Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Multifactorial genetics of exencephaly in the SELH/Bc mouse strain Hoscheit, Julia Lynn 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-0481.pdf [ 10.7MB ]
Metadata
JSON: 831-1.0092219.json
JSON-LD: 831-1.0092219-ld.json
RDF/XML (Pretty): 831-1.0092219-rdf.xml
RDF/JSON: 831-1.0092219-rdf.json
Turtle: 831-1.0092219-turtle.txt
N-Triples: 831-1.0092219-rdf-ntriples.txt
Original Record: 831-1.0092219-source.json
Full Text
831-1.0092219-fulltext.txt
Citation
831-1.0092219.ris

Full Text

M U L T I F A C T O R I A L GENETICS OF E X E N C E P H A L Y IN T H E SELH/Bc M O U S E STRAIN  by JULIA L Y N N HOSCHEIT B . S c , Colorado State University, 2002  A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES (Medical Genetics)  T H E UNIVERSITY OF BRITISH C O L U M B I A October 2005  © Julia Lynn Hoscheit, 2005  Abstract  The SELH/Bc mouse strain, a model for multifactorial human neural tube defects, has 10-30% exencephaly depending on the maternal diet. Developmental morphological studies have found that the mesencephalon fold elevation is delayed, compared to other normal strains, in all SELH/Bc embryos and that they omit Closure 2, a normal site of initiation of closure at the mesencephalon/prosencephalon boundary during cranial neural tube closure. Previous studies have found the genetic exencephaly risk to be due to a combination of three genes, Exenl, Exen2, and Exen3, acting additively. Exenl and Exen2 had been mapped to regions on Chr 13 and Chr 5, respectively, and Exen3 provisionally to Chr 11. M y studies further investigated the genetic cause of this defect using F2 exencephaly panels and congenic lines, as well as the diet effect that is observed in SELH/Bc. The frequencies of exencephaly were observed in F2 segregants from crosses between SELH/Bc and a normal unrelated strain, L M / B c . There were two F2 exencephaly panels collected, one from mice fed the regular diet Purina Laboratory Rodent Diet #5001 and one from mice fed the high-risk diet Purina Mouse Diet #5015 ("PMD #5015") that increases exencephaly frequency in SELH/Bc. There was a significant increase in exencephaly frequency in the F2 embryos when mice were fed P M D #5015 and genotypes of the F2 ex encephalic embryos from both panels supported the hypothesis that there is gene-diet interaction between the Exenl alleles from SELH/Bc and P M D #5015. A new locus on Chr 7 that contributed to the risk of exencephaly in the F2 embryos, named Exen4, was established as well.  The congenic lines that had been created by transferring the normal Exenl and Exen2 alleles from L M / B c into the SELH/Bc background and the Exenl and Exen2 alleles from SELH/Bc into the L M / B c background confirmed the locations of the Exenl and Exen2 loci, and demonstrated that exencephaly in SELH/Bc is a multifactorial threshold trait. The congenic lines with Exenl alleles from SELH/Bc responded to P M D #5015 supporting the hypothesis of a gene-diet interaction. Morphological studies of the cranial neural tube closure patterns in congenic line embryos supported the hypothesis that the liability trait for exencephaly (multifactorial threshold defect) is timing of mesencephalon fold elevation as the congenic line embryos showed detectable delay or acceleration of elevation of mesencephalon folds compared to their respective parental strain.  Table of Contents  Abstract  ii  Table of Contents  iv  List of Tables  vii  List of Figures  ix  Abbreviations  xi  Acknowledgement  xii  Chapter One: General Introduction  1  Neurulation  4  Mouse models for NTDs  8  The genetic basis for exencephaly in SELH/Bc mice  13  Multifactorial threshold model  16  Rationale and approach to my studies  19  Chapter Two: General Materials and Methods  23  Mice maintenance and conditions  23  Mouse stocks  23  Polymerase Chain Reaction and viewing products in gels  25  SSLPs  26  DNA preparation  27  Chapter Three: F2 exencephaly frequencies  32  Introduction  32  Materials and Methods  33  Results  37  iv  Discussion Chapter Four: F2 genotypes  48  Introduction  48  Materials and Methods  51  Results  54  Discussion  65  Chapter Five: Congenic line exencephaly frequencies  Chapter Six:  44  69  Introduction  69  Materials and Methods  72  Results  78  Discussion  88  Recombinant congenic lines  97  Introduction  97  Materials and Methods  98  Results  102  Discussion  106  Chapter Seven: Neural tube closure patterns in the congenic lines  109  Introduction  109  Materials and Methods  115  Results  123  Discussion  142  Chapter Eight: Conclusions and general discussion  152  References  162  V  Online references and sources  169  Appendix 1  170  vi  List of Tables Table 2.1: Locations o f the markers used in my studies i n the subsequent chapters.  30  Table 3.1: Frequency o f exencephaly and post-implantation mortality rates ("moles") i n F 2 generations from crosses between S E L H / B c and L M / B c i n all F 2 exencephaly studies.  38  Table 3.2: Exencephaly frequencies and post-implantation mortality rates ("moles") i n F 2 generations from crosses between S E L H / B c and L M / B c .  41  Table 3.3: Exencephaly frequencies within the "grandmother groups," subdivided by diet.  43  Table 4.1: Genotype summaries o f F2 ex encephalic embryos and their corresponding values at various genetic markers i n the EX-5001 study.  55  Table 4.2: Genotype summaries o f control normal F2 embryos and their corresponding x values at selected genetic markers i n the EX-5001 study.  57  Table 4.3: Genotype summaries o f F2 exencephalic embryos and their corresponding x values at various genetic markers i n the J H E X - 5 0 1 5 study.  61  Table 4.4: Genotype summaries o f control normal F2 embryos and their corresponding x values at selected genetic markers i n the J H E X - 5 0 1 5 study.  61  Table 5.1: Exencephaly frequencies and post-implantation mortality rates ("moles") i n the S E L H / B C and L M / B c strains and congenic lines on P M D #5015.  81  Table 5.2: Exencephaly frequencies and post-implantation mortality rates ("moles") i n the S E L H / B C and L M / B c strains and congenic lines onPLRD#5001.  84  Table 5.3: Exencephaly frequencies i n parental strains and congenic lines on P M D #5015 and P L R D #5001 for Sets 1 and 2.  85  Table 5.4: Values o f x (obtained from Falconer and Mackay, 1996) related to the combined exencephaly frequencies from Sets 1 and 2.  92  2  2  2  Vll  Table 6.1: Exencephaly frequencies and post-implantation mortality rates ("moles") i n the recombinant congenic lines on P M D #5015.  105  Table 7.1: Summary table o f the number o f litters collected i n the parental strains and congenic lines.  149  Table7.2: The number o f litters that were collected at various gestational ages.  149  Table 7.3: Mean somite count i n parental strains and congenic lines at all the stages.  150  Table 7.4: M e a n somite count i n the parental strains and congenic lines i n the sub-stages o f Stage 4 o f cranial neural tube closure.  151  viii  List of Figures Figure 1.1: Conceptual summary o f the normal process o f neural fold elevation i n the mouse.  7  Figure 1.2: Diagrammatic side-view o f a Day 8/9 embryo showing the locations o f independent zones o f neural fold elevation.  9  Figure 1.3: Side-view o f SELFI/Bc embryos with exencephaly.  14  Figure 1.4: Multifactorial threshold model using the frequency o f exencephaly in S E L H / B c as an example.  18  Figure 1.5 A : Schematic representation o f the creation o f a congenic line.  21  Figure 1.5B: Graphical representation o f the gain o f homozygosity back to the recipient strain and loss o f heterozygosity at backcross generations.  21  Figure 3.1: Conceptual diagram illustrating the genotype distributions for any particular locus o f the sex chromosomes i n F 2 segregants from reciprocal grandmother crosses.  36  Figure 5.1: A conceptual picture o f the construction o f the congenic lines.  70  Figure 5.2: The current chromosomal segments that are i n each congenic line.  79  Figure 5.3: Testing the multifactorial threshold model with the congenic lines.  91  Figure 5.4: Additivity o f the Exen loci.  94  Figure 6.1: Conceptual diagram o f the creation o f the recombinant congenic lines.  100  Figure 6.2: The current intervals on Chr 13 in the recombinant congenic lines.  103  Figure 7.1: Diagrammatic side-view representation o f the pattern o f cranial neural tube closure i n normal mouse strains.  110  Figure 7.2: Scanning electron micrographs o f cranial neural tube closure i n D a y 8/9 embryos.  112  Figure 7.3: Diagrammatic representations o f the sub-stages o f Stage 4 and the novel conformation that was observed i n sub-stage 4 B ' i n the congenic line embryos.  120  Figure 7.4: Comparison between L M / B c and S E L H / B c mean number o f somite pairs present at different stages o f closure.  125  Figure 7.5: Comparison between L M / B c and S E L H / B c mean number o f somite pairs present at the sub-stages for Stage 4.  126  Figure 7.6: Comparison between L M / B c and 1 3 S / L M mean number o f somite pairs present at different stages o f closure.  129  Figure 7.7: Comparison between L M / B c and 1 3 S / L M mean number o f somite pairs present at the sub-stages for Stage 4.  130  Figure 7.8: Light microscope photographs o f congenic line embryos i n certain stages and sub-stages.  131  Figure 7.9: Comparison between L M / B c and 5 S / L M mean number o f somite pairs present at different stages o f closure.  133  Figure 7.10: Comparison between L M / B c and 5 S / L M mean number o f somite pairs present at the sub-stages for Stage 4.  134  Figure 7.11 C o m p a r i s o n between S E L H / B c and 1 3 L / S E L H mean number o f somite pairs present at different stages o f closure.  136  Figure 7.12:Comparison between S E L H / B c and 1 3 L / S E L H mean number o f somite pairs present at the sub-stages for Stage 4.  137  Figure 7.13 C o m p a r i s o n between S E L H / B c and 5 L / S E L H mean number o f somite pairs present at different stages o f closure.  139  Figure 7.14:Comparison between S E L H / B c and 5 L / S E L H mean number o f somite pairs present at the sub-stages o f Stage 4.  140  X  Abbreviations A N P - anterior neuropore Chr - chromosome c M - centimorgans D L H P - dorsolateral hinge point E T n - early transposon E T n H - early transposon type II Exenl or Exen2 - Exenl or Exenl alleles from S E L H / B c EX-5001 - the F2 exencephaly panel collected from mice fed P L R D #5001, the normal diet. Collected by Drs. Diana Juriloff and M u r i e l Harris. JHEX-5015 - the F 2 exencephaly panel collected from mice fed P M D #5015, the highrisk diet. Collected by me, Julia Hoscheit. J N K - c- Jun N-terminal kinase M b - megabases M G I - Mouse Genome Informatics M H P - median hinge point N T D - neural tube defect P L R D #5001 - Purina Laboratory Rodent Diet #5001 P M D #5015 - Purina Mouse Diet #5015 Shh - Sonic hedgehog Somites - somite pairs S S L P - simple sequence length polymorphism T G Panel - the original F 2 exencephaly panel collected by Teresa Gunn. U C S C - University o f California Santa Cruz genome browser 5,  s  SS - homozygous S E L H / B c alleles L L - homozygous L M / B c alleles S L - heterozygous S E L H / B c and L M / B c alleles 1 3 L / S E L H - the congenic line that had been created by transferring the normal Exenl alleles from L M / B c into the S E L H / B c background. 5 L / S E L H - the congenic line that had been created by transferring the normal Exen2 alleles from L M / B c into the S E L H / B c background. 1 3 S / L M - the congenic line that had been created by transferring the S E L H / B c Exenl alleles into the L M / B c background. 5 S / L M - the congenic line that had been created by transferring the S E L H / B c Exen2 alleles into the L M / B c background. 7 S / L M - the congenic line that was created by transferring the S E L H / B c Exen4 alleles into the L M / B c background.  xi  Acknowledgement First and foremost, I would like to thank m y supervisors, D r . Diana Juriloff and Dr. M u r i e l Harris, for their time, support, guidance, and expertise. I am very appreciative o f the tools that were made available to me when I first came into their lab, such as the congenic lines. They had created them before I came into their lab, which I know was a task that demanded a lot o f time and effort and I feel fortunate to have been able use them in my research. I would also like to acknowledge the help o f former and present labmates, Diana M a h and Sarah Kennedy. Diana M a h prepared the D N A for the F2 exencephaly panel collected by Drs. Juriloff and Harris and genotyped the F 2 exencephalic embryos for some markers on Chrs 13 and 5. Sarah Kennedy was involved in the creation o f the recombinant congenic lines, genotyping many mice to look for recombinants to create the lines, as well as involved in typing some mice in the 13L/Rec6 line recombinant congenic line for a couple markers to further define boundaries on C h r 13.1 would also like to thank them for their conversations and friendship. Finally I would like to thank m y family and friends for their love and support, especially m y father Greg and his wife Pam, m y mother Cate and her husband Kevan, and m y sister Emily.  xii  Chapter 1: General Introduction Neurulation, the process o f forming the neural tube, is a fundamental event o f vertebrate embryogenesis. The neural tube is the early precursor to the brain and spinal cord and its proper closure is essential for the differentiation o f the central nervous system. This process is extremely complex involving numerous cell mechanisms and the combined effort o f several morphogenetic forces (Sadler, 2005; Colas and Schoenwolf, 2001; Copp et al., 2003). Failure o f the neural tube to close properly results in what are called neural tube defects ("NTDs"). Neural tube defects, the most frequent being spina bifida and anencephaly, are common birth defects that occur often i n the general population. Failure o f the posterior neural tube to close completely results in spina bifida, which, at birth, is characterized by congenital cleft o f the spinal column with hernial protrusion o f the meninges (coverings of the spinal cord) and sometimes the spinal cord (http://www.nlm.nih.gov/medlineplus/ mplusdictionary.html). Failure o f the cranial neural tube to close completely results i n anencephaly, which, at birth, is characterized by congenital absence o f all or a major part o f the brain. Spina bifida is debilitating and anencephaly is fatal. Together, the incidence of spina bifida and anencephaly is approximately 1/1000 but varies considerably worldwide (Boyles et al., 2005). Geographic location, ethnic background, and socioeconomic status affect the incidence o f N T D s . Although some N T D s can appear as part o f a syndrome, the majority are isolated, non-syndromic cases that do not have an identifiable cause but are attributed to a combination o f genetic and environmental variables. The non-Mendelian pattern o f increased recurrence risk with number affected in a family has pointed to a genetically multifactorial basis o f N T D . The recurrence risk  1  in first-degree relatives o f an affected child is approximately 3-5% and the recurrence risk for future children increases even more within a family with two affected children with an N T D , meaning the parents can be assumed to be carrying even more risk alleles across several gene loci (Fraser and Nora, 1975). Environmental variables, such as folate nutrition, have also been shown to influence the risk o f N T D s . The risk o f both recurrence and occurrence o f spina bifida and anencephaly has been observed to drop by up to 70% with folic acid supplementation to the maternal diet before conception and through the first trimester ( M R C Vitamin Study Research Group, 1991). Given the evidence for both genetic and environmental factors influencing the risk o f N T D s i n the human population, the majority o f N T D cases are considered to be multifactorial and they are best explained by the multifactorial threshold model (Fraser and Nora, 1975). To date no major genes for risk o f common human N T D s have been identified. Part o f the difficulty i n the genetic studies o f human N T D is limited availability o f families with several affected members, low reproductive fitness for people with spina bifida, fatality o f anencephalic newborns, and the unfeasibility o f studying human embryos. W i t h respect to genetics, the nature o f the genetic variants that interact to cause risk o f N T D s remains a mystery, adding to the difficulty o f genetic studies o f human N T D s . Whether the genetic variants that cause risk o f N T D are mutations or polymorphisms is not known, as w e l l as whether the alleles that contribute to multifactorial nonsyndromic N T D s are at the same loci as the syndromic mutations (Juriloff et al., 2001). Furthermore, it is not known i f multifactorial N T D s arise from cumulative effects o f the genetic variants in unrelated genes expressed i n interacting tissues i n neural tube closure or whether they arise from cumulative effects o f the genetic  2  variants from genes that are members o f the same gene regulatory pathway (Juriloff et al., 2001). M a n y candidate genes have been investigated in human N T D populations using case-control and family-based association studies. One o f the difficulties i n these types o f studies is having sufficient sample sizes to have the power to detect significance o f the genetic variant o f interest, especially since the cause o f N T D s appears to be highly heterogeneous. Generally the candidate genes that are studied are i n biochemical pathways, o f most interest the folate metabolic pathway, developmental pathways, and those that cause N T D s i n mice when "knocked out." Several o f the studies fall short o f showing significant results, some have been suggestive, and only a few have yielded positive results (Reviewed i n Boyles et. al, 2005). A polymorphism o f 5,10Methylenetetrahydrofolate reductase (MTHFR)  (677C—>T), a key enzyme i n the  methylation cycle (folate metabolic pathway) that has 50-60% lower activity than the wild-type variant, has been associated with N T D s i n some populations, specifically the Irish and Dutch (Shields et al., 1999; van der Put et a l , 1995), but not i n others (Rampersaud et al., 2003). Recently, a single nucleotide polymorphism (SNP) within the neural cell adhesion molecule 1 (NCAM1) gene was significantly associated with risk o f N T D s i n American Caucasian spina bifida families suggesting variations i n NCAM1 may influence risk for human N T D s (Deak et al., 2005). Most recently, a genome-wide linkage screen on families having at least two related individuals with N T D s revealed regions o f interest on chromosomes 7 and 10, possibly, as Rampersaud et al. state, "providing valuable positional data for prioritization o f candidate gene assessment i n future studies o f N T D s " (Rampersaud et al., 2005). Failure to identify major causal genes  3  for human multifactorial N T D s and lack o f positive results coming from the human N T D field have led the way for animal models to provide valuable insight into the critical pathways and cellular mechanisms that are involved i n proper neural tube development and the causes o f N T D s .  Neurulation In humans, development and closure o f the neural tube is usually completed within 28 days o f conception. Given that this is a time that is inaccessible to study the developing neural tube i n human embryos and given the ethical concerns, neurulation has been extensively studied i n the mouse (as well as other animal models). Neurulation is typically divided into two phases called primary and secondary neurulation. Primary neurulation creates the brain and most o f the spinal cord and w i l l be discussed below. Secondary neurulation occurs at the more caudal levels and creates the lowest portion o f the spinal cord and tail and w i l l not be discussed here, as it is not relevant to this thesis. Described by Copp et al., "this process involves mesenchymal cells in the dorsal part o f the tail bud undergoing condensation and epithelialization to form the secondary neural tube, the lumen o f which is continuous with that o f the primary neural tube" (Copp et al., 2003). These neurulation processes take place across embryonic days 7.5 through 10 in the mouse. Given here is a general description o f primary neurulation with emphasis on the cranial region. Primary neurulation is a process that involves shaping, folding, and fusing o f bilateral neural folds (Copp et al., 2003). Conceptually, it is described as a developmental process that results i n the rolling up o f a flat layer o f ectodermal cells into an elongated  4  hollow tube (Colas and Schoenwolf, 2001). O n a tissue level, primary neurulation begins with the formation o f the neural plate followed by shaping o f the neural plate, bending o f the neural plate (elevation o f neural folds), and fusion o f the bilateral neural folds. The neural plate is derived from ectoderm, one o f the three germ layers that are formed during gastrulation, differentiating v i a induction from the underlying notochordal plate (Sadler, 2005). Cells destined to form the neural plate elongate i n an apical-basal direction to form a thickened region o f specialized ectoderm, hence forming the neural plate. When viewed dorsally or ventrally, the neural plate is shaped like a "spade," being relatively wide at the anterior end and narrowing posteriorly. Shaping o f the neural plate involves it continuing to thicken apicobasally and, additionally, undergoing elongation by convergent extension (Colas and Schoenwolf, 2001). Described by Keller et al. convergent extension is "a process i n which laterally placed cells move toward and are intercalated into the midline narrowing and lengthening the neural plate" (Keller et al., 2000). The first initiation o f neural tube closure, occurring i n the cervical region, requires convergent extension, as shown by mouse knockouts for genes involved i n the planar cell-polarity pathway that have severe N T D s (Reviewed i n Copp et al., 2003). A s the neural plate undergoes elongation, the underlying mesoderm does so as w e l l by similar cell movements so that the entire body axis lengthens (Sadler, 2005). Bending o f the neural plate is initiated as its shaping is underway. Firstly, bending involves the formation o f the neural folds, comprised o f neuroepithelium (neuroectoderm) and underlying mesenchyme within a surface ectoderm (Harris and Juriloff, 1999), at the lateral borders o f the neural plate (Colas and Schoenwolf, 2001). This involves expansion o f the cranial mesenchyme, which i n turn begins the elevation  5  phase o f the cranial neural folds (Copp et al., 2003). The neural folds elevate from the median hinge point ( " M H P " ) , the first principal site o f bending, and establish a troughlike space called the neural groove, which becomes the lumen o f the primary neural tube after fusion o f the bilateral neural folds (See Fig. 1.1). The M H P overlies the notochord and extends the entire neural axis. The second principle sites o f bending are the paired dorsolateral hinge points ( " D L H P s " ) along the sides o f the neural folds. Although they do not form i n most regions o f the spine, they form i n the cranial region and intermediate spine (Copp et al., 2003; Sadler, 2005). This is required for the neural folds to turn inwards (flip around) so that they can approach each other and eventually make contact. Various mechanisms, such as contraction o f sub-apical actin microfilaments, emigration o f the neural crest, maintenance o f a proliferative neuroepithelium, and programmed cell death, are all thought to have a role i n shaping (elevation) and dorsolateral bending (Copp et al., 2003). Bending o f the neural plate and formation and elevation o f the bilateral neural folds ultimately brings them into apposition with each other at the dorsal midline where they make contact and fuse, establishing the " r o o f o f the neural tube, a continuous layer o f cells across the neural groove (Colas and Schoenwolf, 2001). The overlying surface ectoderm o f the neural folds fuses across the midline, as well, to become the overlying epidermis, separating it from the neural tube. Adhesion o f the neural folds is thought to involve cellular protrusions extending from the cells on the tips o f the neural folds, interdigitating as the folds come into contact (Geelen and Langman, 1979). A s discussed in Copp et al., once the neural folds adhere to one another, programmed cell death is  6  Figure 1.1: Conceptual summary o f the normal process o f neural fold elevation in the mouse. In panel a, the neural plate has bent forming the bilateral neural folds and they have started elevating. In panel b, the median hinge point and dorsolateral hinge points have formed so that the folds are completing elevation and the tips are apposing each other. In panel c, the neural folds have fused and have gone under epithelial remodeling to establish separate surface ectodermal and neuroepithelial continuity across the midline. This figure is from Harris and Juriloff (1999).  neural crest cells  riewotctodefm mcMfKfiprta surfact ectoderm  1} basal laminae  dorsolateral Mnpt point median hinge point notodhord  pH  lumen of neural tube  involved i n "remodeling the epithelia i n order to break the continuity between the neuroepithelium and surface ectoderm on each side, therefore establishing separate surface ectodermal and neuroepithelial continuity across the midline" (Copp et al., 2003).  Mouse models for NTDs Most o f the current knowledge about N T D s comes from animal models. There are over 100 single-gene mutations (and counting) i n mice that have been reported to affect neurulation and cause N T D s . The majority o f these mutations are knockouts o f genes (null mutations), each o f which has a homologous human gene (Harris, 2001; reviewed i n Harris and Juriloff, 1999, and Juriloff and Harris, 2000). This shows that the misregulation o f any number o f different genes, related directly or indirectly to neurulation, may result i n N T D s . This diversity demonstrates the extent to which the cause o f N T D s can be heterogeneous. In the mouse mutants, the locations o f N T D s appear to be related to specific zones suggesting that there are different mechanisms specific to each type o f N T D (Juriloff and Harris, 2000) (See F i g . 1.2). Failure o f elevation i n the specific zones leads to split face (failure o f elevation i n Zone A ) , exencephaly (failure o f elevation i n Zone B ) , rachischisis (failure o f elevation i n Zone D), or spina bifida (failure o f elevation o f caudal Zone D ) . The regional location o f N T D s (brain versus spine) differs between the mutants; most mutations cause exencephaly, several cause exencephaly and spina bifida and a few only cause spina bifida. For those mutations that cause only exencephaly or cause only spina bifida, these suggest that there are region-specific differences i n neurulation-related gene expression (Copp et al., 2003).  8  Figure 1.2: Diagrammatic side-view o f a Day 8/9 embryo showing the locations o f independent zones o f neural fold elevation. Locations o f N T D s in mouse mutants appear to be related to these specific zones. This figure has been modified from a figure in Harris and Juriloff (1999).  Exencephaly  Split face -f M  "  \  /  JJ- i  ^ Rachischisis  Spina bifida  From these knockout mutants, several crucial molecular pathways have been identified for normal neurulation, as well as many cellular mechanisms. Knockouts for genes i n the planar cell-polarity pathway (Reviewed i n Copp et al., 2003) such as, but not limited to, disheveled, loop-tail, crash, and circletail, have made it clear that convergent extension is required during the initiation o f neural tube closure (Curtin et al., 2003; Hamblet et al., 2002; K i b a r et al., 2001; Montcouquiol et al., 2003; Murdoch et al., 2001; Murdoch et al., 2003). These mouse mutants fail to initiate contact i n the cervical region (Closure 1), which subsequently leads to craniorachischisis (most o f the neural tube remains open). Knockouts for genes i n the sonic hedgehog (Shh) pathway (Reviewed in Copp et al., 2003) provide evidence that these genes regulate neural plate bending. Shh is thought to negatively regulate dorsolateral hinge point formation ( " D H L P " ) in the neural folds down the neural axis (Copp et al., 2003) and this has been supported by Shh-null mutants, where D L H P s are observed i n the neural folds and the neural tube closes successfully (Ybot-Gonzalez et al., 2002). In addition, Shh over-expression has been shown to produce N T D s (Echelard et al., 1993). The Patchedl (Ptcl) receptor, responsible for maintaining Shh signaling in an off-state in the absence o f Shh ligand, when knocked-out produces severe N T D s in Ptcl-mx\\ mice (Goodrich et al., 1997). Other mutants such as extra-toes, open-brain, and Zic2 disrupt Shh signaling and have N T D s as well (Gunther et al., 1994; H u i et al., 1993; Nagai et al., 2000). Mouse mutants have been successful in identifying cellular mechanisms important for cranial neurulation (Reviewed in Copp et al., 2003). T w o mouse knockouts for genes expressed i n the mesenchyme that are important for proliferation, Twist and Cartl, have cranial N T D s , providing support that expansion o f the cranial mesenchyme is  10  necessary for cranial neural tube closure (Chen and Behringer, 1995; Zhao et a l , 1996). Mouse knockouts i n cytoskeleton-related genes such as Mena, profilin (Pfnl) (Lanier et al., 1999), and non-receptor tyrosine kinases Abl and Arg (Koleske et al., 1998), the functions o f which are actin-related, have exencephaly, providing support that cranial neurulation is highly dependent on the actin cytoskeleton. Knockouts i n apoptosis-related genes have N T D s as w e l l providing support that programmed cell death is important i n neurulation. Interestingly, knockouts o f genes that lead to increased or decreased apoptosis are attributed to the cause o f N T D s . The Apafl, caspase9, and p53 (to name a few) knockouts all have reduced apoptotic cell death i n association with the development o f exencephaly (Harris (B.S.) et al., 1997; Honarpour et al., 2001; K u i d a et al., 1998; Sah et al., 1995) whereas some knockouts such as AP-2, ApoB, Tulpl (to name a few) have increased apoptosis i n the cranial neural folds presumably associated with development o f exencephaly (Schorle et al., 1996; Zhang et al., 1996; Kohlbecker et al., 2002; Homanics et al., 1995; Ikeda et al., 2001). Although many mouse mutants have been helpful i n identifying critical pathways and cellular mechanisms/molecules involved i n neurulation, many o f them are syndromic; the N T D i n the mouse (most likely exencephaly) is part o f a severe syndrome that includes other primary defects in other developmental systems as w e l l (Harris & Juriloff, 1999; Juriloff & Harris, 2000). Thus, these mouse mutants with single-gene mutations do not reflect what is observed i n the majority o f human N T D cases, the majority being non-syndromic and multifactorial, and therefore may be considered o f limited value as a model for human N T D . O n the plus side, there are some established  11  mouse models for spontaneous non-syndromic, multifactorial N T D s . T w o o f them are discussed below.  Curly tail mouse model Curly tail (ct) is the most w e l l understood mouse model for nonsyndromic N T D s , especially for posterior N T D s , and has been extensively studied for the past 50 years (van Straaten and Copp, 2001). This mouse model is characterized b y a high frequency o f tail flexion (considered to be a m i l d N T D ) (50%), some spina bifida (10%), and a low frequency o f exencephaly (3%). N T D s i n ctlct mice are similar i n location and form to those occurring i n humans, as they have spina bifida and/or exencephaly. The principle ct gene maps to distal chromosome 4, and the phenotype and incidence o f N T D are influenced b y several modifier genes and by environmental factors (Reviewed i n van Straaten and Copp, 2001), making this genetic system "multifactorial." The incompletely penetrant ct phenotype is mainly observed i n ct homozygotes (recessive), but can be observed i n some heterozygous mice i n the presence o f particular combinations o f modifier alleles. It is a hierarchical genetic system (major gene and modifiers) where embryos need ct i n order to have a N T D and therefore the polygenic additive version o f the multifactorial threshold model, based on the idea that the genetic factors are interchangeable and additive, cannot be applied to this model to explain the cause o f N T D s i n curly tail.  12  SELH/Bc mouse strain The S E L H / B c strain, the focus o f this study, is a well-established mouse model for common N T D s , specifically exencephaly (the equivalent to human anencephaly). This mouse stock was developed by Drs. Diana Juriloff and M u r i e l Harris i n the Department o f M e d i c a l Genetics at the University o f British Columbia and the origin o f the strain is described i n detail i n Juriloff et al. (1989), as w e l l as briefly i n General Materials and Methods (Chapter 2). S E L H / B c is characterized by a high frequency o f exencephaly (10-30%) and all S E L H / B c embryos exhibit an abnormal mechanism o f neural tube closure whether they become exencephalic or not (Discussed further in Chapter 7). The exencephaly i n S E L H / B c is generally over the midbrain region (See F i g . 1.3). The genetic cause is attributed to at least 3 interchangeable loci acting together additively, closely resembling common human N T D s . In addition to the multiple loci, environmental factors also influence the incidence o f exencephaly i n S E L H / B c . Different diets alter the risk o f exencephaly i n S E L H / B c . Harris and Juriloff state that "dietary supplementation o f folic acid, methionine, zinc, niacin, brewers' yeast,  riboflavin,  vitamin B12, or inositol does not significantly reduce the risk o f exencephaly" (Harris and Juriloff, 2005).  The genetic basis of exencephaly in SELH/Bc mice Previous studies have shown the genetic cause o f exencephaly to be attributed to the additive effects o f at least three genes (Juriloff et al., 1989; G u n n et al., 1992; Juriloff et al., 2001). The first study determining the mode o f inheritance o f the exencephaly-  13  Figure 1.3: Side-view o f S E L H / B c embryos with exencephaly. Picture A shows an S E L H / B c embryo with exencephaly at 14 days o f gestation ( E l 4 ) ; note that the exencephaly spans the midbrain region. Picture B shows an S E L H / B c fetus ( E l 8 ) where the brain tissue has degraded. This figure is from Macdonald et al. (1989).  14  causing genes i n the S E L H / B c strain was based on segregants from a cross o f S E L H / B c to a normal related strain ICR/Be. Exencephaly frequencies from classical crosses such as reciprocal crosses between the two strains, and backcrosses, indicated that the risk o f exencephaly was due to the additive effects o f two or three loci and that it was probably a multifactorial threshold trait (Juriloff et al., 1989). The next study further investigating the genetic basis o f exencephaly in S E L H / B c involved classical crosses between S E L H / B c and an unrelated normal strain, S W Y / B c . Similar results were obtained from these segregants, as well, providing support for the hypothesis o f two to three additive genes contributing to the risk o f exencephaly i n S E L H / B c (Gunn et al., 1992). The subsequent study that investigated the genetic basis o f exencephaly i n S E L H / B c was designed to identify the chromosomal locations o f the S E L H / B c exencephaly risk genes and is described i n Juriloff et al. (2001). This involved typing o f markers across the genome i n the extremes o f F2 males (those that produced the most exencephaly i n offspring and those that produced none), derived from crosses to a nonrelated strain, L M / B c . In addition, an F2 exencephaly panel was collected to examine genotypes o f F2 exencephalic embryos at markers i n the provisional regions for exencephaly risk genes from S E L H / B c revealed i n the F2 sire screen. Together these studies confirmed that the risk o f exencephaly i n the S E L H / B c strain was due to the cumulative effect o f two or three loci. They were named Exenl, Exenl, and Exen3. The Exenl and Exen2 loci were mapped to m i d Chr 13 and distal C h r 5, respectively, and the Exen3 locus provisionally to Chr 11.  15  Multifactorial threshold model  The majority o f human N T D s can best be explained by the multifactorial threshold model (Fraser and Nora, 1975). They are not inherited i n a simple Mendelian manner. The concept is that there is a normally distributed continuous developmental variable i n a population, referred to as "liability," and a developmental threshold separating the population into those that are affected with a birth defect and those that are normal (unaffected) (Falconer and Mackay, 1996; Fraser, 1976) (See F i g . 1.4). One either has an N T D or does not; there is no intermediate phenotype. Combinations o f several genetic factors and environmental factors determine the liability to a birth defect, making the system multifactorial (Fraser, 1976). A s the number o f factors that contribute to the trait increases, the liability for the defect increases. When it reaches a threshold, the liability is so great that the defect results. Falconer and Mackay state that "the continuous variation o f liability may be thought o f as the rate o f some developmental process or it may be a compound o f several different physiological or developmental processes, but it is not necessary to know how these are combined to give the liability, or even to know what they really are" (Falconer and Mackay, 1996), which is the part o f the beauty o f the model. Not only do genetic and environmental factors influence the position o f the distribution relative to the threshold, they can influence the position o f the threshold itself as well (Fraser, 1976). For N T D s , such as exencephaly i n the S E L H / B c strain and anencephaly i n humans, the continuous variable or liability trait is most likely to be timing o f cranial neural fold elevation. Embryos with delayed elevation w i l l have a higher probability o f not completing elevation i n their "window o f opportunity" when they should be capable  16  o f doing so. This biological limit, the latest point in development when the neural folds are able to complete elevation and fuse, is the developmental threshold. Normal distribution curves are used to represent the population o f interest and it is assumed the population has a normal distribution for the liability. For genetic analyses, the proportion or percentage o f individuals that are affected, the incidence, must be converted to mean liabilities because on a percentage scale variances differ according to the mean and changes i n percentages may appear to be non-additive when i n actuality they are additive on the multifactorial threshold scale. In order to make the transformation it is necessary to define the liability as being normally distributed within a population (Falconer and Mackay, 1996). If the liability is normally distributed, then the unit o f liability is its standard deviation (a) and the mean liability is related to the incidence by the normal deviate x, which is the deviation o f the threshold from the mean in standard deviation units o f liability (See Fig. 1.4). Values o f x for different incidences have been tabulated and can be obtained from Falconer and M a c k a y (1996). When comparing two or more populations the threshold must be defined as being fixed and the liability is assumed to be variable.  17  Figure 1.4: Multifactorial threshold model using the frequency o f exencephaly i n S E L H / B c as an example. 2 5 % o f S E L H / B c embryos cross the developmental threshold and become exencephalic. U s i n g the table i n Falconer and M a c k a y (1996) that shows the tabulated x values, the proportion 25% relates to 0.674 standard deviation units (a). In other words, the threshold o f S E L H / B c is 0.674 a o f liability away from the mean.  18  Rationale and approach to my studies M y studies further investigated several aspects o f the multifactorial nature o f exencephaly i n the S E L H / B c strain using two core tools, F 2 exencephaly panels and congenic lines.  F2 exencephaly panels Given that exencephaly is a complex trait i n the S E L H / B c strain, F 2 exencephaly panels, derived from crosses between S E L H / B c and the normal non-related strain, L M / B c , were collected to examine the genotypes from F 2 exencephalic embryos across the Exen loci. F 2 exencephaly panels are useful because they can pick up all genotype combinations that contribute to the exencephaly. Basically, Exen alleles from S E L H / B c are expected to be transmitted more than the normal Exen alleles from L M / B c i n the F 2 exencephalic embryos. The F 2 exencephalic embryos would have more Exen alleles from S E L H / B c across a combination o f the Exen loci, as well. Past genetic analyses, which involved the original F2 exencephaly panel, mapped the Exenl, Exen2, and Exen3 loci, but the chromosomal regions where the Exen loci were mapped to remained fairly large. T w o new F2 exencephaly panels were collected, one from mice fed the regular diet P L R D #5001 and one from mice fed the high-risk diet P M D #5015, to refine the mapping o f the Exen loci and to examine the diet effect i n F2 segregants.  19  Congenic lines Congenic lines are a powerful tool to map genes i n a complex trait, and additionally, they are used to study the effects o f the individual genes that contribute to it. Darvasi describes a complex trait as, "a trait in which a one-to-one relationship between genotype and phenotype does not exist" (Darvasi, 1998). Congenic lines dissect the complex trait so that a "one-to-one relationship between genotype and phenotype" can exist. This can be specifically useful for multifactorial threshold traits, to determine the amount o f risk each gene contributes. Congenic lines are created by repeated backcrosses to the background recipient strain. W i t h each subsequent backcross, the residual heterozygosity is reduced by about 50% due to random segregation, eliminating alleles from the donor strain at all the other loci affecting the trait, and therefore increasing the homozygosity for the recipient strain (Silver, 1995) (See Fig. 1.5). A t each generation, only those offspring who have received the donor allele from the donor strain at the locus o f interest were selected for the next round of backcrossing. A t a certain selected generation, the mice with the desired genotype are intercrossed and homozygotes for the desired chromosomal segment are selected to establish the interval-specific congenic line. Congenic lines had been made by transferring the Exenl and Exen2 alleles, respectively, from S E L H / B c into a normal strain background, L M / B c , and reciprocally substituting normal L M / B c Exenl and Exen2 alleles, respectively, into the S E L H / B c background. They were created to serve multiple purposes. One o f the main objectives o f the congenic lines was to refine the mapping o f the Exen genes. This partly involved defining the current intervals o f the transferred chromosomal segments by typing various  20  Figure 1.5: A ) Schematic representation o f the creation o f a congenic line. A t each backcross generation to the recipient strain (black boxes) the residual heterozygosity (checkerboard pattern) is reduced by 50%. A t each generation the donor allele/differential segment is selected for with genetic markers. B ) Graphical representation o f the gain o f homozygosity back to the recipient strain (Y-axis) and loss o f heterozygosity at backcross generations. Both figures from Silver (1995).  markers to determine the borders between the transferred chromosomal segment and the recipient background. The other part entailed creating recombinant congenic lines from one o f the main congenic lines to refine the mapping specifically for the Exenl locus on Chr 13. The other general objective o f the congenic lines was to study the effects o f the individual Exen genes. Firstly, they were used to investigate the individual effect o f the Exen alleles from SELFf/Bc on exencephaly frequency, testing the multifactorial threshold model. It predicts that each congenic line with part o f the liability to exencephaly from S E L H / B c w i l l express part o f the exencephaly frequency o f S E L H / B c . Secondly, the congenic lines were used to test whether certain Exen loci respond to diet. Thirdly, they were used to isolate the individual effect o f each Exen locus i n the process of cranial neural fold elevation i n early embryos.  This thesis presents the work investigating the multifactorial cause o f exencephaly in the S E L H / B c strain using F2 exencephaly panels and congenic lines. This work w i l l advance the general understanding o f the development o f exencephaly i n the S E L H / B c strain, show that the multifactorial threshold model can be used to explain complex traits such as exencephaly, and provide insight into the mechanisms underlying human neural tube defects.  22  Chapter 2: General Materials and Methods  This chapter describes the Materials and Methods that are common to more than one o f the studies described i n the subsequent chapters.  Mice Maintenance and conditions The mice were maintained i n the animal unit i n the Department o f Medical Genetics at the University o f British Columbia in windowless rooms. The mice were housed in standard polycarbonate "shoe-box" cages (5" x 11" x 7") with stainless steel lids and microisolator filter tops and supplied with acidified water ad libitum (pH - 3.1 by HC1) and food ad libitum. The room was maintained on a light cycle from 6:00 A M to 6:00 P M with a temperature o f approximately 70°F. Density o f mice was 2-5 per cage. They were kept specific pathogen free.  Mouse stocks SELH/Bc S E L H / B c is an inbred strain at F48+. The S E L H / B c mice are non-agouti, black chinchilla ( a a B B c c ) . The S E L H stock was derived from a cross o f a partially inbred ch  ch  stock o f mixed genetic background ( B A L B / c G a , 129/-, C B A / - ) to "random-bred B L U : H a (ICR)." The mixed genetic background was homozygous for the lidgap-Gates (lg ) Ga  mutation that was backcrossed onto the I C R background b y intercross-backcross to N 3 . A t that time a new recessive mutation, spherocytosis British Columbia, sph  2Bc  (Unger et  al., 1983) appeared i n the second intercross generation ( N 3 X N 3 ) . The parents o f the  23  affected pups were used to begin a new stock segregating for sph maintained by brother-sister inbreeding with selection for sph  2Bc  lg  c  and it was  carriers and against the  mutation. It was i n this stock, i n the F5 newborns, that exencephaly was observed.  Ga  The exencephaly-producing parents were selected for i n subsequent generations, while sph  2Bc  was selected against, and it eventually became an inbred strain by sister/brother  inbreeding. Most recently, the lg  Ga  mutation i n B A L B / c G a mice was found to be an  eight-exon deletion i n the mouse Map3kl gene on distal C h r 13 (Juriloff et al., 2005). In addition to exencephaly, there have been more independent mutations observed in S E L H / B c over the past two decades. They include one at the nude locus, nuBc (Koehn et al., 1988), three at the albino locus (Juriloff et al., 1992; Juriloff et al., 1994; Hofmann et al., 1998), one called whiskers amiss (warn) that causes breakable whiskers and disheveled-appearing body hair that has been mapped to the region o f the type 1 keratin cluster on C h r 11 (Taylor et al., 2000), one that causes a white belly spot that has been mapped to the Kit gene on C h r 5 (Kit ), w  and one that causes the lens o f the eye to go  opaque that has not been mapped (Juriloff et al., 2005). A l l new mutations that arose had been purged from the main S E L H / B c lineage. Ovarian teratomas have also been observed in this strain (personal communication). The apparently high spontaneous mutation rate in S E L H / B c indicates that this stock may be genetically unstable. Interestingly, the S E L H / B c strain appears to have a high level o f retrotransposition o f early transposons ("Etas"). One o f the mutations at the albino locus mentioned above was found to be due to an E T n U insertion i n exon 1 (Hofmann et al., 1998). Most recently, Baust et al. found that the S E L H / B c strain has a highly transcribed  24  variant o f the E T n l l retrotransposon suggesting that early retrotransposition occurs relatively frequently i n the S E L H / B c strain (Baust et al., 2003).  LM/Bc L M / B c is a highly inbred strain that has been used as the control strain for S E L H / B c i n previous studies as they have virtually no spontaneous production o f exencephaly. They were derived by sib inbreeding o f mice that carried 1/8 o f their genes from the S W V / B c strain and 7/8 o f their genes from an inbred strain that had been derived from an unpedigreed stock o f " C 3 H " mice (Juriloff et al., 1991). These mice are homozygous for lg , m  a mutation that causes open eyes at birth. The animals used i n this  study were at F84 or higher.  Polymerase Chain Reaction and viewing products in gels Simple sequence length polymorphisms ("SSLPs") were used as genetic markers to genotype target D N A . P C R primers designed to detect S S L P s were obtained from Research Genetics, Inc. (Huntsville, Alabama, U S A ) , N A P S (University o f British Columbia), or Sigma (Oakville, Ontario ). Some S S L P primers were designed i n our lab using the website http://seq.yeastgenome.org/cgi-bin/web-primer (See below). Each P C R reaction was carried out i n a 25 u l volume overlaid with mineral o i l i n a 650 u.1 reaction tube. Each reaction contained 100 ng o f D N A and 0.14 u M o f each forward and reverse primer. The rest o f the reaction mixture contained "master m i x " , consisting o f d A T P , d G T P , d C T P , and d T T P (final concentration, 50 u M each; Qiagen), Taq D N A polymerase (final concentration, approximately 1.25 Units; Invitrogen), 10X P C R Buffer  25  (final concentration, 10 u M ; Invitrogen), and magnesium chloride (final concentration usually 1.5 m M , range 1.5-2.5 m M ; Invitrogen). P C R was usually performed i n a PerkinElmer 4600 thermocycler, usually under the following conditions: 4.5 min. at 94°C (denaturation), followed by 35 cycles o f 30 seconds at 94°C (denaturation), 30 seconds at 55°C (annealing) and 30 seconds at 72°C (extension), and then followed by 7 minutes at 72°C. Some primers required different annealing temperatures ranging from 52°C - 55°C. Some markers required a "hot-start," where the master m i x was added at 94°C after the D N A and primers were denatured. The marker dye bromphenol blue-xylene cyanol F F (5 pi) was added to the P C R product and approximately one third o f the mixture (10 ul) was run by electrophoresis on 4% NuSieve 3:1 Agarose horizontal gels containing 0.5 ug/ml o f ethidium bromide. Gels were run i n I X T A E at about 150 V for 1-1.5 hours, then observed and photographed under U V light using Polaroid 667 film.  SSLPs Simple sequence length polymorphisms ("SSLPs") were used to type the F2 exencephalic embryos and all other mice involved i n m y studies. They are highly polymorphic between inbred strains o f mice and can be typed quickly and easily using the Polymerase chain reaction. The markers used for this study were designed i n our lab, with the exception o f the M i t markers. Generally, we used University o f California Santa Cruz assemblies (http://genome.ucsc.edu/cgi-bin/hgGateway') to look up genes i n the vicinity o f the area we were interested i n acquiring primers for. The sequence o f the gene was retrieved and simple dinucleotide repeats, repeated at least 20 times, were searched  26  for. If there was good sequence that surrounded the repeat, consisting o f a random distribution o f all nucleotides, it was copied into a web program that would give us some primer sequences (http://seq.yeastgenome.org/cgi-bin/web-primer). The primer sequences were ordered from N A P S or Sigma. The primer sequences that were made i n our lab can be found i n Appendix 1.  DNA preparation Phenol chloroform method D N A was prepared from stored frozen embryos by standard phenol chloroform extractions by Diana M a h for the EX-5001 F2 exencephaly panel (See Chapter 4). Approximately half o f the embryo was used. It was washed i n sterile saline solution before being frozen. O n ice, the embryonic tissue was transferred to a 15 m l tube and broken apart. 3.5-5 m l o f Lysis Buffer was added to each tube and then 200 p i o f 10 mg/ml Proteinase K was added next and shaken gently to m i x everything together. The tubes were then taped to a metal rack and put into a 60°C water-bath set at 70-80 rpm. They were left there overnight. The next day 1 volume o f phenol chloroform was added to the tube containing the embryonic tissue and centrifuged for 15 minutes at 2200 rpm. The top aqueous layer was then transferred to a tube containing phenol chloroform (1:1 mixture) and then centrifuged again for 15 minutes at top speed. The top layer was transferred to a tube containing chloroform and then centrifuged again for another 15 minutes. The new top layer was then transferred to an empty tube and mixed with 2 volumes o f 100% ethanol. After gentle mixing the ethanol was poured out and the D N A was washed with 70% ethanol. The ethanol was decanted and the D N A was left to dry in  27  the fumehood for about 10 minutes. It was then transferred to a 1.5 m l tube and suspended in 100-300 p i 1 X T E depending on the size o f the tissue.  QIAamp DNA Mini kit (Tissue Protocol) D N A was prepared from stored frozen embryos by using the QIAamp D N A M i n i K i t (Tissue Protocol) for the J H E X - 5 0 1 5 F2 exencephaly panel. Approximately 25 mg o f the embryo was used and the other extra tissue was kept frozen for later use and back up. The embryonic tissue was cut with a razor blade and transferred to a 1.5 m l microcentrifuge tube and 180 p i o f Buffer A T L was added. In addition, 20 p i o f Proteinase K was added and mixed by vortexing and incubated at 56°C until the tissue was completely lysed. Lysis lasted 1-3 hours. After lysis, 200 p i o f Buffer A L was added to the sample and mixed by vortexing and incubated at 70°C for 10 minutes. After incubation, 200 ul o f 95% Ethanol was added to the sample and mixed to create a homogenous sample. The sample was then transferred to a QIAamp spin column and centrifuged at 8000 rpm for 1 minute. The spin column was then transferred to another tube and 500 p i o f Buffer A W 1 was added and then centrifuged at 8000 rpm for 1 minute. This step was repeated with Buffer A W 2 and the whole spin column was centrifuged at top speed (13,200 rpm) for 3 minutes. The spin column was then transferred to a 1.5 m l tube and 200 p i o f Buffer A E was added and incubated at room temperature for 5 minutes. After 5 minutes the sample was centrifuged at 8000 rpm for 1 minute and then placed i n the refrigerator for storage i n a 1.5 m l tube.  28  QIAamp DNA Mini kit (Crude lysis protocol) This method was generally used to prepare D N A from stored frozen tail-tip tissue samples. The tail tissue was placed i n a 1.5 m L microcentrifuge tube, and smashed with the "hard end" o f yellow inoculating loops. 200 u l o f lysis buffer A T L was added, followed by 20 u l Proteinase K and 200 ul Buffer A L . The mixture was vortexed for 15 seconds and then incubated at 56°C for 10 minutes. 400 u l o f 9 5 % ethanol was added and vortexed for 15 seconds to homogenize the lysate. Subsequently 635 u l o f the lysate was applied to a QIAamp spin column (in a 2-ml collection tube) and centrifuged at 8000 rpm for 1 min. The spin column was transferred to a new collection tube and the remaining lysate was loaded and the centrifuge was repeated again. A t this point 500 ul Buffer A W 1 was loaded into the spin column and centrifuged at 8000 rpm for 1 min. The spin column was then transferred to another new collection tube where 500 u l Buffer A W 2 was loaded into the spin column and it was centrifuged at the maximum rpm, 13,200 rpm, for 3 minutes. The spin column was transferred to a new L 5 m l tube and 50 u l Buffer A E was added and then incubated at room temperature for 5 minutes. The tube (including the spin column) was then centrifuged at 8000 rpm for 1 minute. The D N A , suspended i n Buffer A E , was stored i n the refrigerator for future use i n the 1.5 m l tube.  29  Table 2.1: Locations o f the markers used i n m y studies i n the subsequent chapters. Megabase values (Mb) were obtained from the University o f California Santa Cruz ( U C S C ) genome browser (generally the M a y 2004 contig) (http://genome.ucsc.edu/cgibin/hgGatewav) and centimorgan (cM) values were obtained from Mouse Genome Informatics ( M G I ) (http://www.informatics.iax.org/mgihome/nomen/strains.shtml).  Marker  Chr  D13MU3  M b (UCSC)  C M * (MGI)  13  19.8  10.0  D13MU117  13  37.0  20.0  D13MU39  13  —  37.0  Fdg3-C/D  13  47.8  31.0  Fgfr4-E/F  13  54.2  33.0  Gprk6-C/D  13  54.5  35.0  D13MU13  13  55.6  35.0  Ntrk2-C/D  13  57.9  36.0  Fancc-C/D  13  61.6  36.0  Ptcl-A/B  13  61.8  36.0  Adcy2-A/B  13  64.6  41.0  Nkd2-A/B  13  69.9  —  Nr2fl-A/B  13  74.3  45.0  G2151-A/B  13  82.0  E4U3-A/B  13  85.0  — —  D13MU193  13  88.4  43.0  D13MU30  13  98.2  52.0  D13MU76  13  107.4  61.0  D13MU78  13  115.8  75.0  D5MU13  5  35.9  20.0  D5Nds2  5  71.4  41.0  D5MU24  5  111.4  60.0  D5MU95  5  122.4  68.0  D5MU30  5  127.1  72.0  Gats-C/D  5  131.8  —  D 5MM 68  5  134.5  78.0  D5MU122  5  147.6  85.0  D11MU14  11  98.4  57.0  DllMitlO  11  104.3  63.0  Scn4a-C/D  11  106.0  64.0  Other  Not mapped in U C S C  Not i n M G I Not i n M G I Not i n M G I  Not i n M G I  30  D7MU75  7  5.9*  1.7  * Feb '03 UCSC contig  Rshll-C/D  7  10.6  —  Not i n M G I  D7MU79  7  22.8  16.0  Ccnel-A/B  7  33.5  16.0  Gnefr-C/D  7  40.5  —  D7MU62  7  78.5  42.6  D10MU180  10  118.0  64.0  D10MU164  10  67.5  Not mapped i n U C S C  D17MU10  17  — —  24.5  Not mapped i n U C S C  D19MU68  19  3.4  6.0  Capnl-A/B  19  5.8  3.0  Not i n M G I  * 1 c M on average relates to 1.8 M b . I obtained this value by finding each chromosome's length (except the Y chromosome) i n megabases (Mb) from the Ensemble Genome browser (http://www.ensembl.org/Mus_musculus/index.html), as well as the last mapped known gene on each chromosome. I then went to M G I to get the c M position for that gene. If there was not a value, which was the case for some chromosomes, I went to the next to last known gene and so forth until there was a gene that had a c M position in M G I . Generally, the known genes were 2 M b within the end o f the chromosome. I then divided the length o f the chromosome (in M b ) by the c M position o f the last known gene (with a c M position) for each chromosome and then added the 20 values up and divided that number by 20 to get the average ratio between M b and c M , which was calculated to be 1 c M : 1 . 8 M b .  31  Chapter 3: F2 exencephaly frequencies Introduction In the original studies o f S E L H / B c , a panel o f F2 exencephaly segregants from a cross between S E L H / B c to a normal non-related strain, L M / B c , was collected as part o f a study to determine the number o f genes that act together to cause exencephaly i n S E L H / B c and to map them. This original F2 exencephaly panel ( " T G Panel") was collected by Teresa G u n n (Gunn, 1995 (thesis); Juriloff et al., 2001) i n 1994 when S E L H / B c was not yet a fully inbred strain. The rationale for collecting a second F2 exencephaly panel was partly to repeat the experiment, now from a fully inbred S E L H / B c strain, partly to refine the mapping o f the Exen loci with a larger panel, and partly to serve as a control i n diet studies. This new study ("EX-5001") was collected from mice fed Purina Laboratory Rodent Diet #5001 ( " P L R D #5001") by Diana Juriloff and M u r i e l Harris in 2003. Previous studies have showed that the exencephaly frequency i n the S E L H / B c strain is affected by maternal diet (Harris and Juriloff, 2005). This observation has been consistent over the years i n numerous studies since it was first observed i n 1995. O n P L R D #5001, the diet the mice are normally fed, the exencephaly frequency is approximately 5-10%. O n Purina Mouse Diet #5015 ( " P M D #5015") the exencephaly frequency is approximately 20-30%. It was unknown whether the diet effect is a direct effect on embryos or requires mediation by S E L H / B c maternal genotype. Therefore, a second new F 2 exencephaly study ("JHEX-5015") was collected on P M D #5015 to test the diet effect i n segregants i n a n o n - S E L H / B c mother.  32  Previous studies i n the S E L H / B c strain show that there is an excess o f females among exencephalic embryos. For example, i n one study o f S E L H / B c , 66% o f exencephalic embryos were female (Juriloff et al., 1989). The sexes o f the T G F2 exencephalic embryos are not known. Therefore, whether there is an excess o f females among the F 2 exencephalic embryos from the EX-5001 and J H E X - 5 0 1 5 studies was examined.  Materials and Methods Mouse maintenance A l l mice originated from and were maintained i n our animal unit i n the Department o f M e d i c a l Genetics at the University o f British Columbia under standard conditions previously described (See General Materials and Methods). A l l breeding colonies that produced adult mice used i n this study were maintained on Purina Laboratory Rodent Diet #5001 ( " P L R D #5001").  Breeding design Reciprocal crosses between S E L H / B c and L M / B c were made to create the F l generation. To generate a cohort o f F2 embryos, nulliparous F l females (age 2-4 months) were bred with F l male sibs. For both studies, each male received 2 females at a time (occasionally 1 or 3) and some males received a second set o f females. Overall, the E X 5001 study derived from 143 females and 73 males and the J H E X - 5 0 1 5 study derived from 55 females and 27 males. Breeding for the EX-5001 study aimed to obtain 50 exencephalic embryos. Breeding for the J H E X - 5 0 1 5 study aimed to obtain about 30  33  exencephalic embryos. These sample sizes were expected to show a highly significant difference (p< 0.01) i n exencephaly frequency i f P M D #5015 caused at least a doubled rate o f exencephaly. A panel o f F2 exencephalic embryos from the EX-5001 study was collected by saving tissue from all exencephalic embryos from all litters. A panel o f F2 exencephalic embryos from the J H E X - 5 0 1 5 study was collected by saving tissue from all exencephalic embryos from all litters. A panel o f 36 normal control F2 embryos from the EX-5001 study was collected by saving tissue from all embryos from 3 random, consecutive litters. A panel o f 39 control F2 embryos from the J H E X - 5 0 1 5 study was collected by saving tissue from all embryos from 3 random litters.  Scoring exencephaly Pregnancies were judged by eye and palpation. Developmental stage o f embryos was judged by morphological characteristics. Females were euthanized by carbon dioxide on day 14 o f gestation (range E 1 0 (1 litter) - E l 6 (1 litter)) for examination o f embryos. The uterus was removed, pinned to a black wax substrate, immersed i n physiological saline (0.85% N a C l ) solution, and cut open to reveal the conceptuses. Dead post implantation embryos ("moles") were recorded. The morphology o f scoreable embryos was examined under a dissection microscope and recorded. The following defects were screened: exencephaly, spina bifida, rachischisis, major abnormalities o f the limbs and tail, gastroschisis, cleft lip, and severe edema. The F2 exencephalic embryos were  34  individually washed i n ice-cold sterile saline solution and stored i n cryovials at -20°C for later preparation o f D N A .  Test of Diet Effect For the EX-5001 study, the F l female mice were fed P L R D #5001 from weaning until autopsy. For the J H E X - 5 0 1 5 study, the F l females were placed on P M D #5015 2-4 weeks before introduction o f males and maintained on P M D #5015 until autopsy. Conception was always within a week o f introduction o f males.  Cytoplasmic and X-linked effects After collecting the J H E X - 5 0 1 5 study, we realized we could look for "grandmother effects" (cytoplasmic and X - l i n k e d effects) on the risk o f exencephaly in the F2 embryos by subdividing the data by strain o f grandmother (See Fig. 3.1). In addition, we tested for a diet effect on exencephaly frequencies within the subdivided data. W e also tested whether there was a "grandmother effect" on gender o f exencephalic embryos. This would reflect an X - l i n k e d exencephaly risk factor i n S E L H / B c or L M / B c , i f any (See Fig. 3.1). W i t h i n the grandmother groups, a diet effect on the sex o f the embryos was also tested for. Following along with this type o f analysis, we tested whether there was a "grandmother effect" on post-implantation mortality rates within each panel and tested for a diet effect on post-implantation mortality rates subdivided by grandmother strain.  35  Figure 3.1: Conceptual diagram illustrating the genotype distributions for any particular locus on the sex chromosomes i n F 2 segregants from reciprocal grandmother crosses. The difference i n X chromosome genotypes o f the females i n the F 2 generation (circled) from reciprocal grandmothers would result i n an impact on exencephaly i f there were an X-linked recessive factor increasing risk o f exencephaly.  LM/Bc cytoplasm  LM/Bc 9 x S E L H / B c rv  F xF 1  X X L  L  xX Y  X X  S  xX Y  L  1  : X X S  S  L  L  : X Y : X Y ! L  S  SELH/Bc cytoplasm  SELH/Bc 9 x LM/Bc r ? X X S  F  1  x F  X X S  1  X X S  S  L  xX Y  L  : X X L  xX Y  S  S  S  : X Y :X Y S  L  36  Sex genotyping For the E X - 5 0 0 1 exencephaly panel, D N A had been prepared from stored frozen embryos by standard phenol chloroform extraction and banked (See General Materials and Methods). For the J H E X - 5 0 1 5 F2 exencephaly panel and both F 2 normal control panels, D N A was prepared from stored frozen embryos by using the QIAamp D N A M i n i kit (See General Materials and Methods). Sex o f embryos was identified by P C R using the SmcX-1 and SmcY-1 primers (Personal Communication from D a v i d Threadgill). The primer sequences were obtained from the Smc gene and they can be found i n Appendix I. The products were resolved electrophoretically on 4% NuSieve 3:1 Agarose gels, stained by ethidium bromide, and photographed over U V light.  Statistical Methods To compare groups i n various experiments the chi-square test o f independence was used (http://www.georgetown.edu/faculty/ballc/webtools/web_chi.html). To compare observed sex ratios against a 1:1 expected F : M sex ratio a  goodness-of-fit test was  used. The p-value 0.05 was chosen for statistical significance.  Results Diet effect on exencephaly The original T G study contained 3.7% exencephaly (43 i n 1151 embryos) and the new EX-5001 study contained 2.8% exencephaly (52 i n 1873 embryos) (See Table 3.1). These frequencies were not significantly different from each other (% = 2.15;p> 0.5). In 2  the J H E X - 5 0 1 5 study, the frequency o f exencephaly was 4.4% (34 i n 773 embryos),  37  Table 3.1: Frequency o f exencephaly and post-implantation mortality rates ("moles") in F 2 generations from crosses between S E L H / B c and L M / B c i n all F2 exencephaly studies.  # litters  # implants  % moles  # embryos  # exens  % Exen  Sex ratio (F:M)  T G Panel  ?  1194  3.6  1151  43  3.7  ?:?  EX-5001  143  1971  5.0  1873  52  2.8  39:13  JHEX-5015  45  818  5.5  773  34  4.4  24:10  significantly higher than the 2.8% observed in the EX-5001 study tf= 4.58; p< 0.05), indicating that P M D #5015 caused an increased risk of exencephaly in F2 embryos, a 1.6 X increase. The frequency of exencephaly (4.4%) from the JHEX-5015 study was not significantly different from the T G study (3.7%) (^= 0.53; p> 0.5).  Other defects observed in the F2 embryos Among the 1873 embryos in the EX-5001 study, there were few defects other than exencephaly observed. There were two embryos with "odd neural tube defects," but they were not included in this F2 exencephaly panel. One embryo had "overgrown" brain tissue, splayed to either side, coming out of a hole just rostral to the cerebellum. The other had a tiny amount of exencephaly behind the Closure 2 site, slightly offset to the left. No rachischisis or spina bifida neural tube defects were observed. One embryo was observed with sirenomelia (no tail) and gastroschisis and one embryo was observed with no lower jaw. Among the 773 embryos in the JHEX-5015 study, no defects other than exencephaly were observed.  Sex effect on exencephaly In the EX-5001 panel, 75% (39:13) of exencephalic embryos were female. In the JHEX-5015 panel, 71% (24:10) of the exencephalic females were female (See Table 3.1 on pg. 38). Both F2 exencephaly panels showed an excess of females with a significant deviation from a 1:1 sex ratio ( x = 13.0;p< 0.001, x - 5.8;p< 0.025). The exencephalic 2  2  sex ratio was not significantly different between the two panels ( x = 0.20; p> 0.5), 2  showing that there was no interaction between a diet effect and gender effect on risk of  39  exencephaly. The normal control sex ratios ( F : M ) for the E X - 5 0 0 1 and J H E X - 5 0 1 5 panels were 23:13 and 15:24. Neither control panel showed a significant deviation from a 1:1 sex ratio ( x = 2.78 and 2.08, respectively; p> 0.1). 2  Cytoplasmic and X-linked effects on exencephaly The F2 exencephaly panels were subdivided by strain o f grandmother to test for cytoplasmic and X - l i n k e d effects (See Table 3.2). In the EX-5001 study, the exencephaly frequency was significantly higher i n the F2 embryos from L M / B c grandmothers (3.5%) than from S E L H / B c grandmothers (1.8%) ( x = 4.58; p< 0.05). In the J H E X - 5 0 1 5 study, 2  this "grandmother effect" followed the same trend with a higher frequency o f exencephaly i n F 2 embryos from L M / B c grandmothers (4.8%) than from S E L H / B c grandmothers (3.5%) but it was not statistically significant ( x = 0.56; p> 0.5). 2  The pattern suggesting higher exencephaly frequencies i n F 2 embryos with L M / B c cytoplasm could alternatively be caused by an X - l i n k e d exencephaly liability gene from L M / B c (See F i g . 3.1 for concept on pg. 36). X-linkage is distinguishable from cytoplasmic effects because it would alter the sex ratio among F 2 exencephalic embryos. A l l F2 males from either reciprocal cross would have the same m i x o f genotypes. In F2 exencephalic female embryos, theoretically half o f them would be genetically the same in both reciprocal crosses, whereas the other half would be genetically different i n X-linked genes between the crosses. In the EX-5001 panel, the exencephalic sex ratio ( F : M ) was 11:4 (2.75:1) within the S E L H / B c grandmother group and 28:9 (3.1:1) within the L M / B c grandmother group, fairly similar and not significantly different from each other ( £ = 0.03;p> 0.975) (See Table 3.2). In the J H E X - 5 0 1 5 panel, the exencephalic sex ratio was  40  T P ^ P ^ ™ y « * * ("moles") i n F2 generations from crosses between SELH/Bc and LM/Bc. The panels have been subdivided by strain o f grandmother within each study on different diets. Ind  T^Mm^ThTnf^^  Study  eq  CieS  1  Grandmother # litters  m o r t a l i t  #  #  implants  % moles  embryos  # exens  % Exen  855  5.03  812  15  1.8  Sex ratio (F:M) * ••»•/  EX-5001  SELH/Bc  EX-5001  LM/Bc  81  1116  4.93  1061  37  3.5  28:9  SELH/Bc  17  245  7.76  226  8  3.5  7:1  LM/Bc  28  573  4.54  547  26  4.8  17:9  JHEX5015 JHEX5015  62  11:4  7:1 within the SELH/Bc grandmother group and 17:9 (1.89:1) within the L M / B c grandmother group, again not significantly different from each other (5^= 1.44; p> 0.1). The results indicate that the "grandmother effect" is most likely due to L M / B c cytoplasm, not to X-linked exencephaly genes.  Diet effect on exencephaly frequency within "grandmother groups " Following along with this analysis, we looked for a diet effect on exencephaly frequency within the subdivided SELH/Bc and L M / B c grandmother groups. The response to P M D #5015 appeared to be about the same in each grandmother group; it increased the exencephaly rate in F2 embryos from L M / B c grandmothers but it was not statistically significant (See Table 3.3). B y subdividing the data in half we lost statistical power to pick up a diet effect, i f any, within grandmother groups. We also tested for a diet effect on the gender of the F2 exencephalic embryos subdivided by strain of grandmother. We found no evidence of a diet effect on the gender of the F2 exencephalic embryos as the excess of females stays comparably consistent within these subcategories.  Post-implantation mortality The post-implantation mortality rate in the JHEX-5015 study, 5.5%, was similar to the 5.0% observed in the EX-5001 study {£= 0.30;p> 0.5). This shows that there was no effect of diet on post-implantation mortality rates. Neither post-implantation mortality rate from the two new studies, EX-5001 and JHEX-5015, was significantly different from the T G study ( x = 2.16 and 0.53, respectively). 2  42  Table 3.3: Exencephaly frequencies within the "grandmother groups," subdivided by diet.  SELH/Bc Grandmothers  LM/Bc Grandmothers  P L R D #5001  1.8%  3.5%  PMD #5015  3.5%  4.8%  Diet  43  Whether there was any "grandmother effect" on post-implantation mortality rates ("moles") was explored (Table 3.2). In the EX-5001 study, the post-implantation mortality rates from S E L H / B c (5.03%) and L M / B c (4.93%) grandmothers were essentially the same (% = 0.01;/?= .995). In the J H E X - 5 0 1 5 study, 7.76% o f total 2  implants were "moles" within S E L H / B c grandmothers and 4.54% o f total implants were "moles" within L M / B c grandmothers (%= 3.02; p> 0.1), not significantly different from 2  each other. A diet effect on post-implantation mortality rates within grandmother groups was checked for. From S E L H / B c grandmothers, the post-implantation mortality rates from the two diet panels were not significantly different from each other (%= 2.34; p> 0.1). A 2  similar result was seen from L M / B c grandmothers across the two diets (% = 0.11; p> 0.9). 2  This shows that the post-implantation mortality rates are not affected by strain o f grandmother, diet, or an interaction between the two.  Discussion The first major finding i n this study was the diet effect i n the F 2 embryos. Given that the exencephaly rate doubles i n the S E L H / B c strain on P M D #5015, we were looking for a similar increased rate o f exencephaly in the F 2 embryos that would double the 2.8% rate o f exencephaly observed i n the EX-5001 study. The exencephaly frequency from the J H E X - 5 0 1 5 study was significantly higher than the E X - 5 0 0 1 study, a 1.6 X increase. First and foremost, this suggests that the diet effect observed i n the F2 studies does not require a S E L H / B c mother, a 100% S E L H / B c genetic background. This does not rule out though that the diet effect is mediated by certain maternal S E L H / B c  44  dominant alleles. The question that arises is whether only maternal factors account for the diet effect, only embryonic genetic factors account for the diet effect, or whether both account for the diet effect. To account for the diet effect, the first hypothesis involves a gene-diet interaction, specifically with one o f the Exen loci. The exencephaly risk o f S E L H / B c has been found to be due to a combination o f about three genes acting together additively and they are named Exenl, Exen2, and Exen3 (Juriloff et al., 2001). The hypothesis is that one or more of the Exen loci specifically interact with P M D #5015 to produce the increased exencephaly frequency. In the F 2 segregants, i f an embryo is homozygous or heterozygous for the S E L H / B c Exen alleles that interact with P M D #5015, they would have higher risk o f becoming exencephalic than they would i f the mice were fed P L R D #5001, so that more F 2 embryos with that particular genotype on P M D #5015 would become exencephalic that would maybe not on P L R D #5001. Typing the D N A from the individual F2 exencephalic embryos i n the EX-5001 and J H E X - 5 0 1 5 studies to compare genotypic ratios at various genetic markers across the Exen loci would test whether there is a gene-diet interaction. Another explanation to account for the increase in exencephaly frequency observed between the E X - 5 0 0 1 and J H E X - 5 0 1 5 studies involves general maternal effects. Firstly, we could be dealing with a general maternal effect that affects many strains o f mice. One noticeable observation is that the mice become fatter faster when fed P M D #5015 compared to when they are fed P L R D #5001. M a n y mice might respond the same when eating P M D #5015. S E L H / B c mice also become noticeably fatter on P M D  45  #5015. This is interesting because neural tube defects have been associated with obesity (Anderson et al., 2005). Given that nonsyndromic exencephaly is considered to be a multifactorial threshold trait involving various combinations of genes acting together, along with environmental factors, to create varying levels of risk of developing exencephaly, the multifactorial threshold model can be applied here to help explain the F2 response to maternal diet, specifically P M D #5015. P M D #5015 would be considered an environmental factor that contributes to exencephaly, increasing the liability to developing it. This would then influence the position of the distribution relative to the threshold, so that the "F2 population" distribution shifts towards the right so that more individuals cross the threshold and are affected. Whether this gene-diet interaction or diet effect in general is independent of maternal SELH/Bc genotype remains to be resolved. Given that the F l dams are part SELH/Bc, we could be seeing part of a maternal SELH/Bc diet effect, a semidominant diet effect per se. The mothers have one Exen allele leaving the possibility of a maternal effect of a semidominant Exen allele from SELH/Bc, as well as a direct effect in embryos with the genotype that interacts with P M D #5015. The second major finding in this study was the excess of females among the F2 exencephalic embryos in the EX-5001 and JHEX-5001 panels. In the EX-5001 and JHEX-5015 panels, 75% (39:13) and 71% (24:10) of exencephalic embryos were female, respectively. This is in agreement with previous SELH/Bc studies. Females express exencephaly more easily than males and it is unknown why this occurs. This trend is also observed in human cases of anencephaly and in other mouse mutants such as curly tail  46  (Brook et al., 1994). Some authors have attributed the female predominance i n neural tube defects to a differential embryonic growth rate but this difference was not seen i n other studies (Rittler et al., 2004). Brook et al. reported that the difference seemed more likely that male and female embryos differ in some specific aspect o f the neurulation process that increases the susceptibility o f females to exencephaly (Brook et al., 1994). W i t h the removal o f the F 2 exencephalic embryos (mostly female) from the samples, one might expect to see slightly skewed sex ratios i n the normal F2 panels showing more males than females. When the total sample size is considered, though, removing 52 or 34 F 2 exencephalic embryos (from the EX-5001 and J H E X - 5 0 1 5 studies, respectively) makes such a small difference to the sample that we do not expect to be able to detect an excess o f males i n normal embryos. The third major finding i n this study was that the L M / B c cytoplasm appeared to facilitate the expression o f S E L H / B c genes that contribute to exencephaly. A s previously described, we analyzed "grandmother effects" within the F 2 exencephalic embryos for both new panels. In the E X - 5 0 0 1 study there was a significantly higher percentage o f exencephaly from L M / B c grandmothers compared to S E L H / B c grandmothers. Although this L M / B c "grandmother effect" was not statistically significant i n the J H E X - 5 0 1 5 study, it followed the same trend. This effect is attributed to the L M / B c cytoplasm and not X-linkage because the sex ratio o f the F2 exencephalic embryos does not significantly change between grandmother groups (See Fig. 3.1 on pg. 36 and Table 3.2 on pg. 41). This L M / B c cytoplasmic effect is an unexpected finding as it possibly suggests an epigenetic contribution to the risk o f exencephaly.  47  C h a p t e r 4: F2 genotypes Introduction  Previous studies have found the genetic exencephaly risk of SELH/Bc to be due to a combination of about three genes, Exenl, Exen2, and Exen3, acting additively (Juriloff et al., 2001). The Exenl and Exen2 loci were mapped to regions on Chr 13 and 5, respectively, and the Exen3 locus provisionally to Chr 11. One strategy employed to determine these locations was the examination of genotypes from F2 exencephalic embryos between crosses of SELH/Bc to a normal non-related strain, L M / B c . This original F2 exencephaly panel ("TG Panel") was collected in 1994 by Teresa Gunn. The rationale behind using this panel to help map the Exen loci was that F2 exencephalic embryos would have more Exen alleles from SELH/Bc across a combination of the those loci. The risk for an F2 embryo to develop exencephaly would increase with increasing number of SELH/Bc alleles across loci, demonstrating the additivity of the Exen loci in creating risk of this trait. Given that the F2 exencephalic embryos came from a segregating generation, they would have different allele combinations across the exencephaly-risk loci and the risk for exencephaly would not be associated with only one particular allele combination (Juriloff et al., 2001). As the SELH/Bc alleles at each locus were demonstrated to be semidominant, the risk of exencephaly would not only be associated with a SELH/Bc homozygous genotype, but also to a lesser extent with heterozygosity (Juriloff et al., 2001). The rationale for collecting a new F2 exencephaly panel ("EX-5001") was to repeat the experiment with a fully inbred SELH/Bc strain, to improve the mapping of Exen loci with a larger panel, and to serve as a control in diet studies. This panel was  48  collected from mice fed Purina Laboratory Rodent Diet #5001 ("PLRD #5001"), the diet the mice are normally fed, by Diana Juriloff and Muriel Harris in 2003. A second new F2 exencephaly panel ("JHEX-5015") was collected from mice fed Purina Mouse Diet #5015 ("PMD #5015") to test for a diet effect that is observed in the SELH/Bc strain. A 1.6 X significant increase in exencephaly frequency was observed in the JHEX-5015 study (See Chapter 3). Anticipating that a diet effect would be observed in the F2 embryos, part of the rationale for creating the JHEX-5015 panel was to test the hypothesis that specific Exen loci interact with P M D #5015. This involved typing individual F2 exencephalic embryos from the EX-5001 and JHEX-5015 panels with genetic markers for genotype at the Exenl, Exen2, and Exen3 loci on Chr 13, 5, and 11, respectively, and also for markers on additional chromosomes of interest. Given that the diet was the same for the EX-5001 and T G studies, we expected the EX-5001 F2 exencephalic embryos' genotype distributions across the Exen loci to be comparable to those of the T G Panel, confirming the contribution of risk of exencephaly from SELH/Bc alleles across these loci. By comparing the JHEX-5015 F2 exencephalic embryos' genotypes at specific Exen loci to those in the EX-5001 panel, we hoped to see a segregation ratio that was significantly different at an Exen locus between the two new panels suggesting a gene-diet interaction. For example, i f the role of one Exen locus weakened or disappeared on P M D #5015 in the F2 embryos, we would expect to see a random Mendelian genotypic ratio, whereas on P L R D #5001, the segregation ratio would significantly deviate from random showing an excess of SELH/Bc alleles at the same locus. This would suggest that this Exen locus became irrelevant, contributing little or no risk of exencephaly on P M D #5015.  49  The alternative hypothesis that there is no Exen gene-diet interaction was concurrently tested. F o r example, i f P M D #5015 did not interact specifically with one particular Exen gene, we would expect the genotype ratios and allele combinations across the liability loci to be similar between the EX-5001 and J H E X - 5 0 1 5 panels. This would suggest that the diet simply added to the genetic effect. Test crossing o f segregants is a powerful way to estimate the number o f loci contributing to a trait between inbred strains. Teresa Gunn's study indicated the number o f major loci to contribute to the risk o f exencephaly i n S E L H / B c to be about three loci. The roles o f the Exenl, Exen2, and Exen3 loci seem to be w e l l established. However, the location o f the Exen3 locus and possibly another locus are not w e l l established. The previous F2 sire screen pointed towards regions on Chr 7, 10, and 17 possibly containing a locus contributing to the risk o f exencephaly, but the significance was not sustained in the T G F2 exencephaly panel (Juriloff et al., 2001). For this reason they were revisited in the EX-5001 and J H E X - 5 0 1 5 studies, along with a new region o f interest from Chr 19 that contains a M u s D transposon insertion i n the S E L H / B c strain (Personal communication from D r . D i x i e Mager). Therefore, the genotypes o f the F2 exencephalic embryos from the EX-5001 and JHEX-5015 studies were compared at various genetic markers across the Exen loci and additionally, on other chromosomes o f interest to test the hypotheses stated above.  50  Materials and Methods  Experimental strategy The first part o f this study involved confirming the locations o f Exenl (Chr 13), Exen2 (Chr 5), and Exen3 (Chr 11). This involved comparing the segregation ratios from the EX-5001 F 2 exencephaly panel against the random genotype distribution o f 13 SS:26 SL:13 L L at these loci (52 F 2 exencephalic embryos). In addition, the segregation ratios at putative loci on C h r 7, 10, 17, and 19 were compared against the random distribution segregation ratio to test for a locus that contributed to the risk o f exencephaly, i f any. The second part o f this study involved comparing the segregation ratios from the J H E X - 5 0 1 5 F2 exencephaly panel against the random genotype distribution o f 8.5 SS:17 S L : 8.5 L L to test for contribution to exencephaly (34 F2 exencephalic embryos). The segregation ratios from the putative loci on Chr 7, 10, 17, and 19 were also compared against random distribution. This set up the framework so that the segregation ratios from the EX-5001 and J H E X - 5 0 1 5 F2 exencephaly panels could be compared to each other to test what effect, i f any, P M D #5015 had on the F2 exencephalic embryos' genotypic ratios. The last part o f this study involved comparing the segregation ratios from the two F2 exencephaly panels at the Exen and putative loci to test hypotheses for the increase i n exencephaly frequency observed i n the J H E X - 5 0 1 5 study; the hypotheses being either an interaction between the diet and genotype or the diet having an additive effect to the genotype already present. The patterns o f the segregation ratios would be interpreted to infer the possible mechanism o f interaction present between the diet and gene(s).  51  Mice, breeding design, scoring exencephaly The descriptions o f inbred strains used and their standard conditions, along with the breeding design and scoring exencephaly have been previously described i n General Materials and Methods.  DNA preparation D N A from the EX-5001 F 2 exencephalic panel ("EX-5001") had previously been prepared from stored frozen embryos by standard phenol chloroform extractions and banked (See General Materials and Methods). D N A from the J H E X - 5 0 1 5 F 2 exencephalic panel ("JHEX-5015") and the normal control panels was prepared from stored frozen embryos by using the QIAamp D N A M i n i kit (See General Materials and Methods).  Rationale for choice of markers used For the E X - 5 0 0 1 study, the markers used on Chr 13, 5, and 11 were chosen because the Exen loci were originally mapped there. For the J H E X - 5 0 1 5 study, the Chr 13 markers were chosen based on the best markers from the EX-5001 study that gave highly significantly skewed segregation ratios and that were easy to visualize on a gel. Fgfr4-E/F, Ntrk2-C/D, and Ptchl-A/B, spaced evenly apart, were initially chosen because they represented the area that had the highest probability o f Exenl being located there. Further on i n this study, additional markers, FgdS-C/D, D13MU193, D13MU30, and D13MU76, were typed on the J H E X - 5 0 1 5 F2 exencephalic embryos for analysis. For the  52  Chr 5 and 11 markers chosen, the best markers from the EX-5001 study were used to type the J H E X - 5 0 1 5 panel. The Chr 7 markers used for the EX-5001 and J H E X - 5 0 1 5 studies were chosen to revisit an old result observed i n the previous F2 sire screen (Juriloff et al., 2001). In that study, D7MU75 appeared to approach significance. The C h r 10 and 17 markers used for both EX-5001 and J H E X - 5 0 1 5 studies were also chosen to revisit regions significant in the F2 sire screen (Juriloff et al., 2001). The Chr 19 marker used i n both studies, D19MU68, was chosen because our collaborator, D r . D i x i e Mager, mapped a "master" M u s D element i n S E L H / B c to Chr 19 at 5.7 M b ( U C S C online genome browser) (Personal communication from D r . D i x i e Mager). D19MU68 is approximately located at 3.5 M b from the centromere, 2.2 M b (1-2 c M ) proximal to the M u s D element. The hypothesis for looking at this marker was that the M u s D element contributes to the risk o f exencephaly and causes genetic instability i n S E L H / B c . For the E X - 5 0 0 1 panel, another Chr 19 marker, Capnl-A/B at 5.8 M b , was used to confirm the results o f D19MU68.  PCR and electrophoresis The genotypes o f the F 2 exencephalic embryos i n both new panels were identified by P C R . The products were resolved by electrophoresis on 4% NuSieve 3:1 Agarose gels, stained by ethidium bromide, and photographed over U V light (See General Materials and Methods).  53  Statistical Methods To compare segregation ratios at genetic markers between the E X - 5 0 0 1 and JHEX-5015 studies the chi-square test o f independence was used (http://www.georgetown.edu/faculty/ballc/webtools/web_chi.html). T o compare observed segregation ratios against a 1:2:1 expected S S : S L : L L segregation ratio a chi-square goodness-of-fit test was used. The/?-value 0.05 was chosen for statistical significance. Exactp-values were obtained using a Hewlett-Packard 1 I C calculator. A technique developed by R. A . Fisher was used to combine probabilities o f independent chi-square tests that suggested but did not establish statistical significance. This technique combined the probabilities to create an overall test for significance (Sokal andRohlf, 1995).  Results The EX-5001 study The locations o f Exenl and Exen2 were further supported by the genotypes o f the F2 exencephalic embryos i n the EX-5001 study. A l l the C h r 13 and C h r 5 markers used in this study gave segregation ratios that significantly deviated from the random distribution o f 13 SS:26 SL:13 L L , showing an excess o f S E L H / B c alleles across these loci. The x values for genotypic ratios ranged from 9.9 - 27.0 for markers on m i d Chr 13 2  (See Table 4.1). The C h r 13 markers spanned 59.6 M b (30 c M ) , from Fgd3-C/D (47.8 M b ) to D13MU76 (107.4 M b ) (See Fig. 5.2 on pg. 79 for map o f Chr 13). The most significant Chr 13 markers, Fgfr4-E/F, D13MU13, and Gprk6-C/D, gave segregation ratios o f 29:14:9 (x ^ 27.0; p< .001) (See Table 4.1) showing an excess o f SS 2  54  Table 4.1: Genotype summaries o f F2 exencephalic embryos and their corresponding values at various genetic markers i n the EX-5001 study.  Marker  #SS 28 29 29 29 28 28 21 21 22  #SL 15 14 14 14 16 16 26 27 24  #LL 9 9 9 9 8 8 5 4 6  X value 23.19 26.46 26.46 26.46 23.08 23.08 9.84 11.19 10.15  P-value 0.00001 0.000002 0.000002 0.000002 0.00001 0.00001 0.007 0.004 0.006  D5MH95 D5Mit30 Gats-C/D  29 28 28  20 21 20  3 3 4  28.76 25.96 24.92  0.000001 0.000002 0.000004  D11Mit14 D11Mit10 Scn4a-C/D  15 16 17  30 28 27  7 8 8  3.70 2.76 3.19  0.157 0.252 0.203  D10Mit164 D10Mit180  12 11  26 28  14 13  0.16 0.46  0.923 0.795  D17MH10  13  21  18  2.88  0.237  D7MH75 Rshl1-C/D D7MH79  22 20 18  20 22 27  10 10 7  8.30 5.08 4.73  0.016 0.079 0.094  D19MH68 Capn1-A/B  3 4  24 23  25 25  18.92 17.66  0.00008 0.0001  Fdg3-C/D Fgfr4-E/F Gprk6-C/D D13MM3 Fancc-C/D Ptd-A/B D13MH193 D13MH30 D13MH76  2  homozygotes. The ratio o f S S : S L was approximately 2:1 from markers Fdg3-C/D Ptchl-A/B, supporting previous data showing that SS and S L were not equivalent in effect on exencephaly and that Exenl appears to be semidominant (Juriloff et al., 2001). Interestingly, the 2:1 S S : S L ratio disappeared at D13MU193 (21:26:5) as the number o f heterozygotes at this marker increased to the expected number and only the homozygous genotypes were significantly deviated from expected (See Table 4.1 on pg. 55). The lowest chi-square value (x =9.9; p< 0.01) from D13MU193 was still highly significant 2  showing considerable deviation from randomness. A t D13MU13, the normal control panel gave a segregation ratio o f 7:17:12, not straying far from Mendelian segregation (X = 1.50; p> 0.1) (See Table 4.2). There appears to be a slight under-representation o f SS 2  homozygotes, but we w o u l d expect this as a number o f SS homozygotes at the Exenl locus become exencephalic. A s expected, the F2 exencephalic panel and normal panel were significantly different from each other at D13MU13 ( x  2 =  11.64; p< 0.01).  The genotype distributions from the distal Chr 5 markers significantly deviated from randomness i n the E X - 5 0 0 1 F 2 exencephaly panel providing strong evidence for the location o f the Exen2 locus. The three markers spanned 9.4 M b , from D5MU95 (122.4 M b ) to Gats-C/D (131.8 M b ) . D5MU95 gave a segregation ratio o f 29:20:3 ( S S : S L : L L ) ( X ^ 29.0;p< 0.001), and D5MU30 and Gats-C/D had segregation ratios o f 28:21:3 and 2  28:20:4, respectively ( x = 26.0, 25.0; p < 0.001) (See Table 4.1 on pg. 55) (See F i g . 5.2 2  on pg. 79 for map o f C h r 5). Here, at the Exen2 locus, the significance is coming from the excess o f SS homozygotes and the deficiency o f L L homozygotes. In the normal control panel, D5MU95 gave no significant deviation from random segregation, the ratio being 7:17:12 ( x = 1.50; p> 0.1) (See Table 4.2). Expectedly, at D5MU95, the F 2 exencephaly 2  56  Table 4.2: Genotype summaries o f control normal F 2 embryos and their corresponding x values at selected genetic markers i n the EX-5001 study.  2  Marker D13MU13 D5MU95 D7MU75 DllMitlO D19MU68  S S : S L : L L ratio 7:17:12 7:17:12 3:22:11 9:18:9 12:18:6  X value, 1.50, p> 1.50. p> 5.33, p> 0.00, p= 2.00, p> 2  p-value 0.1 0.1 0.05 1.0 0.1  57  panel and the F 2 normal control panel gave significantly different segregation ratios ( x = 2  16.73;/>< 0.001). Whether the ExenS locus on Chr 11 contributed to the risk o f exencephaly was not resolved by this study. The markers spanned 7.6 M b (7 c M ) from D11MU14 at 98.4 M b to Scn4a-C/D at 106.0 M b . In the EX-5001 F2 exencephaly panel, D11MU14,  DUMitlO,  and Scn4a-C/D gave segregation ratios o f 15:30:7, 16:28:8, and 17:27:8, respectively ( x  2  = 3.7, 2.8, 3.1) (See Table 4.1 on pg. 55). Although not statistically significant (bestp< 0.16), these segregation ratios suggested a small effect coming from the Exen3 locus as there is an excess o f S E L H / B c alleles at this locus (e.g. D11MU14 - 60 S:44 L ) . If any, the putative effect o f Chr 11 was too weak to be statistically significant i n this sample size. A t DUMitlO,  the F 2 normal control panel gave a segregation ratio o f 9:18:9  showing complete Mendelian segregation ( x = 0.00) (See Table 4.2 on pg. 57). The F2 2  exencephalic panel and normal control panel were not significantly different from each other 0r= 1.33;p> 0.5). Surprisingly, proximal Chr 7 marked by D7MU7'5 appeared to be involved i n the risk o f exencephaly (See Table 4.1 on pg. 55). In the EX-5001 exencephaly panel, D7MU75 gave a segregation ratio o f 22:20:10 showing an excess o f SS homozygotes at this locus ( x = 8.3;p< 0.025). The two other Chr 7 markers, Rshll-C/D and D7MU79, 2  gave segregation ratios o f 20:22:10 and 18:27:7 ( x = 5.1 and 4.8) showing that i f there is 2  an exencephaly-risk locus on Chr 7, it appears to be closer to D7MU75. C h r 7 markers spanned 16.9 M b (14 c M ) from D7MU75 at 5.9 M b to D7MU79 at 22.8 M b . In the normal control panel D7MU75 gave a segregation ratio o f 3:22:11 suggesting a deficiency o f SS homozygotes ( x  2 =  5.33; p> 0.05) (See Table 4.2 on pg. 57). The F 2 exencephalic panel  58  and normal control panel were significantly different from each other (3^= 12.07; p< 0.01). Chr 19 marked by D19MU68 and Capnl-A/B gave segregation ratios that significantly deviated from randomness i n the EX-5001 F 2 exencephaly panel. The markers spanned 2.4 M b from D19MU68 at 3.4 M b to Capnl-A/B at 5.8 M b . D19MU68 gave a segregation ratio o f 3:24:25 ( x = 19.0; p< 0.001) and Capnl-A/B gave a 2  segregation ratio o f 4:23:25 ( x - 17.6; p< 0.001) showing a severe deficiency o f SS 2  homozygotes and a strong excess o f L L homozygotes (See Table 4.1 on pg. 55). Surprisingly, this result suggests that this region on Chr 19 from L M / B c may contribute to the risk o f exencephaly i n F2 embryos. A t D19MU68, the normal control panel gave a segregation ratio o f 12:18:6  2.00,p> 0.1) that did not significantly deviate from  randomness and that had a trend opposite to the F2 exencephaly panel that showed a numerical excess o f SS homozygotes (See Table 4.2 on pg. 57). A s expected, the F2 exencephalic panel and normal control panel were significantly different from each other (X =15.51;/>< 0.001). 2  Little statistical support was present for any role o f genes on C h r 10 and Chr 17 marked by D10MU164, D10MU180, and DI 7MU10. They gave segregation ratios o f 12:26:14,11:28:13, and 13:21:18, respectively (x = 0.16, 0.46, and 2.88) (See Table 4.1 2  on pg. 55). A l l markers segregated randomly i n the F 2 exencephalic embryos and therefore showed no evidence o f involvement i n the risk o f exencephaly.  59  The JHEX-5015 study The expected random segregation ratio o f genotypes o f the 34 F 2 exencephalic embryos was 8.5:17:8.5 ( S S : S L : L L ) for all genetic markers. The seven Chr 13 markers spanned 59.6 M b (30 c M ) from Fgd3-C/D at 47.8 M b to D13MU76 at 107.4 M b . In the J H E X - 5 0 1 5 F2 exencephaly panel, Fgd3-C/D, the most proximal C h r 13 marker used, gave a segregation ratio o f 14:15:5 suggesting an excess o f SS homozygotes  5.24;  p< 0.089) (See Table 4.3). The next three Chr 13 markers, Fgfr4-E/F, Ntrk2-C/D, and Ptchl-A/B, gave segregation ratios o f 13:15:6  3.36;/?> 0.1). D13MU193 gave a  segregation ratio o f 14:14:6 (3^= 4.83;p< 0.09), D13MU30 gave a segregation o f 15:12:7 6.7; p< 0.05), and D13MU76 gave a segregation ratio o f 15:11:8 ( £ = 7.12;p< 0.05), all showing an excess o f SS homozygotes. The Chr 13 markers used i n this study that did not significantly deviate from random showed a trend towards an excess o f SS homozygotes and their lack o f statistical significance was most likely due to the relatively small sample size. D13MU30 and D13MU76 were the only C h r 13 markers that gave segregation ratios that significantly deviated from random. The F 2 normal control panel gave a random segregation ratio o f 7:20:12 at Ptchl-A/B ( x = \ .2>\\p> 0.5) (See Table 2  4.4). The segregation ratios o f the F2 exencephalic panel and normal control panel at Ptchl-A/B, the representative marker from this region, were not significantly different from each other ( x = 4.19; /?> 0.1). 2  Interestingly, the Exen2 locus on Chr 5 appeared to not be involved i n the risk o f exencephaly i n the J H E X - 5 0 1 5 F2 exencephaly panel when mice are fed P M D #5015. D5MU95 gave a segregation ratio o f 8:20:6 (3^= 1.30; p> 0.5) and D5MU30 and D5MU168 gave segregation ratios o f 9:18:7 and 10:16:8, respectively (x ^ 0.35;p> 0.5), 2  60  Table 4.3: Genotype summaries o f F2 exencephalic embryos and their corresponding x values at various genetic markers i n the J H E X - 5 0 1 5 study.  2  Marker Fgd3-C/D Fgfr4-E/F Ntrk2-C/D Ptd-AJB D13Mit193 D13Mit30 D13MH76  #SS 14 13 13 13 14 15 15  #SL 15 15 15 15 14 12 11  # LL 5 6 6 6 6 7 8  X value 5.24 3.36 3.36 3.36 4.83 6.70 7.12  P-value < 0.089 0.186 0.186 0.186 0.089 < 0.05 <0.05  D5MH95 D5MH30 D5MM68  8 9 10  20 18 16  6 7 8  1.30 0.35 0.35  0.522 0.839 0.839  D7Mit75 Rshl1-C/D D7MH79  12 10 10  18 19 18  4 5 6  3.88 1.94 1.06  0.144 0.379 0.589  D11Mit10 Scn4a-C/D  8 11  20 18  6 5  1.30 2.24  0.522 0.326  D10Mit180  5  20  9  2.00  0.368  D17Mit10  8  15  11  1.01  0.604  D19MH68  4  16  14  6.00  0.050  2  Table 4.4: Genotype summaries o f control normal F2 embryos and their corresponding x values at selected genetic markers i n the J H E X - 5 0 1 5 study.  2  Marker  Ptcl-A/B D5MU95 D7MU75 Scn4a-C/D D19MU68  S S : S L : L L ratio 7:20:12 7:15:17 5:23:11 13:21:5 6:15:18  X value 1.31, p> 0.5 2  7.21, p< 0.05 3.10, p> 0.1 3.51,p>0.1  9.46, p< 0.025  61  all showing random segregation o f L M / B c and S E L H / B c alleles at this locus (See Table 4.3). This suggested that the effect from P M D #5015 replaced the effect from the Exen2 alleles from S E L H / B c i n terms o f contribution to the risk o f exencephaly. Interestingly, the normal control panel gave a segregation ratio o f 7:15:17, showing an excess o f L L homozygotes (% 7.21; p< 0.05) (See Table 4.4). The F 2 exencephalic panel and normal 2=  control panel were almost significantly different from each other (3^= 5.73;p< 0.09). In the F2 exencephaly panel, C h r 11 markers, DllMitlO  and Scn4a-C/D, gave segregation  ratios o f 8:20:6 and 11:18:5, respectively (3^= 1.30 and 2.24;p> 0.1), not significantly different from random segregation (See Table 4.3 on pg. 61). In the F 2 normal control panel, Scn4a-C/D gave a segregation ratio o f 13:21:5, also not significantly different from random segregation (3^= 3.51; p> 0.1) (See Table 4.4 on pg. 61) and the segregation ratios from the F2 exencephalic panel and the normal control panel at Scn4a-C/D were similar to each other ( x = 0.06; p> 0.9). 2  The proximal C h r 7 markers, D7MU75, Rshll-C-D,  and D7MU79, gave  segregation ratios o f 12:18:4,10:19:5, and 10:18:6, respectively ( x = 3.88; 1.94, and 2  1.06; p> 0.1) in the J H E X - 5 0 1 5 exencephaly panel (See Table 4.3 on pg. 61). D7MU75 suggested an excess o f SS homozygotes at this "locus," (from here on to be referred as Exen4) providing support to the data from the EX-5001 study that there may be a gene on Chr 7 that contributes to the risk o f exencephaly i n the F2 embryos. In contrast, the normal control panel gave a segregation ratio o f 5:23:11 at D7MU75, suggesting a deficiency o f SS homozygotes (3^= 3.10; p> 0.1) (See Table 4.4 on pg. 61). The F 2 exencephalic and normal control panels were significantly different from each other at this locus ( x = 6.45; p< 0.05). 2  62  Chr 10 and 17 markers, D10MU180 and D17MU10, gave random segregation ratios suggesting no involvement i n the risk o f exencephaly i n the F 2 embryos. They were 5:20:9 and 8:15:11, respectively (% = 2.00;p> 0.1, tf= 1.01;p> 0.5) (See Table 4.3 on pg. 61). Chr 19 marker D19MU68 gave a segregation ratio o f 4:16:14 (j^= 6.00; p< 0.05), showing a deficiency o f SS homozygotes and an excess o f L L homozygotes i n the F2 exencephalic embryos (See Table 4.3 on pg. 61). A s found i n the E X - 5 0 0 1 study, this result suggested that this region o f Chr 19 from L M / B c may contribute to the risk o f exencephaly i n F2 embryos. Surprisingly, however, the F2 normal control panel gave a similar segregation ratio o f 6:15:18, also showing an excess o f L L homozygotes ( x = 9.46; p< .025) (See Table 4.4 on pg. 61), suggesting segregation distortion. The segregation ratios at D19MU68 from the F 2 exencephalic panel and normal control panel were similar to each other (yr = 0.59; p> 0.975). 2  EX-5001 vs. JHEX-5015 For C h r 13, the segregation ratios from the EX-5001 and J H E X - 5 0 1 5 F2 exencephalic panels showed an excess o f S E L H / B c alleles at the Exenl locus, most o f the significance coming from a surplus o f SS homozygotes. The markers used on the two F 2 exencephalic panels, Fgd3-C/D, Fgfr4-E/F, Ptchl-A/B, D13MU193, D13MU30, and D13MU76 showed genotype distributions that were not significantly different from each other (x?= 2.14, 2.43, 2.13, 1.38, 3.99, and 2.79 respectively). This suggests the contribution to risk o f exencephaly from Exenl is independent o f diet.  63  The segregation ratios from the markers on C h r 5, on the other hand, were very different from each other between the EX-5001 and J H E X - 5 0 1 5 F 2 exencephalic panels. For D5MU95, the genotype distribution for the EX-5001 panel was 29:20:3 ( S S : S L : L L ) , significantly different from 8:20:6 from the J H E X - 5 0 1 5 panel (x= 9.57; p< 0.01). For 2  D5MU30, the segregation ratios from EX-5001 and J H E X - 5 0 1 5 were 28:21:3 and 9:18:7, respectively (x= 8.18; p< 0.025). The J H E X - 5 0 1 5 segregation ratios were not 2  significantly different from Mendelian segregation, where as the EX-5001 segregation ratios showed a great excess o f SS homozygotes at the Exen2 locus. This difference between segregation ratios o f the two F2 exencephalic panels suggests that the Exen2 alleles from S E L H / B c had little or no role in risk o f exencephaly on P M D #5015 and had a strong role on P L R D #5001. Comparing the markers on Chr 11 between the EX-5001 and J H E X - 5 0 1 5 panels, the segregation ratios from DUMitlO  and Scn4a-C/D were very similar to each other  (X = 0.54 and 0.01; p> 0.9). Scn4a-C/D, i n both panels, approached significance and 2  showed an excess o f S E L H / B c alleles at the Exen3 locus, following the same trend as the T G Panel (Juriloff et al., 2001). This suggested that there was a small effect coming from the Exen3 alleles from S E L H / B c , but the individual sample sizes were not big enough to be statistically significant. Because all three F 2 exencephaly panels showed a similar segregation pattern at Scn4a-C/D that suggested statistical significance but did not establish it, we used a technique developed by Fisher to combine the probabilities o f the independent x tests to create an overall test. Although this test did not provide a 2  statistically significant x value 2  (x  2==  10.36; p= 0.11), it suggested there might be a small  effect coming from the Exen3 locus.  64  Comparing the markers on Chr 7 between the two F2 exencephalic panels, the segregation ratios from D7MU75, Rshll-C/D,  and D7MU79 were not significantly  different from each other (x = 1.94, 1.52, and 0.41, respectively; p> 0.1). This provides 2  support that a locus on Chr 7 (Exen4) contributes to the risk o f exencephaly on both diets as the two panels show significant deviations from randomness, showing an excess o f SS homozygotes. The Chr 19 marker, D19MU68, gave similar segregation ratios i n the EX-5001 and J H E X - 5 0 1 5 F 2 exencephalic panels (x = 0.59; p> 0.5) showing deficiencies o f SS 2  homozygotes and excesses o f L L homozygotes. However, the control F 2 panel for the JHEX-5015 F 2 exencephaly panel showed an excess o f L L homozygotes for this region so this remains unresolved. The Chr 10 marker, D10MU180, and the C h r 17 marker, Dl 7Mill 0, gave segregation ratios that did not significantly deviate from randomness i n the two F 2 exencephalic panels (x = 0.57 and 0.12, respectively; p< 0.9). This suggests that these 2  regions on Chr 10 and 17 are not involved i n the risk o f exencephaly i n the F2 embryos on P L R D #5001 or P M D #5015.  Discussion The first major finding was that the hypothesis that one o f the Exen loci interacts with P M D #5015 was supported by this study. The segregation ratios o f the F 2 exencephalic embryos at the Exen2 locus on Chr 5 significantly differed between the E X 5001 and J H E X - 5 0 1 5 studies. In the EX-5001 study, the C h r 5 markers gave non-random genotype distributions showing a huge excess o f SS homozygotes suggesting  65  involvement i n the risk o f exencephaly i n the F2 embryos. In contrast, the genotype distributions o f the C h r 5 markers were quite random i n the J H E X - 5 0 1 5 study showing no evidence o f involvement i n the risk o f exencephaly when mice are fed P M D #5015. This difference between the two panels indicates a possible Exen gene-diet relationship and suggests that the diet effect from P M D #5015 supersedes the effect o f the Exen2 alleles from S E L H / B c on C h r 5, therefore making the Chr 5 genotype virtually irrelevant. The dropping out o f the Exen2 locus when mice are fed P M D #5015 could possibly be attributed to a gene-diet interaction with another Exen locus as well. In this case, embryos with those particular Exen alleles from S E L H / B c would have a greatly elevated risk o f exencephaly on P M D #5015, shifting the other loci to relatively less importance. W e would expect to see the segregation ratios to shift towards randomness at all other loci. Contrary to this prediction, the segregation ratios for the Exenl (Chr 13) and Exen4 (Chr 7) loci did not differ between the two diets. A second alternative hypothesis was that an Exen locus or a locus on C h r 10 or 17 with a weak effect i n the EX-5001 study would interact with P M D #5015 and have a strong effect i n the J H E X - 5 0 1 5 study. This would lead to that locus having a significantly more skewed segregation ratio on P M D #5015 than on P L R D #5001 whereas the segregation ratios at the other Exen loci would remain similarly skewed on both diets. This was not the case as the data does not fit this pattern. A third alternative hypothesis was that the effect from P M D #5015 adds to the genetic effect on risk o f exencephaly. In this case, the segregation ratios for the EX-5001 and J H E X - 5 0 1 5 F 2 exencephaly panels would be similar across all Exen loci. Our data  66  does not support this, though, as we see different segregation ratios at Exen2 between the EX-5001 and J H E X - 5 0 1 5 study. A fourth alternative hypothesis was that heterozygosity across the Exen loci, combined with the effect from P M D #5015, would add more risk o f exencephaly than heterozygosity itself on P L R D #5001. More embryos that are heterozygous across the Exen loci would become exencephalic. W e would expect to see an increase i n heterozygous genotypes across the Exen loci i n the exencephalic embryos i n the J H E X 5015 study. In this case, we did not see this pattern. In summary, the best fitting explanation is that effect from P M D #5015 superseded the effect o f the Exen2 alleles from S E L H / B c . A second major finding i n this study was that a locus on C h r 7, now called Exen4, appeared to contribute to the risk o f exencephaly i n the F 2 embryos. Both F2 exencephaly panels i n the EX-5001 and J H E X - 5 0 1 5 studies gave segregation ratios that showed an excess o f SS homozygotes at this locus. In the previous F 2 sire screen, this region from C h r 7 seemed to have an Exen locus as the /?-value approached statistical significance, but this finding was not supported i n the original T G F 2 exencephaly panel. Despite this, the same C h r 7 region was revisited i n this study and showed that it did have a role i n the risk o f exencephaly. More interesting is that the role o f the Exen4 locus appeared to be more important than the role o f the Exen3 locus. This changes the genetic explanation o f exencephaly i n the S E L H / B c strain, as there are possibly four genes, the most important being on C h r 13, 5, and 7. The third finding i n this study was that a region on C h r 19 from L M / B c may contribute to the risk o f exencephaly i n F2 embryos. The two F 2 exencephaly panels i n  67  the EX-5001 and J H E X - 5 0 1 5 studies showed huge excesses o f L L homozygotes and severe deficiencies o f SS homozygotes at proximal Chr 19 markers. The data is hard to interpret though as the F 2 normal control panel for the J H E X - 5 0 1 5 study showed a similar trend therefore indicating no role for this chromosomal region, but rather segregation distortion. O n the other hand, the F2 normal control panel for the EX-5001 study showed a random distribution o f genotypes at the same marker indicating this region may have a role when compared to the EX-5001 F2 exencephaly panel. Although the L M / B c strain is not focused on i n this study i n regards to whether L M / B c mice contribute anything to the risk o f exencephaly in the F2 studies, it is interesting nonetheless. Therefore, further work is needed to determine the role, i f any, o f this region from Chr 19 in the risk o f exencephaly in the F2 embryos.  68  Chapter 5: Congenic line exencephaly frequencies Introduction Congenic lines are a powerful tool for genetic analysis o f a complex trait. They can be used to precisely map the individual contributing genes and, i n addition, they isolate the individual physiological and developmental effects o f the gene. A s the previous genetic analyses indicated that there are about three additive genes o f moderate effect that contribute to exencephaly i n the S E L H / B c strain (Juriloff et al., 2001), the use of congenic lines seemed an excellent tool to further dissect the trait. The major congenic lines used for m y study were established approximately 5 years ago. One set o f congenic lines had been created by transferring the chromosomal segments that contain the Exenl (Chr 13) and Exen2 alleles from S E L H / B c (Chr 5), respectively, from S E L H / B c into the normal L M / B c strain background (See Fig. 5.1). The formal names, according to mouse c  nomenclature rules, would be L M - S E L H - T i x e n i  c  and L M . S E L H - / i x e « 2  (http://www.informatics.jax.org/mgihome/nomen/strains.shtml). For convenience these lines are referred to as " 1 3 S / L M " and " 5 S / L M . " The reverse congenic lines had been created by transferring the chromosomal segments that contain the normal Exenl and Exenl alleles from L M / B c into the S E L H / B c background (See Figure 5.1). Their formal names would be SELH.LM-Exenl  L  and SELH.LM-Exen2  L  and are referred to as  " 1 3 L / S E L H " and " 5 L / S E L H , " respectively. For the 1 3 S / L M and 1 3 L / S E L H congenic lines, the markers used for selection to transfer the chromosomal segments that contained the Exenl locus were D13MU3, D13MU39, and D13MU76, spanning at least 87 M b on m i d Chr 13. For the 5 S / L M and 5 L / S E L H congenic lines, the markers used for selection to transfer the chromosomal  69  Figure 5.1: A conceptual picture o f the construction o f the congenic lines. Exenl (Chr 13) and Exenl (Chr 5) alleles from S E L H / B c were transferred into the L M / B c background and vice versa. Black bars indicate the 19 S E L H / B c autosomes and the white bars indicate the 19 L M / B c autosomes.  LM/Bc  SELH/Bc  I Chr 5  Exenl  Chr 13  Chr 5  Exenl  BDODDD 5S/LM  i  13S/LM  illlVllllllll "5L/SELH  OOfflooooD i  Chi-13  IllllllllllHflllli 13L/SELH  70  region o f interest that contained the Exen2 locus were D5MU95, D5MU30, and D5MU168, spanning at least 12 M b on distal Chr 5. A t the time the congenic lines were being constructed one o f the purposes was to confirm that the Exenl locus is on m i d Chr 13 and that the Exen2 locus is on distal Chr 5. This purpose has been made redundant by the new EX-5001 F 2 exencephaly panel (See Chapter 4). The second and more important purpose for constructing the congenic lines was to test the multifactorial threshold model, which predicts that each congenic line with part o f the liability to exencephaly from S E L H / B c w i l l express part o f the exencephaly frequency o f S E L H / B c . In addition, this model predicts each o f the four congenic lines w i l l differ from the background strain i n exencephaly frequency, demonstrating the additivity o f the Exen loci. In particular, L M / B c has 0% exencephaly; substituting the Exenl or Exen2 alleles from S E L H / B c ("Exenl " or "Exen2 " respectively) into the L M / B c background would make L M / B c have some exencephaly. Reciprocally, i n the reverse congenic lines, transferring the normal Exenl and Exen2 alleles from L M / B c into the S E L H / B c background would reduce the frequency o f exencephaly compared to that of S E L H / B c itself. Although additive, the Exen loci may not be equal i n effect. The Exenl locus appeared to be the major locus relative to the other participating loci i n previous studies (Juriloff et al., 2001). Based on this, 1 3 S / L M would produce more exencephaly than 5 S / L M and, likewise, 5 L / S E L H (with Exenl alleles from S E L H / B c ) would produce more exencephaly than 1 3 L / S E L H . A third main use o f the congenic lines was to isolate the individual effect o f each Exen locus i n the process o f cranial neural fold elevation i n early embryos and it was  71  hypothesized that these effects would be detectable i n their morphological characteristics (See Chapter 7). A fourth use o f the congenic lines became apparent when the diet effect was discovered (See Chapter 3 for details o f the diet effect), to test whether the congenic lines respond to diet. The congenic lines are useful i n examining i f the diet effect requires a S E L H / B c mother ( S E L H / B c maternal effect) and i f certain Exen loci respond to diet. The data from the congenic lines w i l l be helpful i n interpreting the mechanism o f the diet effect observed i n the S E L H / B c strain and i n the F 2 exencephaly panels. This chapter reports on the use o f the 1 3 S / L M , 5 S / L M , 1 3 L / S E L H , and 5 L / S E L H congenic lines to confirm the chromosomal locations o f the Exenl and Exen2 loci, to test the fit o f observed exencephaly frequencies against the multifactorial threshold model, and to test the genetic basis involved i n elevated frequencies o f exencephaly on diet P M D #5015. In addition, this chapter reports on a new congenic line, 7 S / L M , that isolates the effect o f the Exen4 alleles from S E L H / B c (See Chapter 4), as well as the improved definition o f the chromosomal segments transferred into each congenic line by typing o f additional flanking and intermediate markers.  Materials and Methods Mice A l l mice originated from and were maintained in our animal unit i n the Department o f M e d i c a l Genetics at the University o f British Columbia under standard conditions previously described (See General Materials and Methods). A l l breeding colonies that produced adult mice used i n this study were maintained on Purina  72  Laboratory Rodent Diet #5001 ( " P L R D #5001"). During these experiments, 1 3 L / S E L H was at N F - N F i 2 , 5 L / S E L H was at N F - N F , 1 3 S / L M was at N F 7 - N F i o , 5 S / L M was 5  7  5  8  5  8  9  6  6  at N7F4-N7F7, 1 3 S / L M * was at N7F4-N7F5, and 7 S / L M was at N7F4-N7F5. These congenic lines were homozygous for their respective transferred chromosomal regions. According to Silver, the fraction o f background loci that are still heterozygous at the N t h generation can be calculated as [(l/2) ~ ], with the remaining fraction [1 - (l/2) ~ N  l  l  N  ] homozygous for the inbred strain allele (Silver, 1995). U s i n g this calculation,  1 3 L / S E L H (N5) is identical to S E L H / B c across approximately 94% o f the genome. 5 L / S E L H ( N ) is approximately 99% identical to S E L H / B c . The 1 3 S / L M congenic line 8  (N ) is identical to L M / B c across 97% o f the genome and 1 3 S / L M * , 7 S / L M , and 5 S / L M 6  (all at N7) are approximately 98% identical to L M / B c .  Experimental strategy The first part o f this study involved defining the differential chromosomal segments that were transferred into the S E L H / B c and L M / B c strain backgrounds. This involved typing D N A with flanking and intermediate markers o f the transferred chromosomal segments. The D N A was from mice o f the same lineage the embryo collections were from. To test the multifactorial threshold model, exencephaly frequencies for each congenic line were observed on P M D #5015 to maximize the frequency o f exencephaly in each line to obtain interpretable frequencies i n reasonable sample sizes. This study was done i n part o f two sets i n 2003 and 2004. In general, 6-16 males were used per line to generate embryos.  73  To test for diet effects i n the congenic lines, the other part o f Set 1 (see above) was collected concurrently on P L R D #5001 i n 2003 to examine the genetic basis for the diet effect observed i n the S E L H / B c strain and F 2 exencephaly studies. Based on the data in Set 1, we hypothesized that the presence of Exenl alleles from S E L H / B c ("Exenl ") are required for a response to diet. Set 2 was collected to observe exencephaly frequencies from all four congenic lines at the same time on both diets, to increase sample sizes for some o f the lines because their exencephaly frequencies had been too low to be interpretable, and to confirm the pattern o f data found i n Set 1.  Construction of the 7S/LM and 13S/LM* congenic lines While this study was i n progress, we obtained evidence for the Exen4 locus on Chr 7 from the E X - 5 0 0 1 F 2 exencephaly panel (See Chapter 4). In the construction o f the 1 3 S / L M congenic line, proximal Chr 7 from S E L H / B c had been carried along by purposeful selection i n the backcross generations and later left to chance. Therefore it was possible that proximal C h r 7 from S E L H / B c was still present. After typing the 1 3 S / L M line for Chr 7, we found that proximal Chr 7 was homozygous SS and that this congenic line therefore actually contained the Exen4 alleles from S E L H / B c as well as the Exenl alleles from S E L H / B c . A n experiment was done to separate the loci into two distinct new congenic lines, referred in our lab as " 7 S / L M " and " 1 3 S / L M * . " 1  To create the 7 S / L M and 1 3 S / L M * congenic lines, the 1 3 S / L M congenic line was backcrossed to L M / B c . F o r the 7 S / L M congenic line, the markers used for selection to keep the S E L H / B c chromosomal region o f interest were D7MU75 and D7MU79, while the Chr 13 markers, D13MU3, D13MU39, and D13MU76, were selected against to make 1  13S/LM without the Exen4 allelesfromSELH/Bc 74  sure the Chr 13 segment from S E L H / B c was not carried through. Reciprocally, to make the 1 3 S / L M * line, the S E L H / B c alleles at Chr 13 markers were selected for and the S E L H / B c alleles at C h r 7 markers were selected against. For the 7 S / L M and 1 3 S / L M * congenic lines, the mice with the correct haplotypes for C h r 7 and 13, respectively, were intercrossed to produce homozygous progeny who maintained the lines. The 7 S / L M and 1 3 S / L M * congenic lines were made homozygous after the N 7 generation.  Defining the transferred chromosomal segments After the congenic lines were created, further definition o f the exact segments o f chromosome that were transferred i n each congenic line was necessary. F o r 1 3 S / L M and 1 3 L / S E L H , markers D13MU117, D13MU91, Fgd3-C/D, Fgfr4-E/F, Gprk6-C/D, Lect2C/D, Ntrk2-C/D, Ptcl-A/B, D13MU193, and D13MU30 were used to further define the transferred regions (See Table 2.1 on pg. 30 for locations o f markers). For the 5 L / S E L H and 5 S / L M congenic lines, markers D5MU13, D5Nds2, D5MU24, and D5MU122 were used to further define the transferred chromosomal regions. Markers D7MU75, D7MU79, Ccnel-A/B, Gnefr-C/D, and D7MU62 were used to define the C h r 7 segments i n the 7 S / L M and 5 L / S E L H congenic lines. T w o to four mice were typed for each marker i n each congenic line. D N A was prepared from tail tips using the QIAamp D N A M i n i - k i t and the genotypes o f the mice were identified by P C R using the above markers as described in General Materials and Methods. Primer sequences can be found i n Appendix 1 for primers that were designed i n our lab.  75  Testing the multifactorial threshold model The female mice were fed P M D #5015 to maximize the exencephaly frequencies in attainable sample sizes for each congenic line. Sample sizes were limited by availability o f mice and a target o f 150-200 embryos per line was used. Based on the binomial distribution, this sample size would have the power to detect a frequency o f 2% exencephaly or greater i n 9 5 % o f samples [(1 - 0 . 0 2 ) = 0.048] (Sokal and Rohlf, 1995). 150  In x tests, a sample size o f 150 would detect reductions i n exencephaly frequency from 2  25% in S E L H / B c to less than 15% atp< 0.05. The females were placed on P M D #5015 at the same time as introduction o f the males and maintained on this diet until autopsy for embryo collection. This experiment was done in 2003 and 2004, to make up parts o f Set 1 and 2.  Testing for diet effects in the congenic lines The congenic lines, 1 3 S / L M , 5 S / L M , 1 3 L / S E L H , and 5 L / S E L H , were used to test for diet effects i n Sets 1 and 2. For embryo collection on Purina Lab Rodent Diet #5001 ( " P L R D #5001), the female mice were fed P L R D #5001 from weaning until autopsy. For embryo collection on Purina Mouse Diet #5015 ( " P M D #5015"), the females were placed on P M D #5015 at the same time as introduction o f males and maintained on it until autopsy. In general, most (>80%) o f the females were on P M D #5015 for up to one week before conception. M i c e taking longer to become pregnant typically were on P M D #5015 for two weeks.  76  Scoring exencephaly Pregnancies were judged by eye and palpation. Developmental stage o f embryos was judged by morphological characteristics. Females were euthanized by carbon dioxide on day 14 o f gestation (range E l 1-E18) for examination o f embryos. The uterus was removed, pinned to a black wax substrate, immersed i n physiological saline (0.85% N a C l ) solution, and cut open to reveal the conceptuses. Dead post implantation embryos ("moles") were recorded. The morphology o f scoreable embryos was examined under a dissection microscope and recorded. The following defects were screened: exencephaly, spina bifida, rachischisis, major abnormalities o f the limbs and tail, gastroschisis, cleft lip, and severe edema. The litters were stored in glass jars, i n either B o u i n ' s fixative or 10% Buffered Formalin Acetate.  Statistical methods To compare the exencephaly frequencies on the two diets i n each congenic line, the chi-square test o f independence was used. The p-value 0.05 was chosen for statistical significance. The Spearman rank correlation was used to test for a relationship between congenic line exencephaly frequencies and mole frequencies (post-implantation mortality rates) in each Set, within each diet. A value o f r near one indicates a strong relationship, s  whereas a value near zero indicates independence. (http://nimitz.mcs.kent.edu/~blewis/stat/spearman.html)  77  Results Defining the transferred chromosomal segments (See Figure 5.2) The large C h r 13 segment transferred from S E L H / B c into the L M / B c background ( 1 3 S / L M and 1 3 S / L M * ) spanned at least 87.6 M b (50 c M ) from D13MU3 (19.8 M b ) to D13MU76 (107.4 M b ) . Additional markers between D13MU3 and D13MU76 confirmed that the whole segment was S E L H / B c homozygous (See Figure 5.2). The chromosomal regions that flank D13MU3 and D13MU76 were not typed and therefore their genotypes are not known. The C h r 13 segment transferred from L M / B c into the S E L H / B c background ( 1 3 L / S E L H ) spanned that same distance as the Chr 13 segment described above and the same intermediate markers between D13MU3 and D13MU76 confirmed that the transferred chromosomal segment was homozygous L M / B c (See Figure 5.2). The S E L H / B c C h r 5 segment transferred into the L M / B c background ( 5 S / L M ) used the markers D5MU95 (122.4 M b ) , D5MU30 (127.1 M b ) , and D5MU168 (134.5 M b ) , spanning 12.1 M b (10 c M ) , to select for the S E L H / B c segment. Typing o f additional markers showed that the segment spanned at least 76.2 M b (44 c M ) from D5Nds2 (71.4 M b ) to D5MU122 (147.6 M b ) (See Figure 5.2). The proximal flanking marker, D5MU13 (35.9 M b ) was back to L M / B c homozygosity. The L M / B c Chr 5 segment transferred into S E L H / B c ( 5 L / S E L H ) spanned at least 12.1 M b (10 c M ) from D5MU95 (122.4 M b ) to D5MU168 (134.5 M b ) . The next closest proximal and distal flanking markers at 109.7 M b (D5Mit24) and 147.6 M b (D5MU122) were back to S E L H / B c homozygosity, showing a more precise area where the Exen2 locus is located, compared to the 5 S / L M line. A s L M / B c markers at proximal Chr 7 had been purposely selected i n early generations o f creation o f the 5 L / S E L H congenic line, this line was typed for Chr 7  78  Figure 5.2: The current chromosomal segments that are i n each congenic line. White regions indicate homozygous L M / B c regions, black regions indicate homozygous S E L H / B c regions, and the hatched regions are unknown. Length o f each chromosome is indicated at bottom i n M b . Locations o f the markers can be found i n Table 2.1 on pg. 30.  13L/SELH  5L/SELH  13S/LM'  5S/LM  D7MH75 D5MH13  D5MH13  D13MH3-  D7MH79 Ccne1-A/B Gnefr-C/D  D13MH3  D13MH13  Ptc1-A/B -  D7MH62 D5MH24  D5MH24  D13Mit30-  D13MH76  D13Mit76  D5MH95 D5MH30 D5MH168  D5MH122  D5MH122  Chr 7  Chr 5  •  LL  •  SS  Chr 7 133.1 Mb  149.2 Mb  116.5 Mb  VO  D5MH95 D5MH30 D5MM68  Chr 5  Chr 13  D7MH79 Ccne1-A/B Gnefr-C/D  D7MH62  Rd-A/B  D13Mit193 -  I  D7MH75  D5Nds2  D5Nds2  Fgd3-C/D\ Fgfr4-E/F \ Gprk6-C/D \ D13Mit13^  7S/LM  •  Not typed  Scale __10 Mb  (D7MU75). 5 L / S E L H was homozygous L M / B c for markers on C h r 7; therefore the normal Exen4 alleles from L M / B c were also transferred into this line along with the normal Exen2 alleles, complicating the interpretation o f the data i n this line. The S E L H / B c C h r 7 segment transferred into the L M / B c background ( 7 S / L M ) and the L M / B c Chr 7 segment found i n the 5 L / S E L H line spans at least 16.9 M b from D7MU75 at 5.9 M b to D7MU79 at 22.8 M b (See Figure 5.2). In the 7 S / L M line, typing 2  of additional markers showed that the segment spanned at least 20.6 M b (14.3 c M ) from D7MU75 to Ccnel-A/B.  The next distal marker, Gnefr-C/D at 33.9 M b , was back to  L M / B c homozygosity. In 5 L / S E L H , typing o f additional markers showed that the L M / B c Chr 7 segment spanned at least 16.9 M b (14.3 c M ) from D7MU75 to D7MU79. The next distal marker, Ccnel-A/B,  as w e l l as the other two, were back to S E L H / B c homozygosity.  Testing the multifactorial threshold model Set 1 was the first formal examination o f the frequency o f exencephaly i n the congenic lines and it was not known whether the lines would produce any exencephaly. Therefore, to test the multifactorial threshold model, litters were collected on P M D #5015 to maximize the exencephaly frequency i n each line (See Table 5.1). Set 2 litters for this test were a replicate for the Set 1 data and therefore were also collected on P M D #5015. There was good agreement between the values (exencephaly frequencies) obtained i n Set 1 and Set 2 (See Table 5.1), with the exception o f 1 3 S / L M that had a significantly higher exencephaly frequency i n Set 2 (x= 5.00; p< 0.05). One explanation for this difference i n 2  D7MU75 mapped to 5.9 Mb in the UCSC Feb. 2003 contig. This contig is being used because D7Mit75 appeared to be mis-mapped in the May '04 contig. The physical map on some chromosomes is not complete and there appears to still be mistakes. Chr 7, for example, seems to have created an order that is incompatible with several individual linkage mapping experiments.  2  80  Table 5.1: Exencephaly frequencies and post-implantation mortality rates ("moles") i n the S E L H / B C and L M / B c strains and congenic lines on P M D #5015. Date  Strain/Line  # litters  # implants  % moles  # embryos  #Exen  % Exen  Apr-Sept '04  SELH/Bc  21  217  10.1  195  49  25.1  Feb '02  LM/Bc  11  112  3.6  108  0  0  2003 - Set 1  13L/SELH  16  144  17.4  119  1  0.8  5L/SELH  8  59  15.9  53  4  7.5  13S/LM  22  239  9.6  216  21  9.7  5S/LM  12  145  7.6  134  5  3.7  7S/LM  12  143  7.0  133  0  0  13L/SELH  21  215  4.7  205  0  0  5L/SELH  20  194  10.3  174  17  9.8  13S/LM  20  225  12.4  197  43  21.8  5S/LM  16  192  5.2  182  2  1.1  7S/LM  15  153  6.5  143  4  2.8  13S/LM*  29  295  7.8  280  8  2.9  2004 - Set 2  A  A  The exencephaly frequency for 1 3 S / L M significantly differs between the 2 sets (x 11.5; p< 0.001)  A  2=  Note: Litters o f <5 implants were deleted  81  c  the 1 3 S / L M congenic line was that Chr 7 (Exen4 ) was possibly segregating i n Set 1 and, by chance, went to fixation i n the S E L H / B c allele in the parents o f Set 2. In comparing exencephaly frequencies between the congenic lines and control strains, Set 2 became the best test as it was more complete, having relatively matched sample sizes between the congenic lines and containing the 1 3 S / L M * congenic line. Demonstrated on Table 5.1 (pg. 81), the congenic lines on the L M / B c background each produced exencephaly, whereas L M / B c itself produced 0% exencephaly. Substituting Exenl into the L M / B c background caused L M / B c to have approximately 3% exencephaly ( 1 3 S / L M * ) . Substituting Exen2 into the L M / B c background produced s  1-4% exencephaly ( 5 S / L M , Sets 1 and 2). This supported the prediction from the multifactorial threshold model that these congenic lines would express part o f the liability to exencephaly from S E L H / B c . In the 7 S / L M line, the interpretation was less clear because all the exencephalic embryos obtained (2.8% exencephaly) were from one sire in Set 2, raising the possibility that a non-Chr 7 genetic factor might have been involved. c  c  The combination o f Exenl and Exen4 (13S/LM) produced more exencephaly than either Exen gene alone, but less than S E L H / B c (25%), also i n agreement with the multifactorial threshold model. Reciprocally, the congenic lines on the S E L H / B c background produced less exencephaly than S E L H / B c itself. O n P M D #5015, S E L H / B c had 25.1% exencephaly. The 1 3 L / S E L H congenic line had 0-1% exencephaly and 5 L / S E L H had approximately 10% exencephaly, both sets for both lines significantly less than S E L H / B c itself ( A l l x  2  values > 7.67; all /^-values < 0.01). These data showed that 1 3 L / S E L H and 5 L / S E L H  82  retained some o f the risk to exencephaly from S E L H / B c , again i n agreement with the multifactorial threshold model. When fed the lower risk diet, P L R D #5001, the congenic lines tested, 1 3 S / L M , 5 S / L M , 1 3 L / S E L H , and 5 L / S E L H , all produced some exencephaly (See Table 5.2), but some were at lower frequencies than on P M D #5015.  Testing the genetic basis of the diet effect Diet Effects in Set 1 (Table 5.3, pg. 85) Part o f Set 1 on P L R D #5001 was collected concurrently with the other part o f Set 1 on P M D #5015 to look for diet effects. A s expected, the S E L H / B c strain had a significantly higher exencephaly frequency on P M D #5015 (25%) than P L R D #5001 (12%) tf= 13.0, p< 0.001) (See Table 5.3). L M / B c had 0% exencephaly on both diets. Comparing the exencephaly frequencies on both diets for each congenic line, Table 5.3 shows that the 1 3 S / L M line had a significantly higher exencephaly frequency on P M D #5015 (10%) than on P L R D #5001 (4%) tf= A.5\;p< 0.05). The 1 3 L / S E L H congenic line on both diets had exencephaly frequencies too low for statistical interpretation, but it appeared that this line did not respond to diet. The 5 L / S E L H congenic line showed no significant difference between exencephaly frequencies on the two diets (x= 0.07), with 2  a slightly higher exencephaly frequency on P M D #5015 (7.5%) than on P L R D #5001 (6.4%). From Set 1, the congenic line data suggested that the lines with Exenl ( 1 3 S / L M s  and maybe 5 L / S E L H ) responded to P M D #5015 and that the congenic lines lacking Exenl did not ( 1 3 L / S E L H only). The 5 S / L M line was not compared between the diets because there was no data collected on P L R D #5001 i n this set.  83  Table 5.2: Exencephaly frequencies and post-implantation mortality rates ("moles") i n the S E L H / B C and L M / B c strains and congenic lines on P L R D #5001. Date  Strain/Line  # litters  # implants  % moles  # embryos  #Exen  % Exen  Apr-Sept '04  SELH/Bc  23  249  10.0  224  26  11.6  Oct-Nov '03  LM/Bc  11  125  4.0  120  0  0  2003 - Set 1  13L/SELH  5  49  6.1  46  0  0  5L/SELH  11  101  21.8  79  5  6.4  13S/LM  15  168  8.3  154  6  4.0°  13L/SELH  20  211  6.2  198  2  1.0  5L/SELH  20  192  16.1  161  5  3.1  13S/LM  19  180  10.0  162  22  13.5 "  5S/LM  19  212  4.2  203  1  0.5  2004 - Set 2  A  A  One embryo had cleft face i n this line  The exencephaly frequency for 1 3 S / L M significantly differs between the 2 sets 9.17; p<0.01)  B  Note: Litters o f <5 implants were deleted  84  Table 5.3: Exencephaly frequencies in parental strains and congenic lines on P M D #5015 and P L R D #5001 for Sets 1 and 2.  Exencephaly frequency PLRD #5001 PMD #5015 SELH/Bc LM/Bc  11.6 0.0  25.1 0.0  Set1 13L/SELH 5L/SELH 13S/LM  0.0 6.4 4.0  0.8 7.5 9.7  Set 2 13L/SELH 5L/SELH 13S/LM 5S/LM  1.0 3.1 13.5 0.5  0.0 9.8 21.8 1.1  p< 0.001  NS p< 0.05  p< 0.025 p< 0.05  85  Diet effects in Set 2 (Table 5.3, previous page) Comparing the exencephaly frequencies on both diets for each congenic line, the data from Table 5.3 suggests that the 1 3 S / L M congenic line responded to diet, as in Set 1 (X = 4.08; p< 0.05). O n P M D #5015 the frequency was 22% and on P L R D #5001 the 2  frequency was 14%. The exencephaly frequencies from the 5 L / S E L H lines on the two diets were significantly different from each other (3% and 10%, respectively) ( £ = 6.05; p< 0.025), supporting the hypothesis that the congenic lines with Exenl respond to diet. s  The exencephaly frequencies from the 5 S / L M congenic lines were about 1% and did not differ between the two diets suggesting that this congenic line did not respond to diet. The exencephaly frequencies from 1 3 L / S E L H were also 1% or less on both diets suggesting this line did not respond to diet. Thus both lines lacking Exenl from S E L H / B c did not appear to respond to diet. It is important to note that the 5 S / L M and 1 3 L / S E L H congenic lines had very low exencephaly baseline frequencies on P M D #5015 making the detection o f a decrease i n exencephaly due to the P L R D #5001 diet effect difficult to detect.  Post-implantation mortality rates ("moles ") and other defects S E L H / B c had a 10% post-implantation mortality rate on P M D #5015 and P L R D #5001 (Tables 5.1 and 5.2 on pgs. 81 and 84 respectively). There appeared to be no consistent beneficial effect on post-implantation mortality from removing S E L H / B c Exen alleles from S E L H / B c . Although, the post-implantation mortality rates from 1 3 L / S E L H mostly tended to be lower than S E L H / B c , rates from 5 L / S E L H typically tended to be higher than S E L H / B c . The higher rates i n 5 L / S E L H can be attributed to effects o f  86  specific litters. There appeared to be clustering where 1-3 litters (average 16% o f litters) in each Set (divided by diet) contributed approximately half or more o f the mole count. There was no consistent pattern o f post-implantation mortality rates between the two diets, therefore suggesting this rate was independent o f diet. L M / B c had an approximate 4% post-implantation mortality rate on both diets (Tables 5.1 and 5.2; pgs. 81 and 84, respectively). W i t h the exception o f 5 S / L M , all the other congenic' lines on the L M / B c background tended to have higher post-implantation mortality rates than L M / B c itself. This suggests that transferring Exen alleles from S E L H / B c into L M / B c increases embryonic mortality. The post-implantation mortality rates on both diets appeared to be similar and there was no notable evidence o f a diet effect on post-implantation mortality rate. Whether there was a relationship between exencephaly frequency and mole frequency was assessed by Spearman's rank correlation (http://nimitz.mcs.kent.edu/ ~blewis/stat/spearman.html). The exencephaly and mole frequencies were ranked across all the congenic lines and parental strains in each set, within each diet. O n P M D #5015, the Spearman rank correlation coefficients were calculated to be 0.52 and 0.92 i n Sets 1 and 2 respectively. O n P L R D #5001, they were 0.88 and 0.81 i n Sets 1 and 2 respectively. Three out o f those 4 values were suggestive o f a relationship between exencephaly and mole frequency, but only one ranking assessment was significant (0.92 in Set 2 on P M D #5015). This suggested that as exencephaly frequency goes up, so does post-implantation mortality, which i n general happens before neural tube closure. A m o n g the hundreds o f embryos examined i n the congenic lines, few defects other than exencephaly occurred. 1-4 embryos with severe edema were observed in each  87  congenic line across Sets 1 and 2 with the exception o f 7 S / L M . In addition, 1 embryo in the L M / B c strain had severe edema. Cleft face was observed i n one embryo in the S E L H / B c strain and i n the 5 L / S E L H congenic line. In these two embryos the neural tube failed to close i n the forebrain folds and therefore, the face could not develop normally. Gastroschisis was also observed i n the cleft face embryo i n S E L H / B c .  Discussion The C h r 13 segment from S E L H / B c transferred into L M / B c ( 1 3 S / L M ) , and vice versa ( 1 3 L / S E L H ) , was made purposely big to ensure that the Exenl locus was transferred. The F 2 sire genome screen pointed to a region on m i d C h r 13 that likely contained the Exenl locus (Juriloff et al., 2001), yet this region was a zone o f probability that had unclear boundaries due to sampling variation and the fact that no single specific genotype across the Exen loci is associated with exencephaly, nor does an F 2 sire have to have the Exenl alleles from S E L H / B c to transmit exencephaly. M o r e markers were typed on the original F2 exencephaly panel to help precisely map the Exenl locus, but the difficulty was that F 2 recombinants in this region did not provide valuable information o f the location o f the Exenl locus because F2 embryos do not have to be SS or S L at this locus to have exencephaly. Taken together, these experiments showed a region on m i d Chr 13 that had a high probability o f containing the Exenl locus and i n order to ensure that the Exenl locus was transferred into L M / B c and vice versa, it was necessary to transfer a segment o f that size. The chromosomal segments containing the Exenl locus transferred into the Chr 5 congenic lines, 5 S / L M and 5 L / S E L H , were not as big as the C h r 13 segments transferred  88  into the strain backgrounds. The available markers for distal Chr 5 were used to transfer the Chr 5 segments into S E L H / B c and L M / B c and it was assumed that the segment transferred would be larger than the region indicated by the markers. The segments that are transferred when making congenic lines are often bigger than the region marked by the selection markers due to linkage. According to Silver, the expected average length o f the differential chromosomal segment transferred into a background strain i n c M can be calculated as [200(1-2" )/N] using one marker to select N  for the segment i n the backcrosses. For values o f N greater than five, this equation can be simplified to [200/N] (Silver, 1995). It is important to note, though, that lengths o f the transferred chromosomal segments can vary greatly at the same generation given the inherently random distribution o f crossover sites (Silver, 1995). If two (or more) markers are used to select for the segment, the equation could be adapted (for this thesis) to include the length between the two boundary markers for each side o f the segment so it would look like [(200/N) + length o f segment], figuring that the segment would have generally half o f the expected average length (200/N) flanking it on each side. The distal Chr 5 segment transferred into the 5 L / S E L H congenic line (at Ng) was 10 c M - 25 c M (See Figure 5.2 on pg. 79). U s i n g the adapted equation, the expected average length at N  8  was calculated to be 35 c M , larger than what was observed in 5 L / S E L H . The distal Chr 5 segment transferred into 5 S / L M (at N7) was found to be longer, spanning 44 c M - 65 c M (See Figure 5.2 on pg. 79). A t this generation (N7) the expected average length was calculated to be approximately 39 c M . This shows that the segments were expected to be quite large for 5 S / L M and 5 L / S E L H . Therefore, because o f random crossover sites and linkage, it is not surprising that the lengths o f the transferred segments are what they are.  89  The first major finding i n this study was that the congenic lines produced some exencephaly. One alternative expectation was that none o f the lines would have exencephaly, but that each would have a mild delay i n elevation o f the midbrain folds i n Day 8 embryos insufficient to cause exencephaly. The second major finding was that the data from the congenic lines generally supported the multifactorial threshold model. Transferring i n the Exenl  or Exen2 into the L M / B c background made L M / B c have  some exencephaly; taking out Exenl  or Exenl  and replacing it with the normal gene  reduced the frequency o f exencephaly, but did not make the S E L H / B c strain lose exencephaly completely. W e did not make any Exen3 congenic lines, but the results o f the F2 studies (See Chapter 4) suggested a minor contribution to the risk o f exencephaly. The relative impact o f the Exen4 locus was less easily interpretable because the only exencephaly observed i n this line came from one sire i n Set 2, raising the possibility that a non-Chr 7 genetic factor was contributing to that exencephaly. Applying Falconer's multifactorial threshold normal distribution curves to the exencephaly frequencies provides some insight into the size o f the effect o f substituting normal Exen alleles from L M / B c into S E L H / B c and transferring S E L H / B c Exen alleles into L M / B c (See F i g . 5.3). The normal distribution curves are anchored on the threshold of S E L H / B c ; 2 5 % o f S E L H / B c embryos fall beyond the threshold and are affected (exencephaly). Removing Exenl  had a huge effect on the exencephaly frequency and the  normal distribution curve was shifted approximately 2 standard deviation units ("a") to the left. Removing the S E L H / B c Exenl and Exen4 alleles ( 5 L / S E L H ) shifted the normal distribution curve approximately 2/3 a to the left (Fig. 5.3). In both Sets, the exencephaly frequency from 5 L / S E L H was significantly higher than 1 3 L / S E L H (x= 5.84 and 20.97,  90  Figure 5 . 3 : Testing the multifactorial threshold model with the congenic lines. The congenic lines on the SELH/Bc background show that transferring in normal Exen alleles from LM/Bc retain some liability to exencephaly risk. The congenic lines of the LM/Bc background show that transferring in Exen alleles from SELH/Bc makes LM/Bc express part of the exencephaly frequency of SELH/Bc. See Table 5.4 on next page for the x values. T = threshold.  SELH/Bc background  SELH/Bc  LM/Bc background  LM/Bc  0.08%  Embtyonic liability trait  13S/LM*  5L/SELH  5S/LM  Y-axis = number of embryos  91  Table 5.4: Values of x (obtained from Falconer and Mackay, 1996) related to the combined exencephaly frequencies from Sets 1 and 2. This table accompanies Figure 5.3 on the previous page and Figure 5.4 on pg. 94. The threshold is anchored at 5 to keep the x values positive on the scale and the x values are the distance the mean is from the threshold i n standard deviation units and (5-x) is the mean o f the distribution. The threshold is anchored to S E L H / B c .  Combined (% exen) 25.1  X  -  Set 2 (% exen) 25.1  0.674  Mean of distribution 4.326  0.8  0  0.3  2.748  2.252  7.5  9.8  9.3  1.311  3.689  _  0  0.08  3.156  1.844  2.9  2.9  1.881  3.119  3.7  1.1  2.2  2.014  2.986  9.7  21.8  21.8  0.772  4.228  0  2.8  1.4  2.197  2.803  Set 1 (% exen) SELH/Bc  (Exenl,2,3.4) 13L/SELH  (Exen2,3,4) 5L/SELH  (Exenl,3) LM/Bc  A  (No Exens) 13S/LM*  (Exenl) 5S/LM  (ExenT) 13S/LM  B  (Exenl,4) 7S/LM  (Exen4)  Estimated L M / B c exencephaly frequency from all the combined studies i n the lab, including my studies and past studies.  A  For 1 3 S / L M only the Set 2 value was used because the values from Sets 1 and 2 were significantly different from each other (See Table 5.1 on pg. 81)  B  92  respectively; p< 0.025 and 0.001). This, along with the information from the curves, suggests the Exenl locus is more important than the Exen2 and Exen4 loci i n the context of S E L H / B c background and P M D #5015. The effect o f transferring i n S E L H / B c Exen genes into the L M / B c background is harder to interpret. It is difficult to accurately estimate the exencephaly frequency i n the L M / B c strain because it is so low and small changes i n frequency have a large effect o f location o f a strain relative to the threshold when the strain is at extremely low frequencies. G i v e n this, doing the same sort o f test against the multifactorial threshold model on the L M / B c background is not going to be interpretable when determining the size o f the effects o f Exenl and Exen2 i n the L M / B c strain. The 1 3 S / L M congenic line (containing the Exenl and Exen4 alleles from S E L H / B c ) provided a fortuitous opportunity to test the multifactorial threshold model. In this model, i f two genes act additively, the result, as measured i n terms o f frequency o f exencephaly, may appear to be synergistic depending on the magnitude o f the shifts and the relation o f distribution to the threshold (Fraser, 1976). 1 3 S / L M * had approximately 3% exencephaly; b y adding the effect o f the Exen4 alleles from S E L H / B c (13S/LM) the frequency goes up to approximately 22%. Figure 5.4, using the multifactorial threshold model, demonstrates that this is an additive effect. The 1 3 S / L M * curve shifts approximately 1.1 standard deviation units ("a") to the left from the 1 3 S / L M curve and c  that difference (1.1 a) is the predicted effect o f Exen4 . This predicted value correlates to an approximate 2.3% predicted exencephaly frequency for the 7 S / L M line (Exen4) (Falconer and Mackay, 1996). What was observed i n the 7 S / L M line was 1.4%  93  Figure 5.4: Additivity o f the Exen loci. This figure demonstrates the additivity o f the Exenl and Exen4 loci using the multifactorial threshold model. Indicated i n parentheses under the congenic line names are the exencephaly frequencies and the related x values (See Table 5.4 on pg. 92). The threshold is anchored to S E L H / B c and this figure is based on the L M / B c mean (0% related to 3.090 a). L M / B c mean  Threshold  (3.090)  Calculations 3.090a - 0 . 7 7 2 a - 2 . 3 1 8 a  13S/LM  3.090 a - 1 . 8 8 1 a = 1.209 a  (22% - 0.772)  2.318 a - 1.209 a = 1.109 a  Predicted effect of Exen4 •  13S/LM*  (1.109 s d u ) « 2.3% predicted exencephaly frequency for 7S/LM  (3% - 1.881)  7S/LM (1.4%-2.197)  94  exencephaly, shifting the 7 S / L M curve approximately 0.9 a from the L M / B c mean value. This real value is i n good general agreement with the predicted value, demonstrating very well the additivity o f the Exen genes, again supporting the multifactorial threshold model. The third major finding o f this study was the diet response observed i n the 1 3 S / L M and 5 L / S E L H congenic lines. 1 3 S / L M had a significantly higher exencephaly frequency on P M D #5015 than on P L R D #5001 i n both sets. Not only did this line show a diet effect, it also demonstrated that the diet effect does not require an S E L H / B c mother. Based on Set 2, 5 L / S E L H (containing Exenl ) also showed a diet effect supporting the hypothesis that congenic lines with Exenl respond to diet. Notably, mothers i n both situations had Exenl , not just the embryos, leaving the possibility o f a s  maternal effect o f a semidominant Exenl from S E L H / B c as well as a direct effect in embryos. The interpretation for 1 3 L / S E L H and 5 S / L M was less clear because o f the low exencephaly frequencies observed in these lines. L o w exencephaly frequencies on P M D #5015 indicate that there is little or no diet effect. W e observed no big diet effects i n these congenic lines and huge sample sizes would be needed to detect a small diet effect. It appeared that genotypes with Exenl ( 1 3 S / L M and 5 L / S E L H ) responded to s  diet and genotypes without Exenl ( 5 S / L M and 1 3 L / S E L H ) did not respond. These diet s  response results are complicated by the fact that both the 1 3 S / L M and 5 L / S E L H lines are contaminated with other Exen alleles from either L M / B c or S E L H / B c depending on the strain background. 1 3 S / L M contains Exen4 , as well as Exenl , and 5 L / S E L H contains the normal Exen4 as well as the normal Exen2 from L M / B c . Since Exen4 is not present s  in 5 L / S E L H , then the diet response i n not likely due solely to Exen4. Taken together, the  95  1 3 S / L M and 5 L / S E L H congenic lines indicate that the presence o f Exenl , on either strain background, leads to a diet effect. Future work could test the 1 3 S / L M * congenic line for a diet effect to further investigate the role o f the Exenl alleles from SELFI/Bc. In conclusion, all the congenic lines had some exencephaly and the data generally supported the multifactorial threshold model. In addition, they have provided some insight into the mechanism o f the diet effect, as the congenic lines with Exenl  responded  to diet, suggesting a gene-diet interaction. The use o f the congenic lines proved valuable in further investigating the genetic basis o f exencephaly and the diet effect i n the S E L H / B c strain.  96  Chapter 6: Recombinant congenic lines Introduction Congenic lines are a valuable tool to precisely map the individual contributing genes o f a complex trait. The previous chapter reported the improved definition o f the chromosomal segments transferred into each congenic line that were established approximately 5 years ago. The chromosomal segments that had been transferred into the Chr 13 congenic lines, 1 3 L / S E L H and 1 3 S / L M , spanned at least 87 M b on m i d Chr 13 and were purposely selected to be that big to ensure the Exenl locus was transferred. Since these Chr 13 segments were so big, breaking it up into smaller segments by creating recombinant congenic lines was the next necessary step i n precisely mapping the Exenl locus. The 1 3 L / S E L H congenic line, as previously reported i n Chapter 5, has a very low exencephaly frequency. This line was the most reasonable line to create recombinant congenic lines from because the exencephaly frequency would immediately go up to " S E L H / B c - l i k e " frequencies i f the Exenl alleles (in homozygous state) from S E L H / B c CExenF") were incorporated back into the line. This makes the presence o f Exenl  easy  to detect and we can therefore deduce that the current L M / B c chromosomal segment i n that same line excludes the Exenl locus. I f one o f the recombinant congenic lines has the same low exencephaly frequency as 1 3 L / S E L H , then it can be inferred that the Exenl locus is still located i n the L M / B c chromosomal segment. K n o w i n g whether Exenl was incorporated back into a recombinant congenic line/lines we can then compare the Chr 13 haplotypes o f these lines, side b y side, and conceptually deduce where the Exenl locus is excluded ( L M / B c regions) and where it is predicted to be ( S E L H / B c regions). There are  97  three recombinant congenic lines that have been created i n the lab and they are referred to as "13L/Rec-Line 1," "13L/Rec-Line 6," and "13L/Rec-Line 7." This chapter reports on the use o f the recombinant congenic lines to attempt to more precisely map the Exenl locus on Chr 13.  Materials and Methods Mice A l l mice originated from and were maintained i n our animal unit i n the Department o f M e d i c a l Genetics at the University o f British Columbia under standard conditions previously described (See General Materials and Methods). A l l breeding colonies that produced adult mice used i n this study were maintained on Purina Laboratory Rodent Diet #5001 ( " P L R D #5001"). During these experiments, 13L/RecLine 1 was at N8F2-N8F3 and 13L/Rec-Line 7 was at N F 2 , and 13L/Rec-Line 6 was at 8  N F . 8  2  Experimental strategy Exencephaly frequencies for each recombinant congenic line were observed on P M D #5015 to maximize the frequency o f exencephaly i n each line and to stay consistent with the previous studies that tested the multifactorial threshold model (Chapter 5). For 13L/Rec-Line 1, 6 males were used to generate embryos. For 13L/Rec-Line 7, 1 male were used to generate embryos. For 13L/Rec-Line 6, 2 males were used to generate embryos. To precisely map the Exenl locus, the recombinant congenic lines with different haplotypes, containing different locations o f the L M / B c chromosomal segments, were  98  compared against each other. U s i n g the exencephaly frequency data together with the haplotype data from each recombinant congenic line, we can conceptually deduce the location o f the Exenl locus b y comparing them against each other.  Construction of the 13L/Rec lines Backcrossing the 1 3 L / S E L H congenic line back to S E L H / B c was the first step i n creating the recombinant congenic lines (See Fig. 6.1). The " F l " generation, being heterozygous on C h r 13, was then backcrossed to S E L H / B c to generate recombinants. A l l offspring from these crosses ( " F l " x S E L H / B c ) were genotyped at selected Chr 13 markers to look for recombinants. When a recombinant was found, it was backcrossed again to S E L H / B c (See F i g . 6.1-C) to propagate this haplotype to obtain sibs o f the same genotype since it was unlikely to obtain a male and female that had the same recombinant haplotype from the " F l " x S E L H / B c cross. When sibs with the same haplotype were found ($ and 3) b y genotyping at selected markers, they were crossed to each other to produce animals that were homozygous for the desired segment (See Fig. 6.1-D). When a male and female with the desired homozygous genotype were found, they were then crossed to each other to maintain the homozygous line and to create the mice used i n exencephaly frequency studies (See Fig. 6.1-E). For 13L/Rec-Lines 1, 6 and 7, markers D13MU3, Fgd3-C/D, Fgfr4-E/F, Gprk6C/D, D13MU13, Ptchl-A/B, D13MU193, D13MU30, and D13MU76 had been used to select and further define these lines. Ntrk2-C/D was typed on 13L/Rec - Lines 6 and 7 as well.  99  Figure 6.1: Conceptual diagram o f the creation o f the recombinant congenic lines  13L/SELH X S E L H / B c C h r 13  C h r 13  L ----L  S  S  I  L  S  S  I  L  s lis "F1" C h r 13 S-L-LL  B  S.-.-L S....L  S E L H / B c X "Rec Line A" C h r 13  siis s.-is  ? Sib X  SELH/Bc C h r 13  S-L-I-S  s-is S ..  "Rec Line B" X S E L H / B c  C h r 13  C h r 13  S  S  L  iis  L .  S  I  S  L  S__  X  In generations C, D, and E, the genotypes are selected based on markers and other segregants are not shown.  S  is  C h r 13  S....S  Sib  $ Sib X  s  S i  s is s  S  Sib  C h r 13  C h r 13  C h r 13  C h r 13  S  siis  L --  Liis  L .  L  I  L  L  S....S  ? Sib X  $ Sib X $ Sib C h r 13 S  L LI  IS L L  C h r 13  siis L LI  Homozygous Rec Line A  LS  L _. S .  S  IS Sib  C h r 13  C h r 13  L U L  L U L  L l S  L  IS  L _S  Homozygous Rec Line B 100  Defining the chromosomal segments in the recombinant congenic lines After the recombinant congenic lines were created, further definition o f the exact segments o f chromosome that were transferred i n each line was necessary to create sharper boundaries between the L M / B c segments and S E L H / B c background. For 13L/Rec-Line 1, markers Adcy2-A/B, Nkd2-A/B, Nr2fl-A/B, G2151-A/B, Edil3-A/B, and D13MU78 were used. For 13L/Rec-Line 6, Nr2fl-A/B and D13MU78 were used. Only D13MU78 was additionally typed i n 13L/Rec-Line 7. Three to ten mice were typed for each new marker i n each congenic line. The "non-Mit" markers were designed i n our lab (See General Materials and Methods).  Scoring exencephaly Scoring exencephaly has been described i n the previous chapter (See Chapter 3). Tissue was kept from some exencephalic embryos from 13L/Rec-Line 1. They were individually washed i n ice-cold sterile saline solution and stored i n cryovials at -20°C. In addition, tail tissue was kept from 5 13L/Rec-Line 1 pregnant females that were part o f this study. D N A was prepared by the QIAamp D N A M i n i - k i t (See General Materials and Methods) and was used to further define the chromosomal segments i n this line.  Statistical methods To compare the exencephaly frequencies o f the recombinant congenic lines, the chi-square test o f independence was used (http://www.georgetown.edu/facultv/ballc/ webtools/web chi.html). The/>value 0.05 was chosen for statistical significance.  101  Results Defining the chromosomal segments (See Figure 6.2) The L M / B c chromosomal segment in 13L/Rec-Line 1 spanned at least 7.6 M b (3 c M ) from Fgfr4-E/F'to Ptchl-A/B (See Fig. 6.2). Additional markers were typed between Ptchl-A/B (at 61.8 M b ) and D13MU193 (88.4 M b ) to sharpen the boundary between the L M / B c segment and the part that was back to the S E L H / B c background. Typing showed that the L M / B c segment spanned at least 20.0 M b (12 c M ) from Fgfr4-E/F (54.2 M b ) to Nr2fl-A/B (74.2 M b ) . The next distal marker, G2151-A/B (82.0 M b ) , was back to S E L H / B c background and S E L H / B c homozygosity extended to the end o f the chromosome, as confirmed by D13MU78 (115.8 M b ) . For 13L/Rec-Line 1, 5 mice were typed for the new additional markers that were made in the lab to further sharpen the boundary between Ptcl-A/B and D13MU193 (See Figure 6.2). Surprisingly, 2 o f the 5 mice were heterozygous at Nr2fl-A/B. For that reason, I went back to the archived frozen exencephalic embryos that were kept when collecting frequency data (See Materials and Methods) and prepared D N A from the 4 exencephalic embryos and typed them for Nr2fl-A/B. One o f the 4 exencephalic embryos was found to be heterozygous at this marker and the other three were homozygous L M / B c . Given that three o f the exencephalic embryos were " L L " at this marker, it could be concluded that the Exenl locus is not located at Nr2fl-A/B. Therefore, it does not matter i f some mice i n this line are heterozygous at this marker. The L M / B c chromosomal segment i n 13L/Rec-Line 6 spanned at least 19.0 M b (18 c M ) from D13MU193 to D13MU76. Additional typing showed that it spanned at least  102  Figure 6.2: The current intervals on Chr 13 in the recombinant congenic lines. The exencephaly frequencies observed from each line are indicated below each line's Chr 13. Chr 13 is 116.5 Mb long. The marker locations can be found in Table 2.1 on pg. 30.  13L/SELH  D13Mit3 H  Fgd3-C/D Fgfr4-E/F Gprk6-C/D D13Mit13 Ptd-A/B  \ ^ J  13L/Rec-Line6  13L/Rec- Line 1  D13MH3  D13MH3  D13MH3  Fgd3-C/D Fgfr4-E/F \ Gprk6-C/D D13MH13 Ptc1-A/B Adcy2-A/B  Fgd3-C/D Fgfr4-E/F Gprk6-C/D D13MH13 Ntrk2-C/D' Ptc1-A/B Adcy2-A/B' Nkd2-A/B ' Nr2f1-A/B  Fgd3-C/D Fgfr4-BF v l Gprk6-C/D ^ D13Mit13 Ntrk2-C/D  G2151-A/B Edil3-A/B D13Mit193  D13MH193 -|  D13MH193  D13MH30  D13MH30  2  4  Nkd2-A/B Nr2f1-A/B  D13MH193 D13MH30 D13MH76  Scale 10 Mb  13L/Rec- Line 7  -I  Ftc1-A/B  D13MH30  D13MH76  D13MH76 -  D13MH76-  D13MH78  D13Mit78 _  D13MH78.  8.8%  18.3%  0.3% •  LL  SS  §  Not typed  20%  51.2 M b from Adcy2-A/B (at 64.6 M b ) to the bottom o f the chromosome at D13MU78 (at 115.8 M b ) . The part o f C h r 13 that is back to the S E L H / B c background spanned at least 42.0 M b (26 c M ) from D13MU3 (19.8 M b ) to Ptcl-A/B (61.8 M b ) . Some mice were heterozygous at Ptcl-A/B due to recombination i n the differential segment while being brought to homozygosity and were used for exencephaly data due to timing constraints. The L M / B c chromosomal segment i n 13L/Rec-Line 7 spanned at least 35.8 M b (25 c M ) from D13MU3 (19.8 M b ) to D13MU13 (55.6 M b ) . The next available marker distal of D13MU13, Ntrk2-C/D (at 57.9 M b ) , was back to the S E L H / B c background. Typing at D13MU78 (115.8 M b ) showed that the homozygous S E L H / B c background extended down to the end o f the chromosome.  Exencephaly frequencies from the recombinant congenic lines A s demonstrated i n Table 6.1, 13L/Rec-Line 1 produced approximately 18% exencephaly i n 13 litters, an " S E L H / B c - l i k e " exencephaly frequency. This exencephaly frequency suggests that Exenl was integrated back into this line and that the L M / B c segment, that spans at least 20.8 M b , excludes the Exenl locus. From this line, we can deduce that the Exenl locus is either proximal or distal o f the L M / B c segment. 13L/Rec-Line 7 produced approximately 20% exencephaly, an " S E L H / B c - l i k e " exencephaly frequency. This exencephaly frequency, as well, suggested that Exenl was integrated back into this line and that the L M / B c segment, that spans at least 35.8 M b from D13MU3 to D13MU13, excluded the Exenl locus and that it is located distal o f Ntrk2-C/D. Due to time constraints and mice that would not breed, only 2 litters were collected for this line and therefore the exencephaly frequency obtained may not reflect  104  Table 6.1: Exencephaly frequencies and post-implantation mortality rates ("moles") i n the recombinant congenic lines on P M D #5015. # litters  # implants  % moles  # embryos  # exens  % Exen  13L/Rec-Line 1  16  164  6.7  153  28  18.3*  13L/Rec-Line 7  2  22  9.1  20  4  20.0*  13L/Rec-Line 6  4  37  0.0  34  3  8.8*  *These exencephaly frequencies are not significantly different from each other (x =1.92; p>0.1). 2  105  the most accurate value. Importantly, though, this line did produce exencephaly i n only two litters, more than would be expected i f Exenl had not been incorporated back into this line, meaning more than would be expected for 1 3 L / S E L H . Surprisingly, 13L/Rec-Line 6 produced approximately 9% exencephaly, higher than the 1 3 L / S E L H congenic line, and possibly lower than the S E L H / B c strain. This exencephaly frequency, although possibly not an " S E L H / B c - l i k e " frequency, suggested that the Exenl alleles from S E L H / B c were incorporated back into this line, as well, and that the L M / B c segment that spans at least 41.5 M b from Nr2fl-A/B to D13MU78 (See Fig. 6.2 on pg. 103) excludes the Exenl locus. The exencephaly frequencies from the three recombinant congenic lines were not significantly different from each other  (xM-92; p> 0.1).  Post-implantation mortality rates and other defects A s demonstrated i n Table 6.1 on pg. 105, 13L/Rec-Line 1 had a 6.7% postimplantation mortality rate, 13L/Rec-Line 7 had a 9.0% post-implantation mortality rate, and 13L/Rec-Line 6 had a 0.0% post-implantation mortality rate. The 13L/Rec-Lines 6 and 7 both had small sample sizes due to time constraints. N o other defects were observed i n this line.  Discussion The observation that all the recombinant congenic lines produced exencephaly was surprising. The first available lines to obtain exencephaly frequencies for were 13L/Rec-Lines 1 and 7. Both o f them produced " S E L H / B c - l i k e " frequencies suggesting that the Exenl alleles from S E L H / B c were integrated back into these lines, so that the  106  recombinant congenic line mice were basically like S E L H / B c mice, having all the Exen alleles (Exenl, Exen2, ExenS, and Exen4) from S E L H / B c . 13L/Rec-Line 1 showed that the Exenl locus was located either proximal o f FdgS-C/D at 47.8 M b (31 c M ) or distal o f Nr2fl-A/B at 74.2 M b , since the L M / B c "island" is located between S E L H / B c background on either side (See Figure 6.2 on pg. 103). The use o f the 13L/Rec-Line 7 line was to aid i n figuring out whether the Exenl locus was located proximal or distal o f the L M / B c segment i n 13L/Rec-Line 1. The 13L/Rec-Line 7's C h r 13 is divided nicely into the top half that is homozygous L M / B c (from D13MU13 and up) and the bottom half that is homozygous S E L H / B c (from Ntrk2-C/D and down). B y comparing the two Chr 13 haplotypes from these recombinant congenic lines that both produce S E L H / B c - l i k e exencephaly frequencies, we were able to deduce that the Exenl locus was distal o f Nr2fl-A/B (See Fig. 6.2 on pg. 103). Taking the exencephaly frequencies from the other two recombinant lines into account, it was predicted that 13L/Rec-Line 6 would have a very low exencephaly frequency like 1 3 L / S E L H . This was based on the haplotype data from the recombinant congenic lines, as 13L/Rec-Line 6's haplotype for Chr 13 is the reverse o f 13L/Rec-Line 7 (See Fig. 6.2 on pg. 103). Given that 13L/Rec-Line 6 is homozygous L M / B c throughout the region where 13L/Rec-Lines 1 and 7 predicted the Exenl locus to be  s located (Exenl is integrated back into these lines), the low exencephaly frequency would have confirmed the refined location o f the Exenl locus. Most surprisingly, though, the 13L/Rec-Line 6 recombinant congenic line did produce exencephaly, approximately 9%. This frequency was much higher than the exencephaly frequency produced from the 1 3 L / S E L H congenic line, but possibly not as  107  high as S E L H / B c . Its important to take into account the sample size for 13L/Rec-Line 6, though, as only 4 litters were collected. Assuming that this value was somewhat accurate, this suggested that the Exenl alleles from S E L H / B c were integrated back into this line, as well. The dilemma, though, is that the three recombinant congenic lines do not share a region that is homozygous S E L H / B c (See Fig. 6.2 on pg. 103) that the Exenl locus could be located in. One possible explanation could be that there are actually two genes on Chr 13 that contribute to the risk o f exencephaly i n S E L H / B c . If the exencephaly frequency from 13L/Rec-Line 6 reflects an accurate value for that line (around 9%), perhaps this locus contributes less risk o f exencephaly than the distal Exenl locus. Another possible explanation could be that the U C S C contig used to make these recombinant congenic lines is wrong and that there is actually a region that all three recombinant congenic lines share, but because the markers are mis-mapped we would not know. Although all three recombinant congenic lines have significant levels o f exencephaly, the true values for 2 of the lines, 13L/Rec-Lines 6 and 7, is not known due to the small sample size and it remains possible that the lines differ from each other in rate o f exencephaly. To conclude, 13L/Rec-Lines 1 and 7 suggested that the Exenl locus is below Nr2fl-A/B. The finding that 13L/Rec-Line 6 had a fairly high exencephaly frequency, as well, suggested that there could perhaps be another gene that contributes to the risk o f exencephaly i n S E L H / B c , possibly above Fgd3-C/D when compared with the Chr 13 haplotype from 13L/Rec-Line 1. This question unfortunately could not be resolved by this thesis work and future studies are necessary to determine why all three recombinant congenic lines have a significant level o f exencephaly when they appear to not share a homozygous S E L H / B c region.  108  C h a p t e r 7: N e u r a l tube closure patterns in the congenic lines Introduction  Neural tube closure in mice occurs on days 8 and 9 of gestation. This process is crucial for the survival of a neonate as it is the precursor to the brain and spinal cord. If the cranial neural tube fails to close, exencephaly results (the equivalent to human anencephaly). Previous work has shown that in many various mouse strains cranial neural tube closure initiates at four separate sites, referred to as "Closures 1-4" (See Fig. 7.1), demonstrating "classic" normal neural tube closure. Cranial neural tube closure has been observed initiating at multiple sites in human embryos as well, in a rare study in Japan (Nakatsu et al., 2000). Closures 1, 3, and 4 seem to be invariant among mouse strains, whereas the position of Closure 2 is polymorphic (Juriloff et al., 1991). One particular study showed that the Closure 2 site varied between mouse strains L M / B c , SWV/Bc, and ICR/Be. In L M / B c , Closure 2 initiated at the boundary between the prosencephalon (forebrain) and mesencephalon (midbrain) folds; in SWV/Bc, it initiated more rostrally in the prosencephalon folds and; in ICR/Be it initiated more caudally in the mesencephalon folds. This genetically determined normal variation in the initiation site of Closure 2 might cause strain differences in liability to exencephaly (Juriloff et al., 1991; Fleming and Copp, 2000). During neural tube closure, embryos are synchronously undergoing another fundamental process, the development of somites. Somites are paired segmented "blocks" of mesoderm that develop along side the neural tube (Copp et al., 2003). They are extremely important in organizing the segmental pattern of vertebrate embryos and determining the migration paths of neural crest cells (Gilbert, 2000). They give rise to the  109  Figure 7.1: Diagrammatic side-view representation of the pattern of cranial neural tube closure in normal mouse strains. The numbered black dots indicate the regions where Closures 1, 2, 3, and 4 initiate. Dotted arrows show the direction that the closure spreads and how far it extends from the particular closure site. This figure was modified from a figure in Gunn et al. (1995).  110  vertebral column and other tissues including voluntary muscle, bone, connective tissue, and the dermal layers o f the skin. Somite pairs first appear i n the anterior portion o f the trunk when the neural folds are becoming evident and the formation o f new distinct somite pairs progresses caudally at regular intervals. Being visible on the dorsal side o f an embryo under a microscope, the number o f somite pairs ("somites") present is usually the best reflection o f the overall development o f an embryo and can be used as a measure o f developmental state. During neural tube closure, the neural folds generally elevate and initiate contact at approximate somite stages. The initiation o f neural tube closure (Closure 1) has been shown to require convergent extension (Reviewed i n Copp et al., 2003). Keller at al. states that "convergent extension is a process i n which laterally placed cells move toward and are intercalated into the midline narrowing and lengthening the neural plate" (Keller et al., 2000). In mice, Closure 1 initiates at about the 7-somite stage and spreads bidirectionally from this site rostrally into the rhombencephalon (hindbrain), as well as caudally into the spinal region (See Fig. 7.1). Around the 10-14-somite stage, Closure 2 initiates in the vicinity o f the prosencephalon and mesencephalon boundary (Macdonald et al., 1989; Juriloff et al., 1991; Gunn et a l , 1995) and spreads bidirectionally (See Fig. 7.2). Initiation at Closure 2 is attributed to the appropriate timing o f elevation o f the prosencephalon and mesencephalon folds. For proper elevation o f the mesencephalon folds, they must undergo a morphological change from a convex shape to a concave shape, such that the folds no longer splay outwards but "turn i n " and converge towards the midline (Copp et al., 2003). Closure 3 initiates at the most rostral end o f the prosencephalon folds around the same time Closure 2 initiates, and spreads caudally to  111  Figure 7.2: Scanning electron micrographs o f cranial neural tube closure i n D a y 8/9 embryos. The embryos are being viewed from the front, face-on. Panels b and c show normal neural tube closure. The mesencephalon folds are elevating and Closure 2 is initiating at the prosencephalon/ mesencephalon boundary. Panels e and f show the abnormal mechanism o f neural tube closure i n the S E L H / B c strain. Panel e shows the characteristic "mad-cat" conformation in S E L H / B c due to splayed mesencephalon folds attributed to delayed elevation. In panel f, this S E L H / B c embryo has initiated Closure 3 without Closure 2. M = Mesencephalon folds, P= Prosencephalon folds. These scanning electron micrographs are from Macdonald et al. (1989).  Normal strain  (ICR/Be)  Closure 2  SELH/Bc  Closure 3  11-12 somites  13-14 somites  112  "meet" the closure that is spreading rostrally from Closure 2, therefore closing the "gap" that is called the anterior neuropore ( " A N P " ) . The last closure, Closure 4, is unique in that the neural folds remain separate and closure occurs by growth o f a membrane that eventually covers the rhombencephalon (Geelen and Langman, 1977) (See F i g . 7.1). Cranial neural tube closure is typically completed by the 20-somite stage. The pattern o f normal "classic" mouse neural tube closure discussed above has not been observed i n the S E L H / B c mouse strain. A n early developmental study found that S E L H / B c embryos lacked the initial contact at the Closure 2 region and, instead, Closure 3 was the initiation point for cranial closure (Macdonald et al., 1989). The absence o f Closure 2 is thought to be due to the delayed elevation o f the mesencephalon neural folds, and the defect is present as early as the 3-5-somite stage (Gunn et al., 1993). Under the multifactorial threshold model, timing o f mesencephalon neural fold elevation is considered to be the liability trait. If the timing o f elevation o f the mesencephalon neural folds i n S E L H / B c embryos is too late, then the embryo w i l l become exencephalic. Figure 7.2 shows the characteristic splayed mesencephalon folds that are seen i n the S E L H / B c strain. S E L H / B c embryos close their neural tubes by extension o f Closure 3, seen earliest at the 15-somite stage in the previous developmental study, zipping all the way around to meet Closure 4 i n the rhombencephalon (See Fig. 7.2 on previous page) (Macdonald et al., 1989). A l l S E L H / B c embryos have this abnormal mechanism o f neural tube closure and most embryos complete it successfully and live healthy lives. However, in approximately 10-30% o f S E L H / B c embryos (depending on maternal diet) the mesencephalon folds remain unelevated or not elevated enough and subsequently they become exencephalic. For this study, the pattern o f neural tube closure is again revisited  113  in SELH/Bc and L M / B c , a strain with the normal classic pattern of neural tube closure. In addition, this study is the first formal examination of the neural tube closure patterns in the congenic lines, 13L/SELH, 5L/SELH, 13S/LM, and 5S/LM (See Chapter 5). Chapters 5 and 6 showed how congenic lines were useful in attempting to refine the mapping of the Exen loci, as well as testing the multifactorial threshold model. This chapter focuses on the individual developmental effects of each Exen locus in the process of cranial neural fold elevation in early embryos in the congenic lines. When the congenic lines were being created, it was hypothesized, under the multifactorial threshold model, that the threshold or liability scale is timing of mesencephalic fold elevation and that each Exen locus (genetic factor) contributed to the delay of mesencephalic fold elevation in SELH/Bc. Given that exencephaly is an additive trait, one gene could be acting in the neuroectoderm, one in the mesenchyme, and so forth, each contributing its own delay in mesencephalon fold elevation. Furthermore, the individual effects from the Exen loci would be detectable in the degree of developmental delay of the mesencephalon folds. Introduction of the SELH/Bc Exen alleles into the L M / B c background (13S/LM and 5S/LM) would delay mesencephalon fold elevation relative to L M / B c but would be less delayed than in SELH/Bc. Conversely, introduction of the normal L M / B c Exen alleles into the SELH/Bc background (13L/SELH and 5L/SELH) would cause less delay of mesencephalon fold elevation than SELH/Bc but not as accelerated as L M / B c . In addition to examining the congenic lines for the degree of developmental delay of the cranial neural folds, the occurrence, or lack there of, of Closure 2 was examined in the congenic lines.  114  The neural tube closure patterns i n the congenic lines w i l l give insight into the individual developmental effects o f the Exen loci on cranial neural tube closure and w i l l investigate whether mesencephalon fold elevation could i n fact be considered a continuous liability trait under the multifactorial threshold model. This chapter reports on the first formal examination o f the process o f cranial neural tube closure i n the 1 3 S / L M , 5 S / L M , 1 3 L / S E L H , and 5 L / S E L H congenic lines.  Materials and Methods Mice and breeding design A l l mice originated from and were maintained i n our animal unit i n the Department o f M e d i c a l Genetics at the University o f British Columbia under standard conditions previously described (See General Materials and Methods). A l l breeding colonies that produced adult mice used i n this study were maintained on Purina Laboratory Rodent Diet #5001 ( " P L R D #5001"). During these experiments, S E L H / B c was at F - F 8 , L M / B c was at F - F , 1 3 L / S E L H was at N5F7-N5F12, 5 L / S E L H was at 4  4 6  8 4  8 6  N F - N F , 1 3 S / L M was at N F - N F i o , and 5 S / L M was at N7F4-N7F7. 8  5  8  9  6  7  6  Timed pregnancies were obtained by placing 2 (sometimes 1 or 3) nulliparous females in cages with single males at approximately 3 p.m. The diet o f the mice was switched to Purina Mouse Diet #5015 ( " P M D #5015") upon introduction o f the males and they were fed P M D #5015 until autopsy. Females were checked for vaginal plugs by 10:00 a.m. each morning up to a week after introduction o f the male. Those that had plugs were separated into new cages and kept there until sacrifice. Ovulation generally occurs in relation to the midpoint o f the dark cycle i n mice (Bronson et al., 1966) and  115  therefore midnight of the day before a plug was observed was designated day 0 (DO/Oh), retrospectively. Plugs were generally observed on day 0 hour 9 (D0/9h) of gestation. Females were sacrificed, by carbon dioxide, on day 8 hour 18 (D8/18h, 6:00 p.m.) to D9/10h (10:00 a.m.), the majority between D8/21h and D9/0h (See Table 7.2 on pg. 149).  Experimental strategy The neural tube closure pattern in each major congenic line, along with the parental strains, was observed. Neural tube closure patterns were observed on mice fed the high-risk diet, P M D #5015, to maximize the defect, thought to be delay of the mesencephalic folds. This is based on the increased exencephaly frequency observed in the SELH/Bc strain. Collection of the embryos was done in two sets aimed to be around the time Closure 2 was initiating, when the embryos were around the 12-14-somite stage. Given that this was the first examination of the congenic line embryos, we did not know what the neural tube closure patterns would look like. Therefore, the embryos were scored and data collected in an open-ended way, allowing the data to tell us things other than answering a hypothesis. This method was used in the first formal examination of the neural tube closure pattern in SELH/Bc and proved effective (Macdonald et al., 1989). After examining and scoring the embryos from the parental strains, L M / B c and SELH/Bc, the classification system from Macdonald et al. (1989) was modified to include sub-stages for a specific stage which will be further discussed below. The rate of neural tube closure, as reflected by the average somite count of embryos in the specific stages, was compared between L M / B c and SELH/Bc. In addition  116  the rates were compared between L M / B c and the congenic lines on the L M / B c background ( 1 3 S / L M and 5 S / L M , respectively), and between S E L H / B c and the congenic lines on the S E L H / B c background ( 1 3 L / S E L H and 5 L / S E L H , respectively).  Collection  of Day 8/9  embryos  U p o n autopsy, the uterus was removed and pinned to a black wax substrate, immersed i n physiological saline (0.85% N a C l ) solution, and cut open to reveal the conceptuses. Embryos were immediately collected, intact i n their deciduas, and placed into Bouin's fixative. They were then transferred into 70% ethanol after at least 24 hours in fixative. Each litter was stored i n a separate jar. U p o n scoring, the embryos were dissected out o f their deciduas and chorions (membranes) under 70% ethanol, given an ID number, and stored individually i n cryovials i n 70% ethanol. The number o f embryos collected from each major congenic line and parental strain ranged from 86 to 141.  The classification  system  The system o f classification used here is based from the M a c D o n a l d et al. study (1989). It was developed from examination o f I C R / B e (normal), S W V / B c (normal), and S E L H / B c embryos at various phases o f cranial neural tube formation. The stages are as follows: 1) "Folds evident" - The anterior neural folds are visible as symmetrical thickened bulges i n the anterior half o f the embryo.  117  2) "Folds bent" - T h e anterior neural folds have enlarged and bent ventrally, creating what is called the cranial flexure. Rostral to the flexure is the presumptive prosencephalon and directly caudal to it is the presumptive mesencephalon. 3) "Prosencephalon folds beginning to elevate" - The prosencephalon folds have begun elevation and generally, the mesencephalon folds are still splayed. 4) "Prosencephalon folds completing elevation" - The prosencephalon folds are completely elevated and coming into apposition with each other at the mid-line. 5) "Initial contact at Closure 2 " - The neural folds have initiated contact (Closure 2) in the vicinity o f the mesencephalon/prosencephalon boundary. 6) "Closure 3 begun without Closure 2 " - Closure 3 has obviously begun without Closure 2. This stage is unique to S E L H / B c embryos, where Closure 3 has spread caudally a notable distance through the prosencephalon (Closure 2 and Closure 3 initiate concurrently). 7) " O n l y prosencephalon fused" - self-explanatory. This is the subsequent stage for Stage 6, when Closure 3 has "zipped" the length o f the prosencephalon. 8) "Fused to mid-mesencephalon with A N P open" - A later stage o f Closure 2. Closure 2 has spread caudally and rostrally but has not met Closure 3 yet, leaving an open region o f the rostral prosencephalon, the anterior neuropore ( A N P ) . . 9) "Fused to mid-mesencephalon with A N P closed" - The prosencephalon up to the mid-mesencephalon is fused. This is due either to the caudal spreading o f Closure 3 towards Closure 4 from the rhombencephalon like i n S E L H / B c , or spreading from the site o f Closure 2 as w e l l as from Closure 3, like i n normal strains.  118  10) "Fused to apex with A N P open" - Closure 2 has spread caudally to the top o f the head (apex) and rostrally but has not met Closure 3 yet. 11) "Fused to apex with A N P closed" - The prosencephalon folds have fused as well as the mesencephalon folds through to the top o f the head (apex) so that fusion o f the cranial neural tube i n complete from its most rostral aspect to the apex. 12) "Fused to rhombencephalon" - The prosencephalon and mesencephalon folds have fused and only the rhombencephalon remains open. 13) "Fused completely" - The entire cranial neural tube is closed.  When examining and scoring the parental strains, there were apparent differences between the morphology o f the cranial neural folds, especially the mesencephalon folds, between L M / B c and S E L H / B c i n Stage 4, when the prosencephalon folds were completing elevation. Previously, the mesencephalon folds were not taken into account at this stage, only the prosencephalon folds, so it was necessary to modify this classification system so that the embryos were more accurately scored based on the mesencephalon and prosencephalon fold elevation. Therefore the classification system was modified to include 7 sub-stages o f Stage 4 and they are as follows (See Fig. 7.3):  4 A ' ) "Prosencephalon fold gap wide, mesencephalon folds flat" - The gap between the apposing (or almost apposing) prosencephalon folds is relatively wide and the  119  Figure 7.3: Diagrammatic representations of the sub-stages of Stage 4 and the novel conformation that was observed in substage 4B' in the congenic line embryos. These sub-stages are based on LM/Bc and SELH/Bc embryos when the prosencephalon folds were completing elevation (Stage 4) taking into account mesencephalon fold elevation as well as prosencephalon fold elevation. Tangent lines were imagined on the inside surface of mesencephalon folds and the angle at which they crossed were used to determine the amount of elevation of the mesencephalon folds. Brief descriptions of the substages can be found in Materials and Methods. This figure was produced by Drs. Muriel Harris and Diana Juriloff.  mesencephalon folds are still splayed flat. The angle o f the mesencephalon folds is approximately 140°-180° looking "face-on" at the embryo. 4 A ) "Prosencephalon fold gap wide, mesencephalon folds starting to elevate" - The gap between the apposing (or almost apposing) prosencephalon folds is relatively wide and the mesencephalon folds have started elevating. The angle o f the mesencephalon folds (looking face-on at the embryo) is approximately 110°-130°. 4 B ' ) "Prosencephalon folds are parallel or "pear-shaped," small to medium gap, mesencephalon folds > 9 0 ° " - The prosencephalon folds have completed elevation and they are relatively close to each other (small to medium gap). They are either parallel to each other or "pear-shaped," where the caudal part o f the prosencephalon folds is closer together than the rostral part o f the prosencephalon folds. The mesencephalon folds have started elevating, but looking face-on they are >90°. 4B) "Prosencephalon folds are parallel or "pear-shaped," small to medium gap, mesencephalon folds < 9 0 ° " - The conformation o f the prosencephalon folds are the same as the above sub-stage but angle o f the mesencephalon folds is <90°. 4 C ) "Prosencephalon folds V-shaped, mesencephalon folds < 9 0 ° " - The prosencephalon folds have a small to wide gap between them and are V-shaped, where the most rostral part o f the prosencephalon folds are very close together and get wider as you move caudally up the prosencephalon folds. The angle o f the mesencephalon folds is <90° looking face-on. 4C) "Prosencephalon folds close and parallel or "pear-shaped," mesencephalon folds < 9 0 ° " - The prosencephalon folds are apposing each other and the gap between them is  121  close. They are either parallel to each other or "pear-shaped." The angle o f the mesencephalon folds is <90° looking face-on. 4D) "Verging upon Closure 2, mesencephalon folds < 9 0 ° " - The neural folds are so close together at the Closure 2 site that they are certainly going to initiate contact. The angle o f the mesencephalon folds is <90°.  Scoring of embryos Embryos were scored under a Zeiss dissection microscope at 40x for somite count and neural tube development. Stage o f cranial neural fold elevation and/or closure for each individual embryo was scored, as described above, and the embryo was put into a stage, and sub-stage i f in Stage 4. Graphs were used to show the distribution o f embryos at various somite counts among the stages o f neural tube closure (See Figs. 7.4-7.13). For scoring, an individual embryo was put into a small petri dish containing a black wax substrate that had a tiny divot that the embryo could "sink into," so that it could not wander, and immersed i n 70% ethanol. The somite pairs were counted to determine overall developmental age. Using forceps, the embryo was propped so that it was looked at "face-on," meaning that I was looking directly at the prosencephalon folds, straight into them, and could determine the relative distance they were from each other. In addition, I could determine the angle o f mesencephalic folds, therefore more or less quantifying how much they had elevated. This was done by imagining tangent lines on the surface o f the neural folds and estimating the angle they made when they crossed each other (See F i g . 7.3 on pg. 118).  122  Data  Analysis  The mean somite counts for embryos i n specific stages were compared against each other using graphs. First the parental strains were compared against each other, then the congenic lines to their respective strain backgrounds.  Results Patterns  of Neural  Tube Closure  (See Figs. 7.4-7.14)  It is important to note that stages i n the figures that represent the patterns o f neural tube closure i n L M / B c , S E L H / B c , and the congenic lines are not i n sequential order for any given embryo to go through. While the early and late stages can be considered so, Stages 5-10 are discontinuous, as well as some o f the sub-stages o f Stage 4. For example, for any given line that shows embryos going through Stages 5 ("Initial contact at Closure 2") and 6 ("Closure 3 begun without Closure 2"), this means that there are some embryos that do Closure 2, as well as there are some embryos that do Closure 3 without Closure 2 i n the same line. This does not mean an embryo progresses from Stage 5 to Stage 6, as this is impossible. When looking at the figures, it is important to note the sample sizes i n each stage, as well, to know how much emphasis to put on it because embryos can vary i n somite count at any given stage.  LM/Bc  and SELH/Bc  (See Figs 7.4 and 7.5)  Neural tube closure i n L M / B c and S E L H / B c followed the patterns previously described i n M a c D o n a l d et al. and Juriloff et al. (1989 and 1991, respectively). L M / B c  123  and S E L H / B c embryos went through the early stages o f neural tube development (Stages 1,2, 3) at approximately the same somite count (See Fig. 7.4). A notable divergence began to appear at Stage 4, when the prosencephalon folds were completing elevation. L M / B c embryos had a mean somite count o f 11.5, whereas S E L H / B c embryos had a mean somite count o f 13.2, approximately 2 somites difference. L o o k i n g at Stage 4 more closely (See F i g . 7.5 on pg. 126), the difference between L M / B c and S E L H / B c was quite noticeable. Both strains had embryos that went through 4 A ' and 4 A , but S E L H / B c embryos were approximately 1.5 somites older than the L M / B c embryos going through these sub-stages. L M / B c proceeded quickly to sub-stage 4 B ("Prosencephalon folds parallel or "pear-shaped" and small to medium gap, mesencephalon folds < 90°) around the 11-12-somite stage, then to 4 C ("Prosencephalon folds close and parallel or "pearshaped," mesencephalon folds < 90°) around the 12-somite stage, and finally to 4 D ("Verging upon Closure 2, mesencephalon folds < 90°), the only " L M / B c - specific" substage, around the 12-13-somite stage. S E L H / B c embryos, on the other hand, went through all the sub-stages except for 4D, and went through two " S E L H / B c - s p e c i f i c " sub-stages that L M / B c did not go through, 4 B ' and 4 C (See F i g . 7.5 on pg. 126). This is not surprising given the morphologies o f these two sub-stages are characterized and caused by delayed mesencephalon fold elevation. Some S E L H / B c embryos go through 4 B ' at about 13somite stage, some go through 4 B at about the 14-15-somite stage (approximately 3 somites older than L M / B c at this stage), and some S E L H / B c embryos go through 4 C at about the 15-16-somite stage. A t 4 C , the "oldest" sub-stage L M / B c and S E L H / B c share, the S E L H / B c embryos that go through it do so at around the 15-16-somite stage,  124  13. Fused completely  12. Fused to rhombencephalon  11. Fused to apex with A N P closed 10. Fused to apex with A N P open 9 Fused to mid mesencephalon with A N P closed  Figure 7.4: Comparison between L M / B c and S E L H / B c mean number o f somite pairs present at different stages o f closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. S E L H / B c embryos were not observed in Stages 11 and 13 in this study, but they were observed in these stages in the previous developmental study (MacDonald et al., 1989). In order to make this figure more accurate and complete, they were incorporated and they are indicated by the smaller red circles. In addition, S E L H / B c had two outliers indicated by the stars.  ct)  8 Fused to mid mesencephalon with A N P open 7. Only prosencephalon fused  .4.  6 Closure 3 begun without Closure 2  (3)  5. Initial contact at Closure 2 4 Prosencephalon folds completing elevation 3 Prosencephalon folds beginning to elevate  SELH/Bc  2. Folds bent  LM/Bc  1. Folds evident  (12)  Figure 7.5: Comparison between L M / B c and S E L H / B c mean number o f somite pairs present at the sub-stages for Stage 4. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size.  4D. Verging upon Closure 2, mesencephalon folds < 90°  4C. Prosencephalon folds close and parallel or "pear shaped''. mesencephalon folds <90° 4 C . Prosencephalon folds "V-shaped." mesencephalon folds < 90°  4B Prosencephalon folds parallel or "pear-shaped" and small to medium gap. mesencephalon folds < 90° 4 B \ Prosencephalon folds parallel or "pear-shaped" and small to medium gap, mesencephalon folds > 90° 4A. Prosencephalon folds gap wide, mesencephalon folds starting to elevate  SELH/Bc 4 A ' . Prosencephalon folds gap wide, mesencephalon folds flat  LM/Bc (31)  H o  H  1  10  11  12  13  -I  h  14  15  H 16  17  1  h  18  19  20  21  -I  22  1  23  approximately 4 somites difference from L M / B c . This striking difference demonstrates how far behind S E L H / B c is from L M / B c i n regards to mesencephalon fold elevation. In addition, S E L H / B c and L M / B c show signs that they proceed through closure i n a different way past this point as L M / B c goes through 4 D and S E L H / B c does not. A t this point S E L H / B c and L M / B c proceeded through closure i n a different way, as previously documented (Macdonald et al., 1989). B y the time L M / B c initiated Closure 2 (Stage 5) (See F i g . 7.4 on pg. 125), at around the 13-somite stage, S E L H / B c was still going through Stage 4. Subsequently, L M / B c proceeded through the late closure stages between 14 and 19 somites, and their neural tubes were closed on average by the 19somite stage (Figure 7.4 on pg. 125). In contrast, S E L H / B c initiated Closure 3, without Closure 2 (Stage 6), at around the 17-somite stage and this closure progressed caudally until it met Closure 4. A few embryos were observed i n Stages 7, 9, and 12 that supported this type o f closure. G i v e n that virtually all embryos S E L H / B c do not do Closure 2, there is no anterior neuropore ( A N P ) to close and therefore most S E L H / B c embryos do not go through Stages 8 and 10. One embryo, however, was a bit ambiguous and had a tiny slit A N P at the "chin", and therefore was put in Stage 8. In a previous study, one S E L H / B c embryo was also put i n this stage (Macdonald et al., 1989), so perhaps S E L H / B c does go through this stage on a rare occasion. N o S E L H / B c embryos have been observed initiating Closure 2 though (Stage 5), they have only been observed i n Stage 8. Although no embryos were observed with closed neural tubes i n this study, previous work had shown that S E L H / B c embryos close their neural tubes on average by the 22-somite stage, i f they are going to close.  127  In summary, L M / B c embryos close their neural tubes quicker than the S E L H / B c embryos. L M / B c embryos demonstrate classic normal cranial neural tube closure, whereas S E L H / B c demonstrate an abnormal mechanism for neural tube closure.  LM/Bc  vs. 13S/LM  (See Figs.  7.6 and  7.7)  The 1 3 S / L M congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure o f substituting the Exenl  alleles (as w e l l as the  Exen4 alleles) from S E L H / B c into L M / B c . Demonstrated i n Figure 7.6, 1 3 S / L M consistently went through the neural tube closure stages a little later i n developmental age than L M / B c , but not by much. Some embryos appeared to close at the same rate as L M / B c , whereas others were delayed 1-2 somites i n the closure stages. For all the stages, including the sub-stages o f Stage 4, L M / B c embryos go through, some 1 3 S / L M embryos do so as well. In addition, they gained some " S E L H / B c - s p e c i f i c " stages. This first became apparent when some 1 3 S / L M embryos, around the 11-somite stage, were observed i n sub-stage 4 B ' (See Fig. 7.7 on pg. 130). Furthermore, the majority o f embryos in this sub-stage demonstrated a new conformation o f neural tube closure that paired a more advanced prosencephalic fold elevation with a less advanced mesencephalic fold elevation stage (See Fig. 7.8(D) on pg. 131), not observed i n either L M / B c or S E L H / B c (This conformation was not particular to this line though, as it was observed in all the congenic lines). In addition, one embryo was observed i n 4 C , the other " S E L H / B c - s p e c i f i c " sub-stage. The 1 3 S / L M line maintained Closure 2 i n some embryos (Stage 5), but initiated contact approximately 1 somite later than L M / B c . Interestingly, some embryos were  128  13. Fused completely 12. Fused to rhombencephalon 11 Fused to apex with ANP closed 10 Fused to apex with ANP open 9. Fused lo mid mesencephalon with ANP closed 8 Fused to mid mesencephalon with ANP open 7. Only prosencephalon fused 6. Closure 3 begun without Closure 2 5. Initial contact at Closure 2 4. Prosencephalon folds completing elevation 3. Prosencephalon folds beginning to elevate  to  2 Folds bent  +  1. Folds evident  T 2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  4D Verging upon Closure 2, mesencephalon folds < 90°  4C. Prosencephalon folds close and parallel or "pear shaped", mesencephalon folds < 90°  Figure 7.7: Comparison between L M / B c and 1 3 S / L M mean number o f somite pairs present at the sub-stages for Stage 4. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size.  4 C . Prosencephalon folds "V-shaped."' mesencephalon folds < 90°  4B. Prosencephalon "" folds parallel or "pear-shaped'' and small to medium gap. mesencephalon folds < 90° 4B\ Prosencephalon -folds parallel or "pear-shaped" and small to medium gap. mesencephalon folds >90° 4A. Prosencephalon - folds gap wide, mesencephalon folds starting to elevate  O  4A'. Prosencephalon -p folds gap wide, mesencephalon folds flat  13S/LM LM/Bc H II  1 12  1 13  1 14  1 15  1 16  1 17  118  +  19  20  +  21  +  22  23  Figure 7.8: Light microscope photographs of congenic line embryos in certain stages and sub-stages. Photographs A and B show two different genetically identical 1 3 S / L M embryos; the embryo in A has initiated Closure 2, whereas the embryo in B has started Closure 3 without Closure 2. Photograph C shows an embryo from 5 S / L M in sub-stage 4D, which is verging upon Closure 2. Photograph D shows an embryo from 5 S / L M , in sub-stage 4 B ' , demonstrating the novel conformation that was observed in all the congenic lines. The prosencephalon folds have completely elevated, whereas the mesencephalon folds have just begun elevating. Photograph E shows an embryo from 5 L / S E L H demonstrating the rostral Closure 2 in the prosencephalon region that was observed in both the 5 L / S E L H and 1 3 L / S E L H lines. Photographs were taken by Dr. Diana M . Juriloff.  A  B  D  C  E  observed going through Stages 6 and 7, " S E L H / B c - s p e c i f i c " stages. Given that this line produces a considerable amount o f exencephaly though, this is not surprising. In conclusion, 1 3 S / L M tends to suggest that substituting i n the Exenl  (and Exen4)  alleles from S E L H / B c into L M / B c delays mesencephalon fold elevation as they are a little delayed for any given stage and causes some, but not all, embryos to have delayed mesencephalon fold elevation and omit Closure 2.  LM/Bc  vs. 5S/LM  (See Figs.  7.9 and  7.10)  The 5 S / L M congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure o f substituting the Exen2 alleles from S E L H / B c into L M / B c . A s demonstrated i n Figure 7.9, 5 S / L M consistently went through the main neural tube closure stages slightly later than L M / B c . N o clear pattern was evident for the substages o f Stage 4 (See F i g . 7.10 on pg. 134). For all the stages, including the sub-stages o f Stage 4, L M / B c embryos go through (with the exception o f Stage 10), 5 S / L M embryos do so as well. In addition, this line gained some " S E L H / B c - s p e c i f i c " stages. Demonstrated i n Figure 7.9, some 5 S / L M embryos go through sub-stage 4 B ' at around the 11-12-somite stage and, as discussed previously, the majority o f embryos i n this substage demonstrated a new conformation o f neural tube closure (See F i g . 7.8(D) on pg. 131). The other observed " S E L H / B c - specific" stage was Stage 6, where one embryo was observed doing Closure 3 without Closure 2 at the 17-somite stage. In conclusion, 5 S / L M tends to suggest that introducing the Exen2 alleles from S E L H / B c into L M / B c contributes to a little delay o f mesencephalic fold elevation. The majority o f 5 S / L M embryos appeared to maintain Closure 2 and close their neural tubes  132  13 Fused completely 12. Fused to rhombencephalon  Figure 7.9: Comparison between L M / B c and 5 S / L M mean number o f somite pairs present at different stages o f closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size.  11. Fused to apex with ANP closed II) Fused to apex with ANP open 9. Fused to mid mesencephalon with ANP closed 8. Fused to mid mesencephalon with ANP open 7. Only prosencephalon fused 6. Closure 3 begun without Closure 2 5. Initial contact at Closure 2 4 Prosencephalon folds completing elevation 3 Prosencephalon folds beginning to elevate  5S/LM  2 Folds bent  LM/Bc  1. Folds evident 10  12  13  14  I?  16  17  18  19  20  21  22  23  24  25  26  4D. Verging upon Closure 2. mesencephalon folds < 90°  4C. Prosencephalon folds close and parallel or "'pear shaped'', mesencephalon folds < 90°  Figure 7.10: Comparison between L M / B c and 5 S / L M mean number o f somite pairs present at the sub-stages for Stage 4. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size.  4 C . Prosencephalon folds "V-shaped." mesencephalon folds < 90°  (5)  4B Prosencephalon folds parallel or "pear-shaped" and small to medium gap. mesencephalon folds < 90° 4B" Prosencephalon folds parallel or "pear-shaped'' and small to medium gap. mesencephalon folds >90° 4A. Prosencephalon folds gap wide, mesencephalon folds starting to elevate  4A'. Prosencephalon folds gap wide, mesencephalon folds flat  6  7  8  9  10  11  12  13  14  15  16  17  18  19  %  5S/LM  •  LM/Bc  20  21  22  23  like L M / B c , while some embryos omitted Closure 2 and close their neural tubes like SELH/Bc.  SELH/Bc  vs. 13L/SELH  (See Figs.  7.11 and  7.12)  The 1 3 L / S E L H congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure o f substituting the normal Exenl alleles from L M / B c into S E L H / B c . S E L H / B c and 1 3 L / S E L H progress through the first steps of neural tube closure at about the same developmental age (See Fig. 7.11). They began to diverge at stage 4 where 1 3 L / S E L H went through it one somite ahead o f S E L H / B c . In the sub-stages o f stage 4, 1 3 L / S E L H embryos not only went through all the sub-stages S E L H / B c did, but did so at an earlier developmental age then S E L H / B c , about 1-2 somites difference (See Fig. 7.12 on pg. 137). In addition, 1 3 L / S E L H gained stage 4D, an " L M / B c - s p e c i f i c " sub-stage, going through at about the 13-14-somite stage. The effect o f substituting the normal Exenl  alleles from L M / B c into S E L H / B c appeared to  speed up the elevation o f the mesencephalic folds compared to those o f S E L H / B c . Some 1 3 L / S E L H embryos were observed initiating Closure 2, another " L M / B c specific" stage, at around the 14-somite stage. Interestingly, though, contact was initiated more rostrally than the L M / B c Closure 2 site in some embryos i n Stage 5 (See Fig. 7.8(E)). 1 3 L / S E L H continued to show evidence for Closure 2 as more embryos were found i n Stage 8, a stage S E L H / B c does not go through. Only one embryo was observed in Stage 6, starting Closure 3 without Closure 2, at the 17-somite stage and no embryos were observed i n Stage 7. It is likely that neural tube closure is completed earlier i n  135  Figure 7.11: Comparison between S E L H / B c and 1 3 L / S E L H mean number o f somite pairs present at different stages o f closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. S E L H / B c embryos were not observed in Stages 11 and 13 in this study, but they were observed in these stages in the previous developmental study (MacDonald et al., 1989). In order to make this figure more accurate and complete, they were incorporated and they are indicated by the smaller red circles. In addition, S E L H / B c had two outliers indicated by the stars.  13 Fused completely 12. Fused to rhombencephalon 11. Fused to apex with ANP closed 10. Fused to apex with ANP open 9. Fused to mid mesencephalon with ANP closed 8. Fused to mid mesencephalon with ANP open  !(1)  7. Only prosencephalon fused 6 Closure 3 begun without Closure 2 5. Initial contact at Closure 2 4 Prosencephalon folds completing elevation 3. Prosencephalon folds beginning to elevate 2. Folds bent  4-  1 Folds evident  410  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  4D Verging upon Closure 2, mesencephalon folds < 90°  4C. Prosencephalon folds close and parallel or "pear shaped". mesencephalon folds < 90°  (6),  Figure 7.12: Comparison between S E L H / B c and 1 3 L / S E L H mean number o f somite pairs present at the sub-stages for Stage 4. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size.  4 C . Prosencephalon folds "V-shaped." mesencephalon folds < 90°  4B. Prosencephalon folds parallel or "pear-shaped" and small to medium gap. mesencephalon folds < 90° 4B'. Prosencephalon folds parallel or "pear-shaped" and small to medium gap. mesencephalon folds > 90° 4A. Prosencephalon folds gap wide, mesencephalon folds starting to elevate  13L/SELH SELH/Bc  4A\ Prosencephalon folds gap wide, mesencephalon folds flat  io  12  13  14  15  16  17  19  20  21  22  23  1 3 L / S E L H than i n S E L H / B c , although no embryos were observed with completed neural tube closure. In conclusion, 1 3 L / S E L H suggested that substituting the normal Exenl  alleles  from L M / B c into S E L H / B c accelerated mesencephalon fold elevation and restored Closure 2 i n some, but not all embryos. 1 3 L / S E L H gained some stages, but still maintained "SELH/Bc-specific"  SELH/Bc  vs. 5L/SELH  (See Figs.  7.13 and  "LM/Bc-specific"  stages.  7.14)  The 5 L / S E L H congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure o f substituting the normal Exen2 alleles from L M / B c into S E L H / B c . S E L H / B c and 5 L / S E L H proceeded through the first three stages o f neural tube closure at the same developmental age and began to diverge at Stage 4 (See Fig. 7.13). A t Stage 4, 5 L / S E L H embryos were ahead o f S E L H / B c embryos by about 1 somite. W h e n Stage 4 is blown up into its sub-stages (See F i g . 7.14 on pg. 140), 5 L / S E L H was ahead o f S E L H / B c at all the sub-stages by about 1-2 somites and gained 4 D , an " L M / B c - s p e c i f i c " stage, suggesting that introducing the normal Exen2 alleles into S E L H / B c speeds up the elevation o f the mesencephalic folds. A s i n the other congenic lines, some 5 L / S E L H embryos i n 4 B ' displayed the new neural tube closure conformation with closely apposed prosencephalon folds and widely splayed mesencephalon folds (See F i g . 7.8(D) on pg. 131). Figure 7.13 demonstrated that Closure 2 was restored i n some embryos (Stage 5) and this happened around the 15 to 16-somite stage. Interestingly, contact initiated more rostrally than i n L M / B c (as i n 1 3 L / S E L H ) (See Fig. 7.8(E) on pg. 131). Furthermore, one  138  13 Fused completely 12. Fused to rhombencephalon 11. Fused to apex with ANP closed 10. Fused to apex with AN P open 9. Fused to mid mesencephalon with ANP closed  Figure 7.13: Comparison between S E L H / B c and 5 L / S E L H mean number o f somite pairs present at different stages o f closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. S E L H / B c embryos were not observed in Stages 11 and 13 in this study, but they were observed in these stages in the previous developmental study (MacDonald et al., 1989). In order to make this figure more accurate and complete, they were incorporated and they are indicated by the smaller red circles. In addition, S E L H / B c had two outliers indicated by the stars.  8 Fused to mid mesencephalon with ANP open 7. Only prosencephalon fused 6 Closure 3 begun without Closure 2 5 Initial contact at Closure 2 4. Prosencephalon folds completing elevation 3. Prosencephalon folds beginning to elevate  •  5L/SELH  •  SELH/Bc  2. Folds bent  1. Folds evident —I  8  9  10  11  12  13  14  15  16  17  18  19  1  1  1  1  1  1  1  20  21  22  23  24  25  26  4D. Verging upon Closure 2. mesencephalon folds <90°  4C. Prosencephalon folds close and parallel or "pear shaped". mesencephalon folds < 90°  Figure 7.14: Comparison between S E L H / B c and 5 L / S E L H mean number o f somite pairs present at the sub-stages o f Stage 4. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size.  4C\ Prosencephalon folds "V-shaped." mesencephalon folds < 90°  4B, Prosencephalon folds parallel or "pear-shaped" and small to medium gap. mesencephalon folds < 90° 4B\ Prosencephalon folds parallel or "pear-shaped" and small to medium gap. mesencephalon folds >90° 4A Prosencephalon folds gap wide, mesencephalon folds starting to elevate  o  5L/SELH SELH/Bc  4A\ Prosencephalon folds gap wide, mesencephalon folds flat  io  11  12  13  14  15  16  17  IX  19  20  21  22  23  embryo was observed i n Stage 8, providing evidence for Closure 2. Some embryos began Closure 3 without Closure 2 (Stage 6), like S E L H / B c , and this occurred at approximately the same time that S E L H / B c embryos went through Stage 6, at the 17somite stage. 5 L / S E L H embryos appeared to complete closure o f their neural tubes a bit ahead o f S E L H / B c , although their mean somite count o f 22.4 does not give an accurate value as there was one old litter (D9/10) i n this sample where the majority o f embryos had had their neural tubes closed for quite some time. In conclusion, 5 L / S E L H suggested that substituting the normal Exen2 alleles from L M / B c into S E L H / B c accelerated mesencephalon fold elevation and restored Closure 2 in some, but not all embryos. In addition, 5 L / S E L H gained some " L M / B c - s p e c i f i c " stages, but still maintained the " S E L H / B c - s p e c i f i c " stages.  Summary A l l four o f the congenic lines had distinct heterogeneity o f mechanisms o f neural tube closure where some embryos had Closure 2 like L M / B c and some embryos closed without Closure 2 like S E L H / B c . Introducing S E L H / B c Exen genes into L M / B c tended to delay cranial neural tube closure a bit. In addition, Exenl and Exen2 from S E L H / B c appeared to have similar size effects on cranial neural tube closure when introduced into L M / B c . Conversely, inserting normal L M / B c Exen genes into S E L H / B c seemed to accelerate cranial neural tube closure by about 1-2 somites. Here, inserting the normal Exenl into S E L H / B c appeared to have a slightly bigger effect than inserting the normal Exen2 into S E L H / B c . A l l four congenic lines demonstrated a conformation o f neural tube closure that paired a more advanced prosencephalic fold elevation with a less advanced  141  mesencephalic fold elevation stage. This possibly explains the rostrally offset Closure 2 observed i n some embryos i n 1 3 L / S E L H and 5 L / S E L H .  Discussion Given that the multifactorial threshold model explains the risk o f exencephaly i n S E L H / B c , the hypothesis coming into this study was that the continuous variable underlying the threshold scale was timing o f the mesencephalic fold elevation. If elevation was too delayed, beyond the threshold, exencephaly resulted. The congenic lines provided a good opportunity to test this, to observe whether the individual effects from the Exen loci would be detectable i n the degree o f delay o f mesencephalon fold elevation compared to S E L H / B c and L M / B c . H a d there been no detectable delay i n the congenic lines compared to the parental strains, this would have suggested that the threshold mechanism i n exencephaly, the continuous variable along the liability scale, would have been a different kind o f threshold. The neural tube patterns observed from the congenic lines supported the hypothesis that the liability scale for exencephaly is timing o f the mesencephalic fold elevation, that there is quantitative variation i n mesencephalon fold elevation that is influenced by the Exen loci. Introduction o f the S E L H / B c alleles into the L M / B c background generally caused more delay o f the mesencephalon folds compared to L M / B c but not as much as S E L H / B c as measured by the mean somite counts i n the stages that they shared (Figs. 7.6, 7.7, 7.9, and 7.10). Introduction o f the L M / B c alleles into the S E L H / B c background typically accelerated the mesencephalon fold elevation compared to S E L H / B c as measured by the mean somite counts i n the stages that they shared (Figs.  142  7.11, 7.12, 7.13, and 7.14). The observation that Closure 2 occurred i n some embryos and Closure 3 without Closure 2 occurred i n other embryos within the same congenic line supports this hypothesis as well. This was the second finding o f the study, that all the congenic lines had both types of neural tube closure; some embryos closed their neural tubes like S E L H / B c while other embryos closed their neural tubes like L M / B c . This is interesting as it suggests that initiating "Closure 3 without Closure 2 " can be viewed as a threshold trait as it is most likely related to the timing o f elevation o f the mesencephalon folds, as well. If two genetically identical embryos within the same congenic line can have different neural tube closure patterns, it is most probable that environmental, stochastic, and epigenetic factors influence the extent o f neural fold elevation, as well. A question raised regarding the distinct heterogeneity o f mechanisms o f neural tube closure is whether the threshold for exencephaly would be the same threshold for "Closure 3 beginning without Closure 2" i n the congenic lines, but this is not the case given that nearly all o f S E L H / B c embryos close their neural tubes with Closure 3 (without Closure 2) and only 2 5 % become exencephalic. M o s t likely they are related i n part to each other i n the congenic lines, but it does not necessarily mean that for embryos closing their neural tubes like S E L H / B c that they w i l l be exencephalic. More support for "Closure 3 without Closure 2 " being considered a threshold trait comes from the observation o f one rare S E L H / B c embryo i n Stage 8, where the A N P is open, providing evidence for Closure 2 occurring. One S E L H / B c embryo was observed in Stage 8 i n the previous developmental study as well (Macdonald et al., 1989). In a normally distributed S E L H / B c population, these rare embryos would fall i n the other  143  extreme (the left tail o f the distribution curve), where the mesencephalon folds elevate fairly fast and Closure 2 is able to initiate. Most interestingly, the Closure 2 that was observed i n the 1 3 L / S E L H and 5 L / S E L H congenic lines was more rostrally-offset than the Closure 2 i n L M / B c , 1 3 S / L M , and 5 S / L M , occurring i n the prosencephalon folds (See Fig. 7.8(E) on pg. 131). Transferring i n the normal Exenl or Exen2 alleles from L M / B c into the S E L H / B c background, respectively, accelerated the mesencephalon fold elevation so that Closure 2 was able to initiate. In addition, transferring i n the Exenl alleles from L M / B c appeared to make the congenic line ( 1 3 L / S E L H ) more "normal," like L M / B c , than transferring in the Exen2 alleles from L M / B c ( 5 L / S E L H ) . In 1 3 L / S E L H embryos undergoing Closure 2, some initiated at the prosencephalon/mesencephalon boundary like i n L M / B c , and some initiated more rostrally-offset, whereas only rostrally-offset Closure 2 was observed i n embryos initiating Closure 2 i n 5 L / S E L H . It would be interesting i f S E L H / B c were to have Closure 2 i f it would be more rostrally-offset as i n the 1 3 L / S E L H and 5 L / S E L H congenic lines. The third finding o f this developmental study was a new conformation o f morphology observed i n the congenic lines that paired a more advanced prosencephalon fold elevation with a less advanced mesencephalon fold elevation (See F i g . 7.8(D) on pg. 131). From the earlier developmental study o f S E L H / B c (Macdonald et al., 1989) it seemed that the completion o f prosencephalon fold elevation was i n part assisted b y mesencephalon fold elevation to bring the caudal end o f the prosencephalon folds into apposition and for initiation o f Closure 2 i n normal strains. This was based on the " V shape" that characterized the gap o f the prosencephalon folds i n some S E L H / B c embryos  144  with very delayed mesencephalon fold elevation, and assuming this "V-shaped" gap was due to the delay o f elevation o f the contiguous mesencephalon folds inhibiting the prosencephalon folds from completing elevation i n the caudal end. The new variation observed i n the congenic lines contradicts this and suggests that prosencephalon fold elevation is (or can be) independent o f mesencephalon fold elevation as the entire prosencephalon region can be completely elevated and the folds be i n apposition with each other, yet the mesencephalon folds are still relatively splayed or just beginning to elevate. One explanation for the new conformation observed i n the congenic lines is that the transferring i n o f other genes into a different strain background has disrupted the genetic buffering, called canalization, o f the recipient genetic buffering system so that there is more variation i n the developmental process o f neural tube closure. The idea, proposed by Waddington (1975), is that "developmental reactions are i n general canalized, meaning that they are adjusted so as to bring about one definite end-result regardless o f minor variations i n conditions during the course o f the reaction." Waddington states that, "the genotype can absorb certain amount o f its own variation without exhibiting any alteration i n development and that development is canalized in the naturally selected animal." This canalization, or genetic buffering, o f the genotype is evidenced most clearly by the constancy o f the w i l d type. In this case, the L M / B c strain would be considered the w i l d type. There appears to be very little variation i n neural tube closure among L M / B c embryos; it is amazingly uniform. The S E L H / B c strain, a mutant strain characterized by an abnormal mechanism for neural tube closure, displays slight variations in neural tube closure patterns, but has managed, under natural selection, to  145  canalize its system so that most o f the embryos close their neural tubes. Described by Waddington, "canalization is a feature o f the system, which is built up by natural selection, ensuring the production o f the normal, optimal type i n the face o f unavoidable hazards o f existence and morphological regulation may fail i f the abnormalities are too great or occur too late i n development" (Waddington, 1975). When this canalization is broken down, there is more variation. This is best evidenced i n mutants o f Drosophila where there is scarcely a mutant which is comparable i n constancy with the w i l d type. In this case, the S E L H / B c and L M / B c genetic buffering systems have been disrupted with the introduction o f the L M / B c and S E L H / B c Exen alleles, respectively, and perhaps the other passenger loci due to linkage. This disruption is seen more clearly in the congenic lines on the L M / B c background, where normally the embryos are very uniform i n the way they close their neural tubes. The introduction o f the Exen alleles from S E L H / B c not only introduced the mutant alleles that delay mesencephalon fold elevation, they disrupted the genetic buffering o f the L M / B c system so that perhaps the prosencephalon folds and mesencephalon folds are not quite so i n sync with each other, causing the conformation that pairs the more advanced prosencephalon fold elevation with the less advanced mesencephalon fold elevation. The genetic buffering system o f S E L H / B c could have been disrupted as well, despite becoming more normal like L M / B c with the introduction o f the normal Exen alleles, causing the new variations that were observed as well. Within the same congenic line, some embryos were observed closing their neural tubes like S E L H / B c , some were observed closing their neural tubes like L M / B c , some were observed with the new conformation, and some were observed with variations i n between. It was amazing how variable the congenic line embryos' neural  146  tube closure patterns were from each other given that their respective parental strains' neural tube closure patterns were homogeneous. Whether the Exen loci have specific roles in neural tube closure was not obvious. Introducing both the Exenl and Exen2 alleles, respectively, into L M / B c appeared to delay mesencephalon fold elevation a bit and introducing the normal Exenl and Exen2 alleles, respectively, into S E L H / B c background appeared to accelerate mesencephalon fold elevation. The nature o f the relationship between the mesencephalon and prosencephalon folds was hard to interpret with the findings o f this study. It is unknown whether they are separate entities with different elevation mechanisms, whether the same genes that cause elevation o f the mesencephalon folds cause elevation o f the prosencephalon folds, or whether the genes have similar roles in both tissues but maybe one gene is more important i n one tissue than the other. It is clear that completion o f mesencephalon fold elevation is dependent on prosencephalon fold elevation, but it is not clear whether completion o f prosencephalon fold elevation is dependent on mesencephalon fold elevation. The new conformation observed i n the congenic line embryos suggests that prosencephalon fold elevation can be independent o f mesencephalon fold elevation. The preliminary generally accepted view o f neural tube closure i n mice was that neural tube closure was a continuous bidirectional process where the neural folds elevated and made contact i n the cervical region and then fusion proceeded continuously in both the caudal and rostral directions, forming the neural tube. M o r e work provided insight that the pattern o f cranial neural tube closure was more intermittent, involving multiple closure sites (Geelen and Langman, 1977). This work suggested the cranial  147  neural tube closure may be even more complex as the prosencephalon and mesencephalon folds are semi-distinct units that have independence from one another (at least the prosencephalon folds from the mesencephalon folds). Closures 1 and 4, i n the cervical and rhombencephalon regions, respectively, have been shown to initiate closure using separate mechanisms other than the typical mechanism characterized by "rolling up o f a flat layer into a tube." Closure 1 requires convergent extension, a process requiring cell polarity so that the laterally placed cells move toward and are intercalated into the midline narrowing so that the neural plate lengthens to reduce the distance between the nascent neural folds allowing them to meet and fuse. Closure 4 is unique i n that the neural folds remain separate and closure occurs by growth o f a membrane that eventually covers the rhombencephalon (Geelen and Langman, 1977). Variations o f the cranial neural tube patterns i n the congenic line embryos suggest that the prosencephalon and mesencephalon folds could have slightly different mechanisms o f closure themselves. These findings suggest that neural tube closure may be even more complex than what is currently thought.  148  Table 7.1: Summary table o f the number o f litters collected i n the parental strains and congenic lines. # o f litters  # of embryos  % scoreable  LM/Bc  11  99  92.9  13S/LM  9  93  93.5  5S/LM  8  86  91.9  SELH/Bc  14  141  95.7  13L/SELH  12  94  97.9  5L/SELH  10  89  97.8  Table 7.2: The number o f litters that were collected at various gestational ages .  LM/Bc 13S/LM  8/18  8/21  8/22  8/23  1  3  2  4  2  1  5S/LM  4  SELH/Bc  5  2  13L/SELH  4  4  5L/SELH  1  1  A  A  4  9/0  3  2  4  1  9/9  9/10  0-20  1  1-27 2-19 1  1  4  Range of somites obtained  1  2 3  2  9/1  A  0-23 3-18  1  1  5-25  1 litter with > 27 somites excluded from further analysis 9/9 = 9 A . M . on day 9 o f gestation  149  Table 7.3: M e a n somite count i n parental strains and congenic lines at all the stages. Number i n parentheses indicates the sample size in each group. Stage descriptions can be found in Materials and Methods. Neural tube closure stage l  2  3  4  5  LM/Bc  0.0  5.7  8.3  11.5  13.0  (1)  (10)  (12)  (31)  (3)  13S/LM  1.0  6.5  9.0  11.8  14.1  17.7,14.5  (1)  (4)  (5)  (35)  (7)  (3)  2.0  7.0  8.3  11.7  14.3  (1)  (1)  (6)  (33)  (3)  2.0  5.6  8.5  13.2  17.0  (2)  (5)  (17)  (99)  (3)  3.5  6.0  8.3  12.1  14.3*  (2)  (5)  (15)  (53)  (4)  6.0  8.4  12.2  15.7*  17.0  16.0  16.0  16.0  18.5  19.2  (10)  (38)  22.4  (8)  (3)  (2)  (1)  (1)  (1)  (6)  (5)  (12)  5S/LM  SELH/Bc 13L/SELH 5L/SELH  6  7  o  9  10  11  12  13  14.2  14.8  14.0  16.2  17.0  19.0  (10)  (4)  (1)  (10)  (5)  (5)  20.4  15.5  15.3  16.0  16.5  17.4  24.7  (5)  (2)  (3)  (2)  (4)  (5)  (12)  17.0  14.8  15.0  16.4  18.0  19.0  (1)  (4)  (3)  (14)  (1)  (1)  17.0  15.0  19.0,17.5  (4)  (1)  (3)  17.0  14.5  15.7  16.0  17.5  (1)  (6)  (3)  (1)  (2)  +  *rostral closure 2 T h e mean somite count after the outlier at 24 somites was taken out.  +  8  22.0  +  (1)  Table 7.4: M e a n somite count i n the parental strains and congenic lines i n the sub-stages o f Stage 4 o f cranial neural tube closure. Numbers in parentheses indicates the sample size i n each group. Stage descriptions can be found i n Materials and Methods. Sub-stages o f stage 4 cranial neural tube closure 4A'  4A  LM/Bc  9.6 (8)  11.0(1)  13S/LM  10.7 (6)  10.8 (5)  5S/LM  9.6 (5)  SELH/Bc  4B'  4B  4C  4C  4D  11.4(5)  —  11.8(5)  12.7(12)  10.9 (7)  12.8 (4)  15.0(1)  12.4 (5)  13.1 (7)  10.3 (3)  11.5 (7)  11.5(2)  —  12.5 (4)  12.8 (12)  11.1 (31)  12.3 (9)  13.2(18)  14.6 (28)  15.6 (8)  15.6 (5)  —  13L/SELH  10.0(11)  11.0 (6)  11.7(10)  12.3(9)  15.0 (2)  13.7 (9)  13.7 (6)  5L/SELH  10.0 (5)  11.0 (3)  12.1 (18)  12.5 (4)  15.0(1)  14.0 (5)  13.5 (2)  —  Chapter 8: Conclusions and General Discussion  Unraveling the cause o f non-syndromic, multifactorial human N T D s has proved to be a demanding task. The methods used today, such as case-control and family-based studies, have yielded only a few significant associations between candidate genes o f interest and risk for N T D s . There is clear potential for genetic heterogeneity o f N T D . There are over 90 single-gene mutations that cause N T D s i n mice, most o f the genes having human homologues (Harris, 2001), suggesting that misregulation o f any o f the human homologues could possibly increase risk for N T D s . The non-Mendelian pattern o f increased recurrence risk with number affected i n a family has pointed to a genetically multifactorial basis o f N T D , meaning that individual N T D cases are due to combinations of risk alleles across several gene loci. Environmental factors that have been shown to influence risk o f N T D s include, but are not limited to, maternal folate supplementation, maternal obesity, maternal hyperthermia (fever), maternal diabetes, maternal drugs such as valproate for epilepsy, and other various nutritional deficiencies (Hall et al., 1988). Given that the S E L H / B c strain has at least 3 genetic factors, plus an environmental factor that influences the exencephaly risk, it is a valuable animal model to study human N T D s . A s the molecular defects leading to the delay o f cranial neural fold elevation i n S E L H / B c embryos are not evident, a future step would be to uncover the cause by identifying the loci that contribute to the risk o f exencephaly i n S E L H / B c (Juriloff et al., 2001). Not only would identifying the loci provide insight into the molecular defects in S E L H / B c , it would provide insight into the type o f genes that are involved i n additivity, as the Exen loci that contribute to the risk o f exencephaly i n S E L H / B c act i n an additive fashion with one Exen locus adding its effect to the other Exen loci to create risk. The  152  first step, mapping the loci, had been done previously and the Exenl, Exen2, and Exen3 loci were mapped to m i d C h r 13, distal Chr 5, and Chr 11, respectively. One o f the goals o f this work was to attempt to refine the mapping o f the Exen loci i n the regions that had been previously established. The EX-5001 F 2 exencephaly panel, collected from mice fed the regular diet P L R D #5001, and congenic lines were used to do so. The E X - 5 0 0 1 study confirmed the locations o f Exenl and Exen2, on Chrs 13 and 5, respectively, and there was a slight suggestion for Exen3 on C h r 11. It was i n the EX-5001 study that proximal C h r 7 was first revealed i n m y studies to potentially containing a locus that contributed to the risk o f exencephaly. This region was revisited in the first place because it approached significance i n the previous F 2 sire genome screen, yet was not supported in the original F2 exencephaly panel (Juriloff et al., 2001). The J H E X - 5 0 1 5 study, collected from mice fed the high-risk diet P M D #5015, was collected, for the most part, to test for a diet effect i n F2 segregants. G i v e n that the same markers were used to genotype the J H E X - 5 0 1 5 F 2 exencephalic embryos across the Exen loci to test for gene-diet interactions, their genotypes confirmed the roles o f the Exenl locus and the new locus on Chr 7 that was named Exen4. There has always been the question, when analyzing the F 2 exencephalic embryo genotype summaries, o f whether the genotype summaries that have the greatest deviation from random segregation, hence the highest x value/lowest /?-value, are o f greater 2  significance pointing to where the Exen locus most likely is located, or the genotype summaries with the least L M / B c homozygotes or alleles i n general (including the heterozygous genotypes). For example, i n the EX-5001 study, the markers i n the top half of Chr 13 gave the most significantly deviated segregation ratios (See Table 4.1 on pg.  153  55) showing notable differences between the numbers o f S E L H / B c homozygotes and heterozygous genotypes, almost double the amount. The markers more distal showed less L M / B c homozygotes, and less S E L H / B c homozygotes as well, with the number o f S E L H / B c homozygotes and heterozygotes being fairly similar, although still very significant. Therefore, according to /^-values, the region between Fgd3-C/D and Ptcl-A/B most likely contained the Exenl locus. This was i n good agreement with the results from the F2 sire screen and original F2 exencephaly panel that indicated that the location o f the Exenl locus was on C h r 13 near D13MU13 (p< 0.001) (Juriloff et al., 2001). The J H E X - 5 0 1 5 panel, on the other hand, tells a slightly different story. The same Chr 13 markers that showed twice the number o f S E L H / B c homozygotes than heterozygotes i n the E X - 5 0 0 1 panel showed equal amounts o f the two genotypes and the least number o f L M / B c homozygotes. The more distal markers showed the most deviated segregation genotype summaries and had the lowest /^-values (See Table 4.2 on pg. 61). Although not too much emphasis can be put on this panel i n regards to mapping because o f the limited sample size, it appeared that this panel suggests that the location o f the Exenl locus is more distal, between D13MU193 and D13MU76. G i v e n that the F 2 exencephaly panels suggest different regions where the Exenl locus is most likely located according to ^-values, this shows they may not be so reliable i n determining the exact location. However, the region most significant i n the J H E X - 5 0 1 5 panel was significant i n the EX-5001 panel as w e l l , just not the most significant region. F 2 exencephaly panels use probability to determine the location o f the Exen loci so that there is always the chance o f false positives. They create a zone o f probability that peaks at the most significant region, determined by the goodness o f fit x test for segregation ratios at 2  154  markers typed on the F 2 exencephalic embryos, that teeters off bidirectionally creating blurry boundaries. F o r this reason, the congenic lines were used to help refine the mapping o f the Exenl and Exen2 loci. The advantage o f the congenic lines is that there are sharp boundaries between the transferred chromosomal segments that contain the Exen loci and the recipient strain background it is transferred into, so that the location o f the Exen locus, or the region it is included in, is more definite. Chapter 5 showed that the transferred C h r 13 segments were made purposely large to ensure the Exenl locus was transferred. Because it was so large, recombinant congenic lines were created, specifically from the 1 3 L / S E L H congenic line, to break up this transferred chromosomal segment into smaller sized intervals. Three recombinant congenic lines were created that had differing haplotypes. A l l three o f them produced higher exencephaly frequencies than 1 3 L / S E L H , two o f them producing " S E L H / B c - l i k e " frequencies (18% and 20%), and one producing around 9%, although these frequencies were not significantly different from each other. This was surprising because the three recombinant congenic lines do not share a region where all three o f them are homozygous for S E L H / B c alleles. One possible explanation is that there are two genes on C h r 13 that contribute to the risk o f exencephaly i n S E L H / B c . Its interesting that the F2 exencephaly panels could indeed support this, as the markers that span the length o f Chr 13 all give segregation ratios that significantly deviate from random. If there were two genes on Chr 13, one would likely be below Nr2fl-A/B and another would likely be above Fgd3-C/D (See Figure 6.2 on pg. 103) The location o f the gene that would be below Nr2fl-A/B is i n better agreement with the J H E X - 5 0 1 5 panel that demonstrates this region to be the most likely region to contain the Exenl locus.  155  Interestingly these studies were both done on mice fed P M D #5015. One possible candidate gene is Dhfr (at 88.5 M b ) , because o f its role i n the folate pathway. Candidate genes for the other locus that is possibly above Fgd3-C/D  include Fgd3 itself, which is a  Rho G E F protein and a member o f the F G D 1 family that has a role i n regulating the actin cytoskeleton and activating the c-Jun N-terminal kinase ( I N K ) signaling cascade to regulate cell growth and differentiation (Pasteris et al., 2000). Interestingly the Jnkl/Jnk2 double mutant has exencephaly (Sabapathy et al., 1999). This region also includes some known genes that cause exencephaly i n mice, though as part o f a lethal syndrome (Juriloff et al., 2001), such as Tcfap2a (Ap2) that increases apoptosis when knocked-out (Schorle et al., 1996), Jumonji  (Jmj) that is a binding protein whose possible mechanism  o f N T D is unknown (Takeuchi et al., 1995), and Msx2 which is a homeodomain transcription factor that is expressed i n the developing craniofacial region (Winograd et al., 1997) (Reviewed i n Harris and Juriloff, 1999; Juriloff and Harris, 2000; Copp et al., 2003). Refining the mapping o f the Exen2 locus on Chr 5, as w e l l as the Exen3 and Exen4 loci, was less o f a priority i n m y studies. The suggestions o f the general region likely to contain the Exen2 locus from the EX-5001 F2 exencephaly panel and the Chr 5 congenic lines ( 5 S / L M and 5 L / S E L H ) agreed with each other. To date, no previously known N T D mutations appear to have been mapped to distal C h r 5, but there are some potential candidate genes around D5MU168  involved i n the actin cytoskeleton that could  have implications for a role o f loss o f actin function i n exencephaly (Juriloff et al., 2001). They are Actb (beta actin) and Racl  (a Rho GTPase). The Exen3 and Exen4 loci require  further confirmation before candidate genes are considered.  156  A s the identities o f the Exen loci remain unresolved, the nature o f the genetic variants o f the Exen loci that contribute to the risk o f exencephaly i n S E L H / B c remains unresolved as well. They could be polymorphisms that are maintained by selection that slightly impede the function o f an Exen protein, loss-of function mutations o f the Exen loci, or perhaps gain-of-function mutations o f the Exen loci. One study on mice carrying a spontaneous mutation, Dancer, that exhibit cleft lip, another multifactorial birth defect, showed that the Dancer locus on Chr 19 revealed an insertion o f a sequence containing the 5' region o f the p23 gene into the first intron o f TbxlO, which caused ectopic expression (gain-of-function mutation) o f a p23-TbxlO chimeric transcript, driven b y the p23 gene promoter, that encoded a protein product identical to the normal variant o f the TbxlO protein (Bush et al., 2004). This insertion was thought to be most likely due to a translocation event. Basically, this gain-of-function mutation caused the TbxlO gene to be expressed i n the developing lip, where normally it is not expressed. It is a good demonstration o f how the normal expression domain o f a gene does not necessarily predict its role i n a defect because gain-of-function mutations have ectopic expression, which is their mechanism o f disrupting development. O f significance as well, this study has shown that it is important to consider all genes i n the candidate gene region and not to limit the genes to those known to be expressed i n the tissue o f interest. Another alternative could be that the Exen loci are disrupted b y insertions o f EtnlJ or M u s D elements that disrupt the function o f the Exen loci. S E L H / B c appears to be more prone to early transposon element (EtnU) insertions than other strains. T w o out o f 19 reported mutations i n mouse cell lines caused by insertion o f E t a elements into genes had been found i n the S E L H / B c mice suggesting that Etn retrotransposition occurs relatively  157  frequently i n this strain (Baust et al., 2003). Baust et al. identified a young, codingcompetent M u s D element i n S E L H / B c that contributed to the majority o f M u s D transcripts suggesting that this element was overexpressed i n S E L H / B c embryos. It is possible a MusD/EtnTJ transcript has inserted into an Exen locus and has disrupted the function o f it. Once the Exen loci have been identified, analysis o f the mutations w i l l help unravel the mystery o f the type o f genetic variants that are involved i n multifactorial birth defects. One other goal o f this work was to determine i f the response observed in the S E L H / B c strain to P M D #5015 was due to a gene-diet interaction. Taken together, the F2 exencephaly segregation studies and diet studies i n congenic lines presented i n this thesis indicated that there is a gene-diet interaction with the Exenl alleles from S E L H / B c . The J H E X - 5 0 1 5 F 2 exencephaly study showed that the Exen2 locus had little or no role i n risk o f exencephaly when mice are fed P M D #5015, as demonstrated by the Mendelian (random) segregation ratios at the selected markers from the F 2 exencephalic embryos. When mice are fed P L R D #5001, the diet the mice are regularly fed, the Exen2 locus appeared to strongly contribute to risk o f exencephaly as shown by the "significantly deviated from random" segregation ratio. The conclusion from these F2 exencephaly panels was that the effect from P M D #5015 replaced the effect o f the Exen2 alleles from S E L H / B c . In the EX-5001 and J H E X - 5 0 1 5 studies, both Exenl and Exen4 appeared to contribute to the risk o f exencephaly on either diet. If only one Exen locus appeared to be involved i n the risk o f exencephaly i n the J H E X - 5 0 1 5 study the explanation would have been a gene-diet interaction with that specific Exen locus so that no other Exen locus was important on P M D #5015.  158  The conclusion from Chapter 4 was that the effect from P M D #5015 superceded the effect o f the Exen2 alleles from S E L H / B c so that it had little or no role i n risk o f exencephaly i n the F 2 segregants. From the congenic line studies (Chapter 5), however, it appeared that the congenic lines with Exenl alleles from S E L H / B c ( 1 3 S / L M and 5 L / S E L H ) responded to diet and the congenic lines without Exenl alleles from S E L H / B c ( 5 S / L M and 1 3 L / S E L H ) did not respond. The 1 3 S / L M congenic line provided the strongest support for the Exen gene-diet interaction because it showed that the diet effect does not require a S E L H / B c mother. This suggested that there was a gene-diet interaction between the Exenl alleles from S E L H / B c and P M D #5015 that increased the exencephaly frequency. Tying this conclusion together with the data from the F2 exencephaly lines, it seems that perhaps the Exen2 locus dropped out o f the risk o f exencephaly because o f the strong gene-diet interaction between the Exenl alleles from S E L H / B c and P M D #5015, not that the effect from Exenl was directly replaced by the effect o f P M D #5015. W h y the Exen4 alleles from S E L H / B c showed that they still contributed to the risk o f exencephaly i n the J H E X - 5 0 1 5 study is uncertain. Perhaps the Exen4 alleles from S E L H / B c are involved i n the diet response to P M D #5015, as well, i f they are present. Uncovering the mechanism o f the diet effect is difficult because the Exenl locus, the gene indicated i n the diet response, has not yet been identified. Identification o f the Exenl locus could possibly help i n identifying the aspect i n P M D #5015 that strongly influences exencephaly frequency i n S E L H / B c . Interestingly, P M D #5015 is not the only diet to increase the risk o f exencephaly i n S E L H / B c (Harris and Juriloff, 2005). B y chemical analysis, P L R D #5001 and P M D #5015 differ i n level o f almost every nutrient  159  and P L R D #5001 contains 3 natural ingredients that are not present i n P M D #5015, as well as other diets shown to increase exencephaly frequency i n the S E L H / B c strain (Harris and Juriloff, 2005). It would be interesting to do the same tests with the F2 exencephaly panels and the congenic lines to see i f the Exenl locus also interacted with those high-risk diets to increase exencephaly frequency. Maternal obesity and maternal diabetes have both been independently shown to increase risk for N T D s , suggesting some opportunities for further investigation o f the gene-diet interaction (Anderson et al., 2005; Becerra et al., 1990) observed i n m y studies. This diet effect, attributed to a gene-diet interaction with the Exenl locus, on N T D frequency may point to new nutritional approaches to prevention o f human N T D s . The last main goal o f this study was to determine the individual effects o f the Exen loci on cranial neural tube closure. The congenic lines showed that elevation o f the mesencephalon folds could indeed be considered a liability trait as varying amounts o f delay o f elevation o f the mesencephalon folds were observed as w e l l as both types o f closure ( L M / B c and S E L H / B c ) i n each congenic line. Though the hypothesis that the liability trait for S E L H / B c is likely timing o f mesencephalon fold elevation was generally supported, new variations o f neural tube closure patterns from the congenic lines suggested that neural tube closure is more complex, maybe involving similar yet different elevation mechanisms for the prosencephalon and mesencephalon folds. These findings show how complex this process can be i n humans who are a heterogeneous m i x o f N T D risk genes. In summary, the work presented here demonstrates the complexity o f the risk o f exencephaly i n S E L H / B c mice and that exencephaly can be considered a multifactorial  160  threshold defect. In fact, this was the first empirical test o f the multifactorial threshold model i n vertebrates. G i v e n that the risk o f N T D s appears to be very complex i n the human population, the S E L H / B c mouse strain has proved to be a valuable animal model, and more investigation into the genetic basis o f exencephaly and the diet effect i n the S E L H / B c strain w i l l provide insights into the mechanisms underlying human multifactorial neural tube defects. Identifying the Exen loci, especially the Exenl locus, w i l l be a next important step to unraveling the molecular defects i n S E L H / B c . The interacting gene pathways that participate in the multifactorial S E L H / B c risk can then be identified which are o f potentially broader significance (Juriloff et al., 2001). Uncovering the molecular role o f the Exen loci i n neural fold elevation hopefully w i l l contribute to the understanding o f neural tube closure and the mechanisms by w h i c h N T D s arise i n humans.  161  References Anderson, J.L., D . K . Waller, M . A . Canfield, G . M . Shaw, M . L . Watkins, and M . M . Werler (2005) Maternal obesity, gestational diabetes, and central nervous system birth defects. Epidemiology, 16:87-92. Baust, C , L . Gagnier, G.J. Baillie, M . J . Harris, D . M . Juriloff, and D . L . Mager (2003) Structure and expression o f mobile EtnTJ retroelements and their codingcompetent M u s D relatives i n the mouse. J Virol., 77:11448-11458. Becerra J.E., M . J . Khoury, J.F. Cordero, and J.D. Erickson (1990) Diabetes mellitus during pregnancy and the risks for specific birth defects: a population-based casecontrol study. Pediatrics, 85:1-9. Boyles, A . L . , P. Hammock, and M . C . Speer (2005) Candidate gene analysis i n human neural tube defects. Am J Med Genet C Semin Med Genet., 135C:9-23. Bronson, F . H . , C P . Dagg, and G . D . Snell (1966) Reproduction. In: Biology o f the laboratory mouse, 2 Edition, E . L . Green (ed.), M c G r a w - H i l l Inc., N e w York, pp. 194-196. n d  Brook, F . A . , J.P. Estibeiro, and A . J . Copp (1994) Female predisposition to cranial neural tube defects is not because o f a difference between the sexes i n the rate o f embryonic growth or development during neurulation. JMed Genet., 31:383-387. Bush, J.O., Y . L a n , and R . Jiang (2004) The cleft lip and palate defects i n Dancer mutant mice result from gain o f function o f the TbxlO gene. PNAS, 101:7022-7027. Chen, Z . and R . R . Behringer (1995) twist is required i n head mesenchyme for cranial neural tube morphogenesis. Genes Dev., 9:686-699. Colas, J.F and G . C . Schoenwolf (2001) Towards a cellular and molecular understanding o f neurulation. DevDyn. 221:117-145. Copp, A . J . , N . D . Greene, and J . N . Murdoch (2003) The genetic basis o f mammalian neurulation. Nat Rev Genet., 4:784-793. Curtin, J.A., E . Quint, V . Tsipouri, R . M . A r k e l l , B . Cattanach, A . J . Copp, D . J . Henderson, N . Spurr, P. Stanier, E . M . Fisher, P . M . Nolan, K . P . Steel, S.D. Brown, I.C. Gray, and J . N . Murdoch (2003) Mutation o f Celsrl disrupts planar polarity o f inner ear hair cells and causes severe neural tube defects i n the mouse. Curr Biol. 13:1129-1133. Darvasi, A . (1998) Experimental strategies for the genetic dissection o f complex traits in animal models. Nat Genet., 18:19-23  162  Deak, K . L . , A . L . Boyles, H.C. Etchevers, E.C. Melvin, D.G. Siegel, F.L. Graham, S.H. Slifer, D.S. Enterline, T . M . George, M . Vekemans, D. McClay, A . G . Bassuk, J.A. Kessler, E. Linney, J.R. Gilbert, M . C . Speer, and N T D Collaborative Group (2005) SNPs in the neural cell adhesion molecule 1 gene (NCAM1) maybe associated with human neural tube defects. Hum Genet., 117:133-142. Echelard, Y . , D.J. Epstein, B. St-Jacques, L., Shen, J. Mohler, J.A. McMahon, and A.P. McMahon (1993) Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell, 75:1417-1430. Falconer, D.S and T.F.C Mackay (1996) Introduction to Quantitative Genetics. Fourth Edition, Pearson Education Ltd, Essex, England, pp. 299-310 and pp. 319. Fleming, A., and A.J. Copp (2000) A genetic risk factor for mouse neural tube defects: defining the embryonic basis. Hum Mol Genet., 9:575-581 Fraser, F.C. and J.J. Nora (1975) Genetics of Man. First Edition, Lea and Febiger, Philadelphia, pp. 133-153. Fraser, F.C (1976) The multifactoriaPthreshold concept - uses and misuses. Teratology, 14:267-280. Geelen, J.A.G., and J. Langman (1979) Ultrastructural observations on closure of the neural tube in the mouse. Anat Embryol., 156:73-88. Geelen, J.A.G., and J. Langman (1977) Closure of the neural tube in the cephalic region of the mouse embryo. Anat Rec., 189:625-640. Gilbert, S.F. (2000) Developmental Biology. 6 Edition. Sinaurer Associates, Inc, Sunderland, M A , online edition. th  Goodrick, L.V., L. Milenkovic, K . M . Higgins, and M.P. Scott (1997) Altered neural cell fates and medulloblastoma in mouse patched mutants. Science, 277:1109-1113. Gunn, T.M., D . M . Juriloff, and M.J. Harris (1992) Further genetic studies of the cause of exencephaly in SELH/Bc mice. Teratology, 45:679-686. Gunn, T.M., D . M . Juriloff, W.Vogl, M.J. Harris, and J.E. Miller (1993) Histological study of the cranial neural folds of mice genetically liable to exencephaly. Teratology, 48:459-471. Gunn, T.M., D . M . Juriloff, and M.J. Harris (1995) Genetically determined absence of an initiation site of cranial neural tube closure is causally related to exencephaly in SELH/Bc mouse embryos. Teratology, 52:101-108.  163  Gunn, T . M . (1995) Genetic and developmental studies o f abnormal neural tube closure i n S E L H / B c mice (Thesis). University o f British Columbia. Gunther, T., M . Struwe, A . A g u z z i , and K . Schughart (1994) Open brain, a new mouse mutant with severe neural tube defects, shows altered gene expression patterns i n the developing spinal cord. Development 120:3119-3130. H a l l , J.G., J . M . Friedman, B . A . Kenna, J. Popkin, M . Jawanda, and W . A r n o l d (1988) Clinical, genetic, and epidemiological factors i n neural tube defects. Am JHum Genet., 43:827-837. Hamblet, N . S . , N . Lijam, P. Ruiz-Lozano, J. Wang, Y . Yang, Z . L u o , L . M e i , K . R . Chien, D.J. Sussman, and A . Wynshaw-Boris (2002) Dishevelled 2 is essential for cardiac outflow tract development, somite segmentation and neural tube closure. Development 129:5827-5838. Harris, B . S . , T. Franz, S. U l l r i c h , S. Cook, R . T . Bronson, and M . T . Davisson (1997) Forebrain overgrowth (fog): a new mutation i n the mouse affecting neural tube development. Teratology, 55:231-240. Harris, M . J . and D . M . Juriloff (1999) Mini-review: Toward understanding mechanisms o f genetic neural tube defects i n mice. Teratology, 60:292-305. Harris, M . J . (2001) W h y are the genes that cause risk o f human neural tube defects so hard to find? Teratology, 63:165-166. Harris, M . J . and D . M . Juriloff (2005) Maternal diet alters exencephaly frequency i n S E L H / B c strain mouse embryos. Birth Defects Res A Clin Mol Teratol., 73:532540. Hofmann, M . , M . J . Harris, D . M . Juriloff, and T. Boehm (1998) Spontaneous mutations i n S E L H / B c mice due to insertions o f early transposons: molecular characterization o f null alleles at the nude and albino loci. Genomics 52:107-109_ Homanics, G . E . , N . Maeda, M . G . Traber, H . J . Kayden, D . B . Dehart, and K . K . Sulik (1995) Exencephaly and hydrocephaly i n mice with targeted modification o f the apolipoprotein B (Apob) gene. Teratology, 51:1-10. Honarpour, N . , S.L. Gilbert, B . T . Lahn, X . D . Wang, and J. Herz (2001) Apaf-1 deficiency and neural tube closure defects are found i n fog mice. Proc. Natl. Acad. Sci. USA, 98:9683-9687. H u i , C C . and A . L . Joyner (1993) A mouse model o f Greig cephalopolysyndactyly syndrome: the extra-toes mutation contains an intragenic deletion o f the GH3 gene. Nat Genet. 3:241-246. J  164  Ikeda, A., S. Ikeda, T. Gridley, P . M . Nishina, and J.K. Naggert (2001) Neural tube defects and neuroepithelial cell death in Tulp3 knockout mice. Hum. Mol. Genet., 10:1325-1334. Juriloff, D . M . , K . B . Macdonald, and M.J. Harris (1989) Genetic analysis of the cause of exencephaly in the SELH/Bc mouse stock. Teratology, 40:395-405. Juriloff, D . M . , M.J. Harris, C. Tom, and K . B . Macdonald (1991) Normal mouse strains differ in the site of initiation of closure of the cranial neural tube. Teratology, 44:225-233. Juriloff, D . M . , M.J. Harris, V . Wong, and J.E. Miller (1992) Studies of a spontaneous lethal mutation at the albino locus in SELH/Bc mice. Genome, 35:342-346. Juriloff, D . M . , S.D. Porter, and M.J. Harris (1994) Three spontaneous mutations at the albino locus in SELH/Bc mice. Genome, 37:190-197. Juriloff, D . M . , M.J. Harris (2000) Mouse models for neural tube closure defects. Hum Mol Genet, 9:993-1000. Juriloff, D . M . , T . M . Gunn, M.J. Harris, D.G. Mah, M . K . Wu, and S.L. Dewell (2001) Multifactorial genetics of exencephaly in SELH/Bc mice. Teratology, 64:189200. Juriloff, D . M . , M.J. Harris, and D.G. Mah (2005) The open-eyelid mutation, lidgapGates, is an eight-exon deletion in the mouse Map3kl gene. Genomics 85:139142. Keller, R., L. Davidson, A . Edlund, T. Elul, M . Ezin, D. Shook, and P. Skoglund (2000) Mechanisms of convergence and extension by cell intercalculation. Phil Trans Royal Soc London B: Bio Sci., 355:897-922 Kibar,Z., K.J. Vogan, N . Groulx, M.J. Justice, D.A. Underhill, and P. Gros (2001) Ltap, a mammalian homolog of Drosophila Strabismus/Van Gogh, is altered in the mouse neural tube mutant Loop-tail. Nat Genet., 28:251-255. Koehn, D., D . M . Juriloff, and M . J. Harris (1988) Research News. Mouse Newsletter, 80:151-152. Koleske, A.J., A . M . Gifford, M . L . Scott, M . Nee, R.T. Bronson, K . A . Miczek, and D. Baltimore (1998) Essential roles for the A b l and Arg tyrosine kinases in neurulation. Neuron, 21:1259-1272.  165  Kuida, K., T.F. Haydar, C.Y. Kuan, Y. Gu, C. Taya, H. Karasuyama, M.S. Su, P. Rakic, and R.A. Flavell (1998) Reduced apoptosis and cytochrome c-mediated caspase activation in mice lacking caspase 9. Cell, 94:325-337. Kohlbecker, A., A.E. Lee, and H. Schorle (2002) Exencephaly in a subset of animals heterozygous for the AP-2a mutation. Teratology, 65:213-218. Lanier, L.M., M.A. Gates, W. Witke, A.S. Menzies, A.M. Wehman, J.D. Macklis, D. Kwiatkowski, P. Soriano, and F.B. Gertler (1999) Mena is required for neurulation and commissure formation. Neuron, 22:313-325. Macdonald, K.M., D.M. Juriloff, and M.J. Harris (1989) Developmental study of neural tube closure in a mouse stock with a high incidence of exencephaly. Teratology, 39:195-213. Montcouquiol, M., R.A. Rachel, P.J. Lanford, N.G. Copeland, N.A. Jenkins, and M.W. Kelley (2003) Identification of Vangl2 and Scrbl as planar polarity genes in mammals. Nature, 423:173-177. Murdoch, J.N., K. Doudney, C. Paternotte, A.J. Copp, and P. Stanier (2001) Severe neural tube defects in the loop-tail mouse result from mutation of Lppl, a novel gene involved in floor plate specification. Hum Mol Genet., 10:2593-2601 Murdoch, J.N., D.J. Henderson, K. Doudney, C. Gaston-Massuet, H.M. Phillips, C. Paternotte, R. Arkell, P. Stanier, and A.J. Copp (2003) Disruption of scribble (Scrbl) causes severe neural tube defects in the circletail mouse. Hum Mol Genet., 12:87-98.  MRC Vitamin Study Research Group (1991) Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet, 338:131-137. Nagai, T., J. Aruga, O. Minowa, T. Sugimoto, Y. Ohno, T. Noda, and K. Mikoshiba (2000) Zic2 regulates the kinetics of neurulation. Proc. Natl Acad. Sci. USA, 97:1618-1623 Nakatsu, T., C. Uwabe, and K. Shiota (2000) Neural tube closure in humans initiates at multiple sites: evidence from human embryos and implications for the pathogenesis of neural tube defects. Anat Embryol, 201:455-466. Pasteris, N.G., K. Nagata, A. Hall, and J.L. Gorski (2000) Isolation, characterization, and mapping of the mouse Fgd3 gene, a new Faciogenital Dysplasia (FGD1; Aarskog Syndrome) gene homologue. Gene, 242:237-247.  166  Rampersaud, E., E.C. Melvin, D . Siegel, L. Mehltretter, M . E . Dickerson, T . M . George, D. Enterline, J.S. Nye, and M . C . Speer (2003) Updated investigations of the role of methylenetetrahydrofolate reductase in human neural tube defects. Clin. Genet., 63:210-214. Rampersaud, E., A . G . Bassuk, D.S. Enterline, T . M . George, D.G. Siegel, E.C. Melvin, J. Aben, J. Allen, A . Aylsworth, T. Brei, J. Bodurtha, C. Buran, L.E. Floyd, P. Hammock, B . Iskandar, J. Ito, J.A. Kessler, N . Lasarsky, P. Mack, J. Mackey, D. McLone, E. Meeropol, L. Mehltretter, L.E. Mitchell, W.J. Oakes, J.S. Nye, C. Powell, K . Sawin, R. Stevenson, M . Walker, S.G. West, G. Worley, J.R. Gilbert, and M . C . Speer (2005) Whole genome-wide linkage screen for neural tube defects reveals regions of interest on chromosomes 7 and 10. JMed Genet., Apr 14; [Epub ahead of print]. Rittler, M . , J. Lopez-Camelo, and E.E. Castilla (2004) Sex ratio and associated risk factors for 50 congenital anomaly types: Clues for causal heterogeneity. Birth Defects Res A Clin Mol Teratol, 70:13-19. Rohlf, F.J. and R.R. Sokal (1981) Statistical tables. Second edition, W.H. Freeman and Company, New York, pp: 97-99. Sabapathy, K., W. Jochum, K . Hochedlinger, L. Change, M . Karin, and E.F. Wagner (1999) Defective neural tube morphogenesis and altered apoptosis in the absence of both JNK1 and JNK2. Mech Dev., 89:115-124. Sadler, T.W. (2005) Embryology of Neural Tube Development. Am JMed Genet C SeminMed Genet., 135C:2-8. Sah, V.P., L.D. Attardi, G.J. Mulligan, B.O. Williams, R.T. Bronson, and T. Jacks (1995) A subset of p53-deficient embryos exhibit exencephaly. Nat. Genet., 10:175-180. Schorle, H., P. Meier, M . Buchert, R. Jaenisch, and P.J. Mitchell (1996) Transcription factor AP-2 essential for cranial closure and craniofacial development. Nature, 381:238-241. Shields, D.C., P.N. Kirke, J.L. Mills, D. Ramsbottom, A . M . Molloy, H . Burke, D.G. Weir, J.M. Scott, and A.S. Whitehead (1999) The "thermolabile" variant of methylenetetrahydrofolate reductase and neural tube defects: A n evaluation of genetic risk and the relative importance of the genotypes of the embryo and the mother. Am. J. Hum. Genet., 64:1045-1055. Silver, L . M . (1995) Mouse Genetics: Concepts and Applications. Oxford University Press, Oxford, pp. 43-50.  167  Sokal, R.R. and F.J. Rohlf (1995) Biometry: The principles and practice of statistics in biological research. Third edition, W.H. Freeman and Company, New York, pp: 794-797. Takeuchi, T., Y . Yamazaki, Y . Katoh-Fukui, R. Tsuchiya, S. Kondo, J. Motoyama, and T. Higashinakagawa (1995) Gene trap capture of a novel mouse gene, jumonji, required for neural tube formation. Genes Dev., 9:1211-1222. Taylor, L.A., M.J. Harris, and D . M . Juriloff (2000) Whiskers amiss, a new vibrissae and hair mutation near the K r t l cluster on mouse chromosome 11. Mamm Genome, 11:255-259. Unger, A.E., M.J. Harris, S.E. Bernstein, J.C. Falcone, and S.E. Lux (1983) Hemolytic anemia in the mouse. Report of a new mutation and clarification of its genetics. J. Hered., 74:88-92. Van der Put, N . M . , R.P.M. Steegers-Theunissem, P. Frosst, F.J.M. Trijbels, T.K.A.B. Eskes, L.P. van den Heuvel, E . C . M . Mariman, M . den Heyer, R. Rozen, and H.J. Bloom (1995) Mutated methylenetetrahydrofolate reductase is a risk factor for spina bifida. Lancet, 346:1070-1071. van Straaten H . M . and A.J. Copp (2001) Curly tail: a 50-year history of the mouse spina bifida model. Anat Embryol., 203:225-237. Waddington C H . (1975) The Evolution of an Evolutionist. Cornell University Press, Ithaca, New York, pp. 16-22. Winograd, J., M.P. Reilly, R. Roe, J. Lutz, E. Laughner, X . X u , L. Hu, T. Asakura, D. vander Kolk, J.D. Strandberg, and G.L. Semenza (1997) Perinatal lethality and multiple craniofacial malformations in M S X 2 transgenic mice. Hum Mol Genet., 6:369-379. Ybot-Gonzalez P., P. Cogram, D. Gerrelli, A.J. Copp (2002) Sonic hedgehog and the molecular regulation of neural tube closure. Dev., 129: 2507-2517. Zhang, J., S. Hagopian-Donaldson, G. Serbedzija, J. Elsemore, D. Plehn-Dujowich, A.P. McMahon, R.A. Flavell, and T. Williams (1996) Neural tube, skeletal and body wall defects in mice lacking transcription factor AP-2. Nature, 381:238-241. Zhao Q., R.R. Behringer, and B. De Crombrugghe (1996) Prenatal folic acid treatment suppresses acrania and meroanencephaly in mice mutant for the Cartl homeobox gene. Nature Genet., 13:275-283.  168  Online references and sources http://www.nlm.nih.gov/medlineplus/mplusdicti^^ http://seq.yeastgenome.org/cgi-bin/web-primer http://genome.ucsc.edu/cgi-bin/hgGateway http://www.informatics.jax.org/ http://www.georgetown.edu/faculty/ballc/webtools/web_chi.html http://www.informatics.jax.org/mgihome/nomen/strains.shtml http://nimitz.mcs.kent.edu/~blewis/stat/spearman.html http://www.ensembl.org/Mus_musculus/index.html  Appendix 1: Primer sequences for the primers designed i n our lab. Refer to the S S L P s section i n the General Materials and Methods for how they were designed.  Primer name  Chr  Forward sequence  Reverse sequence  Fgd3-C/D  13  5 '-TTGGATTCCATCCCCAGTAA-3'  5 '-CAGCCTACTCACCACAAACTG-3'  Fgfr4-E/F  13  5' - ACTAAGGGC ATCTGGGGTG AT-3'  5' - ACTGCCCGGC ATCTATATCTA-3'  Gprk6-C/D  13  5' - AGGGAGGGTGCTAACTTC AG A-3'  5' - AC AATGGTGGGATCTGGCTAA-3'  Ntrk2-C/D  13  5' -GCGTGCCTTTTATC AACTGCT-3'  5 '-AGGTCGTCAGGTTTGAGTGAT-3'  Fancc-C/D  13  5 '-TCCAGGCCAACTAAAGCTACA-3'  5' - ACGCTTCCCTCCCTTACTACT-3'  Ptcl-A/B  13  5' -GCTTCCTCCGTTTTGCA ATT-3'  5 '-ACAACTGCCTGCAACTCCAA-3'  Adcy2-A/B  13  5 '-CATACATTCAGGCAACAC ACG-3'  5' -TTG AA ATC ACTGTGGCATGG-3'  Nkd2-A/B  13  5' - ATGTCCC AGG AG ACC A AGCTA-3'  5 '-CAAAACCTCATGACGTGGAA-3'  Nr2fl-A/B  13  5' - ATCGTTCCTCCC AC AG AAAC A-3'  5' -TGA ACTCTCTGTGTGCGTGTG-3'  G2151-A/B  13  5' -CGTGTTCCCTC AG ATCC ATGT-3'  5 '-CAGCCATAGAGCTAATGCACA-3'  EdiB-A/B  13  5 '-TGCTGCTGTTAAATGAAACCA-3'  5 '-TGGAAGCACAGACACTGGAAA-3'  Gats-C/D  5  5 '-CCCAGCCTTTTCCATGTAAA-3'  5 '-GAC AGGATAATGACTGTGCCA-3'  Scn4a-C/D  11  5 '-ACGGATAACTGACTGG AAAGG-3'  5 '-CGACATAC ACCCATCTTCCTT-3'  Rshll-C/D  7  5' - AC AG AGAGACTGAC AG ACTGGC-3'  5' -TCTATGTTCACCGGATCTGA AG-3'  Ccnel-A/B  7  5 '-TGGGTCAGTGTGGGTTTCTTT-3'  5' -CCAAAAGGGTC ACTTCCCTTA-3'  Gnefr-C/D  7  5 '-TGGCTTCTTATGTACCGCTCA-3'  5' - A AC AC ACGAGGCTCCATCATA-3'  Capnl-A/B  19  5' -TTCGCTTCTCCTTCTTCTCCT-3'  5 '-TGGACTGTCCTTGTCTCAAAA-3'  X,Y  5 '-CCGCTGCCAAATTCTTTGG-3'  5 '-TGAAGCTTTTGGCTTTGAG-3'  Smcx-1, y-1  170  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092219/manifest

Comment

Related Items