Somitogenesis

ADVANCES IN EXPERIMENTAL MEDICINE AND BIOLOGY

Editorial Board:

NATHAN BACK, State University ofNew York at Buffalo

IRUN R. COHEN, The Weizmann Institute ofScience

ABEL LAJTHA, N.S. Kline Institute for Psychiatric Research

JOHN D. LAMBRIS, University ofPennsylvania

RODOLFO PAOLETTI, University ofMilan

Recent Volumes in this Series

Volume 630INNOVATIVE ENDOCRINOLOGY OF CANCER

Edited by Lev M. Berstein and Richard J. Santen

Volume 631BACTERIAL SIGNAL TRANSDUCTION

Edited by Ryutaro Utsumi

Volume 632CURRENT TOPICS IN COMPLEMENT II

Edited by John. D. Lambris

Volume 633CROSSROADS BETWEEN INNATE AND ADAPTIVE IMMUNITY II

Edited by Stephen P. Schoenberger, Peter D. Katsikis, and Bali Pulendran

Volume 634HOT TOPICS IN INFECTION AND IMMUNITY IN CIDLDREN V

Edited by Adam Finn, Nigel Curtis, and Andrew J. Pollard

Volume 635GI MICROBIOTAAND REGULATION OF THE IMMUNE SYSTEM

Edited by Gary B. Huffnagle and Mairi Noverr

Volume 636MOLECULAR MECHANISMS IN SPERMATOGENESIS

Edited by C. Yan Cheng

Volume 637MOLECULAR MECHANISMS IN XERODERMA PIGMENTOSUM

Edited by Shamin Ahmadand Fumio Hanaoka

Volume 638SOMITOGENESIS

Edited by Miguel Maroto and Neil V. Whittock

A ContinuationOrder Plan is availablefor this series.A continuation order will bring delivery of each newvolume immediately upon publication. Volwnesare billed only upon actual shipment. For further informationplease contact the publisher.

SomitogenesisEdited by

Miguel Maroto, PhD

University ofDundee, Dundee, Scotland

Neil V.Whittock, PhD

University ofExeter, Devon, England

Springer Science+Business Media, LLC

Landes Bioscience

Springer Science+Business Media, LLCLandes Bioscience

Copyright ©2008 Landes Bioscience and Springer Science-Business Media , LLC

All rights reserved.No part ofthis book may be reproduced or transmitted in any form or by any means, electronic or mechani­cal, including photocopy, recording, or any information storage and retrieval system, without permissionin writing from the publisher, with the exception ofany material supplied specifically for the purpose ofbeing entered and executed on a computer system; for exclusive use by the Purchaser of the work.

Printed in the USA

Springer Science-Business Media, LLC, 233 Spring Street, New York, New York 10013, USAhttp://www.springer.com

Please address all inquiries to the publishers:Landes Bioscience, 1002 West Avenue, Austin, Texas 7870 I, USAPhone : 512/6375060; FAX: 512/6376079http://www.landesbioscience.com

Somitogenesis, edited by Miguel Maroto and Neil V. Whittoek, Landes Bioscience I SpringerSeience+Business Media, LLC dual imprint I Springer series: Advances in Experimental Medicine andBiology

ISBN: 978-0-387-09605-6

While the authors, editors and publisher believe that drug selection and dosage and the specifications andusage of equipment and devices, as set forth in this book, are in accord with current recommendationsand practice at the time of publication, they make no warranty, expressed or implied, with respect tomaterial described in this book. In view of the ongoing research, equipment development, changes ingovernmental regulations and the rapid accumulation ofinformation relating to the biomedical sciences,the reader is urged to carefully review and evaluate the information provided herein.

Library of Congress Cataloging-in-Publication Data

Somitogenesis I edited by Miguel Maroto , Neil V. Whittock.p. ; em. -- (Advances in experimental medicine and biology; v. 638)

Includes bibliographical references and index.ISBN 978-0-387-09605-6I. Somite. I. Maroto, Miguel. II. Whittock, Neil V. III. Series.[DNLM : I . Somites-physiology. 2. Morphogenesis--physiology. WI AD559 v.638 2008 I WQ 205

S69612008]QL971.S5852008571.8 '636--dc22

2008023385

PREFACE

We visualise developmental biology as the study of progressive changes thatoccur within cells, tissues and organisms themselves during their life span. A goodexample ofa field ofdevelopmental biology in which this concept is encapsulated isthat ofsomitogenesis. The somite was identified as the primordial unit underlying thesegmented organisation ofvertebrates more than two centuries ago. The spectaculardiscoveries and achievements in molecular biology in the last fifty years have createda gene-based revolution in both the sorts ofquestions as well as the approaches onecan use in developmental biology today. Largely as a result of this, during the 20thand 21st centuries this simple structure, the somite, has been the focus ofa delugeofpapers addressing multiple aspects ofsomite formation and patterning both at thecellular and molecular level. One ofthe main reasons for such interest in the processofsomitogenesis stems from the fact that it is such an exquisitely beautiful exampleofbiology working under strict temporal and spatial control in a reiterative mannerthat is highly conserved across the vertebrate classes.

Our intention is that this book will be ofinterest to different kinds of scientists,including basic researchers, pathologists, anatomists, teachers and students workingin the fields ofcell and developmental biology. The nine chapters cover a wide arrayof topics that endeavour to capture the spirit of this dynamic and ever-expandingdiscipline by integrating both contemporary research with the classical embryologicalliterature that concentrated on descriptions ofmorphological changes in embryos andthe interactions ofcells and tissues during development. Inso doing they encompassthe main aspects of somitogenesis across four vertebrate classes (frog, fish, mouseand chick) and the hope is that this will enable readers to acquire an appreciation ofthis developmental process in all its facets. Each ofthe different animal models offersalternative strategic approaches (including experimental embryology, genetics andcell biology) to tackle the same process and as such each offers an invaluable andunique insight into different aspects of somitogenesis. The topics described in thesechapters cover the generation of somitic tissue during gastrulation, the molecularmechanisms by which the unsegmented pre-somitic mesoderm becomes segmentedinto somites, the generation of polarity within somites and the means by which thesomite is directed to differentiate into a number of different cell derivatives. There

v

vi Preface

are also two chapters devoted to describing the latest developments on relating spon­taneous mouse mutations and mutations leading to abnormal vertebral segmentationin man to the molecular mechanisms already identified as being crucial for somiteformation in the lower vertebrates.

We would like to heartily thank all ofthe authors that have contributed their timeand effort and whose work has made this book possible. Only they know how difficultit has been to conclude a book that we started more than four years ago, initially withanother publisher. One year ago the project was cancelled and then re-started againwhen the people of Landes accepted for publication. I am especially grateful forthe understanding and patience ofCynthia Conomos and Celeste Carlton who haveworked tirelessly to ensure that this project reached completion. We made it!

Miguel Maroto, PhD

Neil V. Whitlock, PhD

ABOUT THE EDITORS...

MIGUEL MAROTO is a MRC Career Development Fellow and Lecturer at theUniversity ofDundee, UK. He received his PhD in Biochemistry and Molecular Biol­ogy from the Department ofBiochemistry ofthe UniversidadAutonoma ofMadrid,Spain. His research interests include investigating the biochemical basis ofdifferentsignalling mechanisms implicated in the acquisition of specific cell fates duringvertebrate development. In recent years he has been involved in the analysis of themechanism of the molecular clock in the control of the process of somitogenesis.

vii

ABOUT THE EDITORS...

NEIL V. WHITTOCK received his PhD in Human Molecular Genetics whilstworking at Guys' and St Thomas' Hospitals in London, UK. His research focussedon developing diagnostic genetic tests for Duchenne muscular dystrophy before mov­ing on to identifying genes involved in bullous skin disorders. He then continuedhis research as a postdoctoral fellow at the University of Dundee before arrivingat the University of Exeter where he spent three years working alongside Dr PeterTurnpenny. The work at Exeter focussed on the identification of genes involved inhuman genetic disorders that affected the development ofthe spine and ribs, specifi­cally the spondylocostal dysostoses. He now works as an Ambulance Technician inDevon, UK, and runs his own antique clock restoration business.

viii

PARTICIPANTS

Beate Brand-SaberiDepartment of Molecular EmbryologyInstitute for Anatomy

and Cell BiologyFreiburgGermany

Yasumasa BesshoGraduate School ofBiological

SciencesNara Institute of Science

and TechnologyTakayama, IkomaJapan

Gavin ChapmanDevelopmental Biology ProgramVictor Chang Cardiac

Research InstituteUniversity ofNew South WalesSydney, New South WalesAustralia

Bodo ChristInstitut fur Anatomie

und Zellbiologie IIUniversitat FreiburgFreiburgGermany

J. Kim DaleCollege of Life SciencesUniversity of DundeeDundee, ScotlandUK

Sally L. DunwoodieDevelopmental Biology ProgramVictor Chang Cardiac

Research InstituteUniversity ofNew South WalesSydney, New South WalesAustralia

Harun ElmasriDepartment of Physiological

Chemistry I BiocenterUniversity ofWuerzburgGermanyandChildrens HospitalHarvard Medical SchoolBoston, MassachusettsUSA

Anton J. GamelInstitute for Anatomy

and Cell BiologyDepartment of Molecular EmbryologyFreiburgGermany

ix

x

Tadahiro limuraStowers Institute for Medical ResearchKansas City, MissouriUSA

Kenro KusumiSchool of Life SciencesArizona State UniversityTempe, ArizonaUSA

Miguel MarotoCollege of Life SciencesUniversity of DundeeDundee, ScotlandUK

Megan L. O'BrienJohns Hopkins Bloomberg School

of Public HealthBaltimore, MarylandUSA

Stefan RudloffDepartment of Molecular EmbryologyInstitute for Anatomy

and Cell BiologyFreiburgGermany

YumikoSagaDivision of Mammalian DevelopmentNational Institute ofGeneticsMishimaJapan

Martin ScaalInstitut fur Anatomie

und Zellbiologie IIUniversitat FreiburgFreiburgGermany

Participants

William SewellSchool of Life SciencesArizona State UniversityTempe, ArizonaUSA

Duncan B. SparrowDevelopmental Biology ProgramVictor Chang Cardiac Research

InstituteDarlinghurst, New South WalesAustralia

Yu TakahashiCellular & Molecular Toxicology

DivisionNational Institute of Health SciencesTokyoJapan

Peter D. TurnpennyPeninsula Medical SchoolClinical Genetics DepartmentRoyal Devon & Exeter HospitalGladstone RoadExeterUK

Neil V. WhittockUniversity of ExeterDevon, EnglandUK

Christoph WinklerDepartment of Physiological

Chemistry I BiocenterUniversity ofWuerzburgGermanyandDepartment of Biological SciencesNational University of SingaporeSingapore

CONTENTS

1. FORMATION AND DIFFERENTIATION OF AVIAN SOMITEDERIVATIVES.•.•...................•..•.......••••••••••••••.••••••••••.•••••••••••••••••••••••....•. 1

Bodo Christ and Martin Scaal

Abstract 1Introduction 1Dermomyotome 2Sclerotome 14Outlook••••.••.•..•.••••...•..•.••.•..•...•.••.•..•.••.•.•••.•••••.•.••.••.•..•....•..•...••••.•..•.•.•.••.•..•...••••.••.•.•••••••.••. 29

2. AVIAN SOMITOGENESIS: TRANSLATING TIME AND SPACEINTO PATTERN...••••••.••••..•.•.....................................•••••••••••••••••••••••••••• 42

Beate Brand-Saberi, Stefan Rudloff and Anton J. Gamel

Abstract...•..•••••••.•.••.....•..•.••......•••.••••..•.••....••..•.••.•..•.••.•..•..••.••....•.••.•..•••.•.••.•..•....•••....••••.•..• 42Introduction 42Epithelialization of the Segmental Plate 43The Anterior Posterior Polarization of the Paraxial Mesoderm 46Resegmentation of the Somitic Derivatives ....•.••.•..••••••..•••.•.••.•..•.•.•.••....••..•.••.•.•.•..•.••.•... 49Regionalization of Somites and Segmental Plate 49Oscillations in Gene Expression Underlying Somitogenesis 50Conclusion and Future Considerations 52

3. GENETIC ANALYSIS OF SOMITE FORMATIONIN LABORATORY FISH MODELS 58

Christoph Winkler and Harun Elmasri

Abstract 58Introduction.••.•..•.•••....•..•.••.•••....•.••.•..•...•.••••..•.••••..•..•.•.•.•..•.••.•....•••....•••.•..•...•.••.•••....••••.•..• 58Genetically Dissecting the Clock in Zebrafish: The DeltaINotch

Somitogenesis Mutants .•.••.•..••••.•••••.•••••.•.••••••••••••.•.••.••••••••.••.•..•••.•.••.•••.•..•.••••..•••••••.. 60Dissecting the Wave Front in Zebrafish: FGF Signalling and Tbx24 62Other Pathways Implicated in Somitogenesis 64

xi

xii Contents

Medaka: A Model Complementary to Zebrafish 64Somite Formation in the Teleost Medaka .••.•..•....•..•.••.•..•..•....•..•....•...•....•..•.••.•..•.••.•..•.•.• 65Medaka Somitogenesis Mutants 66Medaka Somite Mutants with PSM Prepatterning Defects 67Medaka Mutants with Defective Somite Polarity 67Conclusions and Outlook 69

4. OLD WARES AND NEW: FIVE DECADES OF INVESTIGATIONOF SOMITOGENESIS IN XENOPUS LAEVIS 73

Duncan B. Sparrow

Abstract....••.•..•.•.•.••.•..•.••.•..•.••.•..•.••.•..•.••........•.••.•..•.••.•..•.••.•.......•..•.••.•..•.•......•.....•.•....•••...• 73Introduction 73Structure of the PSM ..••.••...••.•..•...••..•...•..•.•..•.••..•.•.•..•....••.•.••.••.•..•.••.•..•.••.•..•.••.•..•.••.•..•.••. 76Morphological Descriptions of Segmentation in Xenopus •••..••••••••••••••.•••••••••••••••••••••••••• 77A Comparison of Xenopus Segmentation with that of Amniote Vertebrates.••.•..•.••.••.• 78What Controls Where the Somitic Furrow Forms? 78Evidence for Segmental Prepatterning of the PSM 79Cycling Genes--Evidence of a "Clock"? ....•.••.•..•.••.•..•.••.•..•.••.•..•.••.•..•.••.•..•..•.....••....••...• 81Embryological Insights into the Nature of the "Wavefront" 85The Molecular Nature of the Wavefront..•..••••.••.•.•..•..•.••.•..•.••.•..•.••.•..•.••.•..•.••.•..•.••.•..•.••. 86What are the Morphomechanical Mechanisms Required for Somite Separation? .••.• 87EnaNASP 89Conclusion .•••••••••.••.•..•..•.•..•.••.•..•.••.••.•...••.•••..••••.•..•...•••.•.•••••.•.••....•.••....•....•..•..•.....•••..•.•.•.. 90

5. ROLE OF DELTA-LIKE-3 IN MAMMALIAN SOMITOGENESISAND VERTEBRAL COLUMN FORMATION 95

Gavin Chapman and Sally L. Dunwoodie

Abstract 95Introduction..•..••.........•..•.•..•..•.••...........•.•.••.••.•.••...•••.....•.•••...•.•.•..•.....•......•..•.•..•.••.•....•••..••95Somitogenesis•••••.•..•.•..•..•.••••..•....•..•...••••....•.••.•••••.•••..•.•.•.••.•....••.•..•.••.....•..•.•..•.••.••.••.•.••••..• 95The Notch Signalling Pathway 97Notch Signalling and Somite Formation 99DIll and DIl3 Perform Different Functions during Somitogenesis

in Mammals 102Dll3 Conclusions and the Future .......••.•..•.•••.••.•..•.•.••..•.••.••.•.••.•..••......•.••.•.••••......•....••... 106

6. MESP-FAMILY GENES ARE REQUIRED FOR SEGMENTALPATTERNING AND SEGMENTAL BORDER FORMATION••......113

Yumiko Saga and Yu Takahashi

Abstract••.•••.•...•..•.••.•..•.••••.....•.•..•.••.•..•.••.•..•.•.•.••.•..•.••.•..••••.•.......•..•.••.•..•.•..•....••.•.••.••.•..•.• 113Introduction•...•..•.••.•..•..•.•••..••.•••..••.••..•..•.••.•.•••...••...•..•.••...•••.•....•..•.••.•..•..•.•..•.•....••••.•.••... 113Background of Mesp1 and Mesp2 .••.•..•.••.•••.•..••••.••.•.••.••.•..•.•..•.••.•..•.••.•..•.••..•.•.••.•..••••.•• 114Function of Mesp2 during Somitogenesis 114Mesp2 Is also Involved in the Segment Border Formation...•.••.•..•..........•.•..•.••.•••••.••.• 118Regulation of Mesp2 Expression during Somitogenesis••.......••.•.•..•..•.••.•..•.••.•..•.••••••••. 120Functional Redundancy between Mesp1 and Mesp2 120

Contents xiii

Mesp Genes in the Other Vertebrates ••••••••.•.••.•••..•.•.••.•..•.•.•.••.•..•...•.••.•••••..•.•••••.••..•.•••.. 121Perspective 122

7. BHLH PROTEINS AND THEIR ROLE IN SOMITOGENESIS...•.....•. 124

Miguel Maroto, Tadahiro Iimura, J. Kim Dale and Yasumasa Bessho

Abstract 124Introduction•..•.•..•..••...•..••..•.••.•.•••..•....•.•.•..•.••.•..••••.•...•..•.••.•..•...••••••..••••••.•.•..•.••.•..•.••.•..••. 124The HairylE(Spl) Related Factors 125pMesogeninl Family: pMesogeninl, Mespo and cMespo .•.••.•..••..•.•••.•••.•.•...•.••.....•..•.• 127The Mesp Family: Mesp, Meso and Thylacine••.•......•...•.•..•....•...•..•••.•.••.•..•.••.•....••.•..••. 127Paraxis 131Myogenic Regulatory Factors (MRFs).••..•.••.••.•..•.•••..•....•.••.•..•.•••.••.•..••..•.••.•..•.•••...•.••.•. 131bHLH Factors Involved in the Control of Myogenesis..•.•..•..•...•.••.•..••.••.••.•..•..•.•..•.••••. 133Conclusions ..••.•..•.••.•..•.••...•.•.....•.....•..•.••..•.......••.•..•.•.•....•..•.••.•....••..••.•..•...•.••.•..•...•.••..••.•. 134

8. MOUSE MUTATIONS DISRUPTING SOMITOGENESISAND VERTEBRAL PATTERNING 140

Kenro Kusumi, William Sewell and Megan L. O'Brien

Introduction 140Notch Signaling Pathway .••••.•.•.•..•....•...•..•.••.•..•••••..•.•..•..•.•..•......•.•..•••.•.•..••.•......•.••..••.•.. 142Canonical Wnt Signaling Pathway 151Noncanonical Wnt Signaling Pathway••••....•.••.•••..•.........•.••••.••••.•.••.••.••.••.••.•..•.••.•..•.••••• 151FGF (Fibroblast Growth Factor) Signaling Pathway 152RA (Retinoic Acid) Signaling Pathway .••.••.•.••.•.•....•..•.•..•.....••.•..••••.•...•..•.••.•..•.••.•..•.••.•. 152BMPITGF-beta Signaling Pathway.•.••.•........••.•..•...•.......•.••.•........••.•..••.......•......••....•..•. 152Hedgehog Signaling Pathway 152Snail Transcriptional Repressors 153T-BoxTranscription Factors 153Homeobox Transcription Factors 153Paired-Box Transcription Factors 155Forkhead Transcription Factors 155Basic Helix-Loop-Helix (bHLH) Transcription Factors.•.••...•.•..•.•.••••.•..•.••...••.••..••.•..•. 155Protocadherins •••••.•.••••.••.•.•.••.....•..•.••.•......•.••.••.••..•.••....•.••.•..•.••.•...••••..••.••.•..•.••....••.••.••.•. 156Other Genes ..•........•.......•.....••.•.••...•.••.•••....•.••.••••.....••.............•.....•.•..•...•.•.....•.••.•..••..•.••.•. 156Conclusion ..•..•.••.•..•.••.•..•...•.••.•.............•.•................••....•...•.•.••..•..•.•..•...•.••.•..•...••••••..•.•••.• 156

9. DEFECTIVE SOMITOGENESIS AND ABNORMAL VERTEBRALSEGMENTATION IN MAN 164

Peter D. Turnpenny

Abstract 164Introduction 164Notch Signaling Pathway Genes and Spondylocostal Dysostosis 169Concluding Remarks 184

INDEX••••••••••••••••.•.....••.•........•••••••••••••••••••••.•••............•.....•..••...•••••••••••••••••••.•••••• 191

CHAPTERl

Formation and Differentiation ofAvianSomite DerivativesBodo Christ and Martin Scaal"

Abstract

D uring somite maturation, the ventral halfof the epithelial somite disintegrates into themesenchymal sclerotome, whereas the dorsal halfforms a transitory epithelial sheet, thedermomyotome, lying in between the sclerotome and the surface ectoderm. The dermo­

myotome is the source ofthe majorityofthe mesodermal tissues in the body, giving rise to cell typesas different as muscle, connective tissue, endothelium and cartilage. Thus, the dermomyotome isthe most important turntable ofmesodermal cell fate choice in the vertebrate embryo. Sclerotomedevelopment is characterized by a cranio-caudal polarization, resegmentation and axial identity.Its formation is controlled by signals from the notochord, the neural tube, the lateral plate meso­derm and the myotome. These signals and cross-talk between somite cells lead to the separationofvarious subdomains, like the central, ventral, dorsal and lateral sclerotome. Here, we discuss thecurrent knowledge on the formation of the dermomyotome and the mechanisms leading to thedevelopment ofthe various dermomyotomal derivatives,with special emphasis on the developmentofmusculature and dermis. We further discuss the molecular control ofsclerotomal subdomainformation and cell type specification.

IntroductionDuringembryonic development, the nascent organismdevelops from the unicellular zygote into

the fully developed body, which is composed ofa multitude ofdifferent tissues intricately arrangedas fUnctionally interacting organs. Vertebrate embryos differ greatly in their early developmentalstages, e.g., the gastrulae ofamphibia and mammals, but their developmental program convergesin the bottleneck ofthe pharyngula, which represents the general Bauplan ofthe vertebrate bodyand is therefore designated as the vertebrate phylotypic stage, before it diverges again to build thediverse anatomical concepts realized within the chordate phylum.P

The vertebrate phylotypic stage is characerized, among other features, by the segmentalorgani­zation ofthe paraxial mesoderm which is arranged as metameric balls ofepithelial spheres alignedalong both sides ofthe axial organs, the notochord and the neural tube (Fig. 1). These segmentalportions ofparaxial mesoderm, the somites, give rise to the majority ofmesodermal organs ofthebody wall. The term "somite" was introduced by Balfour" to characterize these segmental units ofthe paraxial mesoderm which had been previously described as protovertebrae.4 The first detaileddescription ofthe avian somite was given by Williams.s

The somites bud offfrom the presomitic mesoderm (segmental plate) in cranio-caudal direc­tion.6•

7 Prior to morphologicallyovert segmentation, the cranial part ofthe presomitic mesodermundergoes a mesenchymal-to-epithelial transition (MET), with the exception ofthe cells located

*CorrespondingAuthors: BodoChrist,MartinScaal-Institut fUr Anatomie und Zellbiologie IIUniversitat Freiburg, Albertstr. 17, 79104 Freiburg, Germany.Email: bodo.christ@anat.uni-freiburg.de.martin.scaal@anat.uni-freiburg.de

Somitogenesis, edited by Miguel Maroto and Neil V. Whittock. ©2008 Landes Bioscienceand Springer Science+Business Media.

2 Somitogenesis

inside which retain their mesenchymal organization to form the cells of the somitic core, thesomirocoele cells. The other cells rearrange to surround these core cells as an epithelial ball . Thesomites constitute a metameric pattern within the embryonic body wall that determines the seg­mental arrangement ofthe vertebral column, ribs, muscles, tendons. ligaments. dorsal root ganglia.peripheral nerves and blood vessels .v? The metamerism of these structures is the prerequisite forthe ability ofthe vertebrate body to perform bending and rotating movements.

Each somite gives rise to various cell lineages such as myocytes, chondrocyres, fibroeytes andother cell types which differentiate within different subdomains ofthe somite. lO.l 1 Ifthe early epi­thelial somite is subjected to a dorso-ventral rotation12 or to an exchange ofthe dorsal and ventralmoieties13•

14 or medial and lateral moieties '? the somite develops normally corresponding to theadjacent nonrotared somites indicating that the axes ofthese somites are still labile and the cellsare still rnulripotent. These results clearly show that environmental signals control the formation ofthe dermomyotome and sclerotome as well as their subdomains and the differentiation program oftheir cells.6,l6 The determination ofthe cranio-caudal sclerotome polarization, on the other hand,occurs already in the presomitic mesoderm and is mediated by the Delta-Notch signalingpathway.Lateral to the paraxial mesoderm lies the intermediate and lateral plate mesoderm and dorsallythe surface ectoderm is forming the primitive integument. The ontogenetic conservation of thisembryonic Bauplan is crucial, because the specification and coordinate development ofthe greatvariety ofderivatives originating from the somites is regulated by a complex interplay of signalssecreted by neighboring structures.The term "sclerotome" was created by Hatschek'?who workedas an anatomist at the Charles University ofPrague. Hatschek corrected the assumption by His18

that the cells ofthe sclerotome only contribute to the formation ofthe aortic wall.In the first part of this review, we will focus on the dermomyotorne, which represents the

dorsal portion of the developing somite and we will give an overview on the current knowledgeabout the formation and subsequent differentiation ofthis pivotal embryonic structure into der­mis, muscle, endothelium and cartilage. In the second part, we will address the formation of thesclerotome from the ventral halfofthe epithelial somite and from the mesenchymal somitocoelecells. We will describe the development ofdistinct sclerotomal sub domains and their subsequentdifferentiation.

Dermomyotome

Morphogenesis ofthe DermomyotomeSemites develop progressively in caudal to cranial direction. so that the caudalmost sornites

represent the youngest stages and the oldest stages are gradually aligned more cranially.Accordingly.Christ and Ordahl'' established a dynamic staging system denominating the youngest somite assomite number I and the cranially abutting somites in consecutive roman numbers.

Up to somites III to IV, somites are organized as epithelial spheres enclosing a lumen filledwith somitocoele cells. The outer surface ofthese spheres is formed by the basal pole ofthe somiticepithelium and is covered with a basement membrane, while the apices of the somite cells linethe somitocoele. Shortly afrer the deepithelialization ofthe sclerotome. from stage V to VII, theremaining dorsal somitic epithelium stretches to form a slightly oblique epithelial sheet that roofsthe sclerotomal mesenchyme dorsolaterally, the dermomyotome (Fig. 2) .

In dorsolateral view. the dermornyorome adopts a more and more rectangular, sheet-like form,thus abutting closely to the dermomyorornes ofthe adjacent segments from which it is separatedby a thin intersegmental artery. All four margins ofthe dermomyotomal sheet bend in slightly toform four lip-like structures that bulge towards the underlying sclerotome. Aswill be detailed later,these dermomyotomallips are important morphogenetic centers for dermomyotomal growth andmyotome formation. The central portion ofthe dermomyotome in between the marginal lips hasbeen classically referred to as dermatome, as it contributes to the dermis of the back. However,because the origin of the dermis is more complex and not entirely established and the centralderrnomyoromal cells also gives rise to connective tissue , muscle and endoth elium of the back,"

Formation and Differentiation ofAvian Somite Derivatives 3

Figure 1. Scanning electron microscopic view on a transverse fracture of a 2-day chick embryoshowing an epithelial somite. (Courtesy of Dr. Heinz lurgenjacob, Bochum) Explanation offigurenumbers of all figures in thischapter. 1)Neural tube, 2)Notochord, 3)Aorta, 4) Dorsal somite half, 5)Ventral somite half, 6)Surfaceectoderm, 7) Dermomyotome, 8)Myotome, 9)Central sclerotome,9a) Cranial half of the sclerotome, 9b) Caudal half of the sclerotome, 10) Ventral sclero­tome, 11) Lateral sclerotome, 12) Dorsal scierotome,B) Spinal nerve, 14) v, Ebner's fissure15)Meningotome, 16)Syndetome, 17)Somitocoele cells/Arthrotome, 18) Dermis, 19)Wolffianduct, 20) Intermediate mesoderm, 21)Somatopleure, 22) Splanchnopleure, 23)Coelomic cavity,24) Paranotochordal extracellular matrix, 25) Meninx primitiva, 26) Intersomitic blood vessels,27) Limb muscle, 28) Intercostal/Abdominal muscle, 29) Dermis, 30) Dorsal root ganglion.

this term should be omitted in favour ofcentral dermomyotome (CD). The epithelial nature ofthe dermomyotome depends on beta-catenin activity,which is induced by ectodermal Wnt 6 andmediated by the bHLH-transcription factor paraxis,"

Later, the CD disintegrates into a loose mesenchyme, the cells of which migrate dorsallyas dermal and subcutaneous precursor cells and ventral into the myotome as connective tissueprecursors integrating into the epaxial musculature. In contrast, the dermomyotomallips stayepithelial to enableongoing myotomal growth. The DML ismaintained by the mesoderm-intrinsicepithelialization factor Wnt 11, the VLL by Wnt 6 signaling from the overlying ectoderm whichis down regulated over the medial and central dermomyotome by the inhibiting influence ofWnt1UO Finally, the DML and VLL also disintegrate at around embryonic day 7 after the primarypattern ofthe trunk musculature has been laid down. 21

.22

4 Somitogenesis

Initiation and Growth ofthe DermomyotomeThe dermomyotome is formed from the dorsal half of the epithelial somite, whereas the

ventral half-somite gives rise to the sclerotome. Dermomyotomal fate is thus determined by dor­salizing signals from adjacent tissues. In vivo and in vitro experiments in the chick embryo haveestablished that the dorsalizing signals are Wnts secreted by the dorsal neural tube and the dorsalsurface ectoderm (reviewed in ref. 16) By long-range diffusion, the dorsal Wnt signals reach thedermomyotomal cells, where they are available for receptor binding due to the desulfation ofcell-surface heparan sulfate proteoglycans by the sulfatase Qsulfl,23 which is not active in ventralsomitic cells. Moreover, dorsalizing Wnt signaling are suppressed in ventral somitic cells by theShh-dependent Wnt-inhibitor Sfrp2,24 while the ventralizing Shh signal is blocked in the dorsalsomite by the membrane-bound glycoprotein Gas1.25Within the dorsal comparonent, medial andlateral dermornyotornal cells are specified differently. Medial dermomyotome formation dependson Wnt 1 and Wnt 3a from the dorsal neural tube,I3,26-31 whereas the lateral dermomyotomerequires contact-mediated signaling by Wnt6 and possibly other Wnt proteins from the surfaceectoderm.27,32,28,3IThe specification ofdorsomedial and dorsolateral fate seems to be competitive,as in mice lacking both dorsomedial signals, Wnt 1 and Wnt 3a, the medial compartment ofthedermomyotome isnot formed, whereas the lateral domain, marked by Sim 1, isextended medially."Furthermore, Wnt3a seemsto be a positive regulatorofthe proliferation aetvity ofdermomyotomalcells, which acts upstream ofPax3 and Pax7. Overexpression of Wnt3a leads to a mediolateralexpansion ofthe dermomyotomal epithelium due to increased proliferation."

However, the mode of dermomyotomal growth is still under debate. Experiments usingquail-chick transplantations and BrdU-labellingofthe dermomyotomallips identified the medialand lateral lip of the dermomyotome as blastema-like centers of mediolateral dermomyotomalgrowth,21,22 while Dil-labelling experiments suggested a growth contribution of the entire der­momyotorne.r' The epaxial vs. hypaxial identity of the dermomyotomal cells, which is markedby En 1 and Sim 1 expression, respectively, is maintained during the subsequent cell migrationprocesses in the mature somite."

MuscleSkeletal muscle represents the bulk of the total body mass in vertebrates. In contrast to the

smooth musculature ofthe inner organs, which is ofsplanchnopleural origin, all skeletal musclesofthe vertebrate body originate from the dermornyorome, with the exception ofthe branchiomeric andocular muscles which originate from the unsegmented cranial and prechordal mesoderm. Muscleis the phylogenetically oldest derivative ofthe somites, as a rigid skeleton and a tough dermis werenot ofprimary importance in earlyaquatic vertebrates (reviewed in.") Still, in modern anamniotes,the myotome represents the largest and earliest formed somitic domain.F'" whereas in arnniotes,the myotome develops relativelylate. Here, we willgive an overview on the current knowledge onthe initiation ofmyogenesis and early steps ofmuscular differentiation in sornites.

Initiation ofMyogenesisDuring gastrulation, the paraxial mesoderm arises from cells of the cranial primitive streak

and Hensen's node. Cells determined for myogenic differentiation have been identified from theonset of gastrulation." Cell labelling experiments in mouse suggest that, within the primitivestreak, a population of self-renewing stem cells gives rise to spatially coherent paraxial muscleprecursor cells,which migrate as cohorts to their destination in the forming paraxial mesoderm"Thus, cells originating from the rostralrnost primitive streak contribute largely to epaxial muscle,whereas cells from slightly more caudal regions predominantly give rise to hypaxial muscle.40-42However, even prior to gastrulation, chick epiblast cells express the muscle regulatory factor(MRF) MyoD43and undergo myogenesiswhen cultured in protein-free medium." suggesting thatmyogenesis might be the default differentiation program ofpregastrulation epiblast cells.In spiteofthe spatial coherence ofclonal myogenic precursor cells during normal development, classicalrotation experiments in chick have shown that all cells of the segmental plate and early somitesare competent to differentiate into muscle when positioned appropriately,": 13,45 This led to the

Formation andDifferentiation ofAvian Somite Derivatives 5

Figure 2. Scanning electron microscopic view on a transverse fracture of a 3-day chick em­bryo showing the somite derivatives: dermomyotome and sclerotome. (Courtesy of Dr. HeinzJurgen Jacob, Bochum). See Figure 1 for legends.

assumption that the naive paraxial mesoderm is specified for the myogenic lineage by inductiveinfluences from neighboring tissues.

Ablation experiments identified three structures which are required for myogenesis:The neuraltube,13.46-50 the notochord-floor plare-complex'v" and the dorsal ectoderm.P'Tn molecular terms,it appeared that a delicately balanced level ofboth, Shh from the notochord and Wnts from thedorsal neural tube and the ectoderm, acts to elicit myogenesis, while high levels of Shh or Wntalone lead to sclerotomal or nonmyogenic dermomyotomal differentiation, respectivelyy·53 Shhseems to activate dermomyotomal Myj5 expression via zinc-finger transcription factors of theGli family, which can bind to a proximal Myj5epaxial enhancer'v" The view that Shh is involvedin the induction of myogenesis has been challenged by the finding that mice lacking Shh showclear, albeit low, expression ofthe earliest MRF, Myj5.56 However, in the absence ofShh, somiticcells stop to proliferate and eventually die, suggesting that Shh is more likely to be a trophic ormaintenance factor that keeps the myogenic program upright, rather than an inducer of myo­genesisF59 Likewise, removal of the Wnt-producing dorsal neural tube does not impede theinititation ofmuscle marker gene expression, but rather the subsequent myogenic differentiationprograrn/" In a recent paper, Marcelle and coworkers provide further evidence that myogenesisis a mesoderm-autonomous process, as the segmental plate mesoderm expressesMyj5and MyoDindependent ofany neighbouring tissue in vivo and in vitro." They propose that, in the segmentalplate, MyoD expression is induced byWntSb secretedwithin the segmental plate itself, while Myj5

6 Som;togenes;s

does not seem to depend on any inductive signal Thus, there is accumulating evidence that themyogenic program is intrinsic to the paraxial mesoderm, while its localized realization dependson permissive signals from the embryonic environment.

Once the myogenetic program is initiated, how is it realized? The first molecular manifestationofmuscle differentiation is the expression ofMRFs. In mouse, the first known MRF to be expressedin myogenic cells isMyj5followed byMyoD (Fig. 3), whereas in chick the chronology is viceversa.Knockout experiments in the mouse have shown that MyfS and MyoO have redundant functions,but mice lacking both MRFs lack all skeletal muscles, including the myoblast precursor cells in theparaxial mesoderm.f Murine MyfS is able to induce expression ofAfyoD and is in turn inducedby Pax-3, rendering Pax-3 a key regulator ofsomitic myogenesis in mouse, which can induce themyogenic program even in the absence ofsurrounding tissues.63.64A similar role ofPax-3 in avianmyogenesis, or on the putative myogenic inducer WntSb,61 remains elusive. In addition to thepositive regulation by Pax-3 and MyfS, myogenesis is negatively regulated by BMPs.6~ The onsetofmuscle differentiation in the medial epithelial somites66•672 is likely to be due to the release frominhibitory BMP signaling by synthesis of the BMP-antagonist Noggin in the cranial segmentalplate and the dorsornedial somite.68•61

Early Myotome DevelopmentThe first morphological manifestation ofmyogenesis in the avian embryo is the fomation of

primitive muscle fibers subjacent to the dermomyotome. These primitive myotubes, which aremononucleate, postmitotic from the onset oftheir extension and span the craniocaudal extent ofone somitic segment, represent the myotome (reviewed in.") It is generally accepted that all myo­tornal cells derive exclusivelyfrom the dermomyotome, without any myogenic contribution fromthe sclerotomal mesenchyme," This requires localized deepithelialization ofmyogenic precursorcellsfrom the dermomyotomal epithelium and subsequent translocation ofthe myogenic precursorcells into the myotornal domain. The mechanisms of this myotomal cell recruitment have beenthe topic of an ongoing debate over many years. Early investigators assumed that the myotomedevelops from the entire dermomyorome.W" Later, some authors argued that the myotome formsonly from the 0 ML,71·72 while others believed that the myotome is ofsclerotomal origin ." Morerecently, using Oil-cell labelling techniques. the laboratories of Ordahl and Kalcheim came toconflicting results concerning localization and morphogenetic behavior ofmyotomal precursorcells,?4-82,21 ,22, 16 Briefly,Ordahl and coworkers found that the early myotome is exclusivelyformedby cells from the dorsomedial (OML) and ventromedial (VLL) lip of the dermomyotome,which enter the myotomal domain by direct ventral translocation after deepithelialization andsubsequently elongate bidirectionally towards the cranial and caudal margins of the segment. Inthis view, the OML and VLL are blastema-like growth zones enabling the mcdiolateral growth ofboth, the myotome and the dermomyotomal epithelium, in an incremental mode of growth. Incontrast, Kalcheim and colleagues claimed that the earliest myotomal precursor cells first migratecranially to the craniomedial corner ofthe dermomyotome. From there, they elongate as "pioneercells"unidirectionally in caudal direction until they reach the caudal margin ofthe segment. Thus,they form a scaffold for later immigrating myoblasts over the entire mediolateral extent of thedermomyotome. According to this model, there is no direct contribution of the OML or VLLto myotomal growth, but ongoing intercalatory addition of myotubes, originating from precur­sor cells in the cranial and caudal dermomyotorne, to existing fibres, resulting in mediolateralmyotomal growth. For the later myotome, both models agree on the contribution ofcranial andcaudal dermomyotomallips to myotomal growth as was previously suggested by Christ et a1.69

Recently, Marcelle and coworkers have presented a model based on GFP labelling of myogenicprecursor cellsvia electroporation in OVO,83which provides an unequivocal model for the mechan­ics ofavian myotome formation and represents a synthesis ofaspects ofhoth earlier hypotheses."They demonstrated that in a first step . myotomal cells are only provided by the OML by directingression and bidirectional extension. In a second step, firstly the caudal dermomyotomal border,secondly the cranial dermomyotomal border and lastly the VLL, also start to release myotomalprecursor cells.The cells originatingfrom the 0 ML and VLL contribute exclusivelyto the epaxial

Formation and Differentiation ofAvian Somite Derivatives 7

Figure 3. In situ hybridization for MyoD in a 4-day chick embryo. MyoD expression can beseen in the segmentally arranged myotome, the limbs, the tongue and the branchial arches.(Courtesy of Dr. Ketan Patel, London).See Figure 1 for legends.

and hypaxial domain, respectively,whereas the cellsfrom the cranial and caudal margins populateboth mediolateral compartments. Thus, all four margins ofthe dermomyotome provide a distinctcontribution to the formation and growth of the myotome, in a process combining incrementalgrowth from the DML and VLL and intercalatory growth from the cranial and caudal derrno­myotomal borders. Prior to translocation into the myotome, mitotically active dermomyotomalcells change the plane ofcell division from planar to apico-basal.85•86The basal daughter cells ofthese asymmetric divisions give rise to dermal precursor cells, the apical daughter cells expressN-cadherin and give rise to the rnyotomal lineage.P'"

Late Myotome DevelopmentApproximately two days after the onset ofmyotome formation, ingression of myogenic cells

from the dermomyotome ceasesas the dermomyotome has started to dissociate. Thenceforward,growth ofthe myotome-derived musculature depends on the continuous contribution ofmyogenic

8 Somitogenesis

cells from a proliferative pool ofprecursor cells, which are thought to express the transmembranetyrosine kinase receptor FGFR4 (FREK).88.89.78 The late recruited myotomal cells and satelliteprecursor cells, have been shown to originate from the mature derrnomyotome." However, themechanistic and molecular processes during later myotomal growth are still largely unknown.Duringlater stagesofmyogenesis, the mononucleate, unisegmental myotomal muscle fibres becomemultinucleate myotubes, probablyby fusion with satellite cells.In the epaxial domain, the myotubesgive rise to the intrinsic muscles of the back and, in the superficial strata, fuse longitudinally tobridge the segment borders. A similar process is thought to form the abdominal and intercostalmuscles in the hypaxial domain (see below).

Hypaxial Muscle FormationWithin the hypaxial compartment of the derrnornyotorne, the myogenic lineage develops in

two fundamentally different modes, according to the axial level of the somites: At cervical andtrunk levels, the VLL organizes the formation ofa hypaxial myotome similar to the situation atthe epaxial DML (see above). The lateral growth ofboth, the dermomyotome and the hypaxialmyotome, are coordinated such that the blastema-like VLL extends ventrolaterally into the pro­spective lateral and ventral body wall, forming in its wake the ventrolateral myotome, which willgive rise to the abdominal and intercostal muscles. The molecular regulation ofthis process is stillobscure, the commitment ofVLL cells to the nonmigratory myogenic lineage seems to dependon Paraxis."

In contrast, at limb level, the cells of the lateral dermomyotome do not form a VLL, butdeepithelialize and emigrate into the nascent limb buds to form appendicular, shoulder and hipmuscles (Fig. 4).92.93 Prior to the experimental proofby Christ and coworkers, which was based onquail-chick chimeras, observations by Fischel," Murray" and Grim96already argued for a sorniticorigin ofthe wing musculature, but were challenged by conclusions by other workers who claimeda somatopleural origin ofwing muscles.":"

The lateral plate mesoderm in the limb field secretes the glycoprotein SF/HGF, which activatesthe c-met receptor in the hypaxial dermomyotomes of adjacent somites and thus elicits local­ized depithelialization of myogenic precursor cells at limb level,'?' Mice deficient for SF/HGFor the c-met receptor lack limb muscles because muscle precursor cells do not delaminate fromthe hypaxial dcrmomyotome.P'i''" The same phenotyp is seen in mutants lacking Pax-S, whichis essential for c-met transcription in the dcrmomyotome.P't'?' Following the delamination, theonset ofthe emigration ofprecursor cells into the limb mesenchyme depends on the expression ofthe homeo-domain containing transcription factor Lbxl. In Lbxl mutants, the deepithelializedprecursor cells stay close to the somite of origin and are unable to enter the dorsal premuscularmass in the limb. Intriguingly, the emigration to the ventral premuscular mass seems to be unaf­fecred.'?' During their migratory route, the myogenic cells express Msxl,lOS.lfJ7 Pax3108 and theWnt-antagonist Sftp2109 and are guided on their trajectory by SF/HGF secreted by the station­ary limb mesenchyme.P'"" Once in their target location, the precursor cells stop to proliferatein response to FGFR4 signaling!" and activate the myogenic programme via MyoD and Myj5,113the latter upon induction by ectodermal Wnt6Y4,l1S Prior to Wnt-induced myogenesis, the Wntantagonist Sfp2 is downregulated in the rnyobiasts.':" The subsequent differentiation ofthe myo­blasts into myofibers depends on multiple factors including Myogenin, MefL. and MRF4.116

Similar to the situation in the limb field, in the hypaxial domain ofthe cervical somites 2 to 6,a population ofmigratory cells detach from the VLL and emigrate cranially to form the tonguemuscle.l'"!"

DermisThe dermis, which is also known as corium, is a layer offibrous connective tissue between the

superficial epidermis and the underlying subcutis ofthe amniote integument. The taut yet elasticstructure ofthe skin and the formation ofskin-appendages like hairs, feathers and scales,dependson the proper differentiation ofdermal tissue and its interactions with the overlyingepidermis. Ofall somite derivatives, the development ofthe dermis is least understood. Here, we will summarize