THE USE OF THE SPONTANEOUS BN MOUSE MUTANT, AND TARGETED ALLELES OF SMAD2 AND TGIF TO UNDERSTAND AXIAL SPECIFICATION AND NEURAL DEVELOPMENT
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Tessa Lyn Carrel, B.S., M.S.
*****
The Ohio State University 2004
Dissertation Committee: Approved by Michael WeinsteinPh.D., Advisor
Christine Beattie, Ph.D. ______Heithem El Hodiri, Ph.D. Advisor Program in Molecular, Cellular Amanda Simcox, Ph.D. and Developmental Biology
ABSTRACT
To understand the basis of axial specification and neural development,
mouse models are commonly utilized. Bent tail (Bn) is a spontaneous mutation
on the mouse X chromosome that produces tail deformities and open neural tube defects (NTDs). Analysis of progeny from an intraspecific backcross places the mutation between the microsatellites DXMit166 and DXMit140. Refined genetic
and physical mapping of the Bn critical region demonstrate that the mutation is
associated with a <170 kb submicroscopic deletion, including the entire Zic3 locus.
Human mutations in ZIC3 are associated with left-right axis malformations. The presence of anal and spinal abnormalities in some patients and deletion of Zic3 in
Bn mice support a key role for this gene in neural tube development and closure.
However, mutations in the ZIC3 gene have yet to be identified in families with X-
linked NTDs.
Holoprosencephaly (HPE) results from abnormal development of the forebrain. One gene associated with HPE in humans, TG interacting factor
(TGIF), was identified by its ability to bind to the retinoid X receptor response
element, and has been shown to play a role in regulating TGF-β signaling. HPE
is not evident in mice carrying the targeted null allele of Tgif. To elucidate
ii whether Tgif in conjunction with reductions in TGF-β signaling can cause HPE, mice that have mutations in both Smad2 and Tgif were generated. Results show that of the Smad2+/-; Tgif+/- and Smad2+/-; Tgif-/- embryos, one third display HPE.
Molecular characterization at E9.5 reveals that Shh, Fgf8, Six3 and Zic2 expression are not affected. The forebrain domain of Otx2 expression shows a modest to nearly complete reduction in affected embryos. Published data has shown that introduction of retinoic acid (RA) to cultured embryos or pregnant dams can induce HPE in embryos. Some of these studies have further shown a reduction in rostral Otx2 expression. The known interaction of Tgif and RA signaling led to the evaluation of the possibility that the HPE resulted from altering RA pathways. Analysis of developing embryos provided evidence that genetic deficiencies of Smad2 and Tgif allows for increased sensitivity to RA, suggesting a unique link between genetics and environmental teratogens.
iii DEDICATION
To my husband, my mother and father, and my brother. Thank you for your patience, love and support.
iv ACKNOWLEDGEMENTS
I wish to thank my advisor, Dr. Michael Weinstein, for his intellectual support, zeal, and for his patience and guidance.
I am grateful to my committee members Dr. Christine Beattie, Dr. Heithem
El Hodiri, and Dr. Amanda Simcox for their encouragement.
I thank the members of the Weinstein Laboratory, Ye Liu, Mark Hester,
Samuel Lasse, Maria Festing, J. Chris Thompson, Andy Chow and Sami Jafaar, for proofreading this dissertation and scientific discussions.
I am indebted to those at Columbus Children’s Research Institute for their advice, stimulating conversations and support, especially Dr. David Cunningham,
Dr. LaRae Copley, Dr. Heithem El Hodiri, Ms Marsha Lucas and Ms Marlene
Parker.
I also wish to acknowledge Dr. Gail Herman, whose laboratory the ZIC3 work described in this dissertation was performed.
Finally, I wish to thank the editors of The American Journal of Medical
Genetics, and Molecular and Human Genetics for their permission to use the previously published text and figures that are included in this dissertation.
v VITA
October 16, 1974 Born – Columbus, Ohio.
1995 Beta Beta Beta Biology Honorary.
1/96 – 6/97 Senior Thesis, Muskingum College-Advisor Dr. David Quinn. “Role of Vigabatrin in reversing stimulus- evoked afterdischarges in rat hippocampal slices.”
6/96 – 8/96 Internship, South Carolina School of Medicine - Advisors Dr. Leslie S. Jones and Dr. Marsha Welsh. “Spatial/temporal studies of the extracellular matrix of the developing rat brain.”
1997 Sigma Xi.
1997 Sigma Xi Biology Senior Thesis Award.
1997 Muskingum College, New Concord, Ohio. Biology (Chemistry), BS.
1998 Outstanding Graduate Teaching Assistant Award.
2000 21st Annual Children’s Hospital and Children’s Research Institute Research Forum, 1st place award in basic research.
2003 The Ohio State University, Columbus, Ohio. Program in Molecular, Cellular and Developmental Biology, MS.
2002 Children’s Research Institute Cherry Valley Retreat, 1st place, graduate student oral presentations.
9/97 - 6/98 Graduate Teaching Assistant, Biology 113. The Ohio State University - Supervisor Dr. Courtney Smith. vi 4/98 – 3/03 Graduate Research Associate, The Ohio State University - Advisor Dr. Gail E. Herman. “Identification of a deletion of the Zic3 locus associated with the X-linked mouse mutation Bent tail.”
4/03 – present Graduate Student, The Ohio State University - Advisor Dr. Michael Weinstein. “The TGF-β pathway and holoprosencephaly: understanding the role of Tgif and Smad2.”
6/03 - 9/03 Graduate Teaching Assistant, Molecular Genetics 500. The Ohio State University - Supervisor Dr. Andrea Doseff.
1/04 – 3/04 Graduate Teaching Assistant, Introductory Biology Program 101. The Ohio State University – Supervisors, Dr. Cheryl Johnston, Course Coordinator and Dr. Steve Rissing, Director, Introductory Biology Program.
PUBLICATIONS
Carrel, T., Purandare, S., Harrison, W., Elder, F., Fox, T., Casey, B. and Herman, G. E. “The X-linked mouse mutation Bent tail is associated with a deletion of the Zic3 locus.” Human and Molecular Genetics. 2000. 9(13). 1937-1942.
Carrel, T., Moore G. E., Stanier, P. and Herman, G. E. “Lack of mutations in ZIC3 in three families with neural tube defects.” American Journal of Medical Genetics. 2001. 98:283-285.
SELECTED MEETING ABSTRACTS
Carrel, T., Purandare, S., Harrison, W., Elder, F., Fox, T., Casey, B. and Herman, G. E. “The X-linked mouse mutation Bent tail is associated with a deletion of the Zic3 locus.” Presented at the 50th Annual American Society of Human Genetics. 2000.
FIELD OF STUDY
Major Field: Molecular, Cellular and Developmental Biology
vii TABLE OF CONTENTS
Page Abstract ………………………………………………………………………. ii Dedication …………………………………………………………………….. iv Acknowledgements ………………………………………………………….. v Vita ……………………………………………………………………………. vi List of Tables ………………………………………………………………… x List of Figures ……………………………………………………………….. xi List of Abbreviations ………………………………………………………… xiii
Chapters:
1 Introduction ……………………………………………………………… 1
1.1 Axial Patterning of the Mouse Embryo ………………………………. 2 1.1.1 Anterior-Posterior Axis Formation …………………………. 2 1.1.2 Dorsal-Ventral Axis Formation ……………………………… 5 1.1.3 Left-Right Axis Formation …………………………………… 6 1.2 Neural Development and Disease …………………………………… 9 1.2.1 Neural Tube Formation and Neural Tube Defects ………. 9 1.2.2 Prosencephalic Development and Holoprosencephaly … 12 1.2.3 Neural Maturation and Related Diseases ………………… 15
2 The X-Linked Mouse Mutation Bent Tail Is Associated With a Deletion of the Zic3 Locus ………………………………………………………….. 21
2.1 Introduction ……………………………………………………… 21 2.2 Results …………………………………………………………… 23 2.2.1 Genetic Mapping of Bn ……………………………… 23 2.2.2 Identification of a Submicroscopic Deletion Associated with the Bn Mutation ……………………………….. 25 2.2.3 Phenotypic Characterization of situs defects in Bn mice ……………………………………………… 26 2.3 Discussion ………………………………………………………. 27 2.4 Materials and Methods ………………………………………… 29 viii 2.4.1 Mouse Strains, Crosses and Genetic Mapping …… 29 2.4.2 Physical Mapping of Bn ……………………………… 29 2.4.3 Phenotypic Characterization ………………………… 30
3. Lack of Mutations in ZIC3 in three X-linked pedigrees with neural tube defects …………………………………………………………………… 39
4. Smad2 and Tgif Work in Concert to Cause Murine Holoprosencephaly ……………………………………………………… 44
4.1 Introduction ……………………………………………………… 44 4.2 Results …………………………………………………………… 47 4.2.1 Frequency and Appearance of Holoprosencephaly in Smad2; Tgif Embryos ……………………………….. 47 4.2.1 Molecular Analysis of Smad2; Tgif Holoprosencephalic Embryos ………………………………………………. 50 4.2.2 Retinoic Acid and Holoprosencephaly in Smad2; Tgif Embryos ………………………………………………. 53 4.3 Conclusions …………………………………………………….. 55 4.4 Materials and Methods ………………………………………… 62 4.4.1 Mice and Matings ……………………………………. 62 4.4.2 Genotyping …………………………………………… 63 4.4.3 Histology ……………………………………………… 63 4.4.4 Hematoxylin and eosin staining ……………………. 64 4.4.5 Bromodeoxyuridine Analysis ……………………….. 65 4.4.6 TUNEL Analysis ……………………………………… 67 4.4.7 Whole Mount In Situ Hybridization ………………… 69 4.4.8 Teratogenic Application of Retinoic Acid …………. 72
5. Discussion ……………………………………………………………….. 94
Works Cited …………………………………………………………………. 98
ix LIST OF TABLES
Page
1.1 Order of events of neural development ………………………………… 17
2.1 Nondisjunction in the Bent tail strain …………………………………… 32
2.2 Situs abnormalities in Bn mice …………………………………………. 33
3.1 ZIC3 PCR parameters and primers ……………………………………. 42
4.1 Summary of the evaluation of E10.5 Smad2; Tgif embryos …………. 74
4.2 Malformation frequency in Smad2; Tgif embryos ……………………. 75
4.3 Retinoic acid induced anomalies in Smad2 and Smad3 mutant embryos ………………………………………………………………… 76
4.4 Embryos evaluated via whole mount in situ hybridization ………….. 77
4.5 Genotyping PCR primers and annealing temperatures …………….. 78
x LIST OF FIGURES
Page
1.1 Anterior-posterior specification of an E7.5 embryo …………………… 18
1.2 Molecular pathway of the mouse in the patterning of the left right axis ………………………………………………………………….. 19
1.3 Closure initiation sights of the mouse neural tube …………………… 20
2.1 Tail phenotype and genetic mapping of Bn ………………………….. 35
2.2 Physical mapping of the Bn deletion ………………………………….. 36
2.3 Situs abnormalities in Bn mice …………………………………………. 38
3.1 Pedigrees of unpublished human X-linked neural tube defect family ……………………………………………………………… 43
4.1 The Smad pathway ……………………………………………………… 79
4.2 Craniofacial defects in Smad2+/-; Smad3+/- E 12.5 embryos ……….. 80
4.3 Holoprosencephaly in Smad2; Tgif liveborn pups and cyclopia ……. 81
4.4 Histological evaluation and comparison of sections of Smad2; Tgif embryos …………………………………………………… 82
4.5 Additional anomalies seen in Smad2; Tgif mutant embryos ……….. 84
4.6 Apoptosis was evaluated via TUNEL analysis ………………………. 85
xi 4.7 Cellular proliferation was analyzed via BrdU incorporation at E10.5 ………………………………………………………………….. 86
4.8 Molecular characterization of Smad2; Tgif holoprosencephalic embryos ………………………………………………………………….. 87
4.9 Evaluation of Otx2 expression ………………………………………… 89
4.10 Retinoic acid and Otx2 ………………………………………………… 90
4.11 Retinoic acid induces anomalies in Smad2+/- embryos ……………. 91
4.12 Retinoic acid does not induce anomalies in Smad3 mutant embryos ………………………………………………………………… 92
4.13 Otx2 potential promoter region ………………………………………. 93
xii LIST OF ABBREVIATIONS
A/P anterior-posterior
ATRA all trans retinoic acid
AVE anterior visceral endoderm
Bn bent tail
BMP bone morphogenetic protein cen centromere
CNS central nervous system
D/V dorsal ventral
E embryonic day
FA folic acid
HDAC histone deacetylase
HPE holoprosencephaly
LPM lateral plate mesoderm
MH mad homology
NTD neural tube defect
PFA paraformaldehyde
PBS phosphate buffered saline
RA retinoic acid
xiii RAR retinoic acid receptor
RXR retinoid X receptor
SE standard error
SARA Smad Anchor for Receptor Activation tel telomere
TGF-β transforming growth factor beta
xiv CHAPTER 1
INTRODUCTION
Understanding the genes in human development is accomplished through
the use of animal models. Animals are used to provide insight to the functions of
genes and their role in human disease. The mouse is indispensable in this
regard for many reasons. Genetically, for example, the mouse is similar to
humans, having approximately 85% conservation between the nearly 35,000
genes shared between the two. By using mouse strains harboring single genes with spontaneous mutations, or that have been targeted, generating null alleles, scientists are able to dissect the role of the individual genes in pathways of development, metabolism and cancer.
The spontaneous Bn mouse and targeted alleles of Smad2 and Tgif are examples of strains that can be used to evaluate human neural development in relation to known diseases. In addition, they are models demonstrating that the establishment of proper embryonic axial patterning is essential and if disrupted can lead to disorders that, in the most severe cases, cause embryonic mortality.
1 Together, the aforementioned mice have helped to more precisely define
mechanisms necessary during gestation.
1.1 Axial Patterning of the Mouse Embryo
1.1.1 Anterior-Posterior Axis Formation
Before the embryonic anterior-posterior (A/P) axis is established, the conceptus is oriented in a proximal-distal orientation. The terms “proximal” and
“distal” are references to the tissue fated to develop into the embryo, or epiblast, closeness to the hypoblast, or extraembryonic visceral endoderm (reviewed in
[1]). Molecular analysis of the epiblast has verified that gene expression becomes asymmetrical very early in a developing embryo, which structurally appears to have symmetry. One of the first genes to behave in such a manner is the Hex gene. Hex expression is first detected in the primitive endoderm of implanting blastocysts, and shortly later in a cluster of cells located in the distal tip of the egg cylinder. This cell cluster migrates around the perimeter of the egg
cylinder to the region which becomes the anterior visceral endoderm (AVE) [2].
Once the AVE is specified, asymmetry has been established, and it will help to
influence anterior cell fates in the developing embryo. Work has shown that the
AVE works as an extraembryonic area of induction, and plays a role in inducing
head structures. The mouse AVE can provoke anterior specific gene expression
in many models (reviewed in [3]). These molecular cues are essential for the
2 establishment of the A/P axis as well as patterning the anterior region of the
developing CNS.
Embryos that lack Hex exhibit anterior truncations [4]. Similarly, other
genes expressed in the AVE, namely Lim-1 and Otx2, are essential in early
embryonic development. Lack of Otx2 in mouse results in early embryonic
lethality. Embryos evaluated prior to death lacked tissues predestined to be the
most anterior, namely anterior neural fates [5]. It is believed that Otx2 and Lim-1 are involved in a genetic cascade produced in the AVE, resulting in the restriction of molecular prompts for posteriorizing the embryo [6]. Other genes expressed in the AVE include goosecoid and cerberus-related 1 (reviewed in [1]). Similarly, surgical removal of the AVE, results in the failure of forebrain inducing gene expression [7]. This exemplifies the importance of this extraembryonic region in the stimulation of forebrain development.
Another landmark in the developing mouse embryo that is important for A/P patterning is the node. This group of cells, which has the ability to organize the body plan of the developing embryo, is established once gastrulation has been initiated. The node is located at the anterior end of the primitive streak (reviewed in [8]). Examples of node-specific genes are chordin and noggin. Both encode
molecules that are secreted and function as antagonists of BMP function. Mice
that are homozygous null for chordin and noggin have severe defects in
prosencephalic development [9], reiterating the importance of the node and its
expressed genes in the specification of the anterior end of the embryo. During
gastrulation, cells ingress through the primitive streak, which is established on
3 the future posterior side of the embryo. This involution causes the streak to
lengthen and results in specification of the mesoderm and definitive endoderm
layers. The cells that pass through the primitive streak early in gastrulation are fated to become anterior; cells that pass through later, will take on posterior cell fates. These migrating mesodermal and endodermal cells lose the ability to produce E-cadherin, and can therefore travel as single cells through the primitive streak (reviewed in [10]). Both the AVE and the node are essential for the establishment of anterior cell fates, however, as will be discussed later, the node is important for dictating the overall structure of the body plan. While both are essential for anterior induction, other tissues or structures may play a role in the maintenance of anterior identity.
Hox genes are very important in patterning of A/P axis. The initiation of expression of these genes begins shortly after gastrulation. Although they are not expressed in the most anterior region of developing embryos, they are detected in the presumptive midbrain/hindbrain region to the posterior extreme of the embryo. These genes are located in clusters within the genome, and are expressed in the same sequence in which they are arranged along the chromosomes. The importance of these genes in development is remarkable and well conserved from Drosophila to human. Evolutionarily, these paralogous gene groups have been duplicated. While only one cluster is present in
Drosophila, four are present in mouse and humans (reviewed in [10, 11]). The earliest known expression of Hox gene is in a discrete order that parallels their sequential location on the chromosomes. This allows for the establishment of
4 developmental segments, which specify boundaries for regions such as
forebrain, midbrain and hindbrain. Hox genes are also important for limb
development, organ formation, neural tube patterning, craniofacial development
[10-13].
Once proper patterning events have taken place, definition of the anterior and posterior ends have been established in the embryo. This allows for the initiation of events such as establishment in differences between brain and spinal cord.
1.1.2 Dorsal-Ventral Axis Formation
After gastrulation, the three germ layers, endoderm, mesoderm and ectoderm, are established. Both the mesoderm and the ectoderm are specified
along the dorsal/ventral (D/V) axis by means of similar molecular cues. The
graded expression of noggin, chordin [14, 15] and follistatin [16] on the future
dorsal aspect initiates D/V patterning in both the mesoderm and the ectoderm.
As a result, neural ectoderm will be specified as well as dorsalized mesoderm, a
precursor of muscle and notochord (reviewed in [17]). Acting as an antagonist
to the dorsalizing factors, is the TGF-β molecule BMP-4 [16]. This
antineurogenic signal helps establish precursor tissue that will develop into blood
and mesenchyme from the mesoderm and epidermis from the ectoderm
(reviewed in [17]).
D/V polarity of the neural tube may be better understood than overall D/V
patterning in the embryo. Once neural tube closure has occurred (reviewed in
5 1.2.1), the dorsal aspect, fated to become the spinal cord, will develop the regions to direct sensory information, and the ventral side will guide motor control. Establishment of this polarity is aided ventrally by signals initiated from the notochord, and dorsally by the epidermis. More specifically, graded expression of Shh from the notochord is important for the determination of ventral cell fates. This signaling results in the secretion of molecules that inhibit BMPs.
The net level of expression of the TGF-β pathway (to be discussed in Chapter four) signaling molecules (BMPs and activin) from the epidermis initiates and maintains dorsal identity in the neural tube. In both cases, these gradients induce downstream transcription factors, which initiate the process of determining polarity.
1.1.3 Left-Right Axis Formation
It is well understood that regardless of the outward appearance of most vertebrates, internally they are asymmetrical in regards to position of the visceral organs on the left-right (L/R) axis. The stomach is positioned on the left side, in addition to the spleen. Most commonly, the appendix is in the right lower quadrant of the abdominal cavity. While it appears that the lungs are symmetrical organs with one on each side of the L/R axis, the lobulations of the lungs are very different. The left lung in humans has two lobes while the right has three. (This is different than mouse, which has one and four, respectively.)
Even though the L/R axis is the last of the three to form, it is just as complex and essential as the two previously discussed.
6 Human syndromes of abnormal L/R axis development range in complexity from single organ reversal (situs ambiguus or heterotaxy) to an internal mirror image of normal organ placement (situs totalis). The number of organs involved in the syndrome, however, does not mimic the severity of the anomalies
(reviewed in [18]). Nevertheless, heterotaxy that involves heart and/or great artery malformation commonly leads to problems with survival.
Nodal flow, or movement of fluid across the node in the gastrulating embryo, has been shown to be essential for the establishment of right-left polarity. During this event, monocilia located at the node force extraembryonic fluid in a leftward direction [19, 20]. This is the first incident in which bilateral symmetry is broken.
A role for the kinesin motor family genes has been implicated by the creation of gene knockout and spontaneous mutants. One such example is the iv (inversus viserum) mouse. In this mouse a spontaneous mutation in the left-right dynenin gene leads to a random pattern of L/R patterning due to non-motile cilia [21, 22].
Other related molecules that have similar developmental consequences when mutated are KIF3A and KIF3B, which lack cilia at the node [19, 23]. In addition to the motile cilia at the node, non-motile cilia are also present. These non-motile cilia act as mechanoreceptors, which detect an increase of intracellular calcium
[24]. The effect of the calcium in downstream events is still unclear. An example of cilia related situs defects in humans are found in individuals with Kartagener syndrome [25]. In addition to indiscriminate placement of internal organs, other ciliary defects are present in these persons. Examples include infertility in males, and respiratory defects due to the malformation of tracheal cilia.
7 Nodal, a TGF-β gene, is initially expressed uniformly across the node. (The
TGF-β pathway will be reviewed in chapter four). Once nodal flow is initiated, its expression becomes restricted to the left side of the node [26]. Shortly thereafter, nodal expression on the left side of the developing embryo expands to the left lateral plate mesoderm (LPM). Similarly, lefty-2 and Pitx2 also become localized to the left LPM with nodal. Essential for lateralization is the creation of a specific midline. It is believed that lefty-1 acts as the molecular regulator of the midline. These events act as key initiators for the specification of asymmetric organ development (reviewed in [27]).
Other genes have been implicated in L/R patterning. One of the first genes to be associated with situs anomalies in both mouse and humans, ZIC3 [28-30], will be discussed in Chapter 2. Another such gene is Smad2. While many groups have generated targeted null alleles of Smad2 [31-33] which result in embryonic lethality prior to axial specification, one group has demonstrated through tetraploid rescue of the early lethality that embryonic Smad2 is essential for L/R specification [34]. Failure of any of the aforementioned genes to be expressed correctly during patterning of the L/R axis could result in a range of anomalies. Studies in mouse models have shed an enormous amount of light on the understanding of this process.
One mouse mutant that displays laterality defects is somewhat perplexing.
Unlike every other model, inv, a chromosome 4 insertion of a transgene in mouse, results in complete penetrance of random R/L axial development [35].
50% of all inv mutants have a complete reversal of organ position (situs totalis), 8 and 50% have normal orientation, or situs solitus. It is believed that the gene(s) disrupted by this mutant is one of the first involved in the establishment of the L/R axis, however, the mechanism is not understood.
1.2 Neural Tube Development and Disease.
1.2.1 Neural Tube Formation and Neural Tube Defects
Neural tube development is a dynamic process that is similar between mouse and humans. This process consists of three stages: neural tube formation, prosencephalic development and neuronal maturation. In humans, during the third and fourth week of gestation (murine embryonic day (E) 8-9), primary neuralation, or neural tube formation and closure, occurs. Prior to this stage, during gastrulation, the neural plate is established, when the specified ectoderm is divided into neural and non-neural fields. The area where the neural plate meets the non-neural ectoderm generates the neural folds, which rise, become concave and fuse at the embryonic midline to generate a closed cylindrical form from the neural plate. This new structure is termed the neural
tube (reviewed in [36]). While it seems likely that the closure event could initiate
at one end of the elevating neural plate and, in zipper-like fashion, seal in a
cranial to caudal fashion, closure is initiated in a number of points along the axis,
followed by fusion between the points occurs (Fig. 1.3) (reviewed in [36]).
In mouse, caudal eminence, or neural formation in the tail bud, occurs. The cell
population in this region is comprised of cells that have withdrawn from the 9 primitive streak. Delay or failure of the posterior neuropore to close can result in
kinking or curling of the tail. Occasionally, the tail phenotype is accompanied by
spina bifida [37, 38].
Mutations of genes that aid in primary neuralation can lead to neural tube defects (NTDs). In regards to human congenital malformations, the frequency of
NTDs is second only to congenital heart disease among newborns with an
estimated incidence of 1/1000 pregnancies in the United States, and occurring in
300,000 total births annually [39]. In addition, certain geographical locations
have incidences of NTDs in 3.5 of 1000 live births [40]. This group of disorders
ranges from the rostral end of the neural axis to the caudal end, represented by
exencephaly and spina bifida respectively, in the mouse. Due to the longer
gestational period of humans, those embryos that begin with exencephaly will
generate anencephaly because of the exposure of brain tissue to amniotic fluid.
When larger regions of the neural tube fail to fuse, the resultant phenotype is
chraniorachischisis, but usually, the forebrain and anterior midbrain develop
normally in these cases. The most severe NTD is caused by failure of the entire
neural tube to close resulting in craniorachischisis. While some cases of single
gene disorders or chromosomal abnormalities have been associated with NTDs,
the cause in humans appears to be multifactoral with genetic factors accounting
for only 60% of the risk [41]. Currently, 80 mouse mutants that display NTDs as
part of their phenotype have been identified [36]. These serve as models to for
understanding the mechanisms behind the events that lead to neural tube
formation. 10 It is clear that environmental factors also play an important role in NTD
pathogenesis [42]. Studies have shown that maternal intake of alcohol during
pregnancy has severe effects on the formation of the neural tube (reviewed in
[10]). In addition, periconceptional administration of folic acid (FA) to women of childbearing age has resulted in a dramatic decrease in the incidence and recurrence of human NTDs [43, 44] and is now recommended for all women contemplating pregnancy [45]. However, the mechanism by which FA acts to reduce NTDs (and other birth defects) is not known, and there are numerous studies completed and in progress investigating the role of genetic variations in key proteins in folate metabolism in NTD pathogenesis (reviewed in [42, 46]).
Of the many mouse mutants that display NTDs, only four are known to display sensitivity to FA: crooked tail (Cd), cartilage homeoprotein 1 (Cart1),
Cited2 and splotch (Sp). The gene defect in Cd has not been identified;
however, FA supplementation causes a 55% reduction in the occurrence of
exencephaly in the mutant strain [47]. Cart1 is a homeoprotein that acts as a
transcription factor and has been demonstrated to play a role in craniofacial
development [48]. Cart-/- mice develop acrania and meroanencephaly, which can
be suppressed by prenatal FA supplementation of pregnant heterozygous dams
[49]. Cited2 (formerly known as Mrg1) encodes a novel nuclear protein that acts
as a transcriptional coactivator. Cited2 null animals die late in gestation due to
heart defects and exencephaly [50]; the frequency of the latter malformation is
reduced by FA supplementation. The splotch mutant results from mutations in
the Pax3 transcription factor, which is expressed in developing neural folds [51].
11 FA supplementation of Sp2H heterozygous dams decreases the incidence of NTD
in resulting offspring [52]. However Gefrides et al. [53] saw no reduction in NTDs in the splotch embryos when FA was administered to Sp/+ dams. This conflicting
data may be due to the different alleles used in the two studies. Additionally,
genes involved in the FA pathway have been demonstrated to be involved in
early development [54], resulting in embryonic lethality in some [54, 55].
Occasionally, supplementation of FA reverses the developmental defects caused
by the absence of these genes [54].
1.2.2 Prosencephalic Development and Holoprosencephaly
Once the neural tube has closed, it is segmented into the prosencephalon,
mesencephalon and rhomencephalon. These segments further
compartmentalize into developmental regions between days E9-10.5 in the
mouse or gestational days 25-31 in human development. The most rostral of the
primary segments is the prosencephalon, a precursor to the functional forebrain
and contributor to the shaping of facial structures. And the prosencephalon
forms after the closure of the anterior neuropore (reviewed in [56]). During the fifth and sixth weeks of gestation, it develops and cleaves along three planes.
Horizontal cleavage leads to the formation of paired optic vesicles, and olfactory bulbs and tracts. Cleavage along the transverse plane leads to patterning of the telencephalon and diencephalon. And finally, cleavage along the sagital plane allows for the bifurcation of the prosencephalon into the left and right paired hemispheres (reviewed in [56]). For each of these planes to form properly, axial
12 specification (reviewed in section 1.1) must have been correctly established.
After these segmentation events, thickenings begin to arise within the
prosencephalon, which lead to the development of important structures within the
brain, such as the corpus callosum and pituitary (reviewed in [56]).
Examples of disorders, which arise from abnormal formation of the prosencephalon, include aprosecephaly, atelencephaly and holoprosencephaly.
The most commonly known of this group is holoprosencephaly (HPE), a spectrum of anomalies affected both the brain and facial formation [57]. HPE is a structural abnormality of the brain characterized by failure of the forebrain to cleave into two distinct hemispheres. The incidence of HPE has been determined to be 1:250 early in development [58], but is estimated to occur in 1:10,000 –
16,000 live births [59, 60], thus contributing to a high amount of fetal mortality.
There are three classifications for patients with HPE: lobar, semilobar and a lobar. In the most severe cases of patients with alobar HPE, a single ventricle in the forebrain is present and usually accompanied by cyclopia and a proboscis.
Semilobar patients display milder facial phenotypes such as hypotelorism, cleft lip or palate and a flattened nose with a single nostril, and in addition have partial separation of the anterior ventricles. The most mild form of HPE, lobar, presents near complete separation, with even more mild outward phenotypes (reviewed in
[61, 62]).
While the majority of cases of HPE are sporadic, single gene and chromosomal anomalies have been associated with the disease (reviewed in
[63]). Currently, there are 12 loci that have been implicated in HPE, and 11 13 genes have been associated with these loci. However, genetic predisposition
through mutations in these identified genes do not dictate complete penetrance
of HPE, in that only 70% of obligate carriers show features that are part of the clinical definition [64]. Multiple genes in the hedgehog and TGF-β pathways have been implicated in HPE, illustrating the importance of these two signaling pathways in the development of the forebrain (reviewed in [62, 63]). In addition
to the notochord expression of Shh discussed in 1.2, expression is observed
along the ventral midline of the developing midline [65]. This expression is
important for establishing the ventral aspect of the developing neural tube. Not
only has the result of targeted disruption of Shh been studied in mice, mutations
in the human gene have also been identified. In both circumstances, the
resulting phenotype was HPE ([66] and reviewed in [63]). In addition to Shh, its
receptor and mediators of signaling have been associated with HPE in humans
and/or mouse [62-64, 66]. Mutations within genes of TGF-β pathway in human
and animal models that lead to HPE include nodal homologues, BMP4, BMP5,
and TGIF (reviewed in [62-64]). The latter will be discussed in Chapter 4. In
addition, ZIC2 and SIX3 mutations have been identified in mouse and humans
[67-69].
1.2.3 Neuronal Maturation and Related Diseases
The remaining neuronal events begin during the third month of human
gestation and are necessary for the formation of central nervous system (Table
1.1). While they are each important for a mature and functional CNS, they will 14 only be discussed briefly at this time. However, this conciseness does not reflect
the complexity of the molecular mechanism that controls each process. Initially,
cellular precursors of neurons and glia, arising from the periventricular zones, proliferate (reviewed in [56, 70, 71]). After proliferation, the cells migrate in a methodical sequence and differentiate, as a prelude to the formation of synaptic connections (reviewed in [10]). The migrating cells travel from their origin to locations within the CNS. These loci are then the permanent homes to the developing cells. Many events comprise the maturation element of neural development. During this time, neurons orient and become aligned, the dendrites and axons elongate, and synapses are formed. In addition, apoptotic events occur in a selective manner, and glia differentiate (reviewed in [56]).
Lastly, at birth, myelination of the axonal projections occurs. Myelination is essential for the proper conduction of electrical impulses along the axons. Myelin is itself a very complex molecule, consisting of many proteins and lipids.
Failure of any step of these events can result in a number of developmental anomalies or diseases. Atypical proliferation, either hyper- or hypo-, can lead to abnormal sized or shaped brains (reviewed in [56]). One example of a condition in which proliferation is affected is termed microcephaly.
Patients with microcephaly have a head circumference three standard deviations below the average size. Other symptoms may accompany the smaller head size, most commonly, variable intelligence [72]. Migration defects can be attributed to aberrant molecular signals, and to metabolic disorders. Zellweger syndrome is one such disease [70]. Patients with Zellweger syndrome exhibit defects of
15 neuronal migration, and developmental delay that leads to neonatal lethality [73].
This phenotype is identical to that seen in mouse models, with targeted null alleles of Pex2, Pex5, and Pex11b [74]. A well known example of a disease related to myelination disorders are metachromatic leukodystrophy (MLD) and childhood multiple sclerosis (MS) (reviewed in [71]). MS is a more commonly discussed mylenation disease, which is attributed to demylination due to an autoimmune disorder. MLD is less familiar, and is due to a defect in lysosomal glycolipid degradation, resulting in the slowing of nerve conduction velocity.
Depending on the age of onset in the patients, different motor and physical hindrances arise.
16 Developmental Event Gestational Time (Human) Neural tube formation and closure 3-4 weeks Prosencephalic development 2-3 months Neuronal maturation: Neuronal proliferation 3-4 months Neuronal migration 3-5 months Organization 5 months - years postnatal Mylenation birth - 2 years postnatal Table 1.1. Order of event of neural development. Adapted from [56].
17 Amnion
Embryonic Alantois ectoderm
AVE
Amniotic Cavity Prechordal Mesoderm Endoderm Node Primitive Streak
Figure 1.1. Anterior-posterior specification of an E7.5 embryo. The specialized AVE is specified on the future anterior end of the embryo, while the primitive streak has elongated on the posterior region. (Adapted from [1, 10]).
18 Inv, iv KIF3A, KIF3B
nodal flow
Fgf8
Nodal (lefty2) Lefty 1 LPM
Pitx2 Snail
Left identity Right identity
Figure 1.2. Molecular pathway of the mouse in the patterning of the left-right axis [10, 75].
19 2
4
E
3
B
1
Figure 1.3. Closure initiation sights of the mouse neural tube. After neural tube closure is initiated in the hindbrain region (1), points of closure follow as numerically indicated above. In a zipper-like fashion, the neural folds elevate and seal following the arrows. In humans, a fifth sight is located caudally, beyond the initial sight. E=eye, B=First branchial arch
20 CHAPTER 2
THE X-LINKED MOUSE MUTATION BENT TAIL IS ASSOCIATED WITH A DELETION OF THE ZIC3 LOCUS
2.1 Introduction
Neural tube defects (NTDs) are one of the most common human congenital malformations, second only to congenital heart disease in terms of overall frequency. While chromosomal, monogenic, and teratogenic causes of human NTDs are recognized, the majority are multifactorial in etiology (reviewed in [39]). It has been estimated that genetics factors account for approximately
60% of the risk for human NTDs. The introduction of prenatal screening and periconceptual folic acid supplementation have resulted in a dramatic decline in the incidence of NTDs [76]; however, the basic mechanisms involved in their pathogenesis remain largely unknown.
To help identify genes that may play a role in the pathogenesis of human
NTDs, many researchers have examined mouse mutants that exhibit NTDs as part of their phenotype. Numerous spontaneous mouse mutants with NTDs have been described, and many additional mutations have recently been generated through the technology of transgenic mutagenesis and homologous recombination [77]. Mouse models of NTDs share many features with their 21 human counterparts. These include the location of the defects, possible mechanisms of pathogenesis related to intermittent sites of neural tube closure, and variable expressivity that is based on genetic background and sex [77]. The shorter gestation period in the mouse compared to human is usually reflected in the preservation of brain tissue for cranial malformations, and the resulting defect is exencephaly. Lower NTDs may present as tail defects in the mouse and are represented by mutations such as vestigial tail (vt), fused (Fu) and bent-tail (Bn).
Some mouse mutants display both upper and lower NTDs, and some are
associated with malformations of other organ systems. In particular, eye,
vertebral/rib, and limb malformations are commonly associated with mutants that
display NTDs.
The X-linked bent-tail (Bn) mutation was isolated in 1952 among offspring of
an outbred Namru strain female by a bald (hrba) male [78]. A single allele exists.
Affected hemizygous males and homozygous females have short, kinked tails,
occasional exencephaly or sacral neural tube defects, and reduced viability and
fertility on the inbred Bn background. Heterozygous females have variable, milder
tail abnormalities [78] (Fig. 2.1 A). The presence of an interfrontal bone in the
skulls of affected males and females has also been described [79]. The Bn locus has been regionally mapped in several phenotypic crosses to the middle third of the mouse X chromosome, close to trembly-like (Tyl) [80, 81]. Lack of penetrance in affected Bn females is common, complicating these genetic mapping attempts.
We demonstrate here refined genetic mapping of Bn using molecular markers and further provide evidence that the mutation is caused by a submicroscopic deletion 22 that includes Zic3, a gene whose human ortholog has been previously associated
with malformations of left-right asymmetry and with NTDs.
2.2 Results
2.2.1 Genetic mapping of Bn
To map Bn, an intraspecific backcross was established between heterozygous Bn/+ females and C3H/HeJ males. A prior attempt to generate a backcross using Mus castaneus males resulted in no phenotypically affected animals in over 40 F1 females examined, supporting the influence of modifier loci
on the expression of the tail defects. Affected F1 females from the Bn x C3H cross
were identified at weaning by the presence of one or more kinks in their tail and
crossed to normal males from the inbred Bn stock. Affected males and females and normal appearing male progeny from the backcross were analyzed initially with the markers DXMit140 and DXBay6 [82] that were found to be polymorphic between the Bn and C3H strains and likely to flank the mutant locus. Normal
appearing female progeny were excluded from the analysis because of the
reported [78] and observed incomplete penetrance (see Fig. 2.1 A).
A total of 292 N2 progeny from the backcross were analyzed with the
microscopic microsatellite markers DXMit89 and DXBay6 (Fig. 2.1 B), and fine
mapping of Bn was performed using the 79 animals that had recombination events
between DXMit89 and DXBay6. Pedigree analysis of these recombinants supports
a gene order of cen – DXMit89 – 18.5 cM ± 2.3 cM – DXMit166 – 1.4 ± 0.7 cM – Bn 23 – 1.0 cM ± 0.6 cM – DXMit140 – 4.8 ± 1.3 cM – DXBay6 – tel, where the numbers
represent recombination frequencies calculated in centimorgans ± a standard error
(SE). Complete mapping data for the cross are available through the Mouse
Genome Database (http://www.informatics.jax.org/; accession no. J62602).
During these analyses, we observed several males that had inherited both a
Bn strain and C3H allele at the DXMit140 and DXBay6 loci. The presence of a Y
chromosome was confirmed in each case by PCR genotyping using primers for the
Smcx locus that detect different sized fragments from the X and Y chromosome (H.
Willard, personal communication). Possible explanations for the occurrence of heterozygous males included duplication of a portion of the X chromosome or meiotic nondisjunction. Further analysis using additional polymorphic microsatellite markers spanning the X chromosome supported the latter since large portions of or the entire X chromosome contained both alleles in all of the animals (Table 2.1).
Based on the pattern of crossovers in some of these heterozygous males, it was
clear that both a maternal and paternal X chromosome had been inherited,
consistent with paternal nondisjunction. Since the father’s X chromosome always
contained a normal Bn allele, the nondisjunction is almost certainly unrelated to the mutation and is a characteristic of the background strain. Complete karyotype analysis of two Bn females demonstrated a 39,X karyotype in one and a normal karyotype with a normal X chromosomal banding pattern in the second, suggesting that the nondisjunction probably affects both males and females. From the number of heterozygous backcross males, the frequency of nondisjunction in this strain is estimated to be at least 4.4%. 24 2.2.2 Identification of a submicroscopic deletion associated with the
Bn mutation
To identify additional polymorphic molecular markers within the Bn critical
region, DNA from affected males was examined with 20 microsatellite markers
(DXMit165, 47, 83, 192, 225, 226, 208, 22, 107, 23, 159, 91, 68, 193, 108, 86,
142, 141, 74, and 127) localized on the X chromosomal consensus map between
14 and 20 cM [82]. Although the microsatellite marker DXMit208, localized at
16.5 cM, was not polymorphic between the C3H and Bn strains, it was found to
be deleted in hemizygous Bn males (Fig. 2.2A). None of the other 19 anonymous markers from the region were found to be deleted (data not shown).
Subsequently, additional genes from the region were tested for their presence or
absence in Bn males. These included Hprt, Fgf13, Tnfsf5 (formerly Cd40l), and
Zic3. All of these genes produced a normally sized PCR product with the
exception of Zic3, which was also deleted in Bn males (Fig. 2.2 A). Southern
analysis using several genomic probes from the Zic3 locus confirmed that the
gene was deleted completely in Bn males as examined by RT-PCR (data not
shown).
To determine the extent of the deletion in Bn, 5 BACs were identified by
screening two genomic BAC libraries with DXMit208 and Zic3. All of the BACs
identified contained both markers. Based on PCR and hybridization of end
probes generated from four of these BACs, a physical map of the region deleted
in Bn alleles was constructed (Fig. 2.2 B). As in human, the Zic3 gene spans
<10 kb of genomic DNA. The Bn deletion is estimated to be between 60 and 170
25 kb in size based on amplification pattern of genomic DNA prepared from affected
Bn males using primers for the loci DXChri1-DXChri5 that represent BAC clone end sequences (Fig. 2.2 B).
As part of the efforts of the Human Genome Project, two overlapping PAC clones containing the human ZIC3 locus and spanning 246 kb have been completely sequenced and assembled (Accession Nos. AL035443 and AL022576).
There are no known genes or ESTs within the 37 kb of assembled sequence 3’ of
ZIC3 and the closest gene or EST 5’ is the P21-RAC1 locus, at a distance of 124 kb. P21-RAC1, or RAS-related C3 botulinum toxin substrate 1, is an intronless transcript encoding a protein of 192 amino acids that functions as a small GTP- binding protein and is a member of the Rho family of proteins [83, 84]. Recently, a chick ortholog of P21-RAC1 was shown to be expressed during neural development [85]. Using PCR primers designed from the murine R21-RAC1 cDNA sequence (Accession No.AW318995), we determined that this gene was not deleted in Bn males (Fig. 2.2 A).
2.2.3 Phenotypic characterization of situs defects in Bn mice
Mutations of the human ZIC3 locus have been identified in sporadic males with left-right axis malformations, as well as in several families with similar malformations and x-linked patterns of inheritance [28]. All of the affected males had situs ambiguus with complex congenital heart disease, abnormalities of lung lobulation, and abnormal placement or structure of the stomach, liver, and/or
26 spleen. Therefore, we examined viable Bn males and females, in comparison with normal male littermates, for abnormalities of left-right axis determination. In
the inbred Bn stock, situs ambiguus, in which individual organ position is randomized, was identified in >50% of surviving Bn males and in 38% Bn females (Table 2.2 and Fig 2.3). One Bn female also had an abnormality of lung lobulation, although this finding was not statistically significant (Fig. 2.3 D).
Anomalous vessels, including aortic arch abnormalities or congenital heart defects, were not detected after careful thoracic dissection. However, their occurrence would often be expected to cause pre- or perinatal lethality and, hence, not be present in surviving affected animals.
2.3 Discussion
We have demonstrated that the murine bent-tail mutation is associated with a submicroscopic deletion of the X chromosome that encompasses the Zic3
locus. Zic3 is a C2H2zinc-finger transcription factor that is homologous to the
Drosophila odd-paired gene [86]. Its cellular targets remain unknown although it
is expressed in embryonic mesoderm by 7 days post conception in the mouse
and at later stages (day 9.5-11.5) in the developing nervous system, eye,
dermomyotomes, limb and tail bud [87].
The Bn mutation was originally identified based on its tail phenotype.
In addition to left right axis malformations, 8 of 18 affected males had anal
anomalies, and three additional affected males had lumbosacral spine defects,
27 including sacral agenesis in two and an open neural tube defect in the third [28,
88]. We have now detected abnormalities of situs, similar to those found in
human ZIC3 patients, in affected Bn males and females from the original inbred stock. There is also a significant difference between the number of affected and normal males in our backcross (P<0.01, χ2 analysis), raising the possibility that pre- or perinatal loss of some severely affected males, perhaps with lethal situs
abnormalities. It is also interesting to speculate on possible species-specific
differences in the penetrance of tail and spinal defects versus cardiac and situs
anomalies between human and mouse. Examining the effects of modifier loci on
the Bn phenotype by placing the mutation onto different genetic backgrounds
may begin to address this issue.
During our genetic mapping studies, we detected a high frequency of sex
chromosomal nondisjunction in the background strain on which the Bn mutation
arose. The X-linked, semi-dominant mouse mutation patchy fur (Paf) maps to the
boundary of the pseudoautosomal region and is associated with an increased rate of sex chromosome aneuploidy, perhaps as a result of a small chromosomal rearrangement that impairs X-Y pairing [89]. Hunt and Eicher [90] generated a
strain of males that produce >50% XY sperm. Mice carrying Robertsonian
translocation also show high frequencies of autosomal nondisjunction [91]. The
Bn background strain may, thus, prove useful as another model to study mammalian nondisjunction.
In human, the region immediately surrounding ZIC3 is gene-poor, and it is possible that no additional genes are within the murine deletion. Our finding of
28 left-right axis malformations in Bn mice, in addition to tail defects, is consistent
with the phenotype of ZIC3 mutations in human patients and suggests that
deletion of Zic3 is responsible for most, if not all, of the phenotypic features of
this mutant.
2.4 Materials and Methods
2.4.1 Mouse strains, crosses, and genetic mapping
The Bn strain was obtained from The Jackson Laboratory. It has been
maintained by brother-sister matings between phenotypically affected
heterozygous females and normal males from the original Bn stock. For genetic
mapping, heterozygous Bn females were crossed to C3H/HeJ males (The
Jackson Laboratory), and resultant phenotypically affected F1 females were
backcrossed to normal appearing males from the inbred Bn stock. Microsatellite
markers were purchased from Research Genetics and PCRs performed as
recommended by the manufacturer.
2.4.2 Physical mapping of Bn
Southern hybridizations of Bn and normal genomic DNA were performed as described [92] on SacI- or EcoRI-digested DNA using probes from the 5’, middle and 3’ portions of the murine Zic3 gene. Genomic PCR for Zic3 utilized forward
primer 5’-ATTGCTGCAAGTGCCGGGAACT-3’ and reverse primer 5’-
ACTCACATCTCCTATTTGTAG-3’ from the 3’-untrsnslated region and cycling
29 parametres of 94oC, 55 oC and 72 oC for 30 seconds each for 35 cycles. For
P21-Rac1, the PCR primers were forward primer 5’-GCCAATGTTATGGTAGG-3’
and reverse primer 5’-CAGGACTCACAAGCGAAAA-3’. PCR was performed for
35 cycles with denaturation at 93 oC for 1 minute, annealing at 57 oC for 1 minute, and extension at 72 oC for 1 minute. BACs were obtained by PCR screening a
total mouse genomic BAC library (Research Genetics) with DXMit208 or by
hybridization screening of the RPCI23 library [93] using a mouse Zic3 cDNA
probe. BAC DNA was prepared using a NucleoBond Plasmid Kit (Clontech) and
end sequencing performed with left and right vector-specific primers on an
ABI377 automated sequencer. Primer sequences and PCR conditions for the
BAC end loci DXChri1-DXChri5 (GenBank accession nos AZ044550-AZ044554)
have been entered into the MOUSE Genome Database. For PFG analysis, 1 ug
of BAC DNA was digested with NotI or EagI (New England Biolabs) and
electrophoresis performed on a CHEF Mapper XA System (Bio-Rad) as
recommended by the manufacturer. Digested DNA was transferred to Sure Blot
Nylon Membranes (Intergen) and Southern hybridization performed as described
[94] with radiolabeled vector-specific or gene probes.
2.4.3 Phenotypic characterization
Affected Bn males and females and normal male littermates from the
inbred stock, 2 weeks-5weeks of age, were sacrificed and abdominal and
thoracic viscera examined using a dissecting microscope (Zeiss); magnification
1-5x with assistance from an experienced perinatal pathologist. Cardiac
30 anatomy was ascertained by tracing the pattern of normal blood flow prior to sectioning of the heart to look for internal malformations.
31 DX DX DX DX DX DX DX DX Mit89 Mit166 Mit140 Mit25 Mit1 Bay6 Mit16 Mit130
Animal Phenotype 3.0 15.5 19.0 27.8 29.0 29.5 37.0 55.0 cM cM cM cM cM cM cM cM
2685 nl male B/Ca B/C B/C B/C B/C B/C B B
2729 nl male B B/C B/C B/C B/C B/C B/C B/C
2825 nl male B/C B/C B/C B/C B/C B/C B/C B/C
2848 nl male B/C B/C B/C B/C B/C B B B
2897 nl male B/C B/C B/C B/C B/C B/C B/C B/C
2908 Bn male B/C B/C B/C B/C B/C B/C B/C B
2971 nl male B B/C B/C B/C B/C B/C B/C B/C
3014 Bn male B B B/C B/C B/C B/C B/C B/C
3029 Bn male B B B/C B/C B/C B/C B/C B/C
Table 2.1. Nondisjunction in the Bent tail strain. aB=Inbred Bn allelle, C=C3H/HeJ allele.
32 Abdominal Situs Thoracic Situs normal abnormala normal abnormal
normal male 15 0 15 0 Bn male 6 7b (p=0.001)c 13 0 Bn female 8 5d (p=0.008)c 12 1e
Table 2.2. Situs abnormalities in Bn mice. a One normal male, one Bn female and one Bn male had an isolated left-sided appendix. This finding was not counted abnormal, except in conjunction with other anomalies. b This number represents two males with a midline liver and gall bladder; three males with a left-sided liver and gall bladder; and two males with left-sided liver gall bladder and appendix. c Determined by χ2 analysis. d Includes four females with left-sided liver and gall bladder and one female with left sided liver, gall bladder and appendix. e Partial right sided pulmonary isomerism (Fig. 2.3 D); not statistically significant (p>0.1).
33 Figure 2. 1. Tail phenotype and genetic mapping of Bn. (A) The tails shown are from an affected Bn male offspring of a heterozygous Bn female X normal male (left) and two affected Bn female offspring of a normal female X Bn male (center and right). Note incomplete penetrance of with normal appearing tail in one of the females (center). (B) Ideogram of the mouse X chromosome. The approximate positions of anchor loci and the microsatellite markers used to map Bn are shown. The map positions of the microsatellite markers were obtained from the X chromosome consensus map [82] and do not reflect distances predicted in this cross. The critical region within which the Bn locus must lie is designated with a black bar.
34 A
B Otc DXMit89
Lamp2 A DXMit166 Bn Hprt DXMit140 F9 Fmr1 DXBay6 F8 B Dmd
Zfx D Xist E Plp
F Sts
35 Figure 2.2. Physical mapping of the Bn deletion. (A) Deletion of the Zic3 gene and DXMit208 in Bn males. PCRs were performed for the loci shown using standard conditions with genomic DNA from a normal (N) or Bn (B) male. Control lanes containing no DNA for each primer produced no amplification (not shown). The sizes of the fragments amplified were Chri1 114 bp, Zic3 200 bp, DXMit208 124 bp, Chri4 81 bp, and P21-Rac1 139 bp. Electrophoresis was performed in 2% agarose at 70 V for 2 hr. Bands were visualized using ethidium bromide. (B) Schematic representation of the Bn deletion and BAC contig spanning the deletion. The BACs 131P17, 561K10, and 373A10 were isolated from a 129SV murine BAC library (Research Genetics) and BAC 448L13 from the C57BL/6J RPCI23 library (25). Physical distances are drawn roughly to scale. The markers DXChri1 – DXChri5 (Genbank Accession Nos. AZ044550 – AZ044554) represent STS’s generated from sequenced ends of BAC clones. Sequence from the left ends of BAC 313P17 and 561K10 overlap and have been designated as a single locus (Chri1). The primer sequences have been deposited in the Mouse Genome Database. The orientation of the contig with respect to the centromere is not known. The orientation of the ends of BAC 448L13 with respect to the rest of the contig is not known, and they have not been placed on the contig map.
36 A DXChri1 Zic3 DXMit208 DXChri4 P21-Rac1
N BNBNBNB N B
200 bp
100 bp
80 bp
B 3 5 Chri1 Chri2 Chri Chri4 Chri
Zic3 Bent tail male
313P17 200 kb
561K10 170 kb
448G13 225 kb
373A16 25 kb 60 kb
37 A B
Li Li S
Sp R L R L
C D
H H
R L R L
Figure 2.3. Situs abnormalities in Bn mice. (A) Normal abdominal situs with the liver (Li) on the right (R) and the stomach (S) and spleen (Sp) on the left (L). The gall bladder is under the liver, and the appendix, which is fixed in the right lower quadrant, are not visualized. (B) Situs ambiguus of abdominal organs in a 6- week-old Bn male. Not the left-sided placement of the liver. The gall bladder and appendix are also on the left (not seen). The normal left-sided placement of the stomach and spleen is obscured by the liver. (C) Normal lung lobulation, anterior view, with one lobe on the left and four on the right. H, heart. (D) Partial left pulmonary isomerism in a Bn female with four lobes on the right and two lobes on the left. The fissure separating the two left lobes is shown by an arrow.
38 CHAPTER 3
LACK OF MUTATIONS IN ZIC3 IN THREE X-LINKED PEDIGREES WITH NEURAL TUBE DEFECTS
Neural tube defects (NTDs) are one of the most common human congenital malformations, occurring once in approximately 1000 live births in the US [39, 95].
The majority of human NTDs are presumed to be multifactorial in origin although chromosomal, monogenic, and teratogenic causes are also recognized (reviewed in [39]). While the introduction of prenatal screening and periconceptual folic acid supplementation have resulted in a dramatic decline in the incidence of NTDs, the basic mechanisms involved in their pathogenesis remain largely unknown.
To help identify genes that may play a role in the pathogenesis of human
NTDs, researchers have examined whether genes associated with NTDs in the mouse may also contribute to the genetic risks for human NTDs. One known example is mutations in the Pax3 gene. A murine loss-of-function mutation of
Pax3 results in the Splotch (Sp) phenotype [96], while mutations in the human ortholog cause Waardenburg syndrome types I and III (WSI and WSIII) [97, 98].
Recently, we [29] and others [99] have demonstrated that a deletion of the murine Zic3 gene is associated with the X-linked, semi-dominant mouse mutation
Bent tail (Bn). Affected hemizygous Bn males have a short-kinked tail and 39 reduced viability and fertility [78]. In addition, approximately 10% of the non- viable Bn males have exencephaly, which is equivalent to human anencephaly.
Other birth defects, including omphalocele, minor skeletal anomalies and cleft lip,
have also been reported in Bn males [99]. Affected heterozygous Bn females have a variable phenotype, consistent with random X-inactivation.
The ZIC3 (Zic3) gene is a C2H2 zinc-finger transcription factor that is
related to the Drosophila gene odd-paired [86]. It is expressed in developing
mesoderm in the mouse by day 7.5 post-coitum (pc) and in the developing
nervous system and tail bud [87]. Its cellular targets, however, have not been
elucidated. Mutations in the human ZIC3 gene have been identified in several males with left-right axis malformations [28]. Several of these affected males had anal anomalies; two additional males demonstrated sacral agenesis; and a final affected male had an open lumbosacral NTD [28]. Consistent with the presence of mutations in the human ZIC3 gene in isolated males and in families with X- linked heterotaxy [28], Bn males and females have a variety of abnormalities of situs, including dextrocardia and left-sided or midline liver and gallbladder [29].
Based on the findings of exencephaly and tail defects in Bn mice and the presence of sacral anomalies and NTDs in some males with human ZIC3 mutations, we examined the human gene for mutations in three families with
NTDs and an X-linked pattern of inheritance. Although linkage studies and haplotype analysis in this first family appear to exclude the region containing the
ZIC3 gene (Xq26) [100] they were included in of studies because of the ease of
40 sequencing the gene and to confirm the mapping results. Pedigrees for the other two families, which have not been published previously, are shown in Fig. 3.1.
The human ZIC3 gene contains 3 exons comprised of 1059 bp, 163 bp, and 179 bp, respectively. PCR primers were designed from published intronic genomic sequence (Genbank Accession No. AF028706) to amplify exons 2 and
3. Because of the large size of exon 1, it was subdivided into 6 overlapping PCR fragments (Table 3.1). Genomic DNA was isolated from an obligate carrier female and from a male with a lumbosacral meningomyelocele from the Icelandic pedigree, individual III-2 from the London family, and individuals II-7 and III-3 from the Australian pedigree (see Fig. 3.1). PCR amplification was performed as described in Table 1 and automated DNA sequencing performed on an ABI377 sequencer as described [101].
No mutations or polymorphisms were identified in the ZIC3 gene in the patients studied, excluding it as a cause of the neural tube defects in these pedigrees. It is still possible that mutations or variation in the expression of ZIC3 could be responsible for NTDs in other families with X-linked patterns of inheritance, as well as in selected sporadic males with NTDs. Further, the contribution of variations in ZIC3 to the pathogenesis of common multifactorial
NTDs remains to be determined. The primers and sequencing conditions described here can now be applied to the larger population of sporadic NTD patients to begin to address these issues.
41 PCR segment Forward primer Reverse primer Annealing temperaturea
exon1; segment A GGCACTTCGGAGATCTCCTC GTGATGGTGGTGGTGATGG 55o C
exon1; segment B GCATGGGGCTGAATCCCTTC GAAACTCGCGCGTTGAGTTG 54o C
exon1; segment C CCATCATCACCACACCAGCC GTTGTTGTCCACGTGCCCAG 58o C
exon1; segment D CTGCATGAGCAGGGCGCTG CCTGCTTGATAGGCTGCCGC 58o C
exon1; segment E GAACATGGGAGTGAACGTGG CCCAGTAGCAGACGTGGTTG 57o C
exon1; segment F CAGCACCATGCATGAGCTGG TCTAAGTCCGCTGCGTGACG 55o C
exon2 CGGCGCTGACCGTATTTTAC GCCGACCGCTAATAAAGAGG 50o C
exon3 GCCAGTTCAGGCTATGAATC GGTGCTAGTTTGAACTGCAG 57o C
Table 3.1 ZIC3 PCR parameters and primers.a PCR reactions were carried out in 25 µl total volume with 100 ng patient DNA, 1.5 mM MgCl2, 0.4 µM primers, 0.2 µM dNTPs and 1 unit of Amplitaq Gold (Perkin Elmer). PCR parameters for each set of primers are as follows: 95o C for 12 min for one cycle; denaturation at 95o C for 30 s, then anneal for 30 s and extension at 72o C for 30 s for 35 cycles; and final extension at72o C for 7 min for one cycle.
42 Australian
I 1 2 3
II
1 2 3 4 5 6 7
III 1 2 3
London
I 1 2 2a II 1 2
III
1 2
Unaffected male Unaffected female Spina bifida occulta Club foot and Spina bifida cystica spina bifida occulta
Figure 3.1. Pedigrees of unpublished human X-linked neural tube defect families.
43 CHAPTER 4
SMAD2 AND TGIF WORK IN CONCERT TO CAUSE MURINE HOLOPROSENCEPHALY.
4.1 Introduction
Transforming growth factor beta (TGF-β) superfamily members are responsible for a wide array of biological functions. The more than 30 family members, which include TGF-βs, bone morphogenetic proteins (BMPs), Nodals
and activins, are secreted molecules that signal through complexes of type I and
type II serine/threonine kinase receptors. The type II receptors are responsible
for ligand binding, recruitment, and activation of the type I receptors, which then
activate downstream cascades (Fig. 1).
Smad proteins are substrates for the type I receptors, and perform as
intracellular signaling molecules in the TGF-β pathway. Originally identified in
Drosophila (mad) [102] and C. elegans (sma) [103], the mammalian Smad family
contains eight proteins with similar protein structures, and can be split into three
functional groups. The receptor-mediated (R) Smads, include Smads 1, 2, 3, 5,
and 8. These are phosphorylated by the type I receptor, which allows the R-
Smads to be transported to the nucleus to carry out transcriptional regulation
(Fig. 1). The common-mediator (Co) Smad, Smad4, accompanies this nuclear 44 localization. Finally, Smad6 and Smad7 are classified as inhibitory (I) Smads,
acting through negative feedback mechanisms to regulate TGF-β signaling
(reviewed in [104, 105]).
Structurally, the Smad proteins consist of three primary domains, the MAD
homologous region (MH) 1 domain, the MH2 domain and the linker region. The
N-terminal region, which contains the MH1 domain, acts in nuclear localization
and transcriptional regulation by binding to DNA. This region is highly conserved
between the R-Smads and the Co-Smad, but weakly with the I-Smads. The MH2
domain, on the C-terminus, allows for protein-protein interactions between
Smads, and also between the Smads and type I receptors. Mediators and
cofactors of Smad regulation can also bind to this region, which is highly
conserved between all Smad proteins. Lastly, the linker gives each Smad
protein its unique character due to the variability of this region, and plays a role in
nuclear localization (reviewed in [104-107]).
As illustrated in figure 4.1, activation of Smad proteins occurs after TGF-β or
BMP molecules interact with the type II receptors. This binding recruits the type I receptor, leading to its phosphorylation. Upon activation, the type I receptor can then phosphorylate the R-Smad proteins at a C-terminally positioned SSXS motif. If the ligand that activates this process is a TGF-β, activin or nodal, the R-
Smad that goes on to conduct intracellular signaling is Smad2 or Smad3. In the case of BMP ligands, Smads 1, 5 or 8 mediate signaling. Once phosphorylated, the R-Smad is released from its plasma membrane anchor (SMAD anchor for receptor activation or SARA is one such example) and oligomerizes with Smad4. 45 Accompanied by Smad4, the R-Smads can then be localized into the nucleus to
carry out the function of transcriptional regulation by activating or repressing
target genes (reviewed in [104, 105, 107]). The intracellular signaling of Smad
proteins is critical for a large number of developmental pathways such as
angiogenesis, organogenesis and patterning (reviewed in [107]). Often, the transcriptional regulatory activities of the R-Smads are accompanied by other transcription factors. For instance, CREB, CBP, FoxH1 and p300 act as co- activators, and Sno, Ski, SNIP1 and Tgif act as co-repressors.
5’TG3’-interacting factor, or Tgif, was isolated from a human liver cDNA
expression library, due to its ability to interact with a CRBPII promoter construct
containing a retinoid DNA-responsive element [108]. Through homology of
amino acid sequences, Tgif has been determined to belong to the TALE
superfamily of homeodomain proteins, which are unique due to an insertion of 3
amino acids between the first and second alpha helices of the encoded protein
[109]. In vitro binding assays demonstrated that Tgif was able to compete for
binding with the retinoid X receptor (RXR)-α and therefore repress 9-cis-retinoic
acid (RA) dependent gene activation [108]. Further studies have shown that Tgif
is responsive to the TGF-β pathway, acting as a repressor of transcription by
interacting with activated Smad2 [110] and recruiting histone deacetylases
(HDACs) [111], thus impeding transcriptional machinery from binding to DNA.
Recently, TGIF has been added to the list of genes correlated with
holoprosencephaly (HPE) in humans [112]. The patients evaluated in this study
encompassed a number of mutations, many resulting in loss of TGIF function. 46 Holoprosencephaly (reviewed in section 1.2) is a congenital malformation of the rostral brain, resulting in a spectrum of midline craniofacial defects. Many genes have been associated with human HPE, and further evaluated in murine models
(reviewed in [62, 63]). Phenotypic evaluation of the patients with mutations in
TGIF revealed that they displayed anomalies within the three classes of HPE: lobar, semilobar and alobar.
Currently, there has been little work reported in mouse investigating the role of the TGF-β pathway in the genesis of HPE. As reported in Liu, et al. [113] half of all murine embryos heterozygous for null alleles of both Smad2 and Smad3 display severe craniofacial defects, equivalent to HPE (Fig.4.2). Furthermore, mice double heterozygous for mutations in Smad2 and nodal display craniofacial defects among many other phenotypes [33]. These data further support a role for the TGF-β/Smad pathway in forebrain development. Currently, the link between TGF-β and Tgif in this process is not understood. The following is a summary of our evaluation of this relationship.
4.2 Results
4.2.1 Frequency and appearance of holoprosencephaly in Smad2; Tgif embryos.
The laboratory of David Wotton at the University of Virginia has recently generated a knockout of the Tgif gene. Assessment of these mice, unlike in 47 humans, has yet to yield any evidence of holoprosencephaly (D. Wotton,
personal communication). To determine if reducing TGF-β signaling in
conjunction with a loss of Tgif can lead to HPE, mice carrying mutant alleles of
both Smad2 and Tgif were crossed and the resultant progeny were evaluated at
embryonic day (E) 10.5. No HPE was observed in mice that harbored mutations
in Tgif only after the analysis of 13 litters (85 total offspring) (Table 4.1).
However, the combination of Smad2 and Tgif mutations did in fact result in HPE
(Table 4.1). This occurred at a rate of 35% in Smad2+/-; Tgif+/- and 26% in
Smad2+/-; Tgif-/- embryos (Table 4.2). (Because HPE is evident in both
Smad2+/-; Tgif+/- and Smad2+/-; Tgif-/- embryos, from this point, the phrase
“Smad2; Tgif mutant embryos” will integrate both genotypes.) While it is
noticeable that the frequency is lower in those homozygous for Tgif mutations,
this may be a due to a loss of some of the Smad2+/-; Tgif-/- due to other
abnormalities earlier in development. While it has been established that embryos
that are homozygous for mutations in Smad2 do not survive to gastrulation [31],
most heterozygous animals are viable, fertile and have no obvious anomalies.
(Recently, it has been demonstrated that 10% of Smad2+/- embryos exhibit
gastrulation defects). Furthermore, it has been demonstrated that crosses
between Tgif+/- animals result in the appropriate Mendelian ratio of Tgif+/+, Tgif+/-
and Tgif-/- viable and fertile progeny (M. H. Festing, unpublished results).
Because of this, we are confident that further embryonic evaluation would show that this deficit is due to the sample size, and not due to embryonic lethality.
Statistical analysis supports this hypothesis (p=0.249). Moreover, embryos
48 displaying HPE have also been observed on the day of birth (Fig. 4.3 A) and embryonically at E12.5 (Fig 4.3 B).
All of the phenotypes manifested in human HPE are also evident in
Smad2; Tgif affected embryos (Fig. 4.3 and 4.4). Mild embryos mimic patients
with lobar HPE have variable separation of the forebrain, but are spared of facial
abnormalities (not shown). Evaluating the least severely affected embryos via
histological sections showed that the prosencephalon appears to have initiated the bifurcation process, but it is incomplete (Fig. 4.4C and D). Semilobar HPE was also observed in the mutant embryos. This group of embryos displays slightly less bifurcation of the forebrain (Fig.4.3 D), and mild facial anomalies.
Finally, the presence of a proboscis and cyclopia were observed in some of the
Smad2; Tgif mutant embryos. These also displayed a single holosphere rostrally
(Fig. 4.3 D and 4.4A). Looking histologically at the affected embryos, it was evident that the telencephalon was malformed, and hypercellular. In figure 4.4 A, histological sections reveal that the telencephalon in the mutant embryo appears to be separated from the adjacent midbrain, unlike the normal littermate.
Subsequent slides do show that there is a connection between the telencephalon and the midbrain, however, it is reduced in size (not shown). Likewise, the forebrain region is much smaller in the mutant embryos versus the wildtype embryo, while the midbrains are comparable in size (Fig. 4.4A). This suggests that structurally, the only region of the brain in the holoprosencephalic embryos that is malformed is the forebrain. One embryo that was histologically evaluated did display clefting of the palate (Fig. 4.4 E). This single embryo is not an
49 accurate representation though of the frequency that clefting may occur. Along
with HPE, a few embryos exhibited additional phenotypes, such as hindbrain
defects (Fig. 4.5 A), axial defects (Fig 4.5 B) and hemorrhaging (Figure 4.5 B).
There are at least two possible explanations for the hypercellularity observed in the Smad2; Tgif mutants: reduced apoptosis or increased proliferation. TUNEL assays were utilized to evaluate the level of apoptosis in the mutant embryos. The results of these assays showed that there was only a
2.4% decline in apoptosis when comparing the HPE embryos to normal littermates (Fig. 4.6). This change from 8.1% TUNEL positive cells in aphenotypic embryos, in comparison to 5.7% was shown not to be significant via chi square analysis (p=0.575). Analysis of the mitotic index via BrdU incorporation did reveal a considerable increase in the rostral nervous of the mutant embryos (Fig. 4.7). HPE embryos had 16% more BrdU incorporation than normal littermates, revealing that the hypercellularity was due to an increase in cellular proliferation.
4.2.2 Molecular analysis of Smad2; Tgif holoprosencephalic embryos.
To investigate the molecular mechanisms for the initiation of HPE in the
Smad2; Tgif embryos, a series of molecular markers were assessed via whole mount in situ hybridization at E9.5. Two of the first genes investigated have been associated with HPE in humans, ZIC2 and SIX3 [67, 68]. As discussed previously (Chapters 2 and 3), the Zic family is comprised of presumed transcription factors containing C2H2 type zinc finger domains [86]. A mouse
50 hypomorph of the Zic2 gene, like humans, does display HPE [114]. Similar to the
human patients with SIX3 mutations, mice lacking this homeobox gene have
posteriorized heads, corresponding to a rostral expansion of the Wnt1 domain
[69]. The expression of these two genes was unaffected in Smad2; Tgif embryos
displaying reduced forebrains (Fig. 4.8 A and B). It is worth mentioning that while
the intensity of Six3 expression is equal in the affected embryo in comparison to
its littermate, the area of expression is reduced. While it is possible that this is
due to a partial down regulation of Six3 expression, it is more likely that this
decline is due to a loss of the tissue where endogenous expression occurs.
Previous studies have shown that Fgf8 is required for the specification of
the forebrain [115, 116]. Furthermore Fgf8, expression is lost in the frontonasal
placode of Smad2+/-; Smad3+/- embryos displaying craniofacial abnormalities
[113]. Molecular analysis revealed that the frontonasal placode expression of
Fgf8 in the Smad2; Tgif mutant embryos was unchanged (Fig. 4.8 C). Like Fgf8,
rostral of Shh is lost in Smad2+/-; Smad3+/- embryos [113]. Furthermore,
mutations in SHH and a number of genes which signal downstream of it, have
been associated with HPE in humans and mice (reviewed in [62, 63]).
Evaluation of Shh yielded no modification of expression even in severely effected embryos (Fig. 4.8 D). The unaffected expression of Fgf8 and Shh differ from that of the Smad2+/-; Smad3+/- embryos that exhibit craniofacial defects previously
described, suggesting that the phenotypes arise by different mechanisms.
Arx is a homeobox gene that has been previously shown to be expressed
in the developing forebrain [117, 118]. While expression of arx is decreased in
51 aphenotypic Smad2; Tgif embryos (Fig. 4.8 E), mutations in ARX have been
associated with mental retardation in humans, not HPE [119, 120]. Two groups
have generated a targeted null allele for murine arx. The first reported knockout
of arx resulted in the reduction of the overall size of the brain, olfactory bulbs, and testes, but patterning of the brain was correctly established [120]. In the second report, no brain anomalies were observed. These mice were deficient only in pancreatic development [121]. Furthermore, unlike the Smad2; Tgif
mutant embryos, the lack of Arx in embryos was associated with a decrease in proliferation in the brain [120]. The combination of these reports suggests that
the reduction of arx observed here is not the causative event that leads to the
onset of HPE in the Smad2; Tgif mutant embryos. However, either Smad2 or
Tgif may regulate the expression of Arx, and this loss may lead to the onset of a
later phenotype.
The homeobox gene Otx2 is essential for the proper development of the rostral neuroectoderm (reviewed in [122]). All phenotypic embryos whose Otx2
expression was evaluated through the course of this study had moderate to
complete loss of Otx2 in the forebrain region and, occasionally, a striking loss in
the midbrain (Fig. 4.9). Similarly, a reduction of Otx2 can be induced by
exogenous RA [123, 124]. As mentioned previously, Tgif was initially identified
by its ability to compete with the RXR for its binding site [108], providing a
possible mechanism for the reduction in Otx2 in the Smad2; Tgif embryos.
52 4.2.3 Retinoic acid and holoprosencephaly in Smad2; Tgif embryos.
A derivative of vitamin A, retinoids have been demonstrated to be required for embryonic development, and also play a role in teratogenesis
(reviewed in [125]). Retin-A, Vesanoid and other retinol containing drugs are used pharmacologically for two different purposes. Retin-A and other dermatological ointments have been commonly used for many years as agents to treat acne and other skin problems. Vesanoid on the other hand, is used as a means to “induce cytodifferentiation and decrease proliferation… in remission patients with acute promyelotic leukemia” [126]. A report by Lammer et al. [127] described a group of 59 women who were exposed to a form of RA during pregnancy. Of this group, nearly half had no problems during their pregnancies, which resulted in healthy babies. While 12 of the pregnancies ended by spontaneous abortion, the outcome of 21 was children with severe anomalies.
Most of these incidences were malformations of the midline and central nervous system. The term applied to this group was RA embryopathy [127]. Since this initial report, a number of other groups have given accounts of similar findings
[128-131].
The similarity of the phenotypes of the animals evaluated here and the progeny of the patients receiving RA treatments, as well as the loss of Otx2 gene expression led us to hypothesize that the malformation of the forebrain in the
Smad2; Tgif embryos was linked to RA signaling. We therefore utilized a line of mice that harbors three copies of the retinoic acid receptor β (RAR-β) promoter upstream of the bacterial lacZ gene [132] to evaluate the domain of activity of RA
53 during the developmental stage in which we see a loss of Otx2. In figure 4.10 A,
X-gal staining is evident in the bifurcating telencephalon and developing eye, as
well as dorsally in the region of the somites. Sagittal sections through the
forebrain show that the area of expression of Otx2 (Fig. 4.9 B) and RA signaling
(Fig. 4.10B) is identical in both, suggesting that RA signaling occurs in the region
of the forebrain where Otx2 expression is lost in the Smad2; Tgif mutants.
Previous publications have reported relationships between Tgif and RA,
as well as TGF-β and RA [133-135]. All of these results were derived by using in vitro systems. Currently, there are no reports that investigate the interactions of these pathways during embryonic development. Therefore, we utilized previous reports [136] of RA teratogenesis to evaluate such interactions. Pregnant mice that carried either mutations of Tgif, Smad2, or Smad3 received intraperitoneal
injections at gestational day E7.5 with 7.5 mg all trans RA (ATRA) per kilogram of
maternal weight, and the embryos were evaluated 2 days later at E9.5.
This analysis revealed an increase in sensitivity to RA with the addition of
mutations in Tgif. In addition to other defects, the occurrence of HPE and forebrain reductions were markedly increased with the application of RA in Tgif-/-
embryos in comparison to those that are Tgif+/- (S. H. Lasse, unpublished results). Similarly, Smad2 embryos exposed to retinoic acid also showed a sharp increase in the incidence of HPE in comparison to wildtype littermates (Fig 4.11, and Table 4.3 A). While no wildtype embryos from the litters that received
excessive RA during development had abnormalities of the forebrain, 70% of the
resultant Smad2+/- embryos did display HPE. There was not a significant
54 difference between wildtype and Smad2+/- in regards to frequency of
exencephaly observed in these embryos (Fig. 4.11 A and Table 4.3 A). These
data suggest that there is an interaction between the Smad2 and retinoic acid
pathway, revealing a genetic component allowing for the sensitivity to a teratogen
and leading to HPE.
To determine if this is specific to Smad2 or due to a reduction in TGF-β signaling, the experiment was repeated in the presence of Smad3 mutations. No incidence of HPE was apparent in the Smad3+/- (Fig. 4.12 and Table 4.3 B).
This suggests that the association of Smad signaling and RA is specific to
Smad2. RA does not interact with Smad3 to regulate forebrain development
4.3 Conclusions
The compilation of the results in Table 4.1 obtained by breeding mice with
targeted mutations in Smad2 and Tgif and examining progeny at E10.5 illustrates
that the combination of alleles does indeed result in the occurrence of HPE in
one-third of the ensuing progeny. Unlike that which has been observed in
humans, we have yet to identify HPE in any form in animals that have solely Tgif
mutations. The difference in these observations may be a divergence of
developmental pathways of mice versus humans. On the other hand, the human
patients displaying HPE and harboring Tgif mutations may have defects in TGF-β
signaling, which have yet to be detected. The latter hypothesis was suggested
by the work published by Gripp, et al. [112], in which they speculated that the
TGIF-SMAD2 interaction was disrupted in the patients. 55 In addition to HPE, other defects are occasionally observed in the mutant
animals. The frequency of the occurrence of each of these anomalies (hindbrain
reductions, axial defects and hemorrhaging) is low, but is not surprising, due to
the role of TGF-β pathway in hindbrain patterning [137-139], A/P patterning [31-
33], and vasculogenesis/angiogenesis [140, 141]. Furthermore, hemorrhaging
was reported in 60% of patients taking Vesanoid [126]. It is possible that
disruptions in Smad2 and Tgif signaling may lead to slight alterations in the
signaling of other TGF-β pathway molecules.
It was somewhat surprising that the mutant embryos were observed as being
hypercellular with increased cellular proliferation, because it is common for
retinoid containing drugs to be used as a means to reduce proliferation in some
cancer patients. However, a recent abstract by Bhasin, et al. [142], states that
the exposure of frontonasal mass explants to RA results in an increase of cellular
proliferation. Therefore, it seems that RA may play different roles in regulation of
the cell cycle in a tissue specific manner. Furthermore, it appears that while this
region has an abundance of cells, the cells are not receiving the proper cues for
patterning and differentiation.
To determine if mutations in both Smad2 and Tgif resulted in altering known
HPE genes, Six3, Shh and Zic2 were evaluated in mutant embryos. The lack of
a change in all three of these genes illustrates that the mechanism behind the
development of HPE in Smad2; Tgif mutant embryos is different than those
previously described. Therefore, it was necessary to evaluate these embryos to
determine the source of the onset of HPE. 56 Initially, two hypotheses were developed to decipher the mechanism behind the phenotype. The first hypothesis is that it is possible that the development of
HPE in Smad2; Tgif embryos is due to a failure of A/P axial specification early in development, such as a signaling defect from the AVE. This seems unlikely, because the embryos do develop a proper body structure in terms of head, trunk and tail, and do not show any observable defects by E8.5. To assess this hypothesis molecularly, we evaluated the expression of the Shh gene. Even in the most severely phenotypically affected, Shh expression was unchanged along the notochord and floor plate. This suggests that initiation of the A/P axis is not disrupted in these embryos, hence early initiation of this axis is unaffected.
Evaluation of Hex expression, at the stage where the AVE is established will verify that the AVE is properly established in these mutant embryos.
Even though initiation of the A/P axis was correctly initiated, the second hypothesis of the manifestation of the phenotype could be attributed to a loss of later signals for maintaining anterior identity. To evaluate this hypothesis, genes that are expressed in the anterior/rostral most tissues were examined. Both Fgf8 and Six3 provide evidence that anterior identity is maintained within the rostral forebrain. This statement is supported by the presence of Fgf8 in the frontonasal placode in the Smad2; Tgif phenotypic embryos. The most rostral expression of
Fgf8 is suggestive that while the forebrain of these embryos is severely malformed, the rostral tissues have been specified, but are not receiving the signals for proper development. We observed a potential proboscis in three of eleven (27%) embryos with HPE or severe forebrain reductions. It is possible
57 that in embryos that will later develop a proboscis, Fgf8 expression may be lost.
Fgf8 expression was evaluated in three phenotypic embryos. All three exhibited
normal expression. Perhaps evaluation of more litters may reveal a loss of Fgf8 anteriorly. However, the occurrence of a phenotypically affected embryo at E9.5
(the developmental stage in which all of the whole mounts were preformed) is rare, and represents what we believe to be, the most affected embryos, which includes those that will later possess a proboscis.
Results obtained from the evaluation of the Six3 expression further supports that the cues to maintain anterior cell fates have not been disrupted. Oliver, et al.
[143] they reports that the region in which Six3 expression was observed is analogous to the region that the avian fate map has shown later contributes to
“the most anterior neural derivatives”. Because no change in Six3 expression was observed in the Smad2; Tgif mutant embryos, it can therefore be concluded that, in agreement with the results obtained from Fgf8 analysis, the anterior most tissue was in fact specified and has been maintained.
The above marker analysis was intriguing in light reports regarding Smad2+/-;
Smad3+/- embryos that display craniofacial defects. These mutants are deficient
in Fgf8 expression in the frontonasal placode, and have a truncation in the
anterior Shh expression [113]. In addition, it has been reported that nodal mutations can affect Shh signaling, and, as previously stated heterozygous mutations in Smad2 and nodal results in HPE. Because Smad2; Tgif embryos
present normal Shh expression, it suggests that the HPE observed here does not
arise by the attenuation of nodal signaling. Additionally, Otx2 expression is
58 unaffected in the Smad2+/-; Smad3+/- embryos [113]. These data suggest that
the HPE in the Smad2; Tgif mutant embryos results from a mechanism that is unique from those previously described, therefore, a new avenue had to be explored. Previous reports have shown that exogenous RA is able to downregulate the expression of the Otx2 gene in wildtype mice [123, 124].
Interestingly, Tgif was initially identified by its ability to compete for binding with
elements of retinoid signaling [108]. Merging our data, these earlier findings lead
us to investigate the possibility that RA was involved in the development of HPE
in the Smad2; Tgif mutant embryos.
In the teratogenic studies, the dose of retinoic acid given to the pregnant
animals, 7.5 mg/kg body weight, which is equivalent to 22.5mg/m2/dose [144],
parallels the dosage used by others in similar reports [136]. Moreover, it is a low
dose compared to that given to humans for dermatological or anticancer treatments. The product information sheet provided with Vesanoid states that
the maximal tolerated human pediatric dose is 60 mg/m2/day, and more than three times higher in adults, while the recommended dose is 45 mg/m2/day for
30-90 days [126]. Therefore, we can state that the phenotypes analyzed in the
embryos whose mother had been administered RA were due to teratogenic
affects of a pharmacological amount of RA, and not due to intolerable dose.
However, the means of delivery does differ between humans and this study.
While humans receive topical or oral doses of retinoids, intraperitoneal injections
were used here, making an accurate comparison difficult. By measuring the
59 maternal serum or blood levels of retinoids, we could make a more precise
correlation between the levels od exposure during development.
Here, we have reported that an E7.5 administration of RA to pregnant
mice results in no grossly notable change in the Smad3 mutant embryos. In light
of recent reports, it is surprising that there is no effect on these embryos. In a
paper published by Pendaries and colleagues [134], it was shown that a direct
interaction occurs between Smad3 and RARγ, and this interaction could be
altered by agonist or antagonists of retinoid signaling. Thus, it would seem likely that exogenous RA would result in embryonic anomalies in the Smad3 mutants,
however, as shown here, this is not the case. Taking into consideration these
results, it can be hypothesized that the interaction between Smad3 and RA is not
important in embryonic development. Similarly, there is no evidence of
embryonic abnormalities at E10.5 in Smad3+/-; Tgif+/- or Smad3+/; Tgif-/- mice (S.
H. Lasse, unpublished results), paralleling the teratogenesis data.
Unlike the teratogenesis results with Smad3, RA exposure to Smad2
mutant embryos does result in anomalies. While the same experiment in Tgif
embryos produces a range of RA induced abnormalities (S. H. Lasse,
unpublished results), only HPE is generated in the Smad2+/- embryos. The
manifestation of HPE does suggest that an interaction between RA and Smad2
occurs. Taking into consideration the teratogenesis results, repeating the
experiments of Pendaries, et al. [134] should show similar or enhanced results
with Smad2 in comparison to that obtained with Smad3. In addition, because the
only phenotype observed in the Smad2-RA embryos was HPE, it is likely the
60 Smad2-retinoid interaction is specific to forebrain development, or necessary for
anterior neural specification. Nevertheless, this opens new questions. In what
manner does this interaction occur, and is it dependent on or independent of
Tgif? Regardless of the mechanism of interaction, this is a rare finding in which genetic predisposition can allow for a change in the sensitivity to teratogenic
agents.
In regards to Tgif, it has been shown that Tgif is capable of competing with
the RXR at the RXR binding element in natural and artificial promoters, acting as
a repressor of transcription. Applying this knowledge to define a mechanism for
the occurrence of HPE in the Smad2; Tgif mutant, it is possible that in the
absence of Tgif, RA signaling is increased. Therefore, blocking Tgif would
induce gene expression regulated by retinoid signaling. Initially, reports stated
that Tgif acted as a repressor of transcription [110, 111], suggesting that Tgif
would not directly aid in the regulation of Otx2. However, more recently, work in
Drosophila has shown that Tgif is capable of functioning as an activator [145].
Otx2 is such a gene that could be differentially regulated by Tgif and RA. In this
situation, RXR may bind in place of Tgif, thus blocking the HDAC normally
recruited by Tgif, affecting the transcriptional regulation carried out by Tgif.
Preliminary analysis of a 2119 bp region upstream of the Otx2 gene contains
retinoid response elements as well as the half sites that was shown to allow for
competition between the RXR and Tgif [108] (Fig. 4.13). These domains are
conserved between mouse and humans (data not shown). The presence of
these domains implies a possible role of RA and Tgif in regulating Otx2
61 expression. If Tgif is acting as an activator in this model applies to the model represented here, then it is possible that Tgif directly regulates the expression of
Otx2. However, it is very possible that the repression of Otx2 occurs down
stream of other gene(s) that are transcriptionally regulated by competition
between Tgif and Smad2.
Data presented here is the first to demonstrate an in vivo correlation between
Tgif and retinoic acid. This strengthens the initial finding that this correlation is
competitive, and allows for regulation of gene transcription. Additional data
demonstrate that unlike previous reports, TGF-β signaling, through Smad2, plays
a protective role against retinoic acid during development. While it is likely that
Tgif and Smad2 are working through independent mechanisms, it is probable
that the mutations work in conjunction to alter retinoid signaling, resulting in HPE.
4.4 Materials and Methods
4.4.1 Mice and Matings
Mice carrying either Smad2 [31] or Tgif mutations were intercrossed and
the resultant progeny were used to maintain offspring of the following genotypes:
Smad2+/-; Tgif+/+, Smad2+/-; Tgif+/- or Smad2+/-; Tgif-/-. These breedings were carried out housing one male and two females, given water and food ad libitum.
To examine embryonic development at various stages, matings were timed with the morning of the visualization of the vaginal plug being E0.5. At the
62 appropriate stage of development, the embryos were dissected from their
maternal decidua in 1X phosphate buffered saline (PBS) (pH 7.4), and the
extraembryonic membranes were removed, keeping the extraembryonic portion
for genotyping. The embryos were placed in 4% paraformaldehyde (PFA) and
gently agitated at 4o C overnight. The following morning, the embryos were
dehydrated using an ethanol series (25, 50, 75, 95 and 100%), washing in each
concentration of ethanol for 20 minutes. (Embryos can be stored at -20o C for several months in 95% ethanol.) Gross evaluation of the embryos was preformed using a Zeiss Stemi SVII Apo dissecting microscope. The embryos were photographed employing a MTI 3CCD digital camera attachment and Scion
Series 7 software.
4.4.2. Genotyping
To genotype, standard reaction conditions were utilized in a total volume of 20 µl. The reaction parameters are as follows: 5 min denaturation at 95 O C;
thirty five cycles of 95 O C for 30 seconds, primer specific annealing temperature
(Table 4.5) for 30 seconds and 72 O C for 30 seconds; a final elongation of 10 minutes at 72 O C.
4.4.3. Histology
To prepare the processed embryos for sectioning, the embryos were
passed through two washes of xylene for 20 minutes each. Following, the
embryos were incubated in a 1:1 solution of xylene and paraffin for one hour at
63 57O C. Finally, the embryos were placed into xylene and incubated three times, for one hour each at 57O C. After paraffinization, the embryos were embedded into paraffin blocks, and sectioned to a thickness of 5-7 µm using a Reichart
HistoSTAT Rotary microtome. The sections were placed onto TESPA coated slides, and dried for 2-3 days at 37O C.
4.4.4 Hemotoxylin and Eosin Staining Procedure
Removal of the paraffin and hydration of the slides was carried out by passing them through the following:
Xylene for 5 minutes
Xylene for 5 minutes
100% ethanol for 3 minutes
100% ethanol for 3 minutes
95% ethanol for 3 minutes
70% ethanol for 3 minutes dH2O for 5 minutes
Following the rehydration, the nuclei were stained in hemotoxylin by dipping the slides in the solution briefly, then rinsed in running tap water until the water ran clear. To blue the hemotoxylin, the slides were submerged in acid ethanol until they reach the desired color, and then neutralized in basic ethanol, and finally gently rinsed in running tap water.
64 To stain the cells, the slides were immersed in eosin for 30 seconds, and rinsed in 95% ethanol for 30 seconds. The slides were then dehydrated in preparation for coverslipping by passing through the following series:
100% ethanol for 5 minutes
100% ethanol for 5 minutes
Xylene for 5 minutes
Xylene for 5 minutes
Lastly, the slides were mounted by applying a coverslip adhered with a 1:1 solution of xylene and Permount. After drying the slides in the hood, they were evaluated using Zeiss Axioskop2 light microscope and photographed employing a MTI 3CCD digital camera attachment and Scion Series 7 software.
4.4.5 Bromodeoxyuridine analysis
To analyze the rate of cellular proliferation in developing embryos, pregnant dams were injected intraperitoneally with 1 mg of BrdU per 10 g of maternal weight 2 hours prior to isolating the embryos. The embryos were removed, processed and sectioned as described above. BrdU incorporation was detected using a slightly modified version of the BrdU In Situ Detection Kit II from BD Biosciences protocol, as follows.
Washing in xylene, twice for 5 minutes each deparaffinize slides.
Following the deparaffinization, the slides were dehydrated in a gradient series of ethanol, washing twice in 100% ethanol and once in 90% ethanol, for two
65 minutes each. Next, 3% hydrogen peroxide (in PBS) was utilized to block
endogenous peroxidases, by rinsing the slides for 10 minutes, followed by three
5 minute rinses in 1X PBS.
After placing the working solution of BD Retrieven A (9 ml of solution 1, 41
ml solution 2 in 500 ml distilled water) in a coplin jar and microwaving on high until the solution begins to boil, the slides were inserted into the jar and placed in
a 89O C water bath for 10 minutes. After the incubation, the jar was placed at
room temperature and allowed to cool for 20 minutes, and then rinsed 3 times for
5 minutes each in 1X PBS. Diluted biotinylated anti-BrdU antibody (1:10 in
Diluent Buffer) was placed on top of the tissue and incubated at room temperature in a high humidity environment for 2-3 hours, and again rinsed in 1X
PBS three times, 2 minutes each. The Streptavidin-HRP was applied to each slide, and incubated for 30 minutes at room temperature, and rinsed 3 times, 2 minutes each. The color reaction was carried out with the application of the DAB
chromagen solution (1 drop in 1 ml DAB buffer) to the tissue on the slides.
Reaction time was inconsistent between slides, varying between 1-5 minutes,
depending on the nature of the slide. 1X PBS was used to stop the color
reaction by washing the slides three times, for two minutes each.
To counterstain the non-proliferating cells in comparison to the BrdU
positive cells, the slides were dipped briefly in hematoxylin, and processed and
coverslipped as described above.
66 The slides were evaluated using Zeiss Axioskop2 light microscope and
photographed employing a MTI 3CCD digital camera attachment and Scion
Series 7 software.
4.4.6 TUNEL analysis
To evaluate for the presence of apoptosis, sections of embryos obtained
as described above were examined using the TdT-FragEL DNA Fragmentation
Detection Kit (Oncogene), following the manufacturer’s protocol as follows:
Paraffin sections were deparaffinized and rehydrated as described above to the 70% ethanol step. Following the 70% ethanol, the slides were rinsed in 1X
TBS (20 mM Tris, pH 7.6 and 140 mM NaCl), and the glass was dried without disrupting the tissue. The tissue was next covered in a 1:100 solution of 2 mg/kg proteinase K and 10 mM Tris, pH 8, incubated at room temperature for 20 minutes, and rinsed with 1X TBS. The glass was again dried without disrupting the tissue.
To inactivate endogenous peroxidases, 3% hydrogen peroxide in methanol was applied to the tissue and allowed to process for 5 minutes. Again the slide was rinsed with 1X TBS and dried. Next, the slide was covered with
100 µl of 1X TdT Equilibration Buffer and allowed to incubate for 20 minutes.
Without touching the tissue, the TdT solution was blotted from the slide and 60 µl
of the TdT Labeling Reaction Mixture was placed onto the tissue. To prevent
evaporation or drying, parafilm was placed on top of the tissue. The slides were
incubated in a humidified chamber for 90 minutes at 37O C. 67 To stop the labeling reaction, the slides were rinsed with 1X TBS, followed by applying 100 µl of the stop solution. Thus was allowed to incubate for 5 minutes at room temperature, then again rinsed in 1X TBS and the slide dried.
The tissue was then covered with 100 µl Blocking Buffer for 10 minutes at room temperature, and the Blocking Buffer was carefully blotted away from the tissue, and 100 µl of a 1:50 solution of 50X Conjugate and Blocking Buffer was placed on the tissue. After 30 minutes in a humidified chamber, the slides were rinsed with 1X TBS and again dried. 100µl of DAB dissolved in tap water was applied to the tissue, and incubated for 10 minutes at room temperature, then rinsed with distilled water.
To counterstain the tissue, the slides were covered with methyl green and incubated for 3 minutes. After allowing the dye to run off of the slides, the slides were rinsed well with water. The tissue was covered in glycerol, and a coverslip was placed on top of the tissue. Applying fingernail polish around its perimeter sealed the coverslip.
The slides were evaluated using Zeiss Axioskop2 light microscope and photographed employing a MTI 3CCD digital camera attachment and Scion
Series 7 software.
68 4.4.7 Whole mount in situ hybridization
Preparation of embryos
Embryos were obtained and fixed as described above. The following
morning, the embryos were washed twice in 1X PBS with 0.1% Triton-X 100 for
10 minutes each at 4o C. Dehydration of the embryos was accomplished by
passing them through increasing amounts of methanol (25, 50, 75, 100%) for 10 minutes each. (The embryos can be stored in methanol at this point at –20o C.)
A final wash in 100% methanol is necessary prior to use.
Synthesis of digoxigenin (DIG)-UTP Labeled RNA Probe
In an RNase-free microcentrifuge tube, the following was added: 1µg purified, linear DNA template, 2 µl rNTP labeling mix containing 10mM rATP, rGTP and rCTP, 3.5 mM rUTP and 6.5mM DIG-rUTP, 2 µl 10X transcription buffer, 1 µl RNase inhibitor and 2 µl of the appropriate RNA polymerase (T3, T7 or Sp6). Depc treated water was used to bring the final volume to 20 µl. After mixing well, the reaction was incubated for two hours at 37o C. At the end of the
incubation, 2 µl of DNase I was added and the reaction was incubated for 15
minutes longer.
To precipitate the probe, 2 µl 0.5 M EDTA was added, and the volume of
the reaction was brought to 50 µl with depc water, followed by the addition of 5 µl
of 4 M LiCl, and 130 µl 100% ethanol. Following incubation at -20o C for 30
69 minutes and the reaction was spun in the microcentrifuge. The pellet was then
washed twice with 70% ethanol and dissolved in 100 µl of depc water.
Hybridization of DIG-Labeled probes to embryos
To removed any endogenous RNase activity, the embryos were washed
for two hours in 5% hydrogen peroxide in methanol, and then rehydrated by
passing them through decreasing amounts of methanol (75, 50 and 25%) for 10
minutes each. Next, the embryos were rinsed twice, for 10 minutes each, in
PBTX (PBS, 0.1% TritonX-100). The embryos were then incubated in 2 ml of 10
µg/ml proteinase K in PBTX. The length of these incubation times was variable
depending on the stage of the embryos. (Approximately 7 minutes for E 8.5 and
12-15 minutes for E9.5 embryos.) Following, the embryos were washed twice in
PBTX for 10 minutes each.
To ensure complete fixation, the embryos were placed in fresh 0.2%
glutaraldehyde; 4% PFA, and rocked gently at room temperature for 20 minutes.
Subsequently, the embryos were washed again in PBTX for 10 minutes, twice.
After transferring the embryos to a 2-ml screw cap tube and the tube was filled with hybridization solution (50% formamide, 5XSSC, 2% Blocking Powder, 0.1%
Triton X-100, 0.5% CHAPs, 100 µg/ml of yeast tRNA, 5mM EDTA and 50 µg/ ml heparin). The embryos were allowed to descend to the bottom of the tube, and then the hybridization solution was replaced with a fresh aliquot, and the tubes were incubated for 2 hours at 65o C. At the end of the incubation, 2 ml of pre-
warmed hybridization buffer, containing the labeled riboprobe was added to the
70 tube, replacing the previous solution. Finally, the tubes containing the embryos were incubated overnight at 65o C.
The next morning, the embryos were washed at 65o C for minutes in each of a gradient of Solution I (50% formamide, 5X SSC, 0.1% Triton X-100 and 0.5%
CHAPS), and 2X SSC: 100% Solution I, 75% Solution I/25% 2X SSC, 50%/50% and 25%/75%. The embryos were then washed for 30 minutes twice in 2X
SSC, 0.1% CHAPS and 0.2XSSC, 0.1% CHAPS, at 65o C. Finally, the embryos were washed twice for 10 minutes each in TBTX (50 mM Tris-HCl (pH 7.5),
150mM NaCl and 0.1% Triton X-100) at room temperature.
Anti-DIG antibody (AP-conjugated) Binding
The embryos were pre-blocked for non-specific binding in 1% BSA in
TBTX for 2-3 hours at room temperature. Meanwhile, the pre-absorbed anti-DIG antibody was prepared by first combining 3 mg of embryo powder and 1 µl of anti-DIG-AP FAb (Boehringer) to 0.5 ml of 10% goat serum, 2% BSA and rocking gently at 4o C for 3 hours. After in three hours, the mixture was spun in a microcentrifuge for 10 minutes at 4o C. Finally, the supernatant was diluted to 2 ml with 10% goat serum, 2% BSA.
After the embryos have been pre-blocked for a sufficient amount of time, the blocking solution was removed, and the pre-absorbed antibody solution was added. The anti-DIG antibody binding was incubated at 4o C, overnight. The next morning, the embryos were washed 5 times, in TBTX containing 0.1% BSA
71 for 1 hour each at room temperature. A final wash overnight was facilitated at 4o
C.
Immunohistochemistry
The embryos were washed twice in TBTX containing 2mM levamisole for
15 minutes at room temperature, followed by NTMT (100 mM NaCl, 100mM Tris-
HCl (pH 9.5), 50mM MgCl2 and 0.1% Tween-20) containing 2 mM levamisole, three times for 10 minutes each at room temperature. After the washes, 2 ml of
BM purple AP substrate (Boehringer) was added to each tube containing embryos. The tubes were immediately placed in foil to create a dark environment, and rocked until the color reaction was complete (several minutes to several hours, depending on the abundance of target mRNA.) Washing in
NTMT, followed by multiple washes in PBTX terminated the color reaction. To decrease the background and make the staining sharper, the embryos were washed several times in PBS containing 1% Triton X-100. A final fixation in 0.2% glutaraldehyde, 4% PFA, followed by photographing finished the whole mount procedure. The embryos were either dehydrated by an ethanol series for sectioning, or stored at -20o C in glycerol.
4.4.8 Teratogenic application of retinoic acid
Timed mating were established as described above. On E7.5, pregnant dams were injected intraperitoneally with 7.5 mg of 1 mg/ml retinoic acid in
DMSO per kg of maternal body weight, as recommended in Sulik, et al[136].
72 Embryos were isolated at E 9.5, and photographed and processed as described above.
73 xial defects Number of embryos observed Number of normal embryos Number of abnormal embryo (%) Holoprosencephaly Forebrain/craniofacial defects Hindbrain/midbrain anomalies A Hemorrhaging
Smad2+/+; Tgif+/+ 1 1 0 (0%) 0 0 0 0 0
Smad2+/+; Tgif+/- 27 27 0 (0%) 0 0 0 0 0
Smad2+/+; Tgif-/- 19 19 0 (0%) 0 0 0 0 0
Smad2+/-; Tgif+/+ 1 1 0 (0%) 0 0 0 0 0
Smad2+/-; Tgif+/- 20 12 8 (40%) 6 1 2 0 2
Smad2+/-; Tgif-/- 15 8 7 (47%) 3 1 1 2 2
Table 4.1. Summary of the evaluation of E10.5 Smad2; Tgif embryos.
74 Smad2+/-; Tgif+/- Smad2+/-; Tgif-/-
Holoprosencephaly 30% 20%
Forebrain/craniofacial 5% 6%
Hindbrain reduction 10% 6%
Table 4.2 Malformation frequency in Smad2; Tgif embryos.
75 A HPE exencephaly runted pericardial normal ballooning Smad2+/- abnormal embryos 12 3 1 1 4 percent abnormal 70.6 17.6 5.9 5.9 23.5
WT abnormal embryos 0 5 0 0 12 percent abnormal 0.0 29.4 0.0 0.0 70.6
Total number of litters 5
B HPE exencephaly normal
Smad3+/- abnormal embryos 0 2 18 percent abnormal 0.0 9.1 81.8
WT abnormal embryos 0 1 14 percent abnormal 0.0 6.7 93.3
Total number of litters 6
Table 4.3. Retinoic acid induced anomalies in Smad2 or Smad3 mutant embryos. The columns above represent anomalies observed in embryos exposed to exogenous retinoic acid at E7.5. In some cases, more than one defect was present in the same embryo. Due to the type of breeding used, both wildtype and mutant embryos were evaluated from the same litter. The above tables are accumulation of the litters observed.
76 Molecular Markers Zic2 Six3 Fgf8 Shh Arx Otx2
Smad2+/+; Tgif+/- normal expression 2 1 1 2 altered expression
Smad2+/+; Tgif-/- normal expression 7 1 4 4 8 altered expression
Smad2+/-; Tgif+/- normal expression 2(1) 3(1) 1(1) 2 altered expression
Smad2+/-; Tgif-/- normal expression 3(1) 2(1) 3(2) 4(1) 2 altered expression 5 5(5)
Table 4.4. Embryos evaluated via whole mount in situ hybridization. The values inside of the parentheses represent the number within each group that was phenotypically affected.
77 PRIMER OLIGO SEQUENCE (5’ to 3’) ANNEALING NAME TEMPERATURE
TBW 12 ATCGCCTTCTATCGCCTTCTTG Smad2-4 CATGAATACTACGACGGAGG 55o C
The presence of the targeted Smad2 allele will be indicated by the presence of a band of approximately 650 bp.
Tgif ex 2-5 CCTCAGAGTCACTGCCTGATG int-1 AGTCCAACTGGCCGGAATTGTCAC 60o C GFP low2 TGGTGCAGATGAACTTCAGGGTGAG
The presence of the wildtype Tgif allele is indicated by the amplification of a 558 bp band, while the band of the targeted allele is 690 bp.
Smad3-5 CCACTTCATTGCCATATGCCCTG Smad3-7 CCCGAACAGTTGGATTCACACA 60o C RINA CCAGACTGCCTTGGGAAAAGC
The presence of the wildtype Smad3 allele indicated by the amplification of a 400 bp band, and the mutant , a 300 bp band.
Table 4.5 Genotyping PCR primers and annealing temperature.
78 TGF-β ligand
1 2 3
SARA P- -P -P 4
R-Smad -P 5
I-Smad Smad4 -P
6 Target gene
7
Figure 4.1. The Smad pathway. (1) A TGF-β like ligand binds to a type II receptor, thus (2) recruiting the type I receptor and (3) phosphorylating it. The activated type I receptor is responsible for the events that allow for (4) the R- Smad molecules to be phosphorylated and released into the cytoplasm, where they can (5) bind to Smad4. The heterodimer can then be (6) transported into the nucleus where it can (7) act as a regulator of gene expression. I-Smads function by blocking the activation of the R-Smads by the type I receptor.
79 A B
Figure 4.2. Craniofacial defects in Smad2+/-; Smad3+/- E12.5 embryos. In comparison to a normal littermate (A) the proboscis and cyclopia of the mutant embryo (B) are evident. Less predominant is the semitransparent holosphere of the forebrain (arrow).
80 C D
A B
C D
*
Figure 4.3. Holoprosencephaly in Smad2; Tgif liveborn pups and embryos. (A) A Smad2+/-; Tgif+/- newborn with a proboscis, and perhaps cyclopia. (In all of the remaining pictures the normal embryo is on the left and the mutant embryo is on the right.) (B) This E12.5 mutant embryo displays a reduction of the forebrain (arrow), while the rest of the embryo is unaffected. (C) This E10.5 mutant embryo has a severe forebrain phenotype and mild facial abnormality, in addition to a reduction in the hindbrain (arrow). (D) As with alobar HPE patients, as in D, some E10.5 Smad2; Tgif mutant embryos are cyclopic with a proboscis (asterisk).
81 Figure 4.4. Histological evaluation and comparison of sections of Smad2; Tgif mutant embryos (A,C) and a wildtype littermate (B,D). Note the separation of the forebrain and midbrain in the mutant (A) versus the littermate (B) (sagittal sections). Bifurcation (D, asterisks) in the wildtype embryo is more apparent than in mildly affected embryos (C) (transverse). Clefting of the palate (E, arrow) was observed at a low frequency in Smad2; Tgif embryos. fb=forebrain, mb=midbrain, hb=hindbrain, np=primitive nasal passage, e=eye.
82 A mb B mb fb
e fb
C D hb hb
e e fb fb
*
EFfb np e
e np
83 A B
Figure 4.5. Additional anomalies seen in Smad; Tgif mutant embryos. Occasionally, the forebrain defects are accompanied by reductions in hindbrain defects (A, arrow). A small number of the mutant embryos displayed severe axial defects, which can be accompanied by hemorrhaging, in addition to the HPE anomaly (B).
84 15
12
9
6
3
0 Percent (%) cells TUNEL positive
Figure 4.6. Apoptosis was evaluated via TUNEL assays. Above is an average of the evaluation of three fields within the forebrain of three separate embryos. No significant difference was observed between aphenotypic and phenotypic embryos (p=0.575).
85 Figure 4.7. Cellular proliferation was analyzed via BrdU incorporation at E10.5. To quantitate the proliferation, sections of embryos displaying either (A) a normal forebrain (n=2) or (B) a holoprosencephalic forebrain (n=2) were scored in three separate fields (indicated by boxes, and shown larger in C and D). As a result, a 16% increase in cellular proliferation was detected in the mutant Smad2; Tgif holoprosencephalic embryos in the forebrain.
86 A B
C D
C
E 70
60
50
40
30 Positive Cells Percent (%) Brdu 20
10
0
87 A B
C D
E
Figure 4.8. Molecular characterization of Smad2; Tgif holoprosencephalic embryos. In all of the panels above, the Smad2; Tgif mutant embryos (left) are shown in comparison to a normal littermate (right). (A) Whole mount in situ hybridization shows Zic2 expression is unchanged in a representative Smad2; Tgif mutant embryo. Likewise, the mouse embryonic expression at E9.5 of Six3 (B), Fgf8 (C) and Shh (D) are unaffected. (E) A representative Smad2+/-; Tgif+/- embryo shows decreased expression of arx when compared to a Smad2+/+; Tgif+/- littermate.
88 A
B C
Figure 4.9. Evaluation of Otx2 expression. (A) Otx2 is reduced in phenotypically affected Smad2; Tgif embryos (right) in comparison to wildtype littermates (left). (B&C) Sagittal sections of these embryos reveal that a complete loss of expression is exhibited in the forebrain (C), while expression in the midbrain is still present.
89 A
B
Figure 4.10. Retinoic acid and Otx2. E.9.5 embryos with a retinoic acid responsive transgene upstream of the lacZ gene were stained with X-gal to determine where RA signaling is occurring. In A, LacZ staining is evident in the developing forebrain and eye, and dorsally along the somites, suggestive that RA is important for the development of these regions. Sections through the above embryo (B) reveals that the region of the forebrain where RA is active overlaps that of Otx2 expression in another embryo (Fig. 4.9 B).
90 A B C 80
70
60
50
Smad2+/- 40 Smad2+/+ Percent (%) 30
20
10
0 HPE exencephaly runted pericardial normal ballooning
Figure 4.11. Retinoic acid induces anomalies in Smad2+/- embryos. The graph (A) summarizes the abnormalities observed at E10.5 in embryos exposed to RA at E7.5. Pictured are representative embryos displaying severe HPE (B) and exencephaly (C).
91 A
100
90
80
70
60
50 Smad3+/- Smad3+/+
Percent (%) Percent 40
30
20
10
0 HPE exencephaly axial defects normal with pericardial ballooning
Figure 4.12. Retinoic acid does not induce anomalies in Smad3 mutant embryos. The graph (A) summarizes the abnormalities observed at E10.5 in embryos exposed to RA at E7.5.
92 -2052 catgccgcttttatcaaatgatctgtttaggcgatctaattgggtgagagctagactagtaaataa -1985 atacatgcccatcttaacatcatatataacaggtcagaactgtccaagggaataagagtgtattttc -1918 aaagaattcctctttagagggtatttaaataaggaacgactgggtttttaaagaatcttacgttcac -1851 agttacaaatagctagatcttggaagaaggaacacagaatataaagtgttcttcgcaaaagtaacta -1784 gtttcactcgttaacaatccagttctcttagaaggacatttataaaaatcacattttagggtgggga -1717 aataagacttgtgttctgaatgtaaatctggaagaaatcacagctgttgagtttgtttgtgaagctt -1650 tgcctttgtctcttaagcaaggtaaatatgaaaacctaaggtcttaattttttcattaatgaacttt -1583 atgaagtgacagagaactcagtcttgtatccgccattttgaggaatttctcagagtgtcaaagtttc -1516 aaagcagagagagagagagagagagagagagagagagagagagagagagagagagagagagagcgct -1449 ttaagtgccttattttggaaaataaaagccagccttacacacattgcctgcctgtattaacatcgat -1382 ggggggaggggttcagcaaaggaacacagatgacttttctttttctccccccacccagctccttttc -1315 ttacaaaattgtaggaaatttaaaaagcagagagaagtgcggttagttccattgctttggatgttta -1248 ttgacttcaaacctaaatatatgggaaaggtggctcccgttgcagccccctctccccccacccaaaa -1181 tccaagaaaaagtgtttcccgctggatctgtagcctcctgacataaaaaacatggaagcggcttgct -1114 gtaggaatgcacccctctggacttgcaaaatgtgggggcagcagtttatctgactttacttcagtta -1047 ggattctaacttaagcatttctgtcaagccctgtgctagtcttgaagaaaagacttttttttttccc -980 tctcagtatgtttaaattaaatatttgatgggtgtcgcttaaaagttaaatagttgtttgtttagct -913 atagttctttagacacatgtaatcaaactcctccaaaaccctaataaatctgttacttacttcgaaa -846 tctaattatccagactactaattaggtgaaaatgattactggaaatgacttcaaatgtgggataata -779 gtgattgggcagttttcacaagtttgttgttcagcaagttggtgttttaaatgttgaatgcgttaca -712 ttttaggttcacagcagttgaaagtggtaataatttaaaaaaaaatcagctaagataattaaaatct -645 ctgccagggaaaggcaacagtctgggaagaggtgatttcttgaattgtttctttgtttctcaccaat -578 gggctttgttatcagcattatttatttagccaaagagttctttgtctctgcaaatggccccaatcaa -511 gttttgtttgagacaattagctagtgccagccaatgagtcctatgcaaatgttgggttaggaatgta -444 agcgtgcatttggaggcgtggggttttgttcttggggaaggcagattgtaattgctttcctcgggtc -377 ggttttctttctttctttttttaagtttagggtgggggtggggaagaCAGGTTATCTGGTCTCACTC -310 CATCCCTCTAGTTTTGGAGCTGCTGGGGGGTGGGGGGGACGGCGGGGGTGGGGGACGCATCTGCAAC -243 TCCTTTAAAAGCCTGTGCCCAGCGTCTCCCGGGTTCTTTTTAGTTAGTGCTGGAACGTGGAGGAAGC -179 TGCTCCCTCCGAAGCAGTAAACCAGCATTTCTGTTTGTTTGTTTGCTTTGCCCTTAGTTCCGTCACT -109 CCAAATCTACCCACCAAGGACCCTGACCCTGTCCACTCCAGGCGAATCGAGACCGTCCGGCTGGGTC -42 CCCCCAATTTGGGCCGACTTTGCGCCTCCAAACAACCTTAGCATGATGTCTTATCTAAAGCAACCGC
Figure 4.13. Otx2 potential promoter region. The sequence 5’ of the Otx2 gene contains Smad binding elements (SBE, yellow), core and half site recognized by Tgif in competition assays with RXRs (blue, [108]), Tgif consensus sites (pink) and two tandem retinoic acid response elements (green). The capital letters represent the first part of exon one. The known ATG-start site is in italics and underlined.
93 CHAPTER 5
DISCUSSION
Many human diseases that stem from defects in axial development have been identified. Analysis of mouse models has contributed to the understanding of the cause of these anomalies, due to the conservation of the molecular pathways between mouse and humans. In addition, this analysis has helped to further the understanding on the molecular cues necessary for development.
Bent tail (Bn) is a spontaneous, semi-dominant mutation on the mouse X chromosome that produces tail deformities and, in 10% [99], open neural tube defects. Although this mouse mutant was first described in 1952 [78], little progress had been made on identifying the gene(s) disrupted to cause the tail and neural phenotypes. Our analysis of 292 normal male and affected male and female progeny from an intraspecific backcross involving Bn supports a gene order of cen – DXMit89 – 18.5 cM ± 2.3 cM – DXMit166 – 1.4 ± 0.7 cM – Bn – 1.0 cM ± 0.6 cM – DXMit140 – 4.8 ± 1.3 cM – DXBay6 - tel. During this initial mapping, a high frequency of sex chromosomal nondisjunction, unrelated to the Bn mutation, was also identified in the background strain like others have reported, the Bn strain
94 may provide a model for studying sex chromosome separation during meiosis. In
this study, assays for autosomal nondisjunction was not conducted, and the
possibility of the occurrence of such can not be eliminated at this time.
Refined genetic and physical mapping of the Bn critical region demonstrate
that the mutation is associated with a <170 kb submicroscopic deletion that
includes the anonymous microsatellite marker DXMit208 as well as the entire Zic3
locus. Human mutations in ZIC3 are associated with left-right axis malformations
(MIM #306955, #208530, #207100). Gross evaluation of the abdominal and
thoracic cavities of Bn males and females have revealed that situs anomalies also
resulted from the absence of the Zic3 gene in mouse, with a higher incidence in the
abdomen.
It is possible, due to the size of the deletion that other genes or regulatory
regions could be disturbed, suggesting that Zic3 is not the only gene contributing to
the situs and tail anomalies in the Bn mouse. However, since this data was
published, a report by Purandare, and colleagues [30] has shown that like the Bn
mutant, the Zic3 knockout mouse displays tail anomalies and left-right patterning
defects. This suggests that the absence of Zic3 gene is solely responsible for the
anomalies described here.
The presence of anal and spinal abnormalities in some of the human
patients and the deletion of Zic3 in Bn mice support a key role for this gene in neural tube development and closure. While many forms of X-linked NTDs occur in humans, no genes have yet been associated with any of these. The lack of mutations of the ZIC3 coding region in the families described here does not yet
95 exclude the gene from the cause of NTDs of those examined. It is possible that
evaluating the promoter region of the gene may reveal that regulation of the gene is
disrupted, resulting in the neural phenotypes. It is worth mentioning that like the Bn
mice, Zic3 null mice also display exencephaly. This established a stronger link
between ZIC3 and neural tube development.
Holoprosencephaly (HPE), a spectrum of craniofacial midline anomalies
has been shown to result from a number of teratogenic agents, most notably
retinoic acid. In addition, 12 genetic loci have been implicated in familial or
sporadic HPE, of which seven have been associated with specific genes.
However, there has yet to be reported a correlation between these genetic and
teratogenic factors.
One of the known human HPE genes is TG interacting factor (TGIF),
which encodes a homeodomain protein, and was identified by its ability to bind to
the retinoid X receptor response element. Moreover, it can play a role in TGF-β
mediated Smad2 and Smad3 dependent transcription. Even though mutations in
TGIF have been identified in human HPE patients, HPE is not evident in mice
carrying the targeted null allele of Tgif. Previous work in our lab has established
that Smad2+/-; Smad3+/- embryos have a 50% occurrence of HPE. To elucidate whether Tgif in conjunction with reductions in TGF-β signaling can cause HPE,
we generated mice that have mutations in both Smad2 and Tgif.
Our results show that one-third of the Smad2+/-; Tgif+/- and
Smad2+/-;Tgif-/- embryos, one third display HPE. Molecular characterization at 96 E9.5 via Shh, Fgf8, Six3 and Zic2 expression analysis revealed that the
mechanism behind the manifestation of HPE in the Smad2; Tgif mutant animals
had yet to be described in a genetic model of HPE. The molecular evaluation of
the forebrain domain of Otx2 however did provide a clue to the possible cause of
the phenotype. The reduction of Otx2 expression in affected embryos mimics
what has been reported in embryos exposed to exogenous retinoic acid.
Because Tgif was initially identified by its ability to compete with components of
the RA pathway, and RA is known to cause HPE by maternal exposure during
pregnancy in both mouse and humans, it is likely that RA signaling is playing a
role in the HPE phenotype observed.
While both Tgif and Smad2 have been linked to RA signaling in vitro, the relevance of these interactions in vivo have yet to be demonstrated. Teratogenic analysis of Tgif or Smad2 mutant embryos demonstrated that the absence of the individual gene functions allowed for an increase in the sensitivity of the mutant embryos to RA induced phenotypes. Thus, this provided evidence that RA signaling does interact with Smad2 and Tgif. The results of this seem that the
loss of Tgif leads to a loss of competition in the RA pathway, thus an increase in
RA signaling. Currently, it is unclear whether this is dependent on or independent of Smad2. It is possible that Smad2 has a unique function
contributing to the HPE phenotype. Regardless, this provides evidence that one
role of normal gene function is to provide for protection against environmental
factors that can disturb normal embryonic development.
97 WORKS CITED
1. Beddington, R., and Robertson, E., Anterior patterning in the mouse. Trends in Genetics, 1998. 14: p. 277-284.
2. Thomas, P., Brickman, J. M., Popperl, H., Krumlauf, R., and Beddington, R. S. P., Axis Duplication and Anterior Identity in the Mouse Embryo, in Cold Spring Habor Symposia on Quantitative Biology. 1997, Cold Spring Harbor Laboratory Press: Cold Spring Harbor. p. 115-125.
3. Gardner, R.L., Polarity in early mammalian development. Current Opinion in Genetics and Development, 1999. 9: p. 417-421.
4. Martinez-Barbera JP, B.R., Getting your head around Hex and Hesx1: forebrain formation in mouse. Internation Journal of Developmental Biology, 2001. 45: p. 327-336.
5. Acampora, D., Mazan, S., Lallemand, Y., Avantaggiato, Maury, M., Simeone, S., and Brulet, P., Forebrain and midbrain regions are deletedin Otx2-/- mutants due to defective anterior neuroectoderm specification during gastrulation. Development, 1995. 121: p. 3270-3290.
6. Perea-Gomez A, R.M., Ang SL., Role of the anterior visceral endoderm in restricting posterior signals in the mouse embryo. Internation Journal of Developmental Biology. 45: p. 311-320.
7. Thomas, P., and Beddington, R. S. P., Anterior primitive endoderm may be responsible for patterning the anterior neural plate in the mouse embryo. Current Biology, 1996. 6: p. 1487-1496.
8. DeRobertis, E.M., Fainsod, A., Gont, L. K., and Steinbeisser, H., The evolution of vertebrate gastrulation. Development, 1994. Supplement: p. 117-124.
9. Bachiller, D., Klingensmith, J., Kemp, C., Belo, J. A., Anderson, R. M., May, S. R., McMahon, J. A., McMahon, A. P., Harland, R. M., Rossant, J., and DeRobertis, E. M., The organizer factors chordin and noggin are required for mouse forebrain development. Nature, 2000. 403: p. 658-661.
10. Gilbert, S.F., Developmental Biology. Seventh Edition ed. 2003, Sunderland, Massachusetts: Sinauer Associates, Incorporated, Publishers. 98 11. Veraksa, A., Del Campo, M., and McGinnis, W., Developmental patterning genes and their conserved functions: From model organisms to humans. Molecular Genetics and Metabolism, 2000. 69: p. 85-100.
12. Li, X., and Cao, X., BMP signaling and HOX transcription factors in limb development. Fronts in Bioscience., 2003. Supplement: p. 805-812.
13. Patterson, L.T., and Potter, S.S., Hox genes and kidney patterning. Current Opinon in Nephrology and Hypertension, 2000. 12: p. 19-23.
14. Sasai, Y., Lu, B., Steinbeisser, H., Geissert, D., Gont, L. K., and DeRobertis, E. M., Xenopus chordin: A novel dorsalizing factor activated by organizer-specific homeobox genes. Cell, 1994. 79(779-790).
15. Smith , W.C., and Harland, R. M., Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell, 1992. 70: p. 829-840.
16. Sasai, Y., Lu, B., Steinbeisser, H., and DeRobertis, E. M., Regulation of neural induction by the chd and BMP-4 antagonistic patterning signals in Xenopus. Nature, 1995. 376: p. 333-336.
17. Sasai, Y., and DeRobertis, E. M., Ectodermal patterning in vertebrate embryos. Developmental Biology, 1997. 182: p. 5-20.
18. Casey, B., Genetics of human situs abnormalities. Am J Med Genet, 2001. 101(4): p. 356-8.
19. Nonaka, S., Tanaka, Y., Okada, Y., Takeda, S., Harada, A., Kanai, Y., Kido, M. and Hirokawa, N., Randomization of left-right asymmetrydue to loss of nodal cilia generating leftward flow of extraembryonic fluid in mice lacking KIF3B motor protein. Cell, 1998. 95: p. 829-837.
20. Nonaka, S., Shiratori, H., Saijoh, Y., and Hamada, H., Determination of left-right patterning of the mouse embryo by artificial nodal flow. Nature, 2002. 418(96-99).
21. Okada, Y., Nonaka, S., Tanaka, Y., Nonaka, S., and Hirokawa, N., Abnormal nodal flow proceeds situs inversus in iv and inv mice. Molecular Cell, 1999. 4: p. 459-468.
22. Supp, D.M., Brueckner, M., Kuehn, M. R., Kuehn, M. R., Witte, D. P., Lowe, L. A., McGrath, J., Corrales, J. and Potter, S. S., Targeted deletion of the ATP binding domain of left-right dyenin confirms its role in specifying development of left right assymmetries. Development, 1999. 126: p. 5495-5504.
99 23. Marszalek, J.R., Ruiz-Lozano, P., Roberts E., Chein, K. R. and Goldstein, L. S., Situs inversus and ciliary morphogenesis defects in mouse mutants lacking the KIF3A subunit of kinesin-II. Proceedings of the National Academy of Science, 1999. 96: p. 5043-5048.
24. McGrath, J., Somlo, S., Makova, S., Tian, X., and Brueckner, M., Two populations of node monocilia initiate left-rightasymmetry in the mouse. Cell, 2003. 114: p. 61-73.
25. Afzelius, B.A., A human syndrome caused by immotile cilia. Science, 1976. 193: p. 317-319.
26. Collignon, J., Varlet, I., and Robertson, E. J., RElationship between asymmetric nodal expression and the direction of embryonic turning. Nature, 1996. 381: p. 155-158.
27. Capdevila, J., Vogan, K. J., Tabian, C. J., and Belmonte, J. C. I., Mechanisms of left-right determination in vertebrates. Cell, 2000. 101: p. 9-21.
28. Gebbia, M., Ferrero, G. B., Pilia, G., Bassi, M. T., Aylsworth, A., Penman- Splitt, M., Bird, L. M., Bamforth, J. S., Burn, J., Schlessinger, D., Nelson, D. L., and Casey, B., X-linked situs abnormalities result from mutations in ZIC3. Nat Genet, 1997. 17(3): p. 305-8.
29. Carrel, T., Purandare, S. M., Harrison, W., Elder, F., Fox, T., Casey, B., and Herman, G. E., The X-linked mouse mutation Bent tail is associated with a deletion of the Zic3 locus. Hum Mol Genet, 2000. 9(13): p. 1937-42.
30. Purandare, S.M., Ware, S. M., Kwan, K. M., Gebbia, M., Bassi, M. T., Deng, J. M., Vogel, H., Behringer, R. R., Belmont, J. W., and Casey, B., A complex syndrome of left-right axis, central nervous system and axial skeleton defects in Zic3 mutant mice. Development, 2002. 129(9): p. 2293-302.
31. Weinstein, M., Yang, X., Li, C., Gotay, J., Deng, C-X., Failure of egg cylinder elongation and mesoderm induction in mouse embryos lacking the tumor supressor Smad2. Proceedings of the National Academy of Science, 1998. 95: p. 9378-9383.
32. Waldrip, W.R., Bikoff, E. K., Hoodless, P. A., Wrana, J. L. and Robertson, E. J., Smad2 signaling in extraembryonic tissues determines anterior- posteriorpolarity of the early mouse embryo. Cell, 1998. 92: p. 797-808.
33. Nomura, M., Li, E., Smad2 role in mesoderm formation, left-right patterning and craniofacial development. Nature, 1998. 393: p. 786-790.
100 34. Heyer, J., Escalante,-Alcalde, D., Lia, M., Boettinger, E., Edelmann, W., Stewart, C. L. and Kucherlapati, Post gastrulation Smad2 deficient embryos show defects in embryo turning and anterior morphogenesis. Proceedings of the National Academy of Science, 1999. 96: p. 12595- 12600.
35. Yokoyama, T., Copeland, N., Jenkins, N., Montgomery, C., Elder, F. and Overbeek, P., Reversal of left-right asmmetry: a situs inversus mutation. Science, 1993. 260: p. 679-682.
36. Copp, A.J., Greene, N. D. E. and Murdoch, J. N., The Genetic Basis of Mammalian Neuralation. Nature Reviews Genetics, 2003. 4: p. 784-793.
37. Copp, A.J., Relationship between the timing of the posterior neuropore closure and the development of spinal neural tube defects in mutant (curly tail) and normal mouse embryos in culture. Journal of Embryological and Experimental Morphology, 1985. 88: p. 39-54.
38. Van Straaten, H.W.M., Hekking, J. W. M., Consten, C., and Copp, A. J., Intrinsic and extrinsic facotrs in the mechanism of neurulation: Effect on the curvature of the body axis on the closure of the posterior neuropore. Development, 1993. 117: p. 1163-1172.
39. Botto, L.D., C. A. Moore, M. J. Khoury ,J. D. Erickson., Neural Tube Defects. New England Journal of Medicine, 1999. 341: p. 1509-1519.
40. Copp, A.J., and Bernfield, M., Etiology and pathogenesis of human neural tube defects. Current Opinion in Pediatrics, 1994. 6: p. 624-631.
41. Emery, A.E.H., Methodology in medical genetics. Second edition ed. 1986, Edinburgh: Churchhill Livingstone. 197.
42. Finnell, R.H., et al., Neural tube and craniofacial defects with special emphasis on folate pathway genes. Crit Rev Oral Biol Med, 1998. 9(1): p. 38-53.
43. MRC, V.S.R.G., Prevention of neural-tube defects: results of the Medical Council Vitamin Study. Lancet, 1991. 338(131-187).
44. Czeizel, A.E. and I. Dudas, Prevention of the first occurrence of neural- tube defects by periconceptional vitamin supplementation. N Engl J Med, 1992. 327(26): p. 1832-5.
45. Recommendations for the use of folic acid to reduce the number of cases of spina bifida and other neural tube defects, in MMWR Recomm Rep. 1992, Centers for Disease Control and Prevention. p. 1-7.
101 46. Finnell, R.H., et al., Genetic basis of susceptibility to environmentally induced neural tube defects. Ann N Y Acad Sci, 2000. 919: p. 261-77.
47. Carter, M., et al., Crooked tail (Cd) models human folate-responsive neural tube defects. Hum Mol Genet, 1999. 8(12): p. 2199-204.
48. Qu, S., Tucker S. C., Zhao, Q., deCrombrugghe, B., Wisdom, R., Physical and genetic interactions between Alx4 and Cart1. Development, 1999. 126: p. 359-369.
49. Zhao, Q., R.R. Behringer, and B. de Crombrugghe, Prenatal folic acid treatment suppresses acrania and meroanencephaly in mice mutant for the Cart1 homeobox gene. Nat Genet, 1996. 13(3): p. 275-83.
50. Barbera, J.P., et al., Folic acid prevents exencephaly in Cited2 deficient mice. Hum Mol Genet, 2002. 11(3): p. 283-93.
51. Goulding, M.D., Chalepakis, G., Deutsch, U., Erselius, J.R., Gruss P., Pax-3, a novel murine DNA binding protein expressed during early neurogenesis. EMBO, 1991. 105: p. 1135-1147.
52. Fleming, A., and Copp, A. J., Embryonic folate metabolism and mouse neural tube defects. Science, 1998. 280(5372): p. 2107-9.
53. Gefrides, L.A., G.D. Bennett, and R.H. Finnell, Effects of folate supplementation on the risk of spontaneous and induced neural tube defects in Splotch mice. Teratology, 2002. 65(2): p. 63-9.
54. Piedrahita, J.A., Oetama, B., Bennett, G. D., van Waes, J., Kamen, B. A., Richardson, J., Lacey, S. W., Anderson, R. G., and Finnell, R. H., Mice lacking the folic acid-binding protein Folbp1 are defective in early embryonic development. Nat Genet, 1999. 23(2): p. 228-32.
55. Swanson, D.A., Liu, M. L., Baker, P. J., Garrett, L., Stitzel, Wu, J., Harris, M., Banerjee, R., Shane, B., Brody, L. C., Targeted disruption of the methionine synthase gene in mice. Molecular and Cellular Biology, 2001. 21: p. 1058-1065.
56. Volpe, J.J., Normal and Abnormal Human Brain Development. Mental Retardation and Developmental Disabilities., 2000. 6(1-5).
57. DeMyer, W., Holoprosencephaly (cyclopia-arhinencephaly). Handbook of Clinical Neurology, ed. P.J. Vinken, Bruyn, G. W. 1977, Amsterdam, North-Holland. 465-472.
58. Matsunaga, E., and Shiota, K., Holoprosencephaly in human embryos: epidemiologic studies of 150 cases. Teratology, 1977. 16: p. 261-272. 102 59. Roach, E.D., W. Conneally, P.M., Palmer, C., Merritt, A. D., Holoprosencephaly: birth data, genetic and demographic analyses of 30 families. Birth Defects, 1977. 11(2): p. 294-313.
60. Croen, L.A., Shaw, G. M. and Lammer, E. J., Holoprosencephaly: epidemiologic and clinical characteristics of a California population. American Journal Medical Genetics, 1996. 64: p. 465-472.
61. Rossler, E.a.M., M., Holoprosencephaly: A paradigm for the complex genetics of brain development. Journal of inherited metabolic Disorders, 1998. 21: p. 482-497.
62. Hayhurst, M.a.M., S. K., Mouse models of holoprosencephaly. Current Opinions in Neurology, 2003. 16: p. 135-141.
63. Wallis, D., and Muenke, M., Mutations in Holoprosencephaly. Human Mutation, 2000. 16: p. 99-108.
64. Cohen, M.M., Perspectives on holoprosencephaly: part I. Epidemiology, genetics, and syndromeology. Teratology, 1989. 40(211-235).
65. Roelink, H., Porter, J., Chiang, C., Tanabe, Y. Chang, D., Beachy, P., and Jessell, T., Floor plate and motor neuron inductionby different concentrationsof the amino terminal cleavage product of Sonic Hedgehog proteolysis. Cell, 1995. 81: p. 445-455.
66. Chiang, C., Litingtung, Y., Lee, E., Young, K. E., Cordon, J. L., Westphal, H., and Beachy, P. A., Cyclopia and defective axial patterning in mice lacking Sonic Hedgehog Gene Function. Nature, 1996. 383: p. 407-413.
67. Brown, S.A., Warburton, D. Brown, L.Y., Yu, C. Roeder, E.R., Stengel- Rutkowski, S. Hennekam, R.C.M., Muenke, M., Holoprosencephaly due to mutations in Zic2, a homologue of Drosophila odd-paired. Nature Genetics, 1998. 20: p. 180-183.
68. Wallis, D.E., Roessler, E., Heht, U., Nanni, L., Wiltshire, T., Richeri-Costa, A., Gillessen-Kaesbach G., Zackai, E. H., Rommens, J., and Muenke, M., Mutations in the hoemodomain of the human SIX3 gene cause holoprosencephaly. Nature Genetics, 1999. 22: p. 196-199.
69. Lagutin, O., Zhu CC, Kobayashi D, Topczewski J, Shimamura K, Puelles L, Russell HR, McKinnon PJ, Solnica-Krezel L, Oliver G., Six3 repression of Wnt signaling in the anterior neuroectoderm is essential for vertebrate forebrain development. Genes in Development, 2003. 17: p. 368-379.
103 70. Nissenkorn, A., Michelson, M., Ben-Zeev, B. and Lerman-Sage, T., Inborn errors of metabolism: A cause of abnormal brain develoment. Neurology, 2001. 56: p. 1265-1272.
71. Porter, B.E.a.T., G., Myelin and Disorders That Affect the Formation and Maintenance of This Sheath. Mental Retardation and Developmental Disabilities., 2000. 6: p. 47-58.
72. (OMIM), O.M.I.i.M., Microcephaly, autosomal dominant. 2004, Johns Hopkins University.
73. (OMIM), O.M.I.i.M., Zellweger Syndrome. 2003, Johns Hopkins University.
74. Li, X., Baumgart, E. Marralle, D., and Gould, S. J., PEX11-beta deficiency is lethal amnd impairs neuronal migration but does not abrogate peroxisome function. Molecular and Cellular Biology, 2002. 22(4358- 4365).
75. Wood, W.B., Left-right asymmetry in animal development. Annual Review in Cell and Developmental Biology, 1997. 13: p. 53-82.
76. Mikoshiba, K., Mammalian neural induction and brain pattern formation. Keio J Med, 1999. 48(3): p. 111-23.
77. Harris, M.J., and Juriloff, D.M., Mini-review: toward understanding mechanisms of genetic neural tube defects in mice. Teratology, 1999. 60: p. 292-305.
78. Garber, E.D., Bent-tail, a dominant, sex-linked mutation in the mouse. Proceedings of the Nationall Academy of Science U S A, 1952. 38: p. 876-879.
79. Johnson, D.R., The interfrontal bone and mutant genes in the mouse. Journal of Anatomy, 1976. 121: p. 507-513.
80. Lyon, M.F., Order of loci on the X-chromosome of the mouse. Genetic Research, 1966. 7: p. 130-133.
81. Sweet, H.O.B., R.T., Harris, B.S., and Davisson, M.T., Trembly-like (Tyl) - a new X-linked mutation in the mouse. Mouse Genome, 1990. 87: p. 112.
82. Denny, P., Haynes, A.R., Masson, W., Laval, S.H., Boyd, Y., Disteche, C. and Elliott, R.W., Chromosome X. 2000.
83. Didsbury, J., Weber, R.F., Bokoch, G.M., Evans, T. and Snyderman, R., Rac, a novel ras-related family of proteins that are botulinum toxin substrates. JBC, 1989. 264: p. 16378-16382. 104 84. Drivas, G.T., Shih, A., Coutavas, E., Rush, M.G. and D'Eustachio, P., Characterization of four novel ras-like genes expressed in a human eratocarcinoma cell line. Molecular Cell Biology, 1990. 10: p. 1793-1798.
85. Malosio, M.L., Gilardelli, D., Paris, S., Albertinazzi, C. and de Curtis, I., Differential expression of distinct members of Rho family GTP-binding proteins during neuronal development: identification of RaclB, a new neural-specific member of the family. Journal of Neuroscience, 1997. 17: p. 6717-6728.
86. Aruga, J., Nagai, T., Tokuyama, T., Hayashizaki, Y., Okazaki, Y., Chapman, V. M., and Mikoshiba, K., The mouse zic gene family. Homologues of the Drosophila pair-rule gene odd-paired. J Biol Chem, 1996. 271(2): p. 1043-7.
87. Nagai, T., Aruga, J., Takada, S., Gunther, T., Sporle, R., Schughart, K., and Mikoshiba, K., The expression of the mouse Zic1, Zic2, and Zic3 gene suggests an essential role for Zic genes in body pattern formation. Dev Biol, 1997. 182(2): p. 299-313.
88. Ferrero, G.B., Gebbia, M., Pilia, G., Witte, D., Peier, A., Hopkin, R.J., Craigen, W.J., Shaffer, L.G., Schlessinger, D., Ballabio, A., et al., A submicroscopic deletion in Xq26 associated with familial situs ambiguus. American Journal of Human Genetics, 1997. 61: p. 395-401.
89. Korobova, O., Lane, P.W., Perry, J., Palmer, S., Ashworth, A., Davisson, M.T. and Arnheim, N., Patchy fur, a mouse coat mutation associated with X-Y nondisjunction, maps to the pseudoautosomal boundary region. Genomics, 1998. 54: p. 556-559.
90. Hunt, P.A., and Eicher, E.M., Fertile male mice with three sex chromosomes: evidence that infertility in XXY male mice is an effect of two Y chromosomes. Chromosoma, 1991. 100: p. 293-299.
91. Everett, C.A., Searle, J.B. and Wallace, B.M., A study of meiotic pairing, nondisjunction and germ cell death in laboratory mice carrying Robertsonian translocations. Genetic Research, 1996. 67: p. 239-247.
92. Angel, T.A., Faust, C.J., Gonzales, J.C., Kenwrick, S., Lewis, R.A. and Herman, G.E., Genetic mapping of the X-linked dominant mutations striated (Str) and bare patches (Bpa) to a 600-kb region of the mouse X chromosome: implications for mapping human disorders in Xq28. Mammalian Genome, 1993. 4: p. 171-176.
93. Osoegawa, K., Tateno, M., Woon, P.Y., Frengen, E., Mammoser, A.G., Catanese, J.J., Hayashizaki, Y. and de Jong, P.J., Bacterial artificial
105 chromosome libraries for mouse sequencing and functional analysis. Genome Research, 2000. 10: p. 116-128.
94. Faust, C.J., and Herman, G.E., Physical mapping of the loci Gabra3, DXPas8, CamL1 and Rsvp in a region of the mouse X chromosome homologous to human Xq28. Genomics, 1991. 11: p. 154-164.
95. Campbell, L.R., Dayton, D. H., Sohal, G. S., Neural tube defects: a review of human and animal studies on etiology of neural tube defects. Teratology, 1986. 34: p. 171-187.
96. Epstein, D.J., Vekemans, M., Gros., P., Splotch (Sp2H), a mutation affecting development of the mouse neural tube, shows a deletion within the paired homeodomain of Pax-3. Cell, 1991. 67: p. 767-774.
97. Hoth, C.F., Milunsky, A., Lipsky, N., Sheffer, R., Clarren, S. K., and Baldwin, C. T., Mutations in the paired domain of the human PAX3 gene cause Klein-Waardenburg (WS-III) as well as Waardenburg syndrome type I (WS-I). American Journal of Medical Genetics, 1993. 52: p. 455- 462.
98. Baldwin, C.T., Hoth, C. F., Macine, R. A., Milunsky, A., Mutations in PAX3 that cause Waardenburg syndrome type I: ten new mutations and review of the literature. American Journal of Medical Genetics, 1995. 58: p. 115- 122.
99. Klootwijk, R., et al., A deletion encompassing Zic3 in bent tail, a mouse model for X-linked neural tube defects. Hum Mol Genet, 2000. 9(11): p. 1615-22.
100. Hol, F.A., Geurds, M. P., Jensson, O., Hamel, B. C., Moore, G. E., Newton, R., and Mariman, E. C., Exclusion mapping of the gene for X- linked neural tube defects in an Icelandic family. Hum Genet, 1994. 93(4): p. 452-6.
101. Derry, J.M.J., Gormally, E., Means, G. D., Zhao, W., Meindl, A., Kelley, R. I., Boyd, Y., Herman, G. E., Mutations in D8-D7 sterol isomerase in the tattered mouse and X-linked dominant chondroysplasia punctata. Nature Genetics, 1999. 22: p. 286-290.
102. Sekelsky, J.J., Newfeld, S. J., Raftery, L. A., Chartoff, E. H., and Gelbart, W., Genetic characterization and cloning of mother against dpc, a gene required for decapentapelegic function in Drosophial melanogaster. Genetics, 1995. 139: p. 1347-1358.
103. Savage, C., Das, P., Finelli, A. L., Townsend, S. R., Sun, C. Y., Baird, S., and Padgwtt, R. W., Caenorhabditis Elegans genes sma-2, sma-3 and 106 sma-4 define a conserved family of transforming growth factor-β pathway components. Proceedings of the National Academy of Science, 1996. 93: p. 790-794.
104. Chang, H., Brown, C. W., and Matzuk, M. M., Genetic analysis of the Mammalian Transforming Growth Factor-β Superfamily. Endocrine Reviews, 2002. 23: p. 787-823.
105. Moustakas, A., Souchelntskyi, S., and Heldin, C-H., Smad regulation in TGF-β signal transduction. Journal of Cell Sciences, 2001. 114: p. 4359- 4369.
106. Raftery, L.A., and Sutherland, D. J., TGF-β Family Signal Transduction in Drosophila Development: Mad to Smads. 1999.
107. Weinstein, M., Yang, X., and Deng, C-X., Functions of mammalian Smad genes as revealed by targeted gene disruption in mice. Cytokine and Growth Factor Reviews, 2000. 11: p. 49-58.
108. Bertolino, E., Reimund, B., Wildt-Perinic, D., and Clerc, R. G., A novel homeobox protein which recognizes a TGT core and functionally interferes with a retinoid responsive motif. The Journal of Biological Chemistry, 1995. 270: p. 31178-31188.
109. Burglin, T.R., Analysis of TALE superclass homeobox genes (MEIS, PBC, KNOX, Iroquois, TGIF) reveals a novel domain conserved between plants and animals. Nucleic Acid Research, 1997. 21(4173-4180).
110. Wotton, D., Lo, R. S., Lee, S., and Massague, J., A smad transcriptional corepressor. Cell, 1999. 97: p. 29-39.
111. Wotton, D., Lo, R. S., Swaby, L.-A. C., and Massague, J., Multiple modes of repression by the Smad transcriptional corepressor TGIF. The Journal of Biological Chemistry, 1999. 274: p. 37105-37110.
112. Gripp, K.W., Wotton, D., Edwards, M. C., Roessler, E., Ades, L., Meinecke, P., Richieri-Costa, A., Zackai, E. H., Massague, J., Muenke, M., and Elledge, S. J., Mutations in TGIF cause holoprosencephaly and link NODAL signalling to human neural axis determination. Nature Genetics, 2000. 25: p. 205-208.
113. Liu, Y.F., M., Thompson, J. C., El Hodiri, H. M., and Weinstein, M., Smad2 and Smad3 coordinately regulate craniofacial and endodermal development. Developmental Biology, 2004. in press.
107 114. Nagai T, A.J., Minowa O, Sugimoto T, Ohno Y, Noda T, Mikoshiba K., Zic2 regulates the kinetics of neurulation. Proceedings of the Natational Academy of Science, 2000. 97: p. 1618-1623.
115. Meyers, E.N., Lewandoski, M., and Martin, G. R., An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nature Genetics, 1998. 18: p. 136-141.
116. Frank, D.U., Fotheringham, L. K., Brewer, J. A., Muglia, L. J., Tristani- Firouzi, M., Capecchi, M. R., and Moon, A. M., An Fgf8 mouse mutant phenocopies human 22q11 deletion syndrome. Development, 2002. 129: p. 4591-4603.
117. Miura, H., Yanazawa, M.,Kato, K., and Kitamura, K., Expression of a novel artistaless related homeobox gene 'Arx" in the vertebrate telencephalon, diencephalon and floorplate. Mechanisms of Development, 1997. 65: p. 99-109.
118. El Hodiri, H., Qi, X., and Seufert, D. W., The Xenopus Arx gene is expressed in the developing rostral forebrain. Developmental Genes in Evolution, 2003. 212: p. 608-612.
119. Bienvenu, T., Poirier, K., Friocourt, G., Bahi, N., Beaumont, D., Fauchereau, F., Jeema, L. B., Zemni, R., Vinet, M-C., Francis, F., Couvert, P., Gomot, M., Moraine, C., van Bokhoven, H., Kalscheuer, V., Frints, S., Gecz, J., Ohzaki, K., Chaabouni, H., Fryns, J-P., Desportes, V., Beldjord, C., and Chelly, J., ARX, a novel Prd-class-homeobox gene highly expressed in the telencephalon, is mutated in X-linked mental retardation. Human and Molecular Genetics., 2002. 11: p. 981-991.
120. Kitamura, K., Yanazaawa, M., Sugiyama, N., Miura, H., Iizuka-Kogo, A., Kusaka, M., Omichi, K., Suauki, R., Kato-Fukui, Y., Kamiirisa, K., Matsuo, M., Kamijo, S., Kasahara, M., Yoshioka, H., Ogata, T., Fukuda, T., Kondo, I., Kato, M., Dobyns, W., Yokoyama, M. and Morohashi, K., Mutations of ARX causes abnormal development of forebrain and testes in mice and X- linked lissencephaly with abnormal genitalia in humans. Nature Genetics, 2002. 32: p. 359-369.
121. Collombat, P., Mansouri, A., Hecksher-Sorensen, J., Serup, P., Krull, J., Gradwohl, G. and Gruss, P., Opposing actions of Arx and Pax4 in endocrine pancreas development. Genes and Development, 2003. 17: p. 2591-2603.
122. Simeone, A., Puelles, and Acampora, D., The Otx family. Current Opinions in Genetics and Development, 2002. 12: p. 409-415.
108 123. Clotman, F., Van Maele-Fabry, G., and Picard J. J., Retinoic acid induces a tissue-specific deletion in the expression domain of Otx2. Neurotoxicology and Teratology, 1997. 19: p. 163-167.
124. Clotman, F., Van Maele-Fabry, G., Chu-Wu, L., and Picard J. J., Structural and gene expression abnormalities induced by retinoic acid in the forebrain. Reproductive Toxicology, 1998. 12(169-176).
125. Gilbert, S.A., Retinoic Acid as a Teratogen, in Developmental Biology. 2003, Sinauer Associates, Incorporated, Publishers: Sunderland, Massachusetts. p. 696-697.
126. Roche Laboratories, I., Vesanoid (Tretinoin) Capsules Product Information. 1998-2003.
127. Lammer, E.J., Chen, D. T., Hoar, R. M., Agnish, N. D., Benke, P. J., Braun, J. T., Curry, C. J., Fernhoff, P. M., Grix, A. W., Lott, I. T., Richard, J. M., and Sun, S. C., Retinoic Acid Embryopathy. New England Journalof Medicine, 1985. 313: p. 837-841.
128. Lipson, A.H., Collins, F., and Webster, W. S., Multiple congenital defects associated with maternal use of topical tretinoin. The Lancet, 1993. 341: p. 1352-1353.
129. Rosa, F.W., Retinoic acid embryopathy. New England Journal of Medicine, 1986. 315: p. 262.
130. Vorhees, C.V., Retinoic Acid Embryopathy. New England Journal of Medicine, 1986. 315: p. 262-263.
131. Coberly, S., Lammer, E., and Alashari, M., Retinoic acid embryopathy: Case report and review of the literature. Pediatric Pathology and Laboratory Medicine, 1996. 16: p. 823-836.
132. Balkan, W., Colbert, M., Bock, C., and Linney, E., Transgenic indicator mice for studying activated retinoic acid receptors during development. Proceedings of the National Academy of Sciences, 1992. 89: p. 3347- 3351.
133. Cao, Z., Flanders, K. C., Bertolette, D., Lyakh, L. A., Wurthner, J. U., Parks, W. T., Letterio, J. J., Ruscetti, F. W., and Roberts, A. B., Levels of phospho-Smad2/3 are sensors of the interplay between effects of TGF-β and retinoic acid on monocytic and granulocytic differentiation of HL-60 cells. Blood, 2003. 101: p. 498-507.
109 134. Pendaries, V., Verrecchia, F., Michel, S., and Mauviel, A., Retinoic acid receptors interfere with the TGF-β/Smad signaling pathway in a ligand specific manner. Oncogene, 2003. 22: p. 8212-8220.
135. Degitz, S.J., Morris, D., Foley, G. L., and Francis, B. M., Tole of TGF-β in RA-induced cleft palate in CD-1 mice. Teratology, 1998. 58: p. 197-204.
136. Sulik, K.K., Dehart, D. B., Rogers, J. M., and Chernoff, N., Teratogenicity of low doses of all-trans retinoic acid in presomite mouseembryos. Teratology, 1995. 51: p. 398-403.
137. Arkell, R.a.B., R. S., BMP-7 influences pattern and growth of the developing hindbrain of mouse embryo. Development, 1997. 124: p. 1-12.
138. Unsicker, K., Meier, C., Krieglstein, K., Sartor, B. M., and Flanders, K. C., Expression, localization and function of transforming growth factor-betas in embryonic chick spinal cord, hindbrain, and dorsal root ganglia. Journal of Neurobiology, 1996. 29: p. 262-267.
139. Zhou, Y.X., Zhao, M.,Li, D., Shimazu, K., Sakata, K., Deng, C. X., and Lu, B., Cerebellar deficits and hyperactivity in mice lacking Smad4. Journal of Biological Chemistry, 2003. 278: p. 42313-42320.
140. Schwarte-Waldhoff, I., and Schmiegel, W., Smad4 transcriptional pathways and angiogenesis. International Journal of Intestinal Cancer, 2002. 31: p. 47-59.
141. Deckers, M.M., van Bezooijen, R. L., van der Horst, G., Hoogendam, J., van Der Bent, C., Papapoulos, S. E., and Lowik, C. W., Bone morphogenic proteins stimulate angiogenesis through ostoblast-derived vascular endothelial growth factor A. Endocrinology, 2002. 143: p. 1545- 1553.
142. Bhasin, N., LaMantia, A. S., and Lauder, J. M., Opposing regulation of cell proliferation by retinoic acid and the seratonin(2B) receptor in the mouse frontonasal mass. Anatomy and Embryology, 2004. Epublished ahead of print.
143. Oliver, G., Mailhos, A., Wehr, R., Copeland, N. G., Jenkins, N., and Gruss, P., six3, a murine homologue of the sine oculis gene, demarcates the most anterior border of the developing neural plate and is expressed during eye development. Development, 1995. 121: p. 4045-4055.
144. Guidance for Industry and Revewers. Estimate safe starting dose in clinical trials for therapeutics in healthy adult volunteers. December, 2002, US Department of Health and Human Services, Food and Drug
110 Administration, Center for Drug Evaluation and Research, and Center for Biologics Evaluations and Research.
145. Hyman, C.A., Bartholin, L., Newfield, S. J., and Wotton, D., Drosophila TGIF proteins are transcriptional Activators. Molecular and Cellular Biology, 2003. 23: p. 9262-9274.
.
111