SUBLINEAGE-SPECIFIC CUES REQUIRED FOR EARLY AND LATER NEURAL CREST DEVELOPMENT IN THE ZEBRAFISH, DANIO RERIO
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the
Graduate School of The Ohio State University
By
Brigitte Louise Arduini, B.S.
*****
The Ohio State University 2005
Dissertation Committee: Approved by Dr. Paul D. Henion, Advisor
Dr. Helen Chamberlin Advisor Graduate Program in Molecular Genetics Dr. Mark Seeger
Dr. Harald Vaessin
ABSTRACT
The neural crest (NC) of vertebrate animals gives rise to many derivatives,
including pigment cells, peripheral neurons, glia and elements of the craniofacial
skeleton. The generation of NC-derived cells has been studied extensively to
elucidate mechanisms involved in cell fate specification, differentiation, migration
and survival. Zebrafish trpm7/touchtone, endzone, and foxd3/sympathetic
mutation1 are discrete loci required by subsets of neural crest derivatives.
Severe mutant alleles of the divalent cation channel gene trpm7 are lethal
and cell-autonomously cause reductions in the number and size of crest-derived
melanophores, while sparing other NC lineages. The deficit in cell numbers can
be accounted for at least in part by cell death of melanophore precursors.
Pleiotropic effects in non-crest derived tissues, including altered order of bone
ossification and kidney dysfunction, are observed in homozygotes for semi-viable
alleles.
Mutations in endzone affect all three pigment cell lineages found in
zebrafish. Normally large and stellate, melanophores and xanthophores take on
a rounded, punctate appearance in these mutants. Iridiphores are also reduced in size. These three cell types appear to be similarly reduced in numbers in endzone mutant embryos. While neuronal, glial and ectomesenchymal
ii derivatives of the NC appear to be normal in endzone homozygotes, the non- crest-derived pigmented retinal epithelium is developmentally delayed, pointing to pleiotropism for these mutations, as well. Both trpm7 and endzone act
relatively late during chromatophore development; accordingly molecular
analyses reveal no defects in the early NC cell populations of these mutants.
foxd3sym1 affects multiple derivatives within the NC. FoxD3 is required for
sympathetic and sensory neuron development, but appears to be dispensable for
chromatophore lineages. Anterior elements of the craniofacial skeleton are
reduced, and posterior elements are missing, indicating a role for foxd3 in axial
patterning of the pharyngeal arches. Expression of critical transcription factors in
early crest cells and crest migration are both abrogated, although the
premigratory NC population appears to be induced normally. Simultaneous
abrogation of Foxd3 and Tfap2α function leads to loss of all chromatophores and
most craniofacial cartilages.
Collectively, these data suggest that subpopulations with distinct genetic
requirements exist within the early neural crest and its later sublineages.
iii
DEDICATION
To Mom, for all that you do.
iv
ACKNOWLEDGEMENTS
I would like to thank Paul Henion for excellent guidance and training, for fresh garden vegetables, and for having a sense of humor. I am grateful that his door was (almost) always open, that he taught when needed, and yet allowed me to find my own way. The importance of the mentor-student relationship cannot be overestimated; I am fortunate to have found this one both very challenging and very rewarding.
Thank you also to my lab mates, An Min, Marsha Lucas, Myron Ignatius,
Gao Juan, Smitha Malireddy, Roopa Nambiar, Natalie Gentry, Luo Rushu and
Matt Frieda, for discussions thoughtful and capricious, and for their friendship. I am appreciative of Christine Beattie and the members of her lab, Louise Rodino-
Klapac, Michelle McWhorter, Tessa Carrel, Emily Tansey and Wang Chunping, for sharing expertise, laughter and a penchant for Indian food. I would like to thank Michelle Gray and Anil Challa, who were indispensable to my early training and to my personal growth.
To the members of my committee, Helen Chamberlin, Mark Seeger and
Amanda Simcox, I am grateful for their highly constructive roles in my scientific development, as well as their significant time and efforts on my behalf. I am also greatly appreciative of Harald Vaessin for stepping in when needed. Thank you v to Heithem El-Hodiri for enthusiasm for science and for people, and for organizing decompression time. Very special thanks to Cliff Gebhardt and Martin
Gillard, from whom I first learned to love biology, and to Rhonda Curtis, for so much respect and encouragement. I must also thank the staff of the Molecular
Neurobiology Center and the Department of Molecular Genetics, who routinely uncomplicated the practical necessities of science and education.
I thank my family for innumerable contributions, my mother Judy Leonhart, my father Bino Arduini, my sister Noelle Arduini, and my grandparents Brigitte
Heckmann, Bino and Louise Arduini. Their love, support and help have enabled my achievements at every level.
vi
VITA
March 12, 1977...... Born - Oswego, New York
May 30, 1999...... B.S. Biology Cornell University
1999 - 2000...... Teaching Asst. The Ohio State University.
2000 - 2005...... Research Asst. The Ohio State University.
PUBLICATIONS
Research Publications
1. Luo, R., An, M., Arduini, B.L., and Henion, P.D. (2001) Specific pan-neural crest expression of zebrafish crestin throughout embryonic development. Dev Dyn 220(2): 169-174.
2. Arduini, B.L. and Henion, P.D. (2004) Melanogenic sublineage-specific requirement for zebrafish touchtone during neural crest development. Mech Dev 121(11):1353-64.
3. Elizondo, M.R., Arduini, B.L., Paulsen, J., MacDonald, E.L., Sabel, J.L., Henion, P.D., Cornell, R.A, Parichy, D.M. (2005) Defective skeletogenesis with kidney stone formation in dwarf zebrafish mutant for trpm7. Curr Biol 15(7):667-71.
FIELDS OF STUDY
Major Field: Molecular Genetics
vii
TABLE OF CONTENTS
P a g e
Abstract ...... ii
Dedication ...... iv
Acknowledgements ...... v
Vita ...... vii
List of Tables...... x
List of Figures ...... xi
List of Abbreviations...... xiv
Chapter 1: Introduction...... 1 Neural plate border induction ...... 3 Epithelial-to-mesenchymal transition and neural crest migration...... 9 Specification of sublineages within the neural crest...... 13 Conclusion...... 22 Tables and Figures...... 24
Chapter 2: Specific pan-neural crest expression of zebrafish crestin throughout embryonic development...... 28 Abstract...... 28 Introduction...... 29 Results and Discussion ...... 31 Materials and Methods...... 35 Tables and Figures ...... 37
Chapter 3: Melanogenic sublineage-specific requirement for zebrafish touchtone during neural crest development...... 42 Abstract...... 42 Introduction...... 43
viii Results...... 48 Discussion...... 58 Materials and Methods...... 65 Tables and Figures...... 71
Chapter 4: Defective skeletogenesis with kidney stone formation in dwarf zebrafish mutant for trpm7...... 81 Abstract...... 81 Results and Discussion...... 82 Materials and Methods...... 87 Tables and Figures...... 93
Chapter 5: Zebrafish endzone regulates neural crest-derived chromatophore morphology and differentiation ...... 106 Abstract...... 106 Introduction...... 107 Results...... 112 Discussion...... 121 Materials and Methods...... 125 Tables and Figures...... 130
Chapter 6: sympathetic mutation 1 encodes zebrafish foxd3 and is differentially required for early neural crest development...... 140 Abstract...... 140 Introduction...... 141 Results...... 143 Discussion...... 156 Materials and Methods...... 163 Tables and Figures...... 166
Chapter 7: Genetic interaction of zebrafish foxd3 and tfap2 in neural crest derived pigment and craniofacial development...... 181 Abstract...... 181 Introduction...... 182 Results...... 184 Discussion...... 189 Materials and Methods...... 193 Tables and Figures...... 196
Chapter 8: Discussion...... 202 Sublineage-specific genetic requirements within the neural crest. . . . . 202 Genetic heterogeneity within distinct neural crest sublineages...... 205
List of References...... 208
ix
LIST OF TABLES
Table Page
2.1 Molecular markers for neural crest and neural crest-derived cells . . . . . 37
3.1 Other neural crest derivatives are normal in tct mutant embryos ...... 71
3.2 Melanophores are reduced in tct mutants...... 72
3.3 Depletion of melanoblasts in tct embryos...... 73
3.4 tct acts cell autonomously with respect to melanophore development. . 74
4.1 Relative ossification timing...... 93
5.1 Other neural crest derivatives are normal in enz mutant embryos. . . . .130
5.2 Melanophore and iridiphore numbers are significantly reduced in enz mutant embryos...... 131
5.3 enz acts cell autonomously with respect to melanophore development.132
x
LIST OF FIGURES
Figure Page
1.1 Subsets of neural crest cells arise at distinct axial levels ...... 24
1.2 Schematic of vertebrate neurulation ...... 25
1.3 Zebrafish neural crest development...... 26
1.4 Migration of zebrafish trunk neural crest cells...... 27
2.1 Expression of crestin from the 3-somite stage through 72 hpf in whole- mount preparations and transverse sections...... 38
2.2 Co-expression of neural crest, neural crest sublineage, and differentiated derivative molecular markers in subsets of crestin-expressing cells in whole-mounts and transverse sections...... 39
2.3 The mutant nisakara lacks most iridiphores and displays an early crestin phenotype...... 41
3.1 Visible phenotype of live touchtone mutant embryos...... 75
3.2 tct selectively affects melanophores within the neural crest lineage . . . 76
3.3 Melanophores are reduced in number and size in tct mutant embryos. .77
3.4 Melanoblast numbers are reduced in tct mutant embryos in the second day of development...... 78
3.5 Melanoblasts undergo apoptosis in tct mutant embryos...... 79
3.6 tct is required cell autonomously for melanophore development...... 80
4.1 Retarded growth and altered body proportions in nutria (tct) mutant zebrafish...... 95
xi 4.2 Identification of trpm7 as the gene mutated in tct (nutria) mutant zebrafish ...... 96
4.3 Rescue of trpm7 embryonic defects with divalent cation supplementation ...... 97
4.4 trpm7 expression in wild-type larvae and kidney stone formation in trpm7 mutants ...... 99
4.5 trpm7j124e2 mutants exhibit dramatic differences in skeletal development in comparison to wild-type ...... 100
4.6 Altered sequence and timing of skeletal ossification in nutria compared to wild-type...... 102
4.7 Functional-anatomical units and bone locations examined for ossification sequence and timing...... 104
4.8 Growth retardation alone does not result in precocious ossification. . . 105
5.1 All three neural crest-derived chromatophore cell types are affected by enz mutations...... 133
5.2 enz larvae and adults are undersized compared to wild-type siblings. . 134
5.3 enz selectively affects chromatophores within the neural crest lineage.135
5.4 Chromatophore precursors appear to normal in enz homozygotes at 24 hpf...... 136
5.5 Melanophore cell morphology changes in enz mutant embryos...... 137
5.6 Xanthophores are qualitatively reduced in number and size in enz homozygotes...... 138
5.7 enz acts cell autonomously with respect to melanophore development.149
6.1 Defects in trunk and vagal neural crest derived peripheral neurons, cranial ganglion neurons, glia and lateral line glia in sym1 mutant embryos. . . 166
6.2 Normal development of neural crest-derived chromatophores in sym1 mutants...... 168
6.3 Abnormal pharyngeal arch development in sym1 mutants...... 169
xii 6.4 Abnormal sox9a expression in sym1 mutants...... 170
6.5 sym1 is an inactivating mutation in foxd3...... 171
6.6 Normal somite and floorplate development in wild-type and sym1 mutant embryos ...... 172
6.7 Molecular defects in neural crest specification in sym1 mutants...... 173
6.8 Gene expression abnormalities in premigratory neural crest populations in sym1 mutants...... 174
6.9 Normal development of Rohon-Beard sensory neurons, neural plate and neural plate border in sym1 mutants...... 175
6.10 Gene expression abnormalities in premigratory and early migrating neural crest cells in sym1 mutants...... 176
6.11 Neural crest gene expression in 18-somite sym1 mutants...... 178
6.12 Abnormal cell migration and cell death in sym1 mutant embryos...... 179
7.1 Reduction in total number of melanophores in foxd3sym1 mutant embryos compared to wild-type siblings following sox10 morpholino injection. . .196
7.2 Melanophores are absent in 28 hpf foxd3sym1-tfap2α mutant-morphant embryos...... 197
7.3 Visible phenotypes in live foxd3 sym1-tfap2α mutant-morphant larvae. . .198
7.4 Chromatophore precursors fail to be specified in the absence of foxd3 and tfap2α function...... 199
7.5 Craniofacial cartilages are synergistically affected by simultaneous loss of foxd3 and tfap2α function...... 200
7.6 sox9b expression is reduced in foxd3sym1 mutants compared to wild-type siblings...... 201
xiii
LIST OF ABBREVIATIONS
α alpha
~ approximately
+ positive
AP alkaline phosphatase
AP-2 see Tfap2
β beta bh basihyal
BMP Bone Morphogenetic Protein c centrum
°C degrees Celsius
Ca calcium
CaCl2 calcium chloride
Cad7 Cadherin7
CaP Caudal Primary motor neurons cb ceratobranchial cDNA complementary deoxyribonucleic acid cM centi-Morgans
CNS central nervous system xiv cls colourless
Col2a1 Collagen2-alpha-1 cs corpuscles of Stannius ct collecting tubules, mesonephric ctn crestin
Dct Dopachrome tautomerase
DIG digoxygenin
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid dpf days postfertilization
Dlx2 Distalless homeobox 2
DRG Dorsal root ganglia
E epidermis
ECM extracellular matrix ect ectopterygoid
Edn Endothelin
Ednr Endothelin receptor eh epihyal
EMT epithelial-mesenchymal transition en entopterygoid
ENU ethyl-N-nitrosourea enz endzone
EP early pressure
xv EST expressed sequence tag f frontal
FGF Fibroblast Growth Factor
Fkd6 Forkhead6; currently FoxD3
Flu fluorescein
FP-1 Floor plate-1
Fzd Frizzled ha hemal arches hpf hours postfertilization hm hyomandibula hr hours hyp hypurals
IR immuno-reactive
ISH in situ hybridization
Isl Islet
KCl potassium chloride l lepidotrichia
LFD lysinated fluorescein dextran li liver low lockjaw
LRD lysinated rhodamine dextran
μ micro
M molar
xvi mAb monoclonal antibody
Mitf Microphthalmia-related transcription factor
Mg magnesium
MgCl2 magnesium chloride ml milliliter
MO morpholino mob mont blanc mpt metapterygoid mRNA messenger ribonucleic acid mt mesonephric tubules
MW molecular weight my myotome
N notochord na neural arches
NaCl sodium chloride
NC neural crest
N-CAM Neural-Cell Adhesion Molecule
NCC neural crest cell ng nanograms
Ngn Neurogenin nis nisakara nl nanoliter
NLS nuclear localization signal
xvii NT neural tube
NP neural plate
NPB neural plate border
Pax3 paired box domain 3
PBS phosphate-buffered saline
PBT phosphate-buffered saline-Tween20
PCR polymerase chain reaction
PFA paraformaldehyde
PM presomitic mesoderm
POD chemiluminescent peroxidase
PRE pigmented retinal epithelium ps parasphenoid
PSNS peripheral nervous system
PTHR1 Parathyroid hormone receptor type 1
PTU 1-Phenyl-2-thiourea q quadrate r rhombomere
RB Rohon Beard sensory neurons
RNA ribonucleic acid
RT room temperature
S somite sb swim bladder snai1b snail1b
xviii Sox SRY-related homeobox
SSLP simple sequence length polymorphism sym sympathetic sym1 sympathetic mutation 1 tct touchtone
Tfap2 Transcription factor-activator protein 2
TH Tyrosine hydroxylase
TRP Transient receptor potential
Trpm1 Transient receptor potential-melastatin 1
Trpm7 Transient receptor potential-melastatin 7
TUNEL Terminal dUTP Nicked End Labeling
Tyr Tyrosinase u urostyle and ural 1 + 2
UTR untranslated region v vertebral column
VaP Ventral Primary motor neurons
VEGF Vascular-endothelial growth factor
Wnt Wingless-integrated protein wt wild-type
Xdh Xanthine dehydrogenase
Y tyrosine
xix
CHAPTER 1
INTRODUCTION
The specification, differentiation, proliferation and survival of diverse cell types are of central importance to metazoan development. Through these processes, single cells are transformed into multicellular organisms incorporating various cell populations, tissues and organ systems. As such, the study of these processes is fundamental to our knowledge not only of our own development, but that of members in closely- and distantly-related taxa, as well.
The neural crest is a transient vertebrate embryonic cell population that gives rise to many derivatives, including pigment cells, peripheral neurons, glia and elements of the craniofacial skeleton (Le Douarin and Kalcheim, 1999).
Neural crest cells (NCCs) are induced at the border between neural and non- neural ectoderm, undergo an epithelial-to-mesenchymal transition, and migrate extensively throughout the embryo (Aybar and Mayor, 2002; Huang and Saint-
Jeannet, 2004; Knecht and Bronner-Fraser, 2002; Nieto, 2001). Due to these parallels between neural crest development and embryonic development in general, the neural crest has been a subject of intense interest in the biological sciences. In addition, attainment of a well-defined head, to which the neural crest contributes heavily, is thought to be a key factor in the evolution of vertebrates
1 from a nonvertebrate chordate ancestor (Gans and Northcutt, 1983; Kulesa et al.,
2004). Further, defects in neural crest development lead to a variety of syndromic
afflictions in humans, termed neurocristopathies (Amiel and Lyonnet, 2001;
Bolande, 1997; Nakamura, 1995; Nemecek et al., 2003). Thus, the study of
neural crest development also has direct medical relevance.
Early studies in the chick identified the neural crest as a population of cells that arises from the future central nervous system (CNS) and remains briefly at the “crest” of the neural tube, beneath the overlying ectoderm (His, 1868;
Marshall, 1879). Subsequently, it has been determined that this cell population is induced at the border of the neural plate, the future CNS, that is juxtaposed to non-neural ectoderm, which will become epidermis (see below). Neural crest cells are formed along the length of the embryo except at the most anterior forebrain levels, and can be roughly divided into groups by the axial level at which they arise and the derivatives they form (Le Douarin and Kalcheim, 1999).
Trunk neural crest, generated posterior to the hindbrain, gives rise to neurons
and glia of the peripheral nervous system and pigment cells. Cranial, or cephalic,
crest is generated from the posterior forebrain to the hindbrain and also gives
rise to these cell types, in addition to elements of the craniofacial skeleton (Le
Douarin and Kalcheim, 1999). Vagal, sacral, and cardiac crest distinctions have
been made most clearly in the chick. The vagal neural crest, considered to
develop at the level of somites 1 to 7, and the sacral crest posterior to somite 28,
give rise to neurons and glia of the gut, comprising the enteric nervous system
(Le Douarin and Teillet, 1973; Pomeranz et al., 1991). Cardiac crest is generated
2 at the level of somites 1-3, and contributes to derivatives attributed to the cranial
neural crest, as well as the circulatory system in the region of the heart (Kirby,
1987; Kirby and Waldo, 1990). Further, cells of neural crest origin have been
shown to contribute to the adrenal medulla and other endocrine cell types, as
well as connective tissues of the cornea, tooth papillae, thymus, thyroid and
pituitary glands, and the dermis (Le Douarin and Kalcheim, 1999).
Neural plate border induction
In vivo, the outermost embryonic germ layer, the ectoderm, gives rise to
epidermis and the central nervous system (Figure 1.2). Dorsal ectoderm
develops into neural tissue, while ventral ectoderm becomes epidermis (Gilbert,
c2000). Initially, prospective neural ectoderm is relatively flat and continuous with
non-neural (future epidermal) ectoderm, and is referred to as the neural plate.
During the process of neurulation, the neural plate folds upon itself to form the
neural tube, or future spinal cord. The medial neural plate eventually lies ventrally
and the lateral edges of the neural plate, termed the neural folds, come together dorsally. The neural folds form at the region of the ectoderm between the
developing neural plate and epidermis, and are also referred to as the neural
plate border (NPB). Convergence of the neural folds toward the midline first
occurs rostrally within the embryo, and proceeds caudally over time. As the most
prominent and well-studied cell population within the NPB, induction of the neural
crest is often used to infer induction of the NPB. Neural plate border induction
begins during gastrulation, and neural crest and other NPB cells are already
3 specified in the neural folds as neurulation is occurring (Aybar and Mayor, 2002).
At this time, it remains unclear what the earliest steps of NPB induction involve; induction of the NPB is functionally defined in various organisms by the observation of differentiated neural crest derivatives or by the detection of genes known to be expressed at the NPB. At a given axial level, both epidermis and
neural ectoderm have been shown to give rise to neural crest cells in a variety of
vertebrate organisms (Liem et al., 1995; Mancilla and Mayor, 1996; Moury and
Jacobson, 1990; Selleck and Bronner-Fraser, 1995). Further, the juxtaposition of
these two tissues is sufficient to induce neural crest (Mancilla and Mayor, 1996;
Moury and Jacobson, 1989). The precise molecular mechanism of neural plate
border/neural crest induction has not yet been established, but Bone
Morphogenetic Protein (BMP), Fibroblast Growth Factor and Wnt signaling have
all been implicated in this process.
Evidence from Xenopus and zebrafish suggests that antagonism of BMP
signaling is required to pattern the ectoderm during gastrulation (Luo et al.,
2003b; Luo et al., 2001b; Marchant et al., 1998; Nguyen et al., 1998; Nguyen et
al., 2000; Wilson et al., 1997). BMP inhibitors produced by the dorsal lip of the
blastopore (the organizer), such as Chordin and Noggin, diffuse ventrally,
resulting in a gradient of BMP signaling. Dorsal regions of low BMP activity
become neural plate, while more ventral areas of high BMP activity are specified
as epidermis. The NPB forms between these two tissues where intermediate
levels of BMP signaling are hypothesized to exist. Support for this model comes
from Xenopus explant experiments, in which inhibition of endogenous BMP
4 signaling dose-dependently induces expression of an early neural crest gene,
slug (LaBonne and Bronner-Fraser, 1998; Marchant et al., 1998). Reciprocally,
overexpression of BMPs in embryos reduces slug expression (Mayor et al.,
1995). The neural crest domain of the ectoderm is mispatterned in zebrafish mutants swirl/bmp2b, snailhouse/bmp7 and somitabun/smad5, providing genetic evidence that intermediate levels of BMP signaling are important for NPB formation (Nguyen et al., 1998; Nguyen et al., 2000). However, data from amniotes are less straightforward. Homozygous BMP4 mutant mice, when they survive to neural fold stages, have some neural crest cells (Winnier et al., 1995) and mice carrying homozygous mutations in individual BMP antagonists do not exhibit neural crest defects (Matzuk et al., 1995; McMahon et al., 1998). In chick,
cells expressing BMP antagonists appear to suppress neural crest markers only
when injected into closing neural tube, but not when injected near the open
neural tubes at early time points (Selleck et al., 1998). BMP antagonism has
further been shown to be sufficient to induce neural crest only at the neural plate
border and not elsewhere in the ectoderm (Streit and Stern, 1999). Together,
these data have been interpreted to implicate BMP signaling in the maintenance,
rather than the induction, of the NPB. However, the possibility remains that in
these systems functionally redundant molecules are compensating for the
abrogation of specific factors being tested. Further, a recent study in Xenopus
using conditional knockdown of BMP signaling suggests that specification of the
neural plate and the neural crest are separable during development (Wawersik et
al., 2005). Importantly, these experiments show that the timing of BMP
5 modulation is critical for effects on the neural plate, neural crest, or both, and
may affect experimental outcomes. There appear to be specific temporal windows during which these tissues are competent to respond to BMP signaling or antagonism. In addition, that study and a recent report from zebrafish both suggest that BMP antagonists produced by the mesoderm, such as Chordin,
Noggin and Follistatin, may not be required for exclusion of BMP from the dorsal ectoderm (Ragland and Raible, 2004; Wawersik et al., 2005). Thus, while it has classically been hypothesized that signals from the organizer set up a BMP
gradient in the ectoderm, this process may actually begin earlier, prior to or at the
very initial steps of gastrulation (Ragland and Raible, 2004; Wawersik et al.,
2005).
Whether or not BMP antagonism plays an instructive role in NPB
formation or maintenance, there is evidence that FGF signaling potentiates the
response of NPB cells to BMP signaling (Gajavelli et al., 2004; Streit and Stern,
1999; Wawersik et al., 2005), and that paraxial mesoderm is the embryonic
source of FGF signals important for NPB development (Monsoro-Burq et al.,
2003). In addition, some experiments suggest that FGFs alone may be capable of inducing neural crest fates, albeit incompletely and/or transiently (Mayor et al.,
1997; Monsoro-Burq et al., 2003).
Another signaling pathway that is used widely in development and has
been shown to play key roles in neural crest development is the Wnt pathway.
Several Wnts are expressed at appropriate times and locations to be involved in
NPB induction in a variety of vertebrates, including Wnt6, Wnt7b and Wnt8
6 (Bouillet et al., 1996; Cauthen et al., 2001; Chang and Hemmati-Brivanlou, 1998;
Garcia-Castro et al., 2002; Hume and Dodd, 1993; Schubert et al., 2002). Still other Wnts are expressed in the dorsal neural tube, and have been hypothesized to play a role in expansion of the neural crest population or specification of ectodermal cells to a neural crest fate, if not in induction itself (Brault et al., 2001;
Dunn et al., 2005; Ikeya et al., 1997). In addition, Wnt receptors Frizzled3 (Fzd3) and Frizzled9 (Fzd9) are expressed in the neural folds (Borello et al., 1999;
Deardorff et al., 2001; Shi et al., 1998; Van Raay et al., 2001). Wnts can induce neural crest markers in both neural and epidermal tissue, and mutant and protein knockdown analyses reveal requirements for various Wnt molecules for normal neural crest development (Huang and Saint-Jeannet, 2004). Inactivation of β-
catenin, a downstream Wnt effector, results in neural crest defects in mice (Brault et al., 2001), while overexpression of Fzd3 and β-catenin can induce expression
of the early neural crest gene, slug, in Xenopus animal caps (Deardorff et al.,
2001; LaBonne and Bronner-Fraser, 1998). In addition, the slug promoter contains binding sites for the Wnt effector Tcf/Lef, suggesting that it is a direct target of Wnt signaling. Together, these data clearly suggest that canonical Wnt signaling is important for neural crest induction, although it remains to be seen which Wnt ligands and receptors are specifically involved in vivo.
No single pathway can account for all the functions necessary to pattern
the developing ectoderm. Presently, it appears that integration of these signaling
pathways and others results in the induction of the neural plate border, including
the neural crest, at the lateral edge of the prospective central nervous system. 7 Further, neural crest is not the only cell type that arises from the neural plate
border. Rohon-Beard sensory neurons, found transiently in the dorsal spinal cord
of anamniote vertebrates such as frogs and fish, also arise from the neural plate
border (Lamborghini, 1980; Metcalfe et al., 1990). In fact, Rohon-Beard
progenitors are intermingled and form an equivalence group with neural crest
precursors in the NPB of zebrafish embryos (Artinger et al., 1999; Cornell and
Eisen, 2000), requiring still another signaling pathway to specify their respective fates. Delta-Notch signaling mediates cell fate decisions within this group, with high Delta-expressing cells becoming Rohon-Beard neurons and cells transducing Notch signals contributing to the neural crest (Artinger et al., 1999;
Cornell and Eisen, 2000).
Neural crest cells are visualized within the NPB by expression of specific
genes, particularly transcription factors such as pax3, snail, slug, foxd3, tfap2,
sox9 and sox10 (Knecht and Bronner-Fraser, 2002). Expression of these genes is observed in two bilateral stripes during early somitogenesis stages, representing the neural folds (for example, see Figure 1.3). Induction of neural
crest-specific gene expression, like neurulation, begins in anterior regions of the
embryo and extends posteriorly over time. As neurulation occurs, these stripes
converge on the midline and become a single dorsal domain upon neural tube
closure. Some of these genes continue to be expressed in migrating neural crest cells, and some are re-expressed in specific neural crest sublineage progenitor populations. However, the relationships of these transcription factors to the
upstream signaling pathways that induce neural crest are still being elucidated.
8 Further, only a small subset of the likely downstream genes regulated by these transcription factors has been identified. More work is needed to determine the links between neural crest-inducing molecules and subsequent behavior and gene expression within the neural crest population.
Epithelial-to-mesenchymal transition and neural crest migration
Following induction and specification within the neural plate border, neural crest cells undergo an epithelial-mesenchymal transition (EMT) from the organized sheets of the neural tissue to individual cells capable of migration over long distances (Gilbert, c2000; Nieto, 2001). In some vertebrates, this involves delamination from the neural tube, while in others it appears that neural crest cells may already be outside the neural folds as they converge on the midline.
Evidence from cell culture and chick in vivo studies suggests that the EMT is separable from neural crest induction, although these events are coordinated during development (Newgreen and Minichiello, 1995; Sela-Donenfeld and
Kalcheim, 1999). As with neurulation and neural crest induction, the EMT and
NCC migration are first initiated in the anterior of the embryo, and proceed more caudally as development continues.
At the molecular level, many different pathways are incorporated to achieve the complex result of EMT and migration. BMP signaling once again appears to play an important role in neural crest development, this time with respect to delamination. In the chick, BMP4 is expressed along the dorsal neural tube (Sela-Donenfeld and Kalcheim, 1999). Noggin is also expressed in a
9 gradient along the dorsal neural tube, with higher concentrations caudally and
lower concentrations rostrally, where neural crest emigration is taking place.
Over time, Noggin expression recedes caudally, perhaps presaging the
commencement of the EMT. In vivo and in vitro manipulation of these opposing gradients further supports a role for BMP signaling in delamination (Sela-
Donenfeld and Kalcheim, 1999). The transcriptional repressor slug, a gene expressed in the neural folds, also appears to participate in the EMT. Loss-of- function analyses in both chick and Xenopus result in failure of neural crest migration in vivo (Carl et al., 1999; Nieto et al., 1994). Important targets of Slug in this context may include cell adhesion molecules. The EMT requires down- regulation of cell adhesion molecules expressed in the neural plate, such as N- cadherin, N-CAM and cadherin 6B, as well as dissociation of tight junctions that help hold cells within the neural epithelium (Akitaya and Bronner-Fraser, 1992;
Nakagawa and Takeichi, 1995; Nakagawa and Takeichi, 1998; Savagner et al.,
1997). At the same time, additional molecules like cadherins 7 and 11 must be up-regulated to allow neural crest cells to navigate the extracellular matrix (ECM) and remain associated within specific crest subpopulations (Nakagawa and
Takeichi, 1998; Takeichi et al., 2000). Further, NCCs cannot enter a fully established basal lamina (Erickson, 1987). They have the capacity to produce proteolytic enzymes (Valinsky and Le Douarin, 1985), although it remains unclear whether NCCs actively degrade the ECM in vivo.
After the EMT, neural crest cells (NCCs) remain briefly in the “staging
area,” an extracellular matrix-rich space just dorsal and lateral to the neural tube,
10 before migrating along stereotyped pathways to their eventual locations in the
embryo (see Figure 1.2). NCCs may produce matrix proteins, such as collagen,
to facilitate migration through the extracellular milieu (Kalcheim and Leviel,
1988). Neural crest cells express a variety of Integrins, transmembrane receptors
that bind ECM molecules. Additionally, a number of signals that have been
shown to function in migration of other cell types or in neuronal axon guidance have been implicated in migration of neural crest cells. These include Ephrins,
Semaphorins, Endothelins, and their receptors (McCallion and Chakravarti, 2001;
Osborne et al., 2005; Pla and Larue, 2003; Santiago and Erickson, 2002; Yu and
Moens, 2005). Interestingly, two recent studies have linked canonical Wnt signaling to neural crest delamination (Burstyn-Cohen et al., 2004) and non- canonical Wnt signaling to attractive cues for migrating crest cells (De Calisto et al., 2005). Cytoskeletal components needed for cell motility, such as α-adducin
and palladin, are also expressed in migrating neural crest cells (Gammill and
Bronner-Fraser, 2002).
In the trunk, two major migratory pathways have been defined for neural
crest cells (for example, see Figure 1.4). In most species examined to date, the
first NCCs to migrate follow the medial pathway between the neural tube and
developing somites. These cells will give rise to sympathetic neurons, dorsal
root sensory ganglia and pigment cells. Slightly later, other NCCs emigrate from
the staging area along the lateral pathway between the somite and the overlying
epidermis. Cells following this pathway give rise exclusively to pigment cells.
However, there is evidence from mice that neural crest cells simultaneously enter
11 both pathways (Serbedzija et al., 1990) and in the axolotl (Ambystoma mexicanum, a salamander) the timing of migration along the medial and lateral pathways is reversed (Lofberg et al., 1985). Further, whereas medial migration occurs through the anterior portion of each somite in chick, medially-migrating zebrafish neural crest cells follow the middle of the developing myotome. Hence, many commonalities in neural crest development are apparent between different vertebrates, but some species-specific differences are also evident. Cephalic crest cells migrate primarily in six streams dorsolaterally through the pharyngeal arches to form various aspects of the craniofacial skeleton and other derivatives
of the neural crest. Neural crest cells from the first (mandibular) and second
(hyoid) arches form the anterior portion of the lower jaw, while the third
pharyngeal arch splits into four branchial arches that form the posterior lower jaw
structures. Recent studies in Xenopus indicate that neural crest cells migrating
through the pharyngeal arches also contribute to more dorsal structures in the
skull vault, such as the frontoparietal bone (Gross and Hanken, 2005).
Additionally, vagal neural crest cells that will give rise to the enteric nervous
system migrate caudally along the ventro-medial aspect of the embryo to
colonize the gut. Thus, while neural crest-derived cells are located extensively
throughout the vertebrate body, NCCs often follow distinct, stereotyped pathways
of migration to their final destinations.
12 Specification of sublineages within the neural crest
One of the features that first attracted researchers to the neural crest is
the apparent homogeneity, at least morphologically, of the initial crest population.
Consequently, an issue that has received much attention is the mechanism(s) by
which NCCs choose between the many fates observed in vivo. At one extreme, neural crest cells have been hypothesized to be pluripotent, with each cell having the ability to give rise to any derivative of the neural crest. In this scenario, crest cells all have the same potential, and are directed to specific fates by environmental cues prior to, during, or after migration. At the other extreme, it might be postulated that all neural crest cells are specified to a given fate very early in their ontogeny, even within the neural tube prior to delamination. Thus, the migratory pathway chosen by and eventual location of a given NCC would be dependent on intrinsic cues. Further, important distinctions have been drawn between the specification of an individual neural crest cell, that is, expression of attributes or behaviors that distinguish it from other NCCs, and its commitment, i.e.: its capacity to change fates when challenged in a new environment.
Necessarily, these questions have been approached in many different ways, including transplantation, analysis of explants in vitro, in vivo lineage analysis, immunohistochemistry, gene expression and mutational analyses. While individual experiments have been interpreted as supporting one view of crest specification or another, it seems unlikely that any single mechanism is responsible for the diversity of neural crest cell fates. When the wealth of data is examined globally, several themes emerge: 1) both fate-restricted and pluripotent
13 precursors are present in the premigratory neural crest population, 2) the early
neural crest population is heterogeneous, 3) some neural crest cells are highly
plastic, 4) generation of the full complement of neural crest derivatives requires
both cell-intrinsic mechanisms and environmental factors.
Both fate-restricted and pluripotent precursors are present in the
premigratory neural crest population. In virtually all studies in many different organisms, clonal analysis of single neural crest cells reveal that some NCCs give rise to only one derivative, while others give rise to multiple cell types. In clonal culture methods, neural crest cells that arise from neural tube explants in vitro are collected, diluted to single cell density, and individually plated. Various culture conditions have been shown to permit differentiation into many, though not all, neural crest derivatives (Akira and Ide, 1987; Baroffio et al., 1988;
Stemple and Anderson, 1992). Under conditions that support pigment cell development from Xenopus neural tube explants, Akira and Ide (1987) found that single neural crest cells gave rise to clones containing either black melanocytes or yellow xanthophores alone, or a mixture of these two cell types. Similarly, single NCCs derived from quail neural tube explants have been shown to generate both neuron-only and mixed neuron-melanocyte clones, as well as clones composed of cartilage cells and other non-neuronal cells expressing the crest-specific HNK-1 epitope (Baroffio et al., 1988). Making use of glial-specific antibodies, this system was later used to show that some neural crest cells generate only glial cells, while others give rise to glia and neurons or glia, neurons and melanocytes (Dupin et al., 1990). Similar findings have been
14 reported for clonal culture of rat and mouse neural crest cells (Ito et al., 1993;
Stemple and Anderson, 1992). One caveat of in vitro clonal methods is the
dissociation of potential cell-cell interactions that would normally occur in the
embryo and influence fate choices. An in vitro study using chick neural tube explants investigated the fate restrictions of neural crest cells in the context of interactions with other crest cells (Henion and Weston, 1997). Single cells were labeled as they delaminated from explanted neural tubes and clonal progeny were subsequently observed in the outgrowth. In this context, the majority of clones contained a single derivative, indicating a predominance of fate-restricted precursors. However, there were also a large number of mixed clones, combinations of neurons, glia and melanocytes, highlighting the assortment of fate-restricted and pluripotent precursors within the neural crest population
(Henion and Weston, 1997). Finally, in vivo lineage analysis has been undertaken in a variety of organisms, including chick, frog, mouse and zebrafish
(Bronner-Fraser and Fraser, 1988; Collazo et al., 1993; Dutton et al., 2001b;
Raible and Eisen, 1994; Serbedzija et al., 1994). These studies involve labeling single cells prior to or shortly after delamination from the neural tube, and observation of progeny within the embryo at later stages. In each case, the authors noted examples of single NCCs giving rise to progeny of a single cell type or mixed combinations of NC derivatives. A notable exception to these general findings is found in the lineages of cranial neural crest (CNC) cells in the zebrafish (Schilling and Kimmel, 1994). In this case, the authors found that single
CNC cells only ever gave rise to one identifiable neural crest derivative. This is in
15 contrast to lineage analysis performed at trunk levels in the zebrafish, which
identified an early-migrating population of crest cells that gives rise to clones with
multiple cell types (Raible and Eisen, 1994). These data highlight the likelihood
that there is some variation in fate specification between crest cells at different axial levels or between species. Alternatively, because only the earliest migrating
trunk crest cells in zebrafish displayed multipotency, it remains possible that
differences in timing of events used to characterize premigratory neural crest
cells at different axial levels and between different species renders the
comparison of these analyses less than straightforward. Importantly, the general
finding that these reports collectively show is that under a wide variety of
conditions, both in vitro and in vivo, the premigratory neural crest cell population is composed of both restricted progenitors which will give rise to a single derivative, as well as precursors that have not yet been specified to a single fate and are able to generate multiple cell types.
Not surprisingly, based on the results just described, considerable
heterogeneity exists within the early neural crest population. An obvious example of this is the different complement of derivatives generated at certain axial levels.
At this time, for example, it is believed that only crest cells from hindbrain-anterior trunk levels can contribute to cardiac outflow tract structures (Kirby, 1989; Kirby and Stewart, 1983). Enteric neurons arise from vagal and sacral crest, but not at trunk levels in between (Le Douarin and Teillet, 1973). Crest cells that give rise to different derivatives appear to segregate at a given axial level, as well. Rather than being randomly interspersed, NCCs that generate specific derivatives
16 appear to arise in distinct medio-lateral locations relative to the neural tube
(Dorsky et al., 1998; Schilling and Kimmel, 1994). Further, crest cells that cannot
be morphologically distinguished from one another exhibit different intrinsic
characteristics. Indeed, a variety of monoclonal antibodies, such as R24, E/C8,
A2B5 and MEBL-1, label only subsets of NCCs (Ciment and Weston, 1982;
Girdlestone and Weston, 1985; Kitamura et al., 1992). Further, the populations recognized by these antibodies do not necessarily overlap. Thus, some undifferentiated NCCs can be distinguished from one another by the protein moieties expressed on their cell surfaces. Gene expression has also been used to define subsets of neural crest cells before overt differentiation. A wide variety of lineage-specific genes have now been described (Kelsh and Raible, 2002).
Interestingly, many of these genes are expressed at the level of the dorsal neural tube, supporting the idea that heterogeneity exists at very early stages of neural crest development. Further, there is increasing evidence that genes thought to be expressed uniformly in the premigratory neural crest, such as tfap2, sox10 and foxd3, are required for the development of only a subset of crest-derived cells
(Dutton et al., 2001b; Knight et al., 2003; O'Brien et al., 2004). This suggests at
least two possibilities: 1) subpopulations of neural crest cells can be
distinguished at very early stages by expression of certain gene products, or 2)
these genes are, in fact, expressed in a pan-neural crest fashion, but not all
NCCs that express them require these genes for development. In either case,
these data raise the intriguing possibility that, rather than a single uniform
population, the neural crest is a collection of different subpopulations.
17 Some neural crest cells are highly plastic. Considerable attention has also been given to investigating the level of commitment of NCCs, or their ability to maintain a fate choice cell autonomously. This requires both knowledge of a cell’s normal differentiation behavior, and challenging the cell in a new environment. Consequently, it has become clear that at least some neural crest cells are not committed to a given cell fate, even after overt differentiation, and are able to transfate to other cell types. For example, when quail crest-derived sympathetic ganglia containing overtly differentiated neurons are transplanted to the dorsal neural tube of a chick host, these neurons are able to dedifferentiate, undergo migration, and give rise to glia and sensory, as well as sympathetic, neurons (Le Lievre et al., 1980). Additional experiments using the quail-chick system reveal that vagal NCCs, which normally give rise to enteric neurons, can generate adrenal medullary (endocrine) cells and sympathetic neurons when transplanted to sacral trunk levels (Le Douarin and Teillet, 1974).
Transdifferentiation of melanocytes to glia and vice versa has been demonstrated in quail cell culture (Dupin et al., 2000; Dupin and Le Douarin,
2003) as well as an in vivo mouse injury model (Rizvi et al., 2002). Further evidence comes from studies in amphibians. Anamniotic vertebrates typically produce multiple types of pigment cells, including melanophores, xanthophores and iridiphores (Bagnara, 1998). Akira and Ide (1987) showed that while all three of these cell types could be produced in clonal culture from Xenopus neural tube explants, iridiphores rapidly transdifferentiated to melanocytes. These cells, while at first containing only reflective organelles characteristic of iridiphores, passed
18 through an intermediate phase with both platelets and the melanosomes found in melanocytes. Relationships between chromatophore cell types have also been investigated using the axolotl, Ambystoma mexicanum. Treatment of live animals with guanosine changes the relative proportions of melanophores and xanthophores that make up the pigment pattern (Frost et al., 1987), while in vitro analysis suggests that these shifts in pigment cell type may be at the expense of one another (Thibaudeau and Holder, 1998). In addition to changes between overtly differentiated cells, plasticity may refer to the ability of precursor cells to alter established behaviors in vivo. This has been demonstrated with respect to cranial neural crest cells in the axolotl. Cephalic NCCs at given rostro-caudal levels have been shown to migrate to specific locations and give rise to certain jaw structures. However, when these cells are transplanted short distances along the neural tube, they migrate along the path appropriate for, and contribute to the structures normally generated by, cells from this new location (Epperlein et al.,
2000). Similarly, in both cranial quail-chick chimeras (Baker et al., 1997) and zebrafish trunk NC ablation experiments (Raible and Eisen, 1996), early- emigrating neural crest populations can be replaced by later-emigrating NCCs which would not normally give rise to the same complement of derivatives. In addition, it seems that it is even possible, under some circumstances, for NCCs that have begun migration to be reprogrammed to contribute to non-neural crest- derived tissues (Ruffins et al., 1998). NCCs at early migratory stages were transplanted to the ventral neural tube, which normally gives rise to floor plate and motor neurons (Gilbert, c2000). The transplanted NCCs adopted morphology
19 appropriate to floor plate and motor neurons, as well as expression of at least one gene for each of these cell types (FP-1 and isl-1, respectively). Although this level of plasticity is unlikely to occur widely and its relevance in normal development has not be established, this study and others highlight the ability of a subset of neural crest cells to alter fate choices in response to environmental factors.
Finally, the generation of the full complement of neural crest derivatives found in the embryo is the result of both cell-intrinsic mechanisms and environmental factors. Although the picture is far from complete, relative contributions of these processes have been elucidated in some very specific cases. In most vertebrates, trunk neural crest cells do not give rise to cartilage or bone, even when transplanted to the dorsal neural tube in cranial regions.
However, several studies have found that if the requirement for migration is bypassed, trunk NCCs in the pharyngeal arches do contribute to craniofacial cartilages (Epperlein et al., 2000; McGonnell and Graham, 2002). This suggests that trunk NCCs retain skeletogenic potential, but that the environment in which these cells develop does not possess the necessary signals to direct crest cells to this lineage. Also within the realm of environmental cues are cell-cell interactions between neural crest cells themselves. As mentioned above, later- emigrating NCCs appear to functionally replace ablated early migrating NCCs in vivo, contributing to derivatives they would not normally generate (Baker et al.,
1997; Raible and Eisen, 1996). These results suggest that early migrating cells, which follow the same path as later migrating cells, somehow interact with the
20 later migrating cells to preclude the earlier fates (Baker et al., 1997; Raible and
Eisen, 1996). Further, both in vivo and in vitro data indicate that cell-cell communication is important for the generation of both neuronal and glia cell types within peripheral ganglia (Morrison et al., 2000; Wakamatsu et al., 2000).
Within the dorsal root ganglia (DRG) of chicks, for example, Delta1 is expressed by early neuronal cells, while Notch1 signaling promotes gliogenesis at the expense of neurogenesis in this population (Wakamatsu et al., 2000). Even transient Notch signaling has been shown to irreversibly block neurogenesis in rat clonal culture (Morrison et al., 2000).
On the other hand, cell intrinsic restrictions are also observed. Expression
of mitf has been shown to be both necessary and sufficient for melanocyte
development, and is expressed in a subset of premigratory neural crest cells
(Dorsky et al., 2000; Hornyak et al., 2001; Lister et al., 1999; Tachibana et al.,
1996; Widlund and Fisher, 2003). Further, it appears that NCCs must be
specified as melanoblasts prior to entering the dorsolateral migration pathway
and that other cell types are excluded from this route (Erickson and Goins, 1995).
Thus, inherent differences may determine which NCCs will be exposed to a given environment, in this case, the dorsolateral pathway. Differentiation of sympathetic versus sensory neurons also appears to be influenced by cell autonomous factors (Anderson, 2000). These neuronal subtypes migrate along overlapping pathways through the embryo. As sympathetic precursors must navigate through the region of the embryo where sensory neurons will later develop, it appears that at least some sympathetic neuroblasts are resistant to sensory-inducing
21 signals. Further, sensory neurons have been shown to express neurogenins
(ngns), and loss-of-function studies indicate that ngn2 is required for
development of sensory neurons (Fode et al., 1998; Ma et al., 1998; Ma et al.,
1999; Perez et al., 1999). Sympathetic neuroblasts, on the other hand, do not
appear to express or require ngns, suggesting a possible cell autonomous
mechanism for divergence of these distinct neurogenic populations despite overlapping environmental factors (Anderson, 2000; Ma et al., 1999).
Conclusion
Significant strides have been made in the study of the neural crest over
the last 100 years. A great deal is now known about the embryonic origins,
morphological development and derivatives of the neural crest. Yet despite the
extensive breadth and depth of knowledge regarding this fascinating cell
population, many questions are still unanswered. In particular, much remains to
be learned about the molecules that control various aspects of neural crest
development. The precise roles of genes expressed in premigratory and
migrating neural crest cells, particularly the various transcription factors
expressed widely in the early crest population, must be elucidated. These genes
must, in turn, be linked in pathways to molecules involved in induction and
specification of the early neural crest population, and to their targets within neural
crest cells. The genetic requirements of specific neural crest derivatives and
subpopulations of derivatives also call for further investigation. This knowledge
will lead to a better understanding of the neural crest, including whether or not it
22 is a single homogenous cell population at any point in its ontogeny. Such information may also be enlightening with respect to human medical conditions affecting the neural crest or its derivatives, as well as providing clues to the evolutionary origin of the neural crest. Both forward and reverse genetic analyses continue to be utilized to answer these questions. New methods, tools and organisms are yielding further insights into the development and evolution of the neural crest. Assuredly, these discoveries will be applicable to other embryonic cell populations, as well.
23 TABLES AND FIGURES
Figure 1.1. Subsets of neural crest derivatives arise at distinct axial levels. Schematic diagram of a chick embryo, dorsal view with anterior to the top. FB, forebrain; MB, midbrain; HB, hindbrain; S, somite. (Adapted from Le Douarin et al., 2004)
24
Figure 1.2. Schematic of vertebrate neurulation. (A) Whole embryo, dorsal view with anterior to the top. The neural plate border/neural crest forms at the boundary between the neural plate and the future epidermis. (B) Cross-sections through the trunk, dorsal to the top. As neurulation occurs, the neural plate folds upon itself, bringing the neural folds together at the midline. Neural crest cells delaminate from the dorsal neural tube and migrate along stereotyped pathways. E, epidermis; LP, lateral pathway; M, mesoderm; MP, medial pathway; N, notochord; NC, neural crest; NP, neural plate; NT, neural tube; S, somite. (Adapted from Wu et al., 2003)
25
Figure 1.3. Zebrafish neural crest development. The neural crest can be identified in vivo by expression of neural crest-specific molecules, such as crestin, shown here. (A) Expression of crestin begins in the neural folds, observed as bilateral domains parallel to the anterior-posterior axis (arrow). One somite stage; dorsal view, anterior to the top. NP, neural plate. (B) These domains extend posteriorly over time and converge on the midline during neurulation. Dorsal view, anterior to the top. (C) The neural folds come together at the dorsal midline. Dorsal view, anterior to the top. (D) Neural crest cells begin segmental migration in a rostro-caudal wave (arrowheads). Lateral view, anterior to the left.
26
Figure 1.4. Migration of zebrafish trunk neural crest cells. Cross-section at mid- trunk level, dorsal to the top. Neural crest cells expressing the pan-neural crest marker crestin migrate medially between the neural tube and somites (arrowheads), as well as laterally between the somites and overlying epidermis (arrows). N, notochord; NT, neural tube; S, somites.
27
CHAPTER 2
SPECIFIC PAN-NEURAL CREST EXPRESSION OF ZEBRAFISH CRESTIN THROUGHOUT EMBRYONIC DEVELOPMENT1
ABSTRACT
Zebrafish crestin was identified in a screen for genes dependent on
cyclops function and is a member of a family of retroelements (Rubinstein et al.,
2000). We report here a detailed description of crestin mRNA expression during
zebrafish embryogenesis. Crestin expression was first observed during the onset
of somitogenesis in cells of the neural crest domain of the ectoderm. Crestin
expression was subsequently observed in premigratory cranial and trunk neural
crest cells and then in actively migrating crest cells. Cell counts of crestin
expressing premigratory trunk neural crest cells strongly suggest that crestin is
expressed by all neural crest cells at this stage. Crestin expression co-localized with a battery of markers for premigratory neural crest cells, developmentally distinct neural crest-derived precursor sublineages, and overtly differentiated neural crest-derived cell types. Expression of crestin is gradually downregulated
1 Reprinted from Developmental Dynamics, Vol. 220, Issue 2, Luo, R., An, M., Arduini, B.L. and Henion, P.D., Specific pan-neural crest expression of zebrafish crestin during embryonic development, pp169-174, Copyright 2001 Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc, with permission from John Wiley & Sons, Inc. Contributions by B.L. Arduini include all data pertaining to the nisakara mutant. 28 in overtly differentiated cells. Our results indicate that crestin is a specific pan-
neural crest marker throughout zebrafish embryogenesis.
INTRODUCTION
The neural crest of vertebrate embryos has been studied extensively to
learn about mechanisms that regulate embryonic development (Le Douarin and
Kalcheim, 1999). The neural crest arises as bilateral domains of the neuroectoderm that subsequently converge on the midline, undergo an epithelial-
mesenchymal transition and form a discrete population of mesenchymal cells
located along the dorsolateral aspect of the neural tube. Neural crest-derived
cells then undergo a temporally and spatially stereotyped migration to a variety of
embryonic locations. Ultimately, neural crest gives rise to a variety of
differentiated cell types including peripheral neurons, glia, pigment cells, and
elements of the craniofacial skeleton. In zebrafish embryos, neural crest
development appears to occur in a very similar way compared with other
vertebrate embryos (Eisen and Weston, 1993; Raible et al., 1992). However,
because zebrafish embryos are initially optically clear and their neural crest cells
are larger and fewer in number than in other species, individual premigratory
neural crest cells can be identified and manipulated in living embryos (Raible and
Eisen, 1994; Raible and Eisen, 1996; Schilling and Kimmel, 1994). In addition,
the ability to efficiently generate and identify mutations in zebrafish, and
specifically mutants with phenotypes involving neural crest and it’s derivatives
(Henion et al., 1996; Neuhauss et al., 1996; Odenthal et al., 1996; Piotrowski et
29 al., 1996; Schilling et al., 1996) provides a powerful system for determining the genetic mechanisms that control neural crest development.
One tool that has been lacking for the analysis of zebrafish neural crest development has been a robust marker for all neural crest cells that is expressed throughout embryonic development. One gene, forkhead 6 (Odenthal and
Nusslein-Volhard, 1998), appears to be expressed by all neural crest cells in the ectoderm and prior to migration. However, fkd6 expression is rapidly downregulated in neural crest cells once they begin to migrate, and fkd6 expression is initiated in other tissues (Odenthal and Nusslein-Volhard, 1998).
Because the vast majority of migratory neural crest cells are not identifiable in living embryos unless previously labeled with lineage dye, a critical period of neural crest development, from the initiation of migration until the expression of molecular markers of overt differentiation, has been difficult to analyze in detail.
Although many of the genes that have been identified that are expressed by subsets of neural crest cells (see Table 2.1 and references therein) provide information about this critical period, none of these markers identifies the entire neural crest population.
Zebrafish crestin was identified in a screen for genes dependent on cyclops function and was shown to be expressed by at least some neural crest and crest derived cells (Rubinstein et al., 2000). Here we report the detailed expression of crestin throughout embryonic development. We have also determined that crestin is expressed by all neural crest cells and by all identifiable neural crest sublineages and differentiated derivatives. Therefore,
30 crestin is a specific pan-neural crest marker from the specification of the neural
crest domain of the ectoderm through the overt differentiation of neural crest
derivatives. In a way similar to the use of the HNK-1 epitope in avian embryos,
crestin should prove to be a very useful tool for studying neural crest
development in wild-type and mutant zebrafish.
RESULTS AND DISCUSSION
We have previously reported the cloning of zebrafish crestin and a brief analysis of expression (Rubinstein et al., 2000). Here we report in detail the expression of crestin during embryogenesis from the beginning of somitogenesis through 72 hours postfertilization (hpf; Fig. 2.1A–N). Also, we have used a battery of markers for neural crest cells and specific subsets of neural crest- derived cells to define crestin expression during neural crest development (Table
2.1; Fig. 2.2).
During embryogenesis, crestin expression was first observed in two narrow rows of cells in the ectoderm at the beginning of somitogenesis (Fig.
2.1A, J). These rows of expressing cells elongated, converged on midline and occupied the dorsal region of the neural keel as somitogenesis proceeded (Fig.
2.1B, C, L, N). Thus, by the 12–14 somite stage, crestin was expressed exclusively by presumptive premigratory neural crest cells.
Beginning at approximately the 14-somite stage, a subpopulation of crestin-expressing cells in the first few trunk segments were observed in ventral rows between the neural keel and somites consistent with their being migratory
31 neural crest cells (Fig. 2.1D, E, K). Subsequently, crestin-expressing cells were also observed along the presumptive lateral neural crest migratory pathway (Fig.
2.1F, G, M). This pattern of expressing migratory neural crest cells progressed along the axis until approximately 30 hours post-fertilization (hpf) when crestin- expressing cells were distributed in all areas of the embryo known to be occupied by neural crest-derived cells (Fig. 2.1H, I). Subsequently, crestin expression was gradually downregulated in some overtly differentiated (melanized) melanophores and was eventually extinguished in most cells after 72 hpf.
Since the early pattern of expression suggested that most or all neural
crest cells express crestin, we quantified the number of premigratory crestin+ cells in the trunk. To determine the number of crestin-expressing premigratory neural crest cells in each trunk segment, 14–16-somite stage embryos were double-labeled with crestin and islet2, which is expressed by CaP and VaP primary motor neurons in each segment (Appel et al., 1995), among other neuronal subtypes. We used the distance between islet2-positive CaP/CaP 1
VaP neurons as a measure of unit segment distance and counted the number of crestin-expressing cells on the dorsal aspect of the neural keel in each of these segments (Fig. 2.1O, P). Counts from 14 segments of 4 different embryos revealed that 21+/-4.9 crestin-positive premigratory neural crest cells were present in each segment. This number is quite similar to other estimates (Raible et al., 1992) and suggests that crestin is expressed by all premigratory trunk neural crest cells.
32 At the beginning of somitogenesis, crestin was expressed by two rows of
cells in the neuroectoderm and subsequently by presumptive premigratory neural
crest cells. Double in situ hybridization experiments revealed that the same cells
at both developmental stages expressed fkd6 (Fig. 2.2A, B), which has been
shown to mark the neural crest domain of the ectoderm and premigratory neural
crest cells (Odenthal and Nusslein-Volhard, 1998), confirming the neural crest
identity of crestin-expressing cells. In addition, many crestin-expressing cells also
expressed AP-2 (Fig. 2.2C, D), msxb (Fig. 2.2E, F), and dlx2 (Fig. 2.2G, H) which
are also expressed by some premigratory and early migrating neural crest cells
(Akimenko et al., 1994; Akimenko et al., 1995; Nguyen et al., 1998).
All identifiable neural crest sublineages expressed crestin mRNA. crestin was expressed by all melanogenic neural crest cells identified with MITF, c-kit, dct, and fms (Fig. 2.2I–P; (Kelsh et al., 2000b; Lister et al., 1999; Parichy et al.,
2000b; Parichy et al., 1999). Co-expression of fms by a subset of crestin- expressing cells also indicates that xanthophore precursors are marked by crestin expression (Parichy et al., 2000b). In addition, precursors for iridiphores
(endrb1; not shown; see Fig. 2.3), branchial arch derivatives (Akimenko et al.,
1994), fin mesenchyme (Akimenko et al., 1995) and enteric neurons (Bisgrove et al., 1997) also expressed crestin. In every case, and as expected for a pan- neural crest marker, crestin was expressed by all cells expressing a particular sublineage marker as well as by additional neural crest cells.
All overtly differentiated neural crest-derived cells also appeared to
express crestin. Melanized melanophores transiently expressed crestin (Fig.
33 2.2R) as did iridiphores and presumptive xanthophores (not shown). All neuronal
derivatives of the neural crest in the head and trunk expressed crestin. For
example, dorsal root ganglion sensory neurons (Henion et al., 1996), autonomic
sympathetic neurons (TH; Fig. 2.2V), enteric neurons (see Fig. 2.2Q) and cranial
ganglion neurons (not shown) were found to express crestin. Also, non-neuronal
(anti-Hu negative), presumptive glial cells associated with ganglionic neurons also expressed crestin (see Fig 2.2S, T). Lastly, branchial arch structures were also identifiable based on crestin expression (Fig. 2.2U).
After 72 hpf, crestin expression is downregulated in most neural crest derivatives, although ventral cells associated with and in the region of the dorsal aorta which are presumably late differentiating sympathetic neurons (An et al.,
2002), glial cells and pigment cells continue to express crestin. In addition, some branchial arch associated cells and scattered superficial, presumptive late differentiating pigment cells also continue to express crestin until approximately 5 dpf. This late expression pattern is similar to earlier stages (ca. 30–72 hpf) in that crestin expression is maintained in neural crest-derived cells until several hours
after overt differentiation (e.g., melanization of melanocytes) and is then
downregulated. This results in a progressive decrease in the number of crestin-
positive cells as development proceeds.
Taken together, our results suggest that crestin is expressed by neural crest cells as soon as they are specified, is expressed by all neural crest cells, and is expressed continuously until after neural crest-derived cells undergo overt
34 differentiation. Thus, crestin is a specific pan-neural crest marker throughout zebrafish embryogenesis.
Lastly, crestin should prove to be a useful tool to routinely assess neural
crest development in any mutant. For example, mutations that result in the
specification of little or no neural crest are easily identifiable based on crestin
expression (Rubinstein et al., 2000). Importantly, mutants that display much more
subtle neural crest phenotypes are also readily identifiable. For example, in a
screen for mutations affecting retina development (T. Vihtelic, personal
communication), a mutant was identified that lacked most iridiphores which we
have named nisakara (Sanskrit for one who makes darkness). The iridiphore
phenotype is first detectable at 2–3 dpf (see Fig. 2.3A) and is specific for
iridiphores since no other visible defects in other neural crest derivatives, the eye
or elsewhere are observed. Nevertheless, analysis of crestin expression at much
earlier stages, beginning at early migratory stages and continuing as migration
proceeds, revealed a consistent phenotype that was easily detected (Fig. 2.3B),
even though iridiphore precursors are a minor constituent of the neural crest
population. Therefore, even subtle phenotypes which are otherwise undetectable
can be revealed by analysis of crestin expression.
MATERIALS AND METHODS
All zebrafish (Oregon *AB) were maintained in the Ohio State University
Zebrafish Facility. Embryos were raised at 28.5°C. Immunocytochemistry, riboprobe synthesis, and single- and double-in situ hybridization procedures were
35 performed as previously described (Henion et al., 1996; Rubinstein et al., 2000).
Detailed protocols are available upon request.
For all in situ hybridizations of embryos performed after 26hpf, embryos
were raised in fish water containing 200 mM 1-Phenyl-2-thiourea (PTU) to
prevent melanin synthesis in melanocytes. This allowed us to assess crestin
expression in late stages without interference from melanin-containing
melanocytes. PTU did not affect the number or distribution of melanocytes,
detected with molecular markers in treated fish (e.g., c-kit), for any duration of treatment we used.
36 TABLES AND FIGURES
Marker Neural crest-related expression Reference fkd6 Crest domain of ectoderm; premigratory crest Odenthal et al., 1998 AP2 Premigratory/early migrating neural crest See Nguyen et al., 1998 dlx2 Branchial arch crest Akimenko et al., 1994 msxb Some premigratory crest; fin mesenchyme Akimenko et al., 1995 MITF Early melanogenic sublineage Lister et al., 1999 c-kit Early melanogenic sublineage Parichy et al., 1999 dct Late melanogenic sublineage Kelsh and Eisen, 2000 fms Melanophore and xanthophore precursors Parichy et al., 2000 c-ret Enteric neurons/precursors Bisgrove et al., 1997 anti-Hu All neurons see Henion et al., 1996 melanin Melanophores TH Sympathetic neurons; chromaffin cells see Raible et al., 1992
Table 2.1. Molecular markers for neural crest and neural crest-derived cells.
37
Figure 2.1. Expression of crestin from the 3-somite stage through 72 hpf in whole-mount preparations and transverse sections. For whole-mounts, A, B, D, F are dorsal views with anterior up. C, E, G, H and I are lateral views with anterior to the left and dorsal up. For the sections, all sections are from the trunk region, dorsal is up and use K for orientation to prominent structures. 3-somite stage (A, J); 14-somite stage (B, C, L, N-higher magnification of B with anterior up); 18- somite stage (D, E, K), 24 hpf (F, G, M); 48 hpf (H); 72 hpf (I). Dorsal view, anterior to the left, of 14-somite stage embryo trunk in focal plane of primary motoneurons (O; islet2, blue) and premigratory neural crest (P; crestin, red). nt, neural tube; nc, notochord; s, somite. 38
Figure 2.2. Co-expression of neural crest, neural crest sublineage, and differentiated derivative molecular markers in subsets of crestin-expressing cells in whole-mounts and transverse sections. In all cases except S and T, crestin expression is red and other marker expression is blue. For all whole-mounts except A (dorsal view, anterior up), anterior is to the left and dorsal is up. All sections except H (rostral hindbrain) are trunk region and use D for orientation to prominent structures. In all sections and R, arrowheads indicate double-positive cells. fkd6, 8-somite stage (A, B); AP-2 (C, D), msxb (E, F), dlx2 (G, H), 22- somite stage; MITF, 18-somite stage (I, J); c-kit, 22-somite stage (K, L); dct, 24 hpf (M, N); fms, 22-somite stage (O, P). Enteric neuron precursors, 36 hpf (Q, c- ret); melanocytes, 28 hpf (R, melanin); dorsal root ganglion sensory neurons (T, anti-Hu, and corresponding cells in S, crestin) and satellite glia (S, crestin+ and T anti-Hu-), 48 hpf. Branchial arches (U; arrowhead, crestin) 66 hpf and sympathetic neurons (V; arrow, TH) 48 hpf. nt, neural tube; nc, notochord; s, somites.
39
Figure 2.2
40
Figure 2.3. The mutant nisakara lacks most iridiphores and displays an early crestin phenotype. (A) Live 8 dpf wild-type (bottom) and nisakara (top) embryos. nisakara embryos lack most iridiphores, most prominently in the eye and dorsal stripe. A crestin phenotype is apparent in one-quarter of embryos from matings of identified heterozygous nisakara carriers at 22 hpf (B) Note fewer crestin+ cells in the eye and dorsal stripe of the mutant embryo (top; arrowheads) compared to wild-type sibling (bottom).
41
CHAPTER 3
MELANOPHORE SUBLINEAGE-SPECIFIC REQUIREMENT FOR ZEBRAFISH TOUCHTONE DURING NEURAL CREST DEVELOPMENT2
ABSTRACT
The specification, differentiation and maintenance of diverse cell types are
of central importance to the development of multicellular organisms. The neural
crest of vertebrate animals gives rise to many derivatives, including pigment
cells, peripheral neurons, glia and elements of the craniofacial skeleton. The
development of neural crest-derived pigment cells has been studied extensively
to elucidate mechanisms involved in cell fate specification, differentiation,
migration and survival. This analysis has been advanced considerably by the
availability of large numbers of mouse and, more recently, zebrafish mutants with
defects in pigment cell development. We have identified the zebrafish mutant
touchtone (tct), which is characterized by the selective absence of most neural
crest-derived melanophores. We find that although wild-type numbers of
melanophore precursors are generated in the first day of development and
migrate normally in tct mutants, most differentiated melanophores subsequently
fail to appear. We demonstrate that the failure in melanophore differentiation in
2 Reprinted from Mechanisms of Development, Vol. 121, Arduini, B.L. and Henion, P.D., Melanophore sublineage-specific requirement for zebrafish touchtone during neural crest development, pp1353-64, Copyright 2004, with permission from Elsevier. All data reported herein, unless otherwise cited, were generated by B.L. Arduini under the guidance of P.D. Henion. 42 tct mutant embryos is due at least in part to the death of melanoblasts and that tct function is required cell autonomously by melanoblasts. The tct locus is
located on chromosome 18 in a genomic region apparently devoid of genes
known to be involved in melanophore development. Thus, zebrafish tct may
represent a novel as well as selective regulator of melanoblast development
within the neural crest lineage. Further, our results suggest that, like other neural
crest-derived sublineages, melanogenic precursors constitute a heterogeneous
population with respect to genetic requirements for development.
INTRODUCTION
The neural crest is a transient vertebrate embryonic cell population that
gives rise to a wide variety of cell types, including pigment cells, craniofacial
cartilage, and neurons and glia of the peripheral nervous system (Le Douarin and
Kalcheim, 1999). This array of neural crest-derived cell types has long been of
interest in studying the mechanisms of cell diversification among embryonic cell
populations. The development of neural crest-derived pigment cells in particular
has been studied extensively, and many important insights have resulted from
the analysis of mouse and zebrafish mutants (Bennett and Lamoreux, 2003; Le
Douarin and Kalcheim, 1999; Lister, 2002; Quigley and Parichy, 2002; Silvers,
1979).
Vertebrate chromatophore populations are readily observed, as they carry
their own intrinsic markers. In addition, pigment cells are not strictly required for
viability (Lister et al., 1999; Parichy et al., 1999; Silvers, 1979). As a result, these 43 pigmented cell types have long been used for studying developmental processes
such as cell fate specification, proliferation, migration, differentiation, and survival. Mice and other mammals have a single chromatophore cell type termed melanocytes (Nordlund et al., 1998). Hundreds of mouse coat color mutants have been identified, covering over 100 loci, which affect multiple cellular processes (Bennett and Lamoreux, 2003; Silvers, 1979). Further, many of these mutations in mice have proved to be medically relevant as models for human diseases involving the same genes (Jackson, 1997). Besides the melanocytes
(melanophores) also found in mammals, zebrafish possess neural crest-derived yellow xanthophores and iridescent iridiphores (Bagnara, 1998; Raible et al.,
1992). In addition to the isolation of several zebrafish pigment mutants that arose spontaneously (Johnson et al., 1995; Streisinger et al., 1986), numerous mutagenesis screens have yielded over 100 mutations affecting various processes in the development of different combinations of the three pigment cell types (Henion et al., 1996; Kelsh et al., 1996; Lister et al., 1999; Odenthal et al.,
1996; Rawls et al., 2003). Because melanophores are common to major vertebrate research animal models, much attention has been focused on this particular pigment cell type.
Studies from several vertebrates, including zebrafish, have led to the
extensive characterization of melanophore development (Bennett and Lamoreux,
2003; Dupin and Le Douarin, 2003; Lister, 2002; Nakamura et al., 2002;
Nordlund et al., 1998). Prior to overt differentiation, melanophore precursors are referred to as melanoblasts, and can be identified by expression of specific
44 genes. Induction of mitf expression, the earliest known marker for melanoblasts, has been shown in multiple systems to require Wnt signaling (Dorsky et al.,
2000; Dunn et al., 2000; Jin et al., 2001; Takeda et al., 2000; Widlund and
Fisher, 2003), as well as Sox10 and other molecules (Bondurand et al., 2000;
Elworthy et al., 2003; Potterf et al., 2000). Sox10, mutations in which cause
Waardenburg-Hirschsprung Syndrome in humans, is required for development of nonectomesenchymal neural crest derivatives, such as most pigment cells and many peripheral neurons and glia (Kelsh and Eisen, 2000; Pingault et al., 1998).
Mitf is both necessary and sufficient for melanophore development (Hornyak et al., 2001; Lister et al., 1999; Tachibana et al., 1996). Mitf upregulates expression
of the receptor tyrosine kinase C-kit (Opdecamp et al., 1997), which in mice
affects primordial germ cell development and hematopoiesis in addition to
melanocytes (Silvers, 1979), but in zebrafish appears to be specifically required
by neural crest-derived melanophores (Parichy et al., 1999). Further, evidence
from melanoma cell lines suggests that C-kit may subsequently activate Mitf in a
positive feedback loop (Hemesath et al., 1998; Price et al., 1998), as well as
ultimately targeting Mitf for ubiquitin-mediated degradation (Wu et al., 2000; Xu et
al., 2000). C-kit and its ligand Steel factor are required for both migration and
survival of melanoblasts and melanophores (Jordan and Jackson, 2000; Rawls et
al., 2003; Steel et al., 1992; Wehrle-Haller et al., 2001). Sox10 has also been
implicated in melanoblast survival (Dutton et al., 2001b; Mollaaghababa and
Pavan, 2003). Dopachrome tautomerase (Dct) and Tyrosinase (Tyr) are
enzymes in the melanin synthesis pathway (Jackson et al., 1990; Tsukamoto et
45 al., 1992), and are relatively late melanoblast markers in zebrafish (Kelsh and
Eisen, 2000). Both are expressed in neural crest-derived melanophores, as well
as non-neural crest-derived melanized cells of the pigmented retinal epithelium
(Camp and Lardelli, 2001; Kelsh et al., 2000b; Steel et al., 1992). Expression of dct therefore can be used to identify neural crest-derived cells that lack melanin as melanoblasts in zebrafish embryos. Similarly, c-kit expression is also diagnostic of melanoblasts among neural crest-derived cells.
In spite of the large volume of data available regarding melanophore
development, many questions remain. For example, although mitfa/nacre
zebrafish mutants have no melanophores, all other zebrafish pigment mutants
have at least some melanophores (Kelsh 1996; Odenthal 1996; Lister 1999).
This raises the possibility that there are subpopulations of melanophores with
distinct genetic requirements. In addition, while several genes have been
hypothesized to play roles in maintenance of melanophore populations, only
sox10 and the c-kit/steel factor signaling pathway have been implicated in
melanoblast survival (Wehrle-Haller 1995; Kelsh 2000; Dutton 2001;
Mollaaghababa 2003). Since neither c-kit/sparse nor sox10/colourless zebrafish
mutants are entirely devoid of melanophores (Parichy 1999; Dutton 2001),
additional genes are likely to be required for the regulation of melanoblast
development. Finally, although over 100 pigmentation loci have been identified in
both mice and fish, forward genetic studies continue to yield novel insights into
the regulation of pigment cell development. For example, fms, a gene that plays
no known role in pigmentation in amniotes, is required for development of 46 embryonic xanthophores and adult melanophores in zebrafish (Parichy et al.,
2000b; Parichy and Turner, 2003). Function of Ednrb and its ligand Endothelin-3,
on the other hand, although required by all melanocytes in mice (Baynash et al.,
1994), are dispensable for some adult and all embryonic melanophores in
zebrafish (Parichy et al., 2000a). Whereas zebrafish mutants with defects in genes previously known to be involved in pigment cell development have been identified, the majority of the zebrafish mutants with pigment phenotypes have not yet been molecularly characterized (Henion et al., 1996; Kelsh et al., 1996;
Odenthal et al., 1996). Thus, many zebrafish pigment mutants are likely to reveal new insights into the mechanisms of melanophore development.
The zebrafish touchtone (tct) mutant was isolated because homozygous
mutant embryos lack the majority of neural crest-derived melanophores (Henion
et al., 1996). In this report, we present our analysis of the development of tct
mutant embryos. Our results indicate that tct is selectively required for
melanophore development among neural crest-derived cells. We provide evidence that in the absence of normal tct function a subpopulation of melanoblasts die and that tct is required cell autonomously by melanoblasts.
Thus, the absence of the majority of melanophores in tct mutant embryos appears to result at least in part from the death of melanoblasts prior to overt differentiation. Further, we report the identification of multiple tct alleles and progress toward the molecular identification of the tct locus. We suggest tct is a novel regulator of the development of a subpopulation of neural crest-derived
47 melanoblasts and is indicative of precursor heterogeneity within the melanogenic sublineage.
RESULTS
Live phenotype and isolation of multiple tct alleles
touchtoneb508 (tctb508) was identified in a chemical (ethyl-nitrosourea, ENU) mutagenesis screen for mutations affecting neural crest derivatives as previously described (Henion et al., 1996). The mutant was identified based on its altered pigment pattern. At 48 hours post-fertilization (hpf), dramatically fewer melanophores are present in tctb508 mutant embryos compared to wild-type siblings (Fig. 3.1A, B). Those that are present have a small, punctate morphology, as opposed to the large, stellate morphology of wild-type melanophores (Fig. 3.1A, B insets). This phenotype does not result from a general developmental delay since the rate of development of a variety of other tissues and cells analyzed, including the pigmented retinal epithelium, is indistinguishable from wild-type embryos (see, for example, Fig. 3.1 and not shown). Subsequently, three additional ENU-induced mutations have been identified as tct alleles by complementation analysis (not shown) and genetic mapping (Rawls et al., 2003). As shown in Figure 3.1, these alleles vary in severity of the melanophore phenotype. tctb508 is the strongest mutation with respect to reductions in the number and size of melanophores, followed by tctj124e1 (Rawls et al., 2003), tctos1, and tctos2 in decreasing order of severity. All
48 alleles are recessive and homozygous lethal. Although swim bladders inflate and
the larvae eat, larvae die between 14 and 16 days post-fertilization (dpf). The
reason for lethality is currently unknown, but melanogenesis is not required for
zebrafish viability (Kelsh et al., 2000b; Lister et al., 1999), suggesting that
lethality results from a defect unrelated to the melanophore phenotype. All data
refer to tctb508 unless otherwise specified.
tct selectively affects melanophore development within the neural crest
lineage
Since neural crest-derived melanophore development is clearly disrupted
in tct mutant embryos, we investigated the development of other cell types generated from the neural crest (Table 3.1). Visual inspection revealed that the other classes of chromatophores, xanthophores and iridiphores, both appear normal with respect to number and pattern (Fig. 3.1, arrowheads; Fig. 3.2A, B).
In addition, the expression of molecular markers for precursors of these cell types (Parichy et al., 2000a; Parichy et al., 2000b) is qualitatively indistinguishable between mutant and wild-type embryos (data not shown).
Molecular markers indicated that neural crest-derived neuronal and glial populations develop normally in tct mutant embryos. For example, cervical sympathetic neurons, enteric neurons, and neurons and glia (crestin+/α-Hu-; see
Luo et al., 2001) of the dorsal root ganglia are all present in qualitatively normal numbers and characteristic locations in tct homozygotes (Fig. 3.2E – J and data not shown). Craniofacial cartilage was labeled with Alcian Blue and also found to
49 be normal in terms of individual elements, their sizes and shapes (Fig. 3.2C, D).
In addition, we used the pan-neural crest marker crestin to analyze neural crest
populations at different embryonic stages. In wild-type embryos, crestin is first
expressed in neural crest cells at the boundary of neural and non-neural
ectoderm during gastrulation. Expression continues in premigratory and
migratory neural crest cells, and persists until slightly after overt differentiation of
neural crest derivatives, such that by 24 hpf, crestin-expressing cells are found throughout the embryo (Luo et al., 2001a). crestin expression is normal at all stages in tct embryos up to 27 hpf, suggesting that the early neural crest is unaffected by the tct mutation (data not shown). However, a deficit in crestin expression is observed in tct mutant embryos concomitantly with a reduction in the numbers of cells expressing melanoblast markers (see below). These data indicate that within the embryonic neural crest lineage, tct is required specifically
for melanophore development and only after neural crest dispersal has largely
occurred.
Most melanophores are absent in tct embryos
In wild-type zebrafish embryos raised at 28.5oC, melanization of neural
crest-derived melanophores begins at approximately 25 hpf, in cells just posterior
to the otic vesicle (Kimmel et al., 1995). Similarly, melanophores begin to overtly
differentiate at this time and in the same location in tct mutant embryos.
However, tct mutants are first distinguishable from wild-type siblings at 27 hpf. tct
melanophores are small and punctate, in contrast to wild-type melanophores,
50 which are large and have many processes (Fig. 3.3A, B). In addition, fewer melanized cells have developed over the yolk of tct mutant embryos compared to wild-type siblings at this stage. The number of differentiated melanophores increases dramatically between 25 hpf and 36 hpf in wild-type embryos. In contrast, there is a substantially smaller increase in melanized cells in tct embryos during the same time period (Table 3.2). Specifically, at 27 hpf, there is a 10% depletion in the number of melanophores in tct homozygotes compared to wild-type siblings. By 36 hpf, this depletion in melanophore cell numbers between mutant and wild-type embryos increased to 36% (p<0.0001; Fig. 3.3C, D), and at
4 dpf, tct mutant embryos have an average of 64% fewer melanophores than wild-type siblings (p<0.0001; Table 3.2). These data indicate that a major complement of melanophores that normally differentiate in wild-type embryos fail to develop in tct mutants. As a result, melanophore development is severely disrupted in tct mutant embryos. Interestingly, while the number of melanophores remains reduced in tct mutant embryos, the size and morphology of the remaining melanophores recovers by 6 dpf (Fig. 3.3E, F). By this stage, tct melanophores are morphologically indistinguishable from those of wild-type siblings.
51 Melanoblast numbers in tct mutants are reduced during the second day of development
To determine whether the deficit in melanophore numbers in tct mutant embryos resulted from a decrease in the number of melanophore precursors
(melanoblasts) or simply a failure of melanoblasts to produce melanin (Kelsh et al., 2000b; Lister et al., 1999), we used lineage-specific markers c-kit and dct to identify melanophores and their precursors during development. Expression of c- kit and dct in un-melanized neural crest-derived cells (melanoblasts) in tct mutant embryos was qualitatively indistinguishable from wild-type siblings during the first day of development, and comparable quantitatively at 25 hpf, just as melanophores begin to differentiate (data not shown; Table 3.3). At later stages, however, the number of cells (melanoblasts and melanophores) identified with both markers was reduced in tct mutants compared to wild-type embryos (Fig.
3.4). We quantified both c-kit+ and dct+ cells present in tct mutant embryos and wild-type siblings (see Methods). At 25 hpf, c-kit+ and dct+ cell counts reveal wild- type numbers of melanoblasts and nascent melanophores in tct mutant embryos
(p>0.1; Table 3.3). In contrast, there is approximately a 10% percent reduction of these markers at 27 hpf (Fig. 3.4A, B; Table 3.3). This difference between wild- type and tct embryos becomes more dramatic over time. By 36 hpf, there is nearly a 40% deficit in cells expressing each of these genes in tct mutant embryos (p<0.0001; Fig. 3.4C, D; Table 3.3). Importantly, the numbers of c-kit+ and dct+ cells increase in wild-type embryos between 25 hpf and 36 hpf (Table
52 3.3). Over the same time period, however, the absolute numbers of c-kit+ and
dct+ cells decrease in tct mutant embryos (Table 3.3). In addition, crestin expression in tct mutants appears to be qualitatively similar to the reduction of c- kit- and dct-expressing and melanin-containing cells at 27 hpf (Fig. 3.4E, F). In contrast, mitfa expression in tct mutant embryos is qualitatively normal through
25 hpf. In wild-type embryos, mitfa expression is down-regulated prior to overt differentiation of melanophores (Lister et al., 1999), and mitfa expression appears to be unaffected in tct mutant embryos (data not shown).
Thus, wild-type numbers of melanoblasts are generated in tct
homozygotes during the first day of development, and the number of
differentiated melanophores increases slightly between 25 and 36 hpf (Table
3.2). However, the numbers of c-kit+ and dct+ cells fail to expand, and are
actually reduced during subsequent stages of development while melanoblast
and melanophore numbers are rapidly increasing in wild-type embryos. Together,
these data indicate that melanoblast cell numbers are reduced in tct mutant
embryos.
Melanoblasts undergo apoptosis in tct mutant but not wild-type embryos
The reductions in the numbers of c-kit+ and dct+ cells in tct mutants and
the subsequent absence of the majority of melanophores suggests that the
defect resulting from the tct mutations arises during melanoblast development.
Specifically, the difference in the melanoblast numbers in tct mutant embryos
compared to wild-type siblings indicated that melanoblasts might undergo cell
53 death and/or fail to proliferate in tct embryos. Further, the absolute number of melanoblasts in tct homozygotes is less at 36 hpf compared to 25 hpf. This suggested that at least some melanoblasts die in tct embryos. In order to test this hypothesis, we first examined cell death using TUNEL in tct and wild-type embryos between 27 hpf and 36 hpf. A small number of TUNEL-positive cells located in embryonic positions normally occupied by melanoblasts were observed in some tct mutant embryos but were never observed in wild-type embryos (data not shown). To determine whether any of these apparently dying cells were in fact melanoblasts, tct homozygotes and wild-type siblings were labeled with both an activated Caspase-3 antibody for apoptotic cells and dct riboprobe for melanoblasts. Dying melanoblasts, un-melanized cells labeled with both activated Caspase-3 and dct, were observed in tct embryos at 29 hpf (Fig.
3.5). In contrast, apoptotic melanoblasts were not observed in wild-type embryos.
In order to determine whether differentiated melanophores were also dying in tct mutant embryos, 30 hpf mutant and wild-type embryos, each containing differentiated melanophores, were labeled with the TUNEL assay. No TUNEL+ melanized cells were detected in either tct or wild-type embryos (data not shown). Therefore, the reduction in melanophore cell numbers in tct mutant embryos is due at least in part to death of melanoblasts, but not melanophores.
However, it remains possible that the reduction in melanoblast numbers in tct embryos may also result from other factors such as reduced proliferation or transdifferentiation.
54 tct acts cell autonomously with respect to melanophore development
Genetic mosaics can be used to determine the cell autonomy of a mutation and thus predict the mode of action of a gene product within developmental pathways. Mosaic analysis was performed between embryos from wild-type and tct heterozygote crosses (Table 3.4). Donor embryos were labeled with lysinated rhodamine dextran (LRD) or lysinated fluorescein dextran (LFD).
Cells from these embryos were transplanted into unlabeled hosts, in which melanophore development was subsequently observed. As controls, we noted that wild-type donor cells in wild-type hosts generated large melanophores and that tct donor cells generated melanophores with mutant morphology in mutant hosts (Fig. 3.6E). LRD-labeled cells from wild-type embryos formed large, stellate melanophores when transplanted into tct hosts (n=10; see Fig. 3.6A, B, D).
These results strongly suggested that tct acts cell autonomously within melanoblasts. Further, these results predict that tct donor cells in wild-type hosts should be capable of generating melanophores albeit, based on the mutant phenotype, with reduced frequency due to reduced melanophore numbers (Fig.
3.1, 3.3; Table 3.2) and with abnormal cellular morphology (Fig. 3.1). Therefore, in the reciprocal experiment, cells from LRD-labeled tct embryos and LFD- labeled wild-type embryos were simultaneously transplanted into the same region of a single wild-type host (Table 3.4). tct donor cells did give rise to melanophores with tct morphology in wild-type hosts (Fig. 3.6C). However, in several cases in which LFD-labeled (wild-type) cells gave rise to melanophores,
LRD-labeled cells from tct embryos failed to give rise to melanophores (n=5),
55 qualitatively consistent with the mutant phenotype in which fewer melanophores are generated compared to wild-type embryos. Together with our results indicating that tct function is required by melanoblasts as opposed to melanophores, these data suggest that tct is required cell autonomously for melanoblast development.
tct embryos are touch insensitive
While no other neural crest derivatives are noticeably affected by tct, there
are other defects associated with the mutation. By 26 hpf, wild-type embryos
exhibit three motor behaviors: spontaneous contractions, touch response
(swimming away from a stimulus) and swimming response (turning 180o and
swimming away from a stimulus) (Saint-Amant and Drapeau, 1998). While all of
these behaviors develop initially in tct embryos, mutants are inflexible and touch
insensitive beginning at approximately 39 hpf. tct homozygotes neither respond
to touch, nor swim on their own. Further, the trunks of touch insensitive embryos
are rigid. However, locomotive behavior and touch responses recover completely
after 72 hpf. During the period of touch insensitivity, the heart rate of tct embryos
is also depressed by nearly 30% relative to wild-type siblings (data not shown).
Like lethality, touch insensitivity is characteristic of all tct alleles. The reasons for
immobility and diminished heart rate have not been determined, and while
reduced heart rate could result from immobility and vice versa, these additional
phenotypes may ultimately provide clues to the cause of lethality in tct
homozygotes.
56
The tct locus is located on chromosome 18
As an initial step toward cloning touchtone, we placed alleles tctb508, tctos1,
and tctos2 on the zebrafish genomic linkage map (Postlethwait et al., 1994;
Postlethwait et al., 1998). Simple sequence length polymorphisms (Knapik et al.,
1996; Knapik et al., 1998) between AB* and WIK zebrafish strains were used to
map tct to chromosome 18. Parthogenetic diploid embryos were generated by suppressing the second meiotic division using the early pressure (EP) method
(Streisinger et al., 1981). tct and wild-type embryos were identified by live
phenotype, and PCR was performed on DNA from tct and wild-type embryos.
The tct locus was initially localized to chromosome 18 in EP diploid embryos
using three markers near the proximal telomere, z7654, z9404 and z7426 (see
Methods). tctj124e1 was independently found to cosegregate with SSLP markers z7654 and z11685 (Rawls et al., 2003). Subsequently, tct was found to be tightly
linked to SSLP marker z53176, with 0 recombinants in over 1,400 meioses
analyzed to date (see Methods). Ultimately, identification of the tct locus will not
only be important for precisely elucidating its role in melanogenesis, but will also
provide valuable insight into the cause of larval lethality of tct homozygotes.
Analysis of available genomic sequence
(http://www.pre.ensembl.org/Danio_rerio/) revealed that no other genes with
demonstrated roles in melanophore development in zebrafish are known to be located within close proximity to the marker sequence, raising the possibility that tct represents a novel regulator of melanoblast development.
57
DISCUSSION
Zebrafish tct function is selectively required within the embryonic melanophore neural crest sublineage
The most obvious visible phenotype of live tct mutant embryos is the reduction in melanophore cell number and the abnormal morphology of the melanophores that are present. Because melanophores are unnecessary for zebrafish viability (Lister et al., 1999) and yet all four alleles of tct that we have described are lethal, we examined the development of other neural crest-derived cells, many of which are essential for embryonic development. We found that the development of other neural crest derivatives in tct mutant embryos was indistinguishable from wild-type embryos. Thus, tct appears to be selectively required for the development of the melanophore sublineage. Nevertheless, the development of some other cell type(s) that are required for viability must be abnormal as a result of disrupted tct function, as mutant larvae die during the third week of development. Our results suggest that the affected cell type(s) are not of neural crest origin and may instead be components of one of several internal organ systems that undergo extensive maturation during the period of lethality of tct homozygous larvae. In this vein, the presumptive pleiotropic effects of the tct mutations are reminiscent of human neurocristopathies (Jackson, 1997;
Nordlund et al., 1998). However, it cannot be ruled out that the function, as opposed to embryonic differentiation, of one or more neural crest derivatives is abnormal and leads to larval lethality.
58
Melanoblasts require tct function relatively late during embryogenesis but
prior to overt differentiation
We quantified the reduction in melanophore and melanoblast numbers
during development of tct mutant embryos compared to wild-type embryos. We
found that the majority of the normal melanophore population is absent in tct
mutant embryos. The number of melanoblasts was also found to be lower in tct
embryos compared to wild-type siblings and, strikingly, the absolute number of
melanoblasts declines in tct mutants during the same period of development in
which the number of melanoblasts in wild-type embryos increases. These
observations raised the possibility that the deficit in melanophore numbers in
mutant embryos may result from a defect in melanoblast development that
results in the death of a subpopulation of melanoblasts. Two other observations
in addition to the quantitative data support this notion. First, TUNEL positive cells in positions consistent with the normal locations of melanoblasts during development were observed in tct but not wild-type embryos. Second, we
detected a small number of dying melanoblasts (cells expressing both activated
Caspase-3 and dct mRNA but not melanin) in tct mutant embryos that were
never observed in wild-type embryos. In addition, dying (TUNEL+) melanophores
were not observed in either mutant or wild-type embryos. Taken together, these
results support the view that abnormal tct function in mutant embryos results in
the death of a subset of melanoblasts, which in turn contributes to the observed
deficits in melanophore cell numbers. The death of some melanoblasts in tct
59 mutant embryos may occur for one or more of a variety of reasons. Given the cell autonomous function of tct with respect to melanophore development, it is possible that, analogous to c-kit, tct normally functions as a component of the signaling triggered by a trophic factor. It is equally likely, however, that tct normally is required to complete the melanoblast developmental program and in the absence of normal tct function a subset of melanoblasts undergoes apoptosis as a result of failure to complete this program.
However, it is not possible to conclude that the deficit in melanophore numbers in tct mutant embryos is due exclusively to the death of melanoblasts.
The occurrence of dying melanoblasts in mutant embryos is a relatively infrequent event. This could be due to a variety of reasons, including the small number of neural crest-derived cells in zebrafish generally as compared to other vertebrates, as well as the difficulty in detecting cells that express indicators of cell death while simultaneously maintaining detectable levels of transcripts diagnostic of melanoblasts. Thus, the frequency of detecting any of a total of ca.
30 dying cells over an 11 hour period of development in the entire trunk region of embryos (see Table 3.3) that are verifiably melanoblasts would likely be low. The fact that the total melanoblast population is relatively small also complicates potential approaches to quantifying alterations in melanoblast proliferation between mutant and wild-type embryos. Therefore, while we provide evidence that melanoblast cell death is a result of the tctb508 mutation, we cannot exclude that the mutation also results in additional developmental changes. Melanoblast proliferation or other processes including differentiation (see below) or
60 transdifferentiation may be affected, although we have seen no evidence for the
latter in our analysis.
Our results are consistent with a selective role for tct in melanoblast
development prior to overt differentiation. We suggest that tct function is required
relatively late in melanoblast development. Both the early neural crest cell
population, based on crestin expression, and the early melanoblast population,
identified by mitfa expression, are indistinguishable in tct mutant and wild-type
embryos up to approximately 27 hpf just after the first differentiated
melanophores begin to appear. Likewise, c-kit and dct-expressing melanoblast
populations are indistinguishable up to this stage. Subsequently, a decrease in
the number of melanoblasts, and ultimately the number of melanophores,
becomes progressively more apparent. These results indicate that tct function is
required by a subpopulation of melanoblasts downstream of mitfa, and most
likely downstream, or concomitant with, c-kit. This would place the requirement
for tct function by melanoblasts relatively late in the developmental pathway, but
prior to overt differentiation.
Lastly, it is noteworthy that all of the minority population of melanophores that do develop in tct mutant embryos exhibit an abnormally small, punctate morphology compared to the large, stellate morphology of wild-type melanophores. Analysis of the distribution of dct mRNA relative to melanin
granules within these cells indicates that indeed the cell morphology, in contrast
to melanin distribution, is abnormal in these cells (not shown). Although this
aspect of the tct phenotype eventually recovers, it raises the possibility that tct
61 function is also involved in the cell morphological differentiation of melanophores.
Thus, while some melanophores, albeit abnormal in morphology, differentiate in
tct embryos, some melanoblasts may perish as a result of the inability to
complete the melanophore differentiation program and thereby contribute to the
melanophore phenotype of mutant embryos.
Developmental diversity within the melanophore sublineage
Previous studies indicate that mitf is necessary and sufficient to establish
the melanoblast population during neural crest development (Hornyak et al.,
2001; Lister et al., 1999; Tachibana et al., 1996). Subsequently, melanoblast
development has been shown to require additional genes (Nakamura et al.,
2002). In zebrafish, for example, c-kit/sparse has been shown to be required for
melanoblast and melanophore survival. However, not all melanoblasts require C-
kit function for survival and development. Even presumptive null alleles of c-kit
retain up to 58% of melanophores compared to wild-type siblings at 60 hpf
(Parichy et al., 1999). Further support for the notion of developmentally distinct
melanogenic subpopulations of neural crest cells is found in recent studies of the
zebrafish lockjaw (low) and mont blanc (mob) mutations, which disrupt an ap2α homolog (Knight et al., 2004; Knight et al., 2003). low and mob alleles (Knight et al., 2004; Knight et al., 2003) result in a loss of some but not all early embryonic melanophores, whereas melanophore numbers recover slightly at later stages. In addition, homozygous c-kit/sparse mutant embryos injected with morpholinos complementary to ap2α have fewer melanophores than uninjected c-kit/sparse
62 mutants, consistent with the notion of heterogeneity in the developmental
requirements of subpopulations within the melanophore sublineage (O'Brien et al., 2004). Further, some melanophores do develop in such embryos indicating
that a subpopulation of melanogenic cells develop independently of both ap2α and c-kit. Sox10 has also been implicated in melanoblast survival based on analysis of the colourless mutation (Dutton et al., 2001b; Kelsh and Eisen, 2000).
Likewise, many other zebrafish mutants exhibit phenotypes in which melanophore numbers are reduced but not completely abolished (Henion et al.,
1996; Kelsh et al., 1996; Rawls et al., 2003). Therefore, it is not altogether surprising that tct mutant embryos lack most, but not all melanophores. On the one hand, this could result if all of the alleles identified are hypomorphic. In contrast, and taken together with the phenotypes of other melanophore mutants, the tct mutant phenotype may further support the notion of developmental heterogeneity within the melanophore sublineage. Specifically with respect to tct, whereas most melanoblasts fail to generate melanophores in a tct-dependent manner, other melanoblasts do generate melanophores, albeit transiently abnormal in morphology. This suggests that two distinct populations of melanoblasts, tct-dependent and tct-independent, contribute to the melanophore sublineage. Together with the apparent heterogeneity in the melanoblast population suggested by the analysis of c-kit/sparse and other mutants (Dutton et al., 2001b; Knight et al., 2004; Parichy et al., 1999), it appears that different subsets of melanoblasts have different genetic requirements for development.
63 Further definition of these potential subpopulations should be achieved by the
analysis of melanogenesis in compound mutants.
The concept that the melanogenic neural crest sublineage is comprised of
subsets of precursors cells that differ in their genetic requirements for development is reminiscent of the developmental heterogeneity that exists in different neurogenic neural crest sublineages. For example, the precursors of dorsal root ganglion sensory neurons that will ultimately sub-serve different
sensory modalities have distinct trophic requirements prior to neuronal
differentiation (Ernfors, 2001; Henion et al., 1995; Oakley et al., 1995). In
addition, during sympathetic neuron development, sympathoblasts and nascent
neurons undergo non-uniform temporal shifts in their neurotrophic survival
requirements (Birren et al., 1993; DiCicco-Bloom et al., 1993; Francis and
Landis, 1999). Along these lines, although the function for c-kit signaling in
melanoblast survival has been firmly established (Ito et al., 1999; Morrison-
Graham and Weston, 1993; Rawls et al., 2003; Wehrle-Haller et al., 2001;
Wehrle-Haller and Weston, 1995), it is not clear that the tct locus functions as a
component of a classic trophic system for melanoblasts. Nevertheless, our
results and results from the analysis of melanophore development generally, in
other melanophore mutants and otherwise, raise the interesting possibility of
precursor cell heterogeneity within the melanogenic sublineage.
Finally, we are in the process of identifying the tct locus by positional
cloning. Analysis of genetic mosaics together with our analysis indicating a
requirement for tct function by melanoblasts, suggest that tct is required cell
64 autonomously for melanoblast development. The autonomy of the tct mutation
will undoubtedly prove insightful for the identification of the tct locus, as well as
for the analysis of compound mutations involving other genes required for
melanogenesis. Importantly, the availability of multiple mutant alleles is ultimately
likely to reveal a detailed understanding of the function of the tct locus and its
regulation as well as providing an explanation for the larval lethality of the
mutations.
MATERIALS AND METHODS
Zebrafish
Adult zebrafish and embryos were maintained in the Ohio State University
zebrafish facility. Adults and embryos were reared at ~28.5oC and embryos were
staged based on morphological criteria, according to Kimmel et al., 1995. Mutant
lines were maintained in the AB* and WIK backgrounds. Homozygous mutant
embryos and wild-type siblings were obtained by crossing heterozygous carriers.
Cell counts and statistics
Cells expressing melanin, c-kit or dct were counted in 25 hours post-fertilization
(hpf), 27 hpf, 32 hpf, and 36 hpf embryos. Embryos were mounted laterally on double-bridge cover slips and viewed on a Zeiss Axioplan microscope. Embryos used for melanin+ cell counts were fixed and cells were counted over one entire
65 side of each embryo. Embryos for c-kit+ or dct+ cell counts were processed for in situ hybridization with the corresponding riboprobe. Prior to being fixed, embryos for 32 hpf and 36 hpf c-kit+ and dct+ cell counts were first identified at 27 hpf as either tct mutants or wild-type siblings, then placed in phenylthiourea (PTU, 0.3 g/L) to inhibit melanin deposition until fixation. Melanophores in 4 days post- fertilization (dpf) larvae were counted in the dorsal and ventral stripes. To better visualize distinct melanophores, wild-type and tct larvae were placed in epinephrine (10 mg/ml) at 4 dpf for ~10 minutes, resulting in redistribution of melanosomes to the center of the cell body (Johnson et al., 1995; Rawls and
Johnson, 2000). Larvae were fixed in 4% paraformaldehyde at 4oC overnight, rinsed and stored in 1:1 PBS:glycerol. Larvae were then deyolked, mounted on single bridge cover slips, and viewed on a Zeiss Axioplan microscope.
Melanophores in both the dorsal and ventral stripes were counted from somite 5 to somite 14. Standard errors of mean were calculated for wild-type and tct cell counts at each time point. The mean numbers of cells in wild-type and tct embryos at each time point were also subjected to a one-tailed T-test to determine whether the decrease in cell numbers in tct mutant embryos compared to wild-type siblings was significant.
In situ hybridization and TUNEL
In situ hybridizations were performed as described by Thisse, et al. (1993) with minor modifications. A detailed protocol will be provided upon request. Mitfa and
66 c-ret cDNAs were provided by D. Raible (Bisgrove et al., 1997; Lister et al.,
1999). cDNA clones of c-kit and fms were gifts from D. Parichy (Parichy et al.,
2000b; Parichy et al., 1999). Dct and ednrb1 cDNA clones were provided by R.
Kelsh (Kelsh et al., 2000b; Parichy et al., 2000a). The TUNEL (Terminal
Transferase Assay; 220582; Roche) protocol was modified from the manufacturers suggested protocol and will be provided upon request (Cole and
Ross, 2001).
Immunohistochemistry
Antibody labeling was performed as previously described (An et al., 2002). 48 hpf
embryos were sectioned onto gelatin-subbed slides and stored at -20oC
overnight. All neurons were detected with monoclonal antibody 16A11 that
recognizes neuron-specific Hu RNA binding proteins (An et al., 2002; Henion et
al., 1996), while DRG neurons and enteric neurons were subsequently identified
by position within the embryo. 16A11-immunoreactivity was detected using an
Oregon Green fluorescent secondary antibody (Molecular Probes).
In situ hybridization/antibody double labeling
Embryos were fixed for several hours at room temperature, washed in PBS, and stored at 4oC. In situ hybridization was performed without proteinase K digestion.
Embryos were subsequently stored in PBS at 4oC and then processed for
immunohistochemistry (see above). Anti-activated-Caspase-3 primary antibody
67 (Cell Signaling Technology) and Oregon Green secondary antibody (Molecular
Probes) were used to label apoptotic cells. Negative control embryos were
similarly processed for in situ hybridization and subsequently labeled with secondary antibody only, i.e. without first labeling with the primary anti-Caspase-
3 antibody.
Mosaic analysis
Genetic mosaics were produced using cell transplantation techniques (Ho and
Kane, 1990). Donor embryos obtained from AB* or heterozygous (tct+/b508) crosses were manually dechorionated and injected at the one to two cell stage with 2 – 5 % lysinated rhodamine dextran (10,000 MW, Molecular Probes) or lysinated fluorescein dextran (10,000 MW, Molecular Probes) in 0.2M KCl.
Embryos were then allowed to develop to early blastula stages. For wild-type Æ mutant transplants, 10 – 20 cells were transplanted from LRD-labeled donor embryos into unlabeled host embryos. For mutant Æ wild-type transplants, LFD- labeled wild-type cells and LRD-labeled tct cells were simultaneously transplanted into the same host embryo. Both groups of cells were drawn into the same electrode and placed together in the host. For transplants in both directions, donor-host pairs were kept separate and allowed to develop to >30 hpf, then fixed in 4% paraformaldehyde at 4oC overnight. Donors and hosts were
classified as either wild-type or mutant based on melanophore phenotype, and
subsequently examined using a Zeiss Axioplan microscope with Nomarski optics,
and fluorescein or rhodamine filters to detect donor cells. Punctate melanophores
68 were scored as mutant, while stellate melanophores were considered wildtype.
Wild-type Æ wild-type and mutant Æ mutant transplants were used as controls.
Genetic mapping
tct alleles were maintained in AB* background. For mapping purposes, tct
carriers in this background were crossed to a wild-type WIK line, which is
polymorphic with AB* (Nechiporuk et al., 1999). tct carriers in the WIK
background were then used to generate parthogenetic diploid progeny by
suppressing the second meiotic division with early pressure (Streisinger et al.,
1981). tct embryos and wild-type siblings were identified by live phenotype and
used to obtain DNA. tct was initially placed on the zebrafish genomic map based on PCR amplification of simple sequence length polymorphisms (SSLPs) from
diploid genomes (Knapik et al., 1996; Knapik et al., 1998; Postlethwait et al.,
1994; Postlethwait et al., 1998). Three tct alleles, tctb508, tctos1 and tctos2 were
demonstrated in this laboratory to be linked to SSLP markers near the proximal
telomere of chromosome 18 based on the MGH mapping panel
(http://zebrafish.mgh.harvard.edu/zebrafish/index.htm). tctos1 and tctos2 were
shown to cosegregate with z7654 (map position 14.6 cM), and tctos2 also
cosegregated with z7426 (7.3 cM). tctb508 was linked to z7654, as well as z9404
(10.0 cM). tctj124e1 was independently mapped to chromosome 18 based on
cosegregation with z11685 (7.3 cM) and z7654 (Rawls et al., 2003). Further
linkage analysis was performed on tctb508 using haploid genomes, or meioses
69 (Postlethwait et al., 1994). Analysis with this allele reveals close linkage to microsatellite marker z53176 (6.2 cM), with zero recombination events observed in over 1,400 meioses evaluated.
70 TABLES AND FIGURES
Neural crest-related population Marker Timing Phenotype Early neural crest cells crestin 14s-24hpf - Xanthophores Pteridine pigment 48 hpf - Iridiphores Refractive organelles 72 hpf - Enteric neurons/precursors c-ret 36 hpf - Dorsal Root Ganglion Neurons anti-Hu/location 48 hpf - Sympathetic neurons TH 4 dpf - Neural crest-derived glia crestin+/anti-Hu- 48 hpf - Craniofacial cartilage Alcian Blue 5 dpf -
Table 3.1. Other neural crest derivatives are normal in tct mutant embryos.
71
27 hpf 36 hpf 4 dpf* melanin Wild-type 83.6 + 4.4 144.8 + 5.2 91.8 + 3.4 n=10 n=6 n=10 tct 74.9 + 3.1 92.7 + 3.7 33.5 + 2.9 n=9 n=6 n=10 % compared to wt 89.6 64.0 36.5 p > 0.01 p < 0.0001 p < 0.0001
Table 3.2. Melanophores are reduced in tct mutants.
72
25 hpf 27 hpf 32 hpf 36 hpf Wild-type 120.2 + 2.9 121.1 + 3.4 120.5 + 6.6 148.5 + 7.6 dct n=25 n=11 n=11 n=10 tct 119.0 + 8.3 109.2 + 2.7 90.9 + 2.7 91.4 + 5.1 n=6 n=13 n=15 n=11 p > 0.1 p < 0.01 p < 0.0001 p < 0.0001
Wild-type 100.8 + 2.0 99.9 + 2.3 116.7 + 5.6 ND c-kit n=49 n=29 n=10 tct 95.9 + 4.0 90.4 + 5.1 86.6 + 3.8 ND n=11 n=12 n=10 p > 0.1 p > 0.01 p < 0.001
Table 3.3. Depletion of melanoblasts in tct embryos.
73
Number of hosts Transplantation donor with donor-derived into host melanophores
wild-type into mutant 10 mutant into wild-type* 1 wild-type into wild- type* 5
Table 3.4. tct acts cell autonomously with respect to melanophore development. *Both mutant and wild-type donor cells were transplanted simultaneously into the same host embryo (see Methods).
74
Figure 3.1. Visible phenotype of live touchtone mutant embryos. (A, A inset) At 72 hpf, wild-type embryos have large, stellate melanophores. Yellow xanthophores are apparent along the dorsal aspect of the embryo (arrowhead). (B – E) tct embryos tctb508 (B), tctj124e1 (C), tctos1 (D), tctos2 (E) have fewer melanophores than wild-type embryos, forming an allelic series in which tctb508 is the most severe. Those melanophores that are present are small and punctate (B, inset). Xanthophores, however, are normal in all tct alleles (arrowheads).
75
Figure 3.2. tct selectively affects melanophores within the neural crest lineage. Wild-type (A, C, E, G, I) and tct (B, D, F, H, J) mutant embryos. tct mutant embryos have a wild-type number and pattern of iridiphores at 3 dpf (A, B: lateral view). Craniofacial cartilages revealed with alcian blue staining are normal at 8 dpf (C, D: ventral view). (E, F) Cervical sympathetic neurons, which express TH mRNA (arrowheads in E, F), are indistinguishable between wild-type (E) and tct mutant (F) embryos at 77 hpf. Hu-positive neurons of the dorsal root ganglia (G, H, arrowheads) and the enteric nervous system (arrowheads in I, J) also appear normal in tct mutant embryos (H, J) compared to wild-type siblings (G, I).
76
Figure 3.3. Melanophores are reduced in number and size in tct mutant embryos. Wild-type (A, C) and tct mutant (B, D) embryos at 27 hpf (A, B) and 36 hpf (C, D); dorsal views with anterior to the left. At 27 hpf, tct mutant embryos have fewer melanophores, especially over the yolk, compared to wild-type embryos. By 36 hpf, many more melanophores have differentiated in wild-type embryos than in tct mutant embryos. Although melanophore numbers do not recover in tct homozygotes (F), melanophore morphology (arrowheads) becomes indistinguishable from that of wild-type siblings (E) by 6 dpf.
77
Figure 3.4. Melanoblast numbers are reduced in tct mutant embryos in the second day of development. (A, B) Dorsal views of dct expression in 27 hpf wild- type (A) and tct mutant (B) embryos. tct homozygotes have a slight reduction of dct+ cells compared to wild-type embryos at this stage. (C, D) At 36 hpf, the number of dct+ cells is greater over the yolk (arrowheads) and throughout the body of wild-type embryos. tct mutant embryos (D) have many fewer dct+ cells, especially over the yolk (arrowheads). Expression of c-kit follows similar trends (see Table 3.3). (E, F) Concomitant with the reduction in dct-expressing cells in tct mutant embryos, the number of crestin-expressing neural crest cells is also reduced in tct mutant embryos compared to wild-type siblings at 28 hpf (F).
78
Figure 3.5. Melanoblasts undergo apoptosis in tct mutant embryos. Transverse section through the trunk region of a 29 hpf tct mutant embryo. Un-melanized dct-positive (A, C)/activated Caspase-3-positive (B, C) melanoblasts were observed in tct homozygotes, but never in wild-type siblings.
79
Figure 3.6. tct is required cell autonomously for melanophore development. (A) Nomarski image of a tct mutant host that has received cells from a wild-type donor (anterior to the left). Several large melanophores (arrowheads) are present in addition to punctate melanophores (arrows) characteristic of tct mutant embryos. (B) High magnification of wild-type cells (containing LRD lineage dye; boxed area in A) that formed stellate melanophores in the mutant environment. (C) tct cells transplanted into wild-type host embryos can give rise to punctate melanophores (arrowhead), in contrast to large wild-type host melanophores (arrow). The white outlines approximate the size of each melanophore. (D, E) In tct mutant hosts, both wild-type and tct cells give rise to melanophores. Fluorescein-labeled cell from the wild-type donor (D, arrowhead) and a small rhodamine-labeled melanophore from the tct mutant donor (E, arrowhead) at the same magnification.
80
CHAPTER 4
DEFECTIVE SKELETOGENESIS WITH KIDNEY STONE FORMATION IN DWARF ZEBRAFISH MUTANT FOR TRPM73
ABSTRACT
Development of adult form requires coordinated growth and patterning of multiple
traits in response to local gene activity as well as global endocrine and
physiological effectors. An excellent example of such coordination is the
skeleton. Skeletal development depends on differentiation and morphogenesis of
multiple cell types to generate elements with distinct forms and functions
throughout the body (Fisher et al., 2003; Karsenty and Wagner, 2002;
Kronenberg, 2003). We show that zebrafish touchtone/nutria mutants exhibit
severe growth retardation and gross alterations in skeletal development in
addition to embryonic melanophore and touch response defects (Arduini and
Henion, 2004; Cornell et al., 2004). These alterations include accelerated
endochondral ossification but delayed intramembranous ossification, as well as skeletal deformities. We show the touchtone/nutria phenotype results from
mutations in trpm7, which encodes a transient receptor potential (TRP) family
3 Reprinted from Current Biology, Vol. 15, Elizondo, M.R., Arduini, B.L., Paulsen, J. , MacDonald, E.L., Sabel, J.L., Henion, P.D., Cornell, R.A. and Parichy, D.M. Defective skeletogenesis with kidney stone formation in dwarf zebrafish mutant for trpm7, pp667-671, Copyright 2005, with permission from Elsevier. Contributions of B.L. Arduini to this work include morpholino phenocopy of tct phenotypes, mapping, sequencing portions of trpm7j124e1, trpm7os1 and trpm7os2 alleles, and cation rescue experiments. 81 member that functions as both a cation channel and kinase. We find trpm7
expression in the mesonephric kidney and show that mutants develop kidney
stones, indicating renal dysfunction. These results identify a requirement for trpm7 in growth and skeletogenesis and highlight the potential of forward genetic approaches to uncover physiological mechanisms contributing to the development of adult form.
RESULTS AND DISCUSSION
Genetic screens for ethyl-N-nitrosourea-induced mutations affecting zebrafish
postembryonic development uncovered the nutriaj124e2 mutant, named for its
small size, odd shape, and tendency to swim near the surface (Figures 4.1A and
4.1B). As embryos and early larvae (2–5 days post-fertilization [dpf]), nutria are
comparable in size to wild-type siblings but during later development they exhibit
a severe growth deficit (Figure 4.1C). Craniofacial and trunk body proportions are
altered, although other external features (e.g., scales, fins, complement of adult
pigment cells) are not grossly abnormal. To identify the locus causing this dwarf
phenotype, we mapped the nutria mutation to chromosome 18 in the vicinity of
touchtone (tct) (Arduini and Henion, 2004; Cornell et al., 2004). Both tct and nutria mutants exhibit embryonic melanophore deficiencies and touch unresponsiveness prior to hatching. Complementation tests confirmed that nutria and tct are allelic. Fine mapping of the critical region revealed, among other genes, trpm7, which encodes an orthologue of the transient receptor potential
(TRP) melastatin-7 dual-function cation channel and kinase (0/2889
82 recombinants) (Wolf, 2004). Although TRPM7 has not been implicated in growth
previously, a member of this family, TRPM1 (melastatin), is expressed in human
melanocytic nevi (Duncan et al., 1998), and zebrafish melanophores require tct
cell autonomously (Arduini and Henion, 2004; Cornell et al., 2004).
Our data show that trpm7 corresponds to tct. Sequencing trpm7 cDNAs
revealed premature stop codons in the severe alleles tctj124e1 and tctb508 (Figure
4.2D and 4.2E, see Methods). Moreover, injection of wild-type embryos with a
trpm7 splice-blocking morpholino oligonucleotide results in both melanophore
deficiencies and touch unresponsiveness and thus phenocopies the mutant
(Figure 4.2A–C, see Methods). Finally, TRPM7 acts as an inwardly rectifying
cation channel with broad specificity but high affinity for Mg2+ and Ca2+,
suggesting that manipulating divalent cation availability might rescue embryonic
phenotypes of trpm7 mutants (Monteilh-Zoller et al., 2003; Schmitz et al., 2003).
Supplemental Mg2+ partially rescued melanophore development, whereas supplemental Ca2+ partially rescued melanophore development and touch
responsiveness (Figures 4.1D–F, 4.3). Thus, trpm7 is the gene affected in tct
mutants.
To clarify the mode(s) of trpm7 activity, we examined trpm7 expression in
wild-type embryos and larvae. Consistent with previous observations, we
detected widespread trpm7 expression at embryonic stages, including in the
central nervous system, pronephros, lens and other tissues (Zebrafish
Information Network, 2005 and data not shown). In metamorphic larvae,
transcripts were abundant in liver, mesonephric kidney tubules, and corpuscles
83 of Stannius, which are glands in teleosts that contribute to calcium homeostasis
(Krishnamurthy, 1976). In contrast to severe alleles, trpm7j124e2 (nutria) individuals are viable, allowing us to assess functional consequences of trpm7 mutation during postembryonic development. Consistent with kidney expression of trpm7, histological examination of trpm7j124e2 mutant larvae revealed mineralization within mesonephric tubules (Figures 4.4D and 4.4F). These data suggest that altered trpm7 function in kidney, in corpuscles of Stannius, or in both affects whole organism cation homeostasis and leads to nephrolithiasis.
Because dwarfism syndromes in humans are associated with a range of skeletal defects (Hall, 2002) and cation homeostasis affects bone development and maintenance (Dvorak and Riccardi, 2004), we asked whether the growth deficit and disproportionality of trpm7j124e2 mutants is associated with changes in ossification. A comprehensive analysis of 87 bones between 9 and 51 dpf revealed extensive alterations in ossification sequence between wild-type and mutant larvae (Figures 4.5 and 4.6). Among numerous examples of sequence reordering is the epural bone of the caudal fin, which is the 78th bone to ossify in wild-type but only the 32nd to ossify in trpm7 mutants. Conversely, the maxilla is the 15th bone to ossify in wild-type but the 40th to ossify in trpm7 mutant larvae.
Thus, trpm7+ is essential for the normal sequence of ossification.
To better understand ossification changes, we categorized bones according to function and anatomical location as well as cellular origin. The most dramatic differences were between endochondral bones, which develop through mineralization of a cartilage model, and intramembranous bones, which develop
84 directly, without a cartilage model (Bird and Mabee, 2003; Cubbage, 1996;
Karsenty and Wagner, 2002). In trpm7 mutants, endochondral bones (red
connectors in Figure 4.6A) appeared on average earlier in the ossification
sequence than they did in wild-type larvae, whereas intramembranous bones
(green connectors) appeared on average later than they did in wild-type larvae.
Dissociation of endochondral and intramembranous ossification is exemplified in the caudal complex, in which the endochondral hypurals and epural ossify much earlier in trpm7 mutants compared to the wild-type, whereas the intramembranous urostyle, centra, and other bones ossify much later in trpm7 mutants compared to the wild-type (Figures 4.5A–H). Precocious endochondral ossification is similarly evident in the suspensorium and the branchial arches of trpm7 mutants (Figures 4.5I–P), whereas delayed intramembranous ossification is apparent for several bones of the anterior head (Figures 4.5Q–V). The accelerated ossification of endochondral bones and delayed ossification of intramembranous bones reflects order in the ossification sequence as well as absolute timing. One can most easily visualize these timing differences by plotting how genotype affects the likelihood of bones being ossified at any given age against the statistical significance of these differences (Figures 4.6B and
4.6C). Defects in ossification timing do not simply reflect growth retardation; even severely runted wild-type fish develop skeletal elements of normal shape without premature ossification (Figure 4.8).
Furthermore, functional–anatomical units in close proximity were found to
be differentially affected, even after we controlled for the relative contributions of
85 endochondral and intramembranous bones within these units. For example, otic
bones are accelerated, whereas orbital bones are delayed, despite the
anatomical proximity of these bones and their similar endochondral and
intramembranous compositions (Figure 4.6C; Table 4.1). Finally, trpm7 mutants exhibited skeletal dysplasia, including extreme malformation of the Weberian
(auditory) apparatus and the ribs, compressed vertebrae (of normal number), and kinks in the posterior vertebral column (Figures 4.5D, 4.5H, 4.5S, and 4.5W).
We have demonstrated that trpm7 is essential for growth and
skeletogenesis during the zebrafish larval-to-adult transition, and for
melanophore development and touch response during embryogenesis. trpm7
expression in kidney and corpuscles of Stannius, as well as the presence of
kidney stones in trpm7 mutants, support a model in which effects on growth and
skeletogenesis reflect physiological regulation of cation homeostasis. These
effects may be analogous to parathyroid hormone/parathyroid hormone receptor
type 1 (PTHR1) regulation of calcium homeostasis in mammals, in which
changes in PTHR1 signaling or downstream effectors such as Runx2 can lead to
premature endochondral ossification (Karaplis et al., 1998; Schipani et al., 1995;
Takeda et al., 2001; Ueta et al., 2001). However, this model does not readily
explain the delayed Intramembranous ossification observed in trpm7 mutants.
Although the precise functions of trpm7 in promoting normal ossification
sequence and timing, as well as melanophore development and touch response,
remain to be elucidated, our analyses reveal important roles for trpm7 in physiological regulation of postembryonic growth and skeletogenesis.
86
MATERIALS AND METHODS
Fish Stocks and Maintenance
Mutations were induced in wild-type AB, SJD, and wikut genetic
backgrounds by standard methods of ethyl-N-nitrosourea mutagenesis. Fish were reared at 28.5°C and fed rotifers, paramecia, or dry flake food as
appropriate. For growth series of wild-type and nutria mutants, fish were reared
individually from embryonic through adult stages and fed ad libitum to eliminate
competition. For staining series, wild-type and mutant individuals were sorted at
3 dpf then reared separately at equivalent densities to minimize variation in
growth rates within each genotype.
Skeletal Staining and Classification
Fish were fixed in 4% paraformaldehyde in phosphate-buffered saline
(PBS) for 48 hours and dehydrated over two days to 100% ethanol. They were
then stained with Alcian blue for cartilage, digested with trypsin, and stained with
Alizarin red for bone following the standard protocol for fish larvae (Potthoff,
1984). Fish were transferred to 100% glycerol over 1 week for analysis and
storage. For calcein staining, we fixed larvae with 4% paraformaldehyde in PBS,
embedded them in OCT and collected cryosections on Fisherbrand Superfrost
Plus slides. After they were dried and rinsed in PBS, slides were immersed in
0.1% calcein in PBS for 10 min, rinsed extensively in PBS and coverslipped. The
classification of bones by method of ossification and anatomical-functional units
87 followed that used by references (Bird and Mabee, 2003; Cubbage, 1996;
Fleming et al., 2004).
Genetic Mapping and Sequencing
Meiotic mapping was performed by standard methods using haploid and
diploid mapping panels constructed with AB and wik mapping strains (panel
sizes: trpm7j124e2, 1400 haploid embryos; trpm7b722, 1000 diploid embryos;
trpm7b508, 435 haploid embryos; trpm7b508, trpm7os1, and trpm7os2, 54 diploid
embryos). For identification of mutant lesions, polymerase chain reaction (PCR)
products of trpm7 cDNAs were sequenced directly and compared with
sequences derived from the corresponding, unmutagenized genetic background.
Morpholino-Based Gene Knockdown
For morpholino injection experiments, eggs were obtained from wild-type
fish of the ZDR strain (Scientific Hatcheries, Huntington Beach, California) by natural matings. Morpholino oligonucleotides (GeneTools, Sumerton, Oregon) were diluted to 20 mg/ml in Danieau buffer, then to 0.8 mg/ml in 0.2 M KCl for injection. We designed a morpholino (GTG TGT GAG ATT TAC TCT GCT GTT
C) to interfere with splicing (Draper et al., 2001) at the junction between exon 12 and intron 12. Embryos were injected at the 2-4 cell stage with approximately 5 nl of 0.8 mg/ml trpm7 MO or of the standard negative control morpholino
(GeneTools). Sequencing of morpholino-injected embryos confirmed failed splicing of intron 12, and this failure resulted in a frame shift and premature stop
88 codon (not shown). With all morpholinos, we observe approximately 5% mortality before 12 hpf, presumably from injection wounds. At all stages, embryos injected with standard negative-control morpholino appeared identical to uninjected embryos (not shown). A phenotype of highly reduced melanophore differentiation throughout embryo (shown in Figure 4.2) was observed in 80% of injected embryos (n=60), with other embryos showing a less pronounced phenotype.
Limited morpholino perdurance precluded testing for effects of morpholino knockdown on late-developing kidney and ossification phenotypes.
In Situ Hybridization
Whole-mount in situ hybridization, with minor modifications, followed
(Quigley et al., 2004). In brief, larvae were fixed over two nights in 4% paraformaldehyde and 1% DMSO in PBS, transferred to methanol, and then rehydrated to PBST (PBS with 0.2% Tween-20). Larvae were treated for 30 min with 20 μg/ml proteinase-K in PBST containing 1% DMSO, postfixed for 20 min in
4% paraformaldehyde in PBST, washed in PBST, and then washed three times in a hybridization solution lacking transfer RNA (tRNA) and heparin.
Prehybridizations were performed overnight at 68°C in a hybridization solution
(50% formamide, 5X SSC, 500 mg/ml yeast tRNA, 50 mg/ml heparin, 0.2%
Tween-20, and 9.2 mM citric acid). Hybridizations were performed over two nights at 68°C in a fresh hybridization solution containing antisense digoxygenin- labeled riboprobes fractionated to ~300 nucleotides. Probes derived from 5' and
3' ends of trpm7 transcripts yielded qualitatively similar results. Larvae were then
89 washed twice in 2X SSCT, and three times 2 hr each in 0.2X SSCT at 68°C.
After graded changes were made to maleic acid buffer (MAB; 100 mM maleic
acid [pH 7.5] and 150 mM NaCl), larvae were blocked overnight with Roche
blocking reagent in MAB at 4°C, incubated over three nights in fresh blocking
reagent containing 1:5000 anti-digoxygenin alkaline phosphatase-conjugated
Fab fragments (Roche), and then washed over three additional nights in MAB.
Larvae were then transferred to alkaline phosphate buffer (100 mM Tris [pH 9.5],
50 mM MgCl2, 100 mM NaCl, and 0.1% Tween-20), and the color was developed with NBT/BCIP.
Cation Rescue Experiments
tctb508 mutant embryos and wild-type siblings were obtained from matings
of heterozygous carriers. Chorionated embryos were reared to the 26-somite stage in water and then transferred to embryo medium (Westerfield, 2000)
(17.9mM NaCl; “0” in Figure 4.3) or modified embryo medium (4.2mM NaCl; “-” in
Figure 4.3) with or without the addition of MgCl2 (100mM) or CaCl2 (50mM or
100mM). We tested touch response by prodding embryos still in their chorions
one time with a probe at the midtrunk level; any response to this stimulus was
considered a rescue. Melanophore morphologies were examined over the head
and trunk; the presence of any well-spread melanophores of normal size was considered a rescue. Both melanophore and touch response phenotypes were observed between 48 and 60 hpf. Treatments were replicated across multiple families two to seven times with seven to 35 embryos per condition.
90
Quantitative Analyses of Ossification Sequence and Timing
We reconstructed ossification sequences in wild-type fish and trpm7 mutants by determining the frequency with which each skeletal element was ossified within all individuals of each genotype and then ordering these frequencies into ranks (Mabee et al., 2000). Elements that failed to ossify within the range of ages examined were ordered arbitrarily (ranks 83–87 in Figure
4.6A). Differences in rank between wild-type fish and mutants were highly significant between endochondral and intramembranous bones (Wilcoxon test:
Z=6.03, d.f.=1, p < 0.0001).
To assess likelihoods of bone ossification, we performed separate logistic regression analyses for each bone, developmental mode (endochondral and intramembranous), or functional–anatomical unit with genotype as a main effect and dpf as a covariate. Alternative models, including only genotype or including genotype and other main effects to control for stage and batch variability (e.g., standard length of wild-type siblings), yielded qualitatively similar results. Note that estimates of timing differences between genotypes reflect only the onset of ossification as evidenced by the first indication of Alizarin red staining; they do not account for differences in rate once ossification has started. Overall, trpm7 mutants exhibited greatly accelerated ossification rates for endochondral bones and correspondingly delayed ossification rates for intramembranous bones even after ossification had commenced. Analyses shown here thus underestimate the magnitude of differences between genotypes. Because of the large numbers of
91 individual tests for logistic regression analyses, we assessed overall significance levels by means of a standard Bonferonni correction (α = 0.05, adjusted, critical significance level = 0.000127), which is likely to be overly conservative, as well as a more recent method of assessing false discovery rate (q value) (Storey and
Tibshirani, 2003). We provide both significance thresholds in Figure 4.6B. For the smaller number of tests for developmental mode and functional-anatomical units, we provide only Bonferonni-corrected significance threshold (Figure 4.6C). See
Table 4.1 for a complete listing of effects and significance levels by bone.
Statistical analyses were performed with JMP 5.0.1a for Apple Macintosh computers (SAS Institute, Cary NC) and software of (Storey and Tibshirani,
2003).
92 TABLES AND FIGURES
Table 4.1. Relative ossification timing. (continued) 93 Table 4.1 (continued)
Table 4.1. Relative ossification timing. aFunctional-anatomical grouping. bPartial logistic regression coefficient: relative likelihood of bone being ossified in wild-type as compared to mutant (negative values reflect precocious ossification in mutant; positive values reflect delayed ossification in mutant). cProbability that effect differs significantly from 0 (χ2, 1 d.f.). dSignificance of effect after controlling for multiple comparisons (q, significance level controlled by false-discovery-rate method; a, significance level controlled by Bonferonni method; see text). NS = not significant. *: q < 0.01, a > 0.05. **: q < 0.01, a < 0.05. ***: q < 0.01, a < 0.01. eThe following abbreviations are used: W, Weberian; c, caudal; pre-c, pre-caudal.
94
Figure 4.1. Retarded Growth and Altered Body Proportions in nutria (tct) Mutant Zebrafish with Embryonic Melanophore Defect Rescuable by Divalent-Cation Supplementation. (A) Wild-type and nutria siblings, 50 dpf. (B) Higher- magnification image of a nutria mutant. (C) Diminished growth of nutria compared to wild-type siblings. Each point shows the mean standard length ± standard deviation for seven to 20 individuals. (D–F) Embryonic melanophore defects are rescuable with supplemental Ca2+. (D) Wild-type embryos exhibit well-melanized and well-spread melanophores (arrowheads). (E) Mutants exhibit few poorly melanized and punctate melanophores or melanophore fragments (arrows); tctb508 is shown. (F) In a medium supplemented with Ca2+, mutant melanophores are more numerous, spread, and melanized (arrowheads), although some punctate melanophores remain (arrow). Scale bars represent the following: (A), 2 mm; (B), 1 mm; (D–F), 100 μm.
95
Figure 4.2. Identification of trpm7 as the gene mutated in tct (nutria) mutant zebrafish. (A-C) Morpholino knockdown of trpm7 phenocopies the tct embryonic melanophore defect. (A) In wild-type embryos, well-melanized and well-spread melanophores (arrowheads) cover the dorsum and head at 48 hr. (B) In tct embryos, melanophores are fewer, lighter, and punctate (arrows) (Arduini and Henion, 2004; Cornell et al., 2004). An allele of moderate severity, tctb722, is shown. (C) Wild-type embryos injected with a splice-blocking morpholino to trpm7 exhibit fewer, lighter, and punctate melanophores, as in tct. (D) Schematic of trpm7 cDNAs. Sequencing of trmp7 cDNAs identifies a transcript encoding at least 1773 amino acids (aa), including an ion-transport domain (aa 873-1073) and an α kinase domain (aa 1510-1724). tctj124e1 shows a CÆA transversion resulting in a substitution of a premature stop codon for tyrosine at aa 1545, near the start of the α kinase domain. Sequencing of tctb508 reveals a 68 nucleotide deletion (aa 1410-1432) comprising a single exon and resulting in a frame shift, 16 novel amino acids, and a premature stop codon. Analyses of genomic DNA reveal a deletion corresponding to one exon in addition to parts of flanking introns. (E) Sequence electropherograms from haploid embryos that are wild- type (upper section) or the strong allele tctj124e1 that exhibits a premature stop codon (TAA, lower section). trpm7 Genbank accession number: AY860421. Scale bars in (A)-(C) represent 100 μm.
96
Figure 4.3. Rescue of trpm7 embryonic defects with divalent-cation supplementation. Mg2+ and Ca2+ supplementation differentially rescued melanophore and touch-response phenotypes of tctb508 embryos. High concentrations of CaCl2 were toxic at normal NaCl concentration (“0” in the figure), although embryo viability could be restored at lower NaCl concentrations (“– “ in the figure). (A) Mean + standard error of the mean (SEM) proportion of trpm7b508 embryos exhibiting rescued melanophores (as in Figure 4.1F in the main text). In the standard embryo medium (solution A) and the reduced-NaCl rearing medium (B), no embryos exhibited rescued melanophores. In the medium with supplementary MgCl2 (100 mM; solutions C and D), many embryos exhibited partial melanophore rescue. Although melanophore rescue was not observed in the medium containing low supplementary CaCl2 (50 mM) without MgCl2 (solution E), partial rescue occurred in media containing high supplementary CaCl2 (100 mM) without MgCl2 supplementation (solution F) or low CaCl2 (50 mM) with high MgCl2 (100 mM; solution G). (B) Mean + SEM proportion of trpm7b508 embryos exhibiting touch response. In contrast to melanophore rescue, touch response was not rescued by MgCl2 supplementation alone (solutions C and D). Rather, touch response was maximally rescued with low supplementary CaCl2 (50 mM) without MgCl2 (solution E), although some rescue also occurred with higher CaCl2 (100 mM) supplementation (solution F) or with low CaCl2 (50 mM) in the presence of MgCl2 (100 mM; solution G). Qualitatively similar responses to CaCl2 and MgCl2 supplementation also were observed with trpm7j124e2 (nutria), which exhibit a weaker embryonic melanophore phenotype than trpm7b508. The same media did not rescue similar melanophore defects of sparse (kit) or colorless (sox10) mutants (Dutton et al., 2001b; Parichy et al., 1999), demonstrating specificity of the effect for trpm7 (data not shown). The cellular basis of the touch phenotype is unclear. Because of the timing and recovery of this phenotype, it is unlikely to depend on hair cells of later-line neuromasts, recently shown to require TRP family members NompC and TRPA1 (Corey et al., 2004; Sidi et al., 2003). Toxicity of high-salt solutions precluded testing for rescue of later, larval ossification phenotypes.
97
Figure 4.3
98
Figure 4.4. trpm7 expression in wild-type larvae and kidney stone formation in trpm7 mutants. (A-C) trpm7 expression in metamorphosing wild-type larvae. (A) At early stages of metamorphosis (14 dpf), trpm7 mRNA is present in mesonephric tubules (mt) and corpuscles of Stannius (cs). The following abbreviations are used: v, vertebral column; sb, swim-bladder; and ct, mesonephric collecting tubule. Trpm7 expression is also detectable in more anterior regions of the mesonephros (not shown). (B) A higher magnification trpm7+ mesonephric tubule is shown. (C) Transverse section of late metamorphic (24 dpf) larva. Trpm7 transcripts are present in mesonephric tubules and liver (indicated by “li”). Trpm7 transcripts are not detectable in developing endochondral or intramembranous bones. “my” denotes myotome and “c” denotes centrum. (D-F) trpm7j124e2 (nutria) mutants develop kidney stones. (D) Calcein staining reveals skeletal elements (na, c, ps) and ectopic mineralization (arrows). The following abbreviations are used: na, neural arches; c, centrum; ps, pleural spins; and nt, neural tube. (E) shows left-most ectopic mineralization deposit from panel (D). Epithelium (arrowheads) surrounds the mineralized deposit (arrow). The inset shows mineralized deposit within collecting tubule of another larva. (F) shows calcein staining of mineralized deposits shown in panel (E) and the inset. Scale bars represent the following: (A), 60 μm; (B), 40 μm; (C) and (D), 80 μm; (E) and (F), 20 μm (15 μm for insets).
99
Figure 4.5. trpm7j124e2 mutants exhibit dramatic differences in skeletal development in comparison to wild-type. Alizarin red stains ossified bone and mineralized tissues. Alcian blue stains cartilage. Endochondral bone labels are red, and intramembranous bone labels are green. (A-H) Differences in ossification timing in the caudal complex between wild-type (A-D) and mutant (E- H) siblings (9-44 dpf, left to right). In mutants, endochondral bones ossify earlier [panels (B) and (F): na, u, ha, c, l]. Compression of vertebrae and the caudal complex are evident [panels (D) and (H): brackets]. (I-P) The suspensorium (I-N) and branchial arches (K-P) illustrate precocious endochondral ossification I mutants. Advances in ossification are apparent for several endochondral bones at early stages [17 dpf. Panels (I) and (M): hm. Panels (K) and (O): ch, eh, cb1] and later stages [47 dpf. Panels (J) and (N): mpt, q, sym. Panels (L) and (P): bh]. (Q, U, R, V) Bones of the anterior head illustrate delayed intramembranous ossification in mutants. Intramembranous bones are delayed in mutants at early stages [24 dpf. Panels (Q) and (U): pm, m, ps, ect] and later stages [44 dpf. Panels (R) and (V): f, en]. (S, W) Bones of the Weberian apparatus (brackets) are malformed and compressed in mutants. (T, X) Mineralizations within the mesonephros of mutant larvae [arrowheads in panel (X)]. (S) and (T) are at 35 hpf; (W) and (X) are at 30 dpf. The following abbreviations are used: bh, basihyal; c, centra; cb1, ceratobranchial 1; cb5, ceratobranchial 5; ch, ceratohyal; ect, ectopterygoid; eh, epihyal; en, entopterygoid; ep, epural; f, frontal; ha, hemal arches; hm, hyomandibula; hyp, hypurals; l, lepidotrichia; m, maxilla; mpt, metapterygoid; na, neural arches; pm, premaxilla; ps, parasphenoid; q, quadrate; sym, symplectic; and u, urostyle and ural 1 + 2. Scale bars represent the following: (A-H), 100 μm; (I-X), 250 μm.
100
Figure 4.5
101
Figure 4.6. Altered sequence and timing of skeletal ossification in trpm7j124e2 compared to wild-type. Wild-type larvae (n = 500) and trpm7 mutant siblings (n = 431) were scored for ossification of 87 bones, yielding 80,997 ossification scores. For schematics of bone locations, functional-anatomical units, and effects for each bone, see Figure 4.7 and Table 4.1. (A) In mutants, endochondral bones (red connectors) appear on average earlier in the ossification sequence, whereas intramembranous bones (green connectors) appear on average later in the ossification sequence. Bold connectors show bones in which ossification timing, in addition to sequence, was significantly altered [panel (B)]. (B) In mutants, analyses by developmental mode reveal accelerated ossification for many endochondral bones (red) but delayed ossification for many intramembranous bones (green). The x axis shows the relative likelihood of ossification in the wild- type compared to the mutant: estimates greater than 0 indicate that the bone is more likely to be ossified in wild-type, and ossification is therefore delayed in the mutant; estimates less than 0 indicate that bone is less likely to be ossified in wild-type fish, and ossification is therefore accelerated in the mutant. The y axis shows statistical significance by chi-square value (left axis) and p value (right axis). α = 0.05 and q = 0.01 are the thresholds of statistical significance after multiple comparisons were controlled for. (C) Analyses by functional-anatomical unit and developmental mode. Functional-anatomical units were differentially affected in the mutants even after their relative endochondral and intramembranous compositions were controlled for. Averages: of all bones, diamond; of endochondral bones, large red point; of intramembranous bones, large green point. Error bars in (B) and (C) show one standard error for logistic regression coefficients, β.
102
Figure 4.6
103
Figure 4.7. Functional-anatomical units and bone location examined for ossification sequence and timing. Functional-anatomical units are color coded to match Figures 4.6A-C. (A) shows units of the postcranial skeleton with bones comprising these units listed in small type. (B) shows units and bones of the craniofacial skeleton.
104
Figure 4.8. Growth retardation alone does not result in precocious ossification. Three siblings from a cross segregating trpm7j124e2. (A) Wild-type fish showing endochondral hypurals (denoted by “hyp”) that re not yet ossified, as well as intramembranous vertebral centra (denoted by “c”) and fin lepidotrichia (denoted by “l”) that are already well ossified. (B) Even severely runted wild-type individuals that are from the same family and are indistinguishable in size form trpm7 mutants do not show precocious ossification of endochondral hypurals (or other bones); the ossification of intramembranous centra and lepidotrichia is delayed, reflecting an overall retardation of developmental rate in slow-growing fish. (C) trpm7 mutant has accelerated hypurals ossification and delayed centra and lepidotrichial ossification compared to wild-type. In contrast to runted wild- type individuals, delayed ossification of intramembranous bones does not reflect a general developmental retardation because the absolute timing of other events (cartilage formation, chondrocyte and osteoblast appearance, adult pigment pattern formation, gut looping, swim-bladder bifurcation, etc.) does not differ dramatically form wild-type. Scale bars represent 400 μm. 105
CHAPTER 5
ZEBRAFISH ENDZONE REGULATES NEURAL CREST-DERIVED CHROMATOPHORE MORPHOLOGY4
ABSTRACT
The development of neural crest-derived pigment cells has been studied
extensively as a model for cellular differentiation, disease and environmental
adaptation. Neural crest-derived pigment cells in the zebrafish (Danio rerio)
consist of three types: melanophores, xanthophores and iridiphores. We have
identified the zebrafish mutant endzone (enz), which was isolated in a screen for neural crest mutants based on abnormal melanophore pattern. We have found that although wild-type numbers of chromatophore precursors are generated in the first day of development and migrate normally in enz mutants, the numbers of all three chromatophore cell types is reduced and differentiated melanophores and xanthophores subsequently lose dendricity and iridiphores are reduced in size. We demonstrate that enz function is required cell autonomously by melanophores and that the enz locus is located on chromosome 7. Zebrafish enz appears to selectively regulate chromatophore development within the neural crest lineage since all other derivatives develop normally. Our results suggest
4 Manuscript in preparation. All data reported herein, unless otherwise cited, were generated by B.L. Arduini under the guidance of P.D. Henion. 106 that enz is required relatively late in the development of all three embryonic chromatophore types and is normally necessary for terminal differentiation and the maintenance of cell size and morphology. Thus, although the regulation of the development of different chromatophores in zebrafish are in part genetically distinct, enz provides an example of a common regulator of neural crest-derived chromatophore differentiation and morphology.
INTRODUCTION
The neural crest is a transient vertebrate embryonic cell population that gives rise to a wide variety of cell types, including pigment cells, craniofacial cartilage, and neurons and glia of the peripheral nervous system (Le Douarin and
Kalcheim, 1999). This array of neural crest-derived cell types has long been of interest in studying the mechanisms of cell diversification among embryonic cell populations. The development of neural crest-derived pigment cells in particular has been studied extensively, and many important insights have resulted from the analysis of mouse and zebrafish mutants (Bennett and Lamoreux, 2003; Le
Douarin and Kalcheim, 1999; Lister, 2002; Quigley and Parichy, 2002; Silvers,
1979).
Vertebrate chromatophore populations are readily observed, as they carry their own intrinsic markers. In addition, pigment cells are not strictly required for viability (Lister et al., 1999; Parichy et al., 1999; Silvers, 1979). As a result, these pigmented cell types have long been used for studying developmental processes such as cell fate specification, proliferation, migration, differentiation, and
107 survival. Mice and other mammals have a single chromatophore cell type termed
melanocytes (Nordlund et al., 1998). Hundreds of mouse coat color mutants
have been identified, covering over 100 loci, which affect multiple cellular
processes (Bennett and Lamoreux, 2003; Silvers, 1979). Further, many of these
mutations in mice have proved to be medically relevant as models for human
diseases involving the same genes (Jackson, 1997). Besides the melanocytes
(melanophores) also found in mammals, zebrafish and other ectotherms possess
neural crest-derived yellow xanthophores and iridescent iridiphores (Bagnara,
1998; Raible et al., 1992). In addition to the isolation of several zebrafish pigment
mutants that arose spontaneously (Johnson et al., 1995; Streisinger et al., 1986),
numerous mutagenesis screens have yielded over 100 mutations affecting
various processes in the development of different combinations of the three
pigment cell types (Henion et al., 1996; Kelsh et al., 1996; Lister et al., 1999;
Odenthal et al., 1996; Rawls et al., 2003).
Studies from several vertebrates, including zebrafish, have led to the extensive characterization of melanophore development, and to a lesser extent,
xanthophore and iridiphore development (Bennett and Lamoreux, 2003; Dupin
and Le Douarin, 2003; Lister, 2002; Nakamura et al., 2002; Nordlund et al.,
1998). Prior to overt differentiation, chromatophore precursors are referred to as
chromatoblasts, and can be identified by expression of genes specific to one or
multiple chromatophore sublineages. Sox10, mutations in which cause
Waardenburg-Hirschsprung Syndrome in humans, is required for development of nonectomesenchymal neural crest derivatives, including all pigment cells, as well
108 as many peripheral neurons and glia (Kelsh and Eisen, 2000; Pingault et al.,
1998). Sox10 has been shown to directly regulate expression of microphthalmia-
associated transcription factor (mitf), which is both necessary and sufficient for
melanophore development, and dopachrome tautomerase (dct), an enzyme in the melanin synthesis pathway (Kelsh et al., 2000b; Lister et al., 1999; Steel et al., 1992; Widlund and Fisher, 2003). The receptor tyrosine kinase c-kit is also expressed by melanoblasts, and appears to be necessary for both differentiation and survival of this lineage (Rawls et al., 2001; Wehrle-Haller, 2003). Similarly, the kit ortholog fms, which has no known role in mammalian pigmentation, is required for migration of embryonic xanthophores and specification of a subset of adult melanophores in zebrafish (Parichy et al., 2000b; Parichy and Turner,
2003). fms is expressed by embryonic xanthoblasts and macrophages, which can be distinguished from one another based on location and cellular morphology (Herbomel et al., 1999; Herbomel et al., 2001; Parichy et al., 2000b).
Synthesis of yellow pteridine pigments, found in xanthophores, requires xanthine
dehydrogenase, which is correspondingly expressed by xanthoblasts (Epperlein
and Lofberg, 1990; Parichy et al., 2000b; Reaume et al., 1991). Both fms and
xdh are co-expressed in a subset mitf+ cells in the premigratory neural crest,
which may represent uncommitted precursors of melanophores or xanthophores
(Parichy et al., 2000b). Neither of these genes is co-expressed with c-kit, however, and both have been used as specifically diagnostic of xanthoblasts at migratory and post-migratory stages of pigment cell development (Knight et al.,
2003; Parichy et al., 2000b). The enzyme GTP-cyclohydrolase (Gch) is involved
109 in the conversion of intermediates of both melanin and pteridine synthesis
(Nagatsu and Ichinose, 1999; O'Donnell et al., 1989; Wood et al., 1995).
Accordingly, gch expression is observed in both melanoblasts and xanthoblasts
(Parichy et al., 2000b). A G protein-coupled receptor, endothelin receptor (ednr)
b, is also expressed by neural crest-derived pigment cell precursors (Kumagai et
al., 1998; Lee et al., 2003; Nataf et al., 1996; Parichy et al., 2000a; Shin et al.,
1999) Homozygous ednrb mutant mice are almost completely devoid of
melanocytes (Bennett and Lamoreux, 2003; Silvers, 1979). In contrast, zebrafish
ednrb1/rose mutants display defects in a subset of adult melanophores and
iridiphores (Parichy et al., 2000a). In the zebrafish embryo, ednrb1 is initially
expressed by all chromatophore sublineages, but by late embryonic/early larval
stages, is restricted to iridiblasts and iridiphores (Parichy et al., 2000a).
Morphologically, differentiated melanophores and xanthophores are large
and dendritic, possessing many processes, while iridiphores are rounded in
shape (Nordlund et al., 1998). In ectotherms, considerable attention has been
given to mechanisms of color adaptation, reversible changes in pigmentation
brought on by prolonged exposure to either light or dark environments
(Sugimoto, 2002). Extensive analyses, especially in a variety of fish species, have revealed that this occurs through relocalization of pigment organelles within
cells, changes in cell morphology, and proliferation and apoptosis of chromatophores (Bagnara, 1973; Hogben, 1922; Lister, 1858; Sugimoto, 2002).
In adults, these processes appear to be under hormonal, as well as nervous
control (Bagnara, 1973; Hogben, 1922; Lister, 1858). α-melanophore-stimulating
110 hormone (α-MSH) and melanin-concentrating hormone (MCH) appear to have
mutually antagonizing effects on melanophores, with α-MSH enhancing melanin
and melanophore development, and MCH promoting aggregation of
melanosomes and downregulating secretion of α-MSH (Baker et al., 1986;
Barber et al., 1987; Green et al., 1991; Kawauchi and Baker, 2004). The effects
of α-MSH are enacted in part through influence on mitf expression in mammals
(Busca and Ballotti, 2000; Halaban, 2000; Huber et al., 2003). Rho GTPase-
mediated cytoskeletal rearrangements may play roles in redistribution of pigment
organelles and dendrite collapse (Sugimoto, 2002).
Despite these and other data that have been amassed regarding pigment
cell development in a variety of vertebrate systems, many questions remain. For
example, relatively few genes are known that are required for development of all embryonic chromatophores, and yet are specific to pigment cell sublineages within the neural crest. Sox10 is necessary for development of pigment sublineages, but is also required for some crest-derived neurons and glia, and mammalian ednrb is necessary for both melanocyte and enteric neuron development (Dutton et al., 2001b; Herbarth et al., 1998; Honore et al., 2003;
Lahav, 2005; Lee et al., 2003; Southard-Smith et al., 1998). Further, although
expressed by all three zebrafish chromatophore cell types, ednrb1 appears to be
dispensable for their embryonic development (Parichy et al., 2000a). While
hormonal and neuronal influences have been demonstrated for pigment
distribution, cell morphology, proliferation and survival of differentiated
chromatophores, less is known about the downstream effectors governing these 111 processes. Additionally, these studies focus on adults, and comparatively little is known about control of pigment cell morphology or color adaptation at embryonic stages.
We report here characterization of the zebrafish mutant endzone (enz),
which was isolated based on reduced embryonic pigmentation. Our results
indicate that all three chromatophore cell types are similarly reduced in size and
number in enz mutant embryos. We show that enz is required specifically for
chromatophore sublineages within the neural crest, and that this requirement is
cell autonomous. Further, we report the identification of multiple enz alleles and
progress toward the molecular identification of the enz locus. We suggest that
enz is a relatively late cue required by pigment cell sublineages of the neural crest during embryogenesis, and is indicative of common requirements of chromatophores that are distinct from other neural crest derivatives.
RESULTS
Live phenotype
endzoneb431 (enzb431) was identified in a chemical (ethyl-nitrosourea, ENU)
mutagenesis screen for mutations affecting neural crest derivatives as previously
described (Henion 1996). The mutant was identified based on its altered pigment
pattern. At 48 hours post-fertilization (hpf), the morphology of melanophores is
dramatically smaller and less stellate in enzb431 mutant embryos compared to
112 wild-type siblings (compare insets, Fig. 5.1A and B). In addition, the pigmented
retinal epithelium (PRE) of enz homozygotes, which is not neural crest-derived, is
pale compared to that of wild-type siblings. Unlike crest-derived melanophores,
however, the PRE phenotype recovers by the fourth day post-fertilization (Figure
5.5 and data not shown) At 48 hpf, enzb431 embryos also lack the characteristic
yellow cast caused by xanthophores concentrated dorsally in wild-type embryos
(Figure 5.1). In addition, iridiphores are reduced in number and size in 6 days-
postfertilization (dpf) enz mutant larvae compared to wild-type siblings (Figure
5.1). Recently, three additional ENU-induced mutations have been identified as
enz alleles by complementation and linkage analysis (Figure 5.1 and data not
shown). All four enz alleles are recessive. These alleles vary in the severity of the
developmental defects in all three chromatophore phenotypes. enzos15 is less
acute than enzb431, enzos7, and enzos18, which are similar in expressivity. All data
refer to enzb431 unless otherwise specified. Most enz homozygotes do not develop swim bladders, and subsequently fail to survive past early larval stages
(Figure 5.2). enz mutant larvae that do develop swim bladders survive, but are runted compared to wild-type siblings through adulthood. Qualitative differences in melanophore pigmentation persist through at least 30 dpf (Figure 5.2). As adults, enz homozygotes are fertile and produce both males and females, (not shown), although they remain smaller, on average, than wild-type siblings.
113 Within the neural crest lineage, enz selectively affects chromatophore development.
Due to the defects in neural crest-derived chromatophores in enz mutant
embryos, we investigated the development of other cell types generated from the
neural crest (Table 5.1). Molecular markers indicated that neural crest-derived
neuronal and glial populations develop normally in enz homozygotes. For example, cervical sympathetic neurons, enteric neurons, and neurons and glia of the dorsal root ganglia are all present in qualitatively normal numbers and positions in enz mutant embryos (Figure 5.3 and data not shown). Craniofacial cartilage, stained with Alcian Blue, was also found to be normal in terms of individual elements and their shapes (Figure 5.3). The sizes of all jaw elements are proportionately smaller in enz mutant embryos, in accordance with the overall reduced size of these embryos compared to wild-type siblings (see Figure 5.2).
In addition, the pan-neural crest marker crestin was used to analyze neural crest
populations at different embryonic stages. In wild-type embryos, crestin is first
expressed in neural crest cells at the boundary of neural and non-neural
ectoderm during gastrulation. Expression continues in premigratory and
migratory neural crest cells, and persists until slightly after overt differentiation of
neural crest derivatives, such that by 24 hpf, crestin-expressing cells are found throughout the embryo (Luo et al., 2001a). crestin expression is normal in enz
embryos at all stages investigated, suggesting that the early neural crest is
unaffected by the enz mutation (Figure 5.4 and data not shown). Further, we
examined the expression of genes that, within the neural crest, are specifically
114 expressed by melanophores, (Kelsh et al., 2000b), xanthophores (xdh, Parichy
1999) and all three chromatophore cell types (Parichy et al., 2000a) at 24 hpf.
Expression of each gene is qualitatively normal at this stage, suggesting that
effects on crest-derived chromatophores occur relatively late in the development
of these cell populations (Figure 5.4). These data indicate that within the
embryonic neural crest lineage, enz is required specifically for chromatophore development and only after neural crest dispersal and initial differentiation have largely occurred.
Chromatophore cell morphology is altered in enz mutant embryos
In wild-type zebrafish embryos raised at 28.5oC, neural crest-derived
melanophores normally begin to differentiate at approximately 25 hpf, while
xanthophores and iridiphores begin to overtly differentiate at about 42 hpf and 72
hpf, respectively (Kimmel 1995). By 27 hpf, large, stellate, dark melanophores
are present (anteriorly) in wild-type embryos. In enz mutant embryos, as in wild-
type siblings, melanophores differentiate at ~25 hpf (data not shown). At 27 hpf,
enz melanophores are stellate, but pale compared to those of wild-type siblings
(Figure 5.5). After 27 hpf, enz melanophores begin to transition from the initial
pale, stellate appearance to a dark, punctate form, while wild-type melanophores
remain stellate and dark. The transformation of melanophores occurs in a rostro-
caudal wave, and is complete by about 48 hpf. We hypothesized that this
apparent change in cell morphology of enz melanophores might be the result of
either redistribution of melanosomes within a cell or of an actual cell shape
115 change. To distinguish between these two possibilities, we performed in situ
hybridization of melanized 36 hpf and 48 hpf wild-type and enz mutant embryos, using the melanophore sublineage-specific riboprobes c-kit (Parichy et al., 1999) and dct (Kelsh et al., 2000b). In wild-type embryos, c-kit and dct mRNAs are distributed throughout the cell cytoplasm, including in the processes, reflecting stellate cellular morphology (Figure 5.5 and data not shown). The distribution of dct and c-kit mRNAs is punctate in enz melanophores, similar to the distribution of melanin (Figure 5.5 and data not shown). This is consistent with a cell morphology change, rather than relocalization of melanin within a stellate cell.
Subsequently, we quantified the area of punctate melanophores in enz mutant embryos compared to stellate melanophores in wild-type siblings. At 2 dpf, wild- type melanophores at cephalic axial levels have an average area of 282.1 μm2, while enz melanophores at similar axial levels have a mean area of 21.9 μm2 (P
< 0.0001, see Methods). The change in melanophore cell morphology in enz
mutant embryos further suggested that the apparent absence of yellow
pigmentation might be due at least in part to a reduction in size of xanthophores in these embryos. Individual xanthophores are difficult to distinguish, and even
fms and xdh expression appear as diffuse staining over the dorsal aspect of the
embryo, precluding the quantitative type of analysis performed on melanophores
(Bagnara, 1998; Kelsh et al., 1996; Parichy et al., 2000b). However, qualitative
observations of xanthophore morphology were made using methylene blue,
which is taken up specifically by xanthophores and is concentrated around active
pterinosomes, the organelles that produce pteridine pigments (Le Guyader and 116 Jesuthasan, 2002). Methylene blue staining revealed that while some
xanthophores are present in enz homozygotes, these are much smaller and less stellate than xanthophores in wild-type siblings at 3 dpf (Figure 5.6C, D).
Similarly, iridiphores are reduced in size in enz mutant embryos compared to wild-type siblings. In contrast to melanophores and xanthophores, iridiphores found on the trunks of wild-type embryos have a rounded, rather than stellate morphology at 72 hpf (see Figure 5.1). While this is also true in enz homozygotes, overall iridiphore cell size, as measured by area, is reduced compared to wild-type siblings. At 6 dpf, the average iridiphore area in enz
mutant larvae is reduced by ~40% compared to that in wild-type siblings (P <
0.0001). Together, these data indicate that mutations in enz similarly affect all
three neural crest-derived pigment cell types with respect to cellular morphology
and size.
Chromatophore cell numbers are reduced
To determine whether appropriate numbers of chromatophores are
generated in enz mutant embryos, we counted melanized and iridescent cells in
enz homozygotes and wild-type siblings. At 4 dpf, the number of melanophores
in enz mutant embryos is significantly reduced compared to wild-type siblings
(P<0.0001, Table 5.2). Likewise, iridiphore numbers are reduced in enz
homozygotes compared to wild-type siblings at 6 dpf (P<0.0001, Table 5.2). In
both cases, only about 75% of the wild-type complement of each class of
chromatophores is present in enz mutants at the stages examined. Quantification
117 of xanthophore numbers was again precluded by indistinct boundaries between these cells. However, methylene blue staining for xanthophores at 3 dpf, as well as fms expression between 48 and 53 hpf, revealed that while many are present in enz homozygotes, the number of this chromatophore subpopulation is reduced in a manner qualitatively similar to that of melanophores and iridiphores (Figure
5.6 and data not shown). Thus, enz mutations appear to result in reductions in neural crest-derived chromatophores, while other cell types remain unaffected.
enz acts cell autonomously with respect to melanophore development
Mosaic analysis can be used to determine whether a mutation acts cell autonomously or non-autonomously, and thus predict the mode of action of a gene product within developmental pathways. Genetic mosaics were created between wild-type and enz mutant embryos (Table 5.3). Donor embryos were labeled with lysinated rhodamine dextran (LRD). Cells from donor embryos were then transplanted into unlabeled hosts, in which chromatophore development was subsequently observed. LRD-labeled cells from wild-type embryos formed dark, stellate melanophores at ~32 hpf when transplanted into enz hosts (n=4,
Figure 5.7). These data suggest that enz acts cell autonomously with respect to melanophore, and inferentially all chromatophore, development.
The enz locus is located on chromosome 7
As an initial step towards cloning the endzone gene, we placed allele enzb431 on the zebrafish genomic linkage map (Postlethwait 1994, 1998). Simple
118 sequence length polymorphisms (Knapik 1996, 1998) between AB* and WIK
zebrafish strains were used to map enz to chromosome 7. Parthogenetic diploid
embryos were generated by suppressing the second meiotic division using early
pressure (EP, Streisinger 1981). enz and wild-type embryos were identified by
live phenotype, and PCR was performed on DNA from enz and wild-type
embryos. The enz locus was initially localized to chromosome 7 in EP diploid
embryos using three markers, z10441, z11652 and z6819 (see Methods).
Subsequent recombination analysis established an interval in which enz lies
proximal to z8252 and z10441, and distal to z6819. Chromosomal location of the enz locus was confirmed using haploid embryos from enzos7, enzos15 and enzos18 heterozygous females in AB/WIK mapping lines (data not shown). DNA from wild-type and enz mutant embryos was amplified using SSLP marker z10441.
Each allele was confirmed to be linked to this marker. The enz locus appears to be within approximately 2 cM of z10441 (2 recombinants in 218 meioses).
Further, this region of zebrafish chromosome 7 possesses conserved synteny with human chromosome 11 and murine chromosomes 7 and 19 (Yoder and
Litman, 2000).
Compound trpm7/enz mutant supports a role for enz relatively late in chromatophore development
To investigate the role of enz compared to a known gene, we generated a line of fish doubly heterozygous for enzb431 and trpm7b508, a gene previously
shown to be required for development of the majority of neural crest-derived
119 melanophores (Arduini and Henion, 2004; Cornell et al., 2004; Elizondo et al.,
2005). trpm7b508 appears to affect melanophore development prior to overt
differentiation, as most of these cells do not appear in trpm7b508 homozygotes.
Further, there is evidence that some melanophore precursors undergo apoptosis
in trpm7b508 mutant embryos (Arduini and Henion, 2004). Those melanophores
that do develop in trpm7 single mutant embryos have punctate cell morphology
upon differentiation, in contrast to the initially stellate cell morphology of enz
melanophores. In clutches obtained from double heterozygote crosses, four
classes of phenotypes were observed. As expected, three of these classes were
wild-type, trpm7 single mutants and enz single mutants. The fourth phenotypic
class, approximately 1/16 of the total number of embryos, exhibited defects
resembling both trpm7 and enz mutant embryos, and is presumed to represent
double mutants. trpm7-enz double mutants exhibit an early reduction in
melanophore number that resembles the trpm7 phenotype. Iridiphores and
xanthophores, on the other hand, are depleted as in enz homozygotes. At
approximately 65 hpf, double mutants are touch insensitive, as are trpm7
mutants. Thus, double mutants show a combination of the trpm7 and enz
phenotypes. Depletion of melanophores early in double mutants as in trpm7
homozygotes supports the inference that endzone acts later than trpm7 during
melanophore development.
120 DISCUSSION
zebrafish enz is selectively required for the terminal differentiation of embryonic chromatophores among neural crest-derived cells
The zebrafish mutant enz was first identified based on the reduced size
and numbers of neural crest-derived melanophores, xanthophores and
iridiphores. We observed that melanophores begin to differentiate in enz mutant
embryos at approximately 25 hpf, similar to wild-type siblings. However, these
melanophores were pale compared to wild-type chromatophores, and later
transitioned to a darker, punctate morphology, while wild-type melanophores
remain highly dendritic. Likewise, iridiphores and xanthophores in enz mutants
also begin to differentiate normally but subsequently appear smaller and paler
than corresponding cell types in wild-type embryos. In addition, expression of
genes diagnostic of early neural crest cells, melanoblasts, xanthoblasts and all
chromatoblasts was normal through 24 hpf. Thus, the specification and early
differentiation of neural crest-derived chromatophores is unaffected by the enz
mutation. Double mutant analysis of enz with trpm7b508, a mutation previously
demonstrated to affect melanophores just prior to overt differentiation (Arduini
and Henion, 2004), revealed that embryos homozygous for mutations in both
genes resembled trpm7b508 single mutants with respect to the melanophore
phenotype. This places enz genetically downstream of trpm7b508 and, together
with phenotypic characterization, suggests enz function is required for the
terminal differentiation of chromatophores and specifically for the maintenance of
chromatophore cell morphology. 121 Because of the obvious defects in neural crest-derived chromatophore
differentiation in enz mutants, we examined the development of other neural crest derivatives. We found the development of neuronal, glia and ectomesenchymal neural crest derivatives in enz mutant embryos and larvae to be indistinguishable from that in wild-type embryos. Thus, enz appears to be specifically required for the development of all three chromatophore cell types within the neural crest.
Potential roles of enz in zebrafish development
Although the vast majority of homozygous enz larvae die after 2-3 weeks
of development, a minority of individuals do survive to adulthood and both males
and females are viable. However, homozygous adults were found to be
significantly smaller than wild-type siblings. Because neural crest-derived
chromatophores are not required for viability and are not thought to regulate
growth, enz must function in other cells required for growth and viability.
Interestingly, there is abundant evidence demonstrating roles for pituitary
hormone signaling systems in chromatophore development and growth control
(Fukamachi et al., 2004; Kawauchi and Baker, 2004; Sugimoto, 2002). Notably,
hormones known to regulate chromatophore morphology in teleosts specifically
have been shown in mammals and fish to affect body weight and composition, as
well (Kawauchi and Baker, 2004). Because our data indicate that enz acts cell
autonomously, enz would be predicted to be a downstream effector of such a
pathway.
122 It is also potentially informative to note that recent reports have described
mutants and morphants with live phenotypes similar to enz that disrupt genes
involved in organellar biosynthesis and transport (Amsterdam et al., 2004;
Golling et al., 2002; Pickart et al., 2004). Although more in depth investigations
into the nature of these pigmentation defects have not been described to date,
morpholino-mediated knockdown of zebrafish ATPase 6 subunit v0c (atp6v0c)
and vacuolar protein sorting protein 18 (vps18) result in punctate melanophore
appearance, apparent loss of xanthophores and reduced iridiphore numbers
(Pickart et al., 2004). Similar phenotypes were observed in mutants for a variety
of ATP synthase and vacuolar protein sorting genes generated in a transposon-
based mutagenesis screen (Amsterdam et al., 2004; Golling et al., 2002). We
noted a predicted vps35 ortholog assigned to chromosome 7
(http://www.sanger.ac.uk/), however, sequencing of the four enz alleles reported
here did not reveal any molecular lesions in the vps35 coding sequence. This
does not, however, rule out the possibility that enz encodes vps35, as lesions
may be present in non-coding regions of this gene in enz mutants. Further, the
genomic region in which the enz gene is located contains several other genes
involved in organellar function that may be good candidates for this locus
(http://zfin.org). Thus, it is also possible that the enz locus encodes a protein
required for normal formation or activity of cellular organelles. Further, the
characterization of enz mutant embryos and larvae presented here may provide clues to the mechanisms behind phenotypes described for organellar biosynthesis mutants and morphants described previously. In any case, the
123 potential relevance of the speculation discussed here will be better resolved with
the molecular identification of the enz locus.
Coordinated control of the late development of chromatophore sublineages
Although many loci have been identified that are required for crest-derived
pigment cell development, relatively few of them affect all three chromatophore
cell types found in zebrafish (Golling et al., 2002; Kelsh et al., 1996; Odenthal et
al., 1996; Rawls et al., 2003). Still fewer of these affect all chromatophore
sublineages in the same way (Amsterdam et al., 2004; Golling et al., 2002; Kelsh
et al., 1996). At the same time, other neural crest derivatives appear to develop
normally in enz homozygotes. Together with our finding that the enz mutation
acts cell autonomously with respect to melanophore development, this suggests
that the enz locus does not encode a gene generally required by neural crest- derived cells, but rather a molecule intrinsic to pigment cells themselves. Our
results strongly suggest that the development of all three zebrafish
chromatophore cell types depends on enz and are thus to some extent
coordinately regulated during terminal differentiation. However, since the
development of all three chromatophore lineages is unaffected in enz mutants
until late in their embryonic development and many mutants and identified genes
selectively affect individual or subsets of chromatophore types (Amsterdam et al.,
2004; Golling et al., 2002; Kelsh et al., 1996; Odenthal et al., 1996; Rawls et al.,
2003), the regulation of earlier stages of chromatophore development may be
genetically distinct. Nevertheless, it will be interesting to determine when and to
124 what degree the development of zebrafish chromatophore sublineages is regulated by common and distinct genes. Lastly, the great numbers of zebrafish pigment mutants, including enz, will provide important contributions to ultimately determining the genetic regulation of chromatophore development and leading to potential insights into clinically relevant conditions in humans involving pigment cells.
MATERIALS AND METHODS
Zebrafish
Adult zebrafish and embryos were maintained in the Ohio State University zebrafish facility. Adults and embryos were reared at 28.5oC and embryos were staged based on morphological criteria, according to Kimmel et al., 1995. Mutant lines were maintained in the AB* and WIK backgrounds. Homozygous mutant embryos and wild-type siblings were obtained by crossing heterozygous carriers.
Cell size quantification and analysis
Melanophores in 2 days post-fertilization (dpf) wild-type and enz embryos were imaged on a Zeiss Axioplan compound microscope. Individual cells were outlined and areas calculated using Zeiss Axioplan software with appropriate scalings.
Iridiphores in 6 dpf larvae were imaged on a Leica dissecting microscope.
Individual cells were outlined and areas calculated using Spot Advanced software with appropriate scalings. A minimum of 3 melanophores or iridiphores
125 in at least 3 wild-type and 3 enz mutant individuals were utilized for each set of measurements. Standard errors of mean were calculated for wild-type and enz cell areas, and mean numbers were also subjected to a one-tailed t-test to determine whether differences observed between wild-type and enz chromatophore areas were statistically significant.
Cell counts and statistics
Melanophores in 4 dpf larvae were counted in dorsal stripes. To better visualize distinct melanophores, wild-type and enz mutant larvae were placed in epinephrine (10 mg/ml) at 4 dpf for ~10 minutes, that results in redistribution of melanosomes to the center of the cell body in wild-type melanophores (Johnson
1995, Rawls 2000). Larvae were fixed in 4% paraformaldehyde at 4oC overnight, rinsed and stored in 1:1 PBS:glycerol. Larvae were then deyolked, mounted on single bridge cover slips, and viewed on a Zeiss Axioplan microscope.
Melanophores in the dorsal stripes were counted from somite 5 to somite 14.
Iridiphores in 6 dpf larvae were counted in the dorsal and ventral stripes. Live fish were viewed under incident light using a Zeiss dissecting microscope. Iridiphores were counted posterior of the hindbrain in the dorsal stripe, and caudally from the posterior edge of the yolk ball in the ventral stripe. Standard errors of mean were calculated for wild-type and enz cell counts. The mean numbers of cells in wild- type and enz embryos were also subjected to a one-tailed T-test to determine
126 whether the decrease in cell numbers in enz mutant embryos compared to wild-
type siblings was statistically significant.
In situ hybridization
In situ hybridizations were performed as described by Thisse, et al. (1993)
with minor modifications. A detailed protocol will be provided upon request. mitfa
and c-ret cDNAs were provided by D. Raible (Bisgrove et al., 1997; Lister et al.,
1999). cDNA clones of c-kit and fms were obtained from D. Parichy (Parichy et al., 2000b; Parichy et al., 1999). dct and ednrb1 cDNA clones were provided by
R. Kelsh (Kelsh et al., 2000b; Parichy et al., 2000a).
Immunohistochemistry
Antibody labeling was performed as previously described (An et al., 2002). 7 dpf
larvae were cryo-sectioned onto gelatin-subbed slides and stored at -20oC overnight. All neurons were detected with monoclonal antibody 16A11 that recognizes neuron-specific Hu RNA binding proteins (An et al., 2002; Henion et al., 1996), while DRG neurons and enteric neurons were subsequently identified by position within the larva. 16A11-immunoreactivity was detected using an
Oregon Green fluorescent secondary antibody (Molecular Probes). Sympathetic neurons were identified by detection of tyrosine hydroxylase (TH) using anti-TH polyclonal antibody (Pel-Freeze, Rogers, AZ). Neural crest-derived sympathetic
127 neurons were subsequently assessed by location of TH+ cells within whole-
mount wild-type and endzone larvae.
Mosaic analysis
Genetic mosaics were produced using cell transplantation techniques (Ho
and Kane, 1990). Donor embryos obtained from AB* or heterozygous (enz+/b431) crosses were manually dechorionated and injected at the one to two cell stage with 2 – 5 % lysinated rhodamine dextran (LRD, 10,000 MW, Molecular Probes) or lysinated fluorescein dextran (LFD, 10,000 MW, Molecular Probes) in 0.2M
KCl. Embryos were then allowed to develop to early blastula stages. For wild- type Æ mutant transplants, 10 – 20 cells were transplanted from LRD-labeled donor embryos into unlabeled host embryos. Donor-host pairs were kept separate and allowed to develop to >36 hpf, then fixed in 4% paraformaldehyde
at 4oC overnight. Donors and hosts were classified as either wild-type or mutant based on melanophore phenotype, and subsequently examined using a Zeiss
Axioplan microscope with Nomarski optics, and fluorescein or rhodamine filters to
detect donor cells. Punctate melanophores were scored as mutant, while stellate
melanophores were considered wildtype. Wild-type Æ wild-type and mutant Æ mutant transplants served as controls.
Genetic mapping
enz alleles were maintained in the AB* background. For mapping purposes, enz carriers in this background were crossed to a wild-type WIK line,
128 which is polymorphic to AB* (Nechiporuk et al., 1999). enz carriers in the WIK
background were then used to generate parthogenetic diploid progeny by
suppressing the second meiotic division with early pressure (Streisinger et al.,
1981). enz embryos and wild-type siblings were identified by live phenotype and
used to obtain DNA. enz was initially placed on the zebrafish genomic map based on PCR amplification of simple sequence length polymorphisms (SSLPs)
from diploid genomes (Knapik et al., 1996; Knapik et al., 1998; Postlethwait et al.,
1994; Postlethwait et al., 1998). enzb431 was initially demonstrated to be linked to
SSLP markers near the centromere of chromosome 7 based on the MGH
mapping panel (http://zebrafish.mgh.harvard.edu/zebrafish/index.htm). enz was
shown to cosegregate with z10441 (map position 36.7 cM), z11625 (map position
51.1 cM) and z6819 (map position 45.0 cM). Further linkage analysis was
performed on enz using haploid genomes, or meioses (Postlethwait et al., 1994).
This analysis confirms linkage to microsatellite marker z10441 (within ~2 cM), and further defines an interval for the enz locus proximal of z8252 and z10441
and distal to z6819.
129 TABLES AND FIGURES
Neural crest-related population Marker Timing Phenotype? Early neural crest cells crestin 14s-24hpf - Enteric neurons anti-Hu/location 7 dpf - Dorsal Root Ganglion Neurons anti-Hu/location 7 dpf - Sympathetic neurons anti-TH/location 7 dpf - Neural crest-derived glia foxd3 48 hpf - Craniofacial cartilage Alcian Blue 5 dpf -
Table 5.1 Other neural crest derivatives are normal in enz mutant embryos.
130
Melanophores (4 dpf) Iridiphores (6 dpf) wt 23.9 + 0.7 wt 53.1 + 0.9 enz 17.7 + 1.1 enz 41.4 + 1.7 compared to wt 74.1% compared to wt 78.0% P < 0.0001 P < 0.0001 Larvae were mounted dorsally.
Iridiphores were counted Larvae were mounted dorsally. posterior of somite 1 in the Melanophores were counted in dorsal stripe and posterior of the the dorsal and ventral stripes yolk sac in the ventral stripe. from somites 5 - 14.
Table 5.2 Melanophore and iridiphore numbers are significantly reduced in enz homozygotes.
131
Number of hosts Transplantation donor with donor-derived into host melanophores
wild-type into mutant 4 (dark, stellate) mutant into wild-type 0
Table 5.3. enz acts cell autonomously with respect to melanophore development
132
Figure 5.1. All three neural crest-derived chromatophore cell types are affected by enz mutations. Dorsal views of 54 hpf (A – E) and 82 hpf (F – J) wild-type (A, F) and enz mutant (B – J) embryos. (A, A inset) Wild-type melanophores are stellate and darkly pigmented at 54 hpf. (B – E) In contrast, enzb431 (B), enzos7 (C), enzos15 (D) and enzos18 (E) mutant embryos have small, punctate melanophores at this stage (compare insets in A and B). Yellow xanthophores, observed dorsally in the head of wild-type embryos (arrowhead in A) appear to be absent in enz homozygotes (B – E, arrowheads). (F – J) Iridescent iridiphores are also reduced in size in enz mutants (G – J, arrowheads) compared to wild- type siblings (F, arrowhead) at 82 hpf.
133
Figure 5.2. enz larvae and adults are undersized compared to wild-type siblings. Lateral (A, B) and dorsal (C, D) views of 7 dpf wild-type (A, C) and enz (B, D) larvae. (A) Wild-type zebrafish develop swim bladders between 4 and 6 dpf (arrowhead). (B) The vast majority of enz homozygous larvae do not develop swim bladders (arrowhead). Some melanophores recover in size and morphology in enz homozygotes (D) compared to wild-type siblings (C). (E – L) Those enz homozygous larvae that do develop swim bladders survive, but are runted compared to wild-type siblings (stage-matched wild-type and enz mutant larvae shown at the same magnification). Lateral (E, F) and dorsal (G – J) views of wild-type (E, G, I) and enz mutant (F, H, J) larvae at 21 dpf (E – H) and 30 dpf (I, J). Melanophores continue to be paler in enz mutant embryos than in wild-type siblings through at least 30 dpf (see arrowheads in G – J). (K, L) Lateral views of 119 dpf wild-type (K) and enz homozygous (L) adults. Normal overall morphology and pigmentation of enz mutant adults, as well as nascent fin stripe formation (L, arrowheads), suggests generalized growth retardation in enz mutants compared to wild-type siblings.
134
Figure 5.3. enz selectively affects chromatophores within the neural crest lineage. Wild-type (A, C, E, G) and enz (B, D, F, H) mutant embryos. Craniofacial cartilages revealed with alcian blue staining are normal at 5 dpf (A, B: ventral view). Cervical sympathetic neurons, which express TH immunoreactivity (arrowheads in C, D), are indistinguishable between wild-type (C) and enz mutant (D) embryos at 7 dpf. (E – H) Hu-positive neurons of the dorsal root ganglia (E, F, arrowheads) and the enteric nervous system (arrowheads in G, H) also appear normal in enz mutant embryos (F, H) compared to wild-type siblings (E, G) at 7 dpf.
135
Figure 5.4. Chromatophore precursors appear normal in enz homozygotes at 24 hpf. Lateral views of 24 hpf wild-type (A, C, E, G) and enz mutant (B, D, F, H) embryos. Early neural crest cells (crestin; A, B), xanthoblasts (xdh; C, D), melanoblasts (dct; E, F) and all chromatophore precursors (ednrb1; G, H) are all qualitatively normal at this stage.
136
Figure 5.5. Melanophore cell morphology changes in enz mutant embryos. (A, B) At 27 hpf, wild-type melanophores are large, stellate and well-pigmented (A). enz melanophores are also large and stellate at this stage, but are pale compared to Wildtype (B). (C, D) By 31 hpf, enz melanophore begin to transition to a punctate morphology (D), while wild-type melanophores remain large with many processes (C). (E, F) Wild-type (E) and enz mutant (F) embryos at 34.5 hpf. The morphological transition of melanophores in enz homozygotes continues in a rostro-caudal wave and is complete by approximately 48 hpf. (G – J) At 36 hpf, melanin is distributed throughout the cytoplasm of wild-type melanophores, reflecting the stellate morphology of these cells (G). (H) dct mRNA (red) is likewise distributed in the extensive processes of wild-type cells (arrowheads). Punctate distribution of melanin I) and dct mRNA (J, red) in enzos18 mutant melanophores is identical at 36 hpf.
137
Figure 5.6. Xanthophores are qualitatively reduced in number and size. (A, B) fms expression in 48 hpf wild-type (A) and enz homozygote (B) embryos. Fewer fms+ xanthophores are present in enz mutants than in wild-type siblings at this stage (arrows in A and B). (C, D) Methylene blue-stained xanthophores appear much larger in wild-type (C) embryos than in enz mutants (D) at 3 dpf.
138
Figure 5.7. enz acts cell autonomously with respect to melanophore development. (A) Nomarski image of a 28 hpf enz mutant host that has received cells from a wild-type donor (anterior to the left). Several large, dark melanophores (arrowheads) are present in addition to pale melanophores (arrows) normally observed in enz mutant embryos at this stage (see also Figure 5B). (B) High magnification of a wild-type cell (boxed area in (A)) that formed a dark melanophore in the mutant environment. (C) Melanophore in (B) viewed under a rhodamine filter.
139
CHAPTER 6
ZEBRAFISH FOXD3 IS SELECTIVELY REQUIRED FOR NEURAL CREST SUBLINEAGE DETERMINATION, MIGRATION AND SURVIVAL5
ABSTRACT
Mechanisms that control the patterning of neural crest remain
incompletely understood. Zebrafish homozygous for sympathetic mutation 1
(sym1) have defects in selective neural crest derivatives, including neurons, glia
and the craniofacial skeleton, but retain pigment cells. The sym1 mutation is a
nucleotide deletion that disrupts the forkhead DNA-binding domain of the foxd3
gene, which encodes a conserved winged-helix transcription factor. sym1 mutants have reduced neural crest expression of snail1b and sox10 throughout development, and tfap2a expression is reduced by the 12-somite stage, indicating Foxd3 is an essential regulator of these neural crest transcription factors. A critical role for Foxd3 in the survival of a subset of neural crest cells was shown by TUNEL analysis, which revealed aberrant apoptosis localized to the hindbrain neural crest at the 12- to 15-somite stage. Thus, foxd3 patterns the neural crest by selectively regulating the expression of lineage-associated transcription factors and promoting progenitor cell survival.
5 R.A. Stewart, B.L. Arduini, S. Bergmans, R.E. George, J.P. Kanki, P.D. Henion and A.T. Look, unpublished. Contributions of B.L. Arduini to this work include characterization of all live phenotypes, anti-Hu immunohistochemistry, Alcian blue staining, TUNEL analysis, and in situ hybridizations (Figures 6.2E-J; 6.3E, G; 6.4; 6.6; 6.7; 6.8; 6.9; 6.10; 6.11; 6.12A-H, K, N). 140 INTRODUCTION
The vertebrate neural crest is a transient population of embryonic precursor cells that gives rise to a variety of cell types, including neurons and glia of the peripheral nervous system, chromatophores, and elements of the craniofacial skeleton (Le Douarin and Kalcheim, 1999). As the neural crest develops in the dorsal neural tube, progenitor cells located at the neural plate border undergo an epithelial-mesenchymal transition to form the premigratory neural crest. These cells subsequently migrate along stereotyped pathways to specific locations throughout the embryo, where they differentiate and generate diverse tissue derivatives. While some of the regulatory mechanisms controlling the latter stages of neural crest differentiation have been identified (Le Douarin and Kalcheim, 1999), the genetic pathways mediating the induction of neural crest and the early specification of different neural crest cell sublineages remain unclear. Evidence suggests that the fates of many neural crest cells can be determined very early in development (Henion and Weston, 1997; Luo et al., 2003a; Raible and Eisen, 1994; Schilling and Kimmel, 1994; Wilson et al., 2004), raising the possibility that genes suspected to be involved in the induction and generation of neural crest may also be required for neural crest development in a lineage-restricted fashion.
During gastrulation, cells at the border between neural and non-neural ectoderm are induced to give rise to neural crest cells through the cumulative activities of Wnt, Bmp and Fgf signals originating from adjacent ectodermal and mesodermal tissues (Huang and Saint-Jeannet, 2004; Knecht and Bronner- Fraser, 2002). Neural plate border and/or neural crest cells can express a number of transcription factors, including foxd3, snail/slug, sox10 and tfap2a, that
141 have critical roles in neural crest development (Gammill and Bronner-Fraser, 2003), but how these genes direct the subsequent specification of neural crest sublineages remains unclear.
Zebrafish foxd3 is expressed during gastrulation by cells of the neural plate border, in posterior and ventral tailbud mesoderm and transiently in the developing floor plate (Odenthal and Nusslein-Volhard, 1998). Expression of foxd3 is then observed in premigratory neural crest cells of the head and trunk but is generally extinguished prior to the onset of migration, except for a subset of migrating cranial neural crest cells that transiently express this gene. As expression becomes downregulated in trunk neural crest cells, foxd3 expression is initiated in the somites and, later in some developing peripheral glia (Gilmour et al., 2002; Kelsh et al., 2000a).
FoxD3 has been proposed to have an important conserved role in early neural crest development because it is widely expressed by neural plate border cells and neural crest cells in all vertebrate species examined. Studies in frogs and chicks, using forced expression of wild-type, dominant-negative or antisense forms of foxd3, have shown that aberrant FoxD3 function results in abnormal neural crest development, although the findings have been inconsistent. While
loss-of-function experiments resulted in decreased neural crest gene expression, gain-of-function experiments indicated either the promotion or the inhibition of neural crest formation, perhaps due to foxd3 autoregulatory mechanisms (Dottori et al., 2001; Kos et al., 2001; Pohl and Knochel, 2001; Sasai et al., 2001). Such studies have also failed to address the role of foxd3 at different developmental stages and whether it is required for the development of all neural crest cells or only specific sublineages. Finally, attempts to elucidate the function of foxd3
142 through murine gene-knockout strategies have been hampered by embryonic lethality that prevents full analysis of the developing neural crest (Hanna et al., 2002).
Using an ENU-induced mutagenesis screen to identify genes required for normal development of the zebrafish peripheral sympathetic nervous system (PSNS), we recovered an embryonic lethal mutant, sympathetic mutation 1 (sym1), which lacks sympathetic neurons and displays craniofacial defects. We have determined that the molecular lesion at the sym1 locus results in a functional null allele of the zebrafish foxd3 gene. Unlike foxd3 mutant mice, homozygous sym1 mutant embryos survive throughout all stages of embryonic neural crest development, allowing the in vivo analysis of zebrafish foxd3 function in the generation of neural crest and its derivatives.
RESULTS
Defects in peripheral neuron and glial derivatives in sym1 mutants
To identify genes regulating PSNS development, we performed a forward genetic screen for mutations that disrupt the development of sympathetic neurons in embryos at 5-day postfertilization (dpf). Developing zebrafish sympathetic ganglia were readily identified at 2 dpf, using whole-mount in situ hybridization (ISH) to detect expression of tyrosine hydroxylase mRNA (TH), which is required for catecholamine biosynthesis in noradrenergic and dopaminergic neurons (An et al., 2002; Stewart et al., 2004). The first mutant recovered from this screen, sym1 (sympathetic mutation 1), showed a dramatic reduction or complete loss of TH expression in developing sympathetic neurons 143 (Figs. 6.1E, F). The mutation specifically affected neural crest-derived noradrenergic neurons, as TH expression in dopaminergic neurons of the CNS was not affected (Fig. 6.1E). Analysis of the superior sympathetic cervical ganglion in sym1 mutants between 2-3 dpf indicated that TH was not expressed in this region (Fig. 6.1F). Nor did we detect anti-Hu-positive neurons in the cervical ganglion of mutant embryos (data not shown). To determine whether sympathoadrenal progenitor cells are present in sym1 mutants, we examined the expression of zash1a, the earliest expressed marker of sympathetic neuron differentiation (Stewart et al., 2004). At 40 hours postfertilization (hpf), more than 20 zash1a+-positive cells (sympathoblasts) could be seen in wild-type embryos, which migrate ventrally to the dorsal aorta (Fig. 6.1C). zash1a expression adjacent to the dorsal aorta was not detected in sym1 mutant embryos, although one or two zash1a-positive cells could occasionally be observed displaced dorsally (Fig. 6.1G). At 48 hpf, phox2b is normally expressed in neural crest-derived cells adjacent to the dorsal aorta and at the ventral midline of the anterior gut tube, where sympathetic or enteric precursors form (Shepherd et al., 2004). Phox2b-expressing sympathetic or enteric precursors were not observed in mutant embryos (Fig. 6.1H). Taken together, these results show that the loss of TH expression in the developing sympathetic ganglia of sym1 mutants can be attributed to the absence of sympathetic progenitors, rather than their failure to differentiate into PSNS neurons. We next examined the development of dorsal root (DRG), enteric and cranial ganglion neurons in sym1 embryos. In wild-type embryos, trunk neural crest-derived DRG sensory neurons differentiate and form bilateral ganglia lateral to the ventral neural tube in each somitic segment (An et al., 2002). In most 144 sym1 mutants, DRG neurons were completely absent, although a few mutants had a small, variable number of DRG neurons that appeared unilaterally and dorsally displaced (Figs. 6.1I, M). Zebrafish enteric neurons, derived from vagal neural crest cells, initially differentiate in the anterior gut tube and become distributed antero-posteriorly. In sym1 mutant embryos, only a small minority of enteric neurons developed (Figs. 6.1J, N), and there was a dramatic loss of enteric neuron precursors, defined by the expression of c-ret (data not shown) and phox2b (Bisgrove et al., 1997). The defects seen in the neurons of cranial ganglia (Figs. 6.1K, O) were not as severe as neural crest-derived sympathetic, DRG and enteric neurons in sym1 mutants. In zebrafish, neurons and glia of cranial ganglia arise from both cranial neural crest and placodal origins, although the relative contribution of each cell population to the development of cranial ganglia is currently unknown (Schilling and Kimmel, 1994). We found that the trigeminal, posterior lateral line and vagal ganglion complex, are reduced in size by approximately 30% in sym1 mutant embryos, as indicated by neuron-specific expression of Hu proteins (Marusich et al., 1994). Finally, we analyzed the development of neural crest- derived peripheral glia in the cranial ganglia of sym1 mutant embryos, since foxd3 is expressed by at least some peripheral glia during development (Gilmour
et al., 2002; Kelsh et al., 2000a). Examination of foxd3 expression at 48 hpf in sym1 mutant embryos revealed reduction of peripheral glia within the cranial ganglia as well as those associated with axons of the lateral line system (Figs. 6.1L, P).
145 Neural crest-derived chromatophores are relatively normal in sym1
mutants.
The development of chromatophore sublineages was relatively normal in
sym1 mutants. Indeed, the numbers, morphology and patterning of differentiated
(melanized) melanophores in many mutant embryos were indistinguishable from
wild-type siblings at 3 dpf, although differentiation was delayed by a few hours
compared to that of wild-type embryos (Figs. 6.2A, C). While the expression
patterns of genes diagnostic of melanophore precursors, including the mitfa transcription factor and dct (Kelsh et al., 2000b) appeared to be slightly delayed in mutant embryos (approximately 2-6 hpf), the number and migration patterns of mitfa- (Figs. 6.2E, H) and dct-expressing (Figs. 6.2F, I) melanoblasts were normal.
The initial development of iridiphores and yellow xanthophores in sym1 mutants was also delayed, and these cells exhibited only mild and variable defects. For example, while we observed a slight (10-20%) decrease in the numbers of iridiphores in some mutant embryos at 3 dpf (Figs. 6.2B, D), by 4 dpf iridiphores appear normal when compared to their wild-type siblings. The variable defects observed in these chromatophores correlated with the variable reduction in the expression patterns of ednrb1 (endothelin receptor b1; iridiphore precursors; data not shown) and xdh (xanthine dehydrogenase; xanthophore precursors; data not shown) among individual mutant embryos. Importantly, migratory chromatophore precursors, indicated by sox10 expression (Dutton et al., 2001c), were present in sym1 mutants (Figs. 6.2G, J) by 24 hpf and 146 differentiated chromatophores were identified in their normal postmigratory
locations by 4 dpf(data not shown). Despite the slight delay in the differentiation of iridiphores and xanthophores in sym1 mutants, by 4 dpf the numbers of all chromatophore cell types appeared equivalent in both sym1 and wild-type embryos.
Craniofacial cartilage defects in sym1 mutants
One of the strongest phenotypes of sym1 mutant embryos and larvae is the abnormal size and shape of the jaws (see Fig. 6.2D). The embryonic zebrafish head skeleton is derived from three initial streams of cranial neural crest cells (Schilling and Kimmel, 1994) that form the first (mandibular) and second (hyoid) arches, and ceratobranchials I-V. In sym1 mutants, Alcian blue staining revealed that Meckel’s cartilage was displaced ventrally, the ethmoid plate did not form due to incomplete fusion of the trabeculae, and the hyosymplectic and ceratohyal cartilage structures were either reduced in size or missing (Figs. 6.3A-D). While the severity of defects in 1st and 2nd arch-derived cartilage elements displayed some phenotypic variability, these structures were always abnormal in sym1 mutants. The most severe craniofacial defect was the absence of cartilaginous elements normally derived from ceratobranchial arches I-V in all mutant embryos (Figs. 6.3A, C). The expression of dlx2 is required for jaw development and is expressed by migrating neural crest cells that contribute to the pharyngeal arches as well as by postmigratory arch-associated neural crest cells (Akimenko et al., 1994; Schilling and Kimmel, 1997). In sym1 mutant embryos dlx2 expression was 147 markedly reduced at 24 hpf in all recognizable arch-associated regions, especially the third arch, where dlx2 staining was completely absent (Fig. 6.3G). A similar dlx2 expression pattern was observed at 32 hpf (Fig. 6.3H). Reduced dlx2 expression correlated with the severity of cartilage defects, as neural crest cells in the posterior arch regions give rise to more posterior cartilage elements, such as the ceratobranchials (Schilling and Kimmel, 1994), which are absent in sym1 mutant embryos. Lastly, the reduced expression of sox9a in the arches of sym1 mutant embryos at 24 hpf and 48 hpf was also reduced (Fig. 6.4), and consistent with the reduced dlx2 expression.
The sym1 mutation inactivates the zebrafish foxd3 gene Genetic linkage analysis of the sym1 phenotype to 239 microsatellite markers revealed that markers z5294 and z10183 were linked to within 1-2 cM from the sym1 mutation on chromosome 6 (Fig. 6.5A). Because sym1 mutants had severe defects in neural crest-derived sympathetic neurons and cranial cartilages, we searched the available zebrafish meiotic, radiation hybrid, and heat-shock maps within this region, to identify ESTs and cloned genes expressed in developing neural crest. The most promising candidate gene identified was foxd3, a member of the forkhead family of transcription factors and a marker of
premigratory neural crest cells (Kelsh et al., 2000a; Odenthal and Nusslein- Volhard, 1998). foxd3 resides within 1 cM from the z10183 marker in the zebrafish radiation-hybrid mapping panels (Fig. 6.5A, http://zfin.org). The genomic sequence surrounding the single foxd3 coding exon was determined, and primers flanking this region were used to amplify genomic DNA from the sym1 heterozygous F2 parents, as well as F3 homozygous mutant and wild-type embryos. Alignment of these DNA sequences identified a single guanine 148 nucleotide deletion that induces a frame shift and a premature stop-codon within the DNA-binding domain of foxd3 (Fig. 6.5B). Together with recessive nature of the mutation, these data suggest that sym1 indicate represents a loss-of-function mutation in the foxd3 gene. FoxD3 belongs to a class of winged-helix transcription factors that contain a helix-turn-helix core of three α-helices, flanked by two loops, or “wings” within the DNA-binding domain. The third helix contacts the major groove, while the second (most C-terminal) winged region binds to the phosphate backbone in the minor groove, conferring DNA recognition specificity and containing a highly conserved stretch of 16 amino acids representing the nuclear localization signal (NLS) (Hellqvist et al., 1998). The sym1 mutation predicts loss of the second wing domain, indicating that the residual N-terminal peptide cannot bind DNA, as well as loss of the NLS and the C-terminal transcriptional effector domain (Pani et al., 1992). The recessive nature of the sym1 mutant phenotype, together with the loss of these critical functional domains, predicts that sym1 is a functional null allele of zebrafish foxd3. To confirm that the foxd3 gene was responsible for the sym1 phenotype, we transcribed full-length, wild-type foxd3 RNA in vitro and injected it into one cell-stage embryos from sym1 heterozygous parents. These embryos, together with anti-sense foxd3 RNA-injected (control) embryos, were analyzed for the rescue of sympathetic neurons based on TH expression (Figs. 6.5C-E), and genomic DNA was extracted for genotype confirmation (Methods). foxd3 mRNA rescued the TH expression pattern in the sympathetic cervical complex, defined as >10 TH-positive sympathetic neurons at 4 dpf, in 83% (n=24) of the injected sym1 mutant embryos. The number of TH-positive neurons in sym1 control- injected embryos was identical to that of uninjected sym1 embryos (0-2 TH- 149 positive neurons). To show that the specific loss of foxd3 function leads to the sym1 phenotype, we injected one-cell stage embryos with morpholino phosphorodiamidate oligonucleotides targeting the translation start site of foxd3. Injection of 15ng/embryo was sufficient to specifically phenocopy the PSNS and cartilage defects of sym1 mutants (data not shown). Notably, chromatophore development appeared normal in injected embryos, even at very high concentrations (35-50ng/embryo). Together these experiments demonstrate that disruption of the foxd3 gene is responsible for the selective neural crest phenotypes observed in sym1 mutant embryos.
Normal somite and floor plate development in sym1 mutants During zebrafish embryonic development, foxd3 is expressed by floor plate cells and later in the developing somites (Odenthal and Nusslein-Volhard, 1998). In sym1 mutant embryos somite development appeared normal at all stages examined (1-5 dpf) in terms of the timing and number of somite pairs that developed and somite size and shape (Fig. 6.2A, B; data not shown). The somitic expression of myoD was also indistinguishable between sym1 mutants and wild-type siblings Figs. 6.6A, D) and the pattern of floor plate expression of twhh and col2a in sym1 mutant embryos was unaffected (Figs. 6.6B, C, E, F).
Based on these findings, we conclude that foxd3 is not normally required for somite or floor plate development in zebrafish.
foxd3 function is required for the selective specification of early neural crest cell subpopulations
sym1 mutants were examined for the expression of transcription factors
that exhibit overlapping expression patterns with foxd3 in wild-type embryos and 150 play important roles in early neural crest development. sym1 mutant embryos
and wild-type siblings were examined for the developmental expression of foxd3
itself, tfap2a, snail1b and sox10 (Gammill and Bronner-Fraser, 2003), and the
pan-neural crest marker gene crestin (ctn) (Luo et al., 2001a; Rubinstein et al.,
2000). At the 3- to 5-somite stages, neural crest progenitors were present in the neural plate border in the anterior portion of embryos (Figs. 6.7A-E). At this stage, the expression of foxd3 and tfap2a transcripts in sym1 mutant embryos were indistinguishable from that in wild-type siblings (Figs. 6.7F, G). In contrast, expression level of the transcription factors snail1b and sox10 at the neural plate border were reduced in sym1 mutants (Figs. 6.7H, I), as was the expression of crestin (Figs. 6.7J). These differences in gene expression levels persisted through the 10-somite stage, with decreased expression of sox10, snail1b and
crestin becoming more pronounced and apparent in both cranial and nascent
trunk neural crest (Fig. 6.8). The cranial neural crest expression of foxd3 in wild-
type embryos normally begins to downregulate at the 10-somite stage; however,
sym1 mutants exhibited robust cranial expression of foxd3 at this stage (Fig. 6.8).
Furthermore, the midline convergence of the neural plate borders and formation
of the premigratory neural crest population was not developmentally delayed in
sym1 mutant embryos (Fig. 6.9). These results suggest that early induction of
the neural crest cell population at the neural plate border occurs normally, based
on the expression of tfap2a and foxd3, but fewer of these cells express normal
levels of sox10, snail1b and ctn in sym1 mutants compared to wild-type embryos.
The lineage-specific defects and abnormal gene expression patterns 151 observed in sym1 mutants suggests that foxd3 is differentially required for the
specification and/or survival of a subset of the neural crest sublineages.
Zebrafish harbouring mutations in tfap2a and sox10 also cause selective defects in a subset of neural crest derivatives, suggesting they play critical roles during neural crest specification and development (see Discussion). Thus, we analyzed tfap2a, sox10 and snail1b, during neural crest specification, between the 10- to
12-somite stage (Fig. 6.10) and followed the expression throughout later stages
(see Fig. 6.11). In wild-type embryos at the 10- to 12-somite stage, the transcription factors tfap2a, sox10 and snail1b are normally expressed in both premigratory and postmigratory neural crest cells (Figs. 6.10A-C, G-I, M-O). In sym1 mutants at the same stage, the expression of tfap2a (Figs. 6.10D-F) and sox10 (Figs. 6J-L) was reduced in both cranial and trunk neural crest, and remained reduced throughout development, particularly in the trunk neural crest
(Fig. 6.11). The expression of snail1b continued to be markedly reduced in all neural crest (Figs. 6P-R) relative to stage-matched wild-type embryos (Fig.
6.10M-O). Thus, both sox10 and snail1b expression remained reduced throughout neural crest development, while tfap2a was expressed at normal levels during neural crest induction (3-to 5-somite stage), but begins to be noticeably reduced by the 12-somite stage.
These abnormal neural crest gene expression patterns in sym1 mutants could
be due to the direct requirement for foxd3 in neural crest progenitors, or mis-
patterning of surrounding tissues, such as other ectodermal cell populations.
Based on the expression of HuC at the 3- and 14-somite stages (Fig. 6.9), the 152 development of Rohon-Beard sensory neurons in sym1 mutant embryos was found to be normal. Likewise, the neural plate expression of sox2 and msxb at the neural plate border were indistinguishable between sym1 mutant and wild- type embryos (Fig. 6.9). These results suggest that the developmental requirement for foxd3 function is specific to the developing neural crest. Thus, although the induction of the neural plate, generation of neural crest progenitors
(Fig. 6.7) and formation of the premigratory neural crest population (see below) appears to occur normally in sym1 mutants, the establishment of the normal transcription factor program within these cells is severely disrupted.
Neural crest migration defects in sym1 mutants
The active migration of cranial neural crest in wild-type zebrafish embryos
occurs in three distinct streams, emerging from rhombomeres (r) r2, r4 and r6
and migrating to the mandibular (1), hyoid (2), and posterior arches (3)
respectively. These migrating cells can be easily visualized using the pan-neural
crest marker ctn (Figs. 6.12A, B). Analysis of ctn expression in sym1 mutants at
the 12-somite stage showed a severe reduction in ctn-positive cells contributing
to the 1st stream, and almost no ctn-positive cells in the 2nd stream (Figs. 6.12E,
F, asterisk). While ctn expression levels at the position of the developing 3rd
cranial neural crest stream are relatively normal at the 10-somite in sym1
mutants, the numbers of ctn-positive neural crest cells at the 12-somite stage are reduced compared to wild-type siblings (Fig. 6.12E, F). Analysis of foxd3
153 expression in sym1 mutants at the 12-somite stage revealed that normal
numbers of premigratory cranial neural crest cells were present, but they failed to
express ctn or properly migrate away from the dorsal neural tube (compare Figs.
6.12E with G). This analysis also revealed that sym1 mutants continued to
express foxd3 at very high levels in the cranial neural crest, when expression of
foxd3 is nearly extinguished in wild-type siblings (compare Figs. 6.12C, D with G,
H).
Double ISH analysis of foxd3 and ctn expression in sym1 mutants showed that ctn expression was observed once migration was initiated, starting at the 14- somite stage for the 1st stream and the 17-somite stage for the 2nd stream, which occurred as foxd3 expression was downregulated (data not shown). The most anterior cranial neural crest cells in sym1, which normally contribute to the formation of the olfactory placode, continued to express foxd3 until the 26-somite stage but failed to express ctn (Figs. 6.12L, M black arrowhead). At this stage, ctn-expressing neural crest cells are observed migrating to deep ventral regions in the head in wild-type embryos (Figs. 6.12I, J, black arrow). However, in sym1 mutants the migration pattern of the ctn-positive cells is disrupted, and while some cells do move laterally over the head, they fail to migrate deeper into more ventral head structures (Figs. 6.12L, M, black arrow). In the trunk of sym1 mutants at the 26-somite stage, only a small number of ctn-expressing cells had begun to migrate within the most anterior somites (Fig. 6.12M), while in wild-type embryos, neural crest migration was advanced significantly (Fig. 6.12J). A comparison of ctn-expressing trunk neural crest cells in transverse sections of sym1 mutant and wild-type embryos revealed that the vast majority of sym1 mutant neural crest cells remained localized between the 154 dorsal surface of the neural keel and overlying ectoderm (Fig. 6.12K-N). At 36 hpf, reduced numbers of migratory ctn-expressing neural crest cells were present in sym1 mutants, and may represent migrating chromatophore precursors, as migrating sox10-expressing cells are also observed in sym1 mutants by 24 hpf (see Fig. 6.2G, J; data not shown). Thus, the neural crest progenitors in sym1 mutant embryos can delaminate from the neural keel, but are delayed in their migration, and the majority of these do not follow their correct migration paths, especially in the head (Figs. 6.12L, black arrow).
Loss of foxd3 function in sym1 mutants causes localized neural crest cell death In sym1 mutants, the ctn expression observed in the vagal neural crest at the 10-somite stage is severely reduced by 24 hpf (compare Figs. 6.12I with L, white arrow), possibly due to aberrant cell death. When TUNEL-labeling of apoptotic cells was examined at early stages of development (10-15 hpf) no significant differences were observed between sym1 and wild-type embryos, consistent with correct numbers of presumptive neural crest cells being formed in sym1 mutants at this time. However, beginning at 15 hpf, there is a dramatic increase in the number of TUNEL-positive cells in the dorsal ectoderm of sym1
mutant embryos within the region of the hindbrain, corresponding with the position of the 3rd neural crest stream (compare Figs. 6.12O-P with Q-R, asterisk). Neural crest cell death in this region of sym1 embryos was complete by 17 hpf, after which the number of TUNEL-positive cells in sym1 returned to wild-type levels. We did not observe any difference in the numbers of TUNEL- positive cells in regions other than those containing neural crest cells in the dorsal ectoderm. The appearance of the TUNEL-positive cells correlates with 155 the decreased numbers of ctn-positive cells at this time, and may contribute to the defects in the tissue derivatives normally arising from the 3rd neural crest stream in sym1 mutants. Thus, foxd3 is required for the survival of a subpopulation of neural crest cells in the hindbrain that normally gives rise to neurons, glia and the pharyngeal arches, which are severely affected in sym1 mutants.
DISCUSSION
In this study, we analyzed the first mutant isolated from our zebrafish PSNS screen. The underlying mutation, designated sympathetic mutation 1 (sym1), is a single nucleotide deletion that introduces a premature stop codon within the DNA-binding domain of zebrafish FoxD3, thus predicting a loss of DNA-binding activity. The successful use of forced expression of wild-type zebrafish foxd3 mRNA to rescue the sym1 phenotype, and the ability to phenocopy the mutant using foxd3-specific antisense morpholinos, provides strong confirming evidence to the specific loss of foxd3 function. We show that functional foxd3 is required for downregulating its own expression and that it is selectively required for sublineage fate specification, migration and survival. Our results indicate that foxd3 functions by directing the expression of other essential transcription factors in nascent neural crest cell subpopulations and that it is required for the survival of a subset of neural crest cells.
The role of foxd3 in early neural crest development Previous studies in chickens and frogs have shown that forced misexpression of FoxD3 can promote the expression of neural crest marker genes, suggesting that foxd3 is required for the induction of the neural crest. For 156 example, the forced expression of foxd3 in chick neural epithelium can induce neural crest markers such as Cad7, the HNK1 epitope, Integrinβ1 and Laminin, while its expression in frog animal cap assays induces expression of Slug and AP-2 (Cheung et al., 2005; Dottori et al., 2001; Pohl and Knochel, 2001; Sasai et al., 2001). While these studies indicate that foxd3 is sufficient to induce neural crest gene expression, they do not address whether foxd3 is required for neural crest induction in vivo. Our analysis of sym1 mutants demonstrates that foxd3 is not required for neural crest induction in zebrafish as foxd3- and tfap2a- expressing neural crest progenitors in the neural plate border develop normally in the absence of foxd3 function. Moreover, the correct numbers of early premigratory neural crest cells are generated in sym1 mutants, and delaminate from the dorsal neural keel, indicating that the epithelial-mesenchymal transition does not require foxd3 function. Thus, the in vivo induction and generation of premigratory neural crest cells occur in a foxd3-independent manner. Although neural crest is generated normally in sym1 mutants, the neural crest cell populations expressing sox10 and snai1b are significantly reduced in size at all developmental stages. In addition, while tfap2a expression is initiated normally, its normal sustained expression is aberrantly lost in a substantial proportion of cells in sym1 mutants. These results indicate that foxd3 is required
to induce and/or maintain the normal expression pattern of transcription factors that are essential for proper neural crest patterning.
Sublineage-specific fate determination by foxd3 In sym1 mutants there is a near complete loss of dorsal root ganglion, sympathetic and enteric neurons, and their precursor populations. The peripheral glia associated with the cranial ganglia and lateral line nerves are also 157 severely reduced, and the normal re-expression of foxd3 during later developmental stages is not observed. Since normal numbers of early premigratory cranial and trunk neural crest cells are present in sym1 mutants, these data indicate that foxd3 is required for the subsequent development of neuronal and glial progenitor cells of the neural crest. In contrast, melanocyte development is normal, and chromatophore development in sym1 mutants, while delayed, appears essentially normal by 4 dpf. These results indicate that foxd3 functions selectively to specify neuronal and glial sublineages, but is not required for specifying the development of melanocytes or chromatophore lineages. Our studies indicate that foxd3 is necessary for establishing the normal neural crest subpopulations expressing sox10 and snai1b, and maintaining the expression of tfap2a. Importantly, zebrafish with mutations in either tfap2a or sox10 exhibit defects in neural crest-derived peripheral neuron development, but to a lesser extent than sym1 mutants (Barrallo-Gimeno et al., 2004; Kelsh and Eisen, 2000; Knight et al., 2003). In colourless/sox10 (cls) mutant embryos, sympathetic neurons were not detected, while DRG and enteric neurons were present in reduced numbers (Kelsh and Eisen, 2000). The tfap2a mutants lockjaw (low) and mont blanc (mob) lack sympathetic neurons and exhibit a marked reduction in enteric neurons, but only mildly affect developing DRG
neurons (Barrallo-Gimeno et al., 2004; Knight et al., 2003). Both sox10 and tfap2a mutants exhibit defects in neural crest-derived glia. Thus, a simple explanation for the collective absence of glia, as well as DRG, sympathetic and enteric neurons in sym1 mutants, is the combined reduction of both tfap2- and sox10-expressing neural crest progenitors that results from loss of foxd3. Given that snail/slug family members can affect early neural crest development in other species, it remains possible that the reduction in snai1b expression in sym1 158 neural crest subpopulations may also contribute to the observed neuronal and glial defects in sym1 mutants. The relatively normal development of melanocytes and other chromatophores in sym1 mutants may seem inconsistent with the reductions in the sizes of sox10, tfap2a and snai1b expressing neural crest populations and with the chromatophore defects observed in cls, low and mob mutants. In cls mutants, for example, melanophore and iridiphore sublineages are severely disrupted while xanthophore development persists (Dutton et al., 2001b; Kelsh and Eisen, 2000), and in tfap2a mutants (low, mob) the development of all three chromatophore sublineages is affected (Barrallo-Gimeno et al., 2004; Knight et al., 2003). This apparent discrepancy becomes less striking when one considers that in sym1 mutants, neural crest subpopulations expressing sox10 and tfap2a (and snai1b) are retained, albeit in significantly reduced sizes, and a small number of sox10-expressing neural crest precursors migrate normally. These observations suggest that the subpopulations of neural crest cells expressing sox10 and tfap2a in sym1 mutants may be chromatophore progenitors and in these specific sublineages the expression of such genes may occur independently of foxd3 function. The differential requirements for foxd3 function between neuronal and chromatophore sublineages gain significance in light of abundant evidence from studies in birds and mice indicating that the segregation of neuronal and pigment precursors is one of the earliest events in neural crest cell diversification (Frank and Sanes, 1991; Henion and Weston, 1997; Luo et al., 2003a; Perez et al., 1999; Reedy et al., 1998; Vogel and Weston, 1988; Wilson et al., 2004). In zebrafish, clonal analysis has demonstrated a clear distinction between neurogenic and melanogenic neural crest cells of the head (Schilling and 159 Kimmel, 1994), and a strong clonal bias toward distinct neuronal and chromatophore progenitors in the trunk has been reported (Raible and Eisen, 1994). Our results suggest that foxd3 plays a crucial role, either directly or indirectly, in the establishment of neuronal neural crest sublineages. The loss of foxd3 function in sym1 mutants does not result in the expansion of chromatophore sublineages, nor do neuronal lineages expand in nacre/mitfa mutants in the absence of melanophore progenitors (Dorsky et al., 2000; Lister et al., 1999). These observations suggest that the specification and segregation of these sublineages depend upon instructive signaling events and that neither represents a default developmental pathway.
foxd3 is differentially required for trunk neural crest cell migration While significant numbers of trunk neural crest cells fail to disperse from the dorsal neural tube in sym1 mutants, a subpopulation of trunk neural crest cells show delayed migration. These results indicate that foxd3 is normally required for regulating the overall migratory capacity of most neural crest cells and the timing of trunk neural crest migration. These findings agree with studies showing that foxd3 can regulate neural crest migration by affecting cell adhesion properties (Cheung et al., 2005). In chicks, at least some of these foxd3 functions are mediated through sox10, consistent with the severe defects in trunk neural crest migration observed in cls/sox10 zebrafish mutants (Cheung et al., 2005; Dutton et al., 2001b; Kelsh and Eisen, 2000). snail family members have also been implicated in regulating neural crest migration, suggesting that the defects in neural crest migration may be mediated by the combined reduction in sox10- and snai1b-expressing neural crest populations in sym1 mutants (Aybar et al., 2003; Carl et al., 1999; LaBonne and Bronner-Fraser, 2000). 160 We conclude that foxd3 regulation of neural crest migration is sublineage-
specific, since the migration of chromatophore precursors appears normal and
correlates with the migration of a subset of sox10-expressing cells. These
observations suggest that foxd3 selectively regulates the migration of non-
chromatophore lineages (e. g., neuronal and glial), presumably via sox10, snai1b
and or other downstream regulators (but see below). While migration defects
may contribute to the loss of neurons in sym1 mutants, the ability to observe
neuronal differentiation in other mutants, where most neural crest cells fail to disperse from the dorsal neural tube (for example, spadetail, colgate; Ignatius,
Nambiar and Henion, unpublished data), suggests that cell differentiation may be
independent of their ability to migrate. Together, these results suggest that
sublineage specification by foxd3 temporally precedes or is coincident with trunk
neural crest migration, thus accounting for the selective migration defects in non-
chromatophore lineages of sym1 mutants. These defects in specification may
collectively result from the misregulation of sox10, snai1b and or tfap2a in sym1
mutants.
Molecular specification, patterning and survival of the neural crest by foxd3
Analysis of the craniofacial skeleton in sym1 mutants reveals a differential
requirement for foxd3 function in patterning the cranial neural crest, with the
anterior mandibular and hyoid elements being less affected than the posterior
branchial arches, which are nearly absent. Consistent with this phenotype,
precursors in the 3rd arch that generate the branchial arch elements exhibit the
161 strongest defects in dlx2 and sox9a expression and migration, which may be
correlated with reduced snai1b expression. Analysis of cls mutants suggests that
sox10 is not required for ectomesenchymal crest development in the head (Kelsh
and Eisen, 2000), so the reduced expression of sox10 in sym1 mutants is
unlikely to cause these defects. Since the early expression of tfap2a in the
cranial neural crest is independent of foxd3 function in the sym1 mutant, it is
possible that tfap2a functions genetically upstream of foxd3 in the hindbrain
region and may act through, or in parallel with foxd3 to specify the patterning of
the cranial neural crest. Furthermore, it was recently shown that sox9a and
sox9b play essential and synergistic roles in the patterning and development of
the craniofacial skeleton in zebrafish (Yan et al., 2005). Analysis of mutants and
the effects of sox9 misexpression demonstrate that this gene regulates foxd3 but
not tfap2a expression in pharyngeal arch cells. Thus, it appears that the correct patterning of the pharyngeal arches may depend upon regulated interactions between the tfap2a, sox9a/b and foxd3 pathways.
Our analyses highlight an important mechanism through which foxd3 patterns the neural crest: promoting the survival of a subpopulation of neural crest progenitor cells. Lineage analysis of the neural crest region that undergoes cell death in sym1 mutants indicates that this cell population contributes to a number of neural crest-derived tissues, including cranial neurons, glia, myocardial cells and brachial arches (Raible and Eisen, 1994; Raible and Eisen,
1996; Sato and Yost, 2003; Schilling and Kimmel, 1994). It is likely that the impairment of the 3rdarch stream and its derivatives results from the excessive
162 death of neural crest cells and severe loss of ctn-positive cells in this hindbrain
region of sym1 mutants. Thus, foxd3 may to be specifically required for the
survival of this neural crest subpopulation. Extensive cell death affecting the
second and branchial arches is also observed in the tfap2a mutant low, and may
partly reflect direct regulation of Hox group 2 genes in the hindbrain by tfap2a
(Knight et al., 2003). However, foxd3 expression is reduced in this hindbrain
region of low mutants and thus its craniofacial defects may also partially result
from selective cell death due to loss of foxd3 activity. The regulation of cell death
by foxd3 may be a conserved function, as disruption of foxd3 in mice revealed an
essential role in embryonic stem cell survival (Hanna et al., 2002). Thus, the
localized cell death caused by loss of foxd3 in sym1 mutants may provide an
important example of how selective cell survival contributes to the early
patterning of the zebrafish neural crest.
MATERIALS AND METHODS
Animal husbandry and cloning
Zebrafish were maintained, mutagenized and bred as described (Westerfield, 2000). sym1 mutants were initially identified by phenotype and TH expression, with subsequent confirmation by genotyping. The F1 WIK female harboring the sym1 mutation was outcrossed to the AB strain for genetic mapping and to generate F2 and F3 generations. Genomic DNAs from 40 wild- type and mutant embryos were pooled and screened for linkage to a set of 239
163 CA markers (sequences for markers are available at http://zebrafish.mgh.harvard.edu). The zFoxD3 full length clone used for RNA synthesis and in situ hybridrization experiments is described (Kelsh et al., 2000a; Odenthal and Nusslein-Volhard, 1998). Sequence surrounding the zFoxD3 exon was obtained from the Sanger genomic trace sequence repository and sequence comparisons were performed with SeqMan alignment software (DNASTAR, Inc.).
Whole-mount in situ hybridization, immunostaining and microinjections
Full-length, capped sense and antisense zFoxD3 RNA was generated with Message Machine, and a Poly-A tail was added with a Poly(A) tailing kit (Ambion). Based on the published GenBank sequence (NM_131290) for foxd3, a morpholino (MO) was designed: CACCGCGCACTTTGCTGCTGGAGCA (Gene Tools, Inc.). Whole-mount in situ hybridization was carried out as described (Thisse et al., 1995). Antisense probes were generated as described for dlx2 (Akimenko et al., 1994), ctn (Luo et al., 2001a), foxd3/fkd6 (Kelsh et al., 2000a), zash1a (Allende and Weinberg, 1994), mitf and dct, (Kelsh et al., 2000b; Lister et al., 1999), tfap2a (Knight et al., 2003), snail1b (Thisse et al., 1995) and sox10 (Dutton et al., 2001b). The phox2b probe was a generous gift of Dr. Su Guo and generated by digestion with EcoR1 followed by transcription with T3 RNA polymerase. Cartilage staining with Alcian blue was performed as described (Kimmel et al., 1998). Whole-mount antibody staining with anti-Hu mAb 16A11 (Molecular Probes) was performed on embryos fixed for 2 hr at RT in 4% PFA, as described by (An et al., 2002). Frozen sections of embryos processed for in situ hybridization were generated as described (Luo et al., 2001a).
164 TUNEL assays
Apoptosis was detected in embryos by terminal transferase dUTP nick- end labeling (TUNEL), according to the manufacturer’s protocols (In Situ Cell Death Detection Kit: POD and AP; Roche). After labeling, the embryos were washed with PBS (3x) and blocked in 4% goat serum for 30 min; antifluorescein antibody (1:5000) was added and incubated at 4oC overnight. The embryos were then washed in PBS (3x) and stained using Vector labs VECTORSTAIN® Elite ABC kits to detect AP or POD, activity according to the manufacturer’s recommendations (Vector® labs).
165 TABLES AND FIGURES
Figure 6.1. Defects in trunk and vagal neural crest derived peripheral neurons, cranial ganglion neurons, glia and lateral line glia in sym1 mutant embryos. Lateral views of wild-type (A, B) and sym1 mutant (E, F) embryos showing tyrosine hydroxylase (th) expression at 5 dpf (A, E) and 48 hpf (B, F). sym1 mutant embryos exhibit a dramatic reduction in the number of TH-positive cells (0-2 cells) in the region of the developing sympathetic cervical complex (sym) compared to their wild-type siblings (10-15 cells at 2 dpf, 40-50 cells at 5dpf). The loss of sympathetic neurons is reflected by the reduced number of sympathoblasts, based on zash1a expression at 40 hpf (C, G) and phox2b expression at 48 hpf (D, H). High magnification, ventral views of wild-type embryos showing expression of zash1a in migrating sympathoblasts (arrow in C), whereas in sym1 mutants, only a few scattered zash1a-positive sympathoblasts are evident (arrow in G). Similarly, expression of phox2b by sympathetic and enteric neuron precursors is strongly reduced in sym1 mutants (compare D with H). (I-P) Lateral views of wild-type (I-L) and sym1 mutant (M-P) embryos. Developing trunk DRG (I, M) and enteric (J, N) neurons labeled by anti-Hu immunoreactivity at 3 dpf. The absent neurons in sym1 mutants are indicated by arrowheads in M, and N (black arrows indicate position of developing neurons). (K, O) Lateral view of anti-Hu immunoreactivity reveals cranial ganglion neurons at 3 dpf. White arrows indicate the trigeminal, posterior lateral line and vagal ganglion complex. Black arrows indicate epibranchial ganglia. (L, P) Lateral view of foxd3 expression in 48 hpf wild-type (L) and sym1 mutant (P) embryos. In L, left arrow indicates glial cell expression associated with the posterior lateral line ganglion (right arrow). In P, arrow indicates missing glia in sym1 mutants. Abbreviations: absence of expressing cells (*), amacrine cells (am), dopaminergic neurons (dp), hindbrain (hb), neural crest-derived enteric and sympathetic precursors (NC), sympathetic neurons (sym).
166
Figure 6.1
167
Figure 6.2. Normal development of neural crest-derived chromatophores in sym1 mutants. Lateral views of living wild-type (A, B) and sym1 mutant (C, D) embryos at 3 dpf. Melanophore (black) and xanthophore (yellow; see arrows in A, C) cell numbers and patterns are similar between sym1 mutant and wild-type embryos. Incident light reveals slightly reduced silver iridiphores in some sym1 mutants (D) compared to wild-type embryos (B). Lateral views of developing melanoblasts in wild-type (E, F) and sym1 mutant (H, I) embryos. Expression of mitfa at 19 hpf (E, H) and dct at 36 hpf (F, I) in melanophore precursors in wild-type and sym1 mutant embryos. (G-J) Lateral views of migrating sox10-positive cells in 24 hpf wild-type (G) and sym1 mutant (J) embryos.
168
Figure 6.3. Abnormal pharyngeal arch development in sym1 mutants. Wild-type (A, B) and sym1 mutant (C, D) embryos at 5 dpf, stained with Alcian blue to reveal craniofacial cartilages. Ventral (A, C) and lateral (B, D) views show that the 1st and 2nd arch derivatives are present but abnormal in sym1 mutants, whereas 3rd arch-derived cartilages are missing altogether. dlx2 expression by neural crest-derived pharyngeal arch precursors (E-H). Dorsal views of 24 hpf wild-type (E) and sym1 mutant (G) embryos. Lateral views of 32 hpf wild-type (F) and sym1 mutant (H) embryos. Abbreviations: Meckel’s cartilage (mk), palatoquadrate (pq), ceratohyal arch (ch), ethmoid plate (ep), and ceratobranchial (cb) arches 1-5.
169
Figure 6.4. Abnormal sox9a expression in sym1 mutants. Wild-type (A, B) and sym1 mutant (C, D) embryos. (A, C) Dorsal views of sox9a expression in 24 hpf embryos. Black arrows indicate cranial arch streams of neural crest cells. (B, D) Lateral views of sox9a expression in 48 hpf wild-type (B) and sym1 mutant (D) embryos. White arrows indicate arch neural crest streams. At both 24 and 48 hpf, sox9a arch expression is reduced in sym1 mutants, particularly in posterior (branchial) arches.
170
Figure 6.5. sym1 is an inactivating mutation in foxd3. (A) Linkage of the foxd3 gene to the CA repeat markers on zebrafish chromosome 6. The positions of markers z10183, z6624 and z5294, are highlighted in orange (http://zfin.org) while foxd3 is highlighted in blue. The z10183 marker was mapped to within 1 cM of the foxd3 gene. (B) Schematic diagrams of the wild-type and mutant FoxD3 proteins. In sym1, a guanine (537) deletion results in a frame shift leading to a premature stop codon and disruption of the encoded DNA-binding domain. An acidic domain is highlighted in orange at the N-terminus and the C-terminus transcriptional effector domain is shown in green. Chromatogram traces below the schematic illustrate the nucleotide change (asterisk) affecting the foxd3- coding region. (C-E) Injection of foxd3 RNA rescues the sympathetic TH expression pattern in sym1 mutant embryos. Dorso-lateral views at 4 dpf of wild- type (C) and sym1 mutant embryos (D, E) labeled with the TH riboprobe. Wild- type siblings contain 10-40 TH-expressing cells in the cervical complex at this stage (C), whereas the sym1 mutant embryos injected with antisense foxd3 RNA have 0-2 TH-expressing cells (D). Injection of full-length foxd3 RNA into sym1 mutant embryos results in rescue (E), defined as >10 TH-expressing sympathetic neurons in the region of the cervical complex. Arrows indicate position of the cervical complex.
171
Figure 6.6. Normal somite and floorplate development in wild-type (A-C) and sym1 mutant (D-F) embryos. (A, D) Dorsal views showing normal myoD expression in axial and paraxial mesoderm in 18-somite stage wild-type (A) and sym1 mutant (D) embryos. (B, C, E, F) Lateral views showing normal floor plate expression of twhh (B, E, arrows) in 20-somite stage embryos and col2a1 (C, F, arrows) in 24 hpf embryos.
172
Figure 6.7. Molecular defects in neural crest specification in sym1 mutants. Dorsal views of wild-type (A-E) and sym1 mutant (F-J) embryos. Neural plate border expression of foxd3 (A, F), tfap2a (B, G), sox10 (C, H), snail1b (D, I) and ctn (ctn E, J) in 3-somite-stage embryos. Decreased expression of sox10, snail1b and ctn, but not foxd3 and tfap2a is evident in sym1 mutants. Arrows indicate position of midbrain-hindbrain boundary.
173
Figure 6.8. Gene expression abnormalities in premigratory neural crest populations in sym1 mutants. Dorsal (A-E, A’-E’) and lateral (F-J, F’-J’) views of neural crest expression of foxd3, tfap2a, sox10, snail1b and ctn in 10-somite stage wild-type (A-J) and sym1 mutant (A’-J’) embryos. The neural crest expression of sox10, snail1b and ctn is reduced in sym1 mutants. tfap2a expression is indistinguishable between mutant and wild-type embryos. Neural crest expression of foxd3 appears to be increased in sym1 mutants compared to wild-type embryos.
174
Figure 6.9. Normal development of Rohon-Beard sensory neurons, neural plate and neural plate border in sym1 mutants. Dorsal views of wild-type (A, C, E) and sym1 (B, D, F) mutant embryos. (A, B) huC expression in 3-somite stage embryos showing Rohon-Beard sensory neurons (arrowheads) in wild-type and sym1 mutant embryos. (C-F) The ectodermal neural plate and neural plate border domains, are respectively revealed by sox2 expression at the 6-somite stage (C, D) and msxb expression at the 14-somite stage (E, F) in wild-type and sym1 mutant embryos.
175
Figure 6.10. Gene expression abnormalities in premigratory and early migrating neural crest cells in sym1 mutants. Lateral, cranial and trunk views of wild-type (A-C, G-I, M-O) and sym1 (D-F, J-L, P-R) embryos, showing expression of tfap2a (A-F), sox10 (G-L), and snail1b (M-R) at the 10-to 12-somite stage. tfap2a expression begins to decrease in sym1 mutants at this stage, particularly in the trunk (compare A, C with D, F). The expression of sox10 in sym1 mutants (J-L) and snail1b (P-R) remains reduced in both cranial and trunk neural crest. The reduction in sox10 expression is specific to the neural crest, as sox10 expression in the developing otic placode is normal in wildtype and sym1 mutants (compare H with K). The position of the cranial neural crest anterior to the otic placode is indicated by arrowheads, and trunk neural crest by arrows.
176
Figure 6.10
177
Figure 6.11. Neural crest gene expression in 18-somite sym1 mutants. Dorsal trunk views of wild-type (A-C) and sym1 (D-F) embryos. Trunk neural crest expression of tfap2a (D), sox10 (E) and ctn (F) is strongly reduced in sym1 mutants.
178
Figure 6.12. Abnormal cell migration and cell death in sym1 mutant embryos. Wild-type (A-D) and sym1 mutant (E-H) embryos at the 12-somite stage, showing dorsal (A, E, C, G) and lateral (B,F, D, H) views. ctn expression reveals migration of the three cranial neural crest streams in wild-type embryos (A-B, numbered arrows). In sym1 embryos, ctn expression in cranial neural crest is severely reduced in the 1st neural crest stream (E, F) and almost absent in the 2nd stream (arrow with asterisk in E, F). ctn-positive cells of the 3rd stream are present in sym1 mutants, but only a few cells migrate (E-F). At the same stage, foxd3 expression is normally rapidly downregulated in wild-type embryos upon migration (C-D), but persists in sym1 mutants (G-H), showing that the cranial premigratory crest is present, but does not to migrate. Wild-type (I-J) and sym1 mutant (L-M) embryos double labeled with ctn (black) and foxd3 (red) at the 26- somite stage, showing dorsal (I, L) and lateral (J, M) views. At this stage, ctn is expressed throughout the head in deep ventral positions of wild-type embryos (I, J, black arrow), however, in sym1 mutants, ctn-positive cells fail to migrate ventrally, and spread out more laterally (L, M, black arrow). The expression of ctn is also strongly reduced in the region of the hindbrain posterior to the otic vesicle in sym1 mutants (L-M, white arrows), while foxd3 expression remains high in the anterior-most neural crest cells (L, M, black arrowhead). Numerous, ctn- expressing trunk neural crest cells are seen migrating in wild-type embryos at the 26-somite stage (J; white arrowhead), but few ctn-positive cells are seen in sym1 embryos (M; white arrowhead). Transverse sections at the anterior trunk level of 24 hpf wild-type (K) and sym1 mutant (N) embryos, showing ctn expression in both dorsal and ventral positions in wild-type embryos (K, arrowheads), but only dorsal positions in sym1 mutants (N, arrowheads). Neural crest cell death in sym1 mutants. Dorsal (O, Q) and lateral (P, R) views of TUNEL labeled 15- somite stage wild-type (O-P) and sym1 mutant (Q-R) embryos showing an increase in dorsal TUNEL+ cells (arrow in Q-R) in the hindbrain region in sym1 mutants. The position of the developing otic placode in indicated by asterisks.
179
Figure 6.12
180
CHAPTER 7
GENETIC INTERACTION OF ZEBRAFISH FOXD3 AND TFAP2α IN NEURAL CREST-DERIVED PIGMENT AND CRANIOFACIAL DEVELOPMENT6
ABSTRACT
Regulatory gene networks are used to generate diverse and related cell
populations throughout metazoan development. The vertebrate neural crest
gives rise to numerous cell types, including craniofacial cartilage, pigment cells,
and neurons and glia of the peripheral nervous system. Many transcription
factors have been identified that are expressed in the early neural crest and are
required for normal development of subsets of neural crest derivatives. These
gene products have been studied individually to ascertain their respective
contributions to neural crest development. We report here the effects of
simultaneous loss of foxd3 and tfap2α function during zebrafish pigment and
craniofacial development. We find a total loss of crest-derived pigment cells in
the absence of foxd3 and tfap2α function. We demonstrate that this is a result of
failure in specification of pigment cell precursors. We show that loss of craniofacial cartilages is dramatically more severe in foxd3-tfap2α mutant-
morphants than in either foxd3 mutants or tfap2α morphants alone. These
6 Unpublished. All data described herein, unless otherwise cited, were generated by B.L. Arduini under the guidance of P.D. Henion. 181 defects are preceded by absence of cartilage precursors. Thus, foxd3 and tfap2α together may be required for development of all crest-derived pigment and most or all crest-derived cranial cartilages.
INTRODUCTION
The neural crest is a transient vertebrate embryonic cell population that
gives rise to a wide variety of cell types, including pigment cells, craniofacial
cartilage, and neurons and glia of the peripheral nervous system (Le Douarin and
Kalcheim, 1999). This array of neural crest-derived cell types has long been of
interest in studying the mechanisms of induction and cell fate specification within
embryonic cell populations. As a result, many genes, particularly transcription
factors such as foxd3, tfap2α, sox10, sox9 and snail family members, have been
identified which are expressed in the premigratory and early migratory neural
crest (Knecht and Bronner-Fraser, 2002; Mayor and Aybar, 2001; Nieto, 2001).
These transcription factors are subsequently required for formation of subsets of
neural crest derivatives.
Loss-of-function studies of tfap2α in both mice and zebrafish reveal partial
jaw and pigment defects (Barrallo-Gimeno et al., 2004; Brewer et al., 2004;
Knight et al., 2004; O'Brien et al., 2004; Schorle et al., 1996; Zhang et al., 1996).
Additionally, zebrafish tfap2α mutants lockjaw (low) and mont blanc (mob) exhibit defects in formation of some neuronal neural crest derivatives, particularly sympathetics and enterics, as well as sensory neurons to a lesser extent
(Barrallo-Gimeno et al., 2004; Knight et al., 2003). Expression of other 182 transcription factors required for neural crest development, such as foxd3, sox10,
snai1b and sox9b, appears to be normal at early somitogenesis stages in
zebrafish tfap2α mutants, although expression of these factors are reduced to varying extents later in development. Thus, tfap2α has been hypothesized to be required for maintenance of subpopulations of neural crest derivatives, but only after initial specification of these lineages (Barrallo-Gimeno et al., 2004; Knight et al., 2004; Knight et al., 2003).
Although foxd3 knockout mice die prior to neural crest formation, zebrafish foxd3 mutants survive through larval stages (Hanna et al., 2002; Stewart et al., unpublished). Our previous work with the foxd3sym1 mutation demonstrates
delayed pigment cell development and reduced craniofacial cartilages similar, but
not identical, to tfap2α mutants (Hanna et al., 2002; Stewart et al., unpublished).
Both foxd3 and tfap2α appear to be absolutely required for sympathetic neuron
development (Barrallo-Gimeno et al., 2004; Knight et al., 2003; Stewart et al., unpublished). Whereas some dorsal root sensory neurons are present in tfap2α
mutants, they are abolished in foxd3 mutants. Conversely, some enteric neurons
develop in foxd3 mutants, but none in tfap2α mutants (Barrallo-Gimeno et al.,
2004; Knight et al., 2003; Stewart et al., unpublished). Interestingly, expression of
sox10 and snail1b is reduced during early somitogenesis stages in foxd3/sym1
mutants, while tfap2α expression is normal at these early time points. It seems,
therefore, that neural crest induction occurs appropriately in foxd3 mutants, but
the subsequent transcriptional program is not properly specified (Stewart et al.,
unpublished). 183 Despite all that is currently known, the precise roles of these molecules,
as well as their genetic interactions with other crest-specific transcription factors in forming the full complement of neural crest derivatives, remain elusive. Our data indicate that zebrafish tfap2α and foxd3 together are necessary for neural crest-derived chromatophore and cranial skeleton development. We show that when function of these genes is simultaneously abrogated, pigment cell precursors fail to be specified. In addition, we provide evidence that crest-derived cartilage precursors do not develop properly in the absence of both tfap2α and foxd3 function. We suggest that foxd3 and tfap2α are upstream of genetic cascades required to generate most, if not all, neural crest derivatives.
RESULTS
The foxd3sym1 mutation abrogates melanophore differentiation in sub-
maximal sox10 morpholino-injected embryos
Previous analyses revealed that loss of Foxd3 function results in a delay
of zebrafish chromatophore development (Stewart et al., unpublished).
Subsequently, however, all three pigment cell populations appeared relatively normal (Stewart et al., unpublished). This suggested that chromatophore development might not require foxd3 function. Alternatively, foxd3-dependent and foxd3-independent subpopulations might exist within each pigment cell lineage.
In embryos homozygous for severe sox10/cls alleles, pigment cells either
fail to differentiate (melanophores and iridiphores) or differentiate in reduced 184 numbers and undergo apoptosis (xanthophores), resulting in larvae that
completely lack crest-derived pigment (Dutton et al., 2001b; Kelsh and Eisen,
2000). Thus, all zebrafish chromatophores are sox10-dependent. When sub-
maximal doses of morpholino oligonucleotides against sox10 are injected into
wild-type embryos, some melanophores develop on a delayed time course
(Dutton et al., 2001a). We also found that injection of lower doses of sox10
morpholino into wild-type embryos permitted differentiation of some
melanophores (Figure 7.1). On the other hand, when the same dose of
morpholino was injected into foxd3sym1 mutant embryos, the complement of
melanophores that developed was significantly reduced compared to wild-type
sox10 MO-injected siblings (Figure 7.1). These data support the hypothesis that
distinct foxd3-dependent and foxd3-independent subpopulations exist within the melanophore sublineage, and by inference, in other pigment cell lineages, as well.
Pigment cell types fail to be specified in the absence of foxd3 and tfap2α function
We noted similar delays in melanophore development between foxd3sym1 mutant (Stewart et al., unpublished) and tfap2α mutant or morphant (Barrallo-
Gimeno et al., 2004; Knight et al., 2003; O'Brien et al., 2004) embryos. Together
with our previous report indicating comparable expression profiles for foxd3 and tfap2α in foxd3sym1 mutants (Stewart et al., unpublished), this suggested that
these two transcription factors might play similar roles in chromatophore 185 development. In order to investigate this possibility, we injected a splice site-
blocking morpholino against tfap2α (O'Brien et al., 2004) into foxd3sym1 mutant
embryos and wild-type siblings. Similar to previous reports, we found that tfap2α morphant embryos exhibit delays in melanophore development at 28 hpf (O'Brien et al., 2004). In contrast, foxd3-tfap2α mutant-morphant embryos completely lack
neural crest-derived melanophores at this stage (Figure 7.2). In all cases, non-
crest-derived pigmented retinal epithelium (PRE) developed normally, suggesting
that pigmentation defects observed in the absence of foxd3 and tfap2α function
singly or in combination are not the result of a general delay in development
(Figure 7.2). As reported previously, foxd3sym1 and tfap2α mutants or morphants have relatively normal populations of melanophores, xanthophores and iridiphores by larval stages (Barrallo-Gimeno et al., 2004; Knight et al., 2003;
O'Brien et al., 2004; Stewart et al., unpublished). In contrast, foxd3-tfap2α mutant-morphant larvae are nearly devoid of crest-derived pigment cells through at least 5 dpf, although a small number of xanthophores sometimes differentiate over the head (Figure 7.3 and data not shown).
In order to determine whether these defects were the result of a failure in specification of chromatophore precursors or a defect later in their ontogeny, we examined the expression of pigment cell lineage-specific genes. Mitf has been shown to be both necessary and sufficient for melanophore development in zebrafish, and is the earliest known marker for melanoblasts in this system
(Elworthy et al., 2003; Lister et al., 1999). Expression of mitfa is delayed in foxd3sym1 mutant embryos (Stewart et al., unpublished) but is robustly expressed 186 by the 26-somite stage (Figure 7.4). tfap2α morphants exhibit relatively normal
mitfa expression at this stage (Figure 7.4), consistent with phenotypes previously
reported for tfap2α mutants lockjaw (low) and mont blanc (mob) (Barrallo-
Gimeno et al., 2004; Knight et al., 2003). Strikingly, no mitfa expression is observed in foxd3 mutants injected with tfap2α morpholino (Figure 7.4). Knight,
et al. and Barrallo-Gimeno and colleagues reported slight reduction in expression
of xanthoblast-specific xdh at 24 hpf (Barrallo-Gimeno et al., 2004; Knight et al.,
2003; Parichy et al., 2000b). Wild-type embryos injected with tfap2α morpholino
exhibit reduced expression of xdh similar to mutant phenotypes at this stage
(Barrallo-Gimeno et al., 2004; Knight et al., 2004). Injected siblings homozygous
for the foxd3sym1 mutation, however, exhibit little to no xdh expression at 24 hpf
(Figure 7.4).
In agreement with analyses of tfap2α mutants (Barrallo-Gimeno et al.,
2004; Knight et al., 2003), our wild-type-tfap2α morpholino-injected controls have
relatively normal sox10 expression at the 18 somite stage, although crest
migration is slightly delayed. Our foxd3sym1 uninjected controls have reduced
sox10 expression at this stage, as previously reported (Stewart et al.,
unpublished). In contrast, sox10 expression is almost completely absent in 18
somite foxd3-tfap2α mutant-morphant embryos (Figure 7.4). In contrast,
expression of foxd3 is relatively normal at the 5-somite stage in foxd3-tfap2α
mutant-morphants (data not shown). Together, these data suggest that although
187 neural crest cells are induced in the absence of foxd3 and tfap2α function, these
cells fail to be specified as pigment cell precursors.
Craniofacial cartilages are lost in foxd3sym1-tfap2α mutant-morphants
Loss of function of either tfap2α or foxd3 results in reduction and disorganization of pharyngeal arch derivatives (Barrallo-Gimeno et al., 2004;
Knight et al., 2003; Stewart et al., unpublished). Specific, but distinct, elements of
the lower jaw are missing, while more dorsal structures, such as the
neurocranium, remain intact (Barrallo-Gimeno et al., 2004; Knight et al., 2003;
Stewart et al., unpublished). In contrast, foxd3sym1 mutant embryos injected with tfap2α MO completely lack all lower jaw structures as revealed by Alcian blue
staining (Figure 7.5). All but the most posterior portion of the neurocranium is
absent, as well (Figure 7.5). We examined expression of dlx2, which is diagnostic
of cartilage precursors within the cranial neural crest (Akimenko et al., 1994).
dlx2 expression is abolished in the pharyngeal arches of foxd3-tfap2α mutant-
morphants (Figure 7.5). Normal expression of dlx2 in non-crest-derived forebrain
domains is retained (data not shown). These data suggest that foxd3 and tfap2α interact synergistically in development of cranial crest-derived structures.
188 DISCUSSION
Zebrafish foxd3 and tfap2α are required for chromatophore specification
Both foxd3 and tfap2α have been implicated in the development of a
variety of neuronal and glial neural crest derivatives in the trunk of vertebrate
embryos, but either of these genes alone is dispensable for the majority of neural
crest-derived pigment development (Barrallo-Gimeno et al., 2004; Knight et al.,
2003; O'Brien et al., 2004; Stewart et al., unpublished). Strikingly, however,
simultaneous abrogation of foxd3 and tfap2α function in zebrafish results in
complete absence of chromatophore cell types. We showed that chromatophore
precursors are also absent in foxd3-tfap2α mutant-morphants, indicating failure
of specification of these lineages. Interestingly, this pigment phenotype is
identical to that observed for severe sox10/cls alleles (Dutton et al., 2001b; Kelsh
and Eisen, 2000). Together with our data that sox10 expression is nearly absent
in foxd3-tfap2α mutant-morphants at mid-somitogenesis stages, this suggests
that these two molecules may act genetically upstream of sox10 to specify
pigment cell types within the neural crest. However, both foxd3 and tfap2α have
been proposed by some to play roles exclusively in later neural crest
development, subsequent to initial specification of neural crest sublineages.
Based on analyses in chick, it has been suggested that foxd3 is not required for specification of crest-derived pigment lineages, but for their repression in favor of other lineages (Kos et al., 2001). In support of this hypothesis, foxd3+ cells were
not observed on the dorso-lateral pathway, which is populated exclusively by
melanoblasts (Erickson and Goins, 1995), nor did foxd3+ cells express 189 melanoblast sublineage-specific mitf (Kos et al., 2001). However, the timing of
foxd3 manipulations in this study, after the initial formation of the neural crest cell
population, does not exclude the possibility that foxd3 is required earlier for
specification of crest cells that will later give rise to pigmented derivatives.
Indeed, foxd3+ cells in mice migrate along both the ventro-medial pathway and
the dorso-lateral pathway (Dottori et al., 2001). Thus, foxd3 appears to be
expressed by NCCs that will give rise to pigment cells, as well as those
contributing to the peripheral nervous system (Dottori et al., 2001). Dottori, et al.
note that while migration along the dorso-lateral pathway is delayed in chick
relative to migration along the ventro-medial pathway, neural crest cells enter
these pathways simultaneously in mice, and suggest that the seemingly
contradictory foxd3 results may be the result of differences in timing of
melanoblast specification between chick and mouse (Dottori et al., 2001).
It has similarly been suggested that tfap2α is required for differentiation
and survival, rather than initial specification, of neural crest derivatives (Barrallo-
Gimeno et al., 2004; Brewer et al., 2004; Knight et al., 2003; Schorle et al., 1996;
Zhang et al., 1996). However, deficiencies in expression of some early neural
crest genes, such as snail/slug, sox9, zebrafish crestin, and to a limited extent,
sox10, were noted in tfap2α mutants and morphants (Brewer et al., 2004; Knight
et al., 2003; O'Brien et al., 2004). Expression of these genes has not, to our
knowledge, been examined in tfap2α-null mice (Schorle et al., 1996; Zhang et al.,
1996). tfap2α and sox10 expression, as well as neural crest cell migration, were found to be normal in conditional tfap2α knockout mice using Cre under control 190 of the Wnt1 promoter to drive excision of critical dimerization and DNA binding
domains (Brewer et al., 2004). However, Wnt1 is first expressed in the dorsal neural tube subsequent to neural crest induction in the converging neural folds
(Ikeya et al., 1997), whereas tfap2a is expressed in the neural folds beginning during gastrulation (Barrallo-Gimeno et al., 2004; Knight et al., 2003; Luo et al.,
2002; Mitchell et al., 1991). Thus, it is possible that tfap2α was expressed prior to
the knockout, and that this expression was sufficient to direct sox10 expression
and specification of melanoblasts, but not for later survival of neural crest
derivatives.
Previous analyses have shown that mitf is necessary and sufficient for
melanoblast specification (Hornyak et al., 2001; Lister et al., 1999; Tachibana et
al., 1996) and that sox10 binds and activates the mitf promoter in vitro and in vivo
(Bondurand et al., 2000; Dorsky et al., 2000; Huber et al., 2003; Potterf et al.,
2000). Our data indicate that foxd3 and tfap2α together are necessary for
expression of sox10, suggesting that they are upstream of sox10 in a pathway
required to specify embryonic melanoblasts, and by analogy, other neural crest-
derived pigment cell types. Consistent with this notion, sox10/cls zebrafish
mutant embryos lack all chromatophores, and foxd3 is expressed normally in
mutant embryos (Kelsh and Eisen, 2000). Wnt signaling within the neural crest is also necessary for melanophore development, and has been shown to directly regulate the mitf promoter, as well (Hari et al., 2002; Saito et al., 2003; Takeda et al., 2000). Experiments in Xenopus indicate that Wnt signaling also regulates sox10 expression within the neural crest, but it remains unclear whether this 191 regulation is direct or indirect (Aoki et al., 2003). Thus, Wnt signaling may
modulate sox10 expression through foxd3 and tfap2α. Alternatively, Wnts may
directly regulate the sox10 promoter. However, we found that neural crest
expression of sox10 and mitf are absent in foxd3sym1-tfap2α mutant-morphants, suggesting that Wnt signaling is insufficient to activate either sox10 or mitf expression in the absence of normal foxd3 and tfap2α function.
Zebrafish foxd3 and tfap2α are required for normal development of cranial
neural crest derivatives
Our data also indicate that foxd3 and tfap2α genetically interact in
development of cranial crest-derived ectomesenchyme. While foxd3sym1 mutants and tfap2α mutants or morphants exhibit reductions in and disorganization of
lower jaw structures, abrogation of foxd3 and tfap2α function together results in
complete loss of lower jaw and severe reduction of the neurocranium. The
neurocranium phenotype is intermediate between those of loss-of-function of
zebrafish co-orthologous genes sox9a and sox9b (Yan et al., 2005), while the
lower jaw phenotype is identical to that of sox9a/jellyfish mutants (Piotrowski et
al., 1996; Yan et al., 2005). Interestingly, expression of dlx2, a gene involved in cell adhesion and migration of cranial neural crest cells (Akimenko et al., 1994;
McKeown et al., 2005), is only slightly reduced in sox9b single or sox9a, sox9b double mutants (Yan et al., 2005), but is completely abolished in foxd3-tfap2α
mutant-morphants (this study). Together, these data point to overlapping, but
192 non-identical, requirements for the sox9 orthologs, foxd3 and tfap2α in jaw
development.
Loss of sox9a and sox9b function alone or in combination results in
reduced expression of foxd3 but not tfap2α (Yan et al., 2005). tfap2α mutants
exhibit reduced sox9a and sox9b expression after initiation of cranial neural crest
migration (Barrallo-Gimeno et al., 2004; Knight et al., 2004). Further, tfap2α is not upregulated by forced misexpression of sox9a or sox9b (Yan et al., 2005).
These data suggest that tfap2α is required for formation of subpopulations of
sox9a- and sox9b-expressing cells in cephalic crest regions (Yan et al., 2005).
sox9a and sox9b zebrafish mutants, singly or in combination, exhibit reduction in
foxd3 expression in cranial neural crest. In addition, overexpression of either
sox9a or sox9b upregulates foxd3 expression (Yan et al., 2005). However, loss-
of-function of foxd3 also results in decreases in expression of sox9a (Stewart et
al., unpublished). and sox9b (this study) at cranial crest levels. Thus, it appears
that while tfap2α may be genetically upstream of sox9a and sox9b, sox9a, sox9b
and foxd3 may cross-regulate each other during cranial neural crest
development.
MATERIALS AND METHODS
Zebrafish
Adult zebrafish and embryos were maintained in the Ohio State University
zebrafish facility. Adults and embryos were reared at ~28.5oC and embryos were
193 staged based on morphological criteria, according to Kimmel et al., 1995. Mutant lines were maintained in the AB* and WIK backgrounds. Homozygous mutant embryos and wild-type siblings were obtained by crossing heterozygous carriers.
Whole-mount in situ hybridization and staining
In situ hybridizations were performed as described by Thisse, et al. (1993) with minor modifications. A detailed protocol will be provided upon request. Mitfa cDNA was provided by D. Raible (Bisgrove et al., 1997; Lister et al., 1999). cDNA clones of c-kit and xdh were gifts from D. Parichy (Parichy et al., 2000b;
Parichy et al., 1999). sox10 and ednrb1 cDNA clones were provided by R. Kelsh
(Kelsh et al., 2000b; Parichy et al., 2000a). sox9a cDNA was obtained from J.
Postlethwait (Chiang et al., 2001). A. Fritz provided a dlx2 cDNA clone
(Akimenko et al., 1994). Cartilage staining with Alcian blue was performed as described (Kimmel et al., 1998).
Morpholino injections
As previously described, morpholino (MO) anti-sense oligonucleotides against the sox10 5’-UTR (Dutton et al., 2001a) and the tfap2α exon2-intron2 boundary (O'Brien et al., 2004) were obtained from Gene Tools, Inc., and reconstituted according to company specifications. Stock morpholinos were diluted in ddH2O and 1% phenol red, and injected into the yolks of 1 - 8 cell stage embryos. Altered splicing of tfap2α transcripts was confirmed in 26 somite and
194 72 hpf embryos, as described (O'Brien et al., 2004). 0.5 - 1 ng doses of sox10
MO and 2 - 12 ng doses of ap2E2I2 MO were administered as appropriate.
Results reported for cell counts in sox10 MO-injected embryos, Alcian Blue staining, and in situ hybridizations with mitfa and xdh in tfap2α MO-injected
embryos were observed across two to three families. In situ hybridizations of
tfap2α MO-injected embryos with sox10 and dlx2 riboprobes were performed on
at least 50 embryos from single families for each stage reported. Embryos were
then genotyped to identify foxd3sym1 mutant embryos and wild-type siblings using
the z5294 microsatellite marker
(http://zebrafish.mgh.harvard.edu/zebrafish/index.htm), which is tightly linked to
the foxd3 locus (Stewart et al., unpublished).
Melanophore Cell Counts
sox10 MO-injected embryos were dechorionated and anesthetized at 3 dpf and visualized on a Leica dissecting microscope. Melanophores were counted in the dorsal, ventral and lateral stripes of individual embryos, which were subsequently kept separately. Embryos were subsequently genotyped as above.
195 TABLES AND FIGURES
Figure 7.1 Reduction in total number of melanophores in foxd3sym1 mutant embryos compared to wild-type siblings following sox10 morpholino injection. (A- D) At sub-maximal doses of morpholino, many fewer melanophores develop in foxd3sym1-sox10 mutant-morphant embryos (D) than in wild-type (A), foxd3sym1 mutant (B) or sox10 morphant (C) embryos. (E) On average, the number of melanophores observed in foxd3sym1 mutant embryos injected with sox10 morpholino (purple bars) is less than the number of melanophores observed in wild-type siblings injected with sox10 morpholino (blue bars). X axis, total number of melanophores per embryo; Y axis, number of embryos.
196
Figure 7.2. Melanophores are absent in 28 hpf foxd3sym1-tfap2α mutant- morphant embryos. Lateral views of 28 hpf embryos. (A) At 28 hpf, wild-type embryos have many large, stellate melanophores, especially over the head and in anterior trunk regions (arrows). (B, C) foxd3sym1 mutants and tfap2α morphants have comparable reductions in melanophore numbers (arrows). (D) foxd3sym1 mutants injected with tfap2α MO completely lack neural crest-derived melanophores.
197
Figure 7.3. Visible phenotypes in live foxd3 sym1-tfap2α mutant-morphant larvae. Lateral views of (4 dpf) larvae under transmitted (A – D) and incident (E – H) light. (A) Wild-type larvae have many melanophores in the dorsal and ventral stripes, and dorsally in the head (arrowheads). (B, C) foxd3sym1 mutant and tfap2α morphant embryos have significant numbers of melanophores at 4 dpf (arrowheads). Jaw structures protrude ventrally in foxd3sym1 mutants and tfap2α morphants (arrows, compare to A). (D) foxd3sym1 mutants injected with tfap2α MO completely lack crest-derived melanophores (arrowheads). Jaw structures appear to be missing (arrowhead). (E – H) Reflective iridiphores can be seen in wild-type, foxd3sym1, and tfap2α MO-injected larvae under incident light (E – G, arrows). Yellow xanthophores are apparent along the dorsal aspect of these embryos (E – G, arrowheads). (H) In contrast, both iridiphores and xanthophores are absent in foxd3sym1-tfap2α MO-injected larvae (arrows and arrowhead, respectively).
198
Figure 7.4. Chromatophore precursors fail to be specified in the absence of foxd3 and tfap2α function. (A – D) Lateral views, embryos labeled with mitfa riboprobe at 26 somite-stage (A – D), xdh riboprobe at 24 hpf (E – H) or sox10 riboprobe at 18-somite stage (I – L) and 27 hpf (M – P). (A – D) Both foxd3sym1 mutant (B) and wt-tfap2α morphant (C) embryos have reduced mitfa expression compared to wild-type embryos (A). foxd3sym1-tfap2α mutant-morphant embryos completely lack mitfa+ cells at this stage. (E – H) Expression of xdh is reduced in foxd3sym1 mutant embryos (F), but relatively normal in tfap2α morphant embryos (G) compared to wild-type siblings (E). (H) xdh expression is nearly absent in foxd3sym1 mutant embryos injected with tfap2α morpholino at 24 hpf. (I – L) sox10 is expressed in premigratory and migrating neural crest cells at the 18-somite stage in wild-type embryos (I, arrowheads). foxd3sym1 mutant (J) and to a lesser extent tfap2α morphant (K) embryos have reductions in sox10+ cells, and few of these cells are observed to migrate at this stage. sox10-expressing cells are almost completely absent in foxd3sym1-tfap2α mutant-morphant embryos (L) at this stage, although some persist (arrowheads). (M – P) Some migrating sox10+ cells are observed in foxd3sym1 mutants (N) and tfap2α morphants (O) at 27 hpf, although the numbers of sox10-expressing cells are still reduced compared to wild-type embryos (M). (P) Expression of sox10 is not observed in premigratory or migratory neural crest locations in foxd3sym1-tfap2α mutant-morphants, but is retained in the non-crest derived otic vesicle (arrows in M and P).
199
Figure 7.5. Craniofacial cartilages are synergistically affected by simultaneous loss of foxd3 and tfap2α function. (A - D) 4 dpf larvae stained with Alcian Blue, ventral views, anterior to the left. (A) Wild-type embryos have well-defined lower jaw structures, including the Meckel’s cartilage, ceratohyal and palatoquadrate elements, and ceratobranchial structures. Dorsal elements of the lower jaw are collectively referred to as the neurocranium. (B) foxd3sym1 mutants lack third arch derivatives, and show disorganization of remaining cartilage structures, including failure of neurocranium elements to fuse. (C) Posterior lower jaw elements are also absent in tfap2α morphants, with concomitant disorganization of remaining structures. (D) foxd3-tfap2α mutant-morphants completely lack ventral lower jaw structures, as well as anterior portions of the neurocranium. cb, ceratobranchial; ch, ceratohyal; m, Meckel’s cartilage; ne, neurocranium, pq, palatoquadrate. (E - H) 24 hpf embryos labeled with dlx2 antisense riboprobe. Dorsal views, anterior to the left. dlx2 expression in the pharyngeal arches is reduced in foxd3sym1 mutant (F) and tfap2α morphant (G) embryos compared to wild-type siblings (E) (arrows). (H) foxd3-tfap2α mutant-morphant embryos completely lack pharyngeal arch expression of dlx2.
200
Figure 7.6. sox9b expression is reduced in foxd3sym1 mutants compared to wild-type siblings. Dorsal views of 6-somite stage embryos labeled with sox9b antisense riboprobe, anterior to the top. sox9b is robustly expressed in cranial regions of wild-type embryos (A), but much less so in foxd3sym1 mutant embryos (B). Arrows approximate posterior-most cranial level. e, eye; ov, otic vesicle.
201
CHAPTER 8
DISCUSSION
Within the neural folds and upon delamination from the dorsal neural tube, neural crest cells are morphologically very similar to one another. Thus, the neural crest has sometimes been described as a homogenous population of cells with equivalent potential to generate a variety of derivatives. Using both forward and reverse genetic analyses, we have demonstrated that sublineage-specific genetic requirements exist at early stages, prior to the expression of genes diagnostic of individual sublineages, as well as much later in their development.
Significantly, our data support the notion that premigratory neural crest is, in fact, a collection of subpopulations requiring different but overlapping subsets of genetic cues. Further, we provide evidence for an emerging model suggesting that subpopulations with disparate genetic requirements exist within neural crest sublineages.
Sublineage-specific genetic requirements within the neural crest
Our analysis of enz mutant embryos revealed that enz function is required
for melanophore, xanthophore and iridiphore differentiation, while neuronal, glial
and ectomesenchymal derivatives of the neural crest appear to develop normally.
Together with our finding that the enz mutation acts cell autonomously with
202 respect to melanophore development, this suggests that the enz locus does not
encode a gene generally required by neural crest-derived cells, but rather a molecule intrinsic to pigment cells themselves. We demonstrated that the early neural crest and chromatophore precursor populations in enz homozygotes are indistinguishable from those in wild-type siblings. Additionally, melanophores begin to differentiate at the same time in wild-type and enz mutant embryos. enz melanophores subsequently undergo a cell morphology change to a punctate form, while wild-type melanophores remain stellate. Thus, it appears that enz is required specifically for chromatophore lineages, and relatively late in their ontogeny. Further investigation of the role of enz in pigment cell sublineages will be facilitated by molecular identification of the enz locus.
Many genes, including foxd3, tfap2α, sox9, sox10 and snail/slug family
members, have been described as being expressed in the neural folds and are
used as diagnostic of neural crest induction in the embryo (see Huang and Saint-
Jeannet, 2004; Knecht and Bronner-Fraser, 2002). Our data indicate that at very
early stages of neural crest formation, foxd3 is required for the development of
specific sublineages of the neural crest, including peripheral glia, and sensory
and sympathetic neurons. On the other hand, the larval pigment pattern of
foxd3sym1 mutants was nearly identical to wild-type siblings by 4 dpf, indicating that foxd3 function is dispensable for development of crest-derived chromatophore sublineages. We also showed that expression of zash1a (Allende
and Weinberg, 1994), indicative of early sympathetic neuroblasts, was absent,
while expression of mitfa (Lister et al., 1999), the earliest known gene expressed
203 in melanoblasts, is relatively normal. Thus, foxd3 appears to be required by
specific sublineages prior to their specification within the early neural crest, and
may, in fact, play a role in these fate choices. An increasing number of studies
support the notion of early segregation of sublineages within the premigratory or early migrating neural crest, particularly between neuronal and chromatophore cell types (Frank and Sanes, 1991; Henion and Weston, 1997; Luo et al., 2003a;
Perez et al., 1999; Reedy et al., 1998; Vogel and Weston, 1988; Wilson et al.,
2004), however this is among the first to identify a potential molecular player upstream of sublineage-specific genes like c-kit and trkC. Our findings are not unprecedented, however, as sox10 has been proposed to be required for the formation of nonectomesenchymal sublineages, such as chromatophores, neurons and glia, but dispensable for development of ectomesenchymal derivatives like craniofacial cartilage (Kelsh and Eisen, 2000). Interestingly, our data suggest that foxd3 may act upstream of sox10, as sox10 expression is reduced from very early somitogenesis stages in foxd3sym1 mutant embryos.
Notably, although foxd3, tfap2α, sox9, sox10 and snail/slug have been
studied extensively with respect to neural crest development, none of these
transcription factors alone seems to be necessary for the development of all
neural crest derivatives. Thus, it is possible that these genes are either
expressed by subsets of neural crest cells, or are expressed by all but only
required by some. These alternatives should be explored through simultaneous
examination of transcription factor expression, gain- and loss-of-function studies,
and by lineage analysis in live embryos.
204
Genetic heterogeneity within distinct neural crest sublineages
We showed that crest-derived ectomesenchyme is reduced in foxd3sym1 mutant embryos, but this is particularly true with respect to third pharyngeal arch structures, which are largely absent. Thus, foxd3 function is required by some, but not all, craniofacial cartilage precursors. We also found that chromatophore development is delayed in foxd3sym1 mutant embryos compared to wild-type
siblings. We demonstrated that reducing sox10 expression in a foxd3sym1 mutant
background diminishes the complement of chromatophores that develop to a
greater extent than in wild-type embryos. Similar results have been observed for
another transcription factor, tfap2α, which is expressed in premigratory neural crest cells as well as epidermis (Barrallo-Gimeno et al., 2004; Knight et al., 2003;
O'Brien et al., 2004). tfap2α mutants low and mob exhibit partial reductions in
crest-derived cartilages and chromatophores (Barrallo-Gimeno et al., 2004;
Knight et al., 2003). We found that abrogation of foxd3 and tfap2α together
results in complete loss of pigment cells and nearly all craniofacial cartilages, as
only the posterior neurocranium is present. Further, partial and complete loss of
sox10 expression in foxd3sym1 mutants and foxd3sym1-tfap2α mutant-morphants,
respectively, indicates that these genes are required upstream of sox10 for the
specification of chromatophore sublineages within the neural crest. Loss of dlx2
expression in foxd3sym1 -tfap2α mutant-morphants suggests similar early roles for
foxd3 and tfap2α in craniofacial cartilage development. These results are
consistent with the notion that within different neural crest sublineages,
205 subpopulations having distinct genetic requirements are present at early stages
of neural crest development, for instance in the premigratory neural crest.
Similarly, our analysis of trpm7/tct mutant embryos revealed that while the majority of neural crest-derived melanophores were absent in homozygotes for severe alleles, some melanophores did differentiate, albeit with aberrant cellular
morphology. Thus, some melanophores appear to require trpm7 function
absolutely whereas others develop in at least a partially trpm7-independent
fashion. We also showed that the requirement for trpm7 function occurs relatively
late in the development of the melanogenic sublineage, as expression of genes
diagnostic of early neural crest cells and melanophore precursors is normal
through 24 hpf. These data are consistent with analyses of other zebrafish
pigment mutants, such as c-kit/sparse (Parichy et al., 1999). Even in
homozygotes for presumptive null alleles of c-kit/sparse, more than half the wild-
type complement of melanophores is present (Parichy et al., 1999; Rawls and
Johnson, 2001). The preponderance of murine spotting loci in which some, but
not all, melanophores/melanocytes fail to differentiate is also consistent with the
idea of distinct genetic requirements within the melanocyte sublineage (Bennett
and Lamoreux, 2003; Nakamura et al., 2002; Silvers, 1979). It is possible,
however, that even the severe alleles of trpm7 described here are hypomorphic
rather than amorphic, and thus it cannot be ruled out that all melanophores
require trpm7 function. Nonetheless, the existence of growing numbers of
characterized mutant loci and identified genes that affect subsets of
chromatophores or other neural crest derivatives (Ernfors, 2001; Kelsh and
206 Raible, 2002; Knight et al., 2003; Yan et al., 2005) support an emerging model in which genetically distinct subpopulations exist within these sublineages.
Further elucidation of disparate genetic requirements at both early and later stages of neural crest development will be facilitated by examination of sublineage development in double and triple mutants, and mutant-morphants.
207
LIST OF REFERENCES
Akimenko, M.A., Ekker, M., Wegner, J., Lin, W. and Westerfield, M. (1994) Combinatorial expression of three zebrafish genes related to distal-less: part of a homeobox gene code for the head. J Neurosci 14, 3475-86.
Akimenko, M.A., Johnson, S.L., Westerfield, M. and Ekker, M. (1995) Differential induction of four msx homeobox genes during fin development and regeneration in zebrafish. Development 121, 347-57.
Akira, E. and Ide, H. (1987) Differentiation of neural crest cells of Xenopus laevis in clonal culture. Pigment Cell Res 1, 28-36.
Akitaya, T. and Bronner-Fraser, M. (1992) Expression of cell adhesion molecules during initiation and cessation of neural crest cell migration. Dev Dyn 194, 12-20.
Allende, M.L. and Weinberg, E.S. (1994) The expression pattern of two zebrafish achaete-scute homolog (ash) genes is altered in the embryonic brain of the cyclops mutant. Dev Biol 166, 509-30.
Amiel, J. and Lyonnet, S. (2001) Hirschsprung disease, associated syndromes, and genetics: a review. J Med Genet 38, 729-39.
Amsterdam, A., Nissen, R.M., Sun, Z., Swindell, E.C., Farrington, S. and Hopkins, N. (2004) Identification of 315 genes essential for early zebrafish development. Proc Natl Acad Sci U S A 101, 12792-7.
An, M., Luo, R. and Henion, P.D. (2002) Differentiation and maturation of zebrafish dorsal root and sympathetic ganglion neurons. J Comp Neurol 446, 267-75.
Anderson, D.J. (2000) Genes, lineages and the neural crest: a speculative review. Philos Trans R Soc Lond B Biol Sci 355, 953-64.
208 Aoki, Y., Saint-Germain, N., Gyda, M., Magner-Fink, E., Lee, Y.H., Credidio, C. and Saint-Jeannet, J.P. (2003) Sox10 regulates the development of neural crest-derived melanocytes in Xenopus. Dev Biol 259, 19-33.
Appel, B., Korzh, V., Glasgow, E., Thor, S., Edlund, T., Dawid, I.B. and Eisen, J.S. (1995) Motoneuron fate specification revealed by patterned LIM homeobox gene expression in embryonic zebrafish. Development 121, 4117-25.
Arduini, B.L. and Henion, P.D. (2004) Melanophore sublineage-specific requirement for zebrafish touchtone during neural crest development. Mech Dev 121, 1353-64.
Artinger, K.B., Chitnis, A.B., Mercola, M. and Driever, W. (1999) Zebrafish narrowminded suggests a genetic link between formation of neural crest and primary sensory neurons. Development 126, 3969-79.
Aybar, M.J. and Mayor, R. (2002) Early induction of neural crest cells: lessons learned from frog, fish and chick. Curr Opin Genet Dev 12, 452-8.
Aybar, M.J., Nieto, M.A. and Mayor, R. (2003) Snail precedes slug in the genetic cascade required for the specification and migration of the Xenopus neural crest. Development 130, 483-94.
Bagnara, J.T. (1998) Comparative anatomy and physiology of pigment cells in nonmammalian tissues. Oxford University Press, New York, The Pigmentary System : Physiology and Pathophysiology, pp. Chap.2, pp. 9 - 40.
Bagnara, J.T., Hadley, M.E. (1973) Chromatophores and color change. Prentice- Hall, Englewood Cliffs, NJ.
Baker, B.I., Bird, D.J. and Buckingham, J.C. (1986) Effects of chronic administration of melanin-concentrating hormone on corticotrophin, melanotrophin, and pigmentation in the trout. Gen Comp Endocrinol 63, 62-9.
209 Baker, C.V., Bronner-Fraser, M., Le Douarin, N.M. and Teillet, M.A. (1997) Early- and late-migrating cranial neural crest cell populations have equivalent developmental potential in vivo. Development 124, 3077-87.
Barber, L.D., Baker, B.I., Penny, J.C. and Eberle, A.N. (1987) Melanin concentrating hormone inhibits the release of alpha MSH from teleost pituitary glands. Gen Comp Endocrinol 65, 79-86.
Baroffio, A., Dupin, E. and Le Douarin, N.M. (1988) Clone-forming ability and differentiation potential of migratory neural crest cells. Proc Natl Acad Sci U S A 85, 5325-9.
Barrallo-Gimeno, A., Holzschuh, J., Driever, W. and Knapik, E.W. (2004) Neural crest survival and differentiation in zebrafish depends on mont blanc/tfap2a gene function. Development 131, 1463-77.
Baynash, A.G., Hosoda, K., Giaid, A., Richardson, J.A., Emoto, N., Hammer, R.E. and Yanagisawa, M. (1994) Interaction of endothelin-3 with endothelin-B receptor is essential for development of epidermal melanocytes and enteric neurons. Cell 79, 1277-85.
Bennett, D.C. and Lamoreux, M.L. (2003) The color loci of mice--a genetic century. Pigment Cell Res 16, 333-44.
Bird, N.C. and Mabee, P.M. (2003) Developmental morphology of the axial skeleton of the zebrafish, Danio rerio (Ostariophysi: Cyprinidae). Dev Dyn 228, 337-57.
Birren, S.J., Lo, L. and Anderson, D.J. (1993) Sympathetic neuroblasts undergo a developmental switch in trophic dependence. Development 119, 597- 610.
Bisgrove, B.W., Raible, D.W., Walter, V., Eisen, J.S. and Grunwald, D.J. (1997) Expression of c-ret in the zebrafish embryo: potential roles in motoneuronal development. J Neurobiol 33, 749-68.
Bolande, R.P. (1997) Neurocristopathy: its growth and development in 20 years. Pediatr Pathol Lab Med 17, 1-25.
210 Bondurand, N., Pingault, V., Goerich, D.E., Lemort, N., Sock, E., Caignec, C.L., Wegner, M. and Goossens, M. (2000) Interaction among SOX10, PAX3 and MITF, three genes altered in Waardenburg syndrome. Hum Mol Genet 9, 1907-17.
Borello, U., Coletta, M., Tajbakhsh, S., Leyns, L., De Robertis, E.M., Buckingham, M. and Cossu, G. (1999) Transplacental delivery of the Wnt antagonist Frzb1 inhibits development of caudal paraxial mesoderm and skeletal myogenesis in mouse embryos. Development 126, 4247-55.
Bouillet, P., Oulad-Abdelghani, M., Ward, S.J., Bronner, S., Chambon, P. and Dolle, P. (1996) A new mouse member of the Wnt gene family, mWnt-8, is expressed during early embryogenesis and is ectopically induced by retinoic acid. Mech Dev 58, 141-52.
Brault, V., Moore, R., Kutsch, S., Ishibashi, M., Rowitch, D.H., McMahon, A.P., Sommer, L., Boussadia, O. and Kemler, R. (2001) Inactivation of the beta- catenin gene by Wnt1-Cre-mediated deletion results in dramatic brain malformation and failure of craniofacial development. Development 128, 1253-64.
Brewer, S., Feng, W., Huang, J., Sullivan, S. and Williams, T. (2004) Wnt1-Cre- mediated deletion of AP-2alpha causes multiple neural crest-related defects. Dev Biol 267, 135-52.
Bronner-Fraser, M. and Fraser, S.E. (1988) Cell lineage analysis reveals multipotency of some avian neural crest cells. Nature 335, 161-4.
Burstyn-Cohen, T., Stanleigh, J., Sela-Donenfeld, D. and Kalcheim, C. (2004) Canonical Wnt activity regulates trunk neural crest delamination linking BMP/noggin signaling with G1/S transition. Development 131, 5327-39.
Busca, R. and Ballotti, R. (2000) Cyclic AMP a key messenger in the regulation of skin pigmentation. Pigment Cell Res 13, 60-9.
Camp, E. and Lardelli, M. (2001) Tyrosinase gene expression in zebrafish embryos. Dev Genes Evol 211, 150-3.
211 Carl, T.F., Dufton, C., Hanken, J. and Klymkowsky, M.W. (1999) Inhibition of neural crest migration in Xenopus using antisense slug RNA. Dev Biol 213, 101-15.
Cauthen, C.A., Berdougo, E., Sandler, J. and Burrus, L.W. (2001) Comparative analysis of the expression patterns of Wnts and Frizzleds during early myogenesis in chick embryos. Mech Dev 104, 133-8.
Chang, C. and Hemmati-Brivanlou, A. (1998) Neural crest induction by Xwnt7B in Xenopus. Dev Biol 194, 129-34.
Cheung, M., Chaboissier, M.C., Mynett, A., Hirst, E., Schedl, A. and Briscoe, J. (2005) The transcriptional control of trunk neural crest induction, survival, and delamination. Dev Cell 8, 179-92.
Chiang, E.F., Pai, C.I., Wyatt, M., Yan, Y.L., Postlethwait, J. and Chung, B. (2001) Two sox9 genes on duplicated zebrafish chromosomes: expression of similar transcription activators in distinct sites. Dev Biol 231, 149-63.
Ciment, G. and Weston, J.A. (1982) Early appearance in neural crest and crest- derived cells of an antigenic determinant present in avian neurons. Dev Biol 93, 355-67.
Cole, L.K. and Ross, L.S. (2001) Apoptosis in the developing zebrafish embryo. Dev Biol 240, 123-42.
Collazo, A., Bronner-Fraser, M. and Fraser, S.E. (1993) Vital dye labelling of Xenopus laevis trunk neural crest reveals multipotency and novel pathways of migration. Development 118, 363-76.
Corey, D.P., Garcia-Anoveros, J., Holt, J.R., Kwan, K.Y., Lin, S.Y., Vollrath, M.A., Amalfitano, A., Cheung, E.L., Derfler, B.H., Duggan, A., Geleoc, G.S., Gray, P.A., Hoffman, M.P., Rehm, H.L., Tamasauskas, D. and Zhang, D.S. (2004) TRPA1 is a candidate for the mechanosensitive transduction channel of vertebrate hair cells. Nature 432, 723-30.
Cornell, R.A. and Eisen, J.S. (2000) Delta signaling mediates segregation of neural crest and spinal sensory neurons from zebrafish lateral neural plate. Development 127, 2873-82. 212 Cornell, R.A., Yemm, E., Bonde, G., Li, W., d'Alencon, C., Wegman, L., Eisen, J. and Zahs, A. (2004) Touchtone promotes survival of embryonic melanophores in zebrafish. Mech Dev 121, 1365-76.
Cubbage, C.C., and Mabee, P.M. (1996) Development of the cranium and paired fins in the zebrafish Danio rerio (ostariophysi, Cyprinidae). J Morphol 229, 121-160.
De Calisto, J., Araya, C., Marchant, L., Riaz, C.F. and Mayor, R. (2005) Essential role of non-canonical Wnt signalling in neural crest migration. Development 132, 2587-97.
Deardorff, M.A., Tan, C., Saint-Jeannet, J.P. and Klein, P.S. (2001) A role for frizzled 3 in neural crest development. Development 128, 3655-63.
DiCicco-Bloom, E., Friedman, W.J. and Black, I.B. (1993) NT-3 stimulates sympathetic neuroblast proliferation by promoting precursor survival. Neuron 11, 1101-11.
Dorsky, R.I., Moon, R.T. and Raible, D.W. (1998) Control of neural crest cell fate by the Wnt signalling pathway. Nature 396, 370-3.
Dorsky, R.I., Raible, D.W. and Moon, R.T. (2000) Direct regulation of nacre, a zebrafish MITF homolog required for pigment cell formation, by the Wnt pathway. Genes Dev 14, 158-62.
Dottori, M., Gross, M.K., Labosky, P. and Goulding, M. (2001) The winged-helix transcription factor Foxd3 suppresses interneuron differentiation and promotes neural crest cell fate. Development 128, 4127-38.
Draper, B.W., Morcos, P.A. and Kimmel, C.B. (2001) Inhibition of zebrafish fgf8 pre-mRNA splicing with morpholino oligos: a quantifiable method for gene knockdown. Genesis 30, 154-6.
Duncan, L.M., Deeds, J., Hunter, J., Shao, J., Holmgren, L.M., Woolf, E.A., Tepper, R.I. and Shyjan, A.W. (1998) Down-regulation of the novel gene melastatin correlates with potential for melanoma metastasis. Cancer Res 58, 1515-20.
213 Dunn, K.J., Brady, M., Ochsenbauer-Jambor, C., Snyder, S., Incao, A. and Pavan, W.J. (2005) WNT1 and WNT3a promote expansion of melanocytes through distinct modes of action. Pigment Cell Res 18, 167- 80.
Dunn, K.J., Williams, B.O., Li, Y. and Pavan, W.J. (2000) Neural crest-directed gene transfer demonstrates Wnt1 role in melanocyte expansion and differentiation during mouse development. Proc Natl Acad Sci U S A 97, 10050-5.
Dupin, E., Baroffio, A., Dulac, C., Cameron-Curry, P. and Le Douarin, N.M. (1990) Schwann-cell differentiation in clonal cultures of the neural crest, as evidenced by the anti-Schwann cell myelin protein monoclonal antibody. Proc Natl Acad Sci U S A 87, 1119-23.
Dupin, E., Glavieux, C., Vaigot, P. and Le Douarin, N.M. (2000) Endothelin 3 induces the reversion of melanocytes to glia through a neural crest- derived glial-melanocytic progenitor. Proc Natl Acad Sci U S A 97, 7882-7.
Dupin, E. and Le Douarin, N.M. (2003) Development of melanocyte precursors from the vertebrate neural crest. Oncogene 22, 3016-23.
Dutton, K., Dutton, J.R., Pauliny, A. and Kelsh, R.N. (2001a) A morpholino phenocopy of the colourless mutant. Genesis 30, 188-9.
Dutton, K.A., Pauliny, A., Lopes, S.S., Elworthy, S., Carney, T.J., Rauch, J., Geisler, R., Haffter, P. and Kelsh, R.N. (2001b) Zebrafish colourless encodes sox10 and specifies non-ectomesenchymal neural crest fates. Development 128, 4113-25.
Dutton, K.A., Pauliny, A., Lopes, S.S., Elworthy, S., Carney, T.J., Rauch, J., Geisler, R., Haffter, P. and Kelsh, R.N. (2001c) Zebrafish colourless encodes sox10 and specifies non-ectomesenchymal neural crest fates. Development 128, 4113-25.
Dvorak, M.M. and Riccardi, D. (2004) Ca2+ as an extracellular signal in bone. Cell Calcium 35, 249-55.
214 Eisen, J.S. and Weston, J.A. (1993) Development of the neural crest in the zebrafish. Dev Biol 159, 50-9.
Elizondo, M.R., Arduini, B.L., Paulsen, J., MacDonald, E.L., Sabel, J.L., Henion, P.D., Cornell, R.A. and Parichy, D.M. (2005) Defective skeletogenesis with kidney stone formation in dwarf zebrafish mutant for trpm7. Curr Biol 15, 667-71.
Elworthy, S., Lister, J.A., Carney, T.J., Raible, D.W. and Kelsh, R.N. (2003) Transcriptional regulation of mitfa accounts for the sox10 requirement in zebrafish melanophore development. Development 130, 2809-2818.
Epperlein, H., Meulemans, D., Bronner-Fraser, M., Steinbeisser, H. and Selleck, M.A. (2000) Analysis of cranial neural crest migratory pathways in axolotl using cell markers and transplantation. Development 127, 2751-61.
Epperlein, H.H. and Lofberg, J. (1990) The development of the larval pigment patterns in Triturus alpestris and Ambystoma mexicanum. Adv Anat Embryol Cell Biol 118, 1-99.
Erickson, C.A. (1987) Behavior of neural crest cells on embryonic basal laminae. Dev Biol 120, 38-49.
Erickson, C.A. and Goins, T.L. (1995) Avian neural crest cells can migrate in the dorsolateral path only if they are specified as melanocytes. Development 121, 915-24.
Ernfors, P. (2001) Local and target-derived actions of neurotrophins during peripheral nervous system development. Cell Mol Life Sci 58, 1036-44.
Fisher, S., Jagadeeswaran, P. and Halpern, M.E. (2003) Radiographic analysis of zebrafish skeletal defects. Dev Biol 264, 64-76.
Fleming, A., Keynes, R. and Tannahill, D. (2004) A central role for the notochord in vertebral patterning. Development 131, 873-80.
Fode, C., Gradwohl, G., Morin, X., Dierich, A., LeMeur, M., Goridis, C. and Guillemot, F. (1998) The bHLH protein NEUROGENIN 2 is a 215 determination factor for epibranchial placode-derived sensory neurons. Neuron 20, 483-94.
Francis, N.J. and Landis, S.C. (1999) Cellular and molecular determinants of sympathetic neuron development. Annu Rev Neurosci 22, 541-66.
Frank, E. and Sanes, J.R. (1991) Lineage of neurons and glia in chick dorsal root ganglia: analysis in vivo with a recombinant retrovirus. Development 111, 895-908.
Frost, S.K., Robinson, S.J., Carson, M.K., Thorsteinsdottir, S. and Giesler, J. (1987) Effects of exogenous guanosine on chromatophore differentiation in the axolotl. Pigment Cell Res 1, 37-43.
Fukamachi, S., Sugimoto, M., Mitani, H. and Shima, A. (2004) Somatolactin selectively regulates proliferation and morphogenesis of neural-crest derived pigment cells in medaka. Proc Natl Acad Sci U S A 101, 10661-6.
Gajavelli, S., Wood, P.M., Pennica, D., Whittemore, S.R. and Tsoulfas, P. (2004) BMP signaling initiates a neural crest differentiation program in embryonic rat CNS stem cells. Exp Neurol 188, 205-23.
Gammill, L.S. and Bronner-Fraser, M. (2002) Genomic analysis of neural crest induction. Development 129, 5731-41.
Gammill, L.S. and Bronner-Fraser, M. (2003) Neural crest specification: migrating into genomics. Nat Rev Neurosci 4, 795-805.
Gans, C. and Northcutt, R.G. (1983) Nerual Crest and the Origin of Vertebrates: A New Head. Science 220, 268-274.
Garcia-Castro, M.I., Marcelle, C. and Bronner-Fraser, M. (2002) Ectodermal Wnt function as a neural crest inducer. Science 297, 848-51.
Gilbert, S.F., 1949- (c2000) Developmental biology / Scott F. Gilbert ; with a chapter on plant development by Susan R. Singer. Sinauer Associates,, Sunderland, Mass. :, pp. xviii, 749 p. :.
216 Gilmour, D.T., Maischein, H.M. and Nusslein-Volhard, C. (2002) Migration and function of a glial subtype in the vertebrate peripheral nervous system. Neuron 34, 577-88.
Girdlestone, J. and Weston, J.A. (1985) Identification of early neuronal subpopulations in avian neural crest cell cultures. Dev Biol 109, 274-87.
Golling, G., Amsterdam, A., Sun, Z., Antonelli, M., Maldonado, E., Chen, W., Burgess, S., Haldi, M., Artzt, K., Farrington, S., Lin, S.Y., Nissen, R.M. and Hopkins, N. (2002) Insertional mutagenesis in zebrafish rapidly identifies genes essential for early vertebrate development. Nat Genet 31, 135-40.
Green, J.A., Baker, B.I. and Kawauchi, H. (1991) The effect of rearing rainbow trout on black or white backgrounds on their secretion of melanin- concentrating hormone and their sensitivity to stress. J Endocrinol 128, 267-74.
Gross, J.B. and Hanken, J. (2005) Cranial neural crest contributes to the bony skull vault in adult Xenopus laevis: Insights from cell labeling studies. J Exp Zoolog B Mol Dev Evol 304, 169-76.
Halaban, R. (2000) The regulation of normal melanocyte proliferation. Pigment Cell Res 13, 4-14.
Hall, C.M. (2002) International nosology and classification of constitutional disorders of bone (2001). Am J Med Genet 113, 65-77.
Hanna, L.A., Foreman, R.K., Tarasenko, I.A., Kessler, D.S. and Labosky, P.A. (2002) Requirement for Foxd3 in maintaining pluripotent cells of the early mouse embryo. Genes Dev 16, 2650-61.
Hari, L., Brault, V., Kleber, M., Lee, H.Y., Ille, F., Leimeroth, R., Paratore, C., Suter, U., Kemler, R. and Sommer, L. (2002) Lineage-specific requirements of beta-catenin in neural crest development. J Cell Biol 159, 867-80.
Hellqvist, M., Mahlapuu, M., Blixt, A., Enerback, S. and Carlsson, P. (1998) The human forkhead protein FREAC-2 contains two functionally redundant
217 activation domains and interacts with TBP and TFIIB. J Biol Chem 273, 23335-43.
Hemesath, T.J., Price, E.R., Takemoto, C., Badalian, T. and Fisher, D.E. (1998) MAP kinase links the transcription factor Microphthalmia to c-Kit signalling in melanocytes. Nature 391, 298-301.
Henion, P.D., Garner, A.S., Large, T.H. and Weston, J.A. (1995) trkC-mediated NT-3 signaling is required for the early development of a subpopulation of neurogenic neural crest cells. Dev Biol 172, 602-13.
Henion, P.D., Raible, D.W., Beattie, C.E., Stoesser, K.L., Weston, J.A. and Eisen, J.S. (1996) Screen for mutations affecting development of Zebrafish neural crest. Dev Genet 18, 11-7.
Henion, P.D. and Weston, J.A. (1997) Timing and pattern of cell fate restrictions in the neural crest lineage. Development 124, 4351-9.
Herbarth, B., Pingault, V., Bondurand, N., Kuhlbrodt, K., Hermans-Borgmeyer, I., Puliti, A., Lemort, N., Goossens, M. and Wegner, M. (1998) Mutation of the Sry-related Sox10 gene in Dominant megacolon, a mouse model for human Hirschsprung disease. Proc Natl Acad Sci U S A 95, 5161-5.
Herbomel, P., Thisse, B. and Thisse, C. (1999) Ontogeny and behaviour of early macrophages in the zebrafish embryo. Development 126, 3735-45.
Herbomel, P., Thisse, B. and Thisse, C. (2001) Zebrafish early macrophages colonize cephalic mesenchyme and developing brain, retina, and epidermis through a M-CSF receptor-dependent invasive process. Dev Biol 238, 274-88.
His, W. (1868) Untersuchungen uber die erste Anlage des Wirbeltierleibes. Die erste Entwicklung des Huhnchens im Ei. F.C.W. Vogel, Leipzig.
Ho, R.K. and Kane, D.A. (1990) Cell-autonomous action of zebrafish spt-1 mutation in specific mesodermal precursors. Nature 348, 728-30.
218 Hogben, L.T., Winton, F.R. (1922) The pigmentary effector system. I. Reaction of frog's melanophores to pituitary extracts. Proc Roy Soc B93, 318-329.
Honore, S.M., Aybar, M.J. and Mayor, R. (2003) Sox10 is required for the early development of the prospective neural crest in Xenopus embryos. Dev Biol 260, 79-96.
Hornyak, T.J., Hayes, D.J., Chiu, L.Y. and Ziff, E.B. (2001) Transcription factors in melanocyte development: distinct roles for Pax- 3 and Mitf. Mech Dev 101, 47-59.
Huang, X. and Saint-Jeannet, J.P. (2004) Induction of the neural crest and the opportunities of life on the edge. Dev Biol 275, 1-11.
Huber, W.E., Price, E.R., Widlund, H.R., Du, J., Davis, I.J., Wegner, M. and Fisher, D.E. (2003) A tissue-restricted cAMP transcriptional response: SOX10 modulates alpha-melanocyte-stimulating hormone-triggered expression of microphthalmia-associated transcription factor in melanocytes. J Biol Chem 278, 45224-30.
Hume, C.R. and Dodd, J. (1993) Cwnt-8C: a novel Wnt gene with a potential role in primitive streak formation and hindbrain organization. Development 119, 1147-60.
Ikeya, M., Lee, S.M., Johnson, J.E., McMahon, A.P. and Takada, S. (1997) Wnt signalling required for expansion of neural crest and CNS progenitors. Nature 389, 966-70.
Ito, K., Morita, T. and Sieber-Blum, M. (1993) In vitro clonal analysis of mouse neural crest development. Dev Biol 157, 517-25.
Ito, M., Kawa, Y., Ono, H., Okura, M., Baba, T., Kubota, Y., Nishikawa, S.I. and Mizoguchi, M. (1999) Removal of stem cell factor or addition of monoclonal anti-c-KIT antibody induces apoptosis in murine melanocyte precursors. J Invest Dermatol 112, 796-801.
Jackson, I.J. (1997) Homologous pigmentation mutations in human, mouse and other model organisms. Hum Mol Genet 6, 1613-24.
219 Jackson, I.J., Chambers, D., Rinchik, E.M. and Bennett, D.C. (1990) Characterization of TRP-1 mRNA levels in dominant and recessive mutations at the mouse brown (b) locus. Genetics 126, 451-9.
Jin, E.J., Erickson, C.A., Takada, S. and Burrus, L.W. (2001) Wnt and BMP signaling govern lineage segregation of melanocytes in the avian embryo. Dev Biol 233, 22-37.
Johnson, S.L., Africa, D., Walker, C. and Weston, J.A. (1995) Genetic control of adult pigment stripe development in zebrafish. Dev Biol 167, 27-33.
Jordan, S.A. and Jackson, I.J. (2000) MGF (KIT ligand) is a chemokinetic factor for melanoblast migration into hair follicles. Dev Biol 225, 424-36.
Kalcheim, C. and Leviel, V. (1988) Stimulation of collagen production in vitro by ascorbic acid released from explants of migrating avian neural crest. Cell Differ 22, 107-14.
Karaplis, A.C., He, B., Nguyen, M.T., Young, I.D., Semeraro, D., Ozawa, H. and Amizuka, N. (1998) Inactivating mutation in the human parathyroid hormone receptor type 1 gene in Blomstrand chondrodysplasia. Endocrinology 139, 5255-8.
Karsenty, G. and Wagner, E.F. (2002) Reaching a genetic and molecular understanding of skeletal development. Dev Cell 2, 389-406.
Kawauchi, H. and Baker, B.I. (2004) Melanin-concentrating hormone signaling systems in fish. Peptides 25, 1577-84.
Kelsh, R.N., Brand, M., Jiang, Y.J., Heisenberg, C.P., Lin, S., Haffter, P., Odenthal, J., Mullins, M.C., van Eeden, F.J., Furutani-Seiki, M., Granato, M., Hammerschmidt, M., Kane, D.A., Warga, R.M., Beuchle, D., Vogelsang, L. and Nusslein-Volhard, C. (1996) Zebrafish pigmentation mutations and the processes of neural crest development. Development 123, 369-89.
Kelsh, R.N., Dutton, K., Medlin, J. and Eisen, J.S. (2000a) Expression of zebrafish fkd6 in neural crest-derived glia. Mech Dev 93, 161-4.
220 Kelsh, R.N. and Eisen, J.S. (2000) The zebrafish colourless gene regulates development of non- ectomesenchymal neural crest derivatives. Development 127, 515-25.
Kelsh, R.N. and Raible, D.W. (2002) Specification of zebrafish neural crest. Results Probl Cell Differ 40, 216-36.
Kelsh, R.N., Schmid, B. and Eisen, J.S. (2000b) Genetic analysis of melanophore development in zebrafish embryos. Dev Biol 225, 277-93.
Kimmel, C.B., Ballard, W.W., Kimmel, S.R., Ullmann, B. and Schilling, T.F. (1995) Stages of embryonic development of the zebrafish. Dev Dyn 203, 253-310.
Kimmel, C.B., Miller, C.T., Kruze, G., Ullmann, B., BreMiller, R.A., Larison, K.D. and Snyder, H.C. (1998) The shaping of pharyngeal cartilages during early development of the zebrafish. Dev Biol 203, 245-63.
Kirby, M.L. (1987) Cardiac morphogenesis--recent research advances. Pediatr Res 21, 219-24.
Kirby, M.L. (1989) Plasticity and predetermination of mesencephalic and trunk neural crest transplanted into the region of the cardiac neural crest. Dev Biol 134, 402-12.
Kirby, M.L. and Stewart, D.E. (1983) Neural crest origin of cardiac ganglion cells in the chick embryo: identification and extirpation. Dev Biol 97, 433-43.
Kirby, M.L. and Waldo, K.L. (1990) Role of neural crest in congenital heart disease. Circulation 82, 332-40.
Kitamura, K., Takiguchi-Hayashi, K., Sezaki, M., Yamamoto, H. and Takeuchi, T. (1992) Avian neural crest cells express a melanogenic trait during early migration from the neural tube: observations with the new monoclonal antibody, "MEBL-1". Development 114, 367-378.
Knapik, E.W., Goodman, A., Atkinson, O.S., Roberts, C.T., Shiozawa, M., Sim, C.U., Weksler-Zangen, S., Trolliet, M.R., Futrell, C., Innes, B.A., Koike, G., 221 McLaughlin, M.G., Pierre, L., Simon, J.S., Vilallonga, E., Roy, M., Chiang, P.W., Fishman, M.C., Driever, W. and Jacob, H.J. (1996) A reference cross DNA panel for zebrafish (Danio rerio) anchored with simple sequence length polymorphisms. Development 123, 451-60.
Knapik, E.W., Goodman, A., Ekker, M., Chevrette, M., Delgado, J., Neuhauss, S., Shimoda, N., Driever, W., Fishman, M.C. and Jacob, H.J. (1998) A microsatellite genetic linkage map for zebrafish (Danio rerio). Nat Genet 18, 338-43.
Knecht, A.K. and Bronner-Fraser, M. (2002) Induction of the neural crest: a multigene process. Nat Rev Genet 3, 453-61.
Knight, R.D., Javidan, Y., Nelson, S., Zhang, T. and Schilling, T. (2004) Skeletal and pigment cell defects in the lockjaw mutant reveal multiple roles for zebrafish tfap2a in neural crest development. Dev Dyn 229, 87-98.
Knight, R.D., Nair, S., Nelson, S.S., Afshar, A., Javidan, Y., Geisler, R., Rauch, G.J. and Schilling, T.F. (2003) lockjaw encodes a zebrafish tfap2a required for early neural crest development. Development 130, 5755-68.
Kos, R., Reedy, M.V., Johnson, R.L. and Erickson, C.A. (2001) The winged-helix transcription factor FoxD3 is important for establishing the neural crest lineage and repressing melanogenesis in avian embryos. Development 128, 1467-79.
Krishnamurthy, V.G. (1976) Cytophysiology of corpuscles of Stannius. Int Rev Cytol 46, 177-249.
Kronenberg, H.M. (2003) Developmental regulation of the growth plate. Nature 423, 332-6.
Kulesa, P., Ellies, D.L. and Trainor, P.A. (2004) Comparative analysis of neural crest cell death, migration, and function during vertebrate embryogenesis. Dev Dyn 229, 14-29.
Kumagai, T., Wada, A., Tsudzuki, M., Nishimura, M. and Kunieda, T. (1998) Nucleotide sequence of endothelin-B receptor gene reveals origin of piebald mutation in laboratory mouse. Exp Anim 47, 265-9. 222 LaBonne, C. and Bronner-Fraser, M. (1998) Neural crest induction in Xenopus: evidence for a two-signal model. Development 125, 2403-14.
LaBonne, C. and Bronner-Fraser, M. (2000) Snail-related transcriptional repressors are required in Xenopus for both the induction of the neural crest and its subsequent migration. Dev Biol 221, 195-205.
Lahav, R. (2005) Endothelin receptor B is required for the expansion of melanocyte precursors and malignant melanoma. Int J Dev Biol 49, 173- 80.
Lamborghini, J.E. (1980) Rohon-beard cells and other large neurons in Xenopus embryos originate during gastrulation. J Comp Neurol 189, 323-33.
Le Douarin, N. and Kalcheim, C. (1999) The neural crest. Cambridge University Press, Cambridge; New York, Developmental and cell biology series; 36., pp. xxiii, 445, [18] of plates.
Le Douarin, N.M., Creuzet, S., Couly, G. and Dupin, E. (2004) Neural crest cell plasticity and its limits. Development 131, 4637-50.
Le Douarin, N.M. and Teillet, M.A. (1973) The migration of neural crest cells to the wall of the digestive tract in avian embryo. J Embryol Exp Morphol 30, 31-48.
Le Douarin, N.M. and Teillet, M.A. (1974) Experimental analysis of the migration and differentiation of neuroblasts of the autonomic nervous system and of neurectodermal mesenchymal derivatives, using a biological cell marking technique. Dev Biol 41, 162-84.
Le Guyader, S. and Jesuthasan, S. (2002) Analysis of xanthophore and pterinosome biogenesis in zebrafish using methylene blue and pteridine autofluorescence. Pigment Cell Res 15, 27-31.
Le Lievre, C.S., Schweizer, G.G., Ziller, C.M. and Le Douarin, N.M. (1980) Restrictions of developmental capabilities in neural crest cell derivatives as tested by in vivo transplantation experiments. Dev Biol 77, 362-78.
223 Lee, H.O., Levorse, J.M. and Shin, M.K. (2003) The endothelin receptor-B is required for the migration of neural crest-derived melanocyte and enteric neuron precursors. Dev Biol 259, 162-75.
Liem, K.F., Jr., Tremml, G., Roelink, H. and Jessell, T.M. (1995) Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell 82, 969-79.
Lister, J. (1858) On the cutaneous pigmentary system of the frog. Philos Trans 148, 627-643.
Lister, J.A. (2002) Development of pigment cells in the zebrafish embryo. Microsc Res Tech 58, 435-41.
Lister, J.A., Robertson, C.P., Lepage, T., Johnson, S.L. and Raible, D.W. (1999) nacre encodes a zebrafish microphthalmia-related protein that regulates neural-crest-derived pigment cell fate. Development 126, 3757-67.
Lofberg, J., Nynas-McCoy, A., Olsson, C., Jonsson, L. and Perris, R. (1985) Stimulation of initial neural crest cell migration in the axolotl embryo by tissue grafts and extracellular matrix transplanted on microcarriers. Dev Biol 107, 442-59.
Luo, R., An, M., Arduini, B.L. and Henion, P.D. (2001a) Specific pan-neural crest expression of zebrafish Crestin throughout embryonic development. Dev Dyn 220, 169-74.
Luo, R., Gao, J., Wehrle-Haller, B. and Henion, P.D. (2003a) Molecular identification of distinct neurogenic and melanogenic neural crest sublineages. Development 130, 321-30.
Luo, T., Lee, Y.H., Saint-Jeannet, J.P. and Sargent, T.D. (2003b) Induction of neural crest in Xenopus by transcription factor AP2alpha. Proc Natl Acad Sci U S A 100, 532-7.
Luo, T., Matsuo-Takasaki, M., Lim, J.H. and Sargent, T.D. (2001b) Differential regulation of Dlx gene expression by a BMP morphogenetic gradient. Int J Dev Biol 45, 681-4.
224 Luo, T., Matsuo-Takasaki, M., Thomas, M.L., Weeks, D.L. and Sargent, T.D. (2002) Transcription factor AP-2 is an essential and direct regulator of epidermal development in Xenopus. Dev Biol 245, 136-44.
Ma, Q., Chen, Z., del Barco Barrantes, I., de la Pompa, J.L. and Anderson, D.J. (1998) neurogenin1 is essential for the determination of neuronal precursors for proximal cranial sensory ganglia. Neuron 20, 469-82.
Ma, Q., Fode, C., Guillemot, F. and Anderson, D.J. (1999) Neurogenin1 and neurogenin2 control two distinct waves of neurogenesis in developing dorsal root ganglia. Genes Dev 13, 1717-28.
Mabee, P.M., Olmstead, K.L. and Cubbage, C.C. (2000) An experimental study of intraspecific variation, developmental timing, and heterochrony in fishes. Evolution Int J Org Evolution 54, 2091-106.
Mancilla, A. and Mayor, R. (1996) Neural crest formation in Xenopus laevis: mechanisms of Xslug induction. Dev Biol 177, 580-9.
Marchant, L., Linker, C., Ruiz, P., Guerrero, N. and Mayor, R. (1998) The inductive properties of mesoderm suggest that the neural crest cells are specified by a BMP gradient. Dev Biol 198, 319-29.
Marshall, A.M. (1879) The morphology of the vertebrate olfactory organ. Q J Microsc Sci 19, 300-340.
Marusich, M.F., Furneaux, H.M., Henion, P.D. and Weston, J.A. (1994) Hu neuronal proteins are expressed in proliferating neurogenic cells. J Neurobiol 25, 143-55.
Matzuk, M.M., Lu, N., Vogel, H., Sellheyer, K., Roop, D.R. and Bradley, A. (1995) Multiple defects and perinatal death in mice deficient in follistatin. Nature 374, 360-3.
Mayor, R. and Aybar, M.J. (2001) Induction and development of neural crest in Xenopus laevis. Cell Tissue Res 305, 203-9.
225 Mayor, R., Guerrero, N. and Martinez, C. (1997) Role of FGF and noggin in neural crest induction. Dev Biol 189, 1-12.
Mayor, R., Morgan, R. and Sargent, M.G. (1995) Induction of the prospective neural crest of Xenopus. Development 121, 767-77.
McCallion, A.S. and Chakravarti, A. (2001) EDNRB/EDN3 and Hirschsprung disease type II. Pigment Cell Res 14, 161-9.
McGonnell, I.M. and Graham, A. (2002) Trunk neural crest has skeletogenic potential. Curr Biol 12, 767-71.
McKeown, S.J., Newgreen, D.F. and Farlie, P.G. (2005) Dlx2 over-expression regulates cell adhesion and mesenchymal condensation in ectomesenchyme. Dev Biol 281, 22-37.
McMahon, J.A., Takada, S., Zimmerman, L.B., Fan, C.M., Harland, R.M. and McMahon, A.P. (1998) Noggin-mediated antagonism of BMP signaling is required for growth and patterning of the neural tube and somite. Genes Dev 12, 1438-52.
Metcalfe, W.K., Myers, P.Z., Trevarrow, B., Bass, M.B. and Kimmel, C.B. (1990) Primary neurons that express the L2/HNK-1 carbohydrate during early development in the zebrafish. Development 110, 491-504.
Mitchell, P.J., Timmons, P.M., Hebert, J.M., Rigby, P.W. and Tjian, R. (1991) Transcription factor AP-2 is expressed in neural crest cell lineages during mouse embryogenesis. Genes Dev 5, 105-19.
Mollaaghababa, R. and Pavan, W.J. (2003) The importance of having your SOX on: role of SOX10 in the development of neural crest-derived melanocytes and glia. Oncogene 22, 3024-34.
Monsoro-Burq, A.H., Fletcher, R.B. and Harland, R.M. (2003) Neural crest induction by paraxial mesoderm in Xenopus embryos requires FGF signals. Development 130, 3111-24.
226 Monteilh-Zoller, M.K., Hermosura, M.C., Nadler, M.J., Scharenberg, A.M., Penner, R. and Fleig, A. (2003) TRPM7 provides an ion channel mechanism for cellular entry of trace metal ions. J Gen Physiol 121, 49- 60.
Morrison, S.J., Perez, S.E., Qiao, Z., Verdi, J.M., Hicks, C., Weinmaster, G. and Anderson, D.J. (2000) Transient Notch activation initiates an irreversible switch from neurogenesis to gliogenesis by neural crest stem cells. Cell 101, 499-510.
Morrison-Graham, K. and Weston, J.A. (1993) Transient steel factor dependence by neural crest-derived melanocyte precursors. Dev Biol 159, 346-52.
Moury, J.D. and Jacobson, A.G. (1989) Neural fold formation at newly created boundaries between neural plate and epidermis in the axolotl. Dev Biol 133, 44-57.
Moury, J.D. and Jacobson, A.G. (1990) The origins of neural crest cells in the axolotl. Dev Biol 141, 243-53.
Nagatsu, T. and Ichinose, H. (1999) Regulation of pteridine-requiring enzymes by the cofactor tetrahydrobiopterin. Mol Neurobiol 19, 79-96.
Nakagawa, S. and Takeichi, M. (1995) Neural crest cell-cell adhesion controlled by sequential and subpopulation-specific expression of novel cadherins. Development 121, 1321-32.
Nakagawa, S. and Takeichi, M. (1998) Neural crest emigration from the neural tube depends on regulated cadherin expression. Development 125, 2963- 71.
Nakamura, M., Tobin, D.J., Richards-Smith, B., Sundberg, J.P. and Paus, R. (2002) Mutant laboratory mice with abnormalities in pigmentation: annotated tables. J Dermatol Sci 28, 1-33.
Nakamura, T. (1995) Genetic markers and animal models of neurocristopathy. Histol Histopathol 10, 747-59.
227 Nataf, V., Lecoin, L., Eichmann, A. and Le Douarin, N.M. (1996) Endothelin-B receptor is expressed by neural crest cells in the avian embryo. Proc Natl Acad Sci U S A 93, 9645-50.
Nechiporuk, A., Finney, J.E., Keating, M.T. and Johnson, S.L. (1999) Assessment of polymorphism in zebrafish mapping strains. Genome Res 9, 1231-8.
Nemecek, E.R., Sawin, R.W. and Park, J. (2003) Treatment of neuroblastoma in patients with neurocristopathy syndromes. J Pediatr Hematol Oncol 25, 159-62.
Neuhauss, S.C., Solnica-Krezel, L., Schier, A.F., Zwartkruis, F., Stemple, D.L., Malicki, J., Abdelilah, S., Stainier, D.Y. and Driever, W. (1996) Mutations affecting craniofacial development in zebrafish. Development 123, 357-67.
Newgreen, D.F. and Minichiello, J. (1995) Control of epitheliomesenchymal transformation. I. Events in the onset of neural crest cell migration are separable and inducible by protein kinase inhibitors. Dev Biol 170, 91-101.
Nguyen, V.H., Schmid, B., Trout, J., Connors, S.A., Ekker, M. and Mullins, M.C. (1998) Ventral and lateral regions of the zebrafish gastrula, including the neural crest progenitors, are established by a bmp2b/swirl pathway of genes. Dev Biol 199, 93-110.
Nguyen, V.H., Trout, J., Connors, S.A., Andermann, P., Weinberg, E. and Mullins, M.C. (2000) Dorsal and intermediate neuronal cell types of the spinal cord are established by a BMP signaling pathway. Development 127, 1209-20.
Nieto, M.A. (2001) The early steps of neural crest development. Mech Dev 105, 27-35.
Nieto, M.A., Sargent, M.G., Wilkinson, D.G. and Cooke, J. (1994) Control of cell behavior during vertebrate development by Slug, a zinc finger gene. Science 264, 835-9.
228 Nordlund, J.J., Boissy, R.E., Hearing, V.J., King, R.A. and Ortonne, J.P. (1998) The Pigmentary System : Physiology and Pathophysiology. Oxford University Press, New York, pp. xxiv, 1025 , [48] of plates.
Oakley, R.A., Garner, A.S., Large, T.H. and Frank, E. (1995) Muscle sensory neurons require neurotrophin-3 from peripheral tissues during the period of normal cell death. Development 121, 1341-50.
O'Brien, E.K., d'Alencon, C., Bonde, G., Li, W., Schoenebeck, J., Allende, M.L., Gelb, B.D., Yelon, D., Eisen, J.S. and Cornell, R.A. (2004) Transcription factor Ap-2alpha is necessary for development of embryonic melanophores, autonomic neurons and pharyngeal skeleton in zebrafish. Dev Biol 265, 246-61.
Odenthal, J. and Nusslein-Volhard, C. (1998) fork head domain genes in zebrafish. Dev Genes Evol 208, 245-58.
Odenthal, J., Rossnagel, K., Haffter, P., Kelsh, R.N., Vogelsang, E., Brand, M., van Eeden, F.J., Furutani-Seiki, M., Granato, M., Hammerschmidt, M., Heisenberg, C.P., Jiang, Y.J., Kane, D.A., Mullins, M.C. and Nusslein- Volhard, C. (1996) Mutations affecting xanthophore pigmentation in the zebrafish, Danio rerio. Development 123, 391-8.
O'Donnell, J.M., McLean, J.R. and Reynolds, E.R. (1989) Molecular and developmental genetics of the Punch locus, a pterin biosynthesis gene in Drosophila melanogaster. Dev Genet 10, 273-86.
Opdecamp, K., Nakayama, A., Nguyen, M.T., Hodgkinson, C.A., Pavan, W.J. and Arnheiter, H. (1997) Melanocyte development in vivo and in neural crest cell cultures: crucial dependence on the Mitf basic-helix-loop-helix- zipper transcription factor. Development 124, 2377-86.
Osborne, N.J., Begbie, J., Chilton, J.K., Schmidt, H. and Eickholt, B.J. (2005) Semaphorin/neuropilin signaling influences the positioning of migratory neural crest cells within the hindbrain region of the chick. Dev Dyn.
Pani, L., Overdier, D.G., Porcella, A., Qian, X., Lai, E. and Costa, R.H. (1992) Hepatocyte nuclear factor 3 beta contains two transcriptional activation domains, one of which is novel and conserved with the Drosophila fork head protein. Mol Cell Biol 12, 3723-32. 229 Parichy, D.M., Mellgren, E.M., Rawls, J.F., Lopes, S.S., Kelsh, R.N. and Johnson, S.L. (2000a) Mutational analysis of endothelin receptor b1 (rose) during neural crest and pigment pattern development in the zebrafish Danio rerio. Dev Biol 227, 294-306.
Parichy, D.M., Ransom, D.G., Paw, B., Zon, L.I. and Johnson, S.L. (2000b) An orthologue of the kit-related gene fms is required for development of neural crest-derived xanthophores and a subpopulation of adult melanocytes in the zebrafish, Danio rerio. Development 127, 3031-44.
Parichy, D.M., Rawls, J.F., Pratt, S.J., Whitfield, T.T. and Johnson, S.L. (1999) Zebrafish sparse corresponds to an orthologue of c-kit and is required for the morphogenesis of a subpopulation of melanocytes, but is not essential for hematopoiesis or primordial germ cell development. Development 126, 3425-36.
Parichy, D.M. and Turner, J.M. (2003) Temporal and cellular requirements for Fms signaling during zebrafish adult pigment pattern development. Development 130, 817-33.
Perez, S.E., Rebelo, S. and Anderson, D.J. (1999) Early specification of sensory neuron fate revealed by expression and function of neurogenins in the chick embryo. Development 126, 1715-28.
Pickart, M.A., Sivasubbu, S., Nielsen, A.L., Shriram, S., King, R.A. and Ekker, S.C. (2004) Functional genomics tools for the analysis of zebrafish pigment. Pigment Cell Res 17, 461-70.
Pingault, V., Bondurand, N., Kuhlbrodt, K., Goerich, D.E., Prehu, M.O., Puliti, A., Herbarth, B., Hermans-Borgmeyer, I., Legius, E., Matthijs, G., Amiel, J., Lyonnet, S., Ceccherini, I., Romeo, G., Smith, J.C., Read, A.P., Wegner, M. and Goossens, M. (1998) SOX10 mutations in patients with Waardenburg-Hirschsprung disease. Nat Genet 18, 171-3.
Piotrowski, T., Schilling, T.F., Brand, M., Jiang, Y.J., Heisenberg, C.P., Beuchle, D., Grandel, H., van Eeden, F.J., Furutani-Seiki, M., Granato, M., Haffter, P., Hammerschmidt, M., Kane, D.A., Kelsh, R.N., Mullins, M.C., Odenthal, J., Warga, R.M. and Nusslein-Volhard, C. (1996) Jaw and branchial arch mutants in zebrafish II: anterior arches and cartilage differentiation. Development 123, 345-56.
230 Pla, P. and Larue, L. (2003) Involvement of endothelin receptors in normal and pathological development of neural crest cells. Int J Dev Biol 47, 315-25.
Pohl, B.S. and Knochel, W. (2001) Overexpression of the transcriptional repressor FoxD3 prevents neural crest formation in Xenopus embryos. Mech Dev 103, 93-106.
Pomeranz, H.D., Rothman, T.P. and Gershon, M.D. (1991) Colonization of the post-umbilical bowel by cells derived from the sacral neural crest: direct tracing of cell migration using an intercalating probe and a replication- deficient retrovirus. Development 111, 647-55.
Postlethwait, J.H., Johnson, S.L., Midson, C.N., Talbot, W.S., Gates, M., Ballinger, E.W., Africa, D., Andrews, R., Carl, T., Eisen, J.S. and et al. (1994) A genetic linkage map for the zebrafish. Science 264, 699-703.
Postlethwait, J.H., Yan, Y.L., Gates, M.A., Horne, S., Amores, A., Brownlie, A., Donovan, A., Egan, E.S., Force, A., Gong, Z., Goutel, C., Fritz, A., Kelsh, R., Knapik, E., Liao, E., Paw, B., Ransom, D., Singer, A., Thomson, M., Abduljabbar, T.S., Yelick, P., Beier, D., Joly, J.S., Larhammar, D., Rosa, F. and et al. (1998) Vertebrate genome evolution and the zebrafish gene map. Nat Genet 18, 345-9.
Potterf, S.B., Furumura, M., Dunn, K.J., Arnheiter, H. and Pavan, W.J. (2000) Transcription factor hierarchy in Waardenburg syndrome: regulation of MITF expression by SOX10 and PAX3. Hum Genet 107, 1-6.
Potthoff, T. (1984) Clearing and staining techniques. In Moser, H.G., Richards, W.J., Cohen, D.M., Fahay, M.P., Kendall Jr., A.W., Richardson, S.L. (ed.), Ontogeny and Systematics of Fishes, Allen Press, Lawrence, Kansas, Vol. Special Publication 1, pp. 35-37.
Price, E.R., Ding, H.F., Badalian, T., Bhattacharya, S., Takemoto, C., Yao, T.P., Hemesath, T.J. and Fisher, D.E. (1998) Lineage-specific signaling in melanocytes. C-kit stimulation recruits p300/CBP to microphthalmia. J Biol Chem 273, 17983-6.
Quigley, I.K. and Parichy, D.M. (2002) Pigment pattern formation in zebrafish: A model for developmental genetics and the evolution of form. Microsc Res Tech 58, 442-55. 231 Quigley, I.K., Turner, J.M., Nuckels, R.J., Manuel, J.L., Budi, E.H., MacDonald, E.L. and Parichy, D.M. (2004) Pigment pattern evolution by differential deployment of neural crest and post-embryonic melanophore lineages in Danio fishes. Development 131, 6053-69.
Ragland, J.W. and Raible, D.W. (2004) Signals derived from the underlying mesoderm are dispensable for zebrafish neural crest induction. Dev Biol 276, 16-30.
Raible, D.W. and Eisen, J.S. (1994) Restriction of neural crest cell fate in the trunk of the embryonic zebrafish. Development 120, 495-503.
Raible, D.W. and Eisen, J.S. (1996) Regulative interactions in zebrafish neural crest. Development 122, 501-7.
Raible, D.W., Wood, A., Hodsdon, W., Henion, P.D., Weston, J.A. and Eisen, J.S. (1992) Segregation and early dispersal of neural crest cells in the embryonic zebrafish. Dev Dyn 195, 29-42.
Rawls, J.F., Frieda, M.R., McAdow, A.R., Gross, J.P., Clayton, C.M., Heyen, C.K. and Johnson, S.L. (2003) Coupled mutagenesis screens and genetic mapping in zebrafish. Genetics 163, 997-1009.
Rawls, J.F. and Johnson, S.L. (2000) Zebrafish kit mutation reveals primary and secondary regulation of melanocyte development during fin stripe regeneration. Development 127, 3715-24.
Rawls, J.F. and Johnson, S.L. (2001) Requirements for the kit receptor tyrosine kinase during regeneration of zebrafish fin melanocytes. Development 128, 1943-9.
Rawls, J.F., Mellgren, E.M. and Johnson, S.L. (2001) How the zebrafish gets its stripes. Dev Biol 240, 301-14.
Reaume, A.G., Knecht, D.A. and Chovnick, A. (1991) The rosy locus in Drosophila melanogaster: xanthine dehydrogenase and eye pigments. Genetics 129, 1099-109.
232 Reedy, M.V., Faraco, C.D. and Erickson, C.A. (1998) The delayed entry of thoracic neural crest cells into the dorsolateral path is a consequence of the late emigration of melanogenic neural crest cells from the neural tube. Dev Biol 200, 234-46.
Rizvi, T.A., Huang, Y., Sidani, A., Atit, R., Largaespada, D.A., Boissy, R.E. and Ratner, N. (2002) A novel cytokine pathway suppresses glial cell melanogenesis after injury to adult nerve. J Neurosci 22, 9831-40.
Rubinstein, A.L., Lee, D., Luo, R., Henion, P.D. and Halpern, M.E. (2000) Genes dependent on zebrafish cyclops function identified by AFLP differential gene expression screen. Genesis 26, 86-97.
Ruffins, S., Artinger, K.B. and Bronner-Fraser, M. (1998) Early migrating neural crest cells can form ventral neural tube derivatives when challenged by transplantation. Dev Biol 203, 295-304.
Saint-Amant, L. and Drapeau, P. (1998) Time course of the development of motor behaviors in the zebrafish embryo. J Neurobiol 37, 622-32.
Saito, H., Yasumoto, K., Takeda, K., Takahashi, K., Yamamoto, H. and Shibahara, S. (2003) Microphthalmia-associated transcription factor in the Wnt signaling pathway. Pigment Cell Res 16, 261-5.
Santiago, A. and Erickson, C.A. (2002) Ephrin-B ligands play a dual role in the control of neural crest cell migration. Development 129, 3621-32.
Sasai, N., Mizuseki, K. and Sasai, Y. (2001) Requirement of FoxD3-class signaling for neural crest determination in Xenopus. Development 128, 2525-36.
Sato, M. and Yost, H.J. (2003) Cardiac neural crest contributes to cardiomyogenesis in zebrafish. Dev Biol 257, 127-39.
Savagner, P., Yamada, K.M. and Thiery, J.P. (1997) The zinc-finger protein slug causes desmosome dissociation, an initial and necessary step for growth factor-induced epithelial-mesenchymal transition. J Cell Biol 137, 1403-19.
233 Schilling, T.F. and Kimmel, C.B. (1994) Segment and cell type lineage restrictions during pharyngeal arch development in the zebrafish embryo. Development 120, 483-94.
Schilling, T.F. and Kimmel, C.B. (1997) Musculoskeletal patterning in the pharyngeal segments of the zebrafish embryo. Development 124, 2945- 60.
Schilling, T.F., Piotrowski, T., Grandel, H., Brand, M., Heisenberg, C.P., Jiang, Y.J., Beuchle, D., Hammerschmidt, M., Kane, D.A., Mullins, M.C., van Eeden, F.J., Kelsh, R.N., Furutani-Seiki, M., Granato, M., Haffter, P., Odenthal, J., Warga, R.M., Trowe, T. and Nusslein-Volhard, C. (1996) Jaw and branchial arch mutants in zebrafish I: branchial arches. Development 123, 329-44.
Schipani, E., Kruse, K. and Juppner, H. (1995) A constitutively active mutant PTH-PTHrP receptor in Jansen-type metaphyseal chondrodysplasia. Science 268, 98-100.
Schmitz, C., Perraud, A.L., Johnson, C.O., Inabe, K., Smith, M.K., Penner, R., Kurosaki, T., Fleig, A. and Scharenberg, A.M. (2003) Regulation of vertebrate cellular Mg2+ homeostasis by TRPM7. Cell 114, 191-200.
Schorle, H., Meier, P., Buchert, M., Jaenisch, R. and Mitchell, P.J. (1996) Transcription factor AP-2 essential for cranial closure and craniofacial development. Nature 381, 235-8.
Schubert, F.R., Mootoosamy, R.C., Walters, E.H., Graham, A., Tumiotto, L., Munsterberg, A.E., Lumsden, A. and Dietrich, S. (2002) Wnt6 marks sites of epithelial transformations in the chick embryo. Mech Dev 114, 143-8.
Sela-Donenfeld, D. and Kalcheim, C. (1999) Regulation of the onset of neural crest migration by coordinated activity of BMP4 and Noggin in the dorsal neural tube. Development 126, 4749-62.
Selleck, M.A. and Bronner-Fraser, M. (1995) Origins of the avian neural crest: the role of neural plate-epidermal interactions. Development 121, 525-38.
234 Selleck, M.A., Garcia-Castro, M.I., Artinger, K.B. and Bronner-Fraser, M. (1998) Effects of Shh and Noggin on neural crest formation demonstrate that BMP is required in the neural tube but not ectoderm. Development 125, 4919-30.
Serbedzija, G.N., Bronner-Fraser, M. and Fraser, S.E. (1994) Developmental potential of trunk neural crest cells in the mouse. Development 120, 1709- 18.
Serbedzija, G.N., Fraser, S.E. and Bronner-Fraser, M. (1990) Pathways of trunk neural crest cell migration in the mouse embryo as revealed by vital dye labelling. Development 108, 605-12.
Shepherd, I.T., Pietsch, J., Elworthy, S., Kelsh, R.N. and Raible, D.W. (2004) Roles for GFRalpha1 receptors in zebrafish enteric nervous system development. Development 131, 241-9.
Shi, D.L., Goisset, C. and Boucaut, J.C. (1998) Expression of Xfz3, a Xenopus frizzled family member, is restricted to the early nervous system. Mech Dev 70, 35-47.
Shin, M.K., Levorse, J.M., Ingram, R.S. and Tilghman, S.M. (1999) The temporal requirement for endothelin receptor-B signalling during neural crest development. Nature 402, 496-501.
Sidi, S., Friedrich, R.W. and Nicolson, T. (2003) NompC TRP channel required for vertebrate sensory hair cell mechanotransduction. Science 301, 96-9.
Silvers, W. (1979) The coat colors of mice : a model for mammalian gene action and interaction. Springer-Verlag, New York, pp. xiv, 379 , [2] leaves of plates.
Southard-Smith, E.M., Kos, L. and Pavan, W.J. (1998) Sox10 mutation disrupts neural crest development in Dom Hirschsprung mouse model. Nat Genet 18, 60-4.
Steel, K.P., Davidson, D.R. and Jackson, I.J. (1992) TRP-2/DT, a new early melanoblast marker, shows that steel growth factor (c-kit ligand) is a survival factor. Development 115, 1111-9. 235 Stemple, D.L. and Anderson, D.J. (1992) Isolation of a stem cell for neurons and glia from the mammalian neural crest. Cell 71, 973-85.
Stewart, R.A., Arduini, B.L., Berghmans, S., George, R.E., Kanki, J.P., Henion, P.D. and Look, A.T. (unpublished) Zebrafish foxd3 is required for neural crest survival and differentiation.
Stewart, R.A., Look, A.T., Kanki, J.P. and Henion, P.D. (2004) Development of the peripheral sympathetic nervous system in zebrafish. Methods Cell Biol 76, 237-60.
Storey, J.D. and Tibshirani, R. (2003) Statistical significance for genomewide studies. Proc Natl Acad Sci U S A 100, 9440-5.
Streisinger, G., Singer, F., Walker, C., Knauber, D. and Dower, N. (1986) Segregation analyses and gene-centromere distances in zebrafish. Genetics 112, 311-9.
Streisinger, G., Walker, C., Dower, N., Knauber, D. and Singer, F. (1981) Production of clones of homozygous diploid zebra fish (Brachydanio rerio). Nature 291, 293-6.
Streit, A. and Stern, C.D. (1999) Mesoderm patterning and somite formation during node regression: differential effects of chordin and noggin. Mech Dev 85, 85-96.
Sugimoto, M. (2002) Morphological color changes in fish: regulation of pigment cell density and morphology. Microsc Res Tech 58, 496-503.
Tachibana, M., Takeda, K., Nobukuni, Y., Urabe, K., Long, J.E., Meyers, K.A., Aaronson, S.A. and Miki, T. (1996) Ectopic expression of MITF, a gene for Waardenburg syndrome type 2, converts fibroblasts to cells with melanocyte characteristics. Nat Genet 14, 50-4.
Takeda, K., Yasumoto, K., Takada, R., Takada, S., Watanabe, K., Udono, T., Saito, H., Takahashi, K. and Shibahara, S. (2000) Induction of melanocyte-specific microphthalmia-associated transcription factor by Wnt-3a. J Biol Chem 275, 14013-6.
236 Takeda, S., Bonnamy, J.P., Owen, M.J., Ducy, P. and Karsenty, G. (2001) Continuous expression of Cbfa1 in nonhypertrophic chondrocytes uncovers its ability to induce hypertrophic chondrocyte differentiation and partially rescues Cbfa1-deficient mice. Genes Dev 15, 467-81.
Takeichi, M., Nakagawa, S., Aono, S., Usui, T. and Uemura, T. (2000) Patterning of cell assemblies regulated by adhesion receptors of the cadherin superfamily. Philos Trans R Soc Lond B Biol Sci 355, 885-90.
Thibaudeau, G. and Holder, S. (1998) Cellular plasticity among axolotl neural crest-derived pigment cell lineages. Pigment Cell Res 11, 38-44.
Thisse, C., Thisse, B. and Postlethwait, J.H. (1995) Expression of snail2, a second member of the zebrafish snail family, in cephalic mesendoderm and presumptive neural crest of wild-type and spadetail mutant embryos. Dev Biol 172, 86-99.
Thisse, C., Thisse, B., Schilling, T.F. and Postlethwait, J.H. (1993) Structure of the zebrafish snail1 gene and its expression in wild-type, spadetail and no tail mutant embryos. Development 119, 1203-15.
Tsukamoto, K., Jackson, I.J., Urabe, K., Montague, P.M. and Hearing, V.J. (1992) A second tyrosinase-related protein, TRP-2, is a melanogenic enzyme termed DOPAchrome tautomerase. Embo J 11, 519-26.
Ueta, C., Iwamoto, M., Kanatani, N., Yoshida, C., Liu, Y., Enomoto-Iwamoto, M., Ohmori, T., Enomoto, H., Nakata, K., Takada, K., Kurisu, K. and Komori, T. (2001) Skeletal malformations caused by overexpression of Cbfa1 or its dominant negative form in chondrocytes. J Cell Biol 153, 87-100.
Valinsky, J.E. and Le Douarin, N.M. (1985) Production of plasminogen activator by migrating cephalic neural crest cells. Embo J 4, 1403-6.
Van Raay, T.J., Wang, Y.K., Stark, M.R., Rasmussen, J.T., Francke, U., Vetter, M.L. and Rao, M.S. (2001) frizzled 9 is expressed in neural precursor cells in the developing neural tube. Dev Genes Evol 211, 453-7.
Vogel, K.S. and Weston, J.A. (1988) A subpopulation of cultured avian neural crest cells has transient neurogenic potential. Neuron 1, 569-77. 237 Wakamatsu, Y., Maynard, T.M. and Weston, J.A. (2000) Fate determination of neural crest cells by NOTCH-mediated lateral inhibition and asymmetrical cell division during gangliogenesis. Development 127, 2811-21.
Wawersik, S., Evola, C. and Whitman, M. (2005) Conditional BMP inhibition in Xenopus reveals stage-specific roles for BMPs in neural and neural crest induction. Dev Biol 277, 425-42.
Wehrle-Haller, B. (2003) The role of Kit-ligand in melanocyte development and epidermal homeostasis. Pigment Cell Res 16, 287-96.
Wehrle-Haller, B., Meller, M. and Weston, J.A. (2001) Analysis of melanocyte precursors in Nf1 mutants reveals that MGF/KIT signaling promotes directed cell migration independent of its function in cell survival. Dev Biol 232, 471-83.
Wehrle-Haller, B. and Weston, J.A. (1995) Soluble and cell-bound forms of steel factor activity play distinct roles in melanocyte precursor dispersal and survival on the lateral neural crest migration pathway. Development 121, 731-42.
Westerfield, M. (2000) The Zebrafish Book: A Guide for the Laboratory Use of Zebrafish (Danio rerio). University of Oregon Press, Eugene, Oregon.
Widlund, H.R. and Fisher, D.E. (2003) Microphthalamia-associated transcription factor: a critical regulator of pigment cell development and survival. Oncogene 22, 3035-41.
Wilson, P.A., Lagna, G., Suzuki, A. and Hemmati-Brivanlou, A. (1997) Concentration-dependent patterning of the Xenopus ectoderm by BMP4 and its signal transducer Smad1. Development 124, 3177-84.
Wilson, Y.M., Richards, K.L., Ford-Perriss, M.L., Panthier, J.J. and Murphy, M. (2004) Neural crest cell lineage segregation in the mouse neural tube. Development 131, 6153-62.
Winnier, G., Blessing, M., Labosky, P.A. and Hogan, B.L. (1995) Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev 9, 2105-16. 238 Wolf, F.I. (2004) TRPM7: channeling the future of cellular magnesium homeostasis? Sci STKE 2004, e23.
Wood, J.M., Schallreuter-Wood, K.U., Lindsey, N.J., Callaghan, S. and Gardner, M.L. (1995) A specific tetrahydrobiopterin binding domain on tyrosinase controls melanogenesis. Biochem Biophys Res Commun 206, 480-5.
Wu, J., Saint-Jeannet, J.P. and Klein, P.S. (2003) Wnt-frizzled signaling in neural crest formation. Trends Neurosci 26, 40-5.
Wu, M., Hemesath, T.J., Takemoto, C.M., Horstmann, M.A., Wells, A.G., Price, E.R., Fisher, D.Z. and Fisher, D.E. (2000) c-Kit triggers dual phosphorylations, which couple activation and degradation of the essential melanocyte factor Mi. Genes Dev 14, 301-12.
Xu, W., Gong, L., Haddad, M.M., Bischof, O., Campisi, J., Yeh, E.T. and Medrano, E.E. (2000) Regulation of microphthalmia-associated transcription factor MITF protein levels by association with the ubiquitin- conjugating enzyme hUBC9. Exp Cell Res 255, 135-43.
Yan, Y.L., Willoughby, J., Liu, D., Crump, J.G., Wilson, C., Miller, C.T., Singer, A., Kimmel, C., Westerfield, M. and Postlethwait, J.H. (2005) A pair of Sox: distinct and overlapping functions of zebrafish sox9 co-orthologs in craniofacial and pectoral fin development. Development 132, 1069-83.
Yoder, J.A. and Litman, G.W. (2000) The zebrafish fth1, slc3a2, men1, pc, fgf3 and cycd1 genes define two regions of conserved synteny between linkage group 7 and human chromosome 11q13. Gene 261, 235-42.
Yu, H.H. and Moens, C.B. (2005) Semaphorin signaling guides cranial neural crest cell migration in zebrafish. Dev Biol 280, 373-85.
Zhang, J., Hagopian-Donaldson, S., Serbedzija, G., Elsemore, J., Plehn- Dujowich, D., McMahon, A.P., Flavell, R.A. and Williams, T. (1996) Neural tube, skeletal and body wall defects in mice lacking transcription factor AP-2. Nature 381, 238-41.
239