PLATELETS IN HEMATOLOGIC AND CARDIOVASCULAR DISORDERS…the-eye.eu/public/WorldTracker.org/College Books... · ment in hematologic, cardiovascular, and inﬂamma-tory disorders and

http://www.cambridge.org/9780521881159

PLATELETS INHEMATOLOGIC ANDCARDIOVASCULAR

DISORDERSA Clinical HandbookEdited by

Paolo GreseleUniversity of Perugia, Italy

Valentin FusterMount Sinai School of Medicine, USA

José A. LópezPuget Sound Blood Center and University of Washington, USA

Clive P. PageKing’s College London, UK

Jos VermylenUniversity of Leuven, Belgium

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University PressThe Edinburgh Building, Cambridge CB2 8RU, UK

First published in print format

ISBN-13 978-0-521-88115-9

ISBN-13 978-0-511-37913-0

© Cambridge University Press 2008

Every effort has been made in preparing this book to provide accurate and up-to-date information that is in accord with accepted standards and practice at the time of publication. Nevertheless, the authors, editors and publisher can make no warranties that the information contained herein is totally free fromerror, not least because clinical standards are constantly changing through research and regulation. The authors, editors and publisher therefore disclaimall liability for direct or consequential damages resulting from the use of material contained in this book. Readers are strongly advised to pay carefulattention to information provided by the manufacturer of any drugs or equipment that they plan to use.

2007

Information on this title: www.cambridge.org/9780521881159

This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press.

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

eBook (NetLibrary)

hardback

http://www.cambridge.org/9780521881159http://www.cambridge.org

CONTENTS

List of contributors page vPreface ixGlossary xi

1 The structure and production of bloodplatelets 1Joseph E. Italiano, Jr.

2 Platelet immunology: structure, functions,and polymorphisms of membraneglycoproteins 21Yasuo Ikeda, Yumiko Matsubara, and Tetsuji Kamata

3 Mechanisms of platelet activation 37Lawrence Brass and Timothy J. Stalker

4 Platelet priming 53Paolo Gresele, Emanuela Falcinelli, and Stefania Momi

5 Platelets and coagulation 79José A. López and Ian del Conde

6 Vessel wall-derived substances affectingplatelets 92Azad Raiesdana and Joseph Loscalzo

7 Platelet–leukocyte–endotheliumcross talk 106Kevin J. Croce, Masashi Sakuma, and Daniel I. Simon

8 Laboratory investigation of platelets 124Eduard Shantsila, Timothy Watson,

and Gregory Y.H. Lip

9 Clinical approach to the bleeding patient 147Jos Vermylen and Kathelijne Peerlinck

10 Thrombocytopenia 155James B. Bussel and Andrea Primiani

11 Reactive and clonal thrombocytosis 186Ayalew Tefferi

12 Congenital disorders of plateletfunction 201Marco Cattaneo

13 Acquired disorders of platelet function 225Michael H. Kroll and Amy A. Hassan

14 Platelet transfusion therapy 242Sherrill J. Slichter and Ronald G. Strauss

15 Clinical approach to the patient withthrombosis 261Brian G. Choi and Valentin Fuster

16 Pathophysiology of arterial thrombosis 279Juan José Badimon, Borja Ibanez, and Gemma Vilahur

17 Platelets and atherosclerosis 293Stephan Lindemann and Meinrad Gawaz

18 Platelets in other thrombotic conditions 308David L. Green, Peter W. Marks, and Simon Karpatkin

19 Platelets in respiratory disorders andinflammatory conditions 323Paolo Gresele, Stefania Momi, Simon C. Pitchford,

and Clive P. Page

20 Platelet pharmacology 341Dermot Cox

21 Antiplatelet therapy versus otherantithrombotic strategies 367Nicolai Mejevoi, Catalin Boiangiu, and Marc Cohen

22 Laboratory monitoring of antiplatelettherapy 386Paul Harrison and David Keeling

iii

Contents

23 Antiplatelet therapies in cardiology 407Pierluigi Tricoci and Robert A. Harrington

24 Antithrombotic therapy incerebrovascular disease 437James Castle and Gregory W. Albers

25 Antiplatelet treatment in peripheralarterial disease 458Raymond Verhaeghe and Peter Verhamme

26 Antiplatelet treatment of venousthromboembolism 471Menno V. Huisman, Jaapjan D. Snoep,

Jouke T. Tamsma, and Marcel M.C. Hovens

Index 483

iv

CONTRIBUTORS

Gregory W. Albers, MDDepartment of Neurology and

Neurological SciencesStanford Stroke CenterStanford University Medical CenterStanford, CA, USA

Juan José Badimon, PhD, FACC, FAHACardiovascular InstituteMount Sinai School of MedicineNew York, NY, USA

Catalin Boiangiu, MDDivision of CardiologyNewark Beth Israel Medical CenterNewark, NJ, USA

Lawrence Brass, MD, PhDUniversity of PennsylvaniaPhiladelphia, PA, USA

James Bussel, MDDepartment of Pediatrics, and

Department of Obstetrics and GynecologyWeill Medical College of Cornell

UniversityNew York, NY, USA

James Castle, MDDepartment of Neurology and

Neurological SciencesStanford University Medical CenterStanford, CA, USA

Marco Cattaneo, MDUnità di Ematologia e TrombosiOspedale San PaoloDipartimento di Medicina,

Chirurgia e OdontoiatriaUniversità di MilanoMilano, Italy

Brian G. Choi, MD, MBAZena and Michael A. Wiener

Cardiovascular InstituteMount Sinai School of MedicineNew York, NY, USA

Marc Cohen, MD, FACCDivision of CardiologyNewark Beth Israel Medical Center,and Mount Sinai School of MedicineNew York, NY, USA

Dermot Cox, BSc, PhDMolecular and Cellular TherapeuticsRoyal College of Surgeons in IrelandDublin, Ireland

Kevin J. Croce, MD, PhDDepartment of MedicineCardiovascular DivisionBrigham and Women’s HospitalHarvard Medical SchoolBoston, MA, USA

v

Contributors

Ian del Conde, MDDepartment of Internal MedicineBrigham and Women’s HospitalBoston, MA, USA

Emanuela Falcinelli, PhDDivision of Internal and

Cardiovascular MedicineDepartment of Internal MedicineUniversity of PerugiaPerugia, Italy

Valentin Fuster, MD, PhDZena and Michael A. Wiener

Cardiovascular InstituteMount Sinai School of MedicineNew York, NY, USA

Meinrad Gawaz, MDCardiology and Cardiovascular DiseasesMedizinische Klinik IIIEberhard Karls-Universität TübingenTübingen, Germany

David L. Green, MD, PhDDepartment of Medicine (Hematology)New York University School of MedicineNew York, NY, USA

Paolo Gresele, MD, PhDDivision of Internal and


Robert A. Harrington, MDDuke Clinical Research InstituteDuke University Medical CenterDurham, NC, USA

Paul Harrison, PhD, MRCPathOxford Haemophilia and

Thrombosis CentreChurchill HospitalOxford, UK

Amy A. Hassan, MDMD Anderson Cancer CenterUniversity of TexasHouston, TX, USA

Marcel M. C. Hovens, MDSection of Vascular MedicineDepartment of General Internal

Medicine–EndocrinologyLeiden University Medical CentreLeiden, The Netherlands

Menno V. Huisman, MD, PhDSection of Vascular MedicineDepartment of General Internal


Borja Ibanez, MDCardiovascular InstituteMount Sinai School of MedicineNew York, NY, USA

Yasuo Ikeda, MDDepartment of HematologyKeio University School of MedicineTokyo, Japan

Joseph E. Italiano, Jr, MDHematology DivisionBrigham and Women’s Hospital, andVascular Biology ProgramChildren’s Hospital Boston, andHarvard Medical SchoolBoston, MA, USA

Tetsuji Kamata, MDDepartment of AnatomyKeio University School of MedicineTokyo, Japan

Simon Karpatkin, MDDepartment of Medicine (Hematology)New York University School of MedicineNew York, NY, USA

vi

Contributors

David Keeling, BSc, MD, FRCP, FRCPathOxford Haemophilia and Thrombosis CentreChurchill HospitalOxford, UK

Michael H. Kroll, MDMichael E DeBakey VA Medical Center,and Baylor College of MedicineHouston, TX, USA

Stephan Lindemann, MDCardiology and Cardiovascular DiseasesMedizinische Klinik IIIEberhard Karls-Universität TübingenTübingen, Germany

Gregory Y.H. Lip, MD, FRCPHaemostasis Thrombosis and

Vascular Biology UnitUniversity Department of MedicineCity HospitalBirmingham, UK

José A. López, MDPuget Sound Blood Center, andUniversity of WashingtonSeattle, WA, USA

Joseph Loscalzo, MD, PhDDepartment of MedicineCardiovascular DivisionBrigham and Women’s Hospital, andHarvard Medical SchoolBoston, MA, USA

Peter W. Marks, MDYale University School of MedicineNew Haven, CT, USA

Yumiko Matsubara, PhDDepartment of HematologyKeio University School of MedicineTokyo, Japan

Nicolai Mejevoi, MD, PhDDivision of CardiologyNewark Beth Israel Medical CenterNewark, NJ, USA

Stefania Momi, PhDDivision of Internal and


Clive P. Page, PhDSackler Institute of Pulmonary

PharmacologyDivision of Pharmaceutical SciencesKing’s College LondonLondon, UK

Kathelijne Peerlinck, MD, PhDCenter for Molecular and Vascular Biology,

and Division of Bleeding and VascularDisorders

University of LeuvenLeuven, Belgium

Simon C. Pitchford, PhDLeukocyte Biology SectionNational Heart and Lung InstituteImperial College LondonLondon, UK

Andrea PrimianiDivision of HematologyDepartment of Pediatrics, and Department

of Obstetrics and GynecologyWeill Medical College of Cornell UniversityNew York, NY, USA

Azad Raiesdana, MDDepartment of MedicineCardiovascular DivisionBrigham and Women’s Hospital, andHarvard Medical SchoolBoston, MA, USA

Masashi Sakuma, MDDivision of Cardiovascular MedicineUniversity Hospitals Case Medical CenterCleveland, OH, USA

Eduard Shantsila, MDHaemostasis Thrombosis and

Vascular Biology UnitUniversity Department of MedicineCity HospitalBirmingham, UK

vii

Contributors

Daniel I. Simon, MDDivision of Cardiovascular Medicine,and Heart & Vascular InstituteUniversity Hospitals Case Medical Center,and Case Western Reserve University

School of MedicineCleveland, OH, USA

Sherrill J. Slichter, MDPlatelet Transfusion ResearchPuget Sound Blood Center, andUniversity of Washington

School of MedicineSeattle, WA, USA

Jaapjan D. Snoep, MScSection of Vascular MedicineDepartment of General Internal Medicine–

Endocrinology, andDepartment of Clinical Epidemiology

Leiden University Medical CentreLeiden, The Netherlands

Timothy J. Stalker, PhDUniversity of PennsylvaniaPhiladelphia, PA, USA

Ronald G. Strauss, MDUniversity of Iowa College of Medicine,

and DeGowin Blood CenterUniversity of Iowa HospitalsIowa City, IA, USA

Jouke T. Tamsma, MD, PhDSection of Vascular MedicineDepartment of General Internal


Ayalew Tefferi, MDDivision of HematologyMayo Clinic College of MedicineRochester, MN, USA

Pierluigi Tricoci, MD, MHS, PhDDuke Clinical Research InstituteDuke University Medical CenterDurham, NC, USA

Raymond Verhaeghe, MD, PhDCenter for Molecular and Vascular Biology,

and Division of Bleeding and VascularDisorders

University of LeuvenLeuven, Belgium

Peter Verhamme, MD, PhDCenter for Molecular and Vascular Biology, and

Division of Bleeding and Vascular DisordersUniversity of LeuvenLeuven, Belgium

Jos Vermylen, MD, PhDCenter for Molecular and Vascular Biology, and

Division of Bleeding and Vascular DisordersUniversity of LeuvenLeuven, Belgium

Gemma Vilahur, MSCardiovascular InstituteMount Sinai School of MedicineNew York, NY, USA,and Cardiovascular Research CenterCSIC-ICCC, HSCSP, UABBarcelona, Spain

Timothy Watson, MRCPHaemostasis Thrombosis and Vascular Biology UnitUniversity Department of MedicineCity HospitalBirmingham, UK

viii

PREFACE

Progress in the field of platelet research has acceler-ated greatly over the last few years. If we just considerthe time elapsed since our previous book on platelets(Platelets in Thrombotic And Non-Thrombotic Disor-ders, 2002), over 10 000 publications can be found in aPubMed search using the keyword “platelets.”

Many factors account for this rapidly expandinginterest in platelets, among them an explosive increasein the knowledge of the basic biology of plateletsand of their participation in numerous clinical disor-ders as well as the increasing success of establishedplatelet-modifying therapies in several clinical set-tings. All of this has led to the publication of severalbooks devoted to platelets in recent years. Neverthe-less, it is surprising that none of these is a hand-book that presents a comprehensive and pragmaticapproach to the clinical aspects of platelet involve-ment in hematologic, cardiovascular, and inflamma-tory disorders and the many new developments andcontroversial aspects of platelet pharmacology andtherapeutics.

Based on these considerations, this new book wasnot prepared simply as an update of the previous edi-tion but has undergone a number of conceptual andorganizational changes.

A new editor with a specific expertise in hematol-ogy, Dr. José López, has joined the group of the editors,bringing in a hematologically oriented view. The bookhas been shortened and is now focused on the clini-cal aspects of the involvement of platelets in hemato-logic and cardiovascular disorders. Practical aspectsof the various topics have been strongly empha-sized, with the aim of providing a practical handbookuseful for residents in hematology and cardiology,medical and graduate students, physicians, and alsoscientists interested in the broad clinical implications

of platelet research. We expect that this book will alsobe of interest to vascular medicine specialists, aller-gologists, rheumatologists, pulmonologists, diabetol-ogists, and oncologists.

The book has been organized into four sections, cov-ering platelet physiology, bleeding disorders, throm-botic disorders, and antithrombotic therapy. A totalof 26 chapters cover all the conventional and lessconventional aspects of platelet involvement in dis-ease; emphasis has been given to the recent develop-ments in each field, but always mentioning the key dis-coveries that have contributed to present knowledge.A section on promising future avenues of researchand a clear table with the heading “Take-Home Mes-sages” have been included in each chapter. A groupof leading experts in the various fields covered bythe book, from eight countries on three continents,have willingly agreed to participate; many of them areclinical opinion leaders on the topics discussed. Allchapters have undergone extensive editing for homo-geneity, to help provide a balanced and completeview on the various subjects and reduce overlap to aminimum.

We believe that, thanks to the efforts and continuedcommitment of all the people involved, the result isa novel, light, and quick-reading handbook providingan easy-to-consult guide to the diagnosis and treat-ment of disorders in which platelets play a prominentrole.

Additional illustrative material is available onlinethrough the site of Cambridge University Press(www.cambridge.org/9780521881159).

This book would have not been possible without thehelp of our editorial assistants (M. Sensi, R. Stevens)and of several coworkers in the Institutions of the indi-vidual editors (S. Momi, E. Falcinelli). An excellent

ix

Preface

collaboration with the team at Cambridge Univer-sity Press (Daniel Dunlavey, Deborah Russell, RachaelLazenby, Katie James, Jane Williams, and EleanorUmali) has also been crucial to the successful accom-plishment of what has seemed, at certain moments, adesperate task.

We hope that this book will be interesting and usefulto readers as much as it has been for us.

The Editors

x

Glossary

GLOSSARY

αIIbβ3 αIIbβ3 or glycoprotein IIb-IIIaαIIbβ3, α2β1 Platelet integrinsαMβ2, αLβ2 Leukocyte β2 integrinsαvβ3 Vitronectin receptorβ-TG β-thromboglobulinAA Arachidonic acidACD Citric acid, sodium citrate,

dextroseACS Acute coronary syndromeADP Adenosine-5’-diphosphateAKT Serine/threonine protein

kinaseAPS Antiphospholipid antibody

syndromeASA Acetylsalicylic acidATP Adenosine-5’-triphosphateAVWS Acquired von Willebrand

syndromeBSS Bernard–Soulier syndromeBT Bleeding timeCAD Coronary artery diseasecAMP Cyclic AMPCAMT Congenital amegacaryocytic

thrombocytopeniaCD40L (CD154) CD40 ligandCD62P P-selectinCFU Colony forming unitcGMP Cyclic GMPCHS Chediak–Higashi syndromeCML Chronic myeloid leukemiaCOX-1 Cyclooxygenase-1COX-2 Cyclooxygenase-2CPD Citrate-phosphate-dextroseCRP C-reactive proteinCVID Common variable

immunodeficiency

DDAVP L-deamino-8-O-dargininevasopressin

DIC Disseminated intravascularcoagulation

DTS Dense tubular systemDVT Deep venous thrombosisEC Endothelial cellsECM Extracellular matrixEDHF Endothelium-derived

hyperpolarizing factorEDTA Ethylene diamine tetracetic acidEGF Epidermal growth factoreNOS Endothelial nitric oxide synthaseEP PGE2 receptorEPCs Endothelial progenitor cellsERK Extracellular signal-regulated

kinaseET Essential thrombocytemiaFAK Focal adhesion kinaseFbg FibrinogenFn FibrinGEF Guanine nucleotide exchange

factorGP Glycoprotein (e.g., GP Ib, GP

Ib/IX/V)GPIb Glycoprotein IbGPCR G protein-coupled receptorGPS Gray platelet syndromeGT Glanzmann’s thrombasthenia12-HETE 12-(S)-hydroxyeicosatetraenoic

acidHDL High-density lipoproteinHIT Heparin-induced

thrombocytopeniaHLA Human leukocyte antigenHPA Human platelet antigen

xi

Glossary

HPS Hermansky–Pudlak syndrome5-HT 5-hydroxytryptamineHUS Hemolytic uremic syndromeICAM-1 Intercellular adhesion

molecule-1ICAM-2 Intercellular adhesion

molecule-2ICH Intracranial hemorrhageIFN interferonIL InterleukiniNOS Inducible nitric oxide synthaseIP Prostacyclin receptorITP Idiopathic thrombocytopenic

purpuraIVIG Intravenous immunoglobulinJAK Janus family kinaseJAM Junctional adhesion moleculeJNK c-Jun N-terminal kinaseLDL Low-density lipoproteinLDH lactate dehydrogenaseLFA-1 Leukocyte function-associated

molecule-1LMWHs Low-molecular-weight

heparinsLOX-1 Lectin-like oxLDL-1LPS LipopolysaccharideLT LeukotrieneMAC-1

(CD11b/CD18)

Leukocyte integrin αMβ2

MAIPA Monoclonal antibody-specificimmobilization of plateletantigens

MAPK Mitogen-activated proteinkinase

MAPKKK,MEKK

MAPK kinase kinase

MCP-1 Monocyte chemoattractantprotein-1

MDS Myelodysplastic syndromeMEK, MAPKK MAPK/ERK kinaseMF MyelofibrosisMI Myocardial infarctionMIP-1α Macrophage inflammatory

protein-1αMK MegakaryocyteMMPs Matrix metalloproteinasesMPD Myeloproliferative disorders

MPV Mean platelet volumeNAIT Neonatal allo-immune

thrombocytopeniaNFkB Nuclear factor kBnNOS Neuronal nitric oxide synthaseNO Nitric oxideNSAID Nonsteroidal

anti-inflammatory drugNSTEMI Non-ST–elevation myocardial

infarctionOCS Open canalicular systemPAF Platelet activating factorPAIgG Platelet-associated IgGPAR Protease-activated receptor

(e.g., PAR1, PAR4)PDE inhibitors phosphodiesterase inhibitorsPDGF Platelet-derived growth factorPE Pulmonary embolismPFA-100® Platelet Function

Analyzer-100®

PG ProstaglandinPGH2 Prostaglandin H2PGI2 Prostacyclin (prostaglandin I2)PI PhosphatidylinositolPIP2 Phosphoinositide 4,5

bisphosphatePIP3 Phosphoinositide 3, 4, 5 tris

phosphatePI3K Phosphoinositol-3 kinasePKA Protein kinase APKC Protein kinase CPLA2 Phospholipase A2PLTs PlateletsPMN Polymorphonuclear cellsPMP Platelet microparticlesPNH Paroxysmal nocturnal

hemoglobinuriaPPP Platelet-poor plasmaPR Platelet reactivity indexPRP Plateletrich plasmaPS phosphatidyl serinePSGL-1 P-selectin glycoprotein

ligand-1PT Prothrombin timePTP Posttransfusion purpuraPTT Partial thromboplastin timePUBS Periumbilical blood samplingPV Polycytemia vera

xii

Glossary

RANTES Regulated on activation normalT cell-expressed and secreted

RGD Arg-Gly-AspROS Reactive oxygen speciesSDF-1 Stromal cell-derived factor 1STEMI ST-segment-elevation

myocardial infarctionTAR Congenital thrombocytopenia

with absent radiusTARC Thymus and activation-

regulated chemokineTF Tissue factorTGF Transforming growth

factorTMA Thrombotic microangiopathy

TNF Tumor necrosis factorTNFα Tumor necrosis factor αTP Thromboxane A2 receptorTPO ThrombopoietinTTP Thrombotic thrombocytopenic

purpuraTxA2 Thromboxane A2UFH Unfractionated heparinUVA, UVB Ultraviolet A, ultraviolet BVCAM-1 Vascular cell adhesion

molecule-1VWF von Willebrand factorWAS Wiskott–Aldrich syndromeWBCs White blood cellsWP Washed platelet

xiii

C H A P T E R

1 THE STRUCTURE AND PRODUCTIONOF BLOOD PLATELETS

Joseph E. Italiano, Jr.Brigham and Women’s Hospital; Children’s Hospital Boston; and Harvard Medical School, Boston, MA, USA

INTRODUCTION

Blood platelets are small, anucleate cellular fragmentsthat play an essential role in hemostasis. During nor-mal circulation, platelets circulate in a resting stateas small discs (Fig. 1.1A). However, when challengedby vascular injury, platelets are rapidly activated andaggregate with each other to form a plug on the vesselwall that prevents vascular leakage. Each day, 100 bil-lion platelets must be produced from megakaryocytes(MKs) to maintain the normal platelet count of 2 to3 × 108/mL. This chapter is divided into three sec-tions that discuss the structure and organization ofthe resting platelet, the mechanisms by which MKsgive birth to platelets, and the structural changes thatdrive platelet activation.

1. THE STRUCTURE OF THERESTING PLATELET

Human platelets circulate in the blood as discs thatlack the nucleus found in most cells. Platelets are het-erogeneous in size, exhibiting dimensions of 0.5 × 3.0μm.1 The exact reason why platelets are shaped asdiscs is unclear, although this shape may aid someaspect of their ability to flow close to the endothe-lium in the bloodstream. The surface of the plateletplasma membrane is smooth except for periodicinvaginations that delineate the entrances to the opencanalicular system (OCS), a complex network of inter-winding membrane tubes that permeate the platelet’scytoplasm.2 Although the surface of the plateletplasma membrane appears featureless in most micro-graphs, the lipid bilayer of the resting platelet con-tains a large concentration of transmembrane recep-tors. Some of the major receptors found on the surfaceof resting platelets include the glycoprotein receptor

for von Willebrand factor (VWF); the major serpentinereceptors for ADP, thrombin, epinephrine, and throm-boxane A2; the Fc receptor Fcγ RIIA; and the β3 and β1integrin receptors for fibrinogen and collagen.

The intracellular components of theresting platelet

The plasma membrane of the platelet is separatedfrom the general intracellular space by a thin rim ofperipheral cytoplasm that appears clear in thin sec-tions when viewed in the electron microscope, but itactually contains the platelet’s membrane skeleton.Underneath this zone is the cytoplasm, which con-tains organelles, storage granules, and the specializedmembrane systems.

GranulesOne of the most interesting characteristics of plateletsis the large number of biologically active moleculescontained in their granules. These molecules arepoised to be deposited at sites of vascular injury andfunction to recruit other blood-borne cells. In rest-ing platelets, granules are situated close to the OCSmembranes. During activation, the granules fuse andexocytose into the OCS.3 Platelets have two major rec-ognized storage granules: α and dense granules. Themost abundant are α granules (about 40 per platelet),which contain proteins essential for platelet adhe-sion during vascular repair. These granules are typi-cally 200 to 500 nm in diameter and are spherical inshape with dark central cores. They originate fromthe trans Golgi network, where their characteristicdark nucleoid cores become visible within the bud-ding vesicles.4 Alpha granules acquire their molecu-lar contents from both endogenous protein synthesis

1

Joseph E. Italiano, Jr.

A B

Figure 1.1 The structure of the resting platelet. A. Differential interference contrast micrograph of a field of human discoid resting platelets.

B. Immunofluorescence staining of fixed, resting platelets with Alexa 488-antitubulin antibody reveals the microtubule coil. Coils are

1–3 μm in diameter.

and by the uptake and packaging of plasma pro-teins via receptor-mediated endocytosis and pinocy-tosis.5 Endogenously synthesized proteins such asPF-4, β thromboglobulin, and von Willebrand factorare detected in megakaryocytes (MKs) before endocy-tosed proteins such as fibrinogen. In addition, synthe-sized proteins predominate in the juxtanuclear Golgiarea, while endocytosed proteins are localized in theperipheral regions of the MK.5 It has been well doc-umented that uptake and delivery of fibrinogen to αgranules is mediated by the major membrane glyco-protein αIIbβ3.6,7,8 Several membrane proteins criticalto platelet function are also packaged into alpha gran-ules, including αIIbβ3, CD62P, and CD36. α granulesalso contain the majority of cellular P-selectin in theirmembrane. Once inserted into the plasma membrane,P-selectin recruits neutrophils through the neutrophilcounter receptor, the P-selectin glycoprotein ligand(PSGL1).9 Alpha granules also contain over 28 angio-genic regulatory proteins, which allow them to func-tion as mobile regulators of angiogenesis.10 Althoughlittle is known about the intracellular tracking of pro-teins in MKs and platelets, experiments using ultra-thin cryosectioning and immunoelectron microscopysuggest that multivesicular bodies are a crucial inter-mediate stage in the formation of platelet α gran-ules.11 During MK development, these large (up to0.5 μm) multivesicular bodies undergo a gradual tran-

sition from granules containing 30 to 70 nm internalvesicles to granules containing predominantly densematerial. Internalization kinetics of exogenous bovineserum albumin–gold particles and of fibrinogen posi-tion the multivesicular bodies and α granules sequen-tially in the endocytic pathway. Multivesicular bodiescontain the secretory proteins VWF and β throm-boglobulin, the platelet-specific membrane protein P-selectin, and the lysosomal membrane protein CD63,suggesting that they are a precursor organelle for αgranules.11 Dense granules (or dense bodies), 250 nmin size, identified in electron micrographs by virtueof their electron-dense cores, function primarily torecruit additional platelets to sites of vascular injury.Dense granules contain a variety of hemostaticallyactive substances that are released upon platelet acti-vation, including serotonin, catecholamines, adeno-sine 5′-diphosphate (ADP), adenosine 5′-triphosphate(ATP), and calcium. Adenosine diphosphate is a strongplatelet agonist, triggering changes in the shape ofplatelets, the granule release reaction, and aggrega-tion. Recent studies have shown that the transport ofserotonin in dense granules is essential for the processof liver regeneration.12 Immunoelectron microscopystudies have also indicated that multivesicular bodiesare an intermediary stage of dense granule maturationand constitute a sorting compartment betweenα gran-ules and dense granules.

2

CHAPTER 1: The Structure and Production of Blood Platelets

OrganellesPlatelets contain a small number of mitochondriathat are identified in the electron microscope by theirinternal cisternae. They provide an energy source forthe platelet as it circulates in the bloodstream for 7days in humans. Lysosomes and peroxisomes are alsopresent in the cytoplasm of platelets. Peroxisomesare small organelles that contain the enzyme cata-lase. Lysosomes are also tiny organelles that contain alarge assortment of degradative enzymes, including β-galactosidase, cathepsin, aryl sulfatase, β-glucuroni-dase, and acid phosphatases. Lysosomes function pri-marily in the break down of material ingested byphagocytosis or pinocytosis. The main acid hydrolasecontained in lysosomes is β-hexosaminidase.13

Membrane systems

Open canalicular system

The open canalicular system (OCS) is an elaborate sys-tem of internal membrane tunnels that has two majorfunctions. First, the OCS serves as a passageway to thebloodstream, in which the contents can be released.Second, the OCS functions as a reservoir of plasmamembrane and membrane receptors. For example,approximately one-third of the thrombin receptorsare located in the OCS of the resting platelet, await-ing transport to the surface of activated platelets. Spe-cific membrane receptors are also transported in thereverse direction from the plasma membrane to theOCS, in a process called downregulation, after cell acti-vation. The VWF receptor is the best studied glycopro-tein in this respect. Upon platelet activation, the VWFreceptor moves inward into the OCS. One major ques-tion that has not been resolved is how other proteinspresent in the plasma membrane are excluded fromentering the OCS. The OCS also functions as a sourceof redundant plasma membrane for the surface-to-volume ratio increase occurring during the cell spread-ing that accompanies platelet activation.

Dense tubular system

Platelets contain a dense tubular system (DTS),14

named according to its inherent electron opacity, thatis randomly woven through the cytoplasmic space.The DTS is believed to be similar in function to thesmooth endoplamic reticular system in other cellsand serves as the predominant calcium storage sys-tem in platelets. The DTS membranes possess Ca2+

pumps that face inward and maintain the cytosoliccalcium concentrations in the nanomolar range in theresting platelet. The calcium pumped into the DTS issequestered by calreticulin, a calcium-binding pro-tein. Ligand-responsive calcium gates are also situ-ated in the DTS. The soluble messenger inositol 1,4,5triphosphate releases calcium from the DTS. The DTSalso functions as the major site of prostaglandin andthromboxane synthesis in platelets.15 It is the sitewhere the enzyme cyclooxygenase is located. The DTSdoes not stain with extracellular membrane tracers,indicating that it is not in contact with the externalenvironment.

The cytoskeleton of the resting plateletThe disc shape of the resting platelet is maintainedby a well-defined and highly specialized cytoskeleton.This elaborate system of molecular struts and gird-ers maintains the shape and integrity of the plateletas it encounters high shear forces during circula-tion. The three major cytoskeletal components of theresting platelet are the marginal microtubule coil,the actin cytoskeleton, and the spectrin membraneskeleton.

The marginal band of microtubulesOne of the most distinguishing features of the rest-ing platelet is its marginal microtubule coil (Fig.1.1B).16,17 Alpha and β tubulin dimers assemble intomicrotubule polymers under physiologic conditions;in resting platelets, tubulin is equally divided betweendimer and polymer fractions. In many cell types, theα and β tubulin subunits are in dynamic equilib-rium with microtubules, such that reversible cyclesof microtubule assembly–disassembly are observed.Microtubules are long, hollow polymers 24 nm indiameter; they are responsible for many types of cel-lular movements, such as the segregation of chromo-somes during mitosis and the transport of organellesacross the cell. The microtubule ring of the restingplatelet, initially characterized in the late 1960s byJim White, has been described as a single micro-tubule approximately 100 μm long, which is coiled 8to 12 times inside the periphery of the platelet.16 Theprimary function of the microtubule coil is to maintainthe discoid shape of the resting platelet. Disassemblyof platelet microtubules with drugs such as vincristine,colchicine, or nocodazole cause platelets to roundand lose their discoid shape.16 Cooling platelets to

3


4◦C also causes disassembly of the microtubule coiland loss of the discoid shape.17 Furthermore, elegantstudies show that mice lacking the major hematopoi-etic β-tubulin isoform (β-1 tubulin) contain plateletsthat lack the characteristic discoid shape and havedefective marginal bands.18 Genetic elimination ofβ-1 tubulin in mice results in thrombocytopenia,with mice having circulating platelet counts below50% of normal. Beta-1 tubulin–deficient platelets arespherical in shape; this appears to be due to defec-tive marginal bands with fewer microtubule coilings.Whereas normal platelets possess a marginal bandthat consists of 8 to 12 coils, β-1 tubulin knock-out platelets contain only 2 or 3 coils.18,19 A humanβ-1 tubulin functional substitution (AG>CC) induc-ing both structural and functional platelet alterationshas been described.20 Interestingly, the Q43P β-1-tubulin variant was found in 10.6% of the generalpopulation and in 24.2% of 33 unrelated patientswith undefined congenital macrothrombocytopenia.Electron microscopy revealed enlarged spherocyticplatelets with a disrupted marginal band and struc-tural alterations. Moreover, platelets with this vari-ant showed mild platelet dysfunction, with reducedsecretion of ATP, thrombin-receptor-activating pep-tide (TRAP)–induced aggregation, and impaired adhe-sion to collagen under flow conditions. A more thandoubled prevalence of the β-1-tubulin variant wasobserved in healthy subjects not undergoing ischemicevents, suggesting that it could confer an evolutionaryadvantage and might play a protective cardiovascularrole.

The microtubules that make up the coil are coatedwith proteins that regulate polymer stability.21 Themicrotubule motor proteins kinesin and dynein havebeen localized to platelets, but their roles in restingand activated platelets have not yet been defined.

The actin cytoskeletonActin, at a concentration of 0.5 mM, is the most plenti-ful of all the platelet proteins with 2 million moleculesexpressed per platelet.1 Like tubulin, actin is in adynamic monomer-polymer equilibrium. Some 40%of the actin subunits polymerize to form the 2000 to5000 linear actin filaments in the resting cell.22 The restof the actin in the platelet cytoplasm is maintainedin storage as a 1 to 1 complex with β-4-thymosin23

and is converted to filaments during platelet activa-tion to drive cell spreading. All evidence indicates

that the filaments of the resting platelet are intercon-nected at various points into a rigid cytoplasmic net-work, as platelets express high concentrations of actincross-linking proteins, including filamin24,25 and α-actinin.26 Both filamin and α-actinin are homodimersin solution. Filamin subunits are elongated strandscomposed primarily of 24 repeats, each about 100amino acids in length, which are folded into IgG-likeβ barrels.27,28 There are three filamin genes on chro-mosomes 3, 7, and X. Filamin A (X)29 and filamin B(3)30 are expressed in platelets, with filamin A beingpresent at greater than 10-fold excess to filamin B. Fil-amin is now recognized to be a prototypical scaffoldingprotein that attracts binding partners and positionsthem adjacent to the plasma membrane.31 Partnersbound by filamin members include the small GTPases,ralA, rac, rho, and cdc42, with ralA binding in a GTP-dependent manner32; the exchange factors Trio andToll; and kinases such as PAK1, as well as phosphatasesand transmembrane proteins. Essential to the struc-tural organization of the resting platelet is an inter-action that occurs between filamin and the cytoplas-mic tail of the GPIbα subunit of the GPIb-IX-V com-plex. The second rod domain (repeats 17 to 20) offilamin has a binding site for the cytoplasmic tail ofGPIbα 33, and biochemical experiments have shownthat the bulk of platelet filamin (90% or more) is incomplex with GPIbα.34 This interaction has three con-sequences. First, it positions filamin’s self-associationdomain and associated partner proteins at the plasmamembrane while presenting filamin’s actin bindingsites into the cytoplasm. Second, because a large frac-tion of filamin is bound to actin, it aligns the GPIb-IX-Vcomplexes into rows on the surface of the platelet overthe underlying filaments. Third, because the filaminlinkages between actin filaments and the GPIb-IX-Vcomplex pass through the pores of the spectrin lattice,it restrains the molecular movement of the spectrinstrands in this lattice and holds the lattice in compres-sion. The filamin-GPIbα connection is essential for theformation and release of discoid platelets by MKs, asplatelets lacking this connection are large and frag-ile and produced in low numbers. However, the roleof the filamin-VWF receptor connection in plateletconstruction per se is not fully clear. Because a lownumber of Bernard-Soulier platelets form and releasefrom MKs, it can be argued that this connection is alate event in the maturation process and is not per serequired for platelet shedding.

4


The spectrin membrane skeletonThe OCS and plasma membrane of the resting plateletare supported by an elaborate cytoskeletal system.The platelet is the only other cell besides the ery-throcyte whose membrane skeleton has been visual-ized at high resolution. Like the erythrocyte’s skeleton,that of the platelet membrane is a self-assembly ofelongated spectrin strands that interconnect throughtheir binding to actin filaments, generating triangu-lar pores. Platelets contain approximately 2000 spec-trin molecules.22,35,36 This spectrin network coats thecytoplasmic surface of both the OCS and plasma mem-brane systems. Although considerably less is knownabout how the spectrin–actin network forms and isconnected to the plasma membrane in the platelet rel-ative to the erythrocyte, certain differences betweenthe two membrane skeletons have been defined. First,the spectrin strands composing the platelet mem-brane skeleton interconnect using the ends of longactin filaments instead of short actin oligomers.22

These ends arrive at the plasma membrane originatingfrom filaments in the cytoplasm. Hence, the spectrinlattice is assembled into a continuous network by itsassociation with actin filaments. Second, tropomod-ulins are not expressed at sufficiently high levels, if atall, to have a major role in the capping of the pointedends of the platelet actin filaments; instead, biochemi-cal experiments have revealed that a substantial num-ber (some 2000) of these ends are free in the restingplatelet. Third, although little tropomodulin proteinis expressed, adducin is abundantly expressed andappears to cap many of the barbed ends of the fil-aments composing the resting actin cytoskeleton.37

Adducin is a key component of the membrane skele-ton, forming a triad complex with spectrin and actin.Capping of barbed filament ends by adducin alsoserves the function of targeting them to the spectrin-based membrane skeleton, as the affinity of spectrinfor adducin-actin complexes is greater than for eitheractin or adducin alone.38,39,40

MEGAKARYOCYTE DEVELOPMENTAND PLATELET FORMATION

Megakaryocytes are highly specialized precursorcells that function solely to produce and releaseplatelets into the circulation. Understanding mech-anisms by which MKs develop and give rise toplatelets has fascinated hematologists for over a

century. Megakaryocytes are descended from pluripo-tent stem cells and undergo multiple DNA replica-tions without cell divisions by the unique processof endomitosis. During endomitosis, polyploid MKsinitiate a rapid cytoplasmic expansion phase char-acterized by the development of a highly developeddemarcation membrane system and the accumula-tion of cytoplasmic proteins and granules essentialfor platelet function. During the final stages of devel-opment, the MKs cytoplasm undergoes a dramaticand massive reorganization into beaded cytoplasmicextensions called proplatelets. The proplatelets ulti-mately yield individual platelets.

Commitment to themegakaryocyte lineage

Megakaryocytes, like all terminally differentiatedhematopoietic cells, are derived from hematopoieticstem cells, which are responsible for constant produc-tion of all circulating blood cells.41,42 Hematopoieticcells are classified by their ability to reconstitute hostanimals, surface markers, and colony assays thatreflect their developmental potential. Hematopoi-etic stem cells are rare, making up less than 0.1%of cells in the marrow. The development of MKsfrom hematopoietic stem cells entails a sequenceof differentiation steps in which the developmentalcapacities of the progenitor cells become graduallymore limited. Hematopoietic stem cells in mice aretypically identified by the surface markers Lin-Sca-1+c-kithigh.43,44,45 A detailed model of hematopoiesishas emerged from experiments analyzing the effectsof hematopoietic growth factors on marrow cellscontained in a semisolid medium. Hematopoieticstem cells give rise to two major lineages, a commonlymphoid progenitor that can develop into lympho-cytes and a myeloid progenitor that can develop intoeosinophil, macrophage, myeloid, erythroid, andMK lineages. A common erythroid-megakaryocyticprogenitor arises from the myeloid lineage.46 How-ever, recent studies also suggest that hematopoieticstem cells may directly develop into erythroid–megakaryocyte progenitors.47 All hematopoieticprogenitors express surface CD34 and CD41, and thecommitment to the MK lineage is indicated by expres-sion of the integrin CD61 and elevated CD41 levels.From the committed myeloid progenitor cell (CFU-GEMM), there is strong evidence for a bipotential

5


progenitor intermediate between the pluripotentialstem cell and the committed precursor that can giverise to biclonal colonies composed of megakaryocyticand erythroid cells.48,49,50 The regulatory pathwaysand transcriptional factors that allow the erythroidand MK lineages to separate from the bipotentialprogenitor are currently unknown. Diploid precursorsthat are committed to the MK lineage have tradition-ally been divided into two colonies based on theirfunctional capacities.51,52,53,54 The MK burst-formingcell is a primitive progenitor that has a high prolif-eration capacity that gives rise to large MK colonies.Under specific culture conditions, the MK burst-forming cell can develop into 40 to 500 MKs within aweek. The colony-forming cell is a more mature MKprogenitor that gives rise to a colony containing from3 to 50 mature MKs, which vary in their proliferationpotential. MK progenitors can be readily identifiedin bone marrow by immunoperoxidase and acetyl-cholinesterase labeling.55,56,57 Although both humanMK colony-forming and burst-forming cells expressthe CD34 antigen, only colony-forming cells expressthe HLA-DR antigen.58

Various classification schemes based on morpho-logic features, histochemical staining, and biochem-ical markers have been used to categorize differentstages of MK development. In general, three typesof morphologies can be identified in bone marrow.The promegakaryoblast is the first recognizable MKprecursor. The megakaryoblast, or stage I MK, is amore mature cell that has a distinct morphology.59 Themegakaryoblast has a kidney-shaped nucleus withtwo sets of chromosomes (4N). It is 10 to 50 μmin diameter and appears intensely basophilic inRomanovsky-stained marrow preparations due to thelarge number of ribosomes, although the cytoplasmat this stage lacks granules. The megakaryoblast dis-plays a high nuclear-to-cytoplasmic ratio; in rodents,it is acetylcholinesterase-positive. The promegakary-ocyte, or Stage II MK, is 20 to 80 μm in diameterwith a polychromatic cytoplasm. The cytoplasm of thepromegakaryocyte is less basophilic than that of themegakaryoblast and now contains developing gran-ules.

Endomitosis

Megakaryocytes, unlike most other cells, undergoendomitosis and become polyploid through re-

peated cycles of DNA replication without cell div-ision.60,61,62,63 At the end of the proliferation phase,mononuclear MK precursors exit the diploid state todifferentiate and undergo endomitosis, resulting in acell that contains multiples of a normal diploid chro-mosome content (i.e., 4N, 16N, 32N, 64N).64 Althoughthe number of endomitotic cycles can range from twoto six, the majority of MKs undergo three endomi-totic cycles to attain a DNA content of 16N. How-ever, some MKs can acquire a DNA content as highas 256N. Megakaryocyte polyploidization results ina functional gene amplification whose likely func-tion is an increase in protein synthesis paralleling cellenlargement.65 The mechanisms that drive endomito-sis are incompletely understood. It was initially postu-lated that polyploidization may result from an absenceof mitosis after each round of DNA replication. How-ever, recent studies of primary MKs in culture indi-cate that endomitosis does not result from a com-plete absence of mitosis but rather from a prematurelyterminated mitosis.65,66,67 Megakaryocyte progenitorsinitiate the cycle and undergo a short G1 phase, a typi-cal 6- to 7-hour S phase for DNA synthesis, and a shortG2 phase followed by endomitosis. Megakaryocytesbegin the mitotic cycle and proceed from prophase toanaphase A but do not enter anaphase B or telophaseor undergo cytokinesis. During polyploidization ofMKs, the nuclear envelope breaks down and an abnor-mal spherical mitotic spindle forms. Each spindleattaches chromosomes that align to a position equidis-tant from the spindle poles (metaphase). Sister chro-matids segregate and begin to move toward theirrespective poles (anaphase A). However, the spin-dle poles fail to migrate apart and do not undergothe separation typically observed during anaphase B.Individual chromatids are not moved to the poles, andsubsequently a nuclear envelope reassembles aroundthe entire set of sister chromatids, forming a singleenlarged but lobed nucleus with multiple chromo-some copies. The cell then skips telophase and cytoki-nesis to enter G1. This failure to fully separate sets ofdaughter chromosomes may prevent the formation ofa nuclear envelope around each individual set of chro-mosomes.66,67

In most cell types, checkpoints and feedback con-trols make sure that DNA replication and cell divi-sion are synchronized. Megakaryocytes appear to bethe exception to this rule, as they have managed toderegulate this process. Recent work by a number of

6


laboratories has focused on identifying the signalsthat regulate polyploidization in MKs.68 It has beenproposed that endomitosis may be the consequenceof a reduction in the activity of mitosis-promotingfactor (MPF), a multiprotein complex consisting ofCdc2 and cyclin B.69,70 MPF possesses kinase activ-ity, which is necessary for entry of cells into mitosis.In most cell types, newly synthesized cyclin B binds toCdc2 and produces active MPF, while cyclin degra-dation at the end of mitosis inactivates MPF. Con-ditional mutations in strains of budding and fissionyeast that inhibit either cyclin B or Cdc2 cause themto go through an additional round of DNA replica-tion without mitosis.71,72 In addition, studies using ahuman erythroleukemia cell line have demonstratedthat these cells contain inactive Cdc2 during poly-ploidization, and investigations with phorbol ester–induced Meg T cells have demonstrated that cyclin Bis absent in this cell line during endomitosis.73,74 How-ever, it has been difficult to define the role of MPF activ-ity in promoting endomitosis because these cell lineshave a curtailed ability to undergo this process. Fur-thermore, experiments using normal MKs in culturehave demonstrated normal levels of cyclin B and Cdc2with functional mitotic kinase activity in MKs under-going mitosis, suggesting that endomitosis can be reg-ulated by signaling pathways other than MPF. Cyclinsappear to play a critical role in directing endomito-sis, although a triple knockout of cyclins D1, D2, andD3 does not appear to affect MK development.75 Yet,cyclin E–deficient mice do exhibit a profound defectin MK development.76 It has recently been demon-strated that the molecular programming involved inendomitosis is characterized by the mislocalization orabsence of at least two critical regulators of mitosis:the chromosomal passenger proteins Aurora-B/AIM-1 and survivin.77

Cytoplasmic maturation

During endomitosis, the MK begins a maturation stagein which the cytoplasm rapidly fills with platelet-specific proteins, organelles, and membrane systemsthat will ultimately be subdivided and packaged intoplatelets. Through this stage of maturation, the MKenlarges dramatically and the cytoplasm acquires itsdistinct ultrastructural features, including the devel-opment of a demarcation membrane system (DMS),the assembly of a dense tubular system, and the forma-

tion of granules. During this stage of MK development,the cytoplasm contains an abundance of ribosomesand rough endoplasmic reticulum, where protein syn-thesis occurs. One of the most striking features of amature MK is its elaborate demarcation membranesystem, an extensive network of membrane chan-nels composed of flattened cisternae and tubules. Theorganization of the MK cytoplasm into membrane-defined platelet territories was first proposed by Kautzand DeMarsh,78 and a high-resolution description ofthis membrane system by Yamada soon followed.79

The DMS is detectable in early promegakaryocytesbut becomes most prominent in mature MKs where—except for a thin rim of cortical cytoplasm from whichit is excluded—it permeates the MK cytoplasm. Ithas been proposed that the DMS derives from MKplasma membrane in the form of tubular invagina-tions. 80,81,82 The DMS is in contact with the externalmilieu and can be labeled with extracellular tracers,such as ruthenium red, lanthanum salts, and tannicacid.83,84 The exact function of this elaborate smoothmembrane system has been hotly debated for manyyears. Initially, it was postulated to play a central rolein platelet formation by defining preformed “plateletterritories” within the MK cytoplasm (see below). How-ever, recent studies more strongly suggest that theDMS functions primarily as a membrane reserve forproplatelet formation and extension. The DMS hasalso been proposed to mature into the open canalicu-lar system of the mature platelet, which functions as achannel for the secretion of granule contents. How-ever, bovine MKs, which have a well-defined DMS,produce platelets that do not develop an OCS, sug-gesting the OCS is not necessarily a remnant of theDMS.84

Platelet formation

The mechanisms by which blood platelets are pro-duced have been studied for approximately 100 years.In 1906, James Homer Wright at Massachussetts Gen-eral Hospital began a detailed analysis of how giantprecursor MKs give birth to platelets. Many theo-ries have been suggested over the years to explainhow MKs produce platelets. The demarcation mem-brane system (DMS), described in detail by Yamadain 1957, was initially proposed to demarcate pre-formed “platelet territories” within the cytoplasm ofthe MK.79 Microscopists recognized that maturing

7


MKs become filled with membranes and platelet-specific organelles and proposed that these mem-branes form a system that defines fields for developingplatelets.85 Release of individual platelets was pro-posed to occur by a massive fragmentation of the MKcytoplasm along DMS fracture lines located betweenthese fields. The DMS model proposes that plateletsform through an elaborate internal membrane reor-ganization process.86 Tubular membranes, which mayoriginate from invagination of the MK plasma mem-brane, are predicted to interconnect and branch, form-ing a continuous network throughout. The fusion ofadjacent tubules has been suggested as a mecha-nism to generate a flat membrane that ultimately sur-rounds the cytoplasm of an assembling platelet. Mod-els attempting to use the DMS to explain how the MKcytoplasm becomes subdivided into platelet volumesand enveloped by its own membrane have lost sup-port because of several inconsistent observations. Forexample, if platelets are delineated within the MK cyto-plasm by the DMS, then platelet fields should exhibitstructural characteristics of resting platelets, whichis not the case.87 Platelet territories within the MKcytoplasm lack marginal microtubule coils, one of themost characteristic features of resting platelet struc-ture. In addition, there are no studies on living MKsdirectly demonstrating that platelet fields explosivelyfragment or shatter into mature, functional platelets.In contrast, studies that focused on the DMS of MKsbefore and after proplatelet retraction induced bymicrotubule depolymerizing agents suggest that thisspecialized membrane system may function primar-ily as a membrane reservoir that evaginates to provideplasma membrane for the extensive growth of pro-platelets.88 Radley and Haller have proposed that DMSmay be a misnomer, and have suggested “invagina-tion membrane system” as a more suitable name todescribe this membranous network.

The majority of evidence that has been gatheredsupports the proplatelet model of platelet production.The term “proplatelet” is generally used to describelong (up to millimeters in length), thin cytoplasmicextensions emanating from MKs.89 These extensionsare characterized by multiple platelet-sized beadslinked together by thin cytoplasmic bridges and arethought to represent intermediate structures in themegakaryocyte-to-platelet transition. The actual con-cept of platelets arising from these pseudopodia-like structures occurred when Wright recognized that

platelets originate from MKs and described “thedetachment of plate-like fragments or segments frompseudopods” from MKs.90 Thiery and Bessis91 andBehnke92 later described the morphology of thesecytoplasmic processes extending from MKs duringplatelet formation in more detail. The classic “pro-platelet theory” was introduced by Becker and DeBruyn, who proposed that MKs form long pseudopod-like processes that subsequently fragment to gener-ate individual platelets.89 In this early model, the DMSwas still proposed to subdivide the MK cytoplasm intoplatelet areas. Radley and Haller later developed the“flow model,” which postulated that platelets derivedexclusively from the interconnected platelet-sizedbeads connected along the shaft of proplatelets88; theysuggested that the DMS did not function to defineplatelet fields but rather as a reservoir of surfacemembrane to be evaginated during proplatelet forma-tion. Developing platelets were assumed to becomeencased by plasma membrane only as proplateletswere formed.

The bulk of experimental evidence now supportsa modified proplatelet model of platelet formation.Proplatelets have been observed (1) both in vivoand in vitro, and maturation of proplatelets yieldsplatelets that are structurally and functionally sim-ilar to blood platelets93,94; (2) in a wide range ofmammalian species, including mice, rats, guinea pigs,dogs, cows, and humans95,96,97,98,99; (3) extendingfrom MKs in the bone marrow through junctionsin the endothelial lining of blood sinuses, wherethey have been hypothesized to be released intocirculation and undergo further fragmentation intoindividual platelets100,101,102; and (4) to be absent inmice lacking two distinct hematopoietic transcriptionfactors. These mice fail to generate proplatelets in vitroand display severe thrombocytopenia.103,104,105 Takentogether, these findings support an important role forproplatelet formation in thrombopoiesis.

The discovery of thrombopoietin and the develop-ment of MK cultures that reconstitute platelet for-mation in vitro has provided systems to study MKsin the act of forming proplatelets. Time-lapse videomicroscopy of living MKs reveals both temporal andspatial changes that lead to the formation of pro-platelets (Fig. 1.2).106 Conversion of the MK cytoplasmconcentrates almost all of the intracellular contentsinto proplatelet extensions and their platelet-sizedparticles, which in the final stages appear as beads

8


A B C

Figure 1.2 Formation of proplatelets by a mouse megakaryocyte. Time-lapse sequence of a maturing megakaryocyte (MK), showing the

events that lead to elaboration of proplatelets in vitro. (A) Platelet production commences when the MK cytoplasm starts to erode at one pole.

(B) The bulk of the megakaryocyte cytoplasm has been converted into multiple proplatelet processes that continue to lengthen and form

swellings along their length. These processes are highly dynamic and undergo bending and branching. (C) Once the bulk of the MK cytoplasm has

been converted into proplatelets, the entire process ends in a rapid retraction that separates the released proplatelets from the residual cell body

(Italiano JE et al., 1999).

linked by thin cytoplasmic strings. The transformationunfolds over 5 to 10 hours and commences with theerosion of one pole (Fig. 1.2B) of the MK cytoplasm.Thick pseudopodia initially form and then elongateinto thin tubes with a uniform diameter of 2 to 4 μm.These slender tubules, in turn, undergo a dynamicbending and branching process and develop peri-odic densities along their length. Eventually, the MK istransformed into a “naked” nucleus surrounded by anelaborate network of proplatelet processes. Megakary-ocyte maturation ends when a rapid retraction sep-arates the proplatelet fragments from the cell body,releasing them into culture (Fig. 1.2C). The subse-quent rupture of the cytoplasmic bridges betweenplatelet-sized segments is believed to release individ-ual platelets into circulation.

The cytoskeletal machineof platelet production

The cytoskeleton of the mature platelet plays a cru-cial role in maintaining the discoid shape of the rest-ing platelet and is responsible for the shape changethat occurs during platelet activation. This same set ofcytoskeletal proteins provides the force to bring aboutthe shape changes associated with MK maturation.107

Two cytoskeletal polymer systems exist in MKs: actinand tubulin. Both of these proteins reversibly assem-ble into cytoskeletal filaments. Evidence supports amodel of platelet production in which microtubulesand actin filaments play an essential role. Proplateletformation is dependent on microtubule function, astreatment of MKs with drugs that take apart micro-tubules, such as nocodazole or vincristine, blocks

proplatelet formation. Microtubules, hollow polymersassembled from α and β tubulin dimers, are the majorstructural components of the engine that powers pro-platelet elongation. Examination of the microtubulecytoskeletons of proplatelet-producing MKs providesclues as to how microtubules mediate platelet produc-tion (Fig. 1.3).108 The microtubule cytoskeleton in MKsundergoes a dramatic remodeling during proplateletproduction. In immature MKs without proplatelets,microtubules radiate out from the cell center to thecortex. As thick pseudopodia form during the initialstage of proplatelet formation, membrane-associatedmicrotubules consolidate into thick bundles situatedjust beneath the plasma membrane of these struc-tures. And once pseudopodia begin to elongate (at anaverage rate of 1 μm/min), microtubules form thicklinear arrays that line the whole length of the pro-platelet extensions (Fig. 1.3B). The microtubule bun-dles are thickest in the portion of the proplatelet nearthe body of the MK but thin to bundles of approxi-mately seven microtubules near proplatelet tips. Thedistal end of each proplatelet always has a platelet-sized enlargement that contains a microtubule bundlewhich loops just beneath the plasma membrane andreenters the shaft to form a teardrop-shaped structure.Because microtubule coils similar to those observedin blood platelets are detected only at the ends of pro-platelets and not within the platelet-sized beads foundalong the length of proplatelets, mature platelets areformed predominantly at the ends of proplatelets.

In recent studies, direct visualization of micro-tubule dynamics in living MKs using green fluorescentprotein (GFP) technology has provided insights intohow microtubules power proplatelet elongation.108

9


A B

Figure 1.3 Structure of proplatelets. (A) Differential interference contrast (DIC) image of proplatelets elaborated by mouse megakaryocytes

in culture. Proplatelets contain platelet-sized swellings that decorate their length giving them a beads-on-a-string appearance. (B) Staining

of proplatelets with Alexa 488-anti-tubulin IgG reveals the microtubules to line the shaft of the proplatelet and to form loops at the

proplatelet tips.

End-binding protein three (EB3), a microtubule plusend-binding protein associated only with growingmicrotubules, fused to GFP was retrovirally expressedin murine MKs and used as a marker to follow micro-tubule plus end dynamics. Immature MKs withoutproplatelets employ a centrosomal-coupled micro-tubule nucleation/assembly reaction, which appearsas a prominent starburst pattern when visualized withEB3-GFP. Microtubules assemble only from the cen-trosomes and grow outward into the cell cortex, wherethey turn and run in parallel with the cell edges.However, just before proplatelet production begins,centrosomal assembly stops and microtubules beginto consolidate into the cortex. Fluorescence time-lapse microscopy of living, proplatelet-producingMKs expressing EB3-GFP reveals that as proplateletselongate, microtubule assembly occurs continuouslythroughout the entire proplatelet, including theswellings, shaft, and tip. The rates of microtubulepolymerization (average of 10.2 μm/min) are approx-imately 10-fold faster than the proplatelet elongationrate, suggesting polymerization and proplatelet elon-gation are not tightly coupled. The EB3-GFP studiesalso revealed that microtubules polymerize in bothdirections in proplatelets (e.g., both toward the tipsand cell body), demonstrating that the microtubulescomposing the bundles have a mixed polarity.

Even though microtubules are continuously assem-bling in proplatelets, polymerization does not providethe force for proplatelet elongation. Proplatelets con-

tinue to elongate even when microtubule polymeriza-tion is blocked by drugs that inhibit net microtubuleassembly, suggesting an alternative mechanism forproplatelet elongation.108 Consistent with this idea,proplatelets possess an inherent microtubule slid-ing mechanism. Dynein, a minus-end microtubulemolecular motor protein, localizes along the micro-tubules of the proplatelet and appears to contributedirectly to microtubule sliding, since inhibition ofdynein, through disassembly of the dynactin complex,prevents proplatelet formation. Microtubule slidingcan also be reactivated in detergent-permeabilizedproplatelets. ATP, known to support the enzymaticactivity of microtubule-based molecular motors, acti-vates proplatelet elongation in permeabilized pro-platelets that contain both dynein and dynactin, itsregulatory complex. Thus, dynein-facilitated micro-tubule sliding appears to be the key event in drivingproplatelet elongation.

Each MK has been estimated to release thousandsof platelets.109,110,111 Analysis of time-lapsed videomicroscopy of proplatelet development from MKsgrown in vitro has revealed that ends of proplatelets areamplified in a dynamic process that repeatedly bendsand bifurcates the proplatelet shaft.106 End amplifica-tion is initiated when a proplatelet shaft is bent into asharp kink, which then folds back on itself, forming aloop in the microtubule bundle. The new loop eventu-ally elongates, forming a new proplatelet shaft branch-ing from the side of the original proplatelet. Loops lead

10


the proplatelet tip and define the site where nascentplatelets will assemble and platelet-specific contentsare trafficked. In marked contrast to the microtubule-based motor that elongates proplatelets, actin-basedforce is used to bend the proplatelet in end ampli-fication. Megakaryocytes treated with the actin tox-ins cytochalasin or latrunculin can only extend long,unbranched proplatelets decorated with few swellingsalong their length. Despite extensive characterizationof actin filament dynamics during platelet activation,yet to be determined are how actin participates in thisreaction and the nature of the cytoplasmic signals thatregulate bending. Electron microscopy and phalloidinstaining of MKs undergoing proplatelet formationindicate that actin filaments are distributed through-out the proplatelet and are particularly abundantwithin swellings and at proplatelet branch points.112

One possibility is that proplatelet bending and branch-ing are driven by the actin-based molecular motormyosin. A genetic mutation in the nonmuscle myosinheavy chain-A gene in humans results in a diseasecalled May-Hegglin anomaly,113,114 characterized bythrombocytopenia with giant platelets. Studies alsoindicate that protein kinase Cα (PKCα) associates withaggregated actin filaments in MKs undergoing pro-platelet formation and that inhibition of PKCα orintegrin signaling pathways prevent the aggregationof actin filaments and formation of proplatelets inMKs.112 However, the role of actin filament dynamicsin platelet biogenesis remains unclear.

In addition to playing an essential role in pro-platelet elongation, the microtubules lining the shaftsof proplatelets serve a secondary function: the trans-port of membrane, organelles, and granules into pro-platelets and assemblage of platelets at proplateletends. Individual organelles are sent from the cellbody into the proplatelets, where they move bidirec-tionally until they are captured at proplatelet tips.115

Immunofluorescence and electron microscopy stud-ies indicate that organelles are intimately associatedwith microtubules and that actin poisons do notdiminish organelle motion. Thus, movement appearsto involve microtubule-based forces. Bidirectionalorganelle movement is conveyed in part by the bipolararrangement of microtubules within the proplatelet,as kinesin-coated latex beads move in both direc-tions over the microtubule arrays of permeabilizedproplatelets. Of the two major microtubule motors,kinesin and dynein, only the plus end–directed kinesin

is localized in a pattern similar to that of organellesand granules and is likely responsible for transportingthese elements along microtubules.115 It appears thata two-fold mechanism of organelle and granule move-ment occurs in platelet assembly. First, organelles andgranules travel along microtubules and, second, themicrotubules themselves can slide bidirectionally inrelation to other motile filaments to move organellesindirectly along proplatelets in a piggyback manner.

In vivo, proplatelets extend into bone marrow vascu-lar sinusoids, where they may be released and enter thebloodstream. The actual events surrounding plateletrelease in vivo have not been identified due to the rar-ity of MKs within the bone marrow. The events lead-ing up to platelet release within cultured murine MKshave been documented. After complete conversionof the MK cytoplasm into a network of proplatelets,a retraction event occurs, which releases individualproplatelets from the proplatelet mass.106 Proplateletsare released as chains of platelet-sized particles, andmaturation of platelets occurs at the ends of pro-platelets. Microtubules filling the shaft of proplateletsare reorganized into microtubule coils as platelets arereleased from the end of each proplatelet. Many ofthe proplatelets released into MK cultures remain con-nected by thin cytoplasmic strands. The most abun-dant forms release as barbell shapes composed of twoplatelet-like swellings, each with a microtubule coil,that are connected by a thin cytoplasmic strand con-taining a microtubule bundle. Proplatelet tips are theonly regions of proplatelets where a single micro-tubule can roll into a coil, having dimensions similar tothe microtubule coil of the platelet in circulation. Themechanism of microtubule coiling remains to be elu-cidated but is likely to involve microtubule motorproteins such as dynein or kinesin. Since platelet mat-uration is limited to these sites, efficient platelet pro-duction requires the generation of a large numberof proplatelet ends during MK development. Eventhough the actual release event has yet to been cap-tured, the platelet-sized particle must be liberated asthe proplatelet shaft narrows and fragments.

Platelet formation in vivo

Although MK maturation and platelet production havebeen extensively studied in vitro, studies analyzingthe development of MKs in their in vivo environmenthave clearly lagged behind. Although MKs arise in the

11


bone marrow, they can migrate into the bloodstream;as a consequence, platelet formation may also occurat nonmarrow sites. Platelet biogenesis has been pro-posed to take place in many different tissues, includ-ing the bone marrow, lungs, and blood. Specific stagesof platelet development have been observed in allthree locations. Megakaryocytes cultured in vitro out-side the confines of the bone marrow can form highlydeveloped proplatelets in suspension, suggesting thatdirect interaction with the bone marrow environmentis not a requirement for platelet production. Never-theless, the efficiency of platelet production in cultureappears to be diminished relative to that observed invivo, and the bone marrow environment composedof a complex adherent cell population could play arole in platelet formation by direct cell contact orsecretion of cytokines. Scanning electron micrographsof bone marrow MKs extending proplatelets throughjunctions in the endothelial lining into the sinusoidallumen have been published, suggesting platelet pro-duction occurs in the bone marrow.116,117 Bone mar-row MKs are strategically located in the extravascularspace on the abluminal side of sinus endothelial cellsand appear to send beaded proplatelet projectionsinto the lumen of sinusoids. Electron micrographsshow that these cells are anchored to the endotheliumby organelle-free projections extended by the MKs.Several observations suggest that thrombopoiesis isdependent on the direct cellular interaction of MKswith bone marrow endothelial cells (BMECs), or spe-cific adhesion molecules.118 It has been demonstratedthat the translocation of MK progenitors to the vicin-ity of bone marrow vascular sinusoids was sufficientto induce MK maturation.119 Implicated in this pro-cess are the chemokines SDF-1 and FGF-4, which areknown to induce expression of adhesion molecules,including very late antigen (VLA)-4 on MKs and VCAM-1 on BMECs.120,121 Disruption of BMEC VE-cadherin–mediated homotypic intercellular adhesion interac-tions results in a profound inability of the vascularniche to support MK differentiation and to act as aconduit to the bloodstream.

Whether individual platelets are released from pro-platelets into the sinus lumen or whether MKs pref-erentially release large proplatelet processes intothe sinus lumen that later fragment into individ-ual platelets within the circulation is not fully clear.Behnke and Forer have suggested that the final stagesof platelet development occur solely in the blood cir-

culation.122 In this model of thrombopoiesis, MK frag-ments released into the blood become transformedinto platelets while in circulation. This theory is sup-ported by several observations. First, the presence ofMKs and MK processes that are sometimes beadedin blood has been amply documented. Megakary-ocyte fragments can represent up to 5% to 20% of theplatelet mass in plasma. Second, these MK fragments,when isolated from platelet-rich plasma, have beenreported to elongate, undergo curving and bendingmotions, and eventually fragment to form disc-shapedstructures resembling chains of platelets. Third, sinceboth cultured human and mouse MKs can form func-tional platelets in vitro, neither the bone marrow envi-ronment nor the pulmonary circulation is essentialfor platelet formation and release.123 Last, many ofthe platelet-sized particles generated in these in vitrosystems still remain attached by small cytoplasmicbridges. It is possible that the shear forces encoun-tered in circulation or an unidentified fragmentationfactor in blood may play a crucial role in separatingproplatelets into individual platelets.

Megakaryocytes have been visualized in intravas-cular sites within the lung, leading to the hypothe-sis that platelets are formed from their parent cell inthe pulmonary circulation.124 Ashcoff first describedpulmonary MKs and proposed that they originatedin the marrow, migrated into the bloodstream, and—because of their massive size—lodged in the capillarybed of the lung, where they produced platelets. Thismechanism requires the movement of MKs from thebone marrow into the circulation. Although the size ofMKs would seem limiting, the transmigration of entireMKs through endothelial apertures of approximately3 to 6 μm in diameter into the circulation has beenobserved in electron micrographs and by early livingmicroscopy of rabbit bone marrow.125,126 Megakary-ocytes express the chemokine receptor CXCR4 andcan respond to the CXCR4 ligand stromal cell–derivedfactor 1 (SDF-1) in chemotaxis assays.127 However,both mature MKs and platelets are nonresponsive toSDF-1, suggesting the CXCR4 signaling pathway maybe turned off during late stages of MK development.This may provide a simple mechanism for retainingimmature MKs in the marrow and permitting matureMKs to enter the circulation, where they can liber-ate platelets.128,129 Megakaryocytes are also remark-ably abundant in the lung and the pulmonary circula-tion and some have estimated that 250 000 MKs reach

12


the lung every hour. In addition, platelet counts arehigher in the pulmonary vein than in the pulmonaryartery, providing further evidence that the pulmonarybed contributes to platelet formation. In humans, MKsare 10 times more concentrated in pulmonary arte-rial blood than in blood obtained from the aorta.130

In spite of these observations, the estimated contri-bution of pulmonary MKs to total platelet productionremains unclear, as values have been estimated from7% to 100%. Experimental results using acceleratedmodels of thrombopoiesis in mice suggest that thefraction of platelet production occurring in the murinelung is insignificant.

Regulation of megakaryocytedevelopment and platelet formation

Megakaryocyte development and platelet formationare regulated at multiple levels by many differentcytokines.131 These mechanisms regulate the nor-mal platelet count within an approximately three-foldrange. Specific cytokines, such as IL-3, IL-6, IL-11, IL-12, GM-CSF, and erythropoietin promote prolifera-tion of progenitors of MKs.132,133 Leukemia inhibitoryfactor (LIF) and IL-1α are cytokines that regulate MKdevelopment and platelet release. Thrombopoietin(TPO), a cytokine that was purified and cloned byfive separate groups in 1995, is the principal regula-tor of thrombopoiesis.134 Thrombopoietin regulatesall stages of MK development, from the hematopoieticstem cell stage through cytoplasmic maturation. Kitligand (KL)—also known as stem cell factor, steel fac-tor, or mast cell growth factor—a cytokine that existsin both soluble and membrane-bound forms, influ-ences primitive hematopoietic cells. Cytokines suchas IL-6, IL-11, and KL also regulate stages of MK devel-opment at multiple levels but appear to function onlyin concert with TPO or IL-3. Interestingly, TPO andthe other cytokines mentioned above are not essentialfor the final stages of thrombopoiesis (proplatelet andplatelet production) in vitro. In fact, thrombopoietinmay actually inhibit proplatelet formation by maturehuman MKs in vitro.135

Apoptosis and platelet biogenesis

The process of platelet formation in MKs exhibits somefeatures related to apoptosis, including cytoskeletalreorganization, membrane condensation, and ruf-

fling. These similarities have led to further inves-tigations aimed at determining whether apoptosisis a mechanism driving proplatelet formation andplatelet release. Apoptosis, or programmed cell death,is responsible for destruction of the nucleus in senes-cent MKs.135 However, it is thought that a special-ized apoptotic process may lead to platelet genera-tion and release. Apoptosis has been documented inMKs137 and found to be more prominent in matureMKs as opposed to immature cells. A number of apop-totic factors, both proapoptotic and antiapoptotic,have been identified in MKs (reviewed in Ref. 138).Apoptosis inhibitory proteins such as Bcl-2 and Bcl-xLare expressed in early MKs. When overexpressedin MKs, both factors inhibit proplatelet formation.Bcl-2 is absent in mature blood platelets and Bcl-xlLis absent from senescent MKs,140 consistent with arole for apoptosis in mature MKs. Proapoptotic fac-tors, including caspases and nitric oxide (NO), are alsoexpressed in MKs. Evidence indicating a role for cas-pases in platelet assembly is strong. Caspase activa-tion has been established as a requirement for pro-platelet formation. Caspases 3 and 9 are active inmature MKs and inhibition of these caspases blocksproplatelet formation.139 Nitric oxide has been impli-cated in the release of platelet-sized particles from themegakaryocytic cell line Meg-01 and may work in con-junction with TPO to augment platelet release.141,142

Other proapoptotic factors expressed in MKs andthought to be involved in platelet production includeTGFβ1 and SMAD proteins.143 Of interest is the distinctaccumulation of apoptotic factors in mature MKs andmature platelets.144 For instance, caspases 3 and 9 areactive in terminally differentiated MKs. However, onlycaspase 3 is abundant in platelets,145 while caspase9 is absent.144 Similarly, caspase 12, found in MKs, isabsent in platelets.146 These data support differentialmechanisms for programmed cell death in plateletsand MKs and suggest the selective delivery and restric-tion of apoptotic factors to nascent platelets duringproplatelet-based platelet assembly.

THE STRUCTURE OF THEACTIVATED PLATELET

Platelets, in response to vascular damage, undergorapid and dramatic changes in cell shape, upregulatethe expression and ligand-binding activity of adhesionreceptors, and secrete the contents of their storage

13


A B

Figure 1.4 The resting to active transition of platelets. (A) Differential interference contrast micrographs comparing (A) discoid resting

platelets in suspension to platelets activated by contact to the glass surface and exposure to thrombin. (B) As platelets activate on the surface,

they spread using lamellipodia and form long finger-like filopodia.

granules.147,148 A variety of agonists can activateplatelets, including thrombin, TXA2, ADP, collagen,and VWF.

The platelet shape change

When platelets are exposed to specific agonists, theyconvert from discs to spheres with pseudopodia ina matter of seconds (Fig. 1.4). This shape changeis highly reproducible and follows a sequence ofevents in which the disc converts into a sphere, afterwhich broad lamellipodia and thin finger-like filopo-dia extend from the platelet surface. These shapechanges are driven by the rapid remodeling of theplatelet cytoskeleton. Protrusion of lamellipodia andfilopodia is dependent upon the new assembly of actinfilaments. As the activated platelet sends out pro-cesses, the microtubule coil and intracellular granulesare compressed into the center of the cell.

The conversion of the disc into a roundedshape occurs if cytoplasmic calcium levels rise intothe micromolar levels.149 Resting platelets main-tain cytosolic calcium at 10 to 20 nM.150 Ligandbinding to serpentine receptors activates phos-pholipase Cβ, which hydrolyzes membrane-bound

polyphos phoinositol-4,5-bisphosphate to inositol1,4,5 triphosphate (IP3) and diacylglycerol.151 IP3 thenbinds to receptors on the dense tubular system, induc-ing the release of calcium. The rise in intracellular cal-cium is then used to activate a filament-severing reac-tion that powers the disc to sphere transition. Althoughcalcium can affect the activity of a variety of proteins,one of the key platelet proteins that is activated is gel-solin.152 Gelsolin is an 80-kDa protein present at a con-centration of 5 μm. When calcium binds to gelsolin,it causes gelsolin to attach to an actin filament andsever it.153 The gelsolin then remains bound to thenewly generated filament end. The severing of the fila-ments releases the constraints imposed by the GPIbα-filamin-actin filament linkage and allows the mem-brane skeleton to expand and the platelet to convertto a disc. The critical importance of gelsolin in thisfunction has been demonstrated using platelets frommice that specifically lack gelsolin.152

The rounding of the platelet is followed by the rapidprotrusion of lamellipodia and filopodia. The forma-tion of platelet lamellipodia and filopodia requires theassembly of actin filaments. During platelet activa-tion, the actin filament content doubles from a restingplatelet concentration of 0.22 mM to 0.44 mM. In the

14


TAKE-HOME MESSAGES

� Nearly a trillion platelets circulate in an adult human.� Platelets function as the “band-aids” of the bloodstream� The discoid shape of resting platelets is maintained by a cytoskeleton composed of microtubules, actin filaments,

and a spectrin-based membrane skeleton.� Megakaryocytes undergo endomitosis to increase ploidy.� Megakaryocytes produce platelets by remodeling their cytoplasm into long cytoplasmic projections called pro-

platelets.� Microtubule-based forces power the elongation of proplatelets.� The lamellipodial and filopodial formation that accompanies platelet activation is driven by the actin cytoskeleton.

resting platelet, actin is stored in a monomeric com-plex with β4-thymosin and profilin. Actin assemblyoccurs only from the barbed ends of actin filaments.153

Actin forms polarized filaments that have a clearlydefined directionality. The two ends of an actin fila-ment have different affinities for actin monomer, withthe barbed end having a 10-fold affinity for monomer.This arrangement biases the polymerization reactionfor the barbed end of the growing filament. The fil-ament severing reaction that powers the cell round-ing is followed by the formation of actin nuclei thatinitiate the assembly of new actin filaments beneaththe plasma membrane. This new actin polymerizationprovides the force to push out the finger-like filopo-dia and lamellipodia. The new actin assembly occurswhen gelsolin and other proteins that cap the barbedends of actin filaments are removed and a complexof proteins called the Arp 2/3 complex is activated togenerate new barbed ends.

While the polymerization of actin filaments at theplasma membrane powers the membrane outward, itis the arrangement of the actin filaments that estab-lishes the shape of the protrusion. Filopodia are com-posed of tight bundles of actin filaments that origi-nate near the center of the platelet. The bundles areloosely connected in the middle of the platelet but thenbecome zipped together as they reach the edge of thecell. Filopodia extended by platelets appear to be usedto locate other platelets and strands of fibrin. Plateletshave been observed to rapidly wave and rotate filopo-dia around their periphery and these are also used toapply the myosin-generated contractile force in fib-rin gels. The lamellipodia of the spread platelet areorganized into a dense three-dimensional meshworkof cross-linked actin filaments. This orthogonal net-

work is biologically efficient because it uses the min-imal amount of filament to fill a cytoplasmic volume.The filaments are cross-linked by a protein called fil-amin,31 which binds actin filaments into orthogonalnetworks in vitro and organizes these arrays in theplatelet’s cortex.

Granule secretion

Activation of a platelet is accompanied not only bythe massive reorganization of the actin cytoskeletonbut also by the exocytosis of the platelet storage gran-ules. The contents released from α and dense gran-ules enhance the platelet plug reaction by attractingadditional platelets to the wound. During activation,the majority of granules release their contents intothe open canalicular system. Because of the complextunneling of the open canalicular system, granulesare always positioned in close proximity to the OCS.The fusion and release of granule mediators is depen-dent on a rise of cytosolic calcium into the micro-molar range and is diminished by calcium chelatingagents. Calcium-calmodulin activates myosin light-chain kinase to phosphorylate myosin II.149 The acti-vation of the contractile activity of myosin II generatesa centripetal collapse of the granules into the middleof the cell, promoting the fusion of the granules bybringing them into close contact with the OCS.

FUTURE AVENUES OF RESEARCH

Future research into the biology of MKs and plateletswill undoubtedly provide new insights into how thesecells function and may lead to novel applications.Intravital microscopy of fluorescently labeled MKs

15


should allow us to visualize MKs producing plateletsin the bone marrow. Although many of the majorcytokines that promote MK development have beenidentified, molecules and signals that initiate plateletproduction have not been defined. Identification ofthe signals that instruct MKs to produce plateletsmay yield strategies to promote thrombocytogenesisin vivo. Additional studies into how the bone marrowenvironment nurtures MKs and influences plateletproduction may ultimately lead to the large-scale pro-duction of platelets in vitro.

REFERENCES

1. Nachmias VT, Yoshida K. The cytoskeleton of the blood

platelet: a dynamic structure. Adv Mol Cell Biol

1988;2:181–211.

2. Zucker-Franklin D. The ultrastructure of megakaryocytes

and platelets. In Gordon AS (ed). Regulation of

hematopoiesis. New York: Appleton-Century-Crofts.

1970:1553–86.

3. Flaumenhaft R. Molecular basis of platelet granule

secretion. Arterioscler Thromb Vasc Biol 2003;23:1152–60.

4. Jones OP. Origin of megakaryocyte granules from Golgi

vesicles. Anat Rec 1960;138:105–14.

5. de Larouziere V, Brouland JP, Souni F, Drouet L, Cramer E.

Inverse immunostaining pattern for synthesized versus

endocytosed alpha-granule proteins in human bone

marrow megakaryocytes. Br J Haematol 1998;101:618–25.

6. Coller BS, Seligsohn U, West SM, Scudder LE, Norton KJ.

Platelet fibrinogen and vitronectin in Glanzmann

thrombasthenia: evidence consistent with specific roles for

glycoprotein IIb/IIIa and alpha v beta 3 integrins in platelet

protein trafficking. Blood 1991;78:2603–10.

7. Handagama P, Bainton DF, Jacques Y, Conn MT, Lazajus

RA, Shuman M. Kistrin, an integrin antagonist, blocks

endocytosis of fibrinogen into guinea pig megakaryocyte

and platelet alpha-granules. J Clin Invest 1993;91:193–200.

8. Handagama P, Scarborough RM, Shuman MA, Bainton DF.

Endocytosis of fibrinogen into megakaryocyte and platelet

alpha-granules is mediated by alpha IIb beta 3

(glycoprotein IIb-IIIa). Blood 1993;82:13

PLATELETS IN HEMATOLOGIC AND CARDIOVASCULAR DISORDERS…the-eye.eu/public/WorldTracker.org/College Books... · ment in hematologic, cardiovascular, and inﬂamma-tory disorders and

Documents