PDGF-A Signalling Regulates Radially Oriented Movements of Mesoderm Cells During Gastrulation in

Xenopus

by

Erich William Damm

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy

Graduate Department of Cell and Systems Biology University of Toronto

© Copyright by Erich William Damm 2014

  ii

PDGF-A Signalling Regulates Radially Oriented Movements of Mesoderm Cells During Gastrulation in

Xenopus

Erich William Damm

Doctor of Philosophy

Department of Cell and Systems Biology University of Toronto

2014

Abstract

The molecular regulation and cellular basis of morphogenesis during Xenopus

gastrulation has been a topic of extensive study. In particular, the convergent

extension movements that occur in the chordamesoderm are well understood,

however less is known about the morphogenesis of other mesoderm regions.

Since most of the morphogenetic movements occurring during gastrulation

appear to be region autonomous, it is necessary to understand these region

specific movements and their contribution to the gastrulation process in order to

understand gastrulation as a system.

The organization of cells in a region can be indicative of the cell movements that

are occurring in that region. Using scanning electron microscopy, I have analyzed

the organization of cells in the prechordal mesoderm (PCM) of the gastrula,

which is the sub-region of anterior mesoderm that gives rise to the muscles of the

  iii

head. I found that during the middle of gastrulation, PCM cells were oriented with

their long axes perpendicular to the overlying blastocoel roof (BCR) and that the

orientation of these cells is BCR dependent. By the end of gastrulation, I found

that the PCM had thinned from a multi-layered tissue to single layer of cells in

contact with the BCR, suggesting radial interaction of the PCM. Furthermore,

cells in corresponding regions on the lateral and ventral sides of the embryo were

found to be oriented similarly, suggesting that radially directed movements could

be common throughout the anterior mesoderm.

Three isoforms of the Platelet Derived Growth Factor A (PDGF-A) are expressed

in the BCR and its cognate receptor, PDGFRα, is expressed in the PCM. I have

shown that two isoforms of the PDGF-A isoforms (lf-PDGF-A and int-PDGF-A)

remain associated with the surface of secreting cells and thus signal over a short

range, while a third (sf-PDGF-A) is diffusible and can signal over long distances.

My work has shown that radial intercalation of the PCM is instructively regulated

by a concentration gradient of sf-PDGF-A, which forms in the mesoderm over a

distance of approximately 200 µm by the diffusion of molecules through

intercellular gaps. My work describes for the first time, the molecular regulation of

radial intercalation.

  iv

Acknowledgements First, I would like to thank Rudi Winklbauer. Without you Rudi, none of this work

would have been possible. You are a true scientist, one who is driven by a basic

non-compromising need to explore the mechanics of nature. You are an

inspiration in the way that you relentlessly follow your passion despite any

opposition. I want to thank you for the independence and the support you have

given me throughout this project. It has been instrumental in helping me develop

independence in my research. My time in your lab has been on of the most

influential experiences of my life and I thank you for the opportunities you have

afforded me.

Profs. Ashley Bruce and Tony Harris, thank you both for your guidance, advice

and the criticisms of my work over the years. The passion for research that you

express has been inspirational.

Thank you to the past and present members of the Winklbauer lab for your

suggestions and assistance. Olivia, Tina and Hiro, thank you for training me in

everything Xenopus. Without the three of you, I would never have been able to

start this work let alone finish it.

I would like to thank the members of the 6th floor labs, my vertebrate colleagues

in the Bruce and Tropepe labs especially, for your support and sharing of

  v

reagents. In particular, I’d like to thank Stephanie Lepage for the sharing of ideas

and the many supportive lunch discussions.

To all of my friends, thank you for your support and understanding over the last

six years. In particular, I’d like to thank Jonathan Mitchell who has seen me

through the good and the bad times and who has been a rock solid emotional

support. I’d also like to thank Brough Perkins for his love and support throughout

the writing process.

Lastly, but certainly not least, I’d like to thank my parents and my grandparents

for their unconditional love, trust, support and understanding throughout this

whole process. You raised me and were instrumental in shaping me into the

person I am today, in a way, you own this work as much as I do.

  vi

Table of Contents

 

Chapter One: Introduction ............................................................................................................ 1

1.1: Gastrulation in Xenopus ................................................................................................... 2

1.1.1: Anatomy of the Xenopus gastrula ................................................................................. 6

1.1.2: Convergent Extension: medio-lateral and radial cell intercalation .............................. 10

1.1.3: Epiboly ........................................................................................................................ 17

1.1.4: Vegetal Rotation: Internalizing the vegetal cell mass ................................................. 19

1.1.5: Directional cell migration of the anterior mesoderm .................................................... 20

1.2: Platelet derived growth factor signaling in Xenopus gastrulation ............................. 23

1.2.1: Platelet derived growth factor ligands and receptors .................................................. 24

1.2.2: Platelet derived growth factor ligands and receptors are expressed in adjacent tissue

layers during development .................................................................................................... 32

1.2.3: Signaling downstream of platelet derived growth factor receptors ............................. 34

1.2.4: Platelet derived growth factor signaling during Xenopus gastrulation ........................ 37

1.3: Molecular gradient formation: mechanisms and kinetics. .......................................... 43

1.3.1: Mechanisms of molecular gradient formation ............................................................. 44

1.3.2: Kinetics of gradient formation: tissue level versus molecular level. ............................ 58

1.4: Outline and Objectives .................................................................................................... 60

Chapter Two: The prechordal mesoderm undergoes intercellular migration and radial

intercalation during gastrulation ............................................................................................... 63

2.1: Introduction ...................................................................................................................... 64

2.2: Results .............................................................................................................................. 66

2.2.1 The dorsal, lateral and ventral mesoderm can be subdivided into distinct cell groups

based on gene expression and morphological evidence ...................................................... 66

  vii

2.2.2: During gastrulation, mesoderm cells are unipolar and produce lamelliform protrusions

.............................................................................................................................................. 76

2.2.3: LEM, prechordal and involuted chordamesoderm mesoderm cells are oriented toward

the BCR ................................................................................................................................ 83

2.2.4: Prechordal mesoderm cells undergo radial intercalation at the mesoderm/ectoderm

boundary ............................................................................................................................... 98

2.2.5: Intercellular migration drives prechordal mesoderm radial intercalation ................... 101

2.2.6: Involuted lateral and ventral mesoderm cells are oriented toward the BCR ............. 109

2.3: Discussion ...................................................................................................................... 119

2.3.1 Complementing in vitro explant studies with intact embryo studies ........................... 119

2.3.2 Mesoderm regions undergo region specific morphogenetic movements ................... 122

2.3.3 Intercellular migration and radial intercalation ........................................................... 127

2.4: Materials and Methods .................................................................................................. 131

2.4.1: Embryos and microinjections .................................................................................... 131

2.4.2: In-Situ Hybridization and In-Situ Probe Synthesis .................................................... 131

2.4.3: Scanning Electron Microscopy .................................................................................. 132

2.4.4: Transmission Electron Microscopy ........................................................................... 133

2.4.5: Confocal Microscopy ................................................................................................. 134

2.4.6: Explants .................................................................................................................... 134

2.4.7: Statistical Analysis .................................................................................................... 135

Chapter Three: PDGF-A controls prechordal mesoderm cell orientation and radial

intercalation during gastrulation ............................................................................................. 136

3.1: Introduction .................................................................................................................... 137

3.2: Results ............................................................................................................................ 139

3.2.1: Expression of PDGF-A isoforms is restricted to the inner cell layer of the blastocoel

roof ...................................................................................................................................... 139

  viii

3.2.2: Inhibition of PDGF signalling interferes with pre-chordal mesoderm radial

intercalation, but not with chordamesoderm cell orientation ............................................... 149

3.2.3: sf-PDGF-A is an instructive cue required for radial orientation of prechordal

mesoderm cells ................................................................................................................... 152

3.2.4: sf-PDGF-A is required for directional intercellular mesoderm migration in an explant

system ................................................................................................................................ 164

3.2.5: An instructive role for sf-PDGF-A signaling in directional migration ......................... 167

3.3: Discussion ...................................................................................................................... 178

3.3.1: A mechanism for radial cell intercalation in the prechordal mesoderm ..................... 178

3.3.2: Distinct roles for long and short PDGF-A splice isoforms in cell orientation: contact-

dependent and long-range signaling ................................................................................... 179

3.3.3: Patterns of cell orientation and radial intercalation in Xenopus gastrulation ............ 184

3.4: Materials and Methods .................................................................................................. 185

3.4.1: Embryos and microinjections .................................................................................... 185

3.4.2: Explants .................................................................................................................... 186

3.4.3: Scanning Electron Microscopy .................................................................................. 186

3.4.4: mRNA Isolation/RT-PCR .......................................................................................... 186

3.4.5: Constructs, Morpholinos and mRNA Synthesis ........................................................ 186

3.4.6: In-Situ Hybridization and In-Situ Probe Synthesis .................................................... 187

3.4.7: Statistical Analysis .................................................................................................... 187

Chapter Four: The short splice isoform of PDGF-A forms a chemoattractant gradient by

diffusion of molecules through the extracellular space ....................................................... 188

4.1: Introduction .................................................................................................................... 189

4.2: Results ............................................................................................................................ 192

4.2.1: Short and long PDGF-A isoforms have different extracellular localizations .............. 192

4.2.2: sf-PDGF-A is found distant from its source, in intercellular spaces between mesoderm

cells ..................................................................................................................................... 199

  ix

4.2.3: sf-PDGF-A forms an extracellular gradient that fits a single exponential decay function

............................................................................................................................................ 210

4.2.4: The effective diffusion coefficient Deff is consistent with formation of a sf-PDFG-A-

eGFP gradient by diffusion through extracellular spaces ................................................... 234

4.3: Discussion ...................................................................................................................... 238

4.3.1: The cell retention motif determines if a PDGF-A isoform can form a concentration

gradient ............................................................................................................................... 239

4.3.2: The sf-PDGF-A gradient likely forms by diffusion of molecules through intercellular

spaces ................................................................................................................................ 240

4.3.3: Potential factors influencing the effective diffusion coefficient .................................. 244

4.3.4: The regulation of molecule degradation during gradient formation ........................... 248

4.4: Materials and Methods .................................................................................................. 253

4.4.1: Embryos and Microinjections .................................................................................... 253

4.4.2: eGFP and Myc tagged constructs ............................................................................. 253

4.4.3: mRNA Synthesis ....................................................................................................... 255

4.4.4: Explants .................................................................................................................... 255

4.4.5. Scanning Electron Microscopy .................................................................................. 256

4.4.6: Transmission Electron Microscopy ........................................................................... 256

4.4.7: Antibody Staining ...................................................................................................... 256

4.4.8: Confocal Microscopy ................................................................................................. 256

4.4.9: Image Processing ..................................................................................................... 257

4.5: Appendix ......................................................................................................................... 258

Chapter Five: Final Model and References ............................................................................ 269

5.1: Prechordal mesoderm morphogenesis: the cellular basis and the molecular

regulation ............................................................................................................................... 270

5.2: References ..................................................................................................................... 274

  x

List of Tables

Table 2.1: The proportion of total mesoderm occupied by individual mesoderm subtypes. ......... 74

Table A1: The calculated f1 and f2 values for the range of t values .......................................... 261

Table A2: Error function approximation values for f1 and f2 for the indicated value of t ............ 262

Table A3: Calculated C!"' (τ) values for the indicated value of t ................................................... 263

Table A4: C!"' (τ) and t values with the corresponding values τ!" for experimentally derived  C!. 267

  xi

List of Figures Figure 1.1: Xenopus gastrula fate map ........................................................................................... 5

Figure 1.2: Major cell movements occurring during Xenopus gastrulation ..................................... 8

Figure 1.3: Platelet Derived Growth Factor Family Ligands ......................................................... 27

Figure 1.4: PDGFRa and associated signaling pathways ............................................................ 30

Figure 2.1: The dorsal, lateral and ventral mesoderm are made up of sub-regions ..................... 69

Figure 2.2: Mapping of dorsal, ventral and lateral mesoderm regions on scanning electron

micrographs ................................................................................................................................... 72

Figure 2.3: PCM cells extend lamelliform protrusions toward the BCR in sagittal and transverse

planes ............................................................................................................................................ 79

Figure 2.4: PCM cells become significantly elongated in the sagittal plane during gastrulation ... 82

Figure 2.5: PCM cells become significantly elongated in the transverse plane during gastrulation

....................................................................................................................................................... 86

Figure 2.6: The long axes of mesoderm cells are oriented parallel to the BCR at the start of

gastrulation .................................................................................................................................... 90

Figure 2.7: Internalized PCM and chordamesoderm cells are oriented toward the BCR ............. 94

Figure 2.8: LEM, PCM and CM exhibit distinct cell morphologies when viewed from the BCR

apposed surface ............................................................................................................................ 97

Figure 2.9: PCM thins to a single cell layer by the late gastrula stage ....................................... 100

Figure 2.10: PCM undergoes radially directed cell rearrangements ........................................... 104

Figure 2.11: PCM radial intercalation in vitro .............................................................................. 107

Figure 2.12: PCM cells migrate directionally in vitro ................................................................... 111

Figure 2.13: Directional migration of PCM cells in vitro is BCR dependent ................................ 114

Figure 2.14: Involuted lateral mesoderm and LEM orientation is similar to their dorsal

counterparts ................................................................................................................................. 116

  xii

Figure 2.15: Involuted ventral mesoderm and LEM orientation is similar to their dorsal

counterparts ................................................................................................................................. 118

Figure 2.16: Electron microscropy of the LEM ............................................................................ 124

Figure 3.1: The structure of Xenopus PDGF-A isoforms ............................................................ 141

Figure 3.2: Expression of PDGF-A and PDGFRa in the Xenopus gastrula ................................ 144

Figure 3.3: PDGF-A isoforms are highly conserved between species ....................................... 148

Figure 3.4: PDGF-A inhibition disrupts PCM radial intercalation ................................................ 151

Figure 3.5: PDGF-A knockdown by morpholino oligonucleotides disrupts the orientation of PCM

cells ............................................................................................................................................. 154

Figure 3.6: PDGF-A knockdown by morpholino oligonucleotides disrupts the orientation of

involuted ventral mesoderm cells ................................................................................................ 157

Figure 3.7: PDGF signaling knockdown by dominant negative ligand or receptor disrupts PCM

cell orientation ............................................................................................................................. 160

Figure 3.8: sf-PDGF-A is required for radial orientation of PCM cells ........................................ 163

Figure 3.9: Overexpression of PDGF-A constructs in the BCR does not affect PCM cell

orientation .................................................................................................................................... 166

Figure 3.10: Over-expression of sf-PDGF-A in the marginal zone disrupts PCM cell orientation

..................................................................................................................................................... 169

Figure 3.11: The effect of over-expression of sf-PDGF-A in the marginal can be rescued ........ 171

Figure 3.12: PDGF-A MO inhibits PCM radial intercalation in vitro ............................................ 173

Figure 3.13: sf-PDGF-A is required for directional migration of deep PCM cells ........................ 175

Figure 3.14: sf-PDGF-A is an instructive cue for PCM directional migration .............................. 177

Figure 4.1: Visualizing PDGF molecules .................................................................................... 195

Figure 4.2: sf-PDGF-A-eGFP can rescue the PDGF-A morphant phenotype ............................ 198

Figure 4.3: sf-PDGF-A-eGFP is localized to ectoderm intercellular spaces ............................... 202

Figure 4.4: int-PDGF-A-eGFP is localized to ectoderm cell membranes ................................... 204

Figure 4.5: lf-PDGF-A-myc localization is similar to that of int-PDGF-A-eGFP .......................... 207

  xiii

Figure 4.6: sf-PDGF-A-eGFP forms a visible gradient in mesoderm tissue ............................... 209

Figure 4.7: sf-PDGF-A-eGFP is only observed in intercellular gaps .......................................... 212

Figure 4.8: sf-PDGF-A-eGFP in intercellular spaces .................................................................. 216

Figure 4.9: int-PDGF-A-eGFP remains localized to the mesoderm/ectoderm boundary ............ 218

Figure 4.10: The sf-PDGF-A gradient fits a single exponentional function ................................. 221

Figure 4.11: There is no change in average decay length (l) by increasing sf-PDGF-A expression

..................................................................................................................................................... 223

Figure 4.12: int-PDGF-A-eGFP does not form a gradient .......................................................... 228

Figure 4.13: sf-PDGF-A-eGFP forms a steady state gradient .................................................... 231

Figure 4.14: sf-PDGF-A-eGFP in extracellular spaces moves relative to cells as cells migrate 233

Figure A1: Calculation of the degradation constant for sf-PDGF-A-eGFP ................................. 265

Figure 5.1: A model for the cellular basis of prechordal mesoderm morphogenesis .................. 273

  xiv

List of Appendices Appendix 1………………………………………………………………………..........................…..258

 

  1

Chapter One: Introduction

Sections of this chapter were published in Winklbauer, R. and Damm, E. W. (2011) Internalizing the vegetal cells mass before and during amphibian gastrulation: vegetal rotation and related movements.

WIREs Dev Biol, 1 (2): 301-306

  2

1.1: Gastrulation in Xenopus During the early phase of animal development coordinated cell movements

convert a relatively unstructured ball of cells into a highly organized embryo.

During this process the germ layers (ectoderm, mesoderm and endoderm) are

positioned by active cell movements and passive cell re-arrangements resulting

in the internalization of prospective mesoderm and endoderm and the spreading

of prospective ectoderm over the surface of the embryo. This process is termed

“gastrulation”, initially named after the phase of embryogenesis when the gut

begins to form. Although the set of cell movements that constitute gastrulation

vary from species to species, all animal phyla undergo gastrulation as a process

(Leptin, 2005). Many of these movements are region autonomous but have

influence on the movements of other regions of the embryo. Therefore, because

gastrulation is a process and not a single movement, any understanding of

gastrulation must be a systems level understanding with respect to

morphogenetic movements. That is, it is necessary to understand the individual

movements occurring in all of the embryo regions.

The African clawed frog, Xenopus laevis, is a major model organism for the study

of early development, and is arguably the most well understood example of

gastrulation in a vertebrate. The large externally developing embryos make this

organism ideal for microinjection and reverse genetic experiments as well as

explantation and live imaging of tissues. The embryos are also very well suited

  3

for imaging with scanning and transmission electron microscopy and confocal

microscopy. Gastrulation in Xenopus starts 9 hours after fertilization and lasts for

nearly six hours. Initially, ectoderm begins to undergo epiboly and endoderm is

internalized by vegetal rotation. Mesoderm then undergoes internalization and

subsequently anterior mesoderm undergoes directional migration while the

posterior mesoderm undergoes convergent extension (Figure 1.1).

The cellular basis and molecular regulation of chordamesoderm (CM) convergent

extension has been extensively studied in Xenopus and much focus has been

placed on the role of convergent extension in the elongation of the

anterior/posterior axis of the embryo. The movements of the other mesoderm

regions have not been as extensively studied. In particular, prechordal mesoderm

(PCM) morphogenesis and the contribution of these movements to gastrulation

are not known. My work has focused on describing the cellular and molecular

basis of cell movements of the PCM, a subset of anterior mesoderm cells. I have

found that the PCM undergoes radial intercalation, contributing to the spreading

of the mesoderm layer (see Chapter Two) and that these movements are

instructively regulated by platelet derived growth factor (PDGF) signaling. In

order to emphasize the significance of PCM morphogenesis to the gastrulation

process, each of the known morphogenetic processes occurring during Xenopus

gastrulation will be discussed in detail in this section and are summarized in

Figure 1.1.

  4

Figure 1.1: Xenopus gastrula fate map

(A-B) Fate maps of sagitally (A) and laterally (B) fractured mid-gastrula stage

embryos. The prospective fates of ectoderm, endoderm and mesoderm

subregions as determined from the literature are indicated. 1Keller, 1976; 2Dale

and Slack, 1987; 3Keller, 1991; 4Lane and Smith, 1999; 5Walters et al. 2001;

6Colas et al. 2008; 7Niehrs et al. 1993; 8Bothe et al. 2000; 9Chalmers and Slack,

2000; 10Moody, 1987ab.

  5

Figure 1.1: Xenopus gastrula fate map

  6

1.1.1: Anatomy of the Xenopus gastrula

The Xenopus gastrula can be divided into several regions based on the fate

maps derived from early stage cell labeling and tracing experiments and regional

marker gene expression (Figure 1.2; Keller, 1975; Keller, 1976; Moody, 1987ab;

Dale and Slack, 1987; Niehrs et al. 1993; Bauer et al. 1994; Walters et al. 2001;

Colas et al. 2008). The ectoderm can be divided into two regions based on the

fates of cells in these regions at later stages (Figure 1.2A, B). On the dorsal side

of the embryo, the prospective neuroectoderm overlies the mesoderm and will

differentiate into nervous system structures, while the animal cap, lateral and

ventral ectoderm will give rise to the larval epidermis (Dale and Slack, 1987;

Keller, 1991; Moody, 1987ab).

Several areas of mesoderm have also been identified (Figure 1.2A, B). Axial

mesoderm, including the xBrachyury (XBra) expressing dorsal mesoderm (CM),

and the Goosecoid (Gsc) expressing PCM, produces the notochord (Keller, 1976;

Keller, 1991, Lane and Smith; 1999), and the muscles of the head (Niehrs et al.

1993; Bothe and Dietrich, 2006) respectively. The dorsal, lateral and ventral

leading edge mesendoderm (LEM) regions converge toward ventral side of the

  7

Figure 1.2: Major cell movements occurring during Xenopus gastrulation

(A-D) Schematic diagrams showing the major cell movements that occur over

gastrulation. Stage 10 (A), stage 10.5 (B), stage 11 (C) and stage 12 (D)

embryos are represented. Arrows indicate the directions of cell movements

(black arrows, mesoderm; blue arrows, ectoderm; orange arrows, endoderm).

Dashed orange arrows (A), pre-gastrulation emboly; solid orange arrows (A-D),

vegetal rotation; Red arrows in (D), hypothetical anterior movement of PCM. CM,

chordamesoderm; PCM, prechordal mesoderm; LEM, leading edge

mesendoderm; Ant, anterior; Pos, posterior; V, ventral; D, dorsal.

  8

Figure 1.2: Major cell movements occurring during Xenopus gastrulation

  9

embryo to close the mesoderm mantle during gastrulation and will form blood

islands along with the anterior ventral mesoderm during tailbud stages (Keller,

1976; Keller, 1991; Lane and Smith, 1999; Colas et al. 2008). Somitic mesoderm

(SM) is derived from the dorso-lateral to ventral XBra expressing region (Keller,

1976; Dale and Slack, 1987; Keller, 1991; Lane and Smith, 1999) and the heart

and circulatory system form from the anterior dorso-lateral mesoderm (Keller,

1976; Keller, 1991; Lane and Smith, 1999; Walters et al. 2001). Body wall

muscles are derived from the lateral plate mesoderm, which arises from the

anterior ventro-lateral mesoderm (Lane and Smith, 1999) and the pronephros

develops from the ventral XBra expressing mesoderm.

Endoderm subregions have also been identified (Figure 1.2A). The pharyngeal

endoderm is located in the dorso-anterior endoderm, adjacent to the PCM

(Keller, 1976; Chalmers and Slack, 2000), whereas the small and large intestines

are derived from the ventro-vegetal endoderm (Chalmers and Slack, 2000). The

gut-associated organs such as the liver, gall bladder, pancreas and stomach are

derived from the ventro-anterior endoderm located adjacent to the ventral anterior

mesoderm (Chalmers and Slack, 2000). To date, detailed fate and specification

maps of the gastrula are lacking, therefore only the general location of the groups

of cells that give rise to the structures and organs described above are known.

Further subdividing the gastrula into groups of cells that correspond to their future

fates would be advantageous as it would allow for a region based study of the

  10

specific morphogenetic movements required during gastrulation to build these

tissues.

1.1.2: Convergent Extension: medio-lateral and radial cell intercalation Convergence and extension is a morphogenetic process that narrows and

elongates tissue. This outcome can be accomplished by several mechanisms

such as cell rearrangements (Bertet et al. 2004; Blankenship et al. 2006),

including active cell intercalation (Keller, 1985; Keller and Tibbetts, 1989,

Heisenberg et al. 2000, Sepich et al. 2000, Yamanaka et al. 2007), oriented cell

divisions (Gong et al. 2004) and cell shape changes (Solnica-Krezel and Sepich,

2012). For decades, it has been known that convergence and extension

movements play a role in the elongation of the anterior/posterior axis of the

amphibian embryo (Schechtman, 1942) and this process has since been studied

extensively in Xenopus.

The cellular basis of “convergent extension” In the Xenopus embryo, convergence and extension behaviour is called

“convergent extension” and it occurs most strongly in the posterior dorsal

mesoderm (Keller et al. 1985, Keller and Danilchik, 1988). Studies of this process

have largely been done with in vitro explants in which, superficial (surface of the

embryo) and deep cells from the dorsal side of two embryos are explanted and

  11

then combined (Keller et al. 1985). These “Keller Sandwich” explants include five

types of cells, animal cap cells (ectoderm), prospective neuroectoderm, CM, SM

and PCM. The explants are initially short and wide, however around the middle of

gastrulation, 11 hours post fertilization (h.p.f.), the prospective neuroectoderm

and the chordamesoderm begin to converge and extend, resulting in a dramatic

narrowing and elongation of the explant by the end of gastrulation (15 h.p.f.)

(Keller et al. 1985, Keller and Danilchik 1988). The SM showed less dramatic

convergent extension and the PCM formed a rounded group of mesenchymal

cells (Keller et al. 1985). Thus, convergent extension movements are important

for anterior/posterior axis elongation. Importantly, the other regions of mesoderm

undergo their own distinct tissue autonomous cell movements. This means that

while the effect of convergent extension on explanted tissue is dramatic, it is not

a driver for morphogenesis of other mesoderm regions.

Close analysis of converging and extending chordamesoderm in explants

revealed that the majority of cells were bipolar with their long axes oriented in a

medial to lateral direction with respect to the anterior/posterior midline of the

explant (Keller and Tibbetts, 1989, Wilson and Keller, 1991, Keller, 2004).

Additionally, cells undergoing radial intercalation, that is moving into the plane of

the surface being imaged from deeper in the explant, were observed. This

movement likely plays a role in thinning of the explant (Wilson and Keller, 1991).

Thus, the cellular basis of convergent extension is the intercalation of medially to

  12

laterally oriented cells at the explant midline (dorsal midline in the embryo) with

radial intercalation thinning the tissue and likely further contributing to elongation.

Over the course of gastrulation, cells become stacked behind each other in the

anterior to posterior direction forming the notochord (Keller and Tibbetts 1989,

Keller et al. 1989).

The “convergent extension” mechanism of notochord formation first discovered in

Xenopus is highly conserved amongst vertebrates. At the start of gastrulation in

zebrafish, mesoderm cells located at more lateral positions, converge as

individual migrating cells toward the dorsal side of the embryo (Sepich et al.

2000, Yamashita et al. 2002, Myers et al. 2002). After arriving dorsally, they

become medio-laterally oriented, intercalate at the dorsal midline and become

stacked behind each other in the anterior to posterior direction to form the

notochord (Heisenberg et al. 2000, Sepich et al. 2000, Myers et al. 2002).

Furthermore, in mouse, the trunk notochord forms by a similar convergent

extension process. Trunk notochord precursor cells are internalized through the

node after which the cells become elongated in a medio-lateral direction

(Yamanaka et al. 2007). These cells then intercalate at the dorsal midline forming

two rows of notochord plate cells in the anterior posterior direction (Yamanaka et

al. 2007).

  13

Elongation of a tissue can also be accomplished by radial intercalation and is

thought to be the principle method of elongation in the SM of Xenopus and

zebrafish (Wilson et al. 1989; Yin et al. 2008). In Xenopus SM explants, radial

intercalation occurs in groups of cells (Wilson et al. 1989). That is, it is more likely

that intercalation will occur near other newly intercalated cells. This suggests that

there could be regulation of cell-cell contacts in these cluster sites that make it

easier for deeper cells to move between the more superficial cells (Wilson et al.

1989). Intercalation results in the thinning and elongation of the tissue.

Interestingly, elongation of the explants is not isotropic but rather occurs without

widening of the tissue (Wilson et al. 1989). Similar behaviour is seen in vivo in

zebrafish somitic mesoderm (Yin et al. 2008). In this system, cells intercalating

from deeper in the tissue preferentially separate cells that are anterior/posterior

neighbours resulting in anisotropic spreading of the tissue (Yin et al. 2008). This

works because the intercalating cell preferentially separates two cells that were

initially in contact with each other at their anterior and posterior sides

respectively. The newly intercalated cell forms new anterior and posterior

contacts with the cells that were separated by the insertion of the new cell.

Contacts are also made between the newly intercalated cells and lateral cells.

SM cells are all similar in size, thus the newly intercalated cell occupies the same

lateral area as the cells that were displaced in the anterior/posterior directions

and extension will occur in the anterior/posterior direction without expansion

  14

laterally. A similar process has been proposed to explain the anisotropic

spreading of Xenopus SM explants (Wilson et al. 1989).

The molecular regulation of “convergent extension” Convergent extension involves the intercalation of oriented and polarized cells at

the dorsal midline of the embryo and the cues responsible for orienting, polarizing

and guiding these cells have been investigated. Orientation of cells appears to

require the patterning of the anterior/posterior axis of the CM (Ninomiya et al.

2004). Cells have positional identities along the anterior/posterior axis and only

cells with matching positional identities intercalate with each other (Ninomiya et

al. 2004). The anterior/posterior pattern of the CM can be seen in a

countergradient of Chordin (anterior) and XBra expression (posterior). It is not

known if cell orientation is directly regulated by this countergradient or if cell

orientation and the expression of these genes are the result of patterning by

Nodal and Lefty factors (Branford and Yost, 2002; Dougan et al. 2003; Ninomiya

et al. 2004).

CM cells undergoing convergent extension are bipolar, extending lamelliform

protrusions from opposite poles of the cell (Keller and Tibbetts, 1989; Wallingford

and Harland, 2001). The control of this polarity is not fully understood but likely

involves an interplay between the anterior/posterior patterning of CM tissue and

members of the non-canonical Wnt signaling pathway (Ninomiya et al. 2004;

  15

Tada and Kai, 2009). A core set of non-canonical Wnt signaling (NC-Wnt

signalling) molecules called the planar cell polarity (PCP) proteins, are important

in establishing the polarity of wing hair extension from cells in the Drosophila

wing epithelium (Gubb and Garcia-Bellido, 1982; Wong and Adler, 1993). In

vertebrates, NC-Wnt signaling involves extracellular Wnt molecules, glypicans,

membrane proteins and receptor tyrosine kinase co-receptors that interact with

Frizzled cell surface receptors. Frizzled receptors signal through Dishevelled,

which is membrane localized and subsequently facilitates interactions with

downstream signaling molecules, kinases and GTPases. These interactions

ultimately result in modulation of the actin cytoskeleton and target gene

expression (Tada and Kai, 2009).

Disruption of NC-Wnt signaling components prevents convergent extension.

Dishevelled is localized to the cell membrane of CM cells undergoing convergent

extension and apparently, disruption of Dishevelled function causes these cells to

lose polarity, which results in the inhibition of convergent extension (Wallingford

et al. 2000). Wnt11 in cooperation with glypican 4, a heparin sulfate proteoglycan

(HSPG), frizzled7 and the receptor tyrosine kinase, Ror2, functioning as a co-

receptor interacting with Wnt5a, have been shown to regulate convergent

extension in Xenopus upstream of Disheveled (Djiane et al. 2000; Tada and

Smith, 2000; Oishi et al. 2003; Hikasa et al. 2002). However, the effect of

disrupting NC-Wnt signaling in this system is unclear. A change in cell orientation

  16

as a result of disrupting NC-Wnt signaling would have been missed and could

have been confused with a loss of cell polarity if the change in orientation

resulted in the cell long axis pointing into the explant rather than being oriented in

the plane being imaged. This is because measurements of cell lengths/widths

were made from Keller Explants by looking at the explant surface only

(Wallingford et al. 2000). An analysis of Keller Explants and intact embryos

fractured in multiple planes will be necessary to answer this question.

Interestingly, Wnt11 is a known target gene of XBra (Tada and Smith, 2000). This

may suggest that anterior/posterior patterning of the chordamesoderm is required

for the expression of key NC-Wnt signaling regulators.

Paraxial protocadherin (PAPC) is a transmembrane molecule that regulates

downstream signaling processes. Expression of PAPC in CM is regulated by

Wnt5a signaling through Ror2 (Schambony and Wedlich, 2007) and it has been

shown that signaling through PAPC alongside NC-Wnt signaling is required for

normal convergent extension (Unterseher et al. 2004). Surprisingly, when PAPC

function is blocked, explants elongate normally but fail to constrict (Unterseher et

al. 2004). This suggests that constriction and elongation behaviour is separable

both morphologically and molecularly. Thus, although it has been shown that it is

required, the role of NC-Wnt signaling in convergent extension has not been

clearly defined.

  17

1.1.3: Epiboly During epiboly, the ectodermal blastocoel roof (BCR) undergoes morphogenetic

movements that cause it to spread anisotropically toward the blastopore over the

surface of the embryo so that by the end of gastruation, the embryo surface is

covered by ectoderm (Figure 1.1A-D). In Xenopus, epiboly starts before

gastrulation at stage 9 (7 h.p.f.) on the dorsal side of the embryo (Papan et al.

2007a). Epiboly spreads laterally and begins on the ventral side by stage 10+ (10

h.p.f.) (Papan et al. 2007a). When epiboly starts, all regions of the BCR consist of

multiple cell layers. The dorsal and ventral marginal zone BCR are the thickest,

followed by the prospective neuroectoderm and the apex of the animal cap

(Keller, 1980). The cellular mechanism driving epiboly in Xenopus is not well

understood and is likely dependent on the region of the BCR in question,

although cells in each of these regions appear to undergo intercalation

movements that result in the thinning and spreading of the ectoderm layer

(Keller, 1980). Whether thinning is due to active cell movements or passive cell-

rearrangments has not been fully investigated and appears to be region

dependent (Marsden and DeSimone, 2001). Interestingly, the downward

movement of the ectoderm occurs simultaneously with the clockwise rotation of

the mesodermal dorsal marginal zone (Figure 1.1A), which results in mesoderm

tissue becoming apposed to the inner layer of the BCR (Papan et al. 2007a). The

rotation of mesoderm tissue is independent of BCR morphogenesis (Winklbauer

and Schurfeld, 1999) and the attachment of mesoderm and BCR at the start of

  18

gastrulation suggests that forces generated by this rotation could be involved in

pulling the BCR downward toward the blastopore. The animal cap epithelium

appears to be under tension and behaves in an elastic fashion when tension is

released (Luu et al. 2011). In the embryo, this tension could be a result of dorsal

marginal zone movements, however, this specific possibility has not been

investigated. Thus, epiboly in Xenopus may be a primarily autonomous process

constituting the thinning and spreading of different regions of the BCR over the

embryo though active and passive cell re-arrangements, however, the

involvement of forces generated by the movements of the attached mesoderm

cannot be excluded.

Fibronectin has been implicated in the regulation of intercalation during epiboly in

Xenopus (Marsden and DeSimone, 2001). The inner layer of the BCR is coated

with an extracellular matrix made up of fibrillar fibronectin (Winklbauer, 1998;

Marsden and DeSimone, 2001). When fibronectin fibrillogenesis was inhibited

using function blocking antibodies, epiboly was inhibited and the BCR remained

several cell layers thick by late gastrulation (Marsden and DeSimone, 2001). This

was because cells were mis-oriented and intercalation was unproductive

(Marsden and DeSimone, 2001). Similar effects were seen when inhibiting β1

containing integrins (Marsden and DeSimone, 2001). Therefore, ectoderm cells

require interaction with fibronectin through integrins for normal intercalation

behaviour. This interaction may provide an orienting signal that is required for

  19

cells to intercalate in the right direction during epiboly (Marsden and DeSimone,

2001).

1.1.4: Vegetal Rotation: Internalizing the vegetal cell mass In Xenopus, internalization of endoderm cells that make up the vegetal cell mass

may start as early as cleavage stages and progresses through gastrulation. Prior

to the start of gastrulation, the internalization process is called pre-gastrulation

emboly and involves the inward movement of vegetal cells from the vegetal base

of the embryo toward the blastocoel floor (Figure 1.1A; Bauer et al. 1994; Papan

et al. 2007ab; Wiklbauer and Damm, 2011). As cells are internalized, they leave

the vegetal base region of the embryo resulting in constriction of the base while

internalized cells insert into the blastocoel floor causing expansion and an

increase in the concavity of the floor (Figure 1.1A-D; Papan et al. 2007ab;

Winklbauer and Damm, 2011). It is not clear what drives these movements,

however active migration of cells toward the blastocoel floor and the constriction

of the outer epithelium at the base of them embryo are possible mechanisms

(Winklbauer and Damm, 2011).

After the start of gastrulation, the inward movement of vegetal cells becomes

more intense in the periphery of the dorsal vegetal cell mass (Figure 1.1B;

Winklbauer and Schurfeld, 1999). Like many of the gastrulation movements

discussed up to this point, vegetal rotation is driven by active, tissue autonomous

  20

movements that can be observed in explants of vegetal tissue (Winklbauer and

Schurfeld, 1999). Peripheral cells are elongated in an animal-vegetal direction,

parallel to the BCR (Damm and Winklbauer, 2011). Furthermore, small

protrusions are occasionally seen at the animally pointing ends (Damm and

Winklbauer, 2011). The orientation and morphology of these cells is consistent

with animally oriented active cell migration. Vegetal rotation not only result in the

internalization of endoderm but apparently plays a role in the initial phase of

mesoderm internalization. The rotation of the dorsal marginal zone, which results

in the apposition of the mesoderm to the inner layer of the BCR, appears to be

the result of the expansion of the blastocoel floor and the inward movement of

cells at the base of the embryo (Winklbauer and Schurfeld, 1999). Thus, although

the molecular regulation of vegetal rotation has yet to be elucidated, the role in

gastrulation played by this movement is instrumental to the internalization of

endoderm and possibly mesoderm. Furthermore, epiboly of the dorsal marginal

zone ectoderm, which may be associated with the rotation of mesoderm in the

dorsal marginal zone, may be indirectly influenced by this process.

1.1.5: Directional cell migration of the anterior mesoderm

The anterior mesoderm can be subdivided into two regions, the PCM, which

expresses the transcription factor goosecoid (Gsc) and the leading edge

mesendoderm (LEM) (Winklbauer, 1990; Winklbauer and Nagel, 1991; Niehrs et

al. 1993). Directional migration of the LEM has been well described and is

  21

thought to be an important process in the spreading of mesodermal cells over the

inner surface of the embryo. The region moves autonomously as a cohesive

group of cells toward the animal pole on the adhesive substrate of the inner BCR

(Figure 1.1B,C; Winklbauer, 1990; Winklbauer and Nagel, 1991). Thus, LEM

migration is a model for directional collective cell migration. Cells that are in

contact with the BCR extend lamelliform protrusions in the direction of migration,

toward the animal pole (Winklbauer and Selchow, 1992; Nagel et al, 2004).

Typically, models of LEM morphogenesis have required that the cells use the

BCR inner surface as a substrate for migration. Indeed explants of LEM tissue

can attach and migrate directionally in vitro on explanted BCR or on extracellular

matrix transferred from the inner layer of the BCR to a glass slide (Nakatsuji and

Johnson 1983a; Winklbauer, 1990; Winklbauer and Nagel, 1991; Winklbauer et

al. 1996). However, in vivo it appears that some locomotory protrusions attach to

adjacent mesoderm cells rather than the BCR (Figure 2.8C). Furthermore,

transmission electron micrographs of the mesoderm/ectoderm boundary do not

show protrusions crossing the separation gap (Figure 2.16A). Thus, intercellular

migration (the migration of cells over other cells) may play an important role in

this process. Thus, the model of LEM morphogenesis requires updating to take

these new observations into account.

Fibrillar fibronectin plays an important role in regulating LEM migration. At the

  22

start of gastrulation, inner layer BCR cells assemble a layer of fibrillar fibronectin

on the inner surface of the BCR (Nakatsuji and Johnson, 1983b; Winklbauer and

Nagel, 1991; Winklbauer, 1998). This matrix is required for gastrulation in

urodeles and plays an important role in the regulation of cell migration in

amphibians like Xenopus (Boucaut et al. 1984a; Boucaut et al. 1984b; Darribere

et al. 1988; Nakatsuji and Johnson, 1983a; Nakatsuji and Johnson, 1982). Even

though fibronectin is known to have a role in cell-matrix adhesion in vitro (Schlie-

Wolter et al, 2013), the attachment of LEM cells to the BCR did not depend on

fibronectin, suggesting that cell-cell adhesion molecules such as cadherins may

regulate adhesion here (Winklbauer and Keller, 1996). Interactions between this

fibronectin matrix and LEM cells are required for the extension of lamelliform

protrusions (Winklbauer and Selchow, 1992). Thus, fibronectin may play a

signaling function that regulates the stability of cellular protrusions, which is

required for cell migration. Interestingly, the BCR has an innate polarity, which is

observed in extracellular matrix and associated molecules and can be measured

by cell behaviour in vitro (Nagel and Winklbauer, 1999). The BCR matrix and any

associated molecules can be transferred to a glass slide for in vitro studies

(Nakatsuji et al, 1983a). When explants of LEM cells are placed on this

“conditioned substrate”, the explants move toward the end of the slide that

corresponds to matrix from the animal pole (Nagel and Winklbauer, 1999; Nagel

et al, 2004). This indicates that there is a matrix associated guidance signal. Lf-

PDGF-A has been shown to be involved (see below). Thus, fibronectin and

  23

PDGF-A together may function to stabilize lamelliform protrusions and provide

directional information, regulating the migration of LEM cells.

1.2: Platelet derived growth factor signaling in Xenopus gastrulation

PDGF molecules and their associated receptor tyrosine kinases belong to an

evolutionarily conserved family in vertebrates with diverse functions. PDGF and

PDGF receptors are proto-oncogenes and therefore, PDGF signaling is most

commonly studied for its role in cancer progression. In particular, improperly

regulated autocrine PDGF signaling has been implicated in tumour cell

proliferation, angiogenesis of tumours, epithelial to mesenchymal transition,

cancer metastasis and tumour resistance to drug therapies (Andrae et al. 2008;

Nister et al. 1988; Hermanson et al. 1992; Furuhashi et al. 2004; Jechlinger et al.

2003; Heuchel et al. 1999). However, like many of the genes that regulate cancer

processes, PDGF signaling is essential for a number of developmental processes

such as directional cell migration, programmed cell death, primordial germ cell

migration, neural crest morphogenesis and the organogenesis of several organs

(Andrae et al. 2008). During Xenopus development, PDGF signaling has been

found to regulate multiple processes (Ataliotis et al. 1995; Utoh et al. 2003; Van

Stry et al. 2004; Nagel et al. 2004; Van Stry et al. 2005; Andrae et al. 2008

Damm and Winklbauer, 2011). The diverse functions of PDGF signaling are

correlated with the large number of downstream signaling pathways that can be

  24

activated by the receptors.

I have shown that long range PDGF-A signaling is an essential regulator of

prechordal mesoderm morphogenesis (see Chapter Three), thus in this section I

will discuss the structure of the PDGF ligands and receptors, the dynamics of

their interactions and the long and short range signaling functions of PDGF.

PDGF-A signaling plays multiple roles in the development of the prechordal

mesoderm; therefore in order to put the morphogenetic role played by signaling

into context, the other roles of PDGF signaling during early Xenopus

development will be discussed in detail.

1.2.1: Platelet derived growth factor ligands and receptors PDGF family ligands  The PDGF family of ligands is made up of four members, PDGF-A, B, C and D,

encoded by four separate genes and divided into two subfamilies (Figure 1.3);

PDGF-A and B are type I ligands while PDGF-C and D are type II ligands. Type I

PDGFs are characterized by positively charged arginine and lysine residues in

their C-termini, (Figure 1.3; Raines and Ross, 1992; Kelly et al. 1993; Andersson

et al. 1994). Type II PDGFs have an N-terminal CUB domain, which is cleaved

after secretion (Figure 1.3; Li et al. 2000; Bergsten et al. 2001; Hoch and

Soriano, 2003). The PDGF ligands are closely related to members of the

Vascular Endothelial Growth Factor (VEGF) and Placental Growth Factor (PlGF)

  25

families (Ferrara et al. 2003; Fredricksson et al. 2004a). Homologues of these

molecules are found in species throughout the vertebrate subphylum (Andrae et

al. 2008). Molecules that are similar to the vertebrate PDGF/VEGF family

members have also been identified in invertebrates like Drosophila and C.

elegans. In Drosophila, three homologues called PDGF/VEGF-like ligand 1-3

(Pvf1-3) interact with a PDGF/VEGF-like receptor (McDonald et al. 2003;

Ducheck et al. 2001). Four receptor homologues (VER1-4) and a single ligand

(Pvf1) have been identified in C. elegans (Popovici et al. 2002; Hoch and

Soriano, 2003; Tarsitano et al. 2006). The similarities of the invertebrate ligand

and receptor homologues to vertebrate VEGF ligands and receptors may suggest

that the modern VEGF/VEGFR and PDGF/PDGFR groups diverged from VEGF

and VEGFR like ancestral molecules (Andrae et al. 2008).

PDGF molecules have a conserved PDGF/VEGF core growth factor domain that

contains a set of eight conserved cysteine residues that are important for

receptor binding and ligand dimerization (Hoch and Soriano, 2003; Andrae et al.

2008; Fredricksson et al. 2004a), which occurs by the formation of disulfide

bridges between these conserved cysteine residues (Mercola et al. 1990). PDGF

ligands typically form homodimers, however heterodimers are possible and a

PDGF-AB heterodimer has been identified in human platelets (Stroobant and

Waterfield, 1984). Interestingly, the binding of PDGF-AB has been shown to

  26

Figure 1.3: Platelet Derived Growth Factor Family Ligands

Schematic diagram of platelet derived growth factor (PDGF) family members.

The displayed amino acid (aa) lengths are taken from the Xenopus homologues

of the proteins (Xenbase). N, N-terminus; C, C-terminus.

  27

Figure 1.3: Platelet Derived Growth Factor Family Ligands

  28

activate different downstream signaling pathways from those activated by PDGF-

AA (Ekman et al. 1999).

PDGF family receptors    PDGF receptors are receptor tyrosine kinases that are made up of five

extracellular immunoglobulin loops, a transmembrane domain and a split tyrosine

kinase domain (Figure 1.4; Andrae et al. 2008). These receptors can be semi-

promiscuous with respect to ligand binding. PDGFRα typically binds PDGF-AA,

PDGF-BB and PDGF-CC homodimers as well as PDGF-AB heterodimers

(Andrae et al. 2008). PDGFRβ is less promiscuous than PDGFRα and typically

binds PDGF-BB and DD homodimers (Andrae et al. 2008). Heterodimers of

PDGFRα and PDGFRβ are also possible when expression overlaps (Klinghoffer

et al. 2002). Heterodimerized receptors can bind all of the known PDGF ligand

dimers with varying affinities (Andrae et al. 2008).

  29

Figure 1.4: PDGFRα and associated signaling pathways

Schematic diagram of a PDGFRα homodimer. Tyrosine residues that are

important in interactions with downstream signaling partners are indicated by

numbers representing their amino acid position (from the Xenopus homologue).

  30

Figure 1.4: PDGFRα and associated signaling pathways

  31

Extracellular localization of PDGF ligands  The localization and activation of PDGF ligands in tissues depends on the ability

of the ligand to associate with the matrix surrounding the secreting cells and

proteolytic processing respectively. In mammals and Xenopus, PDGF-A

undergoes alternative splicing to produce long and short PDGF-A variants that

differ in their abilities to bind extracellular matrix (Figure 1.3; Raines and Ross,

1992; Andersson et al. 1994; Hoch and Soriano, 2003; Andrae et al. 2008). In

chick, alternative splicing generates three PDGF-A isoforms, one of which has

matrix binding potential (Horiuchi et al. 2001). PDGF-A long isoforms have a C-

terminal cell retention motif made up of positively charged amino acids that has

been shown to interact with extracellular matrix molecules like fibronectin and

HSPGs (Raines and Ross, 1992; Andersson et al. 1994; Smith et al. 2009). Long

PDGF-A isoforms do not undergo proteolytic processing of their C-termini and

can remain associated with extracellular matrix in order to perform short range

signaling functions (Nagel et al. 2004, Smith et al. 2009; Damm and Winklbauer,

2011). Short PDGF-A isoforms are soluble molecules that can act at a distance

from their source (Damm and Winklbauer, 2011; Raines and Ross, 1992). PDGF-

B contains a retention motif similar to the one found in long PDGF-A isoforms,

however it undergoes extracellular processing by proteases to free it from

extracellular matrix. The processing enzymes involved here have not been

identified, although thrombin type proteases are candidate molecules (Kelly et al.

1993). PDGF-C and PDGF-D contain N-terminal CUB domains that require

  32

proteolytic processing in order for the ligand to become active. Tissue

plasminogen activator has been described as a potential activator of PDGF-C

(Fredriksson et al. 2004b).

1.2.2: Platelet derived growth factor ligands and receptors are expressed in adjacent tissue layers during development

The expression pattern of PDGF ligands and receptors are important in

determining the type of signaling that occurs. During development, paracrine

signaling is apparently the most common form of PDGF signaling since the

expression of PDGF ligands and receptors does not frequently overlap (Damm

and Winklbauer, 2011; Liu et al. 2002ab; Hoch and Soriano, 2002; Ataliotis et al.

1995; Ho et al. 1994; Orr-Urtreger and Lonai, 1992).

A common feature of PDGF ligand and receptor expression during development

is the complimentary expression of ligand and receptor in different but adjacent

cell layers. During mouse development, PDGF-A expression is largely found to

be restricted to epithelial layers such as the primitive ectoderm, surface

ectoderm, myotome, and the olfactory epithelium, among others (Orr-Urtreger

and Lonai, 1992). PDGF-B and PDGF-C are also expressed in cell layers

overlying mesenchymal cell populations (Hoch and Soriano et al. 2003; Aase et

al. 2002). Similarly, in chick embryos PDGF-A is expressed in the overlying

epiblast (Yang et al. 2008). The expression of PDGF receptors appears to be

  33

largely localized to the mesenchymal cells underlying these cell layers (Orr-

Urtreger and Lonai, 1992, Hoch and Soriano, 2003; Yang et al. 2008). However,

during early zebrafish development, PDGF-A and PDGFRα expression appears

to be ubiquitous (Liu et al. 2002a; Liu et al. 2002b). This observation is based on

detection of PDGF-A/Rα RNA in cells and therefore, a more restricted localization

of protein function cannot be excluded. However, at later stages of zebrafish

development the characteristic PDGF/PDGFR expression pattern is observed.

PDGF-A expression is found in overlying ectoderm with PDGFRα expression

localized to the mesenchymal cranial neural crest cells (Liu et al. 2002a; Liu et al.

2002b).

In Xenopus, similar patterns of expression are observed during early and later

development. During gastrulation, PDGF-A expression is found localized to the

inner layer cells of the overlying ectoderm while PDGFRα expression is restricted

to the underlying mesenchymal like mesoderm cells (Ataliotis et al. 1995). During

later stages of Xenopus development, PDGF-A expression is found in the

neuroectoderm, otic vesicle and pharyngeal endoderm while,

PDGFRα expression is found primarily in the cephalic neural crest as they

migrate into the visceral arches (Ho et al. 1994). Effectively, PDGF-A is

expressed along the migratory route of the cephalic neural crest cells, which is

strongly suggestive of a role for PDGF-A in regulating movement in these cells.

The expression of PDGF and PDGFRα in adjacent cell layers allows for

  34

communication between the overlying layers and deeper mesenchymal cells.

During development, this expression pattern permits PDGF signaling to regulate

chemotaxis, cell fate specification, cell survival and other cellular functions of

these mesenchymal populations.

1.2.3: Signaling downstream of platelet derived growth factor receptors

PDGFRα and PDGFRβ are receptor tyrosine kinases that activate a similar

complement of downstream signaling pathways. Binding of PDGF homo/hetero

dimers triggers receptor dimerization (Hoch and Soriano 2003). Typically,

homodimerization occurs because expression of the two receptors rarely

overlaps, however heterodimerization can occur, for example, in mouse postnatal

neurons (Vignais et al. 1995; Hoch and Soriano, 2003). PDGF receptors contain

five extracellular immunoglobulin loops, the outer three of which are important for

interactions with PDGF ligands (Figure 1.4; Heidaran et al. 1990; Yu et al. 1994).

PDGF ligands likely play a role in keeping receptors together during dimerization,

however the receptors themselves have been shown to physically interact

through their fourth immunoglobulin loops (Omura et al. 1997). This may suggest

that ligand independent signaling can occur due to stochastic receptor

interactions. Upon ligand binding and receptor dimerization, autophosphorylation

occurs resulting in the phosphorylation of key tyrosine residues in the kinase

domain of the receptor. These phosphorylated tyrosine residues are Src

  35

homology 2 (SH-2) docking sites for downstream signaling effectors (Figure 1.4).

PDGF associated signaling pathways  PDGFRα and PDGFRβ activate downstream signaling pathways that are

common to a number of growth factor signaling pathways such as Src, PI3K,

PLCγ, as well as indirect links to the Ras-MAPK pathway through activation of

Shc and Grb2 (Heldin et al. 1998). These downstream effectors belong to

signaling pathways that typically function to regulate developmental processes

like cell migration and cell differentiation (Andrae et al. 2008). During gastrulation,

cell migration is a key process and the regulation of this process in this context

by PDGF signaling is of interest. PI3K is an enzyme that phosphorylates

phosphoinositides on the inner surface of the cell membrane, converting

phosphoinositol (4,5)-bisphosphate (PIP2) into phosphoinositol (3,4,5)-

triphosphate (PIP3) (Burgering and Coffer, 1995). PIP3 is a docking site for

pleckstrin homology domain (PH-domain) containing effectors at the cell surface.

The formation of a PIP3 intracellular gradient at the leading edge of migratory

cells has been described as an important consequence of chemotactic gradient

detection (Sasaki et al. 2000; Funamoto et al. 2002). Some of the effectors that

interact with PIP3 include members of the Ras superfamily of small-GTPases

including the RhoGTPases and their associated exchange factors (RasGEFs)

(Hawkins et al. 1995; Fukata et al. 2003). These molecules are known regulators

of the actin cytoskeleton. In this case, activation of GEFs at the leading edge of

  36

the cell through PIP3 interaction can result in the polarized activation of

RhoGTPase family members resulting in modulation of the actin cytoskeleton,

which is important for the cell to form locomotory protrusions (Fukata et al. 2003;

Stephens et al. 2008).

PLCγ function downstream of PDGF signaling has also been described as an

important regulator of directional cell migration. In this case, PLCγ is activated

downstream of the PDGF receptor resulting in the hydrolysis of PIP2, producing

diacyl glycerol (DAG) and inositol triphosphate (IP3), both of which are important

second messenger signaling molecules (Bornfeldt et al. 1995). High levels of IP3

in the cell can result in an increase in cytosolic calcium levels due to release from

cellular stores (Berridge, 1993). This increase in cytosolic calcium level and

decreased levels of PIP2, as a result of PLCγ hydrolysis, is hypothesized to

trigger actin filament disassembly through binding of capping proteins to the

barbed ends of actin filaments (Stossel, 1993; Bornfeldt et al. 1995). This

remodeling of actin filaments is conducive to the formation of new locomotory

protrusions at the leading edge of cells (Bornfeldt et al. 1995).

Regulation of PDGF signaling by receptor mediated endocytosis  In addition to triggering PDGF receptor dimerization and autophosphorylation,

ligand binding is thought to trigger receptor mediated endocytosis of the receptor

ligand complex in some cell types (Kawada et al. 2009). PDGF-

  37

B/Rβ based chemotaxis in fibroblasts has been shown to involve interactions

between PDGFRβ and a ternary protein complex made up of ELMO adaptor

proteins, DOCK4 (Rac1 GEF) and Grb2 (Kawada et al. 2009). PDGFRβ contains

an SH-2 docking site for Grb2, thus following PDGF ligand binding and receptor

autophosphorylation, a ternary complex containing Grb2 interacts with Dynamin

and binds to PDGFRβ. The receptor/ligand complex is subsequently endocytosed

and signaling results in modulation of the cortical actin cytoskeleton, leading to

directional migration of the cells (Kawada et al. 2009). Internalized

PDGFRβ receptors are recycled to the plasma membrane in a Rab4 dependent

process (Kawada et al. 2009). Thus, internalization of the receptor/ligand

complex is required for directional migration in some cell types. It is not known if

receptor mediated endocytosis is required for PDGF-A/ PDGFRα signaling.

1.2.4: Platelet derived growth factor signaling during Xenopus gastrulation

During Xenopus development, PDGF signaling is known to regulate processes

such as directional cell migration of the leading edge mesendoderm, prevention

of apoptosis in mesoderm cells and the epithelial to mesenchymal transition that

occurs during larval skin morphogenesis (Ataliotis et al. 1995; Nagel et al. 2004;

Van Stry et al. 2004; Van Stry et al. 2005; Utoh et al. 2003). In these cases,

signaling through PDGFRα by PDGF-A homodimers is involved. Although,

PDGF-B and PDGFRβ have been cloned in Xenopus, expression of PDGF-B has

  38

not been detected during gastrulation (Mercola et al. 1988). Thus, no role in early

development has been identified for PDGF-B/ PDGFRβ signaling.

PDGF-A signaling is required for gastrulation in Xenopus  Inhibition of PDGF-A signaling with dominant negative PDGF-A ligands and a

dominant negative PDGFRα construct results in severe embryonic phenotypes

(Ataliotis et al. 1995). During gastrulation, blastopore closure was significantly

delayed, although epiboly movements continued (Ataliotis et al. 1995; Nagel et

al. 2004; Damm and Winklbauer, 2011). Disruption of PDGF-A signaling did not

prevent convergent extension in activin induced animal caps, although at larval

stages, the anterior/posterior axis of the embryo was significantly truncated

(Ataliotis et al. 1995). Thus, PDGF-A signaling may not be involved in the

regulation of CM convergent extension, but may be important in regulating

morphogenesis in more anterior tissues, such as the PCM or the LEM.

Additionally, larval head development was severely disrupted. In many cases,

larval heads were smaller than normal, showing abnormally developed structures

and in some cases, heads were completely absent (Ataliotis et al. 1995). This

may suggest potential defects in the morphogenetic movements of the PCM and

cranial neural crest cells given the involvement of these cells in head

development (Bothe et al. 2000). Interestingly, PDGF-A and PDGFRα are

expressed along the migratory routes of both of these cell populations (Ho et al.

1994; Ataliotis et al. 1995; Damm and Winklbauer, 2011). Nervous system

development was also severely disrupted, with many larval stage embryos

  39

having an open back or “spina bifida” phenotype (Ataliotis et al. 1995). The

expression of the PCM marker, Gsc and the CM marker, XBra, was unaffected,

which indicates that mesoderm cell fates were not disrupted by inhibition of

PDGF-A signaling (Damm and Winklbauer, 2011; Ataliotis et al. 1995). Thus, it is

likely that many of the observed defects are due to disruption of morphogenetic

movements in the embryo.

PDGF-A signaling in mesoderm migration  Single mesoderm cells, when seeded on fibronectin extend lamelliform

protrusions as they spread on the substrate (Winklbauer and Selchow, 1992). It

has been shown that aggregates of Activin induced Xenopus mesoderm can

spread on fibronectin by producing lamelliform protrusions around the periphery

of the aggregate, similar to the way single mesoderm cells spread (Symes and

Mercola, 1996). Aggregate spreading requires the inclusion of PDGF-A in the

culture medium, otherwise the cells in the aggregates remain tightly associated

(Symes and Mercola, 1996). Furthermore, aggregate spreading requires PI3K

signaling downstream of PDFGRα. When Wortmannin, a specific PI3K inhibitor is

added to the culture medium, aggregates fail to spread, even in the presence of

PDGF-A (Symes and Mercola, 1996). Thus, as in other systems, signaling

through PI3K downstream of PDGF-A is important for cell motility in Xenopus

mesoderm (Symes and Mercola, 1996; Fukata et al. 2003; Stephens et al. 2008)

The LEM migrates toward the animal pole of the embryo on the inner layer

  40

surface of the BCR roof during gastrulation (Winklbauer and Nagel, et al. 2001;

Davidson et al. 2002; Nagel et al. 2004). This directional migration movement

depends on guidance cues that have been deposited in the BCR extracellular

matrix (Nagel and Winklbauer, 1999; Nagel et al. 2004). It has been shown that

the long, and likely the intermediate isoforms of PDGF-A, are important

regulators of this process (Nagel et al. 2004). LEM cells extend lamelliform

protrusions preferentially in the direction of the embryonic animal pole (Nagel and

Winklbauer, 1991; Nagel et al. 2004). When PDGF-A signaling is disrupted by

morpholino oligonucletoides or by expression of dominant negative constructs,

cells extend protrusions randomly (Nagel et al. 2004). The BCR extracellular

matrix and associated guidance cues can be transferred from the inner layer of

the BCR to a glass slide for in vitro studies (Nakatsuji and Johnson, 1983a; Nagel

and Winklbauer, 1999; Nagel et al. 2004). Aggregates of LEM typically migrate

toward the end of the glass slide that would correspond to the animal pole on the

BCR. This directional movement is disrupted in aggregates from embryos

expressing constructs encoding dominant negative PDGF-A or PDGFRα or

injected with PDGF-A morpholino oligonucleotides (Nagel et al. 2004).

Additionally, consistent with a role for PI3K signaling downstream of PDGF-A in

Xenopus mesoderm migration, Wortmannin treatments were also found to inhibit

directional migration of aggregates (Nagel et al. 2004). The overexpression of

PDGF-A constructs in the embryo or in aggregates of LEM, results in a similar

disruption of directional migration (Nagel et al. 2004). This suggests that PDGF-A

  41

signaling plays an instructive role in the regulation of directional migration of the

LEM on the inner layer of the BCR. This also suggests that a gradient of PDGF-A

may be present within the BCR extracellular matrix. The cell retention motif found

in the long and intermediate PDGF-A isoforms could be interacting with HSPGs

or fibronectin in the extracellular matrix (Andersson et al. 1994; Smith et al.

2009). This gradient could be formed by modulation of PDGF-A expression

between different zones of the BCR. PDGF-A molecules would be deposited in

the BCR extracellular matrix in a graded fashion. Indeed, in situ hybridization for

PDGF-A may provide evidence for this polarization, with increasing expression

towards the animal pole (Ataliotis et al. 1995; See Chapter Three, Figure 3.2).

There appears to be a counter-gradient of PDGFRα expression in mesoderm

tissue, with decreasing expression moving toward the vegetal pole (Ataliotis et al.

1995; See Chapter Three, Figure 3.2). Increased receptor expression in more

posterior cells may increase sensitivity for reduced levels of PDGF-A expression

in that region of the embryo. Thus, PDGF signaling plays an important role in the

regulation of cell movements during gastrulation.

Anti-apoptotic effects of PDGF-A signaling  The effect of inhibiting PDGF-A signaling during gastrulation in Xenopus causes

non-subtle, dramatic phenotypes (Ataliotis et al. 1995). PDGF-A has obvious

effects on cell movements, however some of the observed phenotypes could be

due to the loss of key cell populations in addition to defects in cell movement.

  42

Subsets of cells in the Xenopus gastrula undergo cell death at the early gastrula

transition (Howe et al. 1995). These cells die through Caspase3/7 activation

downstream of the mitochondrial apoptotic pathway (Thornberry and Lazebnik,

1998). PDGF-A signaling appears to be involved in ensuring cell survival in the

prechordal mesoderm population by preventing these cells from undergoing

apoptosis (Van Stry et al. 2004; Van Stry et al. 2005). When PDGF-A signaling is

inhibited by expression of a dominant negative form of PDGFRα, TUNEL-positive

PCM cells with increased Caspase 3 activity were found in the blastocoel (Van

Stry et al. 2004). Apoptosis could be rescued by expression of Caspase 3

inhibitors in these cells, however this did not rescue the dramatic embryonic

phenotypes caused by PDGFRα knockdown (Van Stry et al. 2004). PDGF

receptors that contain tyrosine residue docking sites for PI3K and PLCγ only were

sufficient to rescue apoptosis in PCM cells with defective PDGF signaling (Van

Stry et al. 2005). This shows that the anti-apoptotic effect of PDGF-A signaling

requires signaling by PDGFRα through the PLCγ and PI3K signaling pathways

(Van Stry et al. 2005). Thus, while the anti-apoptotic effects of PDGFRα signaling

may be important in gastrulation, the regulation of cell movements by PDGF-A

signaling is critical. Interestingly, the pathways regulating cell migration and the

anti-apoptotic effects of PDGF signaling appear to overlap. This could suggest a

mechanism where cells that fail to migrate normally because of defective

signaling through the PI3K or PLCγ pathways, undergo programmed cell death to

ensure that they do not accumulate in incorrect locations in the embryo.

  43

1.3: Molecular gradient formation: mechanisms and kinetics.  During embryonic development, gradients of signaling molecules are formed in

embryonic tissues. Morphogen gradients are an example of a molecular gradient.

Different cell fates are specified by different concentrations of molecules along

the gradient and in this way, morphogen gradients play an essential role in the

patterning of embryonic tissues. Chemoattractant gradients are another type of

molecular gradient. In this case, the gradient of molecules provides directional

information so that cells can polarize or migrate in the direction of the source of

the molecules. The morphogen gradients responsible for patterning the

anterior/posterior axis of the Drosophila wing imaginal disc have been subject to

extensive study (Baker et al. 2007). These studies have contributed to the

development of several theories about how gradients form and spread through

tissues.

My work has identified that the short isoform of PDGF-A functions as a long

range signaling molecule and that it forms a molecular gradient that is

responsible for orienting and directing the movements of prechordal mesoderm

cells during gastrulation in Xenopus (See Chapter Three). The mechanism

responsible for forming this type of growth factor gradient is not clear, however

the observations suggest that a mechanism involving the diffusion of molecules

through extracellular spaces fits better than others as potential spreading

  44

mechanisms (See Chapter Four). Thus, the theoretical models for gradient

formation and the techniques used to distinguish between the models in different

systems will be reviewed in this section.

1.3.1: Mechanisms of molecular gradient formation  The patterning of the wing imaginal disc in Drosophila larvae involves several

gradient forming molecules from different molecular families (Baker et al. 2007).

The different properties of these molecules require different mechanisms of

gradient formation and as such the wing imaginal disc has served as an excellent

system for developing models of mechanisms regulating gradient formation.

Since the molecules involved in patterning the wing imaginal disc are conserved

in vertebrates, it is possible that the gradient forming mechanisms are conserved

as well. Regardless of the mechanism involved in the spreading of the molecules,

gradients in tissues form because there is a source of transmittable molecules

that are capable of moving through a tissue (see below). These molecules must

also have a sink, usually uniform degradation, which is required to shape the

gradient (Wolpert, 1969). Without a sink, molecules would continue to move into

the tissue until it becomes saturated, that is, the concentration of molecules at

positions distant from the source would be the same as the concentration at

positions close to the source. This can be seen experimentally when following the

movement of a secreted-GFP constructs, which lack a sink in tissues. In this

case, secreted-GFP continues to move into the tissue until a uniform

  45

concentration is reached (Entchev et al. 2000; Yu et al. 2009).

The mechanisms proposed to regulate gradient formation in the Drosophila wing

imaginal disc include planar transcytosis, the “bucket brigade” mechanism,

morphogen transport by argosomes, morphogen transport by cytonemes, free

and restricted diffusion of molecules, and gradient shaping by extracellular

proteoglycans (Kerszberg and Wolpert, 1998; Entchev et al. 2000; Greco et al.

2001; Panakova et al. 2005; Ramirez-Weber and Kornberg, 1999; Hsiung et al.

2005; Kicheva et al. 2007; Zhou et al. 2012; Yan and Lin, 2009). Each of these

models will be discussed in detail below.

Planar transcytosis  Planar transcytosis involves the spreading of transmittable molecules though

cells rather than through extracellular spaces in order to spread through a tissue.

During planar transcytosis, transmittable molecules are endocytosed by cells

closer to the molecule source, these molecules move through the cytoplasm of

the cell before being secreted through a recycling pathway (Yan and Lin, 2009).

The re-secreted molecules are then subsequently endocytosed by the next layer

of cells and the cycle repeats. Thus, planar transcytosis relies on subsequent

cycles of endocytosis and re-secretion in order to spread the molecules through

the tissue (Yan and Lin, 2009). Decapentaplegic (Dpp) is a signaling molecule

that belongs to the highly conserved transforming growth factor β (TGFβ)

superfamily of signaling molecules. It is produced in the anterior/posterior

  46

compartment boundary of the Drosophila wing imaginal disc and plays an

essential role in determining cell fates along the anterior/posterior axis of the disc

(Basler and Struhl, 1994; Lecuit et al. 1996; Nellen et al. 1996; Baker et al. 2007).

An initial study of Dpp morphogen gradient formation used Dpp molecules tagged

with green fluorescent protein (GFP) (Dpp-GFP) for visualization of the

molecules. Dpp-GFP is typically found intracellularly in wild-type wing imaginal

disc cells (Entchev et al. 2000; Kicheva, et al. 2007; Zhou et al. 2012).

Interestingly, it was observed that mutant shibire clones (shibire mutants have a

mutation in the Drosophila homologue of Dynamin) did not permit the spreading

of Dpp through the clone (Entchev et al. 2000). This result suggested that Dpp

gradient formation required Dynamin function, likely for endocytosis. The

intracellular distribution of Dpp-GFP in wild-type cells, together with the shibire

mutant data, would be consistent with a transcytosis mechanism (Entchev et al.

2000). Kinetics of the Dpp morphogen gradient have been determined using

fluorescence recovery after photobleaching (FRAP) techniques. The diffusion

coefficient derived from this data was three orders of magnitude lower than what

would have been expected for the free diffusion of the Dpp-GFP molecule in

water, but was thought to be consistent with a facilitated transport mechanism

(Kicheva et al. 2007). This was further supporting data for the planar transcytosis

hypothesis. However, more recent data that was gathered using a method called

Fluorescence Correlation Spectroscopy (FCS) (see below) derived diffusion

coefficients that were more in line with what would be expected for free diffusion

  47

of Dpp-GFP (Zhou et al. 2012). Therefore, the diffusion coefficient derived with

the FRAP method may have been an artefact of assumptions made by the

investigators and the techniques used (Zhou et al. 2012). Furthermore, it has

been suggested that the lack of endocytosis in shibire mutant clones could

prevent removal of receptors from the cell surface (Yan and Lin, 2009; Kicheva et

al. 2012). This lack of receptor removal would result in an accumulation of

receptors at the cell surface that could then function as an extracellular sink for

Dpp by binding Dpp molecules. This would prevent the molecules from moving

deeper into the tissue. Thus, a role for planar transcytosis has not been

definitively identified.

The “Bucket-Brigade” mechanism  Signaling by TGFβ family molecules such as Activin, BMPs and TGFβs involves

the binding of the ligand molecule to type II TGFβ receptors with high affinity,

which triggers heterodimerization of the receptor with a type I TGFβ receptor

(Kerszberg and Wolpert, 1998). Heterodimerization, results in

autophosphorylation of the receptors and in a lower affinity interaction with the

bound ligand (Kerszberg and Wolpert, 1998). These observations were used to

design a theoretical model of molecular spreading in tissues (Kerszberg and

Wolpert, 1998). In this model, after binding of the ligand with high affinity to a

receptor monomer, the dimerization of the receptors is triggered. Receptor

heterodimers have a lower affinity for the ligand, thus the ligand is released,

however the receptor dimers persist and continue to signal. The free ligand

  48

molecule will move to a new receptor monomer preferentially because of the high

affinity interaction between ligand and receptor. Cycles of ligand binding, receptor

heterodimerization and ligand release continue along the surface of the same cell

until close apposition of cells allows a free ligand to be passed to a receptor

monomer on an adjacent cell. Thus, in this model ligands are passed around the

cell and to adjacent cells, allowing for signaling and movement of the ligand

through the tissue. Computer simulations of this model have been successful in

forming a molecular gradient, however a role for this model in vivo has not been

identified and thus, remains a theoretical possibility for molecular spreading

(Kerzsberg and Wolpert, 1998).

Molecular transport by cytonemes  Not all molecules have the capability to move freely in a tissue. Structural motifs

and post-translational modifications of proteins can be responsible for retaining a

secreted molecule in the vicinity of the secreting cell. PDGF-A long isoforms for

example, include a positively charged region in the C-terminus that interacts with

the negatively charged sulfate groups of heparin sulfate proteoglycans (HSPGs)

(Andersson et al. 1994; Andrae et al. 2008). This interaction restricts the

molecule to the region near the secreted cells (Andersson et al. 1994; see

Chapter Three). Furthermore, the signaling molecule, hedgehog (Hh), undergoes

post-translational modifications that result in the addition of cholesterol and a

palmitoylation group to its C and N-termini respectively (Porter et al. 1996;

Pepinsky et al. 1998). This post-translational lipid modification of the protein

  49

results in retention of the Hh near the surface of the secreting cells. Thus, these

types of molecules are not likely to be transmitted well through the extracellular

space because they would likely become associated with the surfaces of cells

before a gradient could spread a significant distance.

The cells of the Drosophila wing imaginal disc are known to extend thin but long

filopodia like processes called cytonemes (Ramirez-Weber and Kornberg, 1999).

Similar structures have been observed in cultured mammalian cells when

induced with Fgf4 (Ramirez-Weber and Kornberg, 1999). Thus, these structures

are not unique to invertebrates. Interestingly, cells at varying distances from the

wing disc anterior/posterior compartment boundary extend these cytonemes in

the direction of the compartment boundary and apparently make physical contact

with the cells in the boundary region (Ramirez-Weber and Kornberg, 1999).

Thus, it has been hypothesized that cytonemes facilitate long range signaling

between the compartment boundary cells and cells located at varying distances

from the compartment boundary (Ramirez-Weber and Kornberg, 1999; Hsiung et

al. 2005). Signaling molecules produced by cells at the compartment boundary

could travel the distance to distant cells along cytonemes without the need to

travel through the extracellular space. Therefore the formation of an extracellular

gradient would not be required. Instead, the signaling molecule could travel along

the cytonemes, intracellularly, degrading over time, forming an intracellular

gradient (Ramirez-Weber and Kornberg, 1999). Dpp signaling, which is important

for determining cell fates along the anterior/posterior axis of the wing imaginal

  50

disc is known to form a gradient (Kicheva et al. 2007, Zhou et al. 2012).

However, interestingly, the receptor for Dpp signaling, Thickveins (Tkv), has been

observed as puncta that are motile along the length of the cytonemes (Hsiung et

al. 2005). Thus, if Dpp is associated with the Tkv puncta, it is possible that

cytonemes could play a role in the distribution of Dpp to locations distant from the

source at the compartment boundary (Hsiung et al. 2005). However, an

alternative possibility that has not been ruled out, suggests that Dpp signaling

could play an important role in regulating cytoneme formation or function (Hsiung

et al. 2005). Thus, signal spreading though cytonemes is a potential alternative

mechanism for transport of molecules like the long isoforms of PDGF-A in

vertebrates or lipid modified molecules such as Hh, which would be stopped by

extracellular interactions with cells and extracellular matrix.

Spreading of membrane associated molecules by argosomes and lipoproteins  Cytonemes provide an intracellular solution for the problem of the spreading of

lipid modified molecules and other molecules that become associated with the

cell membrane after secretion. Another potential mechanism for the spreading of

these molecules involves exosomes. In the Drosophila wing imaginal disc,

exosome-like structures have been termed argosomes and are membrane

exovesicles that are produced from the baso-lateral membrane of epithelial cells

(Greco et al. 2001). When argosomes are produced, the membrane and

associated molecules at that location on the cell surface, including any bound

signaling molecules, are formed into exosome like structures. The argosomes

  51

would be released into the extracellular space when the endosome fuses with the

cell plasma membrane and subsequently interact with and are endocytosed by

receiving cells (Greco et al. 2001). Wingless (Wg) and Hedgehog (Hh) lipid

modified molecules that are typically retained at the cell surface could be

transported by argosomes to locations distant from their source. Indeed,

argosome like structures are produced from cells expressing high levels of Wg

(Greco et al. 2001). Interestingly, Wg containing argosomes like structures were

found to move at a similar rate to that of Wg spreading in the wing imaginal disc

(Greco et al. 2001). Argosomes/exosomes could also play a role in the transport

of membrane associated growth factor molecules like cell retention motif

containing isoforms of PDGF-A or membrane associated variants of FGF ligands.

Exosome-like structures are not exclusive to invertebrates. Fibroblasts can

produce exosomes and interestingly, Wnt11, a vertebrate homolog of Wg-like

molecules, has been found to associate with these exosomes through which

intercellular communication between fibroblasts and surrounding breast cancer

cells occurs (Luga et al. 2012). Together, these observations suggest a potential

role for extracellular vesicles in the transport of signaling molecules.

Although argosomes/exosomes are potential candidate structures for the

spreading of membrane tethered proteins, the model describing the spreading of

the Wg and Hh signaling molecules has been updated to involve a different type

of structure. Lipoprotein particles have been shown to be important in the

  52

formation of Wg and Hh morphogen gradients (Panakova et al. 2005).

Lipoprotein particles are secreted structures made up of a core of neutral lipids

surrounded by polar phospholipids; cholesterol and large apolipoproteins are

embedded in these phospholipids (Yan and Lin, 2009; Panakova et al. 2005).

Lipoprotein particles are known to transport lipid molecules like cholesterol and

as such may be involved in the transport of lipid modified signaling molecules

(Yan and Lin, 2009). Lipophorin is a Drosophila apolipoprotein and it has been

found to interact with Wg and Hh in biochemical analyses and it also co-localizes

with both proteins in wing imaginal disc cells (Panakova et al. 2005). Knockdown

of lipophorin resulted in a significant restriction in the range of Hh and Wg

signaling activity, effectively restricting the signaling to near the source of the

molecules (Panakova et al. 2005). Thus lipoprotein particles are required for the

formation of Hh and Wg morphogen gradients. Lipoproteins are also found in

vertebrate embryos (Farese et al. 1996; Willnow et al. 2007), thus this method of

transport for lipid modified signaling molecules could be ubiquitous during

development.

Heparin sulfate proteoglycan (HSPG) interactions  Secreted proteins often contain domains that interact with the negatively charged

sulfate groups found on HSPGs. HSPGs are members of a highly conserved

extracellular protein family in which a core protein is associated with a heparin

sulfate glycosaminoglycan side-chain. The family can be divided into two cell

associated types, the syndecans and glypicans and a diffusible type, the

  53

perlecans (Bernfield et al. 1999). Interactions between glipicans, syndecans and

extracellular signaling proteins have the potential to significantly affect the shape

of a molecular gradient. In vertebrates, the cell retention motif containing

isoforms of PDGF-A are thought to bind to cell surface HSPGs (Andersson et al.

1994) and this interaction likely restricts the signaling role of these isoforms to

near the secreting cells (see Chapter Four). Furthermore, the Drosophila

glypicans, Dally and Dally-like (Dlp), play a major role in the shaping of Wg, Hh

and Dpp gradients in the Drosophila wing imaginal disc. Mutations in genes that

are involved in the biosynthesis or regulation of HSPGs, often results in inhibited

Wg, Hh and Dpp signaling and a lack of extracellular accumulation for these

molecules (Lin, 2004). Thus, a major function of HSPGs in gradient formation

could be to encourage the accumulation of Wg extracellularly where signaling is

required, particularly at locations distant from the source where the concentration

of the molecule would be lower. HSPG’s may also be involved in enhancing the

stability of these extracellular signaling molecules (Lin and Perrimon, 2000).

HSPGs could also function as co-receptors and indeed Dally has been implicated

in such a function in Wg signaling (Baeg et al. 2001).

An interesting possibility that was proposed to explain why Dpp does not cross

cell clones lacking Dally is that HSPGs could function to facilitate the transport of

signaling molecules through the extracellular space. In this model, the side

chains of HSPGs would bind to signaling molecules like Dpp and those

  54

molecules would subsequently be handed to the next side chain, then the next

and so an, effectively passing the molecule along the gradient (Belenkaya et al.

2004) This is similar to the hypothetical bucket brigade mechanism (Kerszberg

and Wolpert, 1998). Thus, the exact role of HSPGs in molecular gradient

formation is not well understood. However HSPGs may play roles in restriction of

movement, facilitation of movement and stabilization extracellular signaling

molecules, which results in the modulation of the shape of molecular gradients.

Gradient formation by free diffusion  Over four decades ago, it was shown theoretically, that simple diffusion of

molecules, over the distances and timescales required when patterning a tissue,

could form a molecular gradient (Crick, 1970). Since then, diffusion of molecules

through extracellular spaces has been described as a mechanism of molecular

gradient formation in several systems, including the formation of the Dpp gradient

in the Drosophila wing imaginal disc, the Fgf8 gradient in the Zebrafish embryo

and the Activin gradient in the early Xenopus embryo (McDowell et al. 1997;

McDowell et al. 2001; Gregor et al. 2007; Kicheva et al. 2007; Hagemann et al.

2009; Yu et al. 2009; Zhou et al. 2012). As discussed above, Dpp gradient

formation has been subject to several hypothetical models of gradient formation.

Recently however, kinetic data gathered using fluorescence correlation

spectroscopy (FCS), has led to a revision of the model (Zhou et al. 2012). FCS is

a single molecule resolution technique that works by measuring the fluctuations

in fluorescent signals over very short time scales within a small half cubic

  55

micrometer volume (Ries et al. 2009; Zhou et al. 2012; Kicheva et al. 2012).

Autocorrelation analysis then allows for determination of diffusion coefficients as

well as providing information about whether molecules move independently or as

aggregates/molecular complexes (Kicheva et al. 2012). Data gathered by FCS

currently supports a model of rapid extracellular diffusion of Dpp molecules,

rather than facilitated transport mechanisms like planar transcytosis, cytoneme

transport or restricted diffusion (Zhou et al. 2012). Diffusion of individual

molecules happens very rapidly (diffusion coefficient of GFP in water is 87µm2/s,

Swaminathan et al. 1997). The majority of Dpp molecules (65%) were found to

have a diffusion coefficient of 21 µm2/s, a rate that is consistent with free diffusion

of these molecules (Zhou et al. 2012). The remaining 35% of extracellular

molecules had a significantly lower diffusion coefficient of 0.03 µm2/s, suggesting

that these molecules are associated with extracellular HSPGs or receptors (Zhou

et al. 2012). Thus, the rapid transport via free diffusion of the majority of Dpp

molecules dominates the spreading process because transport of the Dpp

molecule happens very rapidly compared with binding to HSPGs and receptor

mediated endocytosis (Zhou et al. 2012). As a result of this rapid spreading

process, few molecules need to be motile in the extracellular space because they

are rapidly replaced as they get degraded or become associated with immobile

structures (Zhou et al. 2012). Spreading by diffusion is coupled with intracellular

degradation of Dpp after internalization through receptor mediated endocytosis

(Kicheva et al. 2007; Zhou et al. 2012). Thus, while Dpp interacts with potentially

  56

immobilizing structures such as extracellular HSPG molecules (Belenkaya et al.

2004; Akiyama et al. 2008; Fujise et al. 2003), the interactions with these

structures is considered to be a downstream event, that is, these events occur

slowly and after the molecules have already diffused through the tissue (Zhou et

al. 2012). Thus, while interactions with receptors, co-receptors, and HSPGs at

the cell surface serves to stabilize the gradient through degradation, stabilization

of receptor/ligand interactions, or by concentrating molecules at specific locations

for signaling purposes, the transport rate of Dpp is not affected by these

interactions because diffusion of molecules is several orders of magnitude faster

(Zhou et al. 2012). The interactions with receptors and HSPGs are however,

essential to the formation of a stable gradient. Secreted forms of GFP fail to form

gradients because without stable substrates to bind to, the molecules diffuse out

of the tissue so rapidly that it is not possible to form a stable gradient (Zhou et al.

2012).

Rapid intercellular diffusion of molecules can form gradients in vertebrates as

well. Similar to Dpp gradient formation, Fgf8 in the zebrafish neuroectoderm

spreads by free diffusion of molecules (Scholpp and Brand, 2004; Yu et al. 2009).

Using FCS techniques, it was determined that Fgf8 moves through extracellular

spaces rapidly by free non-directional diffusion (diffusion coefficient of 53µm2/s)

(Yu et al. 2009). A small percentage (~9%) of extracellular Fgf8 was found to

move at a significantly reduced transport rate (Yu et al. 2009). This was due in

  57

part to interactions with HSPG molecules as treatment with heparinase, an

enzyme that degrades HSPGs, resulted in a decrease in the amount of Fgf8 with

the reduced transport rate (Yu et al. 2009). Interestingly, this treatment also

resulted in an expansion of the expression domains of genes downstream of the

FGF signaling pathway, indicating that the range of Fgf molecule influences can

be regulated by interactions with HSPGs (Yu et al. 2009). Additionally, the Activin

gradient, which is important in the specification of mesoderm during Xenopus

development (Asashima et al. 1990), is thought to be formed by free diffusion

through the extracellular space (McDowell et al. 1997; Hagemann et al. 2009).

XBra is a downstream target of Activin/TGFβ signaling and is upregulated at a

distance from the Activin source (McDowell et al. 1997). Interestingly, the Activin

signal is capable of moving through the extracellular spaces of groups of cells

that lack the receptor required for Activin signaling, ruling out a relay mechanism

for signal spreading (McDowell et al. 1997). The range of the gradient can be

modulated by the number of receptors on the surfaces of responding cells

(Hagemann et al. 2009), similar to what is observed in Fgf8 gradient formation in

Zebrafish (Scholpp and Brand, 2004). Higher numbers of receptors results in a

reduced range of the gradient, likely because there are more available receptors

close to the Activin source to bind and capture Activin molecules, reducing the

available pool of freely diffusing molecules. Interestingly, even though

internalized Activin is found in lysosomes, inhibition of endocytosis does not

significantly affect the signaling rage of the gradient (Hagemann et al. 2009).

  58

Although the reason for this is unknown, it has been hypothesized to be an effect

of the available space for diffusion of Activin in the in vitro system used to study

the gradient (Hagemann et al. 2009).

1.3.2: Kinetics of gradient formation: tissue level versus molecular level.

To date, many of the studies of gradient formation have analyzed the kinetics

involved on the tissue level. That is, by analyzing the distribution of molecules

over the entire length of the gradient or by analyzing the range of activation of

signaling processes downstream of gradient forming signaling molecules

(Entchev et al. 2001; Gregor et al. 2007; Kicheva et al. 2007; Wartlick et al. 2011;

Yan et al. 2009; Schwank et al. 2011). These measurements are usually taken

over time periods measured in hours. Typically, tissue level kinetic parameters

are determined with analysis by fluorescence recovery after photobleaching

(FRAP). FRAP works by photobleaching an area of interest, extracellular spaces

for example, and then analyzing the kinetics of the recovery of the fluorescent

molecules back into the photobleached area (Lippincott-Schwartz et al. 2003;

Kicheva et al 2009; Zhou et al. 2012; Kicheva et al. 2012). The kinetic

parameters are derived from the resulting FRAP recovery curves. These

analyses can be used to derive diffusion coefficients and rates of degradation

(Kicheva et al. 2007; Kicheva et al. 2012). Tissue level studies can be particularly

useful for understanding how gradient formation is affected by mutations in

potential regulatory genes. For example, in the Drosophila wing imaginal disc, the

  59

way that induced clones with mutations in genes encoding regulatory proteins

affects the movement of molecules around and through the clones can be readily

studied. This kind of study can provide valuable information about interactions

between cells, extracellular matrix and signaling molecules.

Kinetic parameters, such as the diffusion coefficient, that are derived from tissue

level studies can be significantly different from the same parameters derived from

measurements made on the molecular level. This is particularly the case when

molecules move through tissues by rapid free diffusion, as is the case for Dpp

and Fgf8 (Zhou et al. 2012; Kicheva et al. 2012; Yu et al. 2009; Kicheva et al.

2012). This is because analysis of diffusion through tissues has been shown to

be extremely rapid compared with the rates of receptor mediated endocytosis or

reversible binding from HSPGs (Zhou et al. 2012; Kicheva et al. 2012).

Therefore, the rate kinetics determined reflects the degradation of molecules,

rather than the transport (Zhou et al. 2012). Since the transport and

establishment of the gradient has already occurred before the concentration of

molecules has built up to a detectable level, the fluorescence being analyzed is

likely due to interactions with HSPGs and/or receptors rather than representing

the diffusible fraction of molecules (Zhou et al. 2012). Therefore, the diffusion

coefficient determined from tissue level experiments is an “effective” diffusion

coefficient that likely depends on several different interactions between the

molecule and various structures along the path of movement. The diffusion

  60

coefficients derived from single molecule analysis is a property of the molecule in

question and depends only on the viscosity of the material through which the

molecule is diffusing (Zhou et al. 2012).

Single molecule analysis typically uses FCS and these analyses have revealed

that some gradients form by rapid diffusion of molecules; a conclusion that could

not be reliably made from the data obtained from tissue level analyses. Analysis

with FCS can also detect molecules moving at much slower rates. While the

fraction of molecules moving at a faster rate likely resperesents the transport

mechanism of the molecule (i.e. diffusion, transcytosis), the slower fraction of

molecules likely represents association with receptors and extracellular matrix

molecules (Zhou et al. 2012, Yu et al. 2009). It is unclear, however whether all of

the potential interactions of signaling molecules within their environment can be

sampled within the extremely small volume analyzed by FCS. Thus, combining

high resolution single molecule analysis with results from tissue level analyses

are necessary to understand both the underlying transport mechanism and the

types of interactions of molecules with the environmental factors that can affect

the overall rate of molecular gradient formation.

1.4: Outline and Objectives  Prechordal mesoderm morphogenesis (Chapter Two)  The Xenopus prechordal mesoderm (PCM) is an understudied region of the

Xenopus mesoderm, thus in order to develop an understanding of

  61

morphogenesis in this region, my first objective was to analyze the polarity and

orientation of PCM cells during gastrulation. This was done using scanning

electron microscopy in order to get high resolution views of cells and cellular

processes. I found that the long axes of PCM cells were oriented perpendicular to

the overlying blastocoel roof and that cells extended lamelliform protrusions in

the direction of the blastocoel roof. Futhermore, I analyzed the orientation of cells

in mesoderm regions on the lateral and ventral sides of the embryo and found

that the orientation of these cells were strikingly similar to the corresponding

regions on the dorsal side of the embryo. In order to image live cell movements, I

developed an explant system that was instrumental in confirming that PCM cells

undergo radial intercalation by directional intercellular migration toward the BCR.

Platelet derived growth factor signaling (Chapter Three)  PDGF-A signaling has been implicated in the directional migration of the leading

edge mesendoderm during Xenopus gastrulation (Nagel et al. 2004).

Furthermore, the expression patterns of ligand and receptor are such that PDGF-

A signaling could regulate directional migration of PCM cells. My next objective

was to determine whether PDGF-A signaling played a role in this system. I

showed that the BCR expresses three PDGF-A isoforms, one of which is

considered to be a diffusible molecule. Using a combination of scanning electron

microscopy and live cell imaging of explanted mesoderm and BCR tissue, I

showed that the diffusible PDGF-A isoform plays an instructive role in regulating

the directional migration of deep prechordal mesoderm cells. Thus, I was able to

  62

propose a model in which the diffusible isoform of PDGF-A forms an extracellular

chemoattractant gradient within the PCM tissue that regulates the radial

intercalation of the PCM during gastrulation.

PDGF-A forms an extracellular gradient in PCM tissue (Chapter Four)  My proposed model for the regulation of radial intercalation by PDGF-A signaling

required that a gradient of PDGF-A be established in the PCM. My final objective

was to attempt to visualize and derive kinetic parameters of the putative PDGF-A.

Using eGFP labeled PDGF-A constructs and explants of mesoderm and BCR

tissue, I confirmed that the diffusible short isoform of PDGF-A (sf-PDGF-A)

formed a steady state gradient, while the long and intermediate cell associated

isoforms of PDGF-A remained localized to the mesoderm/ectoderm interface.

Furthermore, I was able to determine the decay length λ, degradataion rate 𝑘,

and the effective diffusion coefficient 𝐷!"" parameters for sf-PDGF-A. Based on

the value determined for 𝐷!"", I suggest that the sf-PDGF-A gradient forms by

diffusion of molecules through extracellular spaces. My work describes a model

for the molecular regulation of radial intercalation in a vertebrate system.

  63

Chapter Two: The prechordal mesoderm undergoes intercellular migration and radial

intercalation during gastrulation

Sections of this chapter were published in Damm, E.W. and Winklbauer, R. (2011) PDGF-A controls mesoderm cell orientation and radial intercalation during Xenopus gastrulation. Development,

138: 565-575                                        

  64

2.1: Introduction Research in the field of mesoderm morphogenesis has resulted in a good

understanding of CM convergent extension and LEM directional migration in the

gastrulating Xenopus embryo. However, morphogenesis of the other regions of

mesoderm, ectoderm and endoderm and the importance of these movements in

gastrulation are not well understood.

Classically, the involuted dorsal mesoderm has been described as being

comprised of two domains, the posterior mesoderm (CM) and the head

mesoderm (including the LEM) (Keller et al. 1985; Winklbauer and Nagel, 1991).

The CM is well known for convergent extension movements (Keller et al. 2000;

Keller, 2004; Solnica-Krezel, 2005; Solnica-Krezel and Sepich; 2012), while the

head mesoderm has been thought to migrate as a whole in the direction of the

animal pole on the inner layer of the BCR, guided by direction determining cues

that are associated with the BCR extracellular matrix (Winklbauer and Nagel

1991; Nagel and Winklbauer, 1999; Davidson et al. 2002; Nagel et al. 2004).

However, more recent evidence, based on gene expression, has shown that

involuted mesoderm can be divided into three motility domains, the involuted part

of the XBra expressing CM, Gsc expressing PCM and the LEM (Smith et al.

1991, Wilkinson et al. 1990, Cho et al. 1991, Niehrs et al. 1994; Saint-Jeannet et

al. 1994, Damm and Winklbauer, 2011). Morphogenesis of the PCM region has

not been subjected to detailed investigation in Xenopus but has been a topic of

  65

discussion in zebrafish studies (Solnica-Krezel et al. 1995; Blader and Strahle,

1998; Chan et al. 2001; Montero et al. 2004). Importantly, cells located in a

region of dorsal mesoderm that likely corresponds to the PCM region have been

described as being oriented with their long axis perpendicular to the BCR (Keller

and Schoenwolf, 1977). This orientation is perpendicular to the direction of

mesoderm movement and suggests that PCM cells may undergo movements

independent of the animally directed migration of the anterior mesoderm. Thus,

the contribution of PCM morphogenesis to gastrulation is unknown.

The opaque nature of the Xenopus embryo makes the imaging of cell

movements challenging. Therefore studies of morphogenesis in Xenopus have

relied heavily on imaging of explanted tissue. The disadvantage of explant

studies is the disruption of the endogenous organization of embryonic tissues. In

order to understand the contribution of different morphogenetic movements to the

gastrulation process, it is essential to combine data from explant experiments

with data obtained from studies using intact embryos in order to ensure that what

is observed in the explant represents the movement in vivo.

Scanning electron microscopy (SEM) of intact specimens provides high

resolution, high contrast and high depth of field images while preserving the

structure of the embryo. In 1977, Keller and Schoenwolf carried out a seminal

study of the gastrulating embryo using SEM. Although this study was primarily

  66

qualitative, they gathered data about cell shapes and organizations in tissues as

well as information about the orientations of cellular protrusions. They found that

the dorsal mesoderm could be divided into two regions with different cell

orientations. These regions correspond to the motility zones of the CM, which

was undergoing convergent extension and the anterior (PCM and LEM)

mesoderm, which was undergoing animally directed cell migration (Keller and

Schoenwolf, 1977; Keller and Tibbetts, 1989; Winklbauer and Nagel, 1991).

Combining endogenous cell orientation data with in vitro data from live imaging of

explanted tissue has improved our understanding of CM and LEM movements

but has also raised questions about whether the movements observed in

explants are representative of the movements occurring in the embryo. In this

chapter, using a combination of whole embryo and in vitro analysis I describe the

morphogenesis of the PCM. I quantitatively show that PCM cells are highly

oriented toward the BCR and that over the course of gastrulation, these cells

undergo radial intercalation, migrating toward the BCR using the surfaces of

other cells as a substrate for migration in a process that is BCR dependent.

 

2.2: Results 2.2.1 The dorsal, lateral and ventral mesoderm can be subdivided into distinct cell groups based on gene expression and morphological evidence

  67

Mesoderm is a non-homogenous tissue, which can be divided into distinct groups

of cells undergoing specific morphogenetic movements. In order to investigate

these movements, the dorsal, lateral and ventral mesoderm was divided into

regions according to the expression of known marker genes, visible

morphological boundaries and differences in cell morphology. The dorsal

mesoderm was divided into three regions (Figure 2.1). XBra marked the CM

(Figure 2.1A-C, D-F) (Wilkinson et al., 1990; Smith et al. 1991) and Gsc marked

the PCM (Figure 2.1A’-C’, D-F) (Cho et al., 1991; Niehrs et al. 1994, Niehrs et al.

1993). The third region, the LEM, was defined as the region between the anterior

boundary of Gsc expression and the leading edge of the mesoderm mantle

(Figure 2.1D-F). I used scanning electron microscopy to analyze morphological

characteristics of cells in each of the dorsal mesoderm sub-regions. In order to

determine the relative positions of XBra and Gsc expressing regions and LEM on

SEM micrographs, the average length and width of these regions were measured

from in situ hybridization experiments (Table 2.1). The average length of the LEM

was measured from the non-Gsc expressing region that was anterior to Gsc

expression on in situ images. Average width measurements from in situ images

were used to directly describe the width of the XBra and Gsc expressing regions

on SEM micrographs (Table 2.1; Figure 2.2A). The average width of the LEM

was assumed to be the same as the Gsc expressing region. Average length data

was used to determine the proportion of total mesoderm length occupied by XBra

or Gsc expression (Table 2.1; Figure 2.2A). These proportions determined the

  68

Figure 2.1: The dorsal, lateral and ventral mesoderm are made up of sub-regions (A-C’) In situ hybridization showing expression of Xbrachyury (A-C) and

goosecoid (A’-C’) in mesoderm of early (NF stage 10), middle (NF stage 11) and

late (NF stage 12) gastrula stage embryos; dashed purple and yellow lines show

the boundary of expression of Xbrachyury and goosecoid respectively. (D-F) Low

magnification scanning electron micrographs of early (NF stage 10) (D), middle

(NF stage 11) (E) and late (NF stage 12) (F) gastrula stage embryos; dorsal and

ventral mesoderm regions are described by coloured areas on scanning electron

micrographs; white arrow indicates the dorsal blastopore lip. Crosshairs show

directions of embryonic axes. Ant., Anterior; Pos., Posterior; D, Dorsal; V,

Ventral.

  69

Figure 2.1: The dorsal, lateral and ventral mesoderm are made up of sub-regions

  70

relative positions of the mesoderm regions expressing these genes according to

the overall length of the curved mesoderm for each SEM image (Table 2.1;

Figure 2.2A). The areas of XBra expression, Gsc expression and the LEM

correlated well with areas of distinct cell orientation on scanning electron

micrographs of sagitally fractured embryos (Figure 2.2A). This correlation

suggests that cell orientation can be an indicator of different mesoderm regions

when marker gene expression data is unavailable.

XBra expression is not confined to the dorsal side but extends laterally to the

ventral side of the embryo (Figure 2.1A-C and Murai et al. 2007); therefore XBra

was used as a general marker for the lateral somitic mesoderm (posterior lateral

mesoderm) (Keller, 1976) and the presumptive pronephric mesoderm (ventro-

lateral posterior mesoderm) (Colas et al. 2008). Unlike the dorsal mesoderm,

specific gene expression markers for lateral and ventral anterior mesoderm and

LEM have not been identified. Cardiogenic marker genes such as Nkx 2.5 and

GATA-4, 5 and 6 (Walters et al. 2001) and the prospective blood island markers,

hepatocyte growth factor and C-met (Koibuchi et al. 2003) are expressed during

gastrulation in the anterior dorso-lateral mesoderm and anterior ventro-lateral

mesoderm respectively. However it is difficult to distinguish a boundary between

the expression of these markers and XBra expression. Thus, cell morphology

was used as an indicator to distinguish between anterior and posterior mesoderm

in these lateral and ventral regions (Figure 2.2B). Similar to the dorsal side, on

  71

Figure 2.2: Mapping of dorsal, ventral and lateral mesoderm regions on scanning electron micrographs (A,B) Average length and width dimensions of dorsal XBra (A), ventral/lateral

XBra (B) and dorsal Gsc (A) expression domains were measured from in situ

hybridization images. Average lengths were determined (See Table 2.1), and the

percentage of total mesoderm length occupied by each domain was mapped

onto the SEM pictures. The widths of domains varied little, and average values

were used directly. The domains defined corresponded well to morphological

boundaries between ectoderm and mesoderm (Brachet’s cleft) on one side, and

large endoderm cells or the archenteron epithelium on the other. No

unambiguous markers or morphological boundaries are available to distinguish

between LEM and deep endoderm. Thus, the PCM-endoderm boundary was

extended anteriorly in parallel to the BCR, to describe the width of the LEM. (B)

The boundary between the involuted mesoderm and ventral/lateral LEM regions

is the transition between different cell morphologies (compare white arrowheads

and red arrowheads). Yellow arrowheads highlight endoderm cells. Average

lengths of these regions were determined and the percentage of total

ventral/lateral mesoderm occupied by each region was mapped on to SEMs.

Purple, chordamesoderm (XBra); yellow, pre-chordal mesoderm (Gsc); orange,

leading edge mesoderm; green, involuted lateral/ventral mesoderm; blue,

lateral/ventral LEM. EN, endoderm; BC, blastocoel.

  72

Figure 2.2: Mapping of dorsal, ventral and lateral mesoderm regions on scanning electron micrographs

  73

Table 2.1: The proportion of total mesoderm occupied by individual mesoderm subtypes. Lengths, widths and proportions of total mesoderm length for early, middle and

late gastrula stages. Blue text indicates that measurments were derived from

marker gene expression data. Asterisks and red text indicate that measurements

were derived from morphological evidence from scanning electron micrographs.

All length and width measurements are in micrometres.

 

Table 2.1: The proportion of total mesoderm occupied by individual mesoderm subtypes.

Stage Region of Interest Length

Width

Width (lower Gsc)

Width (BCR XBra)

Avg. proport. of length

Early Gastrula

(NF Stg. 10)

Chorda. (XBra) 281 +/- 15 146 +/- 8 - - 0.29 +/- 0.01 Prechordal (Gsc) 587 +/- 44 139 +/- 5 71 +/- 0.4 - 0.63 +/- 0.02

Dors. LEM 271 +/- 28 - - - 0.29 +/- 0.02 Avg. Mes. Length 945 +/- 39

Ventral (XBra) 377 +/- 9.5 195 +/- 5 - 81 +/- 1 0.56 +/- 0.03 *Vent. Anterior 110 +/- 9.3 101 +/- 1 - - 0.25 +/- 0.02

*Vent. LEM 103 +/- 26 - - - 0.24 +/- 0.05 Avg. Mes. Length 562 +/- 45

Mid Gastrula (NF Stg. 11)

Chorda. (XBra) 345 +/- 22 142 +/- 5 - - 0.26 +/- 0.02 Prechordal (Gsc) 520 +/- 32 106 +/- 5 - - 0.46 +/- 0.03

Dors. LEM 409 +/- 64 - - - 0.36 +/- 0.05 Avg. Mes. Length 1175 +/- 50

Ventral (XBra) 419 +/- 19 185 +/- 6 - - 0.41 +/- 0.03 *Vent. Anterior 235 +/- 10 87 +/- 2 - - 0.33 +/- 0.01

*Vent. LEM 177 +/- 11 - - - 0.26 +/- 0.01 Avg. Mes. Length 1016 +/- 43

*Lateral Anterior 304 +/- 23 66 +/- 2 - - 0.40 +/- 0.02 Avg. Mes. Length 752.6 +/- 45

Late Gastrula (NF Stg. 12)

Chorda. (XBra) 564 +/- 47 131 +/- 5 - - 0.33 +/- 0.03 Prechordal (Gsc) 507 +/- 28 50 +/- 9 - - 0.27 +/- 0.01

Dors. LEM 644 +/- 36 - - - 0.33 +/- 0.02 Avg. Mes. Length 1782 +/- 68

 

the lateral and ventral sides of the embryo, cell sizes vary between the LEM and

more posteriorly located mesoderm, with LEM cells being obviously larger

(Figure 2.2B). The transition from smaller to larger cell sizes was used to divide

the LEM from the anterior mesoderm (Figure 2.2B). Thus the lateral/ventral

anterior mesoderm was defined as the region between the anterior extent of

XBra expression and the posterior side of the cell size transition (Figure 2.1E,F).

The LEM was therefore the region between the anterior side of the cell size

transition and the leading edge of the mesoderm mantle (Figure 2.1E,F).

The same method used to determine the proportion of the dorsal mesoderm

length occupied by each mesoderm region was used to identify the regions of the

lateral and ventral mesoderm on SEM images. The average proportion of

lateral/ventral mesoderm occupied by XBra expressing posterior mesoderm was

determined from in situ hybridization images for XBra (Table 2.1; Figure 2.2B).

However, since specific gene expression markers for the lateral/ventral anterior

mesoderm were not available, the proportion of lateral/ventral mesoderm

occupied by anterior mesoderm was measured directly from SEM images (Table

2.1; Figure 2.2B). The transition between LEM and anterior mesoderm described

above and the anterior extent of XBra expression were used to mark the anterior

and posterior boundaries of the anterior mesoderm respectively. Furthermore, it

was assumed that the anterior mesoderm was bound by the mesoderm/ectoderm

boundary (Brachet’s cleft) on one side and large endoderm cells on the other

  76

side (Figure 2.2B) and average width measurements were taken directly from

SEM images (Table 2.1). The lateral and ventral LEM made up the remaining

proportion of total mesoderm length (Table 2.1; Figure 2.2B). As with the dorsal

LEM, the width of this region was taken to be the same as that of anterior

mesoderm (PCM dorsally) since there are no known markers that specifically

label LEM, although this may be an underestimation of the width of the region.

Thus, during gastrulation, based on gene expression and morphological

differences, the dorsal, ventral and lateral (left/right) mesoderm can be divided

into XBra expressing posterior mesoderm, anterior mesoderm (Gsc expressing

dorsally) and LEM, which could each undergo specific morphogenetic

movements.

2.2.2: During gastrulation, PCM cells are unipolar and produce lamelliform protrusions In their 1977 paper, Keller and Schoenwolf carried out a detailed scanning

electron microscopy analysis of the dorsal mesoderm in which they qualitatively

described morphological details of non-involuted and involuted cells over the

course of gastrulation. Here, I expand on their original analysis through

quantification of cell morphology on a sub-region basis.

During the mid-gastrula stage, deep cells of the Gsc expressing PCM were

elongated perpendicular to the BCR, were unipolar, with a protrusion-bearing

  77

front end pointing towards the BCR, and a rounded rear in sagittal and transverse

planes of fracture (Figure 2.3A-G). Protrusions were lamelliform or filiform, and

appeared attached to neighbouring cells (Figure 2.3A-F). In particular, filiform

protrusions were extended between cells as previously described for involuted

mesoderm (Figure 2.3F) (Keller and Schoenwolf, 1977). In some instances, cells

could be seen extending multiple protrusions into spaces around surrounding

cells (Figure 2.3D). The sites of protrusion attachment suggest that deep PCM

cells use neighbouring cells as a migration substrate. The preferential orientation

of the long axis of these cells may be due to regulation by an orientating cue.

Alternatively, the orientation of these cells could be passively determined by the

movements of the surrounding regions, such as the endoderm.

The long axis of PCM cells at the mid-gastrula stage is on average 2 times longer

than the short axis. Visually, it appears that PCM cells are preferentially

elongated in the dorsal/ventral axis (Figures 2.3, 2.4, 2.5) and therefore, the

direction of the axis elongation of these cells may be indicative of a cell

movement direction. Since it is unknown if there are changes in PCM mesoderm

cell shape over the course of gastrulation, I analyzed the length/width ratios of

these cells. Length/width ratios have been used as an indicator of cell polarity in

explanted tissue (Wallingford et al. 2000). However, these measurements are

typically done in two dimensions and therefore only provide information about cell

shape in the plane of the image being analyzed. Measurements done in this way

  78

Figure 2.3: PCM cells extend lamelliform protrusions toward the BCR in sagittal and transverse planes (A-F) SEM micrographs of PCM cells from sagittaly fractured (A,C,E) and

transversely fractured (B,D,F) mid-gastrula stage (NF stage 11) embryos,

extending lamelliform protrusions both in (white arrowheads) and out of the plane

of the image (white arrows). Cells can extend multiple protrusions from a single

pole (D, asterisk) in the same or different planes (E, F, asterisk and white

arrowhead, same cell in E and F). (G) Schematic diagram showing the

orientation of fracture planes. (H) Percentage of cells extending protrusions in or

out of the plane of fracture or lacking protrusions measured from transversely

fractured and sagittaly fractured embryos. BCR, blastocoel roof; BC, blastocoel;

AC, archenteron.

  79

Figure 2.3: PCM cells extend lamelliform protrusions toward the BCR in sagittal and transverse planes

  80

are not necessarily reflective of actual cell shape or polarity. I measured the

length/width ratios of the PCM from SEM micrographs of sagittaly and

transversely fractured embryos (Figure 2.3G) before and during gastrulation and

integrated the data from both planes to get a picture of mesoderm cell shape and

orientation in the embryo.

Prior to the start of gastrulation (NF stage 9), cells in the dorsal marginal zone

(Figure 2.4A, red box), which gives rise to the three dorsal mesoderm regions at

later stages of gastrulation, had an average length/width ratio of 1.4 +/- 0.05 in

both the sagittal and transverse fracture planes (Figure 2.4A, 2.5A). However, by

the mid-gastrula stage (NF stage 11), the cell elongation increased significantly

(P-value < 0.0001). Average length/width ratios were 1.9 +/- 0.03 and 1.8 +/- 0.06

in the sagittal and transverse fracture planes respectively (Figure 2.4B, 2.5B).

Interestingly, the adjacent pharyngeal endoderm cells also became visibly

elongated by mid-gastrulation (compare Figure 2.4A and B). There was no

change in cell elongation between mid and late gastrula stages in either plane of

fracture (Figure 2.4B-D and Figure 2.5B-D). This indicates that cells are

elongated along the same axis in both fracture planes. Cells often become

elongated along an axis parallel to their direction of movement (Ridley, 2011),

thus I took the orientation of this long axis to be indicative of the orientation and

the direction of movement of the cell. For quantification purposes it was sufficient

to derive cell orientation from the sagittal plane alone since the length/width ratios

  81

Figure 2.4: PCM cells become significantly elongated in the sagittal plane during gastrulation (A-C) Scanning electron micrographs of stage 9 (A), stage 11 (B) and stage 12

(C) sagittaly fractured embryos; red box in (A) shows the area of the DMZ where

measurements were taken; yellow lines show the length and width axes of PCM

cells along which length/width ratio measurements were taken. (D) Bar graph

comparing the average length/width ratios of PCM cells before and during

gastrulation, asterisk indicates a statistically significant result, P value < 0.0001.

Error bars show standard error of the mean. 9 (84 cells) stage 9, 10 (101 cells)

stage 11 and 7 (71 cells) stage 12 sagitally fractured embryos were used to

generate data. EN, endoderm; EC, ectoderm; BC, blastocoel; DMZ, dorsal

marginal zone; BCR, blastocoel roof.

  82

Figure 2.4: PCM cells become significantly elongated in the sagittal plane during gastrulation

  83

were the same in both planes of fracture. I also determined the positions of PCM

cell protrusions relative to anterior/posterior or lateral neighbours at the mid-

gastrula stage (Figure 2.3H). Cells extended and attached lamelliform protrusions

to neighbouring cells in both the sagittal and transverse fracture planes.

Approximately 50% of cells in both fracture planes extended and attached

protrusions within the plane of fracture (Figure 2.3H). Some cells extended

protrusions in the opposite plane from the one being imaged however, the

majority of remaining cells did not have visible protrusions (Figure 2.3H). These

cells are likely extending protrusions in the plane opposite to the one being

imaged, however these protrusions may be hidden. Some cells extended multiple

lamelliform protrusions from a single pole of the cell in two planes simultaneously

(Figure 2.3E,F). Altogether, PCM cells extend protrusions between neighbouring

cells in any antero-posterior or mediolateral plane.

2.2.3: LEM, PCM and involuted chordamesoderm cells are oriented toward the BCR

The orientation of cells can provide information about the direction of cell

movements in the embryo. At the start of gastrulation (stage 10), cells were

elongated but did not always extend lamelliform protrusions (Figure 2.6). Thus, to

quantify cell orientation, the angle of the cell long axis with respect to the

dorsal/ventral axis (lower PCM) or the BCR (upper PCM and LEM) was

measured from SEM images of sagittaly fractured embryos (Figure 2.3G).

  84

The long axes of deep and more superficial (closer to the embryo surface) LEM

cells in the sagittal plane were predominantly oriented toward the BCR at angles

between 30o and 100o (avg. 71o +/- 3.2) with respect to the BCR (Figure 2.6A,E).

This highly oblique orientation is consistent with the reports of LEM spreading on

the inner BCR surface in the direction of the animal pole and is reflected in a

“shingle” like arrangement when viewed from the surface that is in contact with

the BCR (Figure 2.8A; Winklbauer and Nagel 1991; Davidson et al. 2002; Nagel

et al. 2004).

At the onset of gastrulation, goosecoid expression extends from just above the

embryo equator down to the bottle cell region of the blastopore lip (Figure 2.1A’).

Thus, the PCM can be subdivided into two regions, one that is adjacent to the

BCR (upper PCM) and one that is adjacent to the non-involuted CM (lower PCM),

where a separation boundary between Gsc and XBra expressing cells is

maintained (Figure 2.6G). The cell long axes of upper PCM cells were oriented at

angles between 0o and 100o (avg. 41o +/- 4.7) (Figure 2.6B,F), which is a broader

distribution of cell orientations when compared with the LEM (compare Figure

2.6A and B). At this stage of gastrulation, the lower PCM cells have not been

internalized to the level of the BCR. Thus, it is not possible to relate the

  85

Figure 2.5: PCM cells become significantly elongated in the transverse plane during gastrulation (A-C) Scanning electron micrographs of stage 9 (A), stage 11 (B) and stage 12

(C) transversely fractured embryos; yellow lines show the length and width axes

of PCM cells along which length/width ratio measurements were taken. (D) Bar

graph comparing the average length/width ratios of PCM cells before and during

gastrulation, asterisk indicates a statistically significant result, P value < 0.0001.

Error bars show standard error of the mean. 7 (63 cells) stage 9, 10 (97 cells)

stage 11 and 6 (60 cells) stage 12 transversely fractured embryos were used to

generate the data. EN, endoderm; DMZ, dorsal marginal zone; BCR, blastocoel

roof.

  86

Figure 2.5: PCM cells become significantly elongated in the transverse plane during gastrulation

  87

orientation of these cells to the BCR; instead cell orientation was measured with

respect to the dorsal/ventral axis of the embryo. The orientation of lower PCM

cells was significantly different from cells of the upper PCM (P-value = 0.0007).

Upper PCM cells were oriented at angles between 40o and 90o (avg. 61o +/- 3.7)

with respect to the dorsal/ventral axis of the embryo (Figure 2.6C,G). Thus, the

BCR may produce a signal that can regulate the orientation of cells. The upper

PCM, which is adjacent to the BCR, may be influenced by such a signal. This

could explain the difference in orientation between the upper and lower PCM

regions.

When examined from sagitally fractured embryos, the chordamesoderm was

made up of a mixture of elongated and non-elongated cells. The long axes of the

elongated cells were oriented parallel to the animal/vegetal axis of the embryo,

predominantly at angles between 700 and 900 (avg. 80o +/- 3.3) with respect to

the BCR (Figure 2.6D,G). The orientation of the non-elongated cells could not be

determined from sagitally fractured embryos. These cells may be oriented into

the plane of the image and thus appear as non-elongated when viewed on end in

sagittal fractures (Figure 2.6G). Thus, non-involuted chordamesoderm cells

appear to be oriented in the animal/vegetal axis and the medial/lateral axes of the

embryo.

  88

At the mid-gastrula stage (stage 11), LEM and PCM cells extend lamelliform

protrusions from the leading edge of the cell. To quantify cell orientation, I

measured the angles at which these protrusions were extended from the cell

body with respect to the BCR.

Cells of the dorsal LEM are heterogeneous in size. Individual cells appeared

either mesoderm like or endoderm like in size and morphology (Figure 2.7E).

Endoderm like cells tended to be larger and produced visibly smaller protrusions

with respect to the size of the cell body compared with mesoderm cells (Figure

2.7E). This mixture of morphologies suggests that this region is a mixture of

mesoderm and endoderm cells. LEM cells exhibited a distinct polarity with a

single protrusion-bearing pole. Like the PCM, LEM cells attached their

lamelliform protrusions to neighbouring cells (Figure 2.7E). The orientation of

LEM remained oblique with respect to the BCR, with most cells extending

protrusions at angles between 0 and 70 degrees (avg. 36o +/- 3.9) (Figure

2.7A,E).

Based on the location of Gsc expression, all PCM cells had been internalized

such that they were adjacent to the BCR by the mid-gastrula stage (Figure 2.1B’).

PCM cells were highly oriented toward the BCR (avg. 9o +/- 1.6 with respect to

the BCR) (Figure 2.7B,F). This orientation is perpendicular to the BCR and to the

animally directed movement of mesoderm. In contrast, the long axes of the large

  89

Figure 2.6: The long axes of mesoderm cells are oriented parallel to the BCR at the start of gastrulation (A-D) Rose diagrams showing the percentage orientation of LEM (A) and upper

involuted PCM (B) cells with respect to the BCR and lower PCM (C) and

chordamesoderm (D) cells with respect to the dorsal/ventral axis. 0 degrees,

dorsal BCR or dorsal/ventral axis; 90 degrees, animal pole. (E-G) Scanning

electron micrographs of LEM (E), upper PCM (F) and lower PCM with

chordamesoderm (G), white arrowheads highlight cells with long axis orientations

that are parallel to the BCR (E,F) or dorsal/ventral axis (G), red arrowheads

highlight rounded cells that may be pointing into the plane of the image (G).

Orange, leading edge mesendoderm; yellow, prechordal mesoderm; purple,

chordamesoderm. LEM, leading edge mesendoderm; PCM, prechordal

mesoderm; Chorda, chordamesoderm; BC, blastocoel; BCR, blastocoel roof; EN,

endoderm. “n” is the number of cells scored per region from 10 embryos

analyzed. Crosshairs show directions of embryonic axes; An, animal pole; Ve,

vegtal pole; D, dorsal; V, ventral.

  90

Figure 2.6: The long axes of mesoderm cells are oriented parallel to the BCR at the start of gastrulation

  91

endodermal cells located deep to the PCM were oriented in parallel to the BCR

(avg. 78o +/- 1.7) (Figure 2.7C,F), similar to the orientation of lower PCM cells at

the start of gastrulation. Thus, apparently the lower PCM cells become oriented

toward the BCR as they are internalized and become positioned adjacent to the

BCR. These results support the idea that the BCR has an orienting influence on

PCM cells.

At the mid-gastrula stage the XBra expressing CM can be divided into involuted

and non-involuted sub-regions (Figure 2.1B, 2.6G). The non-involuted region is

located in the lower blastopore lip and involuted cells are located adjacent to the

BCR, posterior to the PCM (Figure 2.7G). The orientation of non-involuted CM

cells was not determined because few cells appeared elongated and many did

not have discernable protrusions. Like at the start of gastrulation, non-involuted

CM cells may have been oriented with the animal/vegetal and medial/lateral axes

of the embryo. Involuted CM often lacked clearly defined protrusions; therefore

the orientation of the cell long axis was measured. Like in the PCM, cells were

oriented perpendicular to the BCR (avg. 5o +/- 2.5 with respect to the BCR)

(Figure 2.7D,G). CM cells appear to behave in a similar way to PCM cells after

being internalized. As cells move into the embryo and become adjacent the BCR,

cells become oriented perpendicular to it. As suggested for the PCM,

chordamesoderm may also be under the influence of an orienting cue from the

BCR.

  92

It has been shown that LEM cells adjacent to the BCR substrate extend

lamelliform protrusions preferentially toward the animal pole (Figure 2.8A and

Winklbauer and Nagel 1991; Nagel et al. 2004). In order to determine if the same

organization occurs in the PCM, the BCR was fractured away and the substrate

adjacent side of the mesoderm was imaged. Numerous protrusions were visible

but there was no obvious orientation and cells had a rounded morphology (Figure

2.8B). Furthermore, the protrusions appeared to be flat against neighbouring

mesoderm cells rather than pointing away from the surface of the embryo as if

they had been crossing Brachet’s cleft to form contacts with the BCR (Figure

2.8B). Thus, it appeared as though cells make contact with adjacent mesoderm

cells rather than the BCR. This observation suggests that unlike LEM cells, PCM

cells at the mesoderm/ectoderm interface may not use the BCR inner surface as

a migration substrate and thus, may use a different mechanism to spread in the

direction of the animal pole.

At the mid-gasrula stage, the XBra expressing blastopore lip is the site of

involution. Like PCM cells, non-involuted and involuted CM cells can extend

lamelliform and filiform protrusions (Figure 2.7G, 2.8C). When the dorsal

blastopore lip was fractured parallel to the embryo surface, medio-laterally

oriented bipolar cells extending protrusions from opposite poles of the cell were

visible (Figure 2.8C), which may be consistent with medio-lateral cell intercalation

(Keller and Tibbetts, 1989; Shih and Keller, 1992).

  93

Figure 2.7: Internalized PCM and chordamesoderm cells are oriented toward the BCR (A-D) Rose diagrams showing the percentage orientation of LEM (A) PCM (B)

endoderm (C) and involuted chordamesoderm (D) cells with respect to the BCR.

0 degrees, dorsal BCR; 90 degrees, animal pole. Blue rose diagrams use the

angle at which protrusions are extended from the cell body with respect to the

BCR as a measure of orientation. Red rose diagrams use the orientation of the

cell long axis with respect to the BCR. (E-G) Scanning electron micrographs of

LEM (E), PCM and endoderm (F) and chordamesoderm (G), red arrowheads in

(E) highlight cells that are orientated parallel to the BCR; red arrowheads in (G)

indicate random orientation of cells; White arrowheads highlight cells oriented

perpendicular to the BCR (F,G). Red line in (G) is the boundary between non-

involuted and involuted chordamesoderm. LEM, leading edge mesendoderm;

PCM, prechordal mesoderm; Chorda, chordamesoderm; AC, archenteron; BC,

blastocoel; BCR, blastocoel roof; EN, endoderm. Orange, leading edge

mesendoderm; yellow, prechordal mesoderm; purple, chordamesoderm; blue,

endoderm. “n” is the number of cells scored per region from 10 embryos

analyzed.

  94

Figure 2.7: Internalized PCM and chordamesoderm cells are oriented toward the BCR

  95

Convergent extension movements are thought to start just prior to the mid-

gastrula stage (Keller et al. 1985; Keller and Danilchik, 1988). It is therefore

unclear how non-involuted medio-laterally oriented superficial cells cooperate

with the deeper non-involuted CM cells and radially oriented involuted CM cells to

narrow and elongate the anterior/posterior axis of the embryo. Internalized CM

cells were oriented like PCM cells, with a long axis orientation perpendicular to

the BCR (Figure 2.7D,G). When viewed from the side previously in contact with

the BCR inner layer, cells resembled the more anterior PCM cells and a strong

medio-lateral orientation was not visibly obvious (Figure 2.8D). The contribution

of involuted chordamesoderm cells to medial-lateral intercalation based

convergent extension is unknown and difficult to visualize given the long axis

orientation of the cells.

There was no change in orientation of LEM or PCM cells by late gastrulation

(compare Figures 2.7A-B, E-F; 2.9A-B, F-G), although the PCM had apparently

thinned to a single cell layer that was in contact with the BCR (Figure 2.9A-B, F-

G). Post-involution chordamesoderm cells were still oriented perpendicular to the

BCR, but cells had become strongly protrusive (Figure 2.9C,D). Unlike the PCM

however, where protrusions are always oriented toward the BCR, I observed that

59% of CM cells extended their lamelliform protrusions toward the BCR while

16% extended lamelliform protrusions toward the archenteron (Figure 2.9C, E).

  96

Figure 2.8: LEM, PCM and CM exhibit distinct cell morphologies when viewed from the BCR apposed surface (A-C) Scanning electron micrographs of the surface of the mesoderm that is

normally apposed to the inner layer of the BCR. Mid-gastrula leading edge

mesendoderm (LEM) (A), white arrowheads point at oriented cells. Mid-gastrula

prechordal mesoderm (PCM) (B), white arrowheads point to cells with rounded

morphology. Pre-involution chordamesoderm (C), white arrowheads point at

protrusions on opposite poles of bipolar cells. Post-involution chordamesoderm

located anterior to the pre-involution chordamesoderm, white arrowheads point at

protrusions on opposite poles of bipolar cells, red arrow heads highlight cells with

rounded morphology, similar to PCM cells (D); the crosshairs show the directions

of embryonic axes. Lat., lateral; Chorda, chordamesoderm.

  97

Figure 2.8: LEM, PCM and CM exhibit distinct cell morphologies when viewed from the BCR apposed surface

  98

PCM cells were never found to extend protrusions toward the archenteron,

therefore different signals may regulate cell body orientation and the direction of

protrusion extension. Thus, chordamesoderm cells may not respond to the same

protrusion orienting cue as the PCM. Interestingly, the long axis of non-involuted

CM cells became oriented radially by late gastrulation (Figure 2.9D, H).

2.2.4: PCM cells undergo radial intercalation as they migrate toward the mesoderm/ectoderm boundary

The PCM dramatically thinned from what was a multi-layered tissue during mid-

gastrulation to a single layer of cells by the end of gastrulation (Figure 2.9G).

Comparison of PCM thickness over the course of gastrulation showed a

significant change in region thickness resulting in an average thinning of 89 µm

over 4.5 hours (P-value < 0.0001) (Figure 2.10A,D). In contrast, there was no

change in the thickness of post-involution chordamesoderm from mid-gastrulation

to late gastrulation (Figure 2.10A).

To determine whether the change in PCM thickness is due to rearrangement or

flattening of cells, we measured the layer index, i.e. tissue thickness in terms of

cell layers (Keller, 1980; Figure 2.10B). The average layer index changed

significantly over the course of gastrulation. It was reduced from 3.1 +/- 0.09

(upper PCM) at the start of gastrulation to 1.5 +/- 0.08 by the late gastrula stage

  99

Figure 2.9: PCM thins to a single cell layer by the late gastrula stage (A-D) Rose diagrams showing the percentage orientation of LEM (A) PCM (B)

chordamesoderm (C) and non-involuted chordamesoderm (D). 0 degrees, dorsal

BCR/outer embryonic epithlium; 90 degrees, animal pole. Blue rose diagrams

use the angle at which protrusions are extended from the cell body with respect

to the BCR as a measure of orientation. Red rose diagrams use the orientation of

the cell long axis with respect to the outer embryonic epithelium. (E) High

magnification scanning electron micrograph of bi-directionally oriented involuted

chordamesoderm cells. White arrowhead, cell oriented toward BCR; red

arrowhead, cell oriented toward the archenteron epithelium (F-H) Scanning

electron micrographs of LEM (F), PCM (G) and chordamesoderm (H), white

arrowheads highlight cells that are orientated perpendicular to the BCR (F-H).

Red arrowheads highlight cells oriented parallel to the BCR (F). LEM, leading

edge mesendoderm; PCM, prechordal mesoderm; Chorda, chordamesoderm;

AC, archenteron; BC, blastocoel; BCR, blastocoel roof; EN, endoderm. Orange,

leading edge mesendoderm; yellow, prechordal mesoderm; purple,

chordamesoderm. “n” is the number of cells scored per region from 10 embryos

analyzed.

  100

Figure 2.9: PCM thins to a single cell layer by the late gastrula stage

  101

(P-value < 0.001) (Figure 2.10C, D-G). There was no change in the layer index of

the post-involution CM (Figure 2.10C, H-I). Occasionally, cells undergo cell

division during gastrulation. If the plane of cell division were parallel to the long

axis of the cell, the resultant daughter cells could increase the number of cell

rows between the BCR and the endoderm, increasing the layer index.

Regardless of cell divisions, the PCM thins to a single cell layer by the end of

gastrulation (Figure 2.10C, D-G) suggesting that all PCM cells, including cells

that appear as a result of cell divisions, undergo radial intercalation. Furthermore,

It is unlikely that cell divisions are responsible for keeping the post-involution CM

multilayered over gastrulation because it has been shown that CM cells are

arrested in the G2 phase during gastrulation (Cooke, 1979). Thus, together with

the presence of oriented protrusions, these results suggest that the PCM

undergoes radial intercalation. Significant insertion of deeper cells into the

superficial layer of the PCM has indeed been observed (Winklbauer and

Schurfeld, 1999). The CM also likely undergoes radial intercalation during the

formation of the notochord (Keller et al. 1989), however these movements may

not be prominent until after the start of neurulation, when convergent extension

occurs most strongly (Wilson and Keller, 1991; Keller et al. 1985).

2.2.5: Intercellular migration drives PCM radial intercalation

PCM cells attach their BCR oriented protrusions to surrounding PCM cells,

suggesting that cells use each other as a substrate for migration. The PCM thins

  102

significantly from an average thickness of 139 +/- 5.1µm during early gastrulation

(NF stage 10+) to 106 +/- 4.6µm during mid-gastrulation (NF stage 11) (P-value =

0.0035) corresponding to an average rate of thinning of 0.3 µm/min (Figure

2.9A,D). The rate of thinning increases at mid-gastrulation, significantly thinning

the PCM to 50 +/- 8.8 µm (P-value < 0.0001) by the late gastrula stage (NF stage

12) at an average rate of 0.6 µm/min (Figure 2.10A,D). Thus, the overall average

rate of thinning of the PCM is 0.45 µm/min (Figure 2.10D). While cell shape

changes such as the flattening of cells can result in thinning of a tissue, it is likely

that it contributes little to the thinning of this region considering that a single,

superficial layer of cells is flattened against the BCR throughout gastrulation

(Figure 2.10D,F). Thus, the average rate of PCM thinning corresponds to the

average radial intercalation rate. Since radial intercalation is occurring by

intercellular migration, the rate of radial intercalation corresponds to the relative

rate of intercellular movement in the PCM.

The opacity of the Xenopus embryo prevents direct observation of cell

movements, therefore in order to analyze live PCM morphogenesis, dorsal

mesoderm and endoderm, including the PCM, was explanted and combined with

BCR (Figure 2.11A). These slice explants preserve the endogenous tissue

arrangement. Based on the SEM study, I expected that PCM cell movement

would be directed towards the BCR. This was indeed observed (Figure 2.11B-C).

Although only cells at the explant surface can be filmed, which may move more

  103

2.10: PCM undergoes radially directed cell rearrangements

(A) Bar graph comparison of changes in Chordamesoderm (XBra) and PCM

region thickness over gastrulation. The PCM thins significantly between stages

10 and 11 (p-value < 0.0001, n = 10) and again between stages 11 and 12 (p-

value < 0.0001, n = 10 embryos/stage) (B) the layer index is a measure of cell re-

arrangement and is calculated as the shortest number of transitions (indicated by

numbers) from cell to cell between the BCR and the endoderm (direction

indicated by red arrows). (C) Comparison of layer index changes in

chordamesoderm (XBra) and PCM over gastrulation. PCM layer index changes

significantly from stage 10 to 11 (p-value = 0.0018, n = 10) and again from stage

11 to 12 (p-value < 0.0001, n = 10) (D) Change in region thickness and layer

index over the 4.25 hour period of gastrulation analyzed. (E-I) High magnification

scanning electron micrographs of PCM (D-F) at stages 10 (D), 11 (E) and 12 (F)

with the span of mesoderm regions indicated with the layer index. Error bars

represent standard error of the mean, asterisks indicated statistically significant

results. PCM, prechordal mesoderm; XBra, chordamesoderm; BCR, blastocoel

roof; AC, archenteron; EN/Endo, endoderm.

  104

Figure 2.10: PCM undergoes radially directed cell rearrangements

  105

easily between deeper cells and the coverslip than between lateral cells in this

artificial situation, instances of radial intercalation were nevertheless observed

(Figure 2.1B-B’), consistent with the active directional migration of individual cells

on each other’s surface.

To further analyze the directional intercellular migration of PCM cells, I developed

a different explant system (Figure 2.12A). The former posterior side of the PCM

region of a mesoderm explant was brought into contact with explanted BCR. This

array recapitulates the in vivo apposition of mesoderm and ectoderm while re-

directing cell movements from radially to posteriorly. After 30 minutes, when

explants had adjusted in response to the mechanical effects of the assembly of

the explants, cell movements were followed.

Similar to in vivo, PCM cells extended large lamelliform protrusions toward the

BCR explant (see Figure 2.12A,B). Cells were observed to exchange neighbors

while they moved towards the BCR, again reinforcing evidence of active

directional migration of individual cells on the cells deeper in the explant (Figure

2.12B-E). Occasionally, cells from deep in the explant appeared at the surface

where they joined the cohort of cells migrating toward the ectoderm explant

(Figure 2.12B-E).

It is unknown if PCM cells in contact with the BCR in vivo (i.e. the superficial layer

  106

Figure 2.11: PCM radial intercalation in vitro

(A) Explant system. (1.) Slice of mesoderm and adjacent endoderm is combined

with (2.) explanted BCR under a coverslip and (3.) filmed to track mesoderm cell

movements. Blue, endoderm; green, ectoderm; orange, mesoderm; red arrow

show direction of cell intercalation; A, anterior; P, posterior; D, dorsal; V, ventral.

(B-B’) Frames from timelapse recording. (B) Cell cluster at 0 min. (B’) Same

cluster after 30min, cells have intercalated at the mesoderm/BCR boundary

(white dotted line). (C) Cell movement tracks from explants during intercalation

movements.

  107

Figure 2.11: PCM radial intercalation in vitro

  108

of PCM cells) use it as a substrate for migration like LEM cells (Winklbauer and

Nagel, 1991; Nagel et al. 2004). In vitro, if PCM cells were to migrate over the

BCR explant, it may appear as though they are migrating toward it. To determine

if mesoderm/BCR separation is maintained in explants, BCR-mesoderm explants

were fractured and analyzed by SEM. Although the BCR-mesoderm boundary

was not perfectly straight in fractures, PCM cells maintained a separation

boundary between the BCR and did not migrate over it (Figure 2.13C). Thus, the

observed movement of PCM cells is toward the BCR explant and not across its

surface.

Cell velocity was measured during consecutive 30-minute intervals (Figure

2.12F-H). Directional movement of mesoderm cells within a five-cell diameter

zone (zone R1) adjacent to the BCR was apparent 30 minutes after assembly of

the explants (Figure 2.12F). Directional movement continued over the next hour

(Figure 2.12F), at an average velocity of 0.8 +/- 0.03 µm/min, which is in the

same range as the rate of intercellular migration in vivo (0.45 µm/min +/- 0.05).

Cells located further from the BCR (zones R2 and R3; Figure 2.12G,H) were

motile but never moved directionally (Figure 2.12G,H). This observation suggests

that the influence of the BCR has a finite distance of approximately five cell

lengths into the mesoderm explant, which corresponds to a distance of

approximately 200 µm (PCM cell length approx. 40 µm; Selchow and Winklbauer,

1997). Directional migration of cells was dependent on the BCR. Explants of

  109

mesoderm alone did not show any directional movement and cells moved

randomly (Figure 2.13D). Furthermore, the directional movement response to the

BCR appeared to be unique to mesoderm cells. When explants of endoderm, in

place of mesoderm, were combined with BCR explants, the endoderm cells

moved randomly over the 90 minute period (Figure 2.13E). These results strongly

suggest that a directional determinant, that mesoderm cells are sensitive to, is

released by the BCR. This determinant may form a chemotactic gradient in the

mesoderm tissue that guides cells to the mesoderm/BCR boundary where radial

intercalation occurs.

2.2.6: Anterior lateral and ventral mesoderm cells are oriented toward the BCR I extended the cell orientation analysis to the lateral and ventral mesoderm at the

mid-gastrula stage. Lateral and ventral LEM, anterior mesoderm and non-

involuted XBra expressing cells were similar in shape to their dorsal counterparts

in sagittally fractured embryos (Figure 2.14; 2.15). Like dorsal LEM and PCM,

ventral and lateral LEM and anterior mesoderm cells extended lamelliform

protrusions (Figure 2.14D-F; 2.15D-F). Smaller protrusions were present on XBra

expressing cells (Figure 2.14F; 2.15F). The orientation of lateral and ventral LEM

cells was similar to dorsal LEM. Cells were oriented highly obliquely and

extended protrusions at angles between 0o and 110o (avg. ventral: 63o +/- 4.6,

  110

Figure 2.12: PCM cells migrate directionally in vitro

(A) Explant system. (1., 2.) Mesoderm and endoderm or ectoderm are explanted,

combined and left to relax for 15 minutes. BCR or endoderm is placed at the

posterior, PCM-containing end of mesoderm explant, to force cells to reorient by

90 degrees (from radially to posteriorly). Explant combinations (3.) are filmed,

and cell velocities measured. R1-R3, regions each five cell diameters wide. Blue,

endoderm; green, ecotoderm; orange, mesoderm; red arrow, redirected cell

movement; A, anterior; P, posterior; D, dorsal; V, ventral. (B-E) Frames from

timelapse recording of a BCR-mesoderm explant. (B) Selected cells at 0 min; (C)

at 30 min, cells close to the BCR (green, light blue, purple and red) are moving

toward the BCR; position where a cell will emerge from below is marked by

turquoise arrowhead; (D) at 60 min, the light blue cell has made contact with the

BCR, the purple cell has divided, the turquoise cell has emerged; (E) at 90 min,

the blue and green cells contact the BCR, the green cell has divided. The dark

blue cell has remained at a constant distance from the BCR. (F-H) Velocity plots

showing average velocities of 50 individual cells collected from five separate

BCR-mesoderm explants. Positive y-axis is toward the BCR. (F) Cells in R1

migrate directionally. (G,H) Cells in R2 and R3 show random migration.

  111

Figure 2.12: PCM cells migrate directionally in vitro

  112

avg. lateral: 51o +/- 3.8) with respect to the BCR (Figure 2.14A,D; 2.15A,D). The

oblique orientation of LEM cells all around the embryo could be due to

mechanical influence by the free margin ahead of the LEM. Alternatively, LEM

cells may not receive the same concentration of orienting cues from the BCR or

may not be able to respond as strongly to these signals, potentially resulting in

the oblique orientation of these cells.

The orientation of anterior mesoderm on the ventral and lateral sides of the

embryo was strikingly similar to the PCM on the dorsal side. Cells were strongly

oriented toward the BCR, predominantly extending protrusions at angles between

0o and 20o (avg. ventral: 11o +/- 1.5, avg. lateral: 11o +/- 1.4) with respect to the

BCR (Figure 2.14B,E; 2.15B,E). Thus, involuted mesoderm all around the

embryo may undergo radial intercalation as the cells migrate toward the

mesoderm/ectoderm boundary.

The long axes of cells in the ventral and lateral XBra expressing regions (non-

involuted at mid-gastrulation) were oriented perpendicular to the animal/vegetal

axis of the embryo when viewed in sagittaly fractured embryos (avg. ventral: 34o

+/- 4, avg. lateral: 20o +/- 2) (Figure 2.14C,F; 2.15C,F). The orientation of these

cells is different from the non-involuted XBra cells on the dorsal side of the

embryo, which appeared largely random at the mid-gastrula stage (Figure 2.7G).

  113

Figure 2.13: Directional migration of PCM cells in vitro is BCR dependent

(A,B) Confocal microscopy images of a mesoderm explant expressing a

membrane bound RFP construct, white arrowheads show BCR oriented

protrusions on the surface of the explant (A), protrusions are not visible beneath

the surface, white arrowheads (B). (C) Scanning electron micrograph of a

mesoderm-BCR combined explant, red dotted line highlights the mesoderm-

ectoderm boundary; green, ectoderm explant; orange, mesoderm explant. (D,E)

Velocity plots showing average velocities of 50 individual cells collected from 5 of

the following explant combinations: (D) Mesoderm alone, no BCR explant and (E)

endoderm combined with a BCR explant.

  114

Figure 2.13: Directional migration of PCM cells in vitro is BCR dependent

  115

Figure 2.14: Involuted lateral mesoderm and LEM orientation is similar to their dorsal counterparts

(A-C) Rose diagrams showing the percentage orientation of lateral LEM (A),

involuted lateral mesoderm (B) and XBra expressing lateral mesoderm (C). Blue

rose diagrams use the angle at which protrusions are extended from the cell

body with respect to the BCR as a measure of orientation. Red rose diagrams

use the orientation of the cell long axis with outer embryonic epithelium. (D-F)

Scanning electron micrographs of LEM (D), lateral involuted mesoderm (E) and

lateral XBra expressing mesoderm (F), red arrowheads highlight cells that are

orientated parallel (D) to the BCR; white arrowheads highlight cells that are

oriented perpendicular to the BCR (E) or outer embryonic epithelium (F) Blue,

lateral leading edge mesendoderm; green, involuted lateral mesoderm; purple,

lateral XBra expressing mesoderm; BCR, blastocoel roof; EN, endoderm; BC,

blatocoel. “n” is the number of cells scored per region from 10 embryos analyzed.

  116

Figure 2.14: Involuted lateral mesoderm and LEM orientation is similar to their dorsal counterparts

  117

Figure 2.15: Involuted ventral mesoderm and LEM orientation is similar to their dorsal counterparts

(A-C) Rose diagrams showing the percentage orientation of ventral LEM (A),

involuted ventral mesoderm (B) and XBra expressing ventral mesoderm (C) at

the mid gastrula stage (NF stage 11). Blue rose diagrams use the angle at which

protrusions are extended from the cell body with respect to the BCR as a

measure of orientation. Red rose diagrams use the orientation of the cell long

axis with outer embryonic epithelium. 0 degrees, lateral BCR/outer embryonic

epithelium; 90 degrees, animal pole. (D-F) Scanning electron micrographs of

LEM (D), ventral involuted mesoderm (E) and ventral XBra expressing mesoderm

(F), red arrowheads highlight cells that are orientated parallel (D) to the BCR;

white arrowheads highlight cells that are oriented perpendicular to the BCR (E) or

outer embryonic epithelium (F) Blue, ventral leading edge mesendoderm; green,

involuted ventral mesoderm; purple, ventral XBra expressing mesoderm; BCR,

blastocoel roof; EN, endoderm; BC, blatocoel. “n” is the number of cells scored

per region from 10 embryos analyzed.

  118

Figure 2.15: Involuted ventral mesoderm and LEM orientation is similar to their dorsal counterparts

  119

This difference may be related to the convergent extension movements on the

dorsal side of the embryo.

 

2.3: Discussion

In this chapter I analyzed cell morphology and orientation in different mesoderm

regions in order to develop a view of the movements in each region. In analyzing

the PCM, I found that these cells, which are oriented perpendicular to the BCR,

undergo radial intercalation as the cells migrate over the surfaces of

neighbouring cells, toward the mesoderm/BCR interface. Furthermore, I showed

that this movement was mesoderm specific and dependent on the BCR.

2.3.1 Complementing in vitro explant studies with intact embryo studies Studies of cell movements during Xenopus gastrulation have made heavy use of

tissue explants. While these experiments have been highly informative, questions

have been raised about how accurately the in vitro observations translate over to

what happens in an intact embryo.

Our understanding of Xenopus CM morphogenesis has been derived from

studies of explanted tissue (Keller et al. 1985; Keller and Danichik, 1988; Keller

and Tibbetts, 1989; Wilson and Keller; 1991). In explants bipolar CM cells are

oriented in a medial to lateral direction and converge and intercalate at the

  120

explant midline, resulting in elongation of the tissue (Keller and Tibbetts, 1989;

Keller et al. 1989). My SEM analysis of CM in the mid-gastrula stage embryo has

identified a layer of non-involuted cells beneath the supra-blastoporal endoderm

that are arranged in a medial-lateral orientation (Figure 2.8C). Prior to the start of

gastrulation, the CM is located in the equatorial region of the embryo. Since the

embryo is spherical, the radius decreases at lines of latitude that are closer to the

blastopore. Thus, as the mass of CM cells rotates vegetally during involution, the

cells must constrict toward the midline in order for the region to move into the

embryo through the blastopore. Therefore, it is not clear if the oriented cells seen

in the embryo are driving tissue elongation or if their orientation is a consequence

of the constriction required for the region to move into the embryo through the

blastopore. At the mid-gastrula stage there is relatively little CM remaining to be

internalized. Therefore, it seems unlikely that these medio-laterally oriented cells

could drive the extension of the anterior/posterior axis through the later gastrula

stages. Further complexity is added when considering that involuted CM cells are

apparently organized perpendicular to the BCR (Figure 2.6G) and that superficial

involuted CM cells are not strongly medio-laterally oriented beneath the BCR

substrate (Figure 2.8D). This organization is not consistent with the dominance of

medio-lateral intercalation that has been observed in explants. Thus, the

apparent medial to lateral orientation of cells in explants has not been directly

observed in an intact embryo and may suggest that elongation occurs through a

different mechanism in vivo; radially directed movements which have also been

  121

observed in explants (Wilson and Keller, 1991), may dominate in the embryo.

Further in vivo analysis must be done to address these questions.

Combining data from in vitro studies of LEM cell migration with studies of intact

embryos has also been problematic. LEM cells are thought to migrate on the

inner surface of the blastocoel roof (BCR) with PDGF signalling directing cells

toward the animal pole (Winklbauer and Nagel, 1991; Nagel et al. 2004). This is

thought to be important in the spreading of mesoderm over the inner BCR and in

the animally directed translocation of the mesoderm as a whole (Winklbauer and

Nagel, 1991; Davidson et al. 2002). Our understanding of this process comes

from analyzing the behaviour of explanted mesoderm on extracellular matrix

(ECM) and associated PDGF guidance cues that were transferred from the BCR

to a glass slide. LEM explants migrate on this substrate toward what would have

been the animal pole in an intact embryo. However, when transmission electron

micrographs are taken of whole embryo sections, there appears to be no contact

between LEM cells and the BCR (Figure 2.16A). It is therefore unclear if these

cells migrate on this surface in the embryo. Furthermore, my SEM analysis of this

region has shown that the leading edge is apparently over taken by a mass of

endoderm cells, which can be identified by their different orientation and cell size

(Figure 2.16B), well before the end of gastrulation. Therefore, the significance of

this process in animally directed movement and spreading of mesoderm needs

further clarification by studies with intact embryos.

  122

In my study of PCM morphogenesis, I have primarily made use of SEM to

analyze cell orientation and morphology over the course of gastrulation. This

approach allowed me to design explant systems, which maintain the in vivo cell

relationships between endoderm, mesoderm and BCR. By basing the design of

my explant systems on what I observed in whole embryos, the observations have

been transferrable to the in vivo situation.

2.3.2 Mesoderm regions undergo region specific morphogenetic movements Convergent extension movements of the chordamesoderm have been

extensively studied with less emphasis being given to the movements occurring

in other regions of the embryo. This has led to an idea that convergent extension

is a driving force of anterior/posterior axis elongation (Keller 2004; Solnica-Krezel

and Sepich, 2012). This is likely due to the dramatic shape change that occurs

when explanted chordamesoderm is cultured over the course of gastrulation.

This is misleading and detracts from the significance of movements in other

regions of mesoderm. Gastrulation is a system of movements, which is highly co-

ordinated; it is unlikely that any one of these movements could be deleted without

seriously disrupting the development of the embryo. In some cases disruption of

a movement may result in a less dramatic phenotype in later stage embryos,

  123

Figure 2.16: Electron microscropy of the LEM

(A) Transmission electron micrograph of Brachet’s cleft indicated by white

arrowheads. The LEM are the cells on the left and the BCR are the cells on the

right. (B) Scanning electon micrograph of the LEM of a stage 12 embryo, red

arrowheads highlight cells that are oriented toward the BCR in the LEM and the

adjacent endoderm. White arrow indicates endoderm mass which has overtaken

the LEM. EN, endoderm; LEM, leading edge mesendoderm.

  124

Figure 2.16: Electron microscropy of the LEM

  125

however to use a superficial observation like the external appearance of an

embryo, as criteria to determine the significance of a gastrulation movement

would be irresponsible. What may seem like a subtle effect in such a superficial

analysis could have a dramatic effect on the adult organism. To improve the

understanding of gastrulation as a system, I investigated the morphogenesis of

the under studied prechordal mesoderm and analyzed the orientations of cells in

lateral and ventral mesoderm regions.

Others and myself have shown through analysis of gene expression patterns and

differences in cell morphology that the dorsal mesoderm can be divided into at

least three types of mesoderm, XBra expressing CM, Gsc expressing PCM and

LEM (Smith et al. 1991, Wilkinson et al. 1991, Cho et al. 1991, Niehrs et al. 1994;

Saint-Jeannet et al. 1994, Damm and Winklbauer, 2004). The CM and PCM

regions can be further subdivided into involuted and non-involuted regions

(Figure 2.6G; 2.7G). I have shown that at the start of gastrulation non-involuted

CM cells are randomly oriented, non-involuted PCM is oriented parallel to the

animal/vegetal axis of the embryo and involuted PCM cells are oriented

perpendicular to the BCR (Figure 2.6B-D,G). LEM cells are oriented obliquely

with respect to the BCR throughout gastrulation (Figure 2.6A; 2.7A; 2.9A). At

later stages of gastrulation, post-involution CM cells are elongated, oriented

perpendicular to the BCR and extend protrusions toward and away from the BCR

while involuted PCM cells are oriented perpendicular to the BCR and extend

  126

protrusions toward the BCR. The PCM ultimately ends up thinning to a single

layer of cells by the end of gastrulation (Figure 2.9G; 2.10C,G). Therefore, these

sub-regions can be defined as independent motility domains because region

specific cell orientations are suggestive of distinct cell movements occurring in

each region. Furthermore, explants of CM, PCM and LEM show autonomy in

their movements suggesting that morphogenesis of these regions is independent

each other (Figure 2.11; Keller et al. 1985; Keller and Tibbetts, 1989; Wilson and

Keller, 1991; Winklbauer and Nagel, 1991; Nagel, et al. 2004; Damm and

Winklbauer, 2011). Depending on the stage of the gastrula, there can be up to

four distinct movement domains (in the early gastrula; LEM, upper PCM, lower

PCM, CM) within the dorsal mesoderm undergoing region specific movements

simultaneously.

The lateral and ventral mesoderm can also be divided into three regions, XBra

expressing posterior mesoderm, anterior mesoderm and LEM. Cell orientations in

the anterior mesoderm and LEM are similar to those found in the PCM and dorsal

LEM respectively (Figure 2.14A-B; 2.15A-B). Unlike on the dorsal side of the

embryo, the non-involuted XBra expressing region was not randomly oriented but

was oriented roughly perpendicular to the animal/vegetal axis of the embryo

(Figure 2.14C,F; 2.15C,F). Movements of lateral and ventral mesoderm have

also been largely neglected in favour of the more dramatic movements of the

  127

dorsal mesoderm. Further analysis of lateral and ventral mesoderm

morphogenesis will be required for a systematic understanding of gastrulation.

It is the sum of movements on the dorsal, lateral and ventral sides of the embryo

that results in the correct positioning of cells at the end of gastrulation.

Intercalation and directional migration of the dorsal PCM and LEM occurs

autonomously in explants of these tissues that have been isolated from the other

mesoderm regions. Thus, it is unlikely that convergent extension of the CM could

act as a driving force to ensure the correct positioning of cells in the PCM, dorsal

LEM or the lateral and ventral mesoderm. Therefore to understand gastrulation,

cell movements must be analyzed on a regional basis.

2.3.3 Intercellular migration and radial intercalation

Anterior mesoderm cells extend large lamelliform protrusions in the absence of

any external cues (Keller and Schoenwolf, 1977; Winklbauer and Nagel, 1991;

Damm and Winklbauer, 2011) and attach these protrusions to neighbouring cells.

This suggests that cells use the surfaces of their neighbours as points of

attachment while they squeeze between adjacent cells. Medial-lateral

intercalation in chordamesoderm explants may also be dependent on intercellular

migration since it is a BCR substrate independent and autonomous process

(Keller et al. 1984; Keller and Tibbetts, 1989; Wilson and Keller, 1991). Thus,

intercellular migration may play a role in several gastrulation movements. The

  128

cell-cell adhesion molecules involved in intercellular migration have not been

characterized. However, cell-cell adhesion molecules like EP/C-cadherin and

XB/U-cadherin are candidate molecules as they are the only classical cadherin

molecules known to be expressed in mesoderm tissue during gastrulation (Muller

et al. 1994; Kuhl and Wedlich, 1996).

The long axes and lamelliform protrusions of PCM cells are strongly oriented

toward the BCR (Figure 2.7B,F). Over the course of gastrulation these cells

migrate toward the BCR and intercalate at the mesoderm/BCR interface in an

instance of radial intercalation. PCM cells are elongated and unipolar, thus the

orientation and protrusive activity of the cells suggests unidirectional radial

intercalation, in the direction of the BCR. The thinning of the PCM is significant.

The region starts out as being 3 to 4 cell layers thick at the start of gastrulation

and ends up as a monolayer by the end of gastrulation (Figure 2.9C; 2.10F). Cell

shapes do not change dramatically over the course of gastrulation and therefore

intercalation of cells would result in an expansion of the area of PCM that is in

contact with the BCR.

Radial intercalation has been observed elsewhere during gastrulation. Although

medio-lateral intercalation is the primary movement observed in CM explants,

radial intercalation is observed during the early phase of convergent extension

when cells from deeper in open-faced CM explants appear at the surface (Wilson

  129

and Keller, 1991; Shih and Keller, 1992; Wallingford et al. 2000; Keller, 2002).

This in vitro observation may fit with my intact embryo observation of radially

oriented cells in the post-involution chordamesoderm (Figure 2.7G). Additionally,

radial intercalation has been reported in ectoderm during epiboly (Keller, 1980;

Marsden and DeSimone, 2001) at the anterior periphery of the LEM (Davidson et

al, 2001) and in the morphogenesis of the SM (Wilson and Keller, 1991; Yin et al.

2008). Radial intercalation of the PCM and the axial mesoderm (somitic and

chordamesoderm) can contribute significantly to elongation of the

anterior/posterior axis of the embryo. In the zebrafish SM, radial interaction

results in an anisotropic elongation of the tissue (Yin et al. 2008). This is because

intercalating cells preferentially insert between anterior/posterior neighbours

rather than lateral neighbours. This results in an extension in the

anterior/posterior direction. The molecular regulation of this process is not well

understood but may involve members of the non-canonical Wnt signalling

pathway (Yin et al. 2008). Furthermore, radial intercalation and anisotropic

spreading of the SM has been observed in Xenopus as well (Wilson and Keller,

1991). Thus, if this polarized anisotropic radial intercalation process is occurring

in SM, CM and PCM, it could have a significant effect on the elongation of the

anterior/posterior axis of the embryo.

I found that the orientation of lateral and ventral involuted mesoderm cells is also

consistent with unidirectional radial intercalation. In a pervious study, ventral

  130

mesoderm cells were labelled with a vital dye prior to involution (Ibrahim and

Winklbauer, 2001). When the BCR was fractured away following involution,

labelled and unlabelled cells were in contact with the BCR, which suggests that

unlabelled cells from deeper in the explant had arrived at the mesoderm/BCR

boundary by radial intercalation.

Unidirectional radial intercalation in the PCM and corresponding lateral and

ventral mesoderm domains is associated with strong tissue spreading in a zone

which has to expand most as the mesoderm moves animally within the spherical

geometry of the gastrula (Keller and Tibbetts, 1989; Winklbauer and Schurfeld,

1999; Ibrahim and Winklbauer, 2001). The expansion of the PCM almost

certainly contributes to elongation of the anterior/posterior axis of the embryo,

however the extent of this contribution is not known. Theoretically the extent of

spreading could be determined by measuring the change in area of mesoderm

occupied by labelled PCM cells over the course of gastrulation. Dorsally, I have

shown that this spreading is dependant on the BCR, but does not require direct

contact with its fibronectin matrix (Winklbauer and Schurfeld, 1999). Furthermore,

this movement is tissue autonomous and therefore does not depend on the

movements occurring in the CM or the LEM. My results suggest that spreading

may be due to a long distance attraction of mesoderm cells by the BCR, leading

to radial intercalation.

  131

2.4: Materials and Methods

2.4.1: Embryos and microinjections  Xenopus laevis embryos were fertilized in vitro in 1/10X Modified Barth’s Solution

(MBS). Fertilized embryos were de-jellied with 2% cysteine solution diluted in

1/10X MBS at pH of 8.0. Embryos were left to develop to the required stages in

1/10X MBS in and incubator at 15o Celsius. Developing embryos were staged

with The Normal Table of Xenopus Development by Nieuwkoop and Faber, 1967.

For confocal microscopy experiments, embryos were injected with 150 pg per

blastomere of mRNA encoding a cell membrane bound RFP molecule in the

equatorial region of the two dorsal blastomeres at the four cell stage in 4% Ficoll

solution (Sigma) using pulled glass needles and a Nanoject II microinjector

(Fisher). Injected emrbyos were kept in 4% Ficoll solution for one hour following

injection to prevent the loss of cytoplasm while the blastomeres healed around

the injection site. Embryos were transferred to 1/10X Modified Barth’s solution

(MBS) and left to develop at 15°C until the required stage.

2.4.2: In-Situ Hybridization and In-Situ Probe Synthesis  The in situ hybridization protocol and solutions were adapted from Harland, 1991.

Embryos were fixed for 1 hour in MEMFA and were then sagitally fractured.

Embryos were fixed for an additional hour in MEMFA. Samples were treated with

Proteinase K (Sigma) for 10 minutes followed by two washes with

  132

triethanolamine and three washes with triethanolamine/acetic anhydride mixture.

Embryos were re-fixed in 4% formaldehyde for 30 minutes followed by washes

with PBS-Tween. Embryos were immersed in hybridization buffer and subjected

to a pre-hybridization period of 4 hours at 60o Celsius before adding the RNA in

situ probe. After addition of probe, embryos were incubated overnight at 60o

Celsius. Embryos were washed in 2X SSC buffer followed by washes in 0.2X

SSC. Embryos were washed in maleic acid buffer and blocked with 2% BMB

blocking buffer plus 20% Gibco Lamb Serum (Life Technologies) in maleic acid

buffer for one hour at room temperature. Subsequently, samples were incubated

in 2% BMB blocking buffer, 20% Gibco Lamb Serum and a 1/2000 dilution of the

anti-digoxygenin antibody in maleic acid buffer for 4 hours at room temperature.

Samples were washed overnight in maleic acid buffer. Samples were washed

with alkaline phosphatase buffer before the addition of BM Purple (Roche).

Embryos were incubated in BM Purple at room temperature for 2.

Plasmids (DB30 for XBra, pBluescript for Gsc) were linearized with BgIII for XBra

(M. Sargent Lab) and EcoRI for Gsc (H. Steinbeisser Lab). Anti-sense mRNA

probes for XBra and Gsc were generated by in vitro transcription using a

mMessage mMachine T7 RNA polymerase kit (Life Technologies) according to

the kit protocol. The kit supplied nucleotide triphosphates (NTPs) were

substituted with digoxigenin-labeled NTPs (Roche).

2.4.3: Scanning Electron Microscopy  

  133

Embryos were fixed in 2.5% glutaraldehyde/0.1M sodium cacodylate solution, pH

7.4, overnight at 4° Celsius before being fractured with a scalpel as required for

the experiment being performed. Embryos were post-fixed in 1% osmium

tetraoxide (SPI or Cedarlane) diluted in 0.1M sodium cacodylate solution (Sigma)

for 1 hour at 4° Celsius before being treated with ethanol/0.1M cacodylate

solution dehydration series. Embryos were treated with 50% ethanol diluted in

0.1M cacodylate solution for 20 minutes, followed by two treatments of 100%

ethanol for 20 minutes each. Embryos were then treated with 50%

hexamethyldisilizane (Sigma) diluted in 100% ethanol for 30 minutes, followed by

two treatments of 100% hexamethydisilizane for 30 minutes each. Specimens

were dried overnight, mounted on SEM stubs (SPI) with carbon tape (SPI) and

sputter coated with gold-palladium. Samples were imaged using a Hitachi S2500

scanning electron microscope.

2.4.4: Transmission Electron Microscopy  The protocol used for the preparation of samples for transmission electron

microscopy is adapted from Kurth et al. 2010. Embryos were fixed in 2.5%

glutaraldehyde/0.1M sodium cacodylate solution, pH 7.4 overnight at 4o Celsius.

For orientation purposes, samples were embedded in 3% low melting

temperature agarose and postfixed in 1% osmium tetraoxide solution for 4 hours

at 4o Celsius. Samples were subsequently subjected to washes in 1X PBS and

water. Samples were dehydrated with an ethanol dehydration series, 30, 50, 70

  134

90 and 96, 100 and 100% ethanol for 30 minutes each. Samples were infiltrated

with Spurr’s resin in solutions of 1:2, 1:1 and 2:1 Spurr’s to 100% ethanol for 1.5

hours each and then in pure Spurr’s resin for 2 hours and again overnight.

Samples were oriented in moulds and baked overnight at 65o Celsius. Semi-thin

sections were stained with toluidine blue and ultra-thin sections were stained with

lead citrate and uranyl acetate. Samples were imaged with a Hitachi H7000

transmission electron microscope.

2.4.5: Confocal Microscopy  Explants were filmed with a Zeiss LSM510 laser scanning confocal system.

mbRFP was visualized by excitation with a 543nm helium/neon laser at 100%

intensity. The following confocal settings were used, pinhole size: 701µm,

detector gain: 940, amplitude offset: -1.22, amplitude gain: 1. The microscope

system was controlled by Zeiss LSM5 software.

2.4.6: Explants  For slice explants, slices of dorsal mesoderm and attached endoderm were

combined with inner ectoderm from the animal BCR and secured under a

coverslip for filming (Figure 2.11A). For mesoderm-BCR or mesoderm-endoderm

combined explants, PCM, inner cells from the animal region of the BCR, or

endoderm from the vegetal cell mass were dissected, combined and secured

under a strip of coverslip, with the mesoderm side originally in contact with the

BCR facing down, and the BCR explant positioned opposite to what would have

  135

been the anterior side of the mesoderm in vivo (Figure 2.12A). Explants were

cultured in 1X MBS on 1% BSA treated glass bottom dishes (Mak Tek) during

filming. Explants were filmed for 90 minutes using a Zeiss Axiovert 200M inverted

microscope. Timelapses were analyzed with the Zeiss Axiovision 4.8 software

suite. For explants filmed with the confocal microscope, mesoderm expressing

the mbRFP construct was combined with a BCR explant from an un-injected

embryo and cultured as above.

2.4.7: Statistical Analysis  Statistical analysis was performed using GraphPad inStat 3. The Mann-Whitney

statistical test at the 95% confidence level was used for all statistical calculations.

  136

Chapter Three: PDGF-A controls prechordal mesoderm cell orientation and radial intercalation

during gastrulation

A version of this chapter was published as Damm, E.W. and Winklbauer, R. (2011) PDGF-A controls mesoderm cell orientation and radial intercalation during Xenopus gastrulation. Development,

138: 565-575

  137

3.1: Introduction Radial intercalation is a basic morphogenetic process, which plays an essential

role in the large-scale rearrangement of germ layers during gastrulation. The

upper half of the amphibian blastula consists of prospective ectoderm, while the

mesoderm forms a narrow ring below the equator. During gastrulation, these

regions expand to eventually cover the endodermal core of the embryo (See

Figure 1.1). Since there is no embryonic growth during gastrulation, expansion

occurs by a respective thinning of these two germ layers, brought about mainly

by an interdigitation of cells in the radial direction, i.e. by radial intercalation

(Keller et al. 2003).

In the Xenopus embryo, radial intercalation has been observed in all regions of

the mesoderm (Keller and Tibbetts, 1989; Wilson and Keller, 1991; Keller, 2002;

Winklbauer and Schurfeld, 1999; Ibrahim and Winklbauer, 2001; Damm and

Winklbauer, 2011). In the CM and the SM, it cooperates with mediolateral cell

intercalation to promote convergent extension (Keller and Tibbetts, 1989; Wilson

and Keller, 1991; Keller, 2002; Yin et al, 2008). In the PCM, radial intercalation

results in thinning and spreading of the region (See Chapeter Two; Damm and

Winklbauer; 2011) Despite its ubiquity, however, radial intercalation is not well

understood at the cellular level, and its molecular control is virtually unknown. In

the Xenopus mesoderm, radial intercalation is an active process (See Chapter

Two; Wilson and Keller, 1991; Winklbauer and Schurfeld, 1999; Ibrahim and

  138

Winklbauer, 2001; Damm and Winklbauer, 2011), but the molecular cues

regulating cell movements have not been determined.

Here, I show that radial intercalation in the prechordal mesoderm (PCM), i.e. in

the region where thinning of the mesoderm layer is greatest (See Figure 2.9G;

2.10C), is driven by an attraction of its cells to the blastocoel roof (BCR) by

PDGF-A. I show that the three PDGF-A isoforms generated by alternative

splicing are expressed in the BCR. The long and intermediate sized variants (lf-

PDGF-A, int-PDGF-A respectively) associate with cell surface and extracellular

components such as heparin sulfate proteoglycans and fibronectin (Smith et al.

2009; Andersson et al. 1994; Raines and Ross, 1992) and regulate the animally

directed migration of the dorsal LEM on the BCR inner surface (Nagel and

Winklbauer; 1999; Nagel et al. 2004). Thus far, no function has been identified for

the short, diffusible isoform (sf-PDGF-A) (Figure 3.1A-B) (Andrae et al. 2008;

Mercola et al. 1988; Raines and Ross, 1992). I also confirm that PDGFR-α, the

cognate receptor for PDGF-A is expressed in the mesoderm (Figure 3.3A-D).

Thus, showing the complementary expression of ligand in the ectoderm and

receptor in the mesoderm that is typical of PDGF/PDGFR-α expression in

vertebrate gastrulae (Ataliotis et al. 1995; Orr-Urtreger and Lonai, 1992; Yang et

al. 2008).

  139

My data shows that sf-PDGF-A acts as a long-range signal to orient and attract

the cells of the multilayered PCM toward the BCR, leading to radial intercalation.

This establishes a dual role for PDGF-A in Xenopus gastrulation, based on

differential splicing, where the long and intermediate sized isoforms regulate

directional mesoderm migration across the BCR surface (Nagel et al. 2004) and

the short form regulates radial intercalation of PCM cells. Furthermore, for the

first time I provide insight into the molecular basis of a radial intercalation

movement in a vertebrate embryo.

3.2: Results 3.2.1: Expression of PDGF-A isoforms is restricted to the inner cell layer of the blastocoel roof PDGF-A is a known chemoattractant and experiments have shown that it

provides directional information to migrating leading edge mesendoderm cells

(Nagel et al. 2004). To determine if PDGF-A could play a role in regulating

intercalation movements of PCM cells, I investigated the expression patterns of

PDGF-A isoforms and PDGFR-α during gastrulation.

Ataliotis et al. 1995 showed that during gastrulation, PDGFR-α expression is

confined to dorsal and ventral involuted mesoderm cells with PDGF-A expression

  140

Figure 3.1: The structure of Xenopus PDGF-A isoforms (A) Clustal comparison of amino acid sequences of the three Xenopus PDGF-A

isoforms. Asterisks indicate 100% amino acid identity match between the three

isoforms. Light blue, N-terminal leader sequence; green, pro-peptide domain;

orange, conserved PDGF growth factor domain; dark blue, hydrophobic amino

acid domain; red, positively charged cell retention motif; brown, short isoform C-

terminus. (B) Schematic diagrams of the three Xenopus PDGF-A isoforms.

  141

Figure 3.1: The structure of Xenopus PDGF-A isoforms

  142

in the overlying ectoderm. I confirmed these results with in situ hybridization

experiments on mid and late gastrula stage embryos using RNA anti-sense

probes targeting PDGF-A and PDGFR-α. As previously reported, PDGF-A

expression is confined to the inner layer of cells of the BCR while expression of

PDGFR-α is confined to involuted XBra expressing posterior mesoderm and

anterior mesoderm (PCM dorsally) on the dorsal and ventral sides of the embryo

(Figure 3.2A-D). It has previously been suggested that PDGF-A may form a

molecular gradient along the inner layer of the BCR (Nagel et al. 2004, Ataliotis

et al. 1995). This gradient would provide a mechanism for providing directional

guidance to migrating LEM cells (Nagel et al. 2004). PDGF-A expression in the

ectoderm appears to taper in an animal to vegetal direction (Figure 3.2A-B). This

expression pattern is consistent with the idea of a PDGF-A expression gradient in

the BCR (Nagel et al. 2004).

The long and short PDGF-A isoforms are nearly identical over their lengths but

differ in amino acid sequence at their C-termini. The long isoform contains a short

stretch of hydrophobic amino acids upstream of a positively charged domain

(Figure 3.1A-B). The function of the hydrophobic domain remains unknown,

however the hydrophobicity of the domain may contribute to cell retention

through interactions with cell membrane phospholipids. The function of the

positively charged domain in cell/matrix retention has been well described

(Raines and Ross, 1992; Kelly et al. 1993; Andersson et al. 1994). The short

  143

Figure 3.2: Expression of PDGF-A and PDGFRα in the Xenopus gastrula

(A-D) In situ hybridization showing the expression domains of PDGF-A (A,B) and

PDGFRα (C,D) at mid (A,C) and late (B,D) gastrula stages. Boundaries of

expression domains are indicated by white arrowheads. (E) Diagram of the long

PDGF-A isoform with PCR primer pair locations indicated; red, cell retention

motif. (F) RT-PCR detection of PDGF-A isoforms and PDGFR-α in BCR,

prechordal mesoderm (PCM) and endoderm (Endo.). Red arrow, 241 b.p. lf-

PDGF-A product; purple arrow, 225 b.p. int-PDGF-A product; blue arrow, 113

b.p. sf-PDGF-A product. RT-PCR results were combined from multiple

experiments. XBra, Xbrachyury; Gsc, goosecoid; ODC, ornithine decarboxylase;

-RT, reverse transcriptase not used in cDNA preparation.

  144

Figure 3.2: Expression of PDGF-A and PDGFRα in the Xenopus gastrula

  145

isoform lacks both of these domains and has a unique C-terminal sequence

(Figure 3.1A-B). Sequence similarity between PDGF isoforms of different species

is strikingly high, particularly in the core PDGF domain and the C-terminus,

highlighting the functional significance of these regions (Figure 3.3A-B). The long

isoform sequences of mammals and chick lack the hydrophobic domain found in

Xenopus, however Anolis carolinensis (Anole lizard) contains a stretch of amino

acids between the PDGF domain and the positively charged retention motif

(Figure 3.3A). Although the sequence of this region differs in residue identity

when compared with Xenopus, the region in both species is made up of

hydrophobic amino acids, which may suggest functional similarity (Figure 3.3A).

The short isoform C-terminus is identical in Xenopus and mammals, but the

sequences of the chick short (sf-PDGF-A2) and intermediate (sf-PDGF-A1)

isoforms differ (Figure 3.3B).

Since sf-PDGF-A is not retained at the cell surface and may penetrate deep into

tissues (Raines and Ross, 1992), I hypothesized that this isoform could control

deep PCM cell movements. The anti-sense RNA probes used for in situ

hybridization experiments were unable to distinguish between the different

PDGF-A isoforms. To determine which splice isoforms are expressed in the BCR,

I designed PCR primers flanking the putative spliced-out exon of the short

isoform mRNA (Figure 3.2E). The long isoform amplicon should be distingishable

from the short isoform by its larger size (Figure 3.2E,F).

  146

RNA from BCR was processed for RT-PCR. Two bands corresponding to

fragments of 240 b.p. (base pairs) (expected for long isoform) and 110 b.p. (short

isoform) were observed (Figure 3.2F). Unexpectedly, a third band was found at

225 b.p. (Figure 3.2F). The 110 b.p. band appeared brightest (Figure 3.2F),

suggesting that the short isoform may be more abundant than the 240 b.p. and

225 b.p. species individually. Expression of sf-PDGF-A in BCR cells is consistent

with a putative role for this isoform in regulating deep PCM cell movements.

Typically, two splice isoforms of PDGF-A have been identified in model systems

(see above). In chick however, a third intermediate sized isoform was found

(Figure 3.2B) (Horiuchi et al. 2001). I hypothesized that the 225 b.p. band could

correspond to an intermediate sized isoform similar to that discovered in chick.

DNA sequencing analysis of the unknown band showed a PDGF-A isoform (int-

PDGF-A), which corresponded to the longer isoform identified in mammalian

systems, containing a cell retention motif but lacking the hydrophobic domain

(Figure 3.2A). The int-PDGF-A cell retention motif sequence was identical to that

of lf-PDGF-A (Figure 3.1A). Despite the lack of a hydrophobic domain in int-

PDGF-A, the sequence similarity between the long and intermediate form C-

termini and their co-expression may suggest a level of functional redundancy

between these isoforms in Xenopus.

  147

Figure 3.3: PDGF-A isoforms are highly conserved between species

(A,B) Clustal alignments of Xenopus lf-PDGF-A/int-PDGF-A (A) and sf-PDGF-A

(B) with the corresponding PDGF-A isoforms in human (Homo sapiens), mouse

(Mus musculus), chicken (Gallus gallus), and the Anole lizard (Anolis carolinesis).

Orange, conserved PDGF growth factor domain; purple, hydrophobic domain;

red, positively charged cell retention motif. Asterisks indicate full amino acid

conservation between species; : indicates amino acid subsitituion between

species with an amino acid of highly similar properties; . indicates amino acid

substitution between species with an amino acid of weakly similar properties.

  148

Figure 3.3: PDGF-A isoforms are highly conserved between species

  149

I confirmed (Ataliotis et al., 1995) the lack of PDGF-A expression in mesoderm

and endoderm, and the presence of PDGFR-α RNA in mesoderm (Figure 3.2F).

Unexpectedly, endoderm cells also expressed PDGFR-α (Figure 3.2F).

Expression of the receptor, which had not been detected by in situ hybridization,

could explain previous findings, which showed that vegetal cells migrate

directionally on BCR matrix, similar to leading edge mesendoderm (Winklbauer

and Nagel, 1991, Nagel et al. 2004).

3.2.2: Inhibition of PDGF signalling interferes with PCM radial intercalation, but not with CM cell orientation The expression of sf-PDGF-A in the BCR and PDGFR-α in the mesoderm makes

PDGF signalling a candidate for orienting PCM cell movements. To test this

notion, I treated stage 11 embryos with the cell-permeable PDGFR-α inhibitor,

AG1296, until DMSO treated controls reached stage 12. Although the PCM

became thinner in treated embryos, it actually consisted of two layers of cells

oriented in parallel to the BCR (Figure 3.4A,I) reflected in a layer index of 2.0 +/-

0.7 (Figure 3.4B). The average thickness of the region was 47.7 µm +/- 3.2

(Figure 3.4A). At the same developmental stage, DMSO treated control embryos

had formed a single cell layer with a layer index 1.3 +/- 0.5 (Figure 3.4B-C,F) and

an average thickness of 32.3 µm +/- 2.6 (Figure 3.4A). In the AG1296 treated

LEM, cells appeared more obliquely oriented compared to controls (Figure

3.4G,J). LEM cells are normally obliquely oriented with respect to the BCR, thus

  150

Figure 3.4: PDGF-A inhibition disrupts PCM radial intercalation

(A) PCM thins significantly (p-value < 0.0001, n = 10 embryos/stage) between

stages 11 and 12; thinning is reduced upon AG1296 treatment (p-value < 0.0001,

n = 10 embryos). (B) Change in layer index of PCM over time and following

AG1296 treatment at stage 11. (C) PCM layer index is significantly higher at

stage 12 following AG1296 treatment (p value = 0.0013, n = 13 embryos) or at

stage 11 in PDGF-A MO injected embryos (p value < 0.0001, n = 12 embryos);

chordamesoderm is unaffected (p value = 0.0886, n = 13 embryos). (A-C) Error

bars represent standard error of the mean, asterisks indicate statistically

significant results. (D) Rose diagrams show the orientation of chordamesoderm

cells, BCR = 0 degrees, n = number of cells. No significant difference between

controls and AG1296 treated embryos (p value = 0.9842). (E-J) Scanning

electron micrographs of sagitally fractured stage 12 embryos treated with DMSO

(E-G) or 10 μM AG1296 (H-J) two hours prior to fixation, coloured dashed lines

indicate region boundaries. (F) Red arrowheads, PCM cells in a single layer in

DMSO controls. (I) Red arrowheads, superficial PCM cells in contact with the

BCR; red arrows, second layer of cells in AG1296 treated embryos. (G) Red

arrowheads, LEM cells oriented perpendicular to the BCR in controls, and (J)

obliquely oriented LEM cells in AG1296 treated embryos. BCR, blastocoel roof;

EN, endoderm; AC, archenteron; BC, blastocoel.

 

  151

Figure 3.4: PDGF-A inhibition disrupts PCM radial intercalation

  152

the increase in obliqueness when PDGF signalling is inhibited may suggest that

the normal orientation of these cells is due to reduced PDGF signalling in this

region. These results confirmed that the increased thickness of the PCM in

embryos with disrupted PDGF-A signalling is primarily due to defects in radially

oriented cell rearrangement rather than changes in cell shape/size or orientation.

The post-involution CM appeared unaffected by PDGF-A inhibition. As in

controls, cells from AG1296 treated embryos were elongated perpendicular to the

BCR (Figure 3.4D,E,H), and extended protrusions either toward or away from the

BCR (Figure 3.4D,E,H). Furthermore, the layer indices of AG1296 and DMSO

treated embryos were similar (Figure 3.4B), suggesting that PDGFR-α inhibition

did not affect movements in this region even though PDGFRα is expressed in the

CM. The results suggest that PDGF-A signalling is required to orient PCM cells

toward the BCR and to prompt radial intercalation, but that radial orientation of

the chordamesoderm is regulated by a different mechanism.

3.2.3: sf-PDGF-A is an instructive cue required for radial orientation of PCM cells To further analyze the role of PDGF-A in radial intercalation, I made use of an

ATG morpholino oligonucleotide (PDGF-A MO) that had previously been shown

to inhibit PDGF-A signalling (Nagel et al. 2004). Following PDGF-A knockdown,

embryo phenotypes resembled those previously described (Ataliotis et al. 1995;

Nagel et al. 2004). Mesoderm patterning was not affected. Both Gsc and XBra

  153

Figure 3.5: PDGF-A knockdown by morpholino oligonucleotides disrupts the orientation of PCM cells

(A,C) Scanning electron micrographs of sagittaly fractured embryos injected with

PDGF-A morpholino. White arrowheads highlight cells oriented parallel to the

BCR in prechordal mesoderm (A) and leading edge mesendoderm (C). (B,D)

Rose diagrams showing the percentage orientation of prechordal mesoderm cells

(B) and leading edge mesendoderm cells (D) with respect to the BCR of embryos

injected with the PDGF-A morpholino. 0 degrees, dorsal BCR; 90 degrees,

animal pole. “n” is number of cells scored from 14 embryos analyzed. (E,F) In situ

hybridization of Xbrachyury (E) and goosecoid (F) expression patterns in

embryos injected with PDGF-A morpholino. (G) Scanning electron micrograph of

the dorsal blastopore lip region of an NF stage 11 gastrula injected with PDGF-A

morpholino, red dotted line outlines the region that has failed to involute, white

arrow points at the dorsal blastopore lip. BCR, blastocoel roof; EN, endoderm;

BC, blastocoel; XBra, Xbrachyury; Gsc, goosecoid. Orange, leading edge

mesendoderm; yellow, prechordal mesoderm.

  154

Figure 3.5: PDGF-A knockdown by morpholino oligonucleotides disrupts the orientation of PCM cells

  155

were still expressed although the positions of the expression domains were

slightly altered (Figure 3.5E-F). The change in expression pattern position is in

agreement with a failure of mesoderm involution and archenteron elongation

(Figure 3.5G) (Nagel et al. 2009). These results show that while specification of

mesodermal regions is not affected by an inhibition of PDGF-A signalling, normal

morphogenetic movements are disrupted.

In morpholino injected embryos, PCM, ventral involuted mesoderm, dorsal LEM

and ventral LEM cells were elongated and extended protrusions from a single

pole (Figure 3.5A,C; 3.6A,C). However, PCM cells were no longer oriented strictly

toward the BCR, but at angles between 0 and 90 degrees, i.e. between the

original orientation and that of the adjacent endodermal cells (Figure 3.5A-B;

3.6A-B). As a result, the PCM layer index of PDGF-A MO injected embryos was

increased compared to controls (Figure 3.4B), which is consistent with an

inhibition of radial intercalation. Similar to chordamesoderm in AG2196 treated

embryos, the ventral XBra expressing mesoderm was unaffected by PDGF-A

inhibition (Figure 3.6E,F). Dorsal and lateral LEM cells from morphant embryos

were also more obliquely oriented with respect to controls (compare Figures 3.5D

and 2.8A; 3.6D and 2.16A). When embryos were injected with mRNA encoding a

dominant negative PDGF-A construct, dnPDGF-A 1308 (Mercola et al. 1990)

(Figure 3.7A,C-D), or a dominant negative version of PDGFRα, PDGFR-37

(Ataliotis et al. 1995) (Figure 3.7B,E-F), the effects on PCM and dorsal LEM cell

orientation were similar to PDGF-A morphant embryos. From these results I

  156

Figure 3.6: PDGF-A knockdown by morpholino oligonucleotides disrupts the orientation of involuted ventral mesoderm cells

(A,C,E) Scanning electron micrographs of the ventral leading edge

mesendoderm (A), involuted ventral mesoderm (C) and ventral XBra expressing

mesoderm (E) of a sagitally fractured mid-gastrula stage (NF stage 11) embryo

injected with PDGF-A MO, white arrowheads highlight cells oriented parallel to

the BCR (A,C) or outer embryonic epithelium (E). (B,D,F) Rose diagrams

showing the percentage orientation of ventral leading edge mesendoderm (B),

ventral involuted mesoderm (D) and ventral XBra expressing mesoderm (F), Blue

rose diagrams use the angle at which protrusions are extended from the cell

body with respect to the ventral BCR as a measure of orientation. Red rose

diagrams use the orientation of the cell long axis with respect to the ventral

external embryonic epithelium. 0 degrees, ventral BCR or ventral external

embryonic epithelium; 90 degrees, animal pole. “n” is number of cells scored

from 9 embryos analyzed. Blue, ventral leading edge mesendoderm; green,

involuted ventral mesoderm; purple, XBra expressing ventral mesoderm. BCR,

blastocoel roof; EN, endoderm; BC, blastocoel.

  157

Figure 3.6: PDGF-A knockdown by morpholino oligonucleotides disrupts the orientation of involuted ventral mesoderm cells

  158

propose that during gastrulation, the mesoderm can be divided into PDGF-A

sensitive (PCM and anterior ventral mesoderm, LEM, non-XBra expressing) and

insensitive zones (XBra expressing zones, including CM) with respect to its role

in regulating cell orientation. PDGF-A signalling is important in determining the

radial orientation of cells in involuted mesoderm all around the embryo and may

influence the dorsal, lateral and ventral LEM, while the orientation of cells in XBra

expressing mesoderm is independent of PDGF-A signalling.

When morpholino resistant mRNA encoding sf-PDGF-A was co-injected with

PDGF-A MO, cell orientation (Figure 3.8A,B) and the layer index of the PCM

were significantly rescued (Figure 3.4B). However, unlike sf-PDGF-A, co-injection

of PDGF-A MO and mRNA encoding morpholino resistant lf-PDGF-A or int-

PDGF-A failed to rescue PCM cell orientation (Figure 3.8C-F). These results

suggest that sf-PDGF-A can influence the orientation of deep PCM cells, possibly

because it is capable of signalling at a distance from its source while the cell-

retention motif containing isoforms may not be able to signal to deeper cells.

Furthermore, as seen from the locations of XBra and Gsc expression cells, sf-

PDGF-A was unable to rescue involution and archenteron elongation (Figure

3.8G-H). This suggests that the orientation of PCM cells and involution are

independent processes.

  159

Figure 3.7: PDGF signaling knockdown by dominant negative ligand or receptor disrupts PCM cell orientation

(A,B) Scanning electron micrographs of a sagittaly fractured embryos with

disrupted PDGF signaling due to expression of dnPDGFRα-37 (A) or dnPDGF-A

1308 (B), white arrowheads highlight cells oriented parallel to the BCR. (C-F)

Rose diagrams showing percentage orientation of prechordal mesoderm (C,E)

and leading edge mesendoderm cells (D,F) with respect to the BCR in embryos

expressing dnPDGFRα-37 (C,D) or dnPDGF-A 1308 (E,F). “n” is number of cells

scored from 9 (dnPDGFRα-37) and 13 (dnPDGF-A 1308) embryos analyzed.

BCR, blastocoel roof; EN, endoderm, BC, blastocoel.

  160

Figure 3.7: PDGF signaling knockdown by dominant negative ligand or receptor disrupts PCM cell orientation

  161

The overexpression of sf-PDGF-A, lf-PDGF-A or int-PDGF-A in the BCR had no

effect on the orientation of PCM cells (Figure 3.9A-C,E,G,I). However,

overexpression of sf-PDGF-A in the BCR caused deep LEM cells to become

more strongly oriented toward the BCR (compare Figures 3.9A,F and 2.8A,E).

LEM cells were unaffected by overexpression of lf-PDGF-A (Figure 3.9B,H) and

int-PDGF-A (Figure 3.9C,J) These results suggest that although lf-PDGF-A and

likely int-PDGF-A are sensed by cells directly in contact with the BCR or its

matrix (Nagel et al. 2004), only the sf-PDGF-A signal is relayed to cells deeper in

the embryo. This is because interactions between the cell retention motif and the

cell surface could anchor these molecules close to the source (Raines and Ross,

1992; Kelly et al. 1993; Andersson et al. 1994). Therefore, it is not surprising that

neither of these molecules are able to function as a directional determinants for

PCM cells.

I hypothesized that if PDGF-A acts as an instructive guidance molecule,

overexpression of sf-PDGF-A in PCM cells should interfere with the ability of cells

to detect an endogenous sf-PDGF-A gradient. Under these circumstances, cells

would be detecting the PDGF-A signal from surrounding cells as well as the

BCR, abolishing any putative gradient. Indeed, in embryos expressing a sf-

PDGF-A construct in the PCM, which is not resistant to the PDGF-A morpholino,

cell orientation was similar to that in PDGF-A morphant embryos (Figure

3.10A,C). LEM cell orientation was unaffected (Figure 3.10B,D). This loss of

  162

Figure 3.8: sf-PDGF-A is required for radial orientation of PCM cells

(A,C,E) Scanning electron micrographs of sagittaly fractured mid-gastrula stage

(NF stage 11) embryos co-injected with PDGF-A morpholino and morpholino

resistant variants of sf-PDGF-A (A), or lf-PDGF-A (C) or int-PDGF-A RNA, white

arrowheads highlight cells oriented perpendicular to the BCR, red arrowheads

highlight cells oriented parallel to the BCR. (B,D,F) Rose diagrams show the

percentage orientation of prechordal mesoderm cells from embryos co-injected

with PDGF-A morpholino and morpholino resistant variants of sf-PDGF-A (B) or

lf-PDGF-A (D) or int-PDGF-A (F) RNA, rose diagrams use the angle at which

protrusions are extended from the cell body with respect to the dorsal BCR as a

measure of orientation. 0 degrees, dorsal BCR; 90 degrees, animal pole. “n” is

number of cells scored from 9 embryos (sf-PDGF-A), 12 embryos (lf-PDGF-A)

and 10 embryos (int-PDGF-A) analyzed. (G,H) In situ hybridization showing the

expression of Xbrachyury (G) and goosecoid (H) in embryos co-injected with

PDGF-A morpholino and sf-PDGF-A RNA, white arrowhead indicates dorsal

blastopore lip. Yellow, prechordal mesoderm. BCR, blastocoel roof; EN,

endoderm; XBra, Xbrachyury; Gsc, goosecoid.

  163

Figure 3.8: sf-PDGF-A is required for radial orientation of PCM cells

  164

orientation was rescued by co-injection with PDGF-A MO (Figure 3.11A-D). This

is consistent with the short isoform of PDGF-A providing an instructive long-range

guidance cue, emitted by the BCR.

3.2.4: sf-PDGF-A is required for directional intercellular mesoderm migration in an explant system  Results from the above experiments support the hypothesis that PDGF-A is an

important regulator of PCM morphogenesis. To visualize the effect of knocking

down PDGF-A on cell movements in the embryo, slice explants (Figure 3.12B)

from embryos injected with PDGF-A MO in the BCR were made. The movements

of PCM cells were erratic and unlike un-injected control explants (see Figure

2.11C), cells did not intercalate at the mesoderm/ectoderm boundary. These

results showed that PDGF-A is important in regulating the movements of cells in

embryo slices which preserve the original organization of tissues in the embryo.

In order to analyze the effect of PDGF-A knockdown on directional migration, I

made use of the explant system I developed for observing live PCM migration

(see Figure 2.12A). This system allowed me to test whether PDGF-A is

responsible for re-orienting cell movements because the BCR explant is placed

on the opposite side from the endogenous direction of cell movement. BCR inner

layer cells injected with PDGF-A MO were placed in contact with un-injected

mesoderm explants; cell tracks were followed and velocity measurements taken.

Mesoderm cells were motile, but not moving directionally (Figure 3.13A,B-B’’),

  165

Figure 3.9: Overexpression of PDGF-A constructs in the BCR does not affect PCM cell orientation

(A-C) Scanning electron micrographs of sagittaly fractured mid gastrula (NF

stage 11) stage embryos over-expressing sf-PDGF-A (A), lf-PDGF-A (B) or int-

PDGF-A (C) in the BCR, (A-C) red arrowheads highlight cells that are oriented

parallel to the BCR, white arrowheads highlight cells oriented perpendicular to

the BCR. (D-I) Rose diagrams showing the percentage orientation of prechordal

mesoderm cells (D,F,H) and leading edge mesendoderm cells (E,G,I) with

respect to the BCR. Rose diagrams use the angle at which protrusions are

extended from the cell body with respect to the dorsal BCR as a measure of

orientation, 0 degrees, dorsal BCR; 90 degrees, animal pole. “n” is number of

cells scored from 11 embryos (sf-PDGF-A), 5 embryos (lf-PDGF-A) and 13

embryos (int-PDGF-A) analyzed. Orange, leading edge mesendoderm; yellow,

prechordal mesoderm, BCR, blastocoel roof; BC, blastocoel; EN, endoderm.

  166

Figure 3.9: Overexpression of PDGF-A constructs in the BCR does not affect PCM cell orientation

  167

until at 90 minutes some cells acquired directionality (Figure 3.13A,B-B’’). This

recovery may be due to an incomplete knockdown of PDGF-A as is typical for

morpholino knockdown. Co-injection of morpholino resistant sf-PDGF-A rescued

the effect of the PDGF-A MO (Figure 3.13C,D-D’’). From these results, I conclude

that PDGF-A determines the direction of PCM cell movement, although cells are

able to migrate randomly in absence of PDGF-A signalling (see Figure 2.14D).

3.2.5: An instructive role for sf-PDGF-A signaling in directional migration To see whether PDGF-A determines the direction of migration instructively, I

carried out gain-of-function experiments. The in vitro migration assay was

modified by combining mesoderm with endoderm in place of BCR (Figure 2.13A).

Since PDGF-A is not normally expressed in endoderm, mesoderm cells moved

randomly (Figure 3.14A,B-B’’). Migration was also random when lf-PDGF-A or

int-PDGF-A was expressed in the endoderm, confirming that these isoforms have

no long-range function (Figure 3.14D-E). However, when endoderm expressing

sf-PDGF-A was combined with mesoderm, cells moved directionally (Figure

3.14C). This argues that the sf-PDGF-A behaves as a long range signal which

instructively determines the orientation of PCM cells, and that a localized source

of sf-PDGF-A is sufficient to act as a directional cue.

  168

Figure 3.10: Over-expression of sf-PDGF-A in the marginal zone disrupts PCM cell orientation

(A,B) Scanning electron micrographs of prechordal mesoderm (A) and leading

edge mesendoderm (B) of sagittaly fractured mid-gastrula stage (NF stage 11)

embryos expressing sf-PDGF-A in the dorsal marginal zone, white arrowheads

highlight cells oriented parallel to the BCR, (C,D) Rose diagrams show the

percentage orientation of prechordal mesoderm cells (C) and leading edge

mesendoderm cells (D) from embryos expressing sf-PDGF-A marginal, rose

diagrams use the angle at which protrusions are extended from the cell body with

respect to the dorsal BCR as a measure of orientation. 0 degrees, dorsal BCR;

90 degrees, animal pole. “n” is number of cells scored from 8 embryos analyzed.

Orange, leading edge mesendoderm; Yellow, prechordal mesoderm; purple,

chordamesoderm. BCR, blastocoel roof; EN, endoderm; BC, blastocoel.

  169

Figure 3.10: Over-expression of sf-PDGF-A in the marginal zone disrupts PCM cell orientation

  170

Figure 3.11: The effect of over-expression of sf-PDGF-A in the marginal zone can be rescued

(A,B) Scanning electron micrographs of prechordal mesoderm from sagittaly

fractured mid-gastrula stage (NF stage 11) embryos expressing sf-PDGF-A in the

dorsal marginal zone (A) or embryos co-injected with PDGF-A morpholino and sf-

PDGF-A RNA in the marginal zone (B), red arrowheads highlight cells oriented

parallel to the BCR (A), white arrowheads highlight cells oriented perpendicular to

the BCR (B). (C,D) Rose diagrams show the percentage orientation of prechordal

mesoderm cells from embryos expressing sf-PDGF-A in the marginal zone (C) or

after rescue by co-injection with PDGF-A morpholino (D), rose diagrams use the

angle at which protrusions are extended from the cell body with respect to the

dorsal BCR as a measure of orientation. 0 degrees, dorsal BCR; 90 degrees,

animal pole. “n” is number of cells scored from 8 embryos analyzed. Yellow,

prechordal mesoderm; purple, chordamesoderm. BCR, blastocoel roof; EN,

endoderm; BC, blastocoel.

  171

Figure 3.11: The effect of over-expression of sf-PDGF-A in the marginal zone can be rescued

  172

Figure 3.12: PDGF-A MO inhibits PCM radial intercalation in vitro

(A) Explant system. (1.) Slice of mesoderm and adjacent endoderm is combined

with (2.) explanted control or PDGF-A MO injected BCR under a coverslip and

(3.) filmed to track mesoderm cell movements. Blue, endoderm; green, ectoderm;

orange, mesoderm; red arrow shows direction of cell intercalation; A, anterior; P,

posterior; D, dorsal; V, ventral. (B-B’) Frames from timelapse recording. (B) Cell

cluster at 0 min. (B’) Same cluster after 30min, cells have intercalated at the

mesoderm/BCR boundary (white dotted line). (C-D) Cell tracks from explants

combined with uninjected BCR (C) or BCR injected with 60ng PDGF-A MO (D).

  173

Figure 3.12: PDGF-A MO inhibits PCM radial intercalation in vitro

  174

Figure 3.13: sf-PDGF-A is required for directional migration of deep PCM cells

(A,C) Velocity plots, average velocities of 50 individual cells collected from five of

each of the following combinations (A) BCR injected with 60 ng PDGF-A

MO/uninjected mesoderm, (C) BCR co-injected with 800 pg of morpholino

resistant sf-PDGF-A mRNA and 60 ng PDGF-A morpholino/uninjected

mesoderm. Positive x-axis toward BCR. (B-B’’; D-D’’) Tracks of individual cells in

mesodermal explant of (A) and (C) respectively.

  175

Figure 3.13: sf-PDGF-A is required for directional migration of deep PCM cells

  176

Figure 3.14: sf-PDGF-A is an instructive cue for PCM directional migration

(A, C-E) Velocity plots, velocities of 50 individual PCM cells from 5 of each of the

following explant combinations: (A) uninjected endoderm/mesoderm, (C)

endoderm injected with 400 pg sf-PDGF RNA/uninjected mesoderm, (D)

endoderm injected with 400 pg intPDGF-A RNA/uninjected mesoderm and (E)

endoderm injected with 400 pg lfPDGF-A RNA/uninjected mesoderm. Positive x-

axis is toward the endoderm. (B-B’’) Tracks of PCM cells from uninjected

endoderm/uninjected mesoderm combination.

  177

Figure 3.14: sf-PDGF-A is an instructive cue for PCM directional migration

  178

3.3: Discussion  By analyzing cell rearrangement in the prechordal mesoderm of the Xenopus

gastrula, I have identified a mechanism for radial intercalation that depends on

long-range PDGF-A signalling. Furthermore, I have identified different functions,

long-range and contact-dependent signalling, for short and long/intermediate

splice variants of PDGF-A, respectively.

3.3.1: A mechanism for radial cell intercalation in the PCM In the PCM of the Xenopus gastrula, a sf-PDGF-A-mediated long range signal

reorients the protrusions of unipolar cells from an orientation parallel to the BCR

to one that is perpendicular to the BCR. The oriented protrusions are in contact

with the cell bodies of adjacent cells. This suggests that cells use the surface of

their neighbours and wedge between each other in an instance of intercellular

migration, leading to radial intercalation and the observed gradual thinning of the

PCM. The ability of the cells to migrate across adjacent cells towards a sf-PDGF-

A source can be directly demonstrated in vitro.

The PCM is separated from the ectodermal BCR by Brachet’s cleft, a tissue

boundary maintained by cycles of attachment and repulsion between ectoderm

and mesoderm cells (Wacker et al. 2000; Rohani et al. 2011). The resulting

dynamic adhesion between germ layers ensures that while the most superficial

PCM cells can contact the BCR, they cannot penetrate into it despite being

attracted by sf-PDGF-A. Lateral mobility at the BCR/mesoderm interface is

  179

essential for cells to efficiently wedge between each other and to intercalate at

the boundary. The sf-PDGF-A signal potentially reaches about 5 cell lengths

deep, but is only recognized by the PCM cells and not by adjacent endodermal

cells. Therefore, given sufficient time, all PCM cells, and only these, will

eventually arrive at the BCR to form a single layer. In summary, PCM thinning

consists of a chemotactic reorientation of unipolar cells by an exogenous tissue,

and intercellular migration, leading to a re-arrangement of the multi-layered tissue

and the formation of a monolayer adjacent to the ectoderm.

During gastrulation in Drosophila, radial intercalation of deep mesoderm cells

depends on signaling by the FGF receptor Heartless and its ligands Pyramus and

Thisbe, which are all required for normal protrusion formation and intercalation

(McMahon et al. 2010; McMahon et al. 2008; Klingeisen et al. 2009; Murray and

Saint, 2007). Dorsal migration of the more superficial mesoderm cells on the

ecotoderm is not affected in heartless mutants (McMahon et al. 2010; Murray and

Saint, 2007). In our system, different isoforms of PDGF-A regulate both the

intercalation of deep mesodermal cells and the directional migration of the more

superficial LEM cells on the BCR (see above and Nagel et al. 2004).

3.3.2: Distinct roles for long and short PDGF-A splice isoforms in cell orientation: contact-dependent and long-range signaling Presence or absence of the positively charged cell retention motif determines

PDGF-A behaviour, i.e. whether the molecule remains bound to the cell surface

  180

or extracellular matrix, or is able to diffuse over larger distances (Andersson et al.

1994; Raines and Ross et al. 1992). In this way, PDGF-A molecules suited either

for long-range or for contact-dependent signaling are generated by differential

splicing (Andrae et al. 2008; Mercola et al. 1988). In Xenopus, previous (Nagel et

al. 2004) and my current results indicate that the long/intermediate and short

isoforms indeed serve distinct functions in cell orientation. The sf-PDGF-A

controls orientation and attracts PCM cells to the BCR over a distance of several

cell lengths. The lf-PDGF-A and int-PDGF-A, on the other hand, associate with

the extracellular matrix of the BCR by binding to proteoglycans or fibronectin

(Smith et al. 2009; Nagel et al. 2004), restricting long range movement of the

molecules. With PDGFR-α being expressed in adjacent mesoderm tissue, this

creates a contact-dependent PDGF signaling mechanism where only mesoderm

cells directly in contact with the BCR receive the signal. Consistent with this, the

lf-PDGF-A (and likely int-PDGF-A given sequence similarity) has been linked to

substrate-dependent guidance cues that direct the migration of the leading edge

mesendoderm on the BCR surface (Nagel et al. 2004). An interesting question

will be how matrix-bound and diffusible PDGF signals are integrated as they

simultaneously impact cells directly at the ectoderm-mesoderm boundary.

As reviewed in Chapter One, a number of signaling pathways are activated

downstream of PDGFRα, some of which are known to be involved in the

regulation of directional cell migration (Heldin et al. 1998; Ronnstrand and Heldin,

  181

2001; Hoch and Soriano, 2003). Interestingly, it has been shown that PDGF-AB

heterodimers and PDGF-AA homodimers activate distinct pathways downstream

of PDGFRα (Ekman et al. 1999). This shows that activation of receptors is not an

all or nothing event and that the activation of individual downstream signaling

pathways can be ligand dependent. Thus, it is possible that homodimers of sf-

PDGF-A, lf-PDGF-A and int-PDGF-A could activate distinct downstream signaling

pathways. Thus, the long range signaling mechanism regulating PCM radial

intercalation and the short range contact dependent signaling mechanism

regulating LEM directional migration could function through distinct signaling

pathways. Cells that are in contact with the BCR receive lf/int-PDGF-A and sf-

PDGF-A signals, thus if signalling downstream of these ligands involves different

signalling pathways, this could explain how LEM cells are able respond to the

short range contact based signal even though they are still exposed to long range

sf-PDGF-A signal.

The downstream signalling pathways involved in short and long range PDGF-A

signalling should be investigated and this could be studied using the approach

used by Van Stry et al. 2005. This approach used a panel of mutant

PDGFRα molecules where single key tyrosine residues or combinations of these

residues were mutated to phenyalanine. These receptors were defective for

individual or groups of signaling pathways. Furthermore, these receptors were

designed so that they could be chemically activated when needed, independently

  182

of the endogenous PDGF-A ligand, which allows for the effect of the receptors on

gastrulation processes to be studied without affecting earlier stages of

development. Over-activation of PDGF signaling in the PCM or the LEM disrupts

cell orientation in these regions (see Section 3.2.3, Damm and Winklbauer, 2011;

Nagel et al. 2004), thus it would be expected that the activation of receptors that

contain intact tyrosine residues for the required signaling pathway(s) would result

in an overexpression phenotype. Focus could be directed to mutant receptors

that when expressed, fail to result in an overexpression phenotype. This effect

would indicate that the missing pathway(s) are required for cells to respond to the

PDGF-A directional signal. Thus, by running through the panel of mutant

receptors that contain different complements of functional signaling pathways

and screening embryos for the PDGF signaling overexpression phenotype, the

required signaling pathways regulating PCM radial intercalation and LEM

directional migration could be identified.

Based on the behaviour of cells in mesoderm explants, the orienting signal from

the BCR spreads about 5 cell layers deep into the mesoderm. Since the

mesoderm does not express PDGF, a relay mechanism involving signal

propagation by PDGF-stimulated PDGF release is excluded. Furthermore, since

lf-PDGF-A and int-PDGF-A are ineffective in BCR rescue experiments and in

endoderm gain-of-function assays, it is unlikely that a PDGF-A induced relay

mechanism is functioning as the orienting cue. Instead, sf-PDGF-A may

  183

effectively diffuse into the mesoderm and form a classical chemoattractant

gradient. Unfortunately, the initial speading of the signal could not be determined

from the migratory response in explants, since irregular movements due to

explant relaxation after explantation were superimposed over directed migration.

After 30 minutes, approximately 5 cell rows were moving directionally,

corresponding to a signal range of about 200 µm. The boundary of this oriented

region advanced little during the next hours. This would be consistent with a

steady state gradient being formed in about an hour, but also with a continuously

expanding gradient whose spreading velocity decreases exponentially.

Compared to these length and time scales, the migration velocity of cells towards

the source, about 20 µm/hour, is small, and will probably not affect the gradient

significantly. (For more on PDGF-A gradients, refer to Chapter Four)

Roles for PDGF-A and PDGFR-α have been identified during early development

of zebrafish, chick and mouse (Yang et al. 2008; Orr-Urtreger and Lonai, 1992;

Soriano, 1997; Montero et al. 2003), but it is not known which isoforms of PDGF-

A are involved, and whether boundary effects or long-range signals are

employed. Also, PDGF-A has functions other than cell orientation in the embryo.

It prevents apoptosis in the mesoderm of Xenopus (Van Stry et al. 2004; Van

Stry et al. 2005), and regulates N-cadherin expression in the chick (Yang et al.

2008). In zebrafish, PDGF-A is part of a signaling pathway involving PI3K and

  184

PKB, which stimulates the formation of PCM cell protrusions during gastrulation

(Montero et al. 2003).

3.3.3: Patterns of cell orientation and radial intercalation in Xenopus gastrulation

Both in PCM and LEM, orientation is influenced by sf-PDGF-A. If signalling is

diminished, PCM cells assume an orientation characteristic of the LEM, and

conversely, when sf-PDGF-A is overexpressed, LEM cells become radially

oriented, similar to the PCM. A lower sensitivity towards the PDGF signal in the

LEM, consistent with the reduced expression of PDGFR-α in this region (Figure

3.2A,B and Ataliotis et al. 1995), could explain the difference between regions. In

the LEM, the matrix-bound lf-PDGF-A orients the lamellipodia of the cells in

contact with the BCR substratum, towards the animal pole (Figure 2.9A; Nagel et

al. 2004).

Several observations suggest that the dorsal pattern of cell orientation continues

laterally and ventrally. Bidirectional orientation extends laterally into the SM,

which takes part in convergent extension. Unidirectional radial orientation and,

anterior to it at the leading edge of the mesoderm mantle, oblique orientation is

seen in lateral and ventral mesoderm (Figure 2.13 and 2.14, Ibrahim and

Winklbauer, 2001). We propose that the whole mesoderm can be subdivided into

a PDGF-insensitive region, characterized by the expression of XBra and a

  185

PDGF-sensitive, unidirectional population located anteriorly to the insensitive

region. The former would comprise the converging and extending mesoderm,

and the latter both the leading edge mesendoderm with its oblique orientation,

and an intermediate region where cells point towards the BCR.

The endodermal vegetal cell mass is not responsive to the sf-PDGF-A long-range

signal. However, cells express PDGFR-α, and endodermal explants migrate

directionally on BCR-conditioned substratum (Figure Winklbauer and Nagel,

1991), consistent with an ability to recognize lf-PDGF-A. Vegetal cells are animal-

vegetally elongated, i.e. in parallel and not perpendicular to the BCR, fitting to the

vegetal rotation movement in which these cells are engaged (Winklbauer and

Schurfeld, 1999). In the absence of a sf-PDGF-A signal, PCM cells take on a

similar orientation. This suggests that in mesoderm and vegetal endoderm, the

basic cell orientation is along the animal/vegetal axis, however the molecular

regulators of this orientation have not been identified. Thus, as mesoderm cells

respond to PDGF-A, they are reoriented towards the BCR, but they fall back into

the animal/vegetal orientation if PDGF-A is not detected.

3.4: Materials and Methods   3.4.1: Embryos and microinjections  See Chapter 2, Section 2.4.1.

  186

3.4.2: Explants  See Chapter 2, Section 2.4.6

3.4.3: Scanning Electron Microscopy  See Chapter 2, Section 2.4.3

3.4.4: mRNA Isolation/RT-PCR  PCM, endoderm and animal caps were dissected at stages 11 and 12. RNA was

purified using TriZol extraction (Life Technologies) and cDNA was synthesized

according to the protocol for the Superscript III reverse transcriptase (Life

Technologies). The following primers were used for PCR reactions: PDGF-A

(long/short form) FWD 5’-GGAATGCACGTGTACAGCAA-3’ and REV 5’-

CGGGAATGTAACATGGCGTA-3’; PDGFR-α FWD 5’-CTCGCAAATGCCACTACAGA-3’ and

REV 5’-CCACAAGGTGTCATTGTTGC-3’; ODC FWD 5’-GTCAATGATGGAGTGTATGGATC-3’

and REV 5’-TCCATTCCGCTCTCCTGAGCAC-3’; Gsc FWD 5’-TGTGGAGCAGTTCAAGCTCT-

3’ and REV 5’-ATCTGGGTACTTGGTTTCTT-3’; XBra FWD 5’-GGATCGTTATCACCTCTG-

3’ and REV 5’-GTGTAGTCTGTAGCAGCA-3’. PCR reactions were assembled

according to the protocol for the Platinum Taq Polymerase (Life Technologies).

Reactions for PDGF-A/PDGFR-α and ODC were run for 30 and 23 cycles

respectively at an annealing temperature of 60°C, reactions for Gsc and XBra for

25 cycles at an annealing temperature of 55°C.

3.4.5: Constructs, Morpholinos and mRNA Synthesis  

  187

RNA was synthesized using mMessage mMachine in vitro transcription kits (Life

Technologies). Plasmids containing lfPDGF-A (pGHE2), sfPDGF-A (pCS2+),

PDGFR-α (pGHE2), PDGFR-37 (pGHE2) and dnPDGF-A 1308 (pGEM) were

prepared for transcription as in Nagel et al. (2004). PDGF-A morpholino (5’

AGAATCCAAGCCCAGATCCTCATTG - 3’) was used as described (Nagel et al. 2004).

Morpholino resistant variants of sf/lf-PDGF-A were generated using the

QuickChange II site directed mutagenesis kit (Stratagene). Five bases (A, G, C,

T, T) within the morpholino hybridization region were changed to (C, A, T, C, A)

respectively. The sense and antisense primers used were (5’-

GCAGCAGGACGCAATGCGAATTTGGGCCTGGATACTGCTGCTAAGCGTCG-3’) and (5’

AGACGCTTAGCAGCAGTATCCAGGCCCAAATTCGCATTGCGTCCTGCTGC-3’) respectively. The

amino acid composition of the PDGF-A protein was preserved.

3.4.6: In-Situ Hybridization and In-Situ Probe Synthesis See Chapter 2, Section 2.4.2.

3.4.7: Statistical Analysis  Statistical analysis was performed using GraphPad inStat 3. The Mann-Whitney

statistical test at the 95% confidence level was used for all statistical calculations.

  188

Chapter Four: The short splice isoform of PDGF-A forms a chemoattractant gradient by diffusion of

molecules through the extracellular space

  189

4.1: Introduction    Chemoattractant and morphogen gradients play essential roles in development.

The wing imaginal discs of Drosophila larvae are patterned by a gradient of

Decapentaplegic, a TGF-β factor that spreads from the compartment boundary

separating the anterior and posterior halves of the disc (Basler and Struhl, 1994;

Entchev et al. 2000; Teleman and Cohen, 2000; Kicheva et al. 2007; Zhou et al.

2012). In Xenopus and zebrafish early development, a hypothesized gradient of

bone morphogenetic proteins (BMPs) produced on the future ventral side of the

embryo is thought to pattern the dorsal/ventral axis of the embryo (Suzuki et al.

1994; Dosch et al. 1997; Kishimoto et al. 1997; Dick et al. 2000; Plouchinec and

DeRobertis, 2011). A chemoattractant gradient of SDF1-α instructively regulates

the directional migration of primordial germ cells during early Zebrafish

development (Doitsidou et al. 2002; Boldajipour et al. 2008). Thus, signalling

molecules from several families have been hypothesized to form gradients. While

experimental evidence has suggested roles for morphogen or chemoattractant

gradients in many systems, the existence of many of these gradients remains

hypothetical, as they have not been directly observed. Several recent studies

have used fluorescently tagged proteins in order to visualize gradients and to

study the kinetics of gradient formation (Kicheva et al. 2007; Yu et al. 2009; Zhou

et al. 2012).

  190

In principle, in order for a stable gradient to form there must be a localized source

of molecules and a sink for the molecules. The sink is usually in the form of

protein degradation, either by extracellular degradation or intracellular

degradation after receptor mediated endocytosis (Kruse et al. 2004; Kitcheva et

al. 2007; Boldajipour 2008; Yu et al. 2009; Zhou et al. 2012). Alternatively,

binding and inactivation of molecules by extracellular receptors or extracellular

antagonists could also function as a sink (Piccolo et al. 1996; Blitz et al. 2000;

Little et al. 2006; Umulis et al. 2009). In order to form a gradient, molecules must

spread from the source through the tissue. The mechanism of spreading seems

to be dependent on the molecule involved. Hypothetically, simple free diffusion of

molecules through cells with a localized sink that is distant from the source can

form a linear gradient (Crick, 1970). Alternatively, free diffusion of molecules

through extracellular spaces with a uniform sink in the tissue and a constant rate

of degradation forms a gradient described by an exponential decay function

(Wartlick et al. 2009). The Dpp morphogen gradient in the Drosophila wing disc

and the Fgf-8 morphogen gradient in the Zebrafish embryo are examples of

gradients that form by free diffusion (Zhou et al. 2012, Yu et al. 2009). Other

potential mechanisms of spreading include the “bucket brigade”, in which

molecules are passed from receptor to receptor on the cell surface, and

transcytosis where molecules internalized by endocytosis are subsequently

secreted and passed on to the next cell (Kerzberg and Wolpert 1998, Strigini,

2005).

  191

PDGFs have been described as both chemoattractants and morphogens (Andrae

et al. 2008). These observations imply that PDGF molecules form a

concentration gradient in tissues. The PCM of the Xenopus embryo undergoes

radial intercalation over the course of gastrulation (see Chapter Two; Damm and

Winklbauer, 2011). Radial intercalation of these cells requires directional

intercellular migration that is instructively regulated by sf-PDGF-A, which is

produced by the overlying BCR (see Chapter Three; Ataliotis et al. 1995; Damm

and Winklbauer, 2011). In vitro observations using explants suggests that the

influence of sf-PDGF-A is stable to a range of approximately 200 µm (see

Chapter Three; Damm and Winklbauer, 2011). Furthermore, sf-PDGF-A dimers

are similar in size to Dpp and Fgf dimers, which form concentration gradients in

tissues (Zhou et al. 2012). Thus, it is possible that sf-PDGF-A, which lacks the C-

terminal cell retention motif domain, could form a chemoattractant gradient in

PCM tissue (Figure 3.1; Raines and Ross 1992; Kelly et al. 1993; Andersson et

al. 1994; Damm and Winklbauer, 2011).

In this chapter, using PDGF-A-eGFP fusion constructs, I show that sf-PDGF-A

forms a gradient in PCM tissue and that these molecules spread through the

tissue by moving through intercellular spaces. Furthermore, I show that lf-PDGF-

A and int-PDGF-A, which contain a C-terminal cell retention motif, do not form a

  192

gradient and remain restricted at the mesoderm/BCR boundary. Lastly, my

results suggest that sf-PDGF-A spreads through the tissue by diffusion.

4.2: Results  4.2.1: Short and long PDGF-A isoforms have different extracellular localizations Several reports have shown that GFP tagged proteins can be used to visualize

concentration gradients of molecules (Kicheva et al. 2007; Yu et al. 2009; Zhou

et al. 2012). Morpholino resistant variants of sf-PDGF-A, lf-PDGF-A and int-

PDGF-A were tagged with eGFP downstream of the N-terminal pro-peptide

cleavage site (Figure 4.1A). Since eGFP is a relatively large molecule and I was

concerned that fusing it with PDGF-A may lead to a non-functional protein, sf-

PDGF-A and lf-PDGF-A constructs were generated with three short Myc epitope

tags downstream of the pro-peptide cleavage site as a contingency (Figure 4.1A).

Because tagging proteins with fluorophores can disrupt protein function, I tested

the ability of the sf-PDGF-A-eGFP to rescue the PDGF-A morphant phenotype.

When a morpholino targeting PDGF-A is injected into BCR cells, PCM cells fail to

orient their protrusions toward the BCR and instead orient their protrusions

between 0o and 90o with respect to the BCR (Figure 4.2A, C), an effect that is

fully rescued by co-expressing sf-PDGF-A with the PDGF-A morpholino (see

chapter three, Figure 3.8A, B). Similar to sf-PDGF-A, the sf-PDGF-A-eGFP

  193

construct was able to rescue the PDGF-A morphant phenotype (P-value <

0.0001) showing that fusion with eGFP did not disrupt function (Figure 4.2B, D).

The Xenopus embryo is opaque during early development; therefore it is not

possible to see the expression of tagged constructs inside the embryo. I made

use of the explant system designed to film intercellular migration of PCM cells

(see chapter two, Figure 2.13A; 4.1B). Since directional migration of PCM cells

can be measured in this explant, and I have shown that sf-PDGF-A instructively

regulates this migration, any putative PDGF-A gradient must be formed in vitro in

this tissue as well as in vivo. Therefore, this explant is a suitable system to study

PDGF-A gradients. Ectoderm (BCR) was injected with sf-PDGF-A-eGFP, lf-

PDGF-A-eGFP or int-PDGF-A-eGFP mRNA and combined with the posterior end

of a mesoderm explant isolated from an un-injected embryo (Figure 4.1B).

Assembled explants were observed with a laser scanning confocal microscope.

PDGF-A is normally expressed by ectoderm cells, therefore I analyzed the

localization of the PDGF-A-eGFP constructs in the ectoderm of

ectoderm/mesoderm explant combinations. In whole embryos, the ectoderm has

large triangular/rectangular intercellular gaps and narrower gaps between parallel

membranes of tightly associated cells (Figure 4.3A,B). Similar gaps are observed

in explanted ectoderm tissue. These gaps open and close as cells break and

  194

Figure 4.1: Visualizing PDGF molecules (A) Schematic diagrams of PDGF-A-eGFP and PDGF-A-myc constructs showing

the location of the eGFP and Myc tags downstream of the pro-peptide domain.

Light blue, leader sequence; orange, PDGF growth factor domain; red, C-terminal

retention motif; light green, propeptide domain; purple, hydrophobic domain;

brown, short form specific C-terminus; green; eGFP tag; blue, 3XMyc tag. (B)

Explant system. (1.) Mesoderm and ectoderm (expressing PDGF-A-eGFP

constructs) are explanted and (2.) combined with the BCR placed at the

posterior, PCM-containing end of mesoderm explant. Explant combinations (3.)

are filmed immediately with a laser scanning confocal system and cell velocities

measured.

  195

Figure 4.1: Visualizing PDGF molecules

  196

re-form contacts over time (Figure 4.3D-L; 4.4A-I). Sf-PDGF-A-eGFP was

detected intracellularly and in existing and newly formed large extracellular gaps

(Figure 4.3E-F, H-I, K-L). Some sf-PDGF-A-eGFP was also co-localized with the

adjacent membranes between cells, which suggests that the molecules could be

accumulating in spaces between cell membranes (Figure 4.3B, F, I;

mbRFP/eGFP co-localization between cells). These observations are consistent

with sf-PDGF-A being secreted but not remaining associated with the cell

surface. Interestingly, int-PDGF-A-eGFP, which contains the C-terminal cell

retention motif, was found to co-localize with the exposed cell membrane of the

large extracellular gaps and with the membranes of adjacent cells (Figure 4.4B-

E, E-F, H-I) and was not found within intercellular gaps. For unknown reasons, lf-

PDGF-A-eGFP failed to show any fluorescent signal. Therefore, I analyzed the

localization of lf-PDGF-A-myc in ectoderm. As with int-PDGF-A-eGFP, lf-PDGF-

A-myc was found to co-localize with the cell membrane (Figure 4.5A-C), however

membrane localization appeared to be stronger. Lf-PDGF-A-myc was observed

in fixed tissue compared with the live imaging of int-PDGF-A-eGFP. The

increased localization of lf-PDGF-A-myc at the membrane could be a fixation

artefact if the accumulation of lf-PDGF-A extracellularly is a highly dynamic

process. Alternatively, the difference in membrane localization could be an

artefact caused by the fusion of the molecules with the eGFP or Myc tags.

Finally, this observation could be the result of real differences in behaviour

between the long and intermediate PDGF isoforms. Regardless of the cause,

  197

Figure 4.2: sf-PDGF-A-eGFP can rescue the PDGF-A morphant phenotype (A-B) Scanning electron micrographs of sagittaly fractured mid-gastrula stage

(NF stage 11) embryos injected with PDGF-A morpholino (A) or co-injection of

PDGF-A morpholino and morpholino resistant sf-PDGF-A-eGFP (B); white

arrowheads in (A) highlight cells oriented parallel to the BCR; white arrowheads

in (B) highlight cells oriented perpendicular to the BCR. (C,D) Rose diagrams

showing the percentage orientation of prechordal mesoderm cells from embryos

injected with PDGF-A morpholino (C) or co-injected with PDGF-A morpholino and

morpholino resistant sf-PDGF-A-eGFP (D); 0 degrees, dorsal BCR; 90 degrees,

animal pole. “n” is number of cells scored from 10 embryos analyzed for each

condition. BCR, blastocoel roof; EN, endoderm.

  198

Figure 4.2: sf-PDGF-A-eGFP can rescue the PDGF-A morphant phenotype

  199

both the intermediate and long PDGF isoforms were membrane localized, which

was expected since the C-terminal cell retention sequence has been shown to be

important for keeping PDGF-A at the surface of the secreting cell (Raines and

Ross, 1992). However, I was unable to determine if, as expected, the tagged

proteins were indeed associated with the extracellular or intracellular side of the

cell membrane.

4.2.2: sf-PDGF-A is found distant from its source, in intercellular spaces between PCM cells

BCR expressing sf-PDGF-A-eGFP was combined with uninjected mesoderm

(Figure 4.1B) and immediately imaged with the confocal microscope. Unlike

ectoderm, mesoderm in the intact embryo has more intercellular spaces due to

sites of cell-cell contact being small, as seen with TEM (Figure 4.3C). In the

embryo, these gaps are potential sites of PDGF-A accumulation. In explants, the

large gaps between mesoderm cells could function similarly. Immediately after

the mesoderm and BCR explants were combined, sf-PDGF-A-eGFP was clearly

visible in the ectoderm, however mesoderm explants were completely clear of

GFP signal (Figure 4.6A-D). When explants were imaged over an hour later, sf-

PDGF-A-eGFP was found in the mesoderm intercellular spaces (Figure 4.6E-H).

Interestingly, there was no sf-PDGF-A-eGFP seen intracellularly in the

mesoderm explant (Figure 4.6E-H). Receptor mediated endocytosis of PDGF-A

has been shown to be important in the regulation of chemotaxis (Kawada et al.

  200

2009; Avrov et al. 2003), thus in order to determine if intracellular sf-PDGF-A-

eGFP might be visible in the mesoderm explant below the surface level filmed in

timelapses, Z-stack images of the explants were generated (Figure 4.7). It was

possible to image explants to a depth of approximately 11 µm. PCM mesoderm

cells range from 15-25 µm in height (Selchow and Winklbauer, 1997), thus the

deepest plane imaged should be approximately halfway through the height of the

cells. Sf-PDGF-A-eGFP was only found extracellularly at all levels of the z-stack

(Figure 4.7A-D). This unexpected result argues against a role for receptor-

mediated endocytosis in this system. Together, these results suggested that sf-

PDGF-A-eGFP molecules had travelled from their source in the BCR explant to

deep in the mesoderm explant through intercellular spaces.

Close examination of mesoderm intercellular spaces revealed that sf-PDGF-A-

eGFP appeared to be punctate (Figure 4.3H,K; 4.8A). Interestingly, when images

of mesoderm intercellular spaces were analyzed, fluorescent structures ranging

in length from 0.5 µm to greater than 1 µm could be seen interspersed by dark

areas (Figure 4.8B,C). Furthermore, eGFP signal of these structures was

obviously more intense in extracellular gaps that were closer to the BCR explant

(PDGF-A source) (Figure 4.8B,C). These observations suggest that extracellular

matrix macromolecules could be creating channels or pores in the extracellular

gaps. Sf-PDGF-A-eGFP could travel as individual dimers or small aggregates

through these channels as it moves through the mesoderm tissue. The decrease

  201

Figure 4.3: sf-PDGF-A-eGFP is localized to ectoderm intercellular spaces

(A-C) Transmission electron micrographs of large ectoderm intercellular spaces,

black arrowhead in (A), close ectoderm cell contacts, between black arrowheads

in (B) and large gaps between mesoderm cells, between white arrowheads and

indicated by asterisk in (C). (D-L) Confocal optical sections of the ectoderm piece

of ectoderm/mesoderm explant combinations, ectoderm expressed membrane

bound RFP (D,G,J) to label cells and sf-PDGF-A-eGFP (E,H,K); Merge images

(F,I,L); white arrowheads indicates accumulation of eGFP signal in the

intercellular spaces at t = 0 (D), t = 3 minutes (E) and t = 6 minutes (F). Scale bar

in merge images (F,I,L) = 50µm.

  202

Figure 4.3: sf-PDGF-A-eGFP is localized to ectoderm intercellular spaces

  203

Figure 4.4: int-PDGF-A-eGFP is localized to ectoderm cell membranes

(A-I) Confocal optical sections of an ectoderm explant expressing membrane

bound RFP (A,D,G) to visualize the cells and int-PDGF-A-eGFP (B,E,H), white

arrow heads show the absence of eGFP signal in intercellular spaces at t = 0 (A-

C), t = 3 minutes (D-F) and t = 6 minutes (G-I). Localization of int-PDGF-A-eGFP

at the cell membrane can be seen in Merge images (C,F,I). Scale bar in merge

images (C,F,I) = 50µm.

  204

Figure 4.4: int-PDGF-A-eGFP is localized to ectoderm cell membranes

  205

in average fluorescence intensity of these structures at increasing distances from

the BCR explant (Figure 4.8B,C) supports the idea that these structures could

represent channels but not large aggregates. Fluorescence intensity is an

indicator of eGFP concentration and the concentration of sf-PDGF-A-eGFP

moving through channels is expected to decrease with increasing distance from

the sf-PDGF-A-eGFP source (BCR explant). It is unlikely that there would be a

decrease in the fluorescence intensity of individual structures with increasing

distance from the BCR explant if these structures represented aggregates of sf-

PDGF-eGFP molecules. Dense extracellular matrix has indeed been observed in

the intercellular gaps of the gastrula (Johnson, 1977ab). The path followed by sf-

PDGF-A-eGFP molecules as they move through the extracellular spaces could

be increased as a result of having to move through channels or around

obstacles, which would reduce the effective rate and range of sf-PDGF-A

spreading in mesoderm tissue.

In order to determine if PDGF-A isoforms containing the C-terminal retention

motif move into the mesoderm like sf-PDGF-A molecules, ectoderm expressing

int-PDGF-A-eGFP was combined with mesoderm. When the explant combination

was imaged immediately after being combined, int-PDGF-A-eGFP was strongly

visible in the ectoderm explant, however no GFP signal was present in

mesoderm intercellular spaces (Figure 4.9A-D). When analyzed later, unlike with

sf-PDGF-A-eGFP, there was still no GFP signal present in the mesoderm explant

  206

Figure 4.5: lf-PDGF-A-myc localization is similar to that of int-PDGF-A-eGFP (A-C) Confocal optical section of an ectoderm explant expressing membrane

bound RFP to visualize the cells and lf-PDGF-A-myc. Sections stained with anti-

myc antibody and FITC conjugated secondary antibody. (A) mbRFP, (B) lf-

PDGF-A-myc, (C) merge image, white arrow heads (B,C) show the localization of

lf-PDGF-A-myc at cell membranes. Scale bar in merge image (C) = 50µm.

  207

Figure 4.5: lf-PDGF-A-myc localization is similar to that of int-PDGF-A-eGFP

  208

Figure 4.6: sf-PDGF-A-eGFP forms a visible gradient in mesoderm tissue

(A-H) Confocal optical sections of ectoderm/mesoderm explant combinations,

ectoderm expressed membrane bound RFP (A,E) to label cells and sf-PDGF-A-

eGFP (B,F); DIC channel (C,G); Merge images (D,H); (A-D) Images taken

immediately after explant assembly; (E-H) images taken after 120 minutes, white

arrowheads (F,H) indicates accumulation of eGFP signal in the intercellular

spaces with visually decreasing fluorescence intensity moving away from the

ectoderm explant/mesoderm interface. Scale bar in merge images (D,H) = 50µm.

  209

Figure 4.6: sf-PDGF-A-eGFP forms a visible gradient in mesoderm tissue

  210

(Figure 4.9E-H). Instead, int-PDGF-A-eGFP signal was strongly present at the

mesoderm-ectoderm boundary, co-localizing with the labelled ectoderm cell

membrane (Figure 4.9H). These results fit with explant and whole embryo

observations that suggest that sf-PDGF-A functions as a long range molecule

while PDGF-A isoforms containing the C-terminal cell retention motif appear to

have short range functions only (see Chapter Three, Damm and Winklbauer,

2011).

4.2.3: sf-PDGF-A forms an extracellular gradient that fits a single exponential decay function

Several lines of indirect evidence have suggested that sf-PDGF-A is forming a

concentration gradient in the mesoderm tissue (see Chapter Three; Figure 4.6).

The presence of sf-PDGF-A-eGFP in mesoderm explants at a distance from the

BCR source is in support of this hypothesis. The spreading of molecules from a

localized source in a non-directional fashion with linear degradation is described

by equation (A1) (see section 4.5). Once a steady state has been reached, this

can be characterized by the exponential decay function:

(1)

𝐶 𝑥 = 𝐶!𝑒!!!

where 𝐶! is the concentration at the source boundary, 𝑥 (µm) is the distance from

the source boundary, and λ (µm) is the decay length, the distance from the

  211

Figure 4.7: sf-PDGF-A-eGFP is only observed in intercellular gaps

(A-D) Confocal optical sections of an ectoderm/mesoderm explant combination z-

stack, ectoderm expressed membrane bound RFP to label cells and sf-PDGF-A-

eGFP; z. position of 0 is the surface of the explant (A); white arrowheads

highlight intercellular spaces that show eGFP signal at increasing depth in the

explants (B,C,D). Scale bar in merge images (A-D) = 50µm.

  212

Figure 4.7: sf-PDGF-A-eGFP is only observed in intercellular gaps

  213

source where the concentration of the molecule has decayed to a factor of 1 𝑒 of

the concentration at the source 𝐶!. Equation (1) shows that 𝐶(𝑥) depends on the

distance from the source, the decay length 𝜆 and the concentration at the source

boundary 𝐶! . Thus, the shape of the exponential function representing a

molecular gradient is determined by 𝜆 and 𝐶! . These parameters in turn are

determined by the degradation rate 𝑘 (s-1) and the diffusion coefficient 𝐷 (µm2s-1)

in the source (ectoderm) and receiving (mesoderm) tissues. In the mesoderm, 𝜆,

𝑘 and 𝐷 are related by:

(2)

𝜆 =𝐷𝑘

(Equations (1) and (2) are adapted from Kicheva et al. 2007)

For the purposes of this study, a relative 𝐶! in arbitrary fluorescence units can be

determined from the average fluorescence intensity of eGFP labelled constructs

at the source boundary. This is possible because the Beer-Lambert law shows

that fluorescence intensity and concentration are proportionally related

(Lakowicz, 2006) providing that saturation of the pixels in the image has not

occurred.

  214

After ectoderm expressing sf-PDGF-A-eGFP and uninjected mesoderm are

combined, a gradient of sf-PDGF-A-eGFP can be seen extending from the

ectoderm explant into the mesoderm explant in confocal images (Figure 4.6). To

determine if the sf-PDGF-A-eGFP distribution is an exponential concentration

gradient, mRNA encoding sf-PDGF-A-eGFP was injected at a concentration of

400pg/blastomere. This concentration of sf-PDGF-A mRNA was used previously

to rescue PDGF-A morphant embryos and thus can mimic endogenous sf-PDGF-

A concentrations (See Chapter Three; Figure 4.2B). Ectoderm expressing sf-

PDGF-A-eGFP was then combined with mesoderm from an uninjected embryo

and z-stack images of the explants were captured after 2 hours. These images

were divided into 13 stripes parallel to the mesoderm-BCR boundary that were

20µm in width, the first of which corresponded to the first 20µm of mesoderm

adjacent to the mesoderm/ectoderm boundary. The average GFP fluorescence

intensity of the intercellular spaces was determined in each stripe and the relative

average 𝐶! was determined by measuring the fluorescence intensity at the

source boundary (i.e. in Brachet’s Cleft) (Figure 4.10A).

Although the general trend was a decay in GFP intensity with increasing distance

from the mesoderm-BCR boundary (Figure 4.10B), fluctuation in fluorescence

intensity occurred from region to region, sometimes increasing slightly when

moving to regions further away from the BCR before dropping back down in

regions even further from the BCR (Figure 4.10B). Similar fluctuations were

  215

Figure 4.8: sf-PDGF-A-eGFP in intercellular spaces

(A) High magnification (1000X), high resolution (1024x1024) image of a

mesoderm-BCR explant showing two intercellular gaps in the mesoderm explant,

(B-C) Digitally magnified images of the intercellular gaps from (A). (B) is the gap

furthest from the BCR explant, (C) is the gap closer to the BCR explant, white

arrowheads highlight sf-PDGF-A-eGFP signal, white lines trace the length of

individual eGFP structures.

  216

Figure 4.8: sf-PDGF-A-eGFP in intercellular spaces

  217

Figure 4.9: int-PDGF-A-eGFP remains localized to the mesoderm/ectoderm boundary (A-H) Confocal optical sections of ectoderm/mesoderm explant combinations,

ectoderm expressed membrane bound RFP (A,E) to label cells and int-PDGF-A-

eGFP (B,F); DIC channel (C,G); Merge images (D,H); (A-D) Images taken

immediately after explant assembly; (E-H) images taken after 120 minutes, white

arrowheads (F,H) indicates accumulation of eGFP signal at the

mesoderm/ectoderm boundary. Scale bar in merge images (D,H) = 50µm.

  218

Figure 4.9: int-PDGF-A-eGFP remains localized to the mesoderm/ectoderm boundary

  219

observed when Dpp-GFP fluorescence intensity was measured in the Drosophila

wing imaginal disc (Kicheva et al. 2007). When averaged over all explants,

however, the sf-PDGF-A-eGFP distribution in mesoderm tissue fit a single

exponential function very well (R2 = 0.95, Figure 4.10B-D). The average decay

length λ was determined from the exponential function plot (Figure 4.10C). Thus,

the average decay length λ was 80 +/- 12 µm. In explants, approximately 4 to 5

rows of PCM cells migrate directionally toward a sf-PDGF-A source (see Chapter

Two, Figure 2.14F, G, H). PCM cells are approximately 40 to 50 µm in length

(Selchow and Winklbauer, 1997). Thus, cells are able to respond to the gradient

up to an approximate distance of 2λ−2.5λ (160-200 µm). This suggests that

mesoderm cells can read the PDGF-A signal to levels of approximately 20% of

the concentration at the mesoderm-ectoderm interface.

In order to ensure that pixels were not saturated in the sf-PDGF-A-eGFP source

where the concentration is highest, which would affect the measurements of

relative average 𝐶! value and thus the shape of the gradient, sf-PDGF-A-eGFP

was also expressed at a high dose (approximately 2.5X higher than the dose

used to rescue PDGF-A morphant embryos). BCR explanted from embryos

injected with 1ng/blastomere of mRNA encoding sf-PDGF-A-eGFP was used.

High concentration explants were imaged under the same conditions and with the

same microscope settings as the lower concentration explants. The average

intracellular fluorescence intensity of the sf-PDGF-A-eGFP source cells in the

  220

Figure 4.10: The sf-PDGF-A gradient fits a single exponentional function

(A) Explanation of the method used to measure average fluorescence intensity in

extracellular gaps and at the source boundary, red boxes indicate examples of

areas where measurements were made from maximum projection z-stack

images, red line represents the approximately 14 pixel distance over which

fluorescence intensity was measured. (B-D) Average eGFP intensity measured

from 20µm wide mesoderm regions to a distance of 260µm from the ectoderm

explant. Measurements were made from confocal maximum intensity projections

of individual mesoderm/ectoderm explant combinations. Ectoderm was co-

injected with 600pg of membrane bound RFP RNA and 1.6ng of sf-PDGF-A-

eGFP RNA (coloured lines, 6 in total) (B), black trace (B,C) is the average GFP

intensity in each 20µm wide mesoderm region of the 6 explants measured, red

line is the curve of best fit for the black trace. The distance from the source

corresponding to the average decay length (λ) is indicated by the red dotted line

(B,C). Error bars show standard error of the mean. (D) Linear regression analysis

of average fluorescent intensity data from the black trace in (B,C), trace is nearly

linear (R2 = 0.95).

  221

Figure 4.10: The sf-PDGF-A gradient fits a single exponentional function

  222

Figure 4.11: There is no change in average decay length (λ) by increasing sf-PDGF-A expression (A) Bar graph comparison of fluorescence intensity at the mesoderm-BCR

boundary and inside ectoderm cells for low and high mRNA conditions. (B,C)

Average eGFP intensity measured from 20µm wide mesoderm regions to a

distance of 260µm from the ectoderm explant. Measurements were made from

confocal maximum intensity projections of individual mesoderm/ectoderm explant

combinations. Ectoderm was co-injected with 600pg of membrane bound RFP

RNA and 4ng of sf-PDGF-A-eGFP RNA (coloured lines, 6 in total) (B), black trace

(B,C) is the average GFP intensity from each mesoderm region of the 6 explants

measured, red line is the curve of best fit for the black trace. The distance from

the source corresponding to the average decay length (λ) is indicated by the red

dotted line (B,C). Error bars show standard error of the mean. (D) Linear

regression analysis of average fluorescent intensity data from the black trace in

(B,C), best fit is close to linear (R2 = 0.84). (E) Comparison of decay lengths of

explants expressing low and high concentration of sf-PDGF-A-eGFP. Error bars

show standard error of the mean. Asterisks indicate statistically significant

results.

  223

Figure 4.11: There is no change in average decay length (λ) by increasing sf-PDGF-A expression

  224

ectoderm was significantly higher (P-value = 0.03) in ectoderm injected with 4ng

sf-PDGF-A-eGFP mRNA compared with lower concentrations of mRNA (Figure

4.11A), although this increase was not proportional to the mRNA dose. The lack

of a proportional increase could be the result of several possibilities. For

example, it is unknown how much of the injected mRNA is actually translated;

very high concentrations of mRNA could overpower the available protein

translation machinery of the cells resulting in a lower than expected protein

concentration. Additionally, translational and post-translational regulation of sf-

PDGF-A-eGFP mRNA and protein could affect protein levels. Fluorescence

intensities were measured using the Zeiss Axiovison 4 microscopy software

package. This software allows pixel intensities to vary between one and 255 for

an 8 bit image. Even with the increased fluorescence intensity, both the average

intracellular (in ectoderm cells) and source boundary (mesoderm-BCR interface)

sf-PDGF-A-eGFP intensities were below the 255 intensity level in the very high

mRNA concentration explants (Figure 4.11A). This result suggests that saturation

of pixels in the image was not an issue and should not affect the shape of the

gradient.

For unknown reasons, fluctuations in GFP intensity were more severe between

regions in mesoderm explants at high concentration of sf-PDGF-A-eGFP,

compared with ectoderm expressing a lower concentration (compare Figures

4.10B and 4.11B). However, similar to the lower concentration situation, on

  225

average the data fit a single exponential function reasonably well (R2 = 0.84,

Figure 4.11B-D). The decay length value was determined to be λ = 111 +/- 19

µm, which is not significantly different from the lower sf-PDGF-A-eGFP mRNA

concentration experiment (P-value = 0.329) (Figure 4.12E). Thus, the higher

source concentration of sf-PDGF-A-eGFP did not increase the decay length (λ)

(Figure 4.11D). This result shows that the relationship described in equation (1) is

valid in this system.

The C-terminal cell retention motif found in the longer PDGF-A isoforms

apparently restricts the PDGF-A-eGFP signal to the mesoderm/ectoderm

boundary (Figure 4.9E-H). In order to quantitatively analyze the dynamics of

PDGF-A isoforms that contain a C-terminal retention motif, 400pg/blastomere of

int-PDGF-A-eGFP mRNA was injected into embryos. Immediately after

combining the mesoderm and ectoderm explants, GFP signal was clearly visible

in ectoderm cells however, no GFP signal appeared in the mesoderm

intercellular spaces over the two hours of filming (Figure 4.12A,B). Thus, PDGF-

A isoforms containing the C-terminal cell retention do not form a gradient and

their signalling functions must be restricted to the region surrounding the

secreting cells.

The relationships described in equations (1) and (2) only describe a gradient at

steady state; therefore equations (1) and (2) can only be applied to spreading of

  226

PDGF-A after the gradient reaches a steady state. The sf-PDGF-A-eGFP signal

continued to spread until approximately one hour after the mesoderm and BCR

explants were combined (Figure 4.13A-D). After that time point, the maximum

distance from the BCR where a GFP signal was visible did not change (Figure

4.13D). This suggests that a steady state gradient was established in about an

hour. To quantify this observation, the GFP fluorescence intensity was measured

in intercellular gaps in a region corresponding to the average decay length λ

(approximately 80µm from the mesoderm/BCR interface). Once a steady state

gradient is established, the GFP fluorescence intensity at this position should

stop increasing and become constant, although some fluctuations in intensity

may be expected. The time point at which GFP fluorescence intensity began to

increase at λ was explant specific, beginning between 20 and 40 minutes after

the start of filming. Fluorescence intensity continued to increase for

approximately another 40 minutes before levelling off in the vicinity of the 70

minute mark (Figure 4.13E). The initial 0 to 40 minute delay in the appearance of

fluorescence in the mesoderm explant is likely due to the time required for the

mesoderm and ectoderm explants to heal and form the mesoderm-ecotoderm

interface (approx. 30 minutes, see Chapter Two). These results indicate that the

sf-PDGF-A-eGFP gradient could reach a steady state within approximately 50

minutes of the first appearance of GFP signal at λ. Interestingly, cells in

mesoderm explants begin moving toward the BCR explant between 30 and 60

  227

Figure 4.12: int-PDGF-A-eGFP does not form a gradient (A,B) Average eGFP intensity in mesoderm regions measured from confocal

maximum intensity projections of individual mesoderm/ectoderm explant

combinations with ectoderm injected with 1.6ng of int-PDGF-A-eGFP RNA

(coloured lines, 6 in total) (A); black trace (A,B) plots the average eGFP intensity

in each mesoderm region of the 6 explants measured. Error bars show the

standard error of the mean.

  228

Figure 4.12: int-PDGF-A-eGFP does not form a gradient

  229

minutes after the assembly of the explants (see Chapter Two, Figure 2.13F, G,

H), thus mesoderm cell movements correlate with the timeline for the formation of

sf-PDGF-A steady state gradient in mesoderm tissue.

The spaces between mesoderm cells are filled with an extracellular matrix

material that includes glycosaminoglycans (Johnson, 1977ab). Since PDGF-A is

found in intercellular spaces, it must be contained within this material.

Interestingly, as cells in mesoderm explants migrated toward the BCR explant, sf-

PDGF-A-eGFP containing extracellular matrix could be seen moving relative to

the cells (Figure 4.14A-D). This is presumably due to forces exerted on the matrix

by the migrating cells. In some cases, gaps fused to make a new gap combining

the PDGF-A containing material from the original gaps (Figure 4.14D). These

observations may explain the fluctuations seen in the GFP intensity

measurements. As described above, measurements are made in stripes

spanning 20 µm in width. If cell movements push material into and out of

measurement zones, it could appear as fluctuations in intensity measurements.

The contribution of these convection processes on the spreading of sf-PDGF-A-

eGFP through mesoderm tissue remains to be analyzed in detail.

  230

Figure 4.13: sf-PDGF-A-eGFP forms a steady state gradient

(A-D) Confocal optical sections from a mesoderm/ectoderm combination

timelapse. (A) Start of timelapse recording, (B) t = 20 minutes, white box and

arrowhead show visually detectable GFP signal at a position corresponding to

the average decay length (λ). Red measurements show the range of the visible

gradient after 70 minutes (C) and 110 minutes (D). White arrow in (C,D) indicates

a sf-PDGF-A-eGFP filled gap at a position corresponding to the average decay

length (λ). Scale bars correspond to 50µm. (E) Plot of the change in average

eGFP fluorescence intensity over the course of the timelapse at position λ as

measured from confocal maximum projections for 5 mesoderm/ectoderm

combinations (coloured lines). Black trace (E) plots the average eGFP intensity in

each mesoderm region from the 5 explants measured over the course of the

timelapse. Red trace (E) plots the average eGFP intensity in ectoderm explant

(PDGF-A-eGFP source) from the 5 explants measured over the course of the

timelapse. Error bars show the standard error of the mean.

  231

Figure 4.13: sf-PDGF-A-eGFP forms a steady state gradient

  232

Figure 4.14: sf-PDGF-A-eGFP in extracellular spaces moves relative to the cells as they migrate (A-D) Confocal optical sections of ectoderm/mesoderm explant combinations,

ectoderm expressed membrane bound RFP to label cells and sf-PDGF-A-eGFP;

white arrowheads follow the merging of two intercellular gaps over six minutes

(A-D). Scale bars = 50µm.

  233

Figure 4.14: sf-PDGF-A-eGFP in extracellular spaces moves relative to cells as cells migrate

  234

4.2.4: The effective diffusion coefficient 𝑫𝒆𝒇𝒇 is consistent with formation of a sf-PDFG-A-eGFP gradient by diffusion through extracellular spaces

The decay length λ, degradation rate 𝑘 and the diffusion coefficient 𝐷 kinetic

parameters provide information about the mechanism of spreading of molecules

in a tissue. The diffusion coefficient 𝐷, which describes the rate of molecule

spreading in a fluid medium as a result of random diffusion, is a property that is

characteristic of the diffusing molecule in a specific medium. However, the

diffusion of a molecule through a tissue is subject to interactions with extracellular

matrix molecules, binding proteins, receptors and factors that can increase the

diffusion path length for the molecule. These interactions can be strongly

reflected in diffusion coefficients determined for molecules when using

techniques like FRAP because they operate over long distance and time scales

(Kitcheva et al. 2012). However, these interactions can also be reflected, to a

lesser degree, in diffusion coefficients determined using techniques that operate

over very short length and time scales, such as FCS (see Chapter 1, section

1.3.1). Thus, diffusion coefficients of molecules in tissues determined using these

techniques are referred to as effective diffusion coefficients 𝐷!"! ,  as they pertain

to the overall rate of gradient formation. The magnitude of the effective diffusion

coefficient 𝐷!"" of a molecule can be indicative of the mechanism of spreading

for that molecule when compared to the diffusion coefficient 𝐷 of that molecule in

water. For example, very low effective diffusion coefficients (0.1 – 0.5 µm2s-1) are

  235

suggestive of active transport mechanisms involving endocytosis or the transport

of molecules along a cell membrane while higher effective diffusion coefficients

(on the order of magnitude of 20 to 86 µm2s-1) could suggest free extracellular

diffusion of molecules (Kicheva et al. 2012; Kicheva et al. 2007; Yu et al. 2009).

In order to determine how sf-PDGF-A-eGFP is transported in mesoderm tissue,

these parameters were calculated. λ,  𝑘 and 𝐷!"" are related by equation (2).

Thus, if two of these parameters are known, the third can be calculated. The

average decay length λ was determined above (Figure 4.10B). In order to

calculate the degradation rate 𝑘, I used the method that was used to analyze the

kinetics of the Dpp gradient of the Drosophila wing imaginal disc (Kicheva et al.

2007; Zhou et al. 2012). The curve generated by measuring the increase in GFP

fluorescence intensity over time at λ in mesoderm-BCR explant combinations

(Figure 4.13E) is analogous to recovery curves generated by FRAP experiments.

In most of the systems where this technique is used, the molecules of interest

have already spread through the area to be analyzed. Therefore, in order to

derive transport rate information about the molecules of interest, an area of pre-

defined size must be photobleached in order to generate a recovery curve. The

degradation rate can then be determined by fitting the experimentally generated

recovery curve to a theoretical recovery curve generated by a mathematical

model of the experimental situation. The mesoderm-BCR explant system that I

have used provides the unique opportunity to observe a molecular gradient as it

forms. The formation of the gradient in mesoderm tissue will occur with the same

  236

kinetics as if the molecules were moving into a photobleached area of

mesoderm. Thus, the method used to derive the degradation rate 𝑘 from FRAP

experiments can be adapted to determine this parameter for sf-PDGF-A-eGFP

from explant experiments without the need for photobleaching. The calculations

and the method used to determine the degradation rate 𝑘 and the effective

diffusion coefficient 𝐷!"" of sf-PDGF-A-eGFP are described in detail in Section

4.5.

The formation of the sf-PDGF-A-eGFP gradient can be modelled as one-

dimensional non-directional transport of molecules with uniform linear

degradation. This is represented by the de-dimensionalized time dependent

solution of the model represented by equation (A1) in Section 4.5:

(4)

𝐶(𝑋, 𝜏) =12𝐶!𝑒

!! 2− Erfc −𝑋 − 2𝜏2 𝜏

− 𝑒!!Erfc𝑋 + 2𝜏2 𝜏

where 𝐶(𝑋, 𝜏)   is the concentration of sf-PDGF-A-eGFP at the relative

dimensionless position  𝑋 = !!  for the corresponding dimensionless time  𝜏 = 𝑘𝑡.

With 𝐶! = !!!

and at 𝑥 =  𝜆 i.e. 𝑋 = 1, equation (4) can be further simplified to (see

Section 4.5 for the full explanation of the simplification):

(5)

𝐶′ =12𝑒 1− 𝑒! + Erf

1− 2𝜏2 𝜏

+ 𝑒!Erf1+ 2𝜏2 𝜏

  237

In order to calculate 𝑘, a theoretical fluorescence “recovery” curve was generated

by plotting the theoretical 𝐶!!! (𝜏) values over the corresponding τ values. 𝐶!!! (𝜏)

was calculated from equation (5) for τ ranging from 0.1 to 10 (see Table A1). The

theoretical recovery curve is shown in Figure A1A, Section 4.5.

The experimental recovery curve (see Figure 4.13E) is fitted to the theoretical

curve by determining the experimental τ, 𝜏!! , that corresponds to the

experimental 𝐶!!!(𝜏) , the values of which were determined by dividing the

average fluorescence intensity at λ at times between 30 and 80 minutes after

mesoderm-BCR explant assembly by 𝐶! (see Table A4, Section 4.5). The

corresponding values for 𝜏!! were determined from the theoretical recovery curve

and are presented in Table A4. The relationship between 𝜏!!, time 𝑡 (s-1) and the

degradation constant 𝑘 (s-1) is by definition:

𝜏 = 𝑘𝑡

Thus, in order to determine a value for the degradation constant 𝑘, the 𝜏!! values

were plotted over the corresponding values for 𝑡 (see Table A4). The resulting

curve (Figure A1B) correlated well with a linear function (R2 = 0.97). The

degradation constant 𝑘 was determined by calculating the slope of this curve:

𝑘 = 1.07×10!! s-1

Since the degradation constant 𝑘 is related to the diffusion coefficient 𝐷 through

equation (2) and the average decay length λ = 80 µm, the effective diffusion

coefficient 𝐷!"" for sf-PDGF-A-eGFP can be calculated as:

  238

𝐷!"" = (80!) 1.07×10!! = 6.8 µm2s-1

Dpp in the Drosophila wing imaginal disc and Fgf8 in the Zebrafish gastrula have

diffusion coefficients of 21 µm2s-1 and 53 µm2s-1 respectively (Zhou et al. 2012;

Yu et al. 2009). Both of these gradients are formed by free diffusion of molecules

through extracellular spaces. Thus, since the effective diffusion coefficient 𝐷!"" of

sf-PDGF-A-eGFP is on the same order of magnitude as the diffusion coefficients

of Dpp and Fgf8, the sf-PDGF-A gradient may form by a similar extracellular

diffusion mechanism (see section 4.3.3 of the Discussion).

4.3: Discussion  In this chapter, I show that sf-PDGF-A is found in intercellular spaces in the

ectoderm and mesoderm and forms a concentration gradient in mesoderm.

Furthermore, I have determined the decay length λ, degradation rate and the

effective diffusion coefficient for the gradient and these parameters are consistent

with formation of the sf-PDGF-A gradient by diffusion of molecules through

intercellular spaces. Interestingly the gradient forms over a range that coincides

with the number of cell rows that migrate directionally toward the BCR in vitro. I

have also confirmed that PDGF-A isoforms containing the C-terminal cell

retention motif do not form a gradient in mesoderm tissue but rather, remain

associated with the ectoderm cell surface and do not cross the

mesoderm/ectoderm boundary.

  239

4.3.1: The cell retention motif determines whether a PDGF-A isoform can form a concentration gradient

In Xenopus, PDGF-A exists as three isoforms. These isoforms differ in their C-

termini where the long and intermediate sized isoforms contain a stretch of

positively charged amino acids known to retain PDGF molecules at the cell

surface (Raines and Ross, 1992; Kelly et al. 1993; Andersson et al, 1994). The

short isoform is lacking this domain. I have shown here that sf-PDGF-A can form

a concentration gradient in mesoderm tissues and thus can act as a long range

signalling molecule. Previously, I had shown that sf-PDGF-A plays an instructive

role in regulating the directional migration and radial intercalation of PCM cells

during gastrulation (see Chapter Three). Therefore, sf-PDGF-A apparently

regulates this process by providing directional information through a

concentration gradient that extends from the BCR source through several

mesoderm cell layers. In contrast, neither int-PDGF-A nor lf-PDGF-A have been

found to regulate PCM cell radial intercalation (see Chapter Three).

The leading edge mesendoderm migrates toward the animal pole of the embryo

during gastrulation, directed by short range contact based PDGF-A signalling

(Winklbauer and Nagel, 1991; Nagel et al. 2004). In this case, lf-PDGF-A is

thought to instructively regulate this process and a gradient of matrix associated

PDGF-A on the BCR has been suggested because LEM explants continue to

migrate directionally in vitro on extracellular matrix transferred from the inner

  240

BCR (Winklbauer and Nagel, 1991; Nagel and Winklbauer, 1999; Nagel et al.

2004). Here, I have shown directly for the first time, that PDGF-A isoforms

containing the C-terminal retention motif (int-PDGF-A-eGFP) remain associated

with and are restricted to the BCR/mesoderm interface. Thus, PDGF-A isoforms

that contain the C-terminal retention motif are not capable of long range

signalling but may require direct contact between the BCR and the adjacent

migrating cells to initiate signalling. Although a cell-associated gradient of int-

PDGF-A-eGFP increasing in a vegetal to animal direction cannot be seen on the

BCR because of uniform over-expression of int-PDGF-A-eGFP mRNA in

ectoderm cells, the accumulation of int-PDGF-A at the mesoderm-ectoderm

boundary is consistent with a role in regulating LEM directional migration.

4.3.2: The sf-PDGF-A gradient likely forms by diffusion of molecules through intercellular spaces Several models have been proposed to explain the formation of concentration

gradients in tissues. The presence of sf-PDGF-A-eGFP in intercellular spaces

indicates that PDGF-A molecules travel through these spaces when forming the

gradient. This makes sense since the gaps between cells are the largest spaces

for molecules to move into without obstruction. In contrast, no GFP signal was

found intracellularly (Figure 4.7A-D). This is unlike Dpp in the Drosophila wing

imaginal disc or Fgf8 in the zebrafish embryo where in both cases, vesicles

containing labelled molecules were found intracellularly (Kitcheva et al. 2007; Yu

et al. 2009; Zhou et al. 2012). The lack of intracellular GFP signal argues against

  241

the idea that the cells themselves traffic PDGF-A. Thus a transcytosis

mechanism, which relies on the endocytosis of molecules and subsequent

secretion of those molecules in order for the signal to be passed on to the next

row of cells, is not likely being employed here.

Another potential mechanism involves the “handing” of molecules from a receptor

on one cell to a receptor on an adjacent cell. This is an active transport

mechanism that has been modelled to describe the formation of TGF-β family

ligand gradients (Kerszberg and Wolpert, 1998). It is unlikely that this “bucket-

brigade” mechanism plays a role in PDGF-A gradient formation because, as it is

modelled, the mechanism depends on two different receptors with low and high

affinities for the ligand molecule (Kerszberg and Wolpert, 1998). PDGF-A binds

strongly to PDGFR-α homodimers and PDGFR-α is the only known receptor to

bind PDGF-A (Andrae et al. 2008). Futhermore, the “bucket brigade” mechanism

depends on continuous receptor binding and therefore ligand molecules would be

strongly associated with cell membranes. In mesoderm explants, GFP signal

does not outline cells, showing that sf-PDGF-A-eGFP is not heavily localized with

the cell membranes. Thus, although binding of sf-PDGF-A to PDGFRα must be

occurring to activate the downstream signalling pathways that regulate directional

cell migration, it is unlikely that there is enough ligand/receptor binding occurring

to spread the amount of sf-PDGF-A-eGFP seen in the mesoderm intercellular

spaces.

  242

It was previously hypothesized that diffusion of molecules through cells can occur

quickly enough to form concentration gradients over distances of approximately

one millimetre (Crick, 1970). Recent studies have implicated the free diffusion of

protein molecules through intercellular spaces as a mechanism for the formation

of gradients of Dpp in the Drosophila wing imaginal disc and Fgf8 in the

Zebrafish embryo (Zhou et al. 2012, Yu et al. 2009). Transport of GFP molecules

in water and in the extracellular space of zebrafish gastrulae occurs by free

diffusion and the effective diffusion coefficients determined for these processes

using FCS analysis are 95 µm2s-1 and 87 µm2s-1 respectively (Petrasek et al.

2008, Yu et al. 2009). Dpp and Fgf8 gradients are also thought to form by free

diffusion because the effective diffusion coefficients as measured by FCS (21

µm2s-1 and 53 µm2s-1 respectively) of these two molecules fall into the same

order of magnitude as the diffusion coefficient of freely diffusing GFP molecules.

Furthermore, the effective diffusion coefficients of these molecules reveal that

transport occurs too rapidly to involve the endocytosis and re-secretion

processes involved in a transcytosis mechanism (Zhou et al. 2012; Yu et al.

2009). Interestingly, in both of the above cases, FCS recorded a subset of Dpp

and Fgf8 molecules with significantly lower effective diffusion coefficients. This

likely reflects reversible and non-reversible interactions between Dpp or Fgf8

molecules and receptors or extracellular matrix molecules (Zhou et al. 2012 and

Yu et al. 2009). Indeed, it was shown that reducing the amount of HSPGs in

extracellular spaces reduced the percentage of Fgf8 molecules with the lower

  243

effective diffusion coefficient (Yu et al. 2009). However, since the large majority

of molecules exhibit the higher transport rate (>60% for Dpp and >90% for Fgf8;

Zhou et al. 2012, Yu et al. 2009), gradient formation is dominated by the rapidly

diffusing molecules.

FCS analyzes transport of molecules over relatively small length scales (500

nm3) and time scales (milliseconds), thus the transport rate information derived

from these experiments approximates the transport rate of individual molecules

well, since interactions between the molecule being transported and other factors

are more limited in the smaller volume than over longer length scales (Kicheva et

al. 2012). Tissue level analysis techniques, such as FRAP, however operate over

much larger length (tens to hundreds of micrometres) and time scales (minutes to

hours) and thus reversible and non-reversible binding events between the

transported molecule and other factors and extracellular degradation will be

reflected in kinetic information from these experiments (Kicheva et al. 2012).

Thus, effective diffusion coefficients derived from these experiments can reflect

the overall kinetics of gradient formation, rather than the transport of individual

molecules. The effective diffusion coefficient for the sf-PDGF-A-eGFP gradient (7

µm2s-1) was determined using a tissue level technique. I propose that sf-PDGF-A

is likely transported through mesoderm extracellular spaces by diffusion based

on the following evidence: The effective diffusion coefficient of sf-PDGF-A-eGFP

is on the same order of magnitude as that for both Dpp and Fgf8. This effective

  244

diffusion coefficient is 70X faster than rates expected for a mechanism involving

endocytosis, such as trancytosis (Kicheva et al. 2009). The exclusive localization

of sf-PDGF-A-eGFP in extracellular spaces also argues against the involvement

of mesoderm cells in the trafficking of sf-PDGF-A. Furthermore, if a “bucket

brigade” type mechanism were involved here, the diffusion coefficient would be

expected to reflect the rate of lateral movement of PDGF receptors in the cell

membrane. These rates have not been determined for PDGF receptors, however

the rates for the smaller sized FGF receptors are known (𝐷 = 0.51 µm2s1 for

zebrafish Fgfr1; Ries et al. 2009). It can be expected that the diffusion coefficient

for PDGFRα in membranes would be similar or smaller. The diffusion coefficient

for the lateral mobility of a receptor tyrosine kinase in a cell membrane is at least

14X slower than the effective diffusion coefficient determined for sf-PDGF-A-

eGFP. Thus, it is likely that the sf-PDGF-A gradient forms by extracellular

diffusion. This question will be further addressed in the future by conducting an

FCS analysis using sf-PDGF-A-eGFP.

4.3.3: Potential factors influencing the effective diffusion coefficient

The effective diffusion coefficient of sf-PDGF-A-eGFP is significantly lower than

the diffusion coefficient of GFP diffusing in water (compare 7 µm2s-1 to 95 µm2s1)

(Petrasek and Schwille, 2008), therefore if sf-PDGF-A-eGFP is moving through

extracellular spaces by diffusion, free movement of the molecules is being

  245

hindered. Movement of molecules through the extracellular space can be

influenced by several factors, including the size of PDGF-A dimers, reversible

and non-reversible binding to cell surface receptors, extracellular matrix

molecules or binding proteins, and extracellular degradation. The potential effects

of these factors on the effective diffusion coefficient of sf-PDGF-A-eGFP will be

discussed below.

Diffusion coefficients are inversely proportional to the radius of the molecules

being transported (Crick, 1970). This is assuming that the diffusing molecules

form a sphere or behave as if they are spherical. Thus, large aggregates of sf-

PDGF-A-eGFP would affect the effective diffusion coefficient. For example, a

1000 fold increase in the molecular weight of sf-PDGF-A-eGFP because of

aggregation would result in a 10 fold decrease in the effective diffusion

coefficient. It is not known how many sf-PDGF-A molecules could be found in a

putative aggregate, however the formation of very large aggregates could

dramatically reduce the diffusion coefficient compared with that of non-

aggregated sf-PDGF-A dimers. It is possible that aggregation could be an

artefact caused by tagging of sf-PDGF-A with a GFP molecule, however I

consider this possibility to be unlikely. There is no indication in the FCS

determined kinetic parameters of either Dpp or Fgf8 that fusing these molecules

to GFP resulted in aggregation (Zhou et al. 2012; Yu et al. 2009). However, the

  246

endogenous formation of aggregates could be a biological mechanism used to

reduce the reach of PDGF-A gradient.

Alternatively, the structure of the extracellular matrix within intercellular gaps

could also explain the effective diffusion coefficient 𝐷!"" and the sub-localization

of sf-PDGF-A-eGFP in these gaps. The spaces between cells in the Xenopus

gastrula are known to be filled with an extracellular matrix that is composed of

proteoglycans (Johnson, 1977ab). Sf-PDGF-A molecules may need to navigate

through pores or channels between extracellular matrix macromolecule

structures as they diffuse through the extracellular spaces. This idea is supported

by the presence of sf-PDGF-A-eGFP containing structures within mesoderm

extracellular gaps (see Figure 4.8). The diffusion of molecules through narrower

irregularly shaped channels rather than straight through a wide extracellular gap

will result in an increase in diffusion path length and a reduction of the effective

diffusion coefficient compared with a non-porous medium like water. Tortuosity

(𝑇) is the increase in the diffusion path length due to geometry that results in a

1𝑇! fold decrease in the diffusion coefficient (Rusakov and Kullman, 1998).

Thus, tortuosity is related to the diffusion coefficient by the expression 𝐷!"" =

𝐷!"##(𝑇!!), where 𝐷!"" is the effective diffusion coefficient of molecules in the

extracellular space, 𝐷!"## is the diffusion coefficient in a non-porous medium like

water and 𝑇 is the tortuosity. The tortuosity factor of mesoderm extracellular gaps

is not known, however this factor has been determined for brain tissue based on

  247

experimental evidence and is approximately equal to 2.2 (Rusakov and Kullman,

1998). If this factor were applicable in mesoderm extracellular gaps, the effective

diffusion coefficient could be reduced by as much as 5 fold due to tortuosity.

Tortuosity factors may explain why FCS measurements typically result in higher

diffusion coefficients than tissue level analyses such as FRAP. Since FCS

measurements are made over very small length scales, the effect of interactions

between the diffusing molecules and extracellular macromolecules may not be

measured, resulting in an effective diffusion coefficient 𝐷!"" that can be closer to

the diffusion coefficient 𝐷 of the molecule in water. Hypothetically, if tortuosity

reduces the effective diffusion coefficient for sf-PDGF-A-eGFP 5 fold, reducing or

eliminating the tortuosity factor could result in an effective diffusion coefficient

that is as high as 𝐷 = 35 µm2s-1. This would place the sf-PDGF-A-eGFP diffusion

coefficient between the FCS determined diffusion coefficients of Dpp and Fgf8

(Zhou et al. 2012; Yu et al. 2009). Thus, tortuosity could be a major factor

influencing the effective diffusion coefficient determined from the FRAP-like

analysis carried out above.

Free diffusion may be a primary mechanism of growth factor spreading, however

these molecules are also subject to binding interactions with extracellular matrix

components and cell surface receptors as gradients form (Yan and Lin, 2009).

Similar to tortuosity, these interactions affect the rate of gradient formation by

modulating the effective diffusion coefficient of the molecule (Rusakov and

  248

Kullman, 1998). PDGF-A molecules are known interactants of heparin sulphate

proteoglycans (HSPGs) (Andersson et al. 1994, Smith et al. 2009). Specifically,

the longer isoforms (lf-PDGF-A and int-PDGF-A in Xenopus) that contain the C-

terminal cell retention motif may be anchored to cells through interactions with

HSPGs (Andersson et al. 1994; Smith et al. 2009). The effect of HSPG

interactions on the movement of sf-PDGF-A is unknown, however the effective

diffusion coefficient of diffusible molecules such as Fgf8 can be modulated by

interactions with HSPGs (Yu et al. 2009). Therefore, in addition to aggregation

and tortuosity, HSPGs can reduce the overall rate of gradient formation by

modulating the effective diffusion coefficient of molecules.

4.3.4: The regulation of molecule degradation during gradient formation

In order for a gradient to reach a steady state, the signalling molecules forming

the gradient must be subjected to degradation. Without degradation of molecules

over the path of signal spreading, molecules would continue to spread from the

source through the tissue until it becomes saturated, provided that the rate of

production at the source remains constant. This is observed when a secreted

GFP molecule, which lacks a sink, diffuses through tissues (Entchev et al. 2000).

Thus, having a sink for signalling molecules is important for the formation of a

stable gradient. Dynamin dependent receptor mediated endocytosis is a

commonly described sink and has been shown to play an important role in the

  249

formation of several signalling gradients such as Fgf, Dpp, SDF-1α (Zhou et al.

2012; Yu et al. 2009; Kicheva et al. 2007; Piddini et al. 2005; Boldajipour et al.

2008). Modulation of Dynamin activities in mutants or by using soluble inhibitors

and dominant negative constructs were shown to change the kinetics of the

gradient, including increases in the decay length λ (Kitcheva et al. 2007;

Boldajipour et al 2008; Yu et al. 2009). This shows that in those cases, receptor

mediated endocytosis plays a role in the degradation of the signal and has an

essential role in shaping the gradient. Interestingly, an orphan non-signalling

chemokine receptor was identified as a sink-generating molecule for SDF-

1α signalling in zebrafish primordial germ cell migration (Boldajipour et al. 2008).

Rather than CXCR4, the signalling receptor for SDF-1α undergoing ligand

induced endocytosis CXCR7, binds to SDF-1α and undergoes receptor mediated

endocytosis (Boldajipour et al. 2008). Thus, non-signalling receptors can bind to

ligands to facilitate degradation.

It is known that PDGF-A ligand binding to PDGFR-α can trigger internalization of

the ligand-receptor complex (Avrov et al. 2003; Kawada et al. 2009). This

receptor mediated endocytosis process is important for the regulation of

directional migration of some cell types (Kawada et al. 2009). Surprisingly,

although vesicles containing labelled molecules have been seen in Xenopus cells

previously (Hagemann et al. 2009), there were no visible sf-PDGF-A-eGFP in the

cells of mesoderm explants (Figure 4.7A-D). Thus, directional migration of

  250

mesoderm may not require internalization of the receptor-ligand complex for

signalling. The lack of intracellular GFP signal argues against the possibility that

a receptor other than PDGFR-α is involved in internalizing sf-PDGF-A for

degradation. However, a role for endocytosis in shaping the sf-PDGF-A-eGFP

gradient cannot be excluded and will be tested for by using cell-permeable

Dynamin inhibitors or dominant negative Dynamin constructs in the future. Thus,

the lack of intracellular sf-PDGF-A-eGFP suggests that degradation of the ligand

for the purpose of forming the concentration gradient may depend on an

alternative mechanism, such as degradation by extracellular proteases or ligand

sequestration by inhibiting molecules.

The degradation of sf-PDGF-A by proteases found in the intercellular gaps could

function as a sink and would be important in shaping the gradient. Some growth

factors are known to be degraded in the extracellular space, however it is

unknown if extracellular proteases interact with PDGF-A. Metalloproteinases are

metal ion dependent enzymes found in the extracellular spaces with known

functions in extracellular matrix remodelling, ectodomain shedding and cleavage

of growth factor molecules (Edwards et al. 2008). In particular, A Disintegrin and

Metalloproteanase (ADAMs) family members are known for their role in cleaving

growth factors, including several that interact with receptor tyrosine kinases such

as, ephrins, epidermal growth factor (EGF) and insulin-like growth factor (IGF)

(White et al. 2003; Edwards et al. 2008; Weber et al. 2012). In most cases, the

  251

ADAMs play an important role in activating the growth factors by removing them

from the surface of the secreting cell and allowing them to become soluble

(Prenzel et al. 1999; Loechel et al. 2000; White et al. 2003). Many ADAM family

members are transmembrane proteins. However, some members of the ADAM

family are soluble proteins suggesting that they may act at a distance from their

source (Edwards et al. 2008). Homologues for the transmembrane ADAMS;

ADAM13, ADAM19 and ADAM 41 have been identified in Xenopus (Neuner et al.

2009; Wei et al. 2010; Xu et al. 2012; Cousin et al. 2012). However, none of the

soluble ADAMs have been identified. Given the known interactions of ADAMs

with growth factors, it is possible that sf-PDGF-A interacts with a soluble ADAM

family member and is inactivated or degraded. It is also possible that soluble or

transmembrane proteases cleave the ectodomain of PDGFRα after binding of

PDGF-A in order to modulate signalling and degradation, similar to ectodomain

shedding. However, proteases would be expected to cleave specific recognition

sites in the PDGF molecule and may not cleave the GFP molecules that are used

experimentally to make the gradient visible. Thus, proteolytic degradation would

potentially release the fused GFP molecules. The released GFP would be

expected to behave like a secreted GFP molecule and diffuse through the tissue.

Thus, a gradient would not be visible. Since a gradient is observed, if

extracaellular degradation is involved in shaping the sf-PDGF-A-eGFP gradient, it

is not clear how the GFP molecule is being removed.

  252

BMPs are members of the TGFβ superfamily of signalling moleculed and are well

known for their involvement in determining ventral cell fates during the dorsal-

ventral patterning of vertebrate embryos (Suzuki et al. 1994; Furthauer et al.

1997; Khoka et al. 2005; Dosch et al. 1997; Stickney et al. 2007). It is

hypothesized that BMPs form a ventral to dorsal morphogen gradient, which is

antagonized by factors such as the soluble protein, Chordin, produced by cells on

the dorsal side of the embryo (Blitz et al. 2000; Hama and Weinstein, 2001;

Piccolo et al. 1996; Umulis et al. 2009). In this case, BMP inhibiting molecules

like Chordin help to shape the BMP gradient by binding to BMP morphogen

molecules and preventing their interaction with cell surface receptors or causing

their degradation, thus functioning as a sink (Umulis et al. 2009). An example of

an inhibitor molecule for PDGF-B is alpha-2-macroglobin (α2M) (Bonner, 2004;

Bonner, 1994). When PDGF-B interacts with an active form α2M, it ends up being

degraded though the LRP receptor (Bonner, 2004). Although, extracellular

inhibitors have not been identified for PDGF-A, it is possible that this type of

molecule could bind to PDGF-A molecules in the intercellular spaces and

inactivate them or prevent receptor binding. In this way, these inhibitors could

shape the gradient as a sink. However, proteolytic degradation after interaction

between PDGF-A and extracellular inhibitors could involve endocytosis, as in the

PDGF-B/α2M example or extracellular degradation and evidence of these

processes here is lacking. Thus the removal of sf-PDGF-A from our system is not

understood. Analysing the shape of the sf-PDGF-A gradient after disrupting

  253

endocytosis or the function of extracellular proteases may reveal roles for these

molecules. Additionally, a potential role for the adjacent pharyngeal endoderm

cells in the degradation of sf-PDGF-A will be investigated in the future.

4.4: Materials and Methods   4.4.1: Embryos and Microinjections  See Chapter Two, section 2.4.1.

4.4.2: eGFP and Myc tagged constructs  The method used to generate eGFP fusion constructs was adapted from

Marjoram and Wright, 2011: PCR was used to add AscI sites to the 5’ and 3’

ends of the eGFP taken from a membrane bound eGFP construct (gift from C.

Heisenberg); primers: FWD 5’-GGCGCGCCGGGGGGGGTGAGCAAGGGC-3’; REV 5’-

GGCGCGCCCCCCTTGTACAGCTC-3’. PCR was also used to add an AscI site and

additional sequence encoding glycine residues upstream of the pro-peptide

cleavage site to lf-PDGF-A (pGHE2), sf-PDGFA (pCS2+) (gifts from K. Symes)

and int-PDGF-A (pCS2+), PCR primers: FWD 5’- GGCGCGCCGCAGTGCCAGCC-3’;

REV 5’-GGCGCGCCCTCCTCAATGCT-3’. The primers used to amplify AscI

containing PDGF-A constructs also added EcoRI and BamHI restriction sites to

the 5’ and 3’ ends of the PDGF-A construct respectively, primers: FWD 5’-

TTGGGAGATTTGTCTGTAGAAAGG-3’ and REV 5’-TGATATCTGATGCCCTCCTCTG-3’.

AscI containing PDGF-A constructs were cloned into pGEM-EasyT using the

  254

pGEM-T Easy TA cloning kit (Promega). AscI-PDGF-A constructs and the AscI-

eGFP PCR product were digested with AscI (NEB). The digested eGFP molecule

was ligated into the digested PDGF-A constructs using a Quick Ligation Kit

(NEB). Ligated constructs were double digested with EcoRI/BamHI and the free

insert was ligated into pCS2+ using the Quick Ligation Kit (NEB).

lf-PDGF-A-myc and sf-PDGF-A-myc constructs were generated by PCR in a two

reaction process in which two Myc tags (2X Myc) were added downstream of the

pro-peptide cleavage site. lf-PDGF-A (pGHE2) and sfPDGFA (pCS2+) plasmids

were used as templates for the initial PCR reaction. Two species were generated

in separate reactions, one containing the 5’ half and added 2X Myc sequence

and one containing the complementary 3’ half and added 2X Myc sequence of

the PDGF-A constructs, PCR Primers:

FWD 5’-GAACAAAAACTTATTTCTGAAGAAGATCTGGAACAAAAACTTATTTCTGAAGA

AGATCTGGTTGAAGAAGCAGTCCCTGCTATC-3’,

REV 5’-CAGATCTTCTTCAGAAATAAGTTTTTGTTCCAGATCTTCTTCAGAAAT

AAGTTTTTGTTCACTTCTTTTTCTGCGACTGGG-3’

5’ and 3’ end primers were also added to amplify the two Myc containing halves

of the PDGF-A constructs, PCR Primers for the 5’ half FWD 5’-

GGATCCCCGTTACTGACTCC-3’, REV 5’-ATCTTCAGAAATAAGTTTTGTT-3’ and for

the 3’ half FWD 5’-GAACAAAACTTATTTCTGAAGAT-3’, REV 5’-

ATCCGAGTCCCAGATGATCC-3’. For the second reaction the species generated in

  255

the first reaction were combined with 5’ and 3’ end primers for the PDGF-A

reaction (see above, the primers used were identical to the primers used to

amplify the AscI containing PDGF-A constructs). The 2X Myc sequence of the 5’

and 3’ species from reaction one are complimentary, thus the PCR reaction

produces a complete PDGF-A molecule containing the 2X Myc sequence. The 5’

and 3’ end primers then amplify the full sized PDGF-A molecule. This reaction

generates unwanted products, therefore these were separated from the full

length PDGF-A molecules using gel electrophoresis and the full length molecules

were purified using a Gel Purification kit (Qiagen). Full length PDGF-A molecules

were cloned into pCS2+ using the Quick Ligation kit (NEB).

4.4.3: mRNA Synthesis  See Chapter 3, Section 3.4.5.

4.4.4: Explants For mesoderm-ectoderm combined explants, PCM explants from uninjected

embryos were combined with inner layer BCR explants expressing

600pg/embryo of mbRFP in addition to one of the following: 1.6ng/embryo or

4ng/embryo of sf-PDGF-A-eGFP or int-PDGF-A-eGFP or 1.6ng/embryo of lf-

PDGF-A-myc depending on the experiment. Combined explants were secured

under a strip of coverslip, with the mesoderm side that was originally in contact

with the BCR facing down, and the BCR explant positioned opposite to what

would have been the anterior side of the mesoderm in vivo (Figure 4.1). Explants

  256

were cultured in 1X MBS on 1% BSA treated glass bottom dishes during filming

(Mak Tek).

4.4.5. Scanning Electron Microscopy  See Chapter 2, Section 2.4.3.

4.4.6: Transmission Electron Microscopy  See Chapter 2, Section 2.4.4.

4.4.7: Antibody Staining  Embryos expressing membrane bound RFP and lf-PDGF-A-myc constructs were

fixed in 4% formaldehyde for one hour after which animal caps were dissected.

Animal caps were washed in 1X PBS three times for 10 minutes each before

blocking with 1% BSA solution for 1 hour. A 1:10 concentration of c-myc antibody

(9E10 from Hybridoma bank) was diluted in 1% BSA solution and applied to the

animal caps for 1 hour. Animal caps were subsequently washed with 1X PBS five

times for 10 minutes each. A FITC conjugated goat anti-mouse secondary

antibody was applied for 1 hour at a concentration of 1:100 diluted in 1% BSA

solution. Subsequently animal caps were washed with 1X PBS five times for 10

minutes each. Animal caps were secured under a coverslip in SlowFade Gold

solution (Invitrogen) before being imaged with the confocal microscope.

4.4.8: Confocal Microscopy  Confocal images and Z-stacks of explants were obtained using a Zeiss LSM510

laser scanning confocal system. The following settings were kept constant for all

  257

experiments involving the eGFP tagged constructs. eGFP was visualized by

excitation with a 488nm argon laser at 25% output, 30% excitation setting,

pinhole size: 701µm, detector gain: 956, Amplitude offset: -0.66, Amplitude gain:

-1. mbRFP was visualized by excitation with a 543nm helium/neon laser at the

100% excitation setting. The following confocal settings were used, pinhole size:

701µm, detector gain: 940, amplitude offset: -1.22, amplitude gain: 1. Images

were taken at a resolution of 512x512 unless otherwise stated with a scan speed

of 7.8 seconds, a 4.48µs pixel time and a line averaging of 4. All z-stacks were

made up of 8 optical sections with a stack thickness of approximately 20 µm.

The microscope system was controlled by Zeiss LSM5 software.

4.4.9: Image Processing  ImageJ was used to process z-stack optical sections into maximum projection

images using the built in z-project function.

  258

4.5: Appendix The formation of the sf-PDGF-A-eGFP gradient can be modeled as a process in

which transport of sf-PDGF-A occurs in a non-directional fashion with linear

degradation along the length of the gradient. The formation of the sf-PDGF-A

gradient in mesoderm-BCR explants can be modeled in one dimension; the

transport of molecules from a localized source (BCR explant) over a large

distance (approx. 200 µm) into the receiving tissue (mesoderm tissue). ImageJ

software was used to generate maximum projection images of the 8 optical

sections of mesoderm-ectoderm explant z-stacks, allowing explants to be treated

as a two-dimensional plane. Furthermore, because the only source of influx of

molecules into the mesoderm is the BCR explant and there is no influx of

molecules from the lateral sides, the formation of the gradient can be modeled in

one dimension. This type of simplification was used to model the formation of the

Dpp gradient in the Drosophila wing imaginal disc (Kicheva et al. 2009; Zhou et

al. 2012). The model used to describe this Dpp gradient was:

(A1, after Zhou et al. 2012)

𝛿𝛿𝑡 𝑐(𝑥, 𝑡) = 𝐷

𝛿!

𝛿𝑥! 𝑐(𝑥, 𝑡)− 𝑘𝑐(𝑥, 𝑡)

where 𝐷 (µm2s-1) is the diffusion coefficient and 𝑘 (s-1) is the degradation rate.

Both of these parameters are constants. The steady state solution for equation

A1 is:

  259

(A2, after Kicheva et al. 2007 and Zhou et al. 2012)

𝐶 𝑥 = 𝐶!𝑒!!!

where  𝐶 𝑥 is the concentration of molecules at distance 𝑥 (µm) from the source

boundary, 𝐶! is the concentration of molecules at the source boundary and λ

(µm) is the decay length, the distance from the source where the concentration of

the molecule has decayed to a factor of 1 𝑒 of the concentration at the source 𝐶!.

Equation (A1) makes predictions about the rate of movement and degradation of

molecules during the formation of a gradient. FRAP analysis can approximate the

rate of degradation and the diffusion coefficient for these molecules by comparing

experimentally derived rate data to the predicted rate data according to equation

(A1). The mesoderm-BCR explant system provides for the unique ability to

observe a gradient as it forms. The influx of fluorescent molecules into a defined

area of the receiving tissue is analogous to the recovery of fluorescent molecules

into an area that was photobleached during a FRAP experiment. Thus, the

methodology used to derive transport rate parameters from FRAP experiments

can be applied to the formation of the sf-PDGF-A gradient in mesoderm-BCR

explants. The sections below describe in detail, the derivation of the degradation

rate and the calculation of the effective diffusion coefficient for the formation of a

sf-PDGF-A-eGFP gradient in mesoderm-BCR explants according to the

methodology used to analyze the kinetics of the Dpp gradient in the Drosophila

wing imaginal disc (Kicheva et al. 2007; Zhou et al. 2012).

  260

A1: Generation of a theoretical “recovery” curve:

The simplified de-dimentionalized time dependent solution to (A1) is shown

below:

(A3, after Zhou et al. 2012)

𝑐(𝑋, 𝜏) =12𝐶!𝑒

!! 2− Erfc −𝑋 − 2𝜏2 𝜏

− 𝑒!!Erfc𝑋 + 2𝜏2 𝜏

where dimensionless position 𝑋 = !! and dimensionless time 𝜏 = 𝑘𝑡. Both sides

of equation (A3) are divided by 𝐶! to give the dimensionless concentration 𝐶!.

𝐶! =𝑐(𝑋, 𝜏)𝐶!

=

12𝐶!𝑒

!! 2− Erfc −𝑋 − 2𝜏2 𝜏

− 𝑒!!Erfc 𝑋 + 2𝜏2 𝑡

𝐶!

Since my measurements are made at 𝑥 = 𝜆 = 80, then 𝑋 = 1 and because

Erfc(𝑥)  =  1− Erf(𝑥), equation (A3) becomes:

(A4)

𝐶′ =12𝑒 2− 1− Erf −

1− 2𝜏2 𝜏

− 𝑒! 1− Erf1+ 2𝜏2 𝜏

which can be re-organized to:

(A5)

𝐶′ =12𝑒 1− 𝑒! + Erf

1− 2𝜏2 𝜏

+ 𝑒!Erf1+ 2𝜏2 𝜏

𝑓1 𝑓2

  261

For τ values ranging between 0 and 10, the 𝑓1  and 𝑓2 error function values were

calculated and are presented in Table A1:

Sample calculation for τ = 0.5:

𝑓2 =1+ 2(0.5)2 0.5

= 1.414

Table A1: The calculated 𝒇1 and 𝒇2 values for the range of t values

τ 𝒇1 𝒇2

0 ∞ ∞

0.1 1.265 1.897

0.2 0.670 1.565

0.4 0.158 1.423

0.5 0 1.414

0.6 -0.129 1.420

1 -0.50 1.50

2 -1.061 1.768

4 -1.751 2.250

6 -2.245 2.654

8 -2.652 3.01

10 -3.00 3.320

  262

𝑓1 and 𝑓2 values from Table A4.1 were used to calculate the error function

approximations using the following equation (Abramowitz and Stegun, 1964):

Erf 𝑓 ≈ 1−  1

1+ 𝑎!𝑓 + 𝑎!𝑓! + 𝑎!𝑓! + 𝑎!𝑓! !

where 𝑎! = 0.2784,𝑎! = 0.2304,𝑎! =  0.00097,𝑎! = 0.0781

Table A2 lists the error function approximations Erf 𝑓1 and Erf 𝑓2 for each

value of τ.

Table A2: Error function approximation values for 𝒇1 and 𝒇2 for the indicated value of t

τ 𝐄𝐫𝐟(𝒇1) 𝐄𝐫𝐟(𝒇2)

0 1 1

0.1 0.926 0.992

0.2 0.656 0.973

0.4 0.176 0.956

0.5 0 0.954

0.6 0.144 0.955

1 0.520 0.966

2 0.866 0.987

4 0.986 0.998

6 0.998 0.999

8 0.999 0.999

10 0.999 0.999

  263

Theoretical recovery curves are plotted as the ratio !(!,!)!!

=  𝐶!!! (𝜏)  over τ, which is

calculated by substituting Erf(𝑓1) and Erf(𝑓2) values into equation A4.

Sample Calculation: for τ = 10

𝐶!!! (10) =12𝑒 1− 𝑒! + 0.999+ 𝑒!0.999 = 0.368  

Table A3 lists 𝐶!!! (𝜏) values for each value of τ and is used to generate the

theoretical fluorescence recovery curve (Figure A1A).

Table A3: Calculated 𝑪𝒕𝒉! 𝝉 values for the indicated value of τ

τ 𝑪𝒕𝒉! (𝝉)

0 0

0.1 0.0029

0.2 0.0268

0.4 0.0919

0.5 0.1123

0.6 0.1503

1 0.2339

2 0.3261

4 0.3628

6 0.3670

8 0.3676

10 0.3678

  264

Figure A1: Calculation of the degradation constant for sf-PDGF-A-eGFP

(A) Theoretical recovery curve for sf-PDGF-A-eGFP. (B) Plot of τ over time (𝑡);

the red line is the linear line of best fit; the R2 value indicates that the black curve

closely approximates a linear function.

  265

Figure A1: Calculation of the degradation constant for sf-PDGF-A-eGFP

  266

A2: Fitting experimental data for sf-PDGF-A-eGFP to theoretical recovery curves: The theoretical recovery curve is shown in Figure A1. In order to determine the

degradation rate 𝑘 for the sf-PDGF-A-eGFP gradient the experimentally derived

data must be fitted to the theoretical curve. Since the experimental

measurements were made at λ, the concentration 𝐶!, should be equal to !!𝐶!

when the gradient reaches a steady state. The experimental recovery curve is

presented in Figure 4.13E. This curve shows that the fluorescence intensity at

λ begins to level off at an average intensity of 73 +/- 1.7 arbitrary intensity units

approximately 70 minutes after the mesoderm and BCR (PDGF-A source)

explants were assembled and filming was started. 𝐶! can be calculated from the

average 𝐶! as follows:

𝐶! = 𝐶!!!

!!, thus 𝐶! = 73 !

!

!!= 198

A 𝐶! value of 198 arbitrary intensity units is in the range of the experimentally

derived average source boundary concentration of 178 +/- 9.8 (see Figure

4.10B).

The experimental data is fitted to the theoretical recovery curve by calculating the

ratio 𝐶!"! (𝜏) =!!!!

and then determining the corresponding value of 𝜏, τex from the

theoretical curve. The time required for fluorescence intensity to increase at

  267

position λ after assembly of the mesoderm-BCR explants is explant specific and

varies between 20 and 40 minutes. I have previously observed that the time

required for mesoderm-BCR explants to heal varies from explant to explant,

however generally it takes approximately 30 minutes for mesoderm-BCR

explants to heal after being combined (See Chapter Two; Damm and Winklbauer,

2011). Thus, the time it takes for fluorescence intensity to begin to increase at

λ could be a result of the time required for the mesoderm-BCR interface to form.

For the purposes of calculating the kinetic parameters 𝑘 and 𝐷, the onset of

molecule movement into the area of λ (t = 0 seconds) was taken to be 30

minutes after mesoderm-BCR explants were combined. The theoretical recovery

curve levels off at 𝐶!!! (𝜏) = 0.3679 as the gradient reaches a steady state;

experimentally the gradient levels off at approximately this value after 50

minutes. The 𝐶!"! (𝜏)  values determined from experiments and the values for τex,

for the 50 minute period are presented in Table A4.

Table A4: 𝑪𝒆𝒙! 𝝉 and 𝒕 values with the corresponding values 𝝉𝒆𝒙 for experimentally derived  𝑪𝝀.

τex 𝒕 (s) 𝑪𝒆𝒙! (𝝉)

0 0 0

0.45 600 0.1174

0.98 1200 0.2285

1.5 1800 0.2576

2.3 2400 0.3422

  268

3.3 3000 0.3560

A3: Determining the degradation rate and the effective diffusion coefficient The relationship between τ and time (t seconds) is:

𝜏 = 𝑘𝑡

where the slope 𝑘 of the function is the degradation constant. The diffusion

coefficient 𝐷 and 𝑘 are related through the equation:

(A7)

𝐷 = 𝜆!𝑘

Thus, if the degradation constant 𝑘 and the decay length 𝜆 are known, the

diffusion coefficient 𝐷 can be calculated. In order to determine 𝑘 for the sf-PDGF-

A-eGFP gradient, τex was plotted over time 𝑡 to produce the linear curve

presented in Figure A1B. Regression analysis showed that the values for τex

over 𝑡 from Table A4 closely approximate a linear function (R2=0.97), as

expected from the relationship between τ and 𝑡. The degradation constant 𝑘 for

the sf-PDGF-A-eGFP gradient is the slope of the linear curve:

𝑘 = 1.07×10!! s-1

Using equation (A7), 𝑘 = 1.07×10!! s-1 and 𝜆 = 80  µm, the effective diffusion

coefficient (𝐷!"") for the sf-PDGF-A-eGFP gradient can be calculated:

𝐷!"" = (80!) 1.07×10!! = 6.8 µm2s-1

  269

Chapter Five: Final Model and References

  270

5.1: Prechordal mesoderm morphogenesis: the cellular basis and the molecular regulation

My work has lead to an understanding of the morphogenetic movements that

occur in the PCM of gastrulating Xenopus embryos. Based on the results of my

work I propose a model of PCM morphogenesis in the following summary (Figure

5.1): After the start of gastrulation, the PCM consists of multi-layered upper and

lower regions with distinct cell orientations. The upper PCM is internalized during

the early stages of gastrulation and is positioned adjacent to the inner layer of the

overlying BCR. The long axes of these cells are perpendicular with respect to the

BCR and this orientation is BCR dependent. The lower PCM region is located in

the dorsal blastopore lip, adjacent to the XBra expressing CM. At this stage, the

long axes of lower PCM cells are oriented parallel to the adjacent CM region. As

gastrulation proceeds, the lower PCM undergoes involution and becomes

internalized to the level of the BCR, likely as a result of a combination of animally

directed intercellular movements within the lower PCM and vegetally directed

epiboly movements of the BCR. Thus, all PCM cells become oriented with their

long axes perpendicular to the BCR by the middle of gastrulation (Figure 5.1A).

My work has shown that the inner layer cells of the BCR produce three PDGF-A

isoforms, sf-PDGF-A, int-PDGF-A and lf-PDGF-A. The cognate receptor for

PDGF-A, PDGFRα, is expressed in the PCM. Thus, this mutually exclusive

expression pattern suggests that the ectoderm signals to the PCM through

  271

PDGF. The int-PDGF-A and lf-PDGF-A isoforms interact with extracellular matrix

molecules and thus remain localized to the vicinity of the secreting cells, with an

accumulation in Brachet’s cleft (Figure 5.1A,B). These isoforms regulate the

animally directed migration of the LEM through a short range contact based

mechanism (Nagel et al. 2004). The sf-PDGF-A isoform participates in long

range signaling. I have shown that sf-PDGF-A appears to form a steady state

extracellular gradient in approximately one hour, by the diffusion of sf-PDGF-A

molecules through PCM intercellular spaces (Figure 5.1A,B; See Chapter Four).

Furthermore, I have shown that in response to the long range sf-PDGF-A signal,

PCM cells orient locomotory protrusions toward the BCR and undergo radial

intercalation as they migrate toward the mesoderm/ectoderm boundary. By late

gastrulation, radial intercalation of the PCM converts a multi-layered region into a

monolayer (Figure 5.1B), resulting in an expansion of the area of PCM that is in

contact with the BCR, while contributing to the extension of the anterior/posterior

axis of the embryo.

  272

Figure 5.1: A model for the cellular basis of prechordal mesoderm morphogenesis (A-B) Schematic diagrams of the change in cell organization of the PCM over the

course of gastrulation. (A) By the middle of gastrulation the PCM is multi-layered

and cells extend locomotory protrusions toward the ectoderm; PDGF-A isoforms

are produced and secreted by ectoderm cells; sf-PDGF-A forms a concentration

gradient in the PCM by diffusing through intercellular spaces. This gradient

instructively orients PDGFRa expressing PCM cells toward the ectoderm, lf-

PDGF-A and int-PDGF-A remain associated with the surfaces of the secreting

cells. (B) By late gastrulation, the PCM has undergone radial intercalation to thin

to a single layer of cells in contact with the ectoderm. The black arrow in (A)

indicates the direction of PCM cell movement; green wedges show the direction

of the sf-PDGF-A concentration gradient.

  273

Figure 5.1: A model for the cellular basis of prechordal mesoderm morphogenesis

  274

5.2: References    Abramowitz M,. Stegun, I. A. (1965). Handbook of Mathematic Functions: with

Formulas, Graphs and Mathematical Tables. ISBN: 0486612724

Akiyama, T., Kamimura, K., Firkus, C., Takeo, S., Shimmi, O. and Nakato, H.

(2008). Dally Regulates Dpp Morphogen Gradient Formation by Stabilizing Dpp

on the Cell Surface. Dev. Biol. 313, 408-419.

Andersson, M., Ostman, A., Westermark, B. and Heldin, C. H. (1994).

Characterization of the Retention Motif in the C-Terminal Part of the Long Splice

Form of Platelet-Derived Growth Factor A-Chain. The Journal of Biological

Chemistry 269, 926-930.

Andrae, J., Gallini, R. and Betsholtz, C. (2008). Role of Platelet-Derived Growth

Factors in Physiology and Medicine. Genes & Development 22, 1276-1312.

Asashima, M., Nakano, H., Uchiyama, H., Davids, M., Plessow, S., Loppnow-

Blinde, B., Hoppe, P., Dau, H. and Tiedemann, H. (1990). The Vegetalizing

Factor Belongs to a Family of Mesoderm-Inducing Proteins Related to Erythroid

Differentiation Factor. Naturwissenschaften 77, 389-391.

Ataliotis, P., Symes, K., Chou, M. M., Ho, L. and Mercola, M. (1995). PDGF

Signalling is Required for Gastrulation of Xenopus Laevis. Development 121,

3099-3110.

  275

Avrov, K. and Kazlauskas, A. (2003). The Role of c-Src in Platelet-Derived

Growth Factor Alpha Receptor Internalization. Exp. Cell Res. 291, 426-434.

Baeg, G. H., Lin, X., Khare, N., Baumgartner, S. and Perrimon, N. (2001).

Heparan Sulfate Proteoglycans are Critical for the Organization of the

Extracellular Distribution of Wingless. Development 128, 87-94.

Baker, N. E. (2007). Patterning Signals and Proliferation in Drosophila Imaginal

Discs. Current Opinion in Genetics {\&} Development 17, 287-293.

Basler, K. and Struhl, G. (1994). Compartment Boundaries and the Control of

Drosophila Limb Pattern by Hedgehog Protein. Nature 368, 208-214.

Bauer, D. V., Huang, S. and Moody, S. A. (1994). The Cleavage Stage Origin of

Spemann's Organizer: Analysis of the Movements of Blastomere Clones before

and during Gastrulation in Xenopus. Development 120, 1179-1189.

Belenkaya, T. Y., Han, C., Yan, D., Opoka, R. J., Khodoun, M., Liu, H. and Lin, X.

(2004). Drosophila Dpp Morphogen Movement is Independent of Dynamin-

Mediated Endocytosis but Regulated by the Glypican Members of Heparan

Sulfate Proteoglycans. Cell 119, 231-244.

Bergsten, E., Uutela, M., Li, X., Pietras, K., Ostman, A., Heldin, C. H., Alitalo, K.

and Eriksson, U. (2001). PDGF-D is a Specific, Protease-Activated Ligand for the

PDGF Beta-Receptor. Nat. Cell Biol. 3, 512-516.

Bernfield, M., Gotte, M., Park, P. W., Reizes, O., Fitzgerald, M. L., Lincecum, J.

  276

and Zako, M. (1999). Functions of Cell Surface Heparan Sulfate Proteoglycans.

Annu. Rev. Biochem. 68, 729-777.

Berridge, M. J. (1993). Cell Signalling. A Tale of Two Messengers. Nature 365,

388-389.

Bertet, C., Sulak, L. and Lecuit, T. (2004). Myosin-Dependent Junction

Remodelling Controls Planar Cell Intercalation and Axis Elongation. Nature 429,

667-671.

Blader, P. and Strahle, U. (1998). Ethanol Impairs Migration of the Prechordal

Plate in the Zebrafish Embryo. Dev. Biol. 201, 185-201.

Blankenship, J. T., Backovic, S. T., Sanny, J. S. P., Weitz, O. and Zallen, J. A.

(2006). Multicellular Rosette Formation Links Planar Cell Polarity to Tissue

Morphogenesis. Developmental Cell 11, 459-470.

Blitz, I. (2000). Is Chordin a Long-Range- Or Short-Range-Acting Factor? Roles

for BMP1-Related Metalloproteases in Chordin and BMP4 Autofeedback Loop

Regulation. Dev. Biol. 223, 120-138.

Boldajipour, B., Mahabaleshwar, H., Kardash, E., Reichmanfried, M., Blaser, H.,

Minina, S., Wilson, D., Xu, Q. and Raz, E. (2008). Control of Chemokine-Guided

Cell Migration by Ligand Sequestration. Cell 132, 463-473.

Bonner, J. C. (1994). Regulation of Platelet-Derived Growth Factor (PDGF) and

Alveolar Macrophage-Derived PDGF by Alpha 2-Macroglobulin. Ann. N. Y. Acad.

  277

Sci. 737, 324-338.

Bonner, J. C. (2004). Regulation of PDGF and its Receptors in Fibrotic Diseases.

Cytokine & Growth Factor Reviews 15, 255-273.

Bornfeldt, K. E., Raines, E. W., Graves, L. M., Skinner, M. P., Krebs, E. G. and

Ross, R. (1995). Platelet-Derived Growth Factor. Distinct Signal Transduction

Pathways Associated with Migration Versus Proliferation. Ann. N. Y. Acad. Sci.

766, 416-430.

Bothe, I. and Dietrich, S. (2006). The Molecular Setup of the Avian Head

Mesoderm and its Implication for Craniofacial Myogenesis. Developmental

Dynamics 235, 2845-2860.

Boucaut, J. C., Darribere, T., Boulekbache, H. and Thiery, J. P. (1984a).

Prevention of Gastrulation but Not Neurulation by Antibodies to Fibronectin in

Amphibian Embryos. Nature 307, 364-367.

Boucaut, J. C., Darribere, T., Poole, T. J., Aoyama, H., Yamada, K. M. and

Thiery, J. P. (1984b). Biologically Active Synthetic Peptides as Probes of

Embryonic Development: A Competitive Peptide Inhibitor of Fibronectin Function

Inhibits Gastrulation in Amphibian Embryos and Neural Crest Cell Migration in

Avian Embryos. J. Cell Biol. 99, 1822-1830.

Branford, W. W. and Yost, H. J. (2002). Lefty-Dependent Inhibition of Nodal- and

Wnt-Responsive Organizer Gene Expression is Essential for Normal

  278

Gastrulation. Current Biology 12, 2136-2141.

Burgering, B. M. and Coffer, P. J. (1995). Protein Kinase B (c-Akt) in

Phosphatidylinositol-3-OH Kinase Signal Transduction. Nature 376, 599-602.

Chalmers, A. D. and Slack, J. M. (2000). The Xenopus Tadpole Gut: Fate Maps

and Morphogenetic Movements. Development 127, 381-392.

Chan, J., Mably, J. D., Serluca, F. C., Chen, J. N., Goldstein, N. B., Thomas, M.

C., Cleary, J. A., Brennan, C., Fishman, M. C. and Roberts, T. M. (2001).

Morphogenesis of Prechordal Plate and Notochord Requires Intact Eph/ephrin B

Signaling. Dev. Biol. 234, 470-482.

Cho, K. W., Blumberg, B., Steinbeisser, H. and Robertis, E. M. D. (1991).

Molecular Nature of Spemann's Organizer: The Role of the Xenopus Homeobox

Gene Goosecoid. Cell 67, 1111-1120.

Colas, A., Cartry, J., Buisson, I., Umbhauer, M., Smith, J. C. and Riou, J. C.

(2008). Mix.1/2-Dependent Control of FGF Availability during Gastrulation is

Essential for Pronephros Development in Xenopus. Dev. Biol. 320, 351-365.

Cooke, J. (1979). Cell Number in Relation to Primary Pattern Formation in the

Embryo of Xenopus Laevis. II. Sequential Cell Recruitment, and Control of the

Cell Cycle, during Mesoderm Formation. J. Embryol. Exp. Morphol. 53, 269-289.

Cousin, H., Abbruzzese, G., McCusker, C. and Alfandari, D. (2012). ADAM13

Function is Required in the 3 Dimensional Context of the Embryo during Cranial

  279

Neural Crest Cell Migration in Xenopus Laevis. Dev. Biol. 368, 335-344.

Crick, F. (1970). Diffusion in Embryogenesis. Nature 225, 420-422.

Dale, L. and Slack, J. M. (1987). Fate Map for the 32-Cell Stage of Xenopus

Laevis. Development 99, 527-551.

Damm, E. W. and Winklbauer, R. (2011). PDGF-A Controls Mesoderm Cell

Orientation and Radial Intercalation during Xenopus Gastrulation. Development

138, 565-575.

Darribere, T., Yamada, K. M., Johnson, K. E. and Boucaut, J. C. (1988). The

140-kDa Fibronectin Receptor Complex is Required for Mesodermal Cell

Adhesion during Gastrulation in the Amphibian Pleurodeles Waltlii. Dev. Biol.

126, 182-194.

Davidson, L. A., Hoffstrom, B. G., Keller, R. and DeSimone, D. W. (2002).

Mesendoderm Extension and Mantle Closure in Xenopus Laevis Gastrulation:

Combined Roles for Integrin α5β1, Fibronectin, and Tissue Geometry. Dev. Biol.

242, 109-129.

Dick, A., Hild, M., Bauer, H., Imai, Y., Maifeld, H., Schier, A. F., Talbot, W. S.,

Bouwmeester, T. and Hammerschmidt, M. (2000). Essential Role of Bmp7

(Snailhouse) and its Prodomain in Dorsoventral Patterning of the Zebrafish

Embryo. Development 127, 343-354.

Djiane, A., Riou, J., Umbhauer, M., Boucaut, J. and Shi, D. (2000). Role of

  280

Frizzled 7 in the Regulation of Convergent Extension Movements during

Gastrulation in Xenopus Laevis. Development 127, 3091-3100.

Doitsidou, M., Reichman-Fried, M., Stebler, J., Koprunner, M., Dorries, J., Meyer,

D., Esguerra, C. V., Leung, T. and Raz, E. (2002). Guidance of Primordial Germ

Cell Migration by the Chemokine SDF-1. Cell 111, 647-659.

Dosch, R., Gawantka, V., Delius, H., Blumenstock, C. and Niehrs, C. (1997).

Bmp-4 Acts as a Morphogen in Dorsoventral Mesoderm Patterning in Xenopus.

Development 124, 2325-2334.

Dougan, S. T., Warga, R. M., Kane, D. A., Schier, A. F. and Talbot, W. S. (2003).

The Role of the Zebrafish Nodal-Related Genes Squint and Cyclops in Patterning

of Mesendoderm. Development 130, 1837-1851.

Duchek, P., Somogyi, K., Jekely, G., Beccari, S. and Rorth, P. (2001). Guidance

of Cell Migration by the Drosophila PDGF/VEGF Receptor. Cell 107, 17-26.

Edwards, D. R., Handsley, M. M. and Pennington, C. J. (2008). The ADAM

Metalloproteinases. Mol. Aspects Med. 29, 258-289.

Ekman, S., Thuresson, E. R., Heldin, C. H. and Ronnstrand, L. (1999). Increased

Mitogenicity of an Alphabeta Heterodimeric PDGF Receptor Complex Correlates

with Lack of RasGAP Binding. Oncogene 18, 2481-2488.

Entchev, E. V., Schwabedissen, A. and Gonzalez-Gaitan, M. (2000). Gradient

Formation of the TGF-Beta Homolog Dpp. Cell 103, 981-991.

  281

Farese, R. V., Cases, S., Ruland, S. L., Kayden, H. J., Wong, J. S., Young, S. G.

and Hamilton, R. L. (1996). A Novel Function for Apolipoprotein B: Lipoprotein

Synthesis in the Yolk Sac is Critical for Maternal-Fetal Lipid Transport in Mice. J.

Lipid Res. 37, 347-360.

Ferrara, N., Gerber, H. and LeCouter, J. (2003). The Biology of VEGF and its

Receptors. Nat. Med. 9, 669-676.

Fredriksson, L., Li, H. and Eriksson, U. (2004a). The PDGF Family: Four Gene

Products Form Five Dimeric Isoforms. Cytokine & Growth Factor Reviews 15,

197-204.

Fredriksson, L., Li, H., Fieber, C., Li, X. and Eriksson, U. (2004b). Tissue

Plasminogen Activator is a Potent Activator of PDGF-CC. EMBO J. 23, 3793-

3802.

Fujise, M., Takeo, S., Kamimura, K., Matsuo, T., Aigaki, T., Izumi, S. and Nakato,

H. (2003). Dally Regulates Dpp Morphogen Gradient Formation in the Drosophila

Wing. Development 130, 1515-1522.

Fukata, M., Nakagawa, M. and Kaibuchi, K. (2003). Roles of Rho-Family

GTPases in Cell Polarisation and Directional Migration. Curr. Opin. Cell Biol. 15,

590-597.

Funamoto, S., Meili, R., Lee, S., Parry, L. and Firtel, R. A. (2002). Spatial and

Temporal Regulation of 3-Phosphoinositides by PI 3-Kinase and PTEN Mediates

  282

Chemotaxis. Cell 109, 611-623.

Furthauer, M., Thisse, C. and Thisse, B. (1997). A Role for FGF-8 in the

Dorsoventral Patterning of the Zebrafish Gastrula. Development 124, 4253-4264.

Furuhashi, M., Sjoblom, T., Abramsson, A., Ellingsen, J., Micke, P., Li, H.,

Bergsten-Folestad, E., Eriksson, U., Heuchel, R., Betsholtz, C. et al. (2004).

Platelet-Derived Growth Factor Production by B16 Melanoma Cells Leads to

Increased Pericyte Abundance in Tumors and an Associated Increase in Tumor

Growth Rate. Cancer Res. 64, 2725-2733.

Gong, Y., Mo, C. and Fraser, S. E. (2004). Planar Cell Polarity Signalling

Controls Cell Division Orientation during Zebrafish Gastrulation. Nature 430, 689-

693.

Greco, V., Hannus, M. and Eaton, S. (2001). Argosomes: A Potential Vehicle for

the Spread of Morphogens through Epithelia. Cell 106, 633-645.

Gregor, T., Wieschaus, E. F., McGregor, A. P., Bialek, W. and Tank, D. W.

(2007). Stability and Nuclear Dynamics of the Bicoid Morphogen Gradient. Cell

130, 141-152.

Gubb, D. and Garcia-Bellido, A. (1982). A Genetic Analysis of the Determination

of Cuticular Polarity during Development in Drosophila Melanogaster. J. Embryol.

Exp. Morphol. 68, 37-57.

  283

Hagemann, A. I., Xu, X., Nentwich, O., Hyvonen, M. and Smith, J. C. (2009).

Rab5-Mediated Endocytosis of Activin is Not Required for Gene Activation Or

Long-Range Signalling in Xenopus. Development 136, 2803-2813.

Hama, J. and Weinstein, D. C. (2001). Is Chordin a Morphogen? Bioessays 23,

121-124.

Harland, R. M. (1991). In Situ Hybridization: An Improved Whole-Mount Method

for Xenopus Embryos. Methods Cell Biol. 36, 685-695.

Hawkins, P. T., Eguinoa, A., Qiu, R. G., Stokoe, D., Cooke, F. T., Walters, R.,

Wennstrom, S., Claesson-Welsh, L., Evans, T. and Symons, M. (1995). PDGF

Stimulates an Increase in GTP-Rac Via Activation of Phosphoinositide 3-Kinase.

Current Biology 5, 393-403.

Heisenberg, C., Tada, M., Rauch, G. J., Saude, L., Concha, M. L., Geisler, R.,

Stemple, D. L., Smith, J. C. and Wilson, S. W. (2000). Silberblick/Wnt11 Mediates

Convergent Extension Movements during Zebrafish Gastrulation. Nature 405, 76-

81.

Heldin, C. H., Ostman, A. and Ronnstrand, L. (1998). Signal Transduction Via

Platelet-Derived Growth Factor Receptors. Biochim. Biophys. Acta 1378, F79-

113.

Hermanson, M., Funa, K., Hartman, M., Claesson-Welsh, L., Heldin, C. H.,

Westermark, B. and Nister, M. (1992). Platelet-Derived Growth Factor and its

  284

Receptors in Human Glioma Tissue: Expression of Messenger RNA and Protein

Suggests the Presence of Autocrine and Paracrine Loops. Cancer Res. 52,

3213-3219.

Heuchel, R., Berg, A., Tallquist, M., Ahlen, K., Reed, R. K., Rubin, K., Claesson-

Welsh, L., Heldin, C. H. and Soriano, P. (1999). Platelet-Derived Growth Factor

Beta Receptor Regulates Interstitial Fluid Homeostasis through

Phosphatidylinositol-3' Kinase Signaling. Proc. Natl. Acad. Sci. U. S. A. 96,

11410-11415.

Hikasa, H., Shibata, M., Hiratani, I. and Taira, M. (2002). The Xenopus Receptor

Tyrosine Kinase Xror2 Modulates Morphogenetic Movements of the Axial

Mesoderm and Neuroectoderm Via Wnt Signaling. Development 129, 5227-5239.

Ho, L., Symes, K., Yordan, C., Gudas, L. J. and Mercola, M. (1994). Localization

of PDGF A and PDGFR Alpha mRNA in Xenopus Embryos Suggests Signalling

from Neural Ectoderm and Pharyngeal Endoderm to Neural Crest Cells. Mech.

Dev. 48, 165-174.

Hoch, R. V. and Soriano, P. (2003). Roles of PDGF in Animal Development.

Development 130, 4769-4784.

Horiuchi, H., Inoue, T., Furusawa, S. and Matsuda, H. (2001). Characterization

and Expression of Three Forms of cDNA Encoding Chicken Platelet-Derived

Growth Factor-A Chain. Gene 272, 181-190.

  285

Howe, J. A., Howell, M., Hunt, T. and Newport, J. W. (1995). Identification of a

Developmental Timer Regulating the Stability of Embryonic Cyclin A and a New

Somatic A-Type Cyclin at Gastrulation. Genes & Development 9, 1164-1176.

Hsiung, F., Ramirez-Weber, F., Iwaki, D. D. and Kornberg, T. B. (2005).

Dependence of Drosophila Wing Imaginal Disc Cytonemes on Decapentaplegic.

Nature 437, 560-563.

Ibrahim, H. and Winklbauer, R. (2001). Mechanisms of Mesendoderm

Internalization in the Xenopus Gastrula: Lessons from the Ventral Side. Dev. Biol.

240, 108-122.

Jechlinger, M., Grunert, S., Tamir, I. H., Janda, E., Ludemann, S., Waerner, T.,

Seither, P., Weith, A., Beug, H. and Kraut, N. (2003). Expression Profiling of

Epithelial Plasticity in Tumor Progression. Oncogene 22, 7155-7169.

Johnson, K. E. (1977a). Extracellular Matrix Synthesis in Blastula and Gastrula

Stages of Normal and Hybrid Frog Embryos. I. Toluidine Blue and Lanthanum

Staining. J. Cell. Sci. 25, 313-322.

Johnson, K. E. (1977b). Extracellular Matrix Synthesis in Blastula and Gastrula

Stages of Normal and Hybrid Frog Embryos. III. Characterization of Galactose-

and Glucosamine-Labelled Materials. J. Cell. Sci. 25, 335-354.

Kuhl, M. and Wedlich, D. (1996). Xenopus Cadherins: Sorting Out Types and

Functions in Embryogenesis. Developmental Dynamics 207, 121-134.

  286

Kawada, K., Upadhyay, G., Ferandon, S., Janarthanan, S., Hall, M., Vilardaga, J.

and Yajnik, V. (2009). Cell Migration is Regulated by PDGF Receptor

Endocytosis. Mol. Cell. Biol. 29, 4508-4518

Keller, R. E. (1975). Vital Dye Mapping of the Gastrula and Neurula of Xenopus

Laevis. I. Prospective Areas and Morphogenetic Movements of the Superficial

Layer. Dev. Biol. 42, 222-241.

Keller, R. E. (1976). Vital Dye Mapping of the Gastrula and Neurula of Xenopus

Laevis. II. Prospective Areas and Morphogenetic Movements of the Deep Layer.

Dev. Biol. 51, 118-137.

Keller, R. E. (1980). The Cellular Basis of Epiboly: An SEM Study of Deep-Cell

Rearrangement during Gastrulation in Xenopus Laevis. J. Embryol. Exp.

Morphol. 60, 201-234.

Keller, R. E. (2004). Developmental Biology: Heading Away from the Rump.

Nature 430, 305-306.

Keller, R. E., Danilchik, M., Gimlich, R. and Shih, J. (1985). The Function and

Mechanism of Convergent Extension during Gastrulation of Xenopus Laevis. J.

Embryol. Exp. Morphol. 89 Suppl, 185-209.

Keller, R. E., Davidson, L., Edlund, A., Elul, T., Ezin, M., Shook, D. and Skoglund,

P. (2000). Mechanisms of Convergence and Extension by Cell Intercalation.

Philosophical Transactions of the Royal Society of London Series, Biological

  287

Sciences 355, 897-922.

Keller, R. (1991). Early Embryonic Development of Xenopus Laevis. Methods

Cell Biol. 36, 61-113.

Keller, R., Cooper, M. S., Danilchik, M., Tibbetts, P. and Wilson, P. A. (1989).

Cell Intercalation during Notochord Development in Xenopus Laevis. J. Exp.

Zool. 251, 134-154.

Keller, R. and Danilchik, M. (1988). Regional Expression, Pattern and Timing of

Convergence and Extension during Gastrulation of Xenopus Laevis.

Development 103, 193-209.

Keller, R. and Tibbetts, P. (1989). Mediolateral Cell Intercalation in the Dorsal,

Axial Mesoderm of Xenopus Laevis. Dev. Biol. 131, 539-549.

Keller, R. (2002). Shaping the Vertebrate Body Plan by Polarized Embryonic Cell

Movements. Science 298, 1950-1954.

Keller, R., Davidson, L. A. and Shook, D. R. (2003). How we are Shaped: The

Biomechanics of Gastrulation. Differentiation 71, 171-205.

Keller, R. and Schoenwolf, G. (1977). An SEM Study of Cellular Morphology,

Contact and Arrangement, as Related to Gastrulation in Xenopus Laevis. Roux's

Archives of Developmental Biology 182, 165-186.

  288

Kelly, J. L., Sanchez, A., Brown, G. S., Chesterman, C. N. and Sleigh, M. J.

(1993). Accumulation of PDGF B and Cell-Binding Forms of PDGF A in the

Extracellular Matrix. J. Cell Biol. 121, 1153-1163.

Kerszberg, M. and Wolpert, L. (1998). Mechanisms for Positional Signalling by

Morphogen Transport: A Theoretical Study. J. Theor. Biol. 191, 103-114.

Khokha, M., Yeh, J., Grammer, T. and Harland, R. (2005). Depletion of Three

BMP Antagonists from Spemann's Organizer Leads to a Catastrophic Loss of

Dorsal Structures. Developmental Cell 8, 401-411.

Kicheva, A., Bollenbach, T., Wartlick, O., Julicher, F. and Gonzalez-Gaitan, M.

(2012). Investigating the Principles of Morphogen Gradient Formation: From

Tissues to Cells. Current Opinion in Genetics & Development.

10.1016/j.gde.2012.08.004.

Kicheva, A., Pantazis, P., Bollenbach, T., Kalaidzidis, Y., Bittig, T., Julicher, F.

and Gonzalez-Gaitan, M. (2007). Kinetics of Morphogen Gradient Formation.

Science 315, 521-525.

Kishimoto, Y., Lee, K. H., Zon, L., Hammerschmidt, M. and Schulte-Merker, S.

(1997). The Molecular Nature of Zebrafish Swirl: BMP2 Function is Essential

during Early Dorsoventral Patterning. Development 124, 4457-4466.

Klingseisen, A., Clark, I. B. N., Gryzik, T. and Muller, H. J. (2009). Differential and

Overlapping Functions of Two Closely Related Drosophila FGF8-Like Growth

  289

Factors in Mesoderm Development. Development 136, 2393-2402.

Koibuchi, N., Kaneda, Y., Taniyama, Y., Matsumoto, K., Nakamura, T., Ogihara,

T. and Morishita, R. (2004). Essential Role of HGF (Hepatocyte Growth Factor) in

Blood Formation in Xenopus. Blood 103, 3320-3325.

Kruse, K., Pantazis, P., Bollenbach, T., J\ulicher, F. and Gonzalez-Gaitan, M.

(2004). Dpp Gradient Formation by Dynamin-Dependent Endocytosis: Receptor

Trafficking and the Diffusion Model. Development 131, 4843-4856.

Kurth, T., Berger, J., Wilsch-Brauninger, M., Kretschmar, S., Cerny, R., Schwarz,

H., Lofberg, J., Piendl, T. and Epperlein, H. H. (2010). Electron Microscopy of the

Amphibian Model Systems Xenopus Laevis and Ambystoma Mexicanum.

Methods Cell Biol. 96, 395-423.

Lane, M. C. and Smith, W. C. (1999). The Origins of Primitive Blood in Xenopus:

Implications for Axial Patterning. Development 126, 423-434.

Lakowicz, J. (2006). Principles of Fluorescence Spectroscopy 3rd Edition.

Springer. ISBN: 9780387463124

Lecuit, T., Brook, W. J., Ng, M., Calleja, M., Sun, H. and Cohen, S. M. (1996).

Two Distinct Mechanisms for Long-Range Patterning by Decapentaplegic in the

Drosophila Wing. Nature 381, 387-393.

Leptin, M. (2005). Gastrulation Movements: The Logic and the Nuts and Bolts.

Developmental Cell 8, 305-320.

  290

Li, X., Ponten, A., Aase, K., Karlsson, L., Abramsson, A., Uutela, M., Backstrom,

G., Hellstrom, M., Bostrom, H., Li, H. et al. (2000). PDGF-C is a New Protease-

Activated Ligand for the PDGF Alpha-Receptor. Nat. Cell Biol. 2, 302-309.

Lin, X. and Perrimon, N. (2000). Role of Heparan Sulfate Proteoglycans in Cell-

Cell Signaling in Drosophila. Matrix Biology : Journal of the International Society

for Matrix Biology 19, 303-307.

Lin, X. (2004). Functions of Heparan Sulfate Proteoglycans in Cell Signaling

during Development. Development 131, 6009-6021.

Lippincott-Schwartz, J., Altan-Bonnet, N. and Patterson, G. H. (2003).

Photobleaching and Photoactivation: Following Protein Dynamics in Living Cells.

Nat. Cell Biol. Suppl, S7-14.

Little, S. C. and Mullins, M. C. (2006). Extracellular Modulation of BMP Activity in

Patterning the Dorsoventral Axis. Birth Defects Research Embryo Today:

Reviews 78, 224-242.

Liu, L., Chong, S., Balasubramaniyan, N. V., Korzh, V. and Ge, R. (2002a).

Platelet-Derived Growth Factor Receptor Alpha (PDGFR-α) Gene in Zebrafish

Embryonic Development. Mech. Dev. 116, 227-230.

Liu, L., Korzh, V., Balasubramaniyan, N. V., Ekker, M. and Ge, R. (2002b).

Platelet-Derived Growth Factor A (PDGF-A) Expression during Zebrafish

Embryonic Development. Dev. Genes Evol. 212, 298-301.

  291

Loechel, F., Fox, J. W., Murphy, G., Albrechtsen, R. and Wewer, U. M. (2000).

ADAM 12-S Cleaves IGFBP-3 and IGFBP-5 and is Inhibited by TIMP-3.

Biochem. Biophys. Res. Commun. 278, 511-515.

Luga, V., Zhang, L., Viloria-Petit, A. M., Ogunjimi, A. A., Inanlou, M. R., Chiu, E.,

Buchanan, M., Hosein, A. N., Basik, M. and Wrana, J. L. (2012). Exosomes

Mediate Stromal Mobilization of Autocrine Wnt-PCP Signaling in Breast Cancer

Cell Migration. Cell 151, 1542-1556.

Luu, O., David, R., Ninomiya, H. and Winklbauer, R. (2011). Large-Scale

Mechanical Properties of Xenopus Embryonic Epithelium. Proc. Natl. Acad. Sci.

U. S. A. 108, 4000-4005

Marjoram, L., Wright, C. (2011). Rapid differential transport of Nodal and Lefty on

sulfated proteoglycan-rich extracellular matrix regulates left-right asymmetry in

Xenopus. Development 138, 475-485

Muller, H. A., Kuhl, M., Finnemann, S., Schneider, S., van der Poel, S. Z.,

Hausen, P. and Wedlich, D. (1994). Xenopus Cadherins: The Maternal Pool

Comprises Distinguishable Members of the Family. Mech. Dev. 47, 213-223.

Marsden, M. and DeSimone, D. W. (2001). Regulation of Cell Polarity, Radial

Intercalation and Epiboly in Xenopus: Novel Roles for Integrin and Fibronectin.

Development 128, 3635-3647.

McDonald, J. A., Pinheiro, E. M. and Montell, D. J. (2003). PVF1, a PDGF/VEGF

  292

Homolog, is Sufficient to Guide Border Cells and Interacts Genetically with

Taiman. Development 130, 3469-3478.

McDowell, N., Zorn, A. M., Crease, D. J. and Gurdon, J. B. (1997). Activin has

Direct Long-Range Signalling Activity and can Form a Concentration Gradient by

Diffusion. Current Biology 7, 671-681.

McMahon, A., Supatto, W., Fraser, S. and Stathopoulos, A. (2008). Dynamic

Analyses of Drosophila Gastrulation Provide Insights into Collective Cell

Migration. Science 322, 1546-1550.

McMahon, A., Reeves, G. T., Supatto, W. and Stathopoulos, A. (2010).

Mesoderm Migration in Drosophila is a Multi-Step Process Requiring FGF

Signaling and Integrin Activity. Development 137, 2167-2175.

Mercola, M., Deininger, P. L., Shamah, S. M., Porter, J., Wang, C. Y. and Stiles,

C. D. (1990). Dominant-Negative Mutants of a Platelet-Derived Growth Factor

Gene. Genes & Development 4, 2333-2341.

Mercola, M., Melton, D. A. and Stiles, C. D. (1988). Platelet-Derived Growth

Factor A Chain is Maternally Encoded in Xenopus Embryos. Science 241, 1223-

1225.

Montero, J., Kilian, B., Chan, J., Bayliss, P. E. and Heisenberg, C. (2003).

Phosphoinositide 3-Kinase is Required for Process Outgrowth and Cell

Polarization of Gastrulating Mesendodermal Cells. Current Biology 13, 1279-

  293

1289.

Moody, S. A. (1987a). Fates of the Blastomeres of the 16-Cell Stage Xenopus

Embryo. Dev. Biol. 119, 560-578.

Moody, S. A. (1987b). Fates of the Blastomeres of the 32-Cell-Stage Xenopus

Embryo. Dev. Biol. 122, 300-319.

Murray, M. J. and Saint, R. (2007). Photoactivatable GFP Resolves Drosophila

Mesoderm Migration Behaviour. Development 134, 3975-3983.

Myers, D. C., Sepich, D. S. and Solnica-Krezel, L. (2002). Convergence and

Extension in Vertebrate Gastrulae: Cell Movements According to Or in Search of

Identity? Trends in Genetics 18, 447-455.

Nagel, M., Tahinci, E., Symes, K. and Winklbauer, R. (2004). Guidance of

Mesoderm Cell Migration in the Xenopus Gastrula Requires PDGF Signaling.

Development 131, 2727-2736.

Nagel, M. and Winklbauer, R. (1999). Establishment of Substratum Polarity in the

Blastocoel Roof of the Xenopus Embryo. Development 126, 1975-1984.

Nakatsuji, N. and Johnson, K. E. (1982). Cell Locomotion in Vitro by Xenopus

Laevis Gastrula Mesodermal Cells. Cell Motil. 2, 149-161.

Nakatsuji, N. and Johnson, K. E. (1983). Comparative Study of Extracellular

Fibrils on the Ectodermal Layer in Gastrulae of Five Amphibian Species. J. Cell.

  294

Sci. 59, 61-70.

Nellen, D., Burke, R., Struhl, G. and Basler, K. (1996). Direct and Long-Range

Action of a DPP Morphogen Gradient. Cell 85, 357-368.

Neuner, R., Cousin, H., McCusker, C., Coyne, M. and Alfandari, D. (2009).

Xenopus ADAM19 is Involved in Neural, Neural Crest and Muscle Development.

Mech. Dev. 126, 240-255.

Niehrs, C., Keller, R., Cho, K. W. and de Robertis, E. M. (1993). The Homeobox

Gene Goosecoid Controls Cell Migration in Xenopus Embryos. Cell 72, 491-503.

Niehrs, C., Steinbeisser, H. and de Robertis, E. M. (1994). Mesodermal

Patterning by a Gradient of the Vertebrate Homeobox Gene Goosecoid. Science

263, 817-820.

Nieukoop, P., Faber, J. (1967). Normal Table of Xenopus laevis (Daudin).

Amsterdam: North-Holland. ISBN: 0815318960

Ninomiya, H., Elinson, R. P. and Winklbauer, R. (2004). Antero-Posterior Tissue

Polarity Links Mesoderm Convergent Extension to Axial Patterning. Nature 430,

364-367.

Nister, M., Libermann, T. A., Betsholtz, C., Pettersson, M., Claesson-Welsh, L.,

Heldin, C. H., Schlessinger, J. and Westermark, B. (1988). Expression of

Messenger RNAs for Platelet-Derived Growth Factor and Transforming Growth

Factor-Alpha and their Receptors in Human Malignant Glioma Cell Lines. Cancer

  295

Res. 48, 3910-3918.

Oishi, I., Suzuki, H., Onishi, N., Takada, R., Kani, S., Ohkawara, B., Koshida, I.,

Suzuki, K., Yamada, G., Schwabe, G. C. et al. (2003). The Receptor Tyrosine

Kinase Ror2 is Involved in Non-Canonical Wnt5a/JNK Signalling Pathway. Genes

to Cells 8, 645-654.

Omura, T., Heldin, C. H. and Ostman, A. (1997). Immunoglobulin-Like Domain 4-

Mediated Receptor-Receptor Interactions Contribute to Platelet-Derived Growth

Factor-Induced Receptor Dimerization. The Journal of Biological Chemistry 272,

12676-12682.

Orr-Urtreger, A. and Lonai, P. (1992). Platelet-Derived Growth Factor-A and its

Receptor are Expressed in Separate, but Adjacent Cell Layers of the Mouse

Embryo. Development 115, 1045-1058.

Panakova Daniela, Sprong, H., Marois, E., Thiele, C. and Eaton, S. (2005).

Lipoprotein Particles are Required for Hedgehog and Wingless Signalling. Nature

435, 58-65.

Papan, C., Boulat, B., Velan, S. S., Fraser, S. E. and Jacobs, R. E. (2007a).

Formation of the Dorsal Marginal Zone in Xenopus Laevis Analyzed by Time-

Lapse Microscopic Magnetic Resonance Imaging. Dev. Biol. 305, 161-171.

Papan, C., Boulat, B., Velan, S. S., Fraser, S. E. and Jacobs, R. E. (2007b). Two-

Dimensional and Three-Dimensional Time-Lapse Microscopic Magnetic

  296

Resonance Imaging of Xenopus Gastrulation Movements using Intrinsic Tissue-

Specific Contrast. Developmental Dynamics 236, 494-501.

Pepinsky, R. B., Zeng, C., Wen, D., Rayhorn, P., Baker, D. P., Williams, K. P.,

Bixler, S. A., Ambrose, C. M., Garber, E. A., Miatkowski, K. et al. (1998).

Identification of a Palmitic Acid-Modified Form of Human Sonic Hedgehog. The

Journal of Biological Chemistry 273, 14037-14045.

Petrasek, Z. and Schwille, P. (2008). Precise Measurement of Diffusion

Coefficients using Scanning Fluorescence Correlation Spectroscopy. Biophys. J.

94, 1437-1448.

Piccolo, S., Sasai, Y., Lu, B. and Robertis, E. M. D. (1996). Dorsoventral

Patterning in Xenopus: Inhibition of Ventral Signals by Direct Binding of Chordin

to BMP-4. Cell 86, 589-598.

Piddini, E., Marshall, F., Dubois, L., Hirst, E. and Vincent, J. (2005). Arrow

(LRP6) and Frizzled2 Cooperate to Degrade Wingless in Drosophila Imaginal

Discs. Development 132, 5479-5489.

Plouhinec, J. and Robertis, E. M. D. (2007). Systems Biology of Embryonic

Morphogens. Molecular BioSystems 3, 454.

Popovici, C., Isnardon, D., Birnbaum, D. and Roubin, R. (2002). Caenorhabditis

Elegans Receptors Related to Mammalian Vascular Endothelial Growth Factor

  297

Receptors are Expressed in Neural Cells. Neurosci. Lett. 329, 116-120.

Porter, J. A., Ekker, S. C., Park, W. J., von Kessler, D. P., Young, K. E., Chen, C.

H., Ma, Y., Woods, A. S., Cotter, R. J., Koonin, E. V. et al. (1996). Hedgehog

Patterning Activity: Role of a Lipophilic Modification Mediated by the Carboxy-

Terminal Autoprocessing Domain. Cell 86, 21-34.

Prenzel, N., Zwick, E., Daub, H., Leserer, M., Abraham, R., Wallasch, C. and

Ullrich, A. (1999). EGF Receptor Transactivation by G-Protein-Coupled

Receptors Requires Metalloproteinase Cleavage of proHB-EGF. Nature 402,

884-888.

Ronnstrand, L. and Heldin, C. H. (2001). Mechanisms of Platelet-Derived Growth

Factor-Induced Chemotaxis. International Journal of Cancer 91, 757-762.

Raines, E. W. and Ross, R. (1992). Compartmentalization of PDGF on

Extracellular Binding Sites Dependent on Exon-6-Encoded Sequences. J. Cell

Biol. 116, 533-543.

Ramirez-Weber, F. A. and Kornberg, T. B. (1999). Cytonemes: Cellular

Processes that Project to the Principal Signaling Center in Drosophila Imaginal

Discs. Cell 97, 599-607.

Ridley, A. J. (2011). Life at the Leading Edge. Cell 145, 1012-1022.

Ries, J., Yu, S. R., Burkhardt, M., Brand, M. and Schwille, P. (2009). Modular

Scanning FCS Quantifies Receptor-Ligand Interactions in Living Multicellular

  298

Organisms. Nature Methods 6, 643-645.

Rohani, N., Canty, L., Luu, O., Fagotto, F. c. and Winklbauer, R. (2011).

EphrinB/EphB Signaling Controls Embryonic Germ Layer Separation by Contact-

Induced Cell Detachment. PLoS Biology 9, e1000597.

Rusakov, D. A. and Kullmann, D. M. (1998). Geometric and Viscous Components

of the Tortuosity of the Extracellular Space in the Brain. Proc. Natl. Acad. Sci. U.

S. A. 95, 8975-8980.

Saint-Jeannet, J. P., Karavanov, A. A. and Dawid, I. B. (1994). Expression of

Mesoderm Markers in Xenopus Laevis Keller Explants. Int. J. Dev. Biol. 38, 605-

611.

Sasaki, M., Kashima, M., Ito, T., Watanabe, A., Izumiyama, N., Sano, M.,

Kagaya, M., Shioya, T. and Miura, M. (2000). Differential Regulation of

Metalloproteinase Production, Proliferation and Chemotaxis of Human Lung

Fibroblasts by PDGF, Interleukin-1beta and TNF-Alpha. Mediators Inflamm. 9,

155-160.

Schambony, A. and Wedlich, D. (2007). Wnt-5A/Ror2 Regulate Expression of

XPAPC through an Alternative Noncanonical Signaling Pathway. Developmental

Cell 12, 779-792.

Schechtman, A. M. (1942). The mechanism of amphibian gastrulation. I.

Gastrulation-promoting interactions between various regions fo an anuran egg

  299

(Hyla regilla), Univ. Calif. Publ. Zool. 51, 1-39.

Schlie-Wolter, S., Ngezahayo, A. and Chichkov, B. N. (2013). The Selective Role

of ECM Components on Cell Adhesion, Morphology, Proliferation and

Communication in Vitro. Exp. Cell Res. 319, 1553-1561.

Scholpp, S. and Brand, M. (2004). Endocytosis Controls Spreading and Effective

Signaling Range of Fgf8 Protein. Current Biology 14, 1834-1841.

Schwank, G., Dalessi, S., Yang, S., Yagi, R., de Lachapelle, A. M., Affolter, M.,

Bergmann, S. and Basler, K. (2011). Formation of the Long Range Dpp

Morphogen Gradient. PLoS Biology 9, e1001111.

Selchow, A. and Winklbauer, R. (1997). Structure and Cytoskeletal Organization

of Migratory Mesoderm Cells from the Xenopus Gastrula. Cell Motil. Cytoskeleton

36, 12-29.

Sepich, D. S., Myers, D. C., Short, R., Topczewski, J., Marlow, F. and Solnica-

Krezel, L. (2000). Role of the Zebrafish Trilobite Locus in Gastrulation

Movements of Convergence and Extension. Genesis 27, 159-173.

Shih, J. and Keller, R. (1992). Cell Motility Driving Mediolateral Intercalation in

Explants of Xenopus Laevis. Development 116, 901-914.

Smith, E., Mitsi, M., Nugent, M. and Symes, K. (2009). PDGF-A Interactions with

Fibronectin Reveal a Critical Role for Heparan Sulfate in Directed Cell Migration

during Xenopus Gastrulation. Proc. Natl. Acad. Sci. U. S. A. 106, 21683-21688

  300

Smith, J. C., Price, B. M., Green, J. B., Weigel, D. and Herrmann, B. G. (1991).

Expression of a Xenopus Homolog of Brachyury (T) is an Immediate-Early

Response to Mesoderm Induction. Cell 67, 79-87.

Solnica-Krezel, L., Stemple, D. L. and Driever, W. (1995). Transparent Things:

Cell Fates and Cell Movements during Early Embryogenesis of Zebrafish.

Bioessays 17, 931-939.

Solnica-Krezel, L. and Sepich, D. S. (2012). Gastrulation: Making and Shaping

Germ Layers. Annu. Rev. Cell Dev. Biol. 10.1146/annurev-cellbio-092910-

154043

Solnica-Krezel, L. (2005). Conserved Patterns of Cell Movements during

Vertebrate Gastrulation. Current Biology 15, R213-28.

Soriano, P. (1997). The PDGF Alpha Receptor is Required for Neural Crest Cell

Development and for Normal Patterning of the Somites. Development 124, 2691-

2700.

Stephens, L., Milne, L. and Hawkins, P. (2008). Moving Towards a Better

Understanding of Chemotaxis. Current Biology 18, R485-94.

Stickney, H., Imai, Y., Draper, B., Moens, C. and Talbot, W. (2007). Zebrafish

bmp4 Functions during Late Gastrulation to Specify Ventroposterior Cell Fates.

Dev. Biol. 310, 71-84.

Stossel, T. P. (1993). On the Crawling of Animal Cells. Science 260, 1086-1094.

  301

Strigini, M. (2005). Mechanisms of Morphogen Movement. J. Neurobiol. 64, 324-

333.

Stroobant, P. and Waterfield, M. D. (1984). Purification and Properties of Porcine

Platelet-Derived Growth Factor. EMBO J. 3, 2963-2967.

Suzuki, A., Thies, R. S., Yamaji, N., Song, J. J., Wozney, J. M., Murakami, K. and

Ueno, N. (1994). A Truncated Bone Morphogenetic Protein Receptor Affects

Dorsal-Ventral Patterning in the Early Xenopus Embryo. Proc. Natl. Acad. Sci. U.

S. A. 91, 10255-10259.

Swaminathan, R., Hoang, C. P. and Verkman, A. S. (1997). Photobleaching

Recovery and Anisotropy Decay of Green Fluorescent Protein GFP-S65T in

Solution and Cells: Cytoplasmic Viscosity Probed by Green Fluorescent Protein

Translational and Rotational Diffusion. Biophys. J. 72, 1900-1907.

Symes, K. and Mercola, M. (1996). Embryonic Mesoderm Cells Spread in

Response to Platelet-Derived Growth Factor and Signaling by

Phosphatidylinositol 3-Kinase. Proc. Natl. Acad. Sci. U. S. A. 93, 9641-9644.

Tada, M. and Smith, J. C. (2000). Xwnt11 is a Target of Xenopus Brachyury:

Regulation of Gastrulation Movements via Dishevelled, but not through the

Canonical Wnt Pathway. Development 127, 2227-2238.

Tada, M. and Kai, M. (2009). Noncanonical Wnt/PCP Signaling during Vertebrate

Gastrulation. Zebrafish 6, 29-40.

  302

Teleman, A. A. and Cohen, S. M. (2000). Dpp Gradient Formation in the

Drosophila Wing Imaginal Disc. Cell 103, 971-980.

Thornberry, N. A. and Lazebnik, Y. (1998). Caspases: Enemies within. Science

281, 1312-1316.

Umulis, D., O'Connor, M. B. and Blair, S. S. (2009). The Extracellular Regulation

of Bone Morphogenetic Protein Signaling. Development 136, 3715-3728.

Unterseher, F., Hefele, J. A., Giehl, K., de Robertis, E. M., Wedlich, D. and

Schambony, A. (2004). Paraxial Protocadherin Coordinates Cell Polarity during

Convergent Extension Via Rho A and JNK. EMBO J. 23, 3259-3269.

Utoh, R., Shigenaga, S., Watanabe, Y. and Yoshizato, K. (2003). Platelet-

Derived Growth Factor Signaling as a Cue of the Epithelial-Mesenchymal

Interaction Required for Anuran Skin Metamorphosis. Developmental Dynamics

227, 157-169.

van Stry, M., Kazlauskas, A., Schreiber, S. L. and Symes, K. (2005). Distinct

Effectors of Platelet-Derived Growth Factor Receptor-Alpha Signaling are

Required for Cell Survival during Embryogenesis. Proc. Natl. Acad. Sci. U. S. A.

102, 8233-8238.

van Stry, M., McLaughlin, K. A., Ataliotis, P. and Symes, K. (2004). The

Mitochondrial-Apoptotic Pathway is Triggered in Xenopus Mesoderm Cells

Deprived of PDGF Receptor Signaling during Gastrulation. Dev. Biol. 268, 232-

  303

242.

Vignais, L., Nait Oumesmar, B. and Baron-Van Evercooren, A. B. (1995). PDGF-

Alpha Receptor is Expressed by Mature Neurones of the Central Nervous

System. Neuroreport 6, 1993-1996.

Wacker, S., Grimm, K., Joos, T. and Winklbauer, R. (2000). Development and

Control of Tissue Separation at Gastrulation in Xenopus. Dev. Biol. 224, 428-439.

Wallingford, J. B. and Harland, R. M. (2001). Xenopus Dishevelled Signaling

Regulates both Neural and Mesodermal Convergent Extension: Parallel Forces

Elongating the Body Axis. Development 128, 2581-2592.

Wallingford, J. B., Rowning, B. A., Vogeli, K. M., Rothb\acher, U., Fraser, S. E.

and Harland, R. M. (2000). Dishevelled Controls Cell Polarity during Xenopus

Gastrulation. Nature 405, 81-85.

Walters, M. J., Wayman, G. A. and Christian, J. L. (2001). Bone Morphogenetic

Protein Function is Required for Terminal Differentiation of the Heart but Not for

Early Expression of Cardiac Marker Genes. Mech. Dev. 100, 263-273.

Wartlick, O., Kicheva, A. and Gonzalez-Gaitan, M. (2009). Morphogen Gradient

Formation. Cold Spring Harbor Perspectives in Biology 1, a001255.

Weber, S. and Saftig, P. (2012). Ectodomain Shedding and ADAMs in

Development. Development 139, 3693-3709.

  304

Wei, S., Xu, G., Bridges, L. C., Williams, P., White, J. M. and DeSimone, D. W.

(2010). ADAM13 Induces Cranial Neural Crest by Cleaving Class B Ephrins and

Regulating Wnt Signaling. Developmental Cell 19, 345-352.

White, J. M. (2003). ADAMs: Modulators of Cell-Cell and Cell-Matrix Interactions.

Curr. Opin. Cell Biol. 15, 598-606.

Wilkinson, D. G., Bhatt, S. and Herrmann, B. G. (1990). Expression Pattern of the

Mouse T Gene and its Role in Mesoderm Formation. Nature 343, 657-659.

Willnow, T. E., Hammes, A. and Eaton, S. (2007). Lipoproteins and their

Receptors in Embryonic Development: More than Cholesterol Clearance.

Development 134, 3239-3249.

Wilson, P. A., Oster, G. and Keller, R. (1989). Cell Rearrangement and

Segmentation in Xenopus: Direct Observation of Cultured Explants. Development

105, 155-166.

Wilson, P. and Keller, R. (1991). Cell Rearrangement during Gastrulation of

Xenopus: Direct Observation of Cultured Explants. Development 112, 289-300.

Winklbauer, R. (1990). Mesodermal Cell Migration during Xenopus Gastrulation.

Dev. Biol. 142, 155-168.

Winklbauer, R. (1998). Conditions for Fibronectin Fibril Formation in the Early

Xenopus Embryo. Developmental Dynamics 212, 335-345.

  305

Winklbauer, R. and Damm, E. W. (2011). Internalizing the Vegetal Cell Mass

before and during Amphibian Gastrulation: Vegetal Rotation and Related

Movements. Wiley Interdisciplinary Reviews: Developmental Biology 1, 301-306.

Winklbauer, R. and Keller, R. E. (1996). Fibronectin, Mesoderm Migration, and

Gastrulation in Xenopus. Dev. Biol. 177, 413-426.

Winklbauer, R. and Nagel, M. (1991). Directional Mesoderm Cell Migration in the

Xenopus Gastrula. Dev. Biol. 148, 573-589.

Winklbauer, R., Nagel, M., Selchow, A. and Wacker, S. (1996). Mesoderm

Migration in the Xenopus Gastrula. Int. J. Dev. Biol. 40, 305-311.

Winklbauer, R. and Schurfeld, M. (1999). Vegetal Rotation, a New Gastrulation

Movement Involved in the Internalization of the Mesoderm and Endoderm in

Xenopus. Development 126, 3703-3713.

Winklbauer, R. and Selchow, A. (1992). Motile Behavior and Protrusive Activity of

Migratory Mesoderm Cells from the Xenopus Gastrula. Dev. Biol. 150, 335-351.

Wolpert, L. (1969). Positional Information and the Spatial Pattern of Cellular

Differentiation. J. Theor. Biol. 25, 1-47.

Wong, L. L. and Adler, P. N. (1993). Tissue Polarity Genes of Drosophila

Regulate the Subcellular Location for Prehair Initiation in Pupal Wing Cells. J.

Cell Biol. 123, 209-221.

  306

Xu, G., Wei, S., White, J. M. and DeSimone, D. W. (2012). Identification and

Characterization of ADAM41, a Novel ADAM Metalloproteinase in Xenopus. Int.

J. Dev. Biol. 56, 333-339.

Yamanaka, Y., Tamplin, O. J., Beckers, A., Gossler, A. and Rossant, J. (2007).

Live Imaging and Genetic Analysis of Mouse Notochord Formation Reveals

Regional Morphogenetic Mechanisms. Developmental Cell 13, 884-896.

Yamashita, S., Miyagi, C., Carmany-Rampey, A., Shimizu, T., Fujii, R., Schier, A.

F. and Hirano, T. (2002). Stat3 Controls Cell Movements during Zebrafish

Gastrulation. Developmental Cell 2, 363-375.

Yan, D. and Lin, X. (2009). Shaping Morphogen Gradients by Proteoglycans.

Cold Spring Harbor Perspectives in Biology, a002493.

Yang, X., Chrisman, H. and Weijer, C. (2008). PDGF Signalling Controls the

Migration of Mesoderm Cells during Chick Gastrulation by Regulating N-Cadherin

Expression. Development 135, 3521-3530.

Yin, C., Kiskowski, M., Pouille, P. A., Farge, E. and Solnica-Krezel, L. (2008).

Cooperation of Polarized Cell Intercalations Drives Convergence and Extension

of Presomitic Mesoderm during Zebrafish Gastrulation. J. Cell Biol. 180, 221-232.

Yu, J. C., Mahadevan, D., LaRochelle, W. J., Pierce, J. H. and Heidaran, M. A.

(1994). Structural Coincidence of Alpha PDGFR Epitopes Binding to Platelet-

Derived Growth Factor-AA and a Potent Neutralizing Monoclonal Antibody. The

  307

Journal of Biological Chemistry 269, 10668-10674.

Yu, S. R., Burkhardt, M., Nowak, M., Ries, J., Petr\'asek, Z., Scholpp, S.,

Schwille, P. and Brand, M. (2009). Fgf8 Morphogen Gradient Forms by a Source-

Sink Mechanism with Freely Diffusing Molecules. Nature 461, 533-536.

Zhou, S., Lo, W., Suhalim, J. L., Digman, M. A., Gratton, E., Nie, Q. and Lander,

A. D. (2012). Free Extracellular Diffusion Creates the Dpp Morphogen Gradient of

the Drosophila Wing Disc. Current Biology 22, 668-675.

  308