgam toxicity-environmental and occupational hazards, J. Dent. 2004, 32, 359. H.K. Yip, D.K. Li, D.C. Yau, Dental amalgam and human health, Int. Dent. J. 2003, 53, 464. Visions of the 21st Century, Time.com, May 2000. www.time.com. G. Spera, Web site facilitates access to biomaterials data, Med. Device Diagn. Ind., September 1998. www.devicelink.com. New site links international biomaterials community, Eur. Med. Device Manufacturer, May/June 2000. Revamped Biomaterials Site Oﬀers Links, Articles, Information, Med. Device Link, February 2000. Making Body Parts (www.biomat.net), Science 2000, 289, 2235. P.M. Gannon, Where materials meet biology, HMS Beagle (BioMedNet), July 2001. www.bmn.com. Biomaterials – The technology of the future, CMBO Perspective, Cell & Molecular Biology Online, October 2001. Connecting with biomaterials, NetWatch, Science, December, 2001. E. Ba¨uerlein, The world of bodybuilding materials, Angew. Chem. Int. Ed. 2002, 41, 1805. ‘‘Pearl of the Web’’. www.swissbiomat.ch, Dental Biomaterials, 2004, 1 (available online).

391

Index a acellular extrinsic ﬁber cementum (AEFC) 172 f. acellular intrinsic ﬁber cementum (AIFC) 172, 174 advanced photon source (APS) 49 ﬀ. allografts 243 ﬀ. alloplasts 245 f. alpha2 -HS glycoprotein 310 ﬀ. alveolar bone 204 ﬀ. ameloblast 164, 170 f., 206 f., 209, 270 amelogenesis 170 ﬀ. amelogenin 171, 207, 249 anchor 117, 121 animal experiments 129 f., 153 f. anisotropy 192 antigenicity 247 apatite 104, 152 f., 191, 354 – composition 103 – structure 103 apatite formation 98 ﬀ., 159 – materials 102 – mechanisms 105 apatite-forming ability 98, 100 apoptosis 114, 303, 307, 311, 333 apposition 203 ﬀ. arteriosclerosis 285, 291 ﬀ. – cellular aspects 301 ﬀ. arteriovenous (AV) shunt 296 ﬀ. artherosclerotic plaques 294, 301, 304, 308 ﬀ. atomic force microscopy (AFM) 177 ﬀ. autografts 243 azimuthal orientation 53

b biliary stones 343 bioactive materials 97 – mechanism of bonding to bone 104 bioactivity 98

biocompatibility 87, 127 f., 137, 142, 246, 271 – eﬀect of particle size 130 biodegradability 83 biodegradation 85, 90, 252 biologic modiﬁers 239 biomaterial 373 ﬀ. biomaterials network 373 ﬀ. – functionalities 386 – goals 384 biomechanical compatibility 41 biomechanical equilibrium 36 ﬀ. biomineralization 82, 145 f., 151, 153, 203 f., 213 f., 216, 219 f., 317 ﬀ., 373 ﬀ. biophysical stimulation 147 bioreactor 150 bone 35 f., 60, 81 ﬀ., 170, 197, 216, 220 – abnormalities 59 – assembly 56 – biomechanics 35 ﬀ. – biomineralization 151 – cell types 152 – defects 81 – development processes 152 – hierarchical architecture 42, 46 – loaded bones 49 ﬀ. – mineralization 3 ﬀ., 13 f., 317 – remodeling behavior 41 bone-bonding ability 97 f., 102 bone-bonding bioactivity 98 ﬀ., 106 bone cements 84 ﬀ. bone formation 19 ﬀ., 30, 60, 65, 74, 101, 120, 249, 285 f., 292 – induction 115 – principles 152 bone grafting 115 f., 242, 253 bone growth 38 – stimulation 109 ﬀ. bone-like apatite 97 bone marrow 20, 148, 268, 286

Handbook of Biomineralization. Edited by M. Epple and E. Ba¨uerlein Copyright 8 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31806-3

392

Index bone mass 40, 61 bone mass density 37 f. – distribution 38, 40 – stress-driven evolution 37 bone mineral density (BMD) 62, 67 bone morphogenetic protein (BMP) 19 ﬀ., 65, 151, 165, 168, 218, 244, 248, 265 f., 269 ﬀ., 306 – activity 30 – basic signaling mechanism 24 – inhibitor 31 – isoforms 20 bone morphogenetic protein (BMP) receptor 23 ﬀ. – aﬃnity 27 ﬀ. – interaction 27 ﬀ. – speciﬁcity 27 ﬀ. bone morphogenetic protein (BMP) signaling 20, 25, 31 bone remodeling 60, 66, 213, 357 – computation 35 ﬀ. – modeling 35 ﬀ. – simulation 39 bone repair 19 f. bone substitutes 97 bone substitution materials – biological origin 87 ﬀ. – clinical need 81 f. – requirements 82 f. – synthetic biomimetic bone substitution material 90 f. – synthetic materials 82 ﬀ. bone tissue engineering 145 ﬀ. – aim 146 – clinical demands 153 f. – components 147 ﬀ. broncholithiasis 345

c calciﬁcation 3, 7, 285 ﬀ., 301 ﬀ., 317 ﬀ., 361 – arterial 291 ﬀ., 296 – classiﬁcation 293 – hypotheses 364 – imaging methods 365 – in ulcera of patients with paraplegia 288 f. – inhibition 322 – metastatic pulmonary calciﬁcation 291 – of synthetic vascular grafts 296 ﬀ. – of the tunica intima (arteriosclerosis) 293

– of the tunica media (Mo¨nckeberg’s arteriosclerosis) 292 – patterns 367 f. – prevention 324 – pulmonary 289 ﬀ. – regulation 287 calciﬁed cartilage 60 f. calcifying vascular cells (CVCs) 309 calciprotein particle (CPP) 322 calcium 4, 9, 74, 102, 166, 204, 207 ﬀ. 213, 285 f., 294, 298, 301, 305 ﬀ. 310, 319 f., 353 f. calcium carbonate 245 calcium hydroxide 271 calcium oxalate 331 ﬀ. calcium phosphate 84 f., 87, 89, 181, 224, 286, 292, 331 calcium stones 333 ﬀ. calibration 52 carbonated apatite (cAp) 50 f., 53 ﬀ., 85 cardiovascular disease 357 f. caries detection 223, 232, 236 caries prevention 223 f. cell adhesion 109 ﬀ., 118 – receptor 110 cell compatibility 149 cell proliferation 114, 118 cell stimulation 145 – biochemical stimulation 151 – biophysical stimulation 150 f. cell survival rate 127, 129, 131, 142 cell therapy 270, 273 ﬀ. cells 301 ﬀ. cellular intrinsic ﬁber cementum (CIFC) 172 ﬀ. cellular mixed stratiﬁed cementum (CMSC) 172, 174 cementoblast 172, 220 cementoclast 217 cementocyte 172, 174 cementogenesis 172 ﬀ., 213, 248 cementoid 173 cementum 159 f., 163 f., 172, 186, 196 f., 204 ﬀ., 213 ﬀ., 216, 220, 225, 242 ceramics 84 ﬀ., 110 chemical force microscopy (CFM) 178 f., 181 chondrocyte 4, 303 citrate 331 coating molecules – structure 117 collagen 42 ﬀ, 47, 50 f., 55, 64, 87 ﬀ., 104, 120, 153, 159, 167, 207, 225, 242, 271, 302, 319

Index compartmentalization 241 computational technique 40, 46 computed tomography (CT) 36, 38 f., 46 cortical bone 36, 42 – meso-scale model 45 f. – structure 42 cross-transplantation experiments 11 crown 183 ﬀ., 225 crown dentin 196 – microstructure 194 – structural material 194 crystal growth 177 ﬀ. cystine stones 338 cytokines 116, 127, 129, 149, 151, 248 f. cytotoxicity 127 f., 137 – particle size-dependence 140 f.

d demineralization 203 f., 212 f., 223 ﬀ ., 231 ﬀ., 242 demineralized freeze-dried bone allograft (DFDBA) 244 dendritic cells 312 dental caries 203 f., 212 ﬀ., 218, 223 ﬀ., 265 f. – mineral changes 225 ﬀ. dental hard tissue 223, 226 dental lamina 207 dental papilla 165 f. dental plaque 224 dental pulp stem/progenitor cells 266 dental trauma 204 – acute dental trauma 220 – chronic dental trauma 220 f. dentin 84, 159 f., 163 ﬀ., 170 ﬀ., 183 ﬀ., 190 ﬀ., 204 ﬀ., 213 ﬀ., 217, 220, 224 ﬀ., 232, 266 – circumpulpal dentin 166 – classiﬁcation 166 – composition 159 – formation 167 ﬀ. – intertubular dentin 168, 197 – mantle dentin 166 ﬀ. – peritubular dentin 168 f. – viscoelastic properties 187 dentin caries 224 ﬀ. dentin matrix protein 1 (Dmp1) 8 f., 14 dentin regeneration 265 ﬀ., 269, 271 ﬀ. dentino-enamel junction (DEJ) interface 184, 186, 194, 198, 200 dentinogenesis 165 ﬀ., 271 detector 51 f., 55, 62 developmental anomalies 204, 210 ﬀ.

diﬀraction patterns 53 dissociation constant 26 f.

e ectopic calciﬁcation 3 f., 6, 10, 14, 317 ﬀ., 349 f., 362 – inhibitor 320 ﬀ. ectopic expression 11, 14 ectopic mineralization 11, 317, 349 ﬀ. – regulators 350 ﬀ. electronic speckle pattern-correlation interferometry (ESPI) 183, 188, 191, 193, 198 enamel 159 ﬀ., 163 ﬀ., 170 ﬀ., 182 ﬀ., 204, 207, 213 ﬀ., 217, 224 f., 228 f., 232 – composition 159, 171 – deformation 194 – formation 177 – maturation 171 – properties 191 – stages of production 209 – structure 192 enamel cap 183 ﬀ., 190 f., 198 – deformation 185 f., 191 – hardness 191 – mechanical behavior 191 ﬀ. – microstructure 193 enamel caries 224 f. enamel crystal 177 ﬀ. enamel hypoplasia 210 enamel matrix derivative (EMD) 249 enamelin 207 enameloid 204 endopeptidase 11 ﬀ. endoprosthetics 35 energy-dispersive x-ray analysis (EDX) 62, 365 evolution 203 extracellular matrix (ECM) 3 ﬀ., 30, 109, 113, 145, 150., 168, 266, 304, 319 f. – ECM proteins 14 – mineralization 4 ﬀ., 9 f., 12, 14 – principles of ECM biomineralization 151 f.

f fetuin 317 ﬀ. ﬁber 44, 159 f., 172, 196 f., 207, 225 ﬁbril 44, 50, 55, 64, 167 ﬁbroblast 173, 249, 266, 286 ﬁbronectin 332 ﬁbrosis 19 ﬁnite element method (FEM) 37

393

394

Index – integration 118 – shape 83 – surface 116 implant technology – biomimetic materials 109 f. implantation 127 f., 134, 140, 146, 149, 269 inﬂammation 127 f., 133, 137 f., 216, 285 ﬀ., 305 inﬂammatory cells 310 ﬀ. g inhibitor 4 ﬀ., 14, 21, 30 gallbladder stone 343 integrin 109 ﬀ. gap junction 287 f. – antagonists 112 gene therapy 151, 153, 265 integrin ligand 118, 122 – ex-vivo BMP cell therapy 273 ﬀ. interleukin 1b (IL-1b) 129, 132, 135, 137 – in-vivo BMP gene therapy 272 f. internet 373 ﬀ. gene transduction 273 intrabony defects 253 geometry 51 f. intrinsic calciﬁcation 362, 367, 369 gingivitis 216 glial-derived neurotrophic factor (GDNF) 20 in-vitro 223, 225, 228, 231 in-vitro calciﬁcation model 364 growth and diﬀerentiation factor (GDF) in-vivo 223, 230 f. 19 ﬀ., 31 iron 128, 134 ﬀ., 141 f. growth factor 147, 149, 248 f. guided tissue regeneration (GTR) 239 ﬀ., 246 ﬀ. k – contraindications 250 knockout mice 321, 324 – factors inﬂuencing GTR success 249 ﬀ. l – indications 250 lactate dehydrogenase (LDH) 127, 132, 135 lithiasis 329 ﬀ. ﬁnite element techniques 38 f., 43, 45 Fourier transform infra-red (FTIR) microscopy 69 fracture 31, 60 freeze-dried bone allograft (FDBA) – mineralized cortical FDBA 244 furcation defects 253

h hard tissue apposition 204 hard tissue resorption 204, 220 healthcare 375 heart valve bioprostheses 361 ﬀ. heterotopic ossiﬁcation (HO) 288 ﬀ. high-resolution electron microscopy (HREM) 177, 179 high-resolution transmission electron microscopy (HRTEM) 177 ﬀ. hip-joint endoprosthetics 38, 40, 46, 103 human tooth function 183 hydroxyapatite 5, 8 f., 42 ﬀ., 64, 84, 89, 105, 116, 153, 159 f., 166, 168, 178, 181, 203, 206 ﬀ., 221, 224, 245 f. hypercalcemia 335 hypercalciuria 334 f., 341 hyperoxaluria 335 f. hyperphosphatemia 287, 350

i imaging 361, 365, 370 implant 35, 46, 81 ﬀ., 87 ﬀ., 109 ﬀ., 115 ﬀ., 120 ﬀ., 130, 277 – bioactivity 97 ﬀ. – coating 119

m macrophage 133, 136, 301 f., 305, 310 ﬀ., 324 mast cells 312 matrix g-carboxyglutamate (Gla) protein (Mgp) 5 f., 8 f., 14 matrix vesicles 303 ﬀ., 350, 357 mechanical stability 83 mechanosensation 36, 42, 46 mechanotransduction 36, 42, 46 membranes 255 – absorbable 246 ﬀ., 252 – non-absorbable 246 ﬀ. mesenchymal stem cells (MSC) 148 meso-scale model 45 f. metals 83, 87, 105, 110 metastatic calciﬁcation 287 microcomputed tomography 62, 321 microparticle 127 ﬀ. microradiography 223, 226, 361 micro-scale model 45 mineralization 36, 60, 64, 115, 159, 164 ﬀ., 216 – bone 3 ﬀ. – inhibitor 14

Index minhibin 12 miscellaneous stones 344 f. Mori-Tanaka technique 44 f. morphogenetic signals 269 f. multipotential diﬀerentiation 269 multi-scale methods 42 f. mutants 20, 31

n nanoparticle 127 ﬀ., 139 f. nanotoxicology 139 f., 143 necrosis 136, 357 neurogenesis 276 f. neutrophils 127 ﬀ., 131 ﬀ. nickel 128, 132 ﬀ., 141 f.

o odontoblast 4, 6, 160 f., 164 ﬀ., 173, 206 f., 213, 220, 266, 270 odontogenesis 163 ﬀ., 203 f. – involved genes 165 – stem cells 165 osseointegration 41, 109, 112, 115 f., 119 f., 122 ossiﬁcation 152, 285, 289 ﬀ. – of synthetic grafts 298 – ossiﬁcation of arteries 294 osteoarthritis 19, 31 osteoblast 3 ﬀ., 8 ﬀ., 59 f., 67, 70 f., 74, 82, 85, 104, 109, 149, 153, 220, 249, 285, 287 f., 303, 309 – adhesion and proliferation on implants 118 ﬀ. – maturation stages 147 f. osteocalcin 5 f., 8, 289 osteoclast 59 ﬀ., 67 ﬀ., 70 ﬀ., 82, 85, 172, 216 f., 285, 312 – diﬀerentiation 67 osteoclastogenesis 72 osteoconduction 20, 115, 242 osteocytes 42, 45, 47, 64, 153 osteodentin 271, 273 osteogenesis 115, 242, 248, 308 f., 317 ﬀ. osteoid 6, 74, 153 osteoidosis 7, 9, 12 osteoinduction 20, 115, 242 osteon 36, 42 ﬀ. – development and mineralization 44 – models 43 osteoneogenesis 289 osteopetrosis 59 ﬀ. – animal models 63 ﬀ., 70 – associated genes 72 – biomineralization 74

– bone mineral properties 62 f. – cellular and molecular bases 72 ﬀ. – clinical features 60 f. – diagnosis 62 – histology 60 – other osteopetrotic models 68 ﬀ. – radiography 61 f. – rodent models 67 osteopontin 8 f., 287, 311, 324, 332, 349, 354 osteoporosis 19, 31, 59 ﬀ., 287, 357 – animal models 63 ﬀ., 70 – biomineralization 74 – bone mineral properties 62 f. – cellular and molecular bases 70 ﬀ. – clinical features 60 f. – diagnosis 62 – histology 60 – non-rodent models 66 – radiography 61 f. – rodent models 63 ﬀ. osteoprogenitor cells 115 ovariectomy 63 f., 66, 70

p pancreatic stones 345 parodontosis 19 particle size 131 ﬀ., 137 ﬀ. – material dependence 135 – toxicity level 138 f. pathological calciﬁcation 283 ﬀ., 361 ﬀ. – examples 286 f. pathological mineralization 3 ﬀ., 9 f., 14, 317 pep-gen p-15 249 peptide growth factor 116 pericytes 266 f., 286, 289, 309 periodontal diseases 203 f., 216 ﬀ., 248, 251 periodontal ligament 216 periodontal regeneration 239 ﬀ. periodontitis 203 f., 216 f., 242 periodontogenesis 203 f. phagocytosis 127, 132, 136, 141, 305, 310, 324 phenomenology 36, 40 Phex 7, 11 ﬀ. phosphate 4, 9, 11, 74, 102, 166, 204, 209, 213, 305, 349, 353 phosphatonin 11 f. phosphophoryn 167, 169 phosphorus 285 f., 294, 298 physiological mineralization 3 f., 9 ﬀ., 14 – activators 6 ﬀ., 14

395

396

Index pit-1 354 polyester-based materials 86 polymers 86 f., 110 polymethylmethacrylate (PMMA)-based materials 86, 118 postoperative care 252, 255 post-translational modiﬁcation 5 precaries 213 predentin 204, 209 protein therapy 271 f. pulmonary alveolar microlithiasis 345 pulmonary calciﬁcation 285 pulp 159 f., 165, 169, 195, 199 pulp exposure 265 f., 272 ﬀ. pulp regeneration 265 f., 269, 271, 276 f. pulp tissue 266 pulpitis 273 f. pyrophosphate 4, 9 ﬀ., 14

q quantitative light-induced ﬂuorescence (QLF) 223, 226, 229 ﬀ. – in-vitro studies 231 ﬀ. – in-vivo studies 234 ﬀ.

r radius 53 Raman spectroscopy 64 receptor speciﬁcity – biochemistry 25 – cell biology 25 regeneration 239 ﬀ., 255 – techniques 241 ﬀ. relative displacement magnitudes (RDM) 187 ﬀ. remineralization 203 f., 212 f., 223 f., 228, 231 f. resorption 203 f. resource 373 ﬀ. RGD (Arg-Gly-Asp) peptides 109 ﬀ., 121 ﬀ. root 196, 225 root caries 232 f. root coverage 253 root dentin 196 f., 228 root surface 240 ﬀ.

s scaﬀold 265 f., 354 – functions 270 scaﬀold design 149 scanning electron microscopy (SEM) 129, 132, 136, 194 ﬀ., 215, 365 scattering pattern analysis 53 sialolithiasis 344

silk 120 simulated body ﬂuid (SBF) 97 ﬀ. – ion concentrations 101 small angle x-ray scattering (SAXS) 49 ﬀ. sodium-dependent phosphate co-transporter 353 stone formation 329 ﬀ. strain 51, 55, 185 – measurement 49 ﬀ. strain-stress conversion 55 stress 51, 55, 187, 192, 197 – deviatoric stress 50 – hydrostatic stress 50 – measurement of internal stresses 49 ﬀ. stress gradient 56 stress-adaptive bone remodeling 36, 46 struvite stones 337 f. superoxide 127, 129, 132, 141 supragingival stones 344 f. surface coating 114 surgery 81, 83, 91, 146 surgical principles 253 ﬀ. synchrotron radiation 49, 367 synthetic graft material 285 f. synthetic materials – biological functionalization 90

t t lymphocytes 312 testicular microlithiasis 343 TGF-b (transforming growth factor beta)superfamily 19 ﬀ., 25, 29, 151, 248, 324 – subfamilies 21 f. tissue engineering 145 f., 265 ﬀ., 373 ﬀ. titanium 120 ﬀ., 127 ﬀ. tooth 157 ﬀ., 183 ﬀ., 265 ﬀ. – cap 190, 197 – crown 185, 191 f., 195, 197 – decay 223 – deformation 184 ﬀ., 187, 190, 199 – development 159, 162 f., 170, 206, 208 – developmental anomalies 206, 210 ﬀ. – disease 203 ﬀ. – evolutionary origin 203 – formation 159 ﬀ. – function 183 f. – malformation 211 – microstructure 187 – nomenclature 186 – structure 177 ﬀ., 183 f., 187 – tissue engineering 265 ﬀ.

Index tooth mineralization – developmental features 207 ﬀ. – elemental analysis 207 ﬀ. transmission electron microscopy (TEM) 99, 103 f., 288, 295, 322 transplant 217 – allogenic transplant 82 – xenogenic transplant 82 transverse microradiography (TMR) 226 ﬀ. trauma 285 ﬀ. tumor 140, 142, 285 ﬀ. tumor necrosis factor alpha (TNF-a) 129, 132, 134 f., 137 ﬀ. type I collagen 8, 10, 65, 71, 166, 168, 207, 225, 247, 270

u uric acid stones 335 f., 341 urinary stones 329 ﬀ. – classiﬁcation 333 ﬀ. – genetic risk factors 341 f. – inhibitors of stone formation 331 – non-genetic risk factors 338 ﬀ. – pathogenesis 330 ﬀ. urolithiasis 330 ﬀ.

v variants 19 f., 27, 29 ﬀ. vascular calciﬁcation 285, 291, 302 ﬀ., 349, 358 – inhibitors 292 vascular cells 149 vascular smooth muscle cells (VSMCs) 301 ﬀ., 319 vascularization 149, 153 vasculogenesis 276 vesicle 166 ﬀ., 207, 303, 313

w whole crown deformation 185 ﬀ. whole-teeth regeneration 277 wide angle x-ray scattering (WAXS) 49 ﬀ. wound healing 240 f.

x xenografts 245, 361 x-radiograph 41 x-ray diﬀraction 59, 62, 66 x-ray scattering 49 ﬀ. x-ray structure 29 ﬀ.

397