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 Dec 16, 2005

You don't seem to have a PDF reader installed, try download the pdf

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

MULTIFACTORIAL GENETICS OF EXENCEPHALY IN THE SELH/Bc MOUSE STRAIN by JULIA LYNN HOSCHEIT B.Sc, Colorado State University, 2002 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Medical Genetics) THE UNIVERSITY OF BRITISH COLUMBIA 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. My 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, LM/Bc. 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 PMD #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 PMD #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 LM/Bc into the SELH/Bc background and the Exenl and Exen2 alleles from SELH/Bc into the LM/Bc 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 PMD #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 v List of Tables viList of Figures ix Abbreviations xi Acknowledgement xiChapter 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 conditionsMouse stocks 23 Polymerase Chain Reaction and viewing products in gels 25 SSLPs 26 DNA preparation 27 Chapter Three: F2 exencephaly frequencies 32 Introduction 3Materials and Methods 33 Results 37 iv Discussion 44 Chapter Four: F2 genotypes 8 Introduction 4Materials and Methods 51 Results 54 Discussion 65 Chapter Five: Congenic line exencephaly frequencies 69 Introduction 6Materials and Methods 72 Results 78 Discussion 88 Chapter Six: Recombinant congenic lines 97 Introduction 9Materials and Methods 98 Results 102 Discussion 6 Chapter Seven: Neural tube closure patterns in the congenic lines 109 Introduction 10Materials and Methods 115 Results 123 Discussion 142 Chapter Eight: Conclusions and general discussion 15References 16V Online references and sources 169 Appendix 1 170 vi List of Tables Table 2.1: Locations of the markers used in my studies in the subsequent chapters. 30 Table 3.1: Frequency of exencephaly and post-implantation mortality rates ("moles") in F2 generations from crosses between SELH/Bc and LM/Bc in all F2 exencephaly studies. 38 Table 3.2: Exencephaly frequencies and post-implantation mortality rates ("moles") in F2 generations from crosses between SELH/Bc and LM/Bc. 41 Table 3.3: Exencephaly frequencies within the "grandmother groups," subdivided by diet. 43 Table 4.1: Genotype summaries of F2 ex encephalic embryos and their corresponding values at various genetic markers in the EX-5001 study. 55 Table 4.2: Genotype summaries of control normal F2 embryos and their corresponding x2 values at selected genetic markers in the EX-5001 study. 57 Table 4.3: Genotype summaries of F2 exencephalic embryos and their corresponding x2 values at various genetic markers in the JHEX-5015 study. 61 Table 4.4: Genotype summaries of control normal F2 embryos and their corresponding x2 values at selected genetic markers in the JHEX-5015 study. 61 Table 5.1: Exencephaly frequencies and post-implantation mortality rates ("moles") in the SELH/BC and LM/Bc strains and congenic lines on PMD #5015. 81 Table 5.2: Exencephaly frequencies and post-implantation mortality rates ("moles") in the SELH/BC and LM/Bc strains and congenic lines onPLRD#5001. 84 Table 5.3: Exencephaly frequencies in parental strains and congenic lines on PMD #5015 and PLRD #5001 for Sets 1 and 2. 85 Table 5.4: Values of x (obtained from Falconer and Mackay, 1996) related to the combined exencephaly frequencies from Sets 1 and 2. 92 Vll Table 6.1: Exencephaly frequencies and post-implantation mortality rates ("moles") in the recombinant congenic lines on PMD #5015. 105 Table 7.1: Summary table of the number of litters collected in the parental strains and congenic lines. 149 Table7.2: The number of litters that were collected at various gestational ages. 149 Table 7.3: Mean somite count in parental strains and congenic lines at all the stages. 150 Table 7.4: Mean somite count in the parental strains and congenic lines in the sub-stages of Stage 4 of cranial neural tube closure. 151 viii List of Figures Figure 1.1: Conceptual summary of the normal process of neural fold elevation in the mouse. 7 Figure 1.2: Diagrammatic side-view of a Day 8/9 embryo showing the locations of independent zones of neural fold elevation. 9 Figure 1.3: Side-view of SELFI/Bc embryos with exencephaly. 14 Figure 1.4: Multifactorial threshold model using the frequency of exencephaly in SELH/Bc as an example. 18 Figure 1.5 A: Schematic representation of the creation of a congenic line. 21 Figure 1.5B: Graphical representation of the gain of homozygosity back to the recipient strain and loss of heterozygosity at backcross generations. 21 Figure 3.1: Conceptual diagram illustrating the genotype distributions for any particular locus of the sex chromosomes in F2 segregants from reciprocal grandmother crosses. 36 Figure 5.1: A conceptual picture of the construction of the congenic lines. 70 Figure 5.2: The current chromosomal segments that are in each congenic line. 79 Figure 5.3: Testing the multifactorial threshold model with the congenic lines. 91 Figure 5.4: Additivity of the Exen loci. 94 Figure 6.1: Conceptual diagram of the creation of 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 of the pattern of cranial neural tube closure in normal mouse strains. 110 Figure 7.2: Scanning electron micrographs of cranial neural tube closure in Day 8/9 embryos. 112 Figure 7.3: Diagrammatic representations of the sub-stages of Stage 4 and the novel conformation that was observed in sub-stage 4B' in the congenic line embryos. 120 Figure 7.4: Comparison between LM/Bc and SELH/Bc mean number of somite pairs present at different stages of closure. 125 Figure 7.5: Comparison between LM/Bc and SELH/Bc mean number of somite pairs present at the sub-stages for Stage 4. 126 Figure 7.6: Comparison between LM/Bc and 13S/LM mean number of somite pairs present at different stages of closure. 129 Figure 7.7: Comparison between LM/Bc and 13S/LM mean number of somite pairs present at the sub-stages for Stage 4. 130 Figure 7.8: Light microscope photographs of congenic line embryos in certain stages and sub-stages. 131 Figure 7.9: Comparison between LM/Bc and 5S/LM mean number of somite pairs present at different stages of closure. 133 Figure 7.10: Comparison between LM/Bc and 5S/LM mean number of somite pairs present at the sub-stages for Stage 4. 134 Figure 7.11 Comparison between SELH/Bc and 13L/SELH mean number of somite pairs present at different stages of closure. 136 Figure 7.12:Comparison between SELH/Bc and 13L/SELH mean number of somite pairs present at the sub-stages for Stage 4. 137 Figure 7.13 Comparison between SELH/Bc and 5L/SELH mean number of somite pairs present at different stages of closure. 139 Figure 7.14:Comparison between SELH/Bc and 5L/SELH mean number of somite pairs present at the sub-stages of Stage 4. 140 X Abbreviations ANP - anterior neuropore Chr - chromosome cM - centimorgans DLHP - dorsolateral hinge point ETn - early transposon ETnH - early transposon type II Exenl5, or Exen2s - Exenl or Exenl alleles from SELH/Bc EX-5001 - the F2 exencephaly panel collected from mice fed PLRD #5001, the normal diet. Collected by Drs. Diana Juriloff and Muriel Harris. JHEX-5015 - the F2 exencephaly panel collected from mice fed PMD #5015, the high-risk diet. Collected by me, Julia Hoscheit. JNK - c- Jun N-terminal kinase Mb - megabases MGI - Mouse Genome Informatics MHP - median hinge point NTD - neural tube defect PLRD #5001 - Purina Laboratory Rodent Diet #5001 PMD #5015 - Purina Mouse Diet #5015 Shh - Sonic hedgehog Somites - somite pairs SSLP - simple sequence length polymorphism TG Panel - the original F2 exencephaly panel collected by Teresa Gunn. UCSC - University of California Santa Cruz genome browser SS - homozygous SELH/Bc alleles LL - homozygous LM/Bc alleles SL - heterozygous SELH/Bc and LM/Bc alleles 13L/SELH - the congenic line that had been created by transferring the normal Exenl alleles from LM/Bc into the SELH/Bc background. 5L/SELH - the congenic line that had been created by transferring the normal Exen2 alleles from LM/Bc into the SELH/Bc background. 13S/LM - the congenic line that had been created by transferring the SELH/Bc Exenl alleles into the LM/Bc background. 5S/LM - the congenic line that had been created by transferring the SELH/Bc Exen2 alleles into the LM/Bc background. 7S/LM - the congenic line that was created by transferring the SELH/Bc Exen4 alleles into the LM/Bc background. xi Acknowledgement First and foremost, I would like to thank my supervisors, Dr. Diana Juriloff and Dr. Muriel Harris, for their time, support, guidance, and expertise. I am very appreciative of 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 of time and effort and I feel fortunate to have been able use them in my research. I would also like to acknowledge the help of former and present lab-mates, Diana Mah and Sarah Kennedy. Diana Mah prepared the DNA for the F2 exencephaly panel collected by Drs. Juriloff and Harris and genotyped the F2 exencephalic embryos for some markers on Chrs 13 and 5. Sarah Kennedy was involved in the creation of 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/Rec-6 line recombinant congenic line for a couple markers to further define boundaries on Chr 13.1 would also like to thank them for their conversations and friendship. Finally I would like to thank my family and friends for their love and support, especially my father Greg and his wife Pam, my mother Cate and her husband Kevan, and my sister Emily. xii Chapter 1: General Introduction Neurulation, the process of forming the neural tube, is a fundamental event of vertebrate embryogenesis. The neural tube is the early precursor to the brain and spinal cord and its proper closure is essential for the differentiation of the central nervous system. This process is extremely complex involving numerous cell mechanisms and the combined effort of several morphogenetic forces (Sadler, 2005; Colas and Schoenwolf, 2001; Copp et al., 2003). Failure of 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 in the general population. Failure of the posterior neural tube to close completely results in spina bifida, which, at birth, is characterized by congenital cleft of the spinal column with hernial protrusion of the meninges (coverings of the spinal cord) and sometimes the spinal cord (http://www.nlm.nih.gov/medlineplus/ mplusdictionary.html). Failure of the cranial neural tube to close completely results in anencephaly, which, at birth, is characterized by congenital absence of all or a major part of 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 of NTDs. Although some NTDs can appear as part of a syndrome, the majority are isolated, non-syndromic cases that do not have an identifiable cause but are attributed to a combination of genetic and environmental variables. The non-Mendelian pattern of increased recurrence risk with number affected in a family has pointed to a genetically multifactorial basis of NTD. The recurrence risk 1 in first-degree relatives of 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 NTD, 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 of NTDs. The risk of both recurrence and occurrence of 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 (MRC Vitamin Study Research Group, 1991). Given the evidence for both genetic and environmental factors influencing the risk of NTDs in the human population, the majority of NTD 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 of common human NTDs have been identified. Part of the difficulty in the genetic studies of human NTD is limited availability of families with several affected members, low reproductive fitness for people with spina bifida, fatality of anencephalic newborns, and the unfeasibility of studying human embryos. With respect to genetics, the nature of the genetic variants that interact to cause risk of NTDs remains a mystery, adding to the difficulty of genetic studies of human NTDs. Whether the genetic variants that cause risk of NTD are mutations or polymorphisms is not known, as well as whether the alleles that contribute to multifactorial nonsyndromic NTDs are at the same loci as the syndromic mutations (Juriloff et al., 2001). Furthermore, it is not known if multifactorial NTDs arise from cumulative effects of the genetic variants in unrelated genes expressed in interacting tissues in neural tube closure or whether they arise from cumulative effects of the genetic 2 variants from genes that are members of the same gene regulatory pathway (Juriloff et al., 2001). Many candidate genes have been investigated in human NTD populations using case-control and family-based association studies. One of the difficulties in these types of studies is having sufficient sample sizes to have the power to detect significance of the genetic variant of interest, especially since the cause of NTDs appears to be highly heterogeneous. Generally the candidate genes that are studied are in biochemical pathways, of most interest the folate metabolic pathway, developmental pathways, and those that cause NTDs in mice when "knocked out." Several of the studies fall short of showing significant results, some have been suggestive, and only a few have yielded positive results (Reviewed in Boyles et. al, 2005). A polymorphism of 5,10-Methylenetetrahydrofolate reductase (MTHFR) (677C—>T), a key enzyme in the methylation cycle (folate metabolic pathway) that has 50-60% lower activity than the wild-type variant, has been associated with NTDs in some populations, specifically the Irish and Dutch (Shields et al., 1999; van der Put et al, 1995), but not in 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 of NTDs in American Caucasian spina bifida families suggesting variations in NCAM1 may influence risk for human NTDs (Deak et al., 2005). Most recently, a genome-wide linkage screen on families having at least two related individuals with NTDs revealed regions of interest on chromosomes 7 and 10, possibly, as Rampersaud et al. state, "providing valuable positional data for prioritization of candidate gene assessment in future studies of NTDs" (Rampersaud et al., 2005). Failure to identify major causal genes 3 for human multifactorial NTDs and lack of positive results coming from the human NTD field have led the way for animal models to provide valuable insight into the critical pathways and cellular mechanisms that are involved in proper neural tube development and the causes of NTDs. Neurulation In humans, development and closure of the neural tube is usually completed within 28 days of conception. Given that this is a time that is inaccessible to study the developing neural tube in human embryos and given the ethical concerns, neurulation has been extensively studied in 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 of the spinal cord and will be discussed below. Secondary neurulation occurs at the more caudal levels and creates the lowest portion of the spinal cord and tail and will 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 of the tail bud undergoing condensation and epithelialization to form the secondary neural tube, the lumen of which is continuous with that of 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 of primary neurulation with emphasis on the cranial region. Primary neurulation is a process that involves shaping, folding, and fusing of bilateral neural folds (Copp et al., 2003). Conceptually, it is described as a developmental process that results in the rolling up of a flat layer of ectodermal cells into an elongated 4 hollow tube (Colas and Schoenwolf, 2001). On a tissue level, primary neurulation begins with the formation of the neural plate followed by shaping of the neural plate, bending of the neural plate (elevation of neural folds), and fusion of the bilateral neural folds. The neural plate is derived from ectoderm, one of the three germ layers that are formed during gastrulation, differentiating via induction from the underlying notochordal plate (Sadler, 2005). Cells destined to form the neural plate elongate in an apical-basal direction to form a thickened region of 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 of 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 in 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 of neural tube closure, occurring in the cervical region, requires convergent extension, as shown by mouse knockouts for genes involved in the planar cell-polarity pathway that have severe NTDs (Reviewed in Copp et al., 2003). As the neural plate undergoes elongation, the underlying mesoderm does so as well by similar cell movements so that the entire body axis lengthens (Sadler, 2005). Bending of the neural plate is initiated as its shaping is underway. Firstly, bending involves the formation of the neural folds, comprised of neuroepithelium (neuroectoderm) and underlying mesenchyme within a surface ectoderm (Harris and Juriloff, 1999), at the lateral borders of the neural plate (Colas and Schoenwolf, 2001). This involves expansion of the cranial mesenchyme, which in turn begins the elevation 5 phase of the cranial neural folds (Copp et al., 2003). The neural folds elevate from the median hinge point ("MHP"), the first principal site of bending, and establish a trough like space called the neural groove, which becomes the lumen of the primary neural tube after fusion of the bilateral neural folds (See Fig. 1.1). The MHP overlies the notochord and extends the entire neural axis. The second principle sites of bending are the paired dorsolateral hinge points ("DLHPs") along the sides of the neural folds. Although they do not form in most regions of the spine, they form in 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 of sub-apical actin microfilaments, emigration of the neural crest, maintenance of a proliferative neuroepithelium, and programmed cell death, are all thought to have a role in shaping (elevation) and dorsolateral bending (Copp et al., 2003). Bending of the neural plate and formation and elevation of the bilateral neural folds ultimately brings them into apposition with each other at the dorsal midline where they make contact and fuse, establishing the "roof of the neural tube, a continuous layer of cells across the neural groove (Colas and Schoenwolf, 2001). The overlying surface ectoderm of the neural folds fuses across the midline, as well, to become the overlying epidermis, separating it from the neural tube. Adhesion of the neural folds is thought to involve cellular protrusions extending from the cells on the tips of the neural folds, interdigitating as the folds come into contact (Geelen and Langman, 1979). As discussed in Copp et al., once the neural folds adhere to one another, programmed cell death is 6 Figure 1.1: Conceptual summary of the normal process of 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 in "remodeling the epithelia in 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 of the current knowledge about NTDs comes from animal models. There are over 100 single-gene mutations (and counting) in mice that have been reported to affect neurulation and cause NTDs. The majority of these mutations are knockouts of genes (null mutations), each of which has a homologous human gene (Harris, 2001; reviewed in Harris and Juriloff, 1999, and Juriloff and Harris, 2000). This shows that the misregulation of any number of different genes, related directly or indirectly to neurulation, may result in NTDs. This diversity demonstrates the extent to which the cause of NTDs can be heterogeneous. In the mouse mutants, the locations of NTDs appear to be related to specific zones suggesting that there are different mechanisms specific to each type of NTD (Juriloff and Harris, 2000) (See Fig. 1.2). Failure of elevation in the specific zones leads to split face (failure of elevation in Zone A), exencephaly (failure of elevation in Zone B), rachischisis (failure of elevation in Zone D), or spina bifida (failure of elevation of caudal Zone D). The regional location of NTDs (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 in neurulation-related gene expression (Copp et al., 2003). 8 Figure 1.2: Diagrammatic side-view of a Day 8/9 embryo showing the locations of independent zones of neural fold elevation. Locations of NTDs 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 " \ / ^ Rachischisis JJ- i 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 in the planar cell-polarity pathway (Reviewed in 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 of neural tube closure (Curtin et al., 2003; Hamblet et al., 2002; Kibar et al., 2001; Montcouquiol et al., 2003; Murdoch et al., 2001; Murdoch et al., 2003). These mouse mutants fail to initiate contact in the cervical region (Closure 1), which subsequently leads to craniorachischisis (most of the neural tube remains open). Knockouts for genes in 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 ("DHLP") in the neural folds down the neural axis (Copp et al., 2003) and this has been supported by Shh-null mutants, where DLHPs are observed in the neural folds and the neural tube closes successfully (Ybot-Gonzalez et al., 2002). In addition, Shh over-expression has been shown to produce NTDs (Echelard et al., 1993). The Patchedl (Ptcl) receptor, responsible for maintaining Shh signaling in an off-state in the absence of Shh ligand, when knocked-out produces severe NTDs in Ptcl-mx\\ mice (Goodrich et al., 1997). Other mutants such as extra-toes, open-brain, and Zic2 disrupt Shh signaling and have NTDs as well (Gunther et al., 1994; Hui 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). Two mouse knockouts for genes expressed in the mesenchyme that are important for proliferation, Twist and Cartl, have cranial NTDs, providing support that expansion of the cranial mesenchyme is 10 necessary for cranial neural tube closure (Chen and Behringer, 1995; Zhao et al, 1996). Mouse knockouts in 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 of which are actin-related, have exencephaly, providing support that cranial neurulation is highly dependent on the actin cytoskeleton. Knockouts in apoptosis-related genes have NTDs as well providing support that programmed cell death is important in neurulation. Interestingly, knockouts of genes that lead to increased or decreased apoptosis are attributed to the cause of NTDs. The Apafl, caspase9, and p53 (to name a few) knockouts all have reduced apoptotic cell death in association with the development of exencephaly (Harris (B.S.) et al., 1997; Honarpour et al., 2001; Kuida et al., 1998; Sah et al., 1995) whereas some knockouts such as AP-2, ApoB, Tulpl (to name a few) have increased apoptosis in the cranial neural folds presumably associated with development of 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 in identifying critical pathways and cellular mechanisms/molecules involved in neurulation, many of them are syndromic; the NTD in the mouse (most likely exencephaly) is part of a severe syndrome that includes other primary defects in other developmental systems as well (Harris & Juriloff, 1999; Juriloff & Harris, 2000). Thus, these mouse mutants with single-gene mutations do not reflect what is observed in the majority of human NTD cases, the majority being non-syndromic and multifactorial, and therefore may be considered of limited value as a model for human NTD. On the plus side, there are some established 11 mouse models for spontaneous non-syndromic, multifactorial NTDs. Two of them are discussed below. Curly tail mouse model Curly tail (ct) is the most well understood mouse model for nonsyndromic NTDs, especially for posterior NTDs, and has been extensively studied for the past 50 years (van Straaten and Copp, 2001). This mouse model is characterized by a high frequency of tail flexion (considered to be a mild NTD) (50%), some spina bifida (10%), and a low frequency of exencephaly (3%). NTDs in ctlct mice are similar in location and form to those occurring in humans, as they have spina bifida and/or exencephaly. The principle ct gene maps to distal chromosome 4, and the phenotype and incidence of NTD are influenced by several modifier genes and by environmental factors (Reviewed in van Straaten and Copp, 2001), making this genetic system "multifactorial." The incompletely penetrant ct phenotype is mainly observed in ct homozygotes (recessive), but can be observed in some heterozygous mice in the presence of particular combinations of modifier alleles. It is a hierarchical genetic system (major gene and modifiers) where embryos need ct in order to have a NTD and therefore the polygenic additive version of 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 of NTDs in curly tail. 12 SELH/Bc mouse strain The SELH/Bc strain, the focus of this study, is a well-established mouse model for common NTDs, specifically exencephaly (the equivalent to human anencephaly). This mouse stock was developed by Drs. Diana Juriloff and Muriel Harris in the Department of Medical Genetics at the University of British Columbia and the origin of the strain is described in detail in Juriloff et al. (1989), as well as briefly in General Materials and Methods (Chapter 2). SELH/Bc is characterized by a high frequency of exencephaly (10-30%) and all SELH/Bc embryos exhibit an abnormal mechanism of neural tube closure whether they become exencephalic or not (Discussed further in Chapter 7). The exencephaly in SELH/Bc is generally over the midbrain region (See Fig. 1.3). The genetic cause is attributed to at least 3 interchangeable loci acting together additively, closely resembling common human NTDs. In addition to the multiple loci, environmental factors also influence the incidence of exencephaly in SELH/Bc. Different diets alter the risk of exencephaly in SELH/Bc. Harris and Juriloff state that "dietary supplementation of folic acid, methionine, zinc, niacin, brewers' yeast, riboflavin, vitamin B12, or inositol does not significantly reduce the risk of exencephaly" (Harris and Juriloff, 2005). The genetic basis of exencephaly in SELH/Bc mice Previous studies have shown the genetic cause of exencephaly to be attributed to the additive effects of at least three genes (Juriloff et al., 1989; Gunn et al., 1992; Juriloff et al., 2001). The first study determining the mode of inheritance of the exencephaly-13 Figure 1.3: Side-view of SELH/Bc embryos with exencephaly. Picture A shows an SELH/Bc embryo with exencephaly at 14 days of gestation (El4); note that the exencephaly spans the midbrain region. Picture B shows an SELH/Bc fetus (El8) where the brain tissue has degraded. This figure is from Macdonald et al. (1989). 14 causing genes in the SELH/Bc strain was based on segregants from a cross of SELH/Bc 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 of exencephaly was due to the additive effects of 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 of exencephaly in SELH/Bc involved classical crosses between SELH/Bc and an unrelated normal strain, SWY/Bc. Similar results were obtained from these segregants, as well, providing support for the hypothesis of two to three additive genes contributing to the risk of exencephaly in SELH/Bc (Gunn et al., 1992). The subsequent study that investigated the genetic basis of exencephaly in SELH/Bc was designed to identify the chromosomal locations of the SELH/Bc exencephaly risk genes and is described in Juriloff et al. (2001). This involved typing of markers across the genome in the extremes of F2 males (those that produced the most exencephaly in offspring and those that produced none), derived from crosses to a non-related strain, LM/Bc. In addition, an F2 exencephaly panel was collected to examine genotypes of F2 exencephalic embryos at markers in the provisional regions for exencephaly risk genes from SELH/Bc revealed in the F2 sire screen. Together these studies confirmed that the risk of exencephaly in the SELH/Bc strain was due to the cumulative effect of two or three loci. They were named Exenl, Exenl, and Exen3. The Exenl and Exen2 loci were mapped to mid Chr 13 and distal Chr 5, respectively, and the Exen3 locus provisionally to Chr 11. 15 Multifactorial threshold model The majority of human NTDs can best be explained by the multifactorial threshold model (Fraser and Nora, 1975). They are not inherited in a simple Mendelian manner. The concept is that there is a normally distributed continuous developmental variable in 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 Fig. 1.4). One either has an NTD or does not; there is no intermediate phenotype. Combinations of several genetic factors and environmental factors determine the liability to a birth defect, making the system multifactorial (Fraser, 1976). As the number of 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 of liability may be thought of as the rate of some developmental process or it may be a compound of 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 of the beauty of the model. Not only do genetic and environmental factors influence the position of the distribution relative to the threshold, they can influence the position of the threshold itself as well (Fraser, 1976). For NTDs, such as exencephaly in the SELH/Bc strain and anencephaly in humans, the continuous variable or liability trait is most likely to be timing of cranial neural fold elevation. Embryos with delayed elevation will have a higher probability of not completing elevation in their "window of opportunity" when they should be capable 16 of 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 of interest and it is assumed the population has a normal distribution for the liability. For genetic analyses, the proportion or percentage of 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 in percentages may appear to be non-additive when in 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 of liability is its standard deviation (a) and the mean liability is related to the incidence by the normal deviate x, which is the deviation of the threshold from the mean in standard deviation units of liability (See Fig. 1.4). Values of x for different incidences have been tabulated and can be obtained from Falconer and Mackay (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 of exencephaly in SELH/Bc as an example. 25% of SELH/Bc embryos cross the developmental threshold and become exencephalic. Using the table in Falconer and Mackay (1996) that shows the tabulated x values, the proportion 25% relates to 0.674 standard deviation units (a). In other words, the threshold of SELH/Bc is 0.674 a of liability away from the mean. 18 Rationale and approach to my studies My studies further investigated several aspects of the multifactorial nature of exencephaly in the SELH/Bc strain using two core tools, F2 exencephaly panels and congenic lines. F2 exencephaly panels Given that exencephaly is a complex trait in the SELH/Bc strain, F2 exencephaly panels, derived from crosses between SELH/Bc and the normal non-related strain, LM/Bc, were collected to examine the genotypes from F2 exencephalic embryos across the Exen loci. F2 exencephaly panels are useful because they can pick up all genotype combinations that contribute to the exencephaly. Basically, Exen alleles from SELH/Bc are expected to be transmitted more than the normal Exen alleles from LM/Bc in the F2 exencephalic embryos. The F2 exencephalic embryos would have more Exen alleles from SELH/Bc across a combination of 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. Two new F2 exencephaly panels were collected, one from mice fed the regular diet PLRD #5001 and one from mice fed the high-risk diet PMD #5015, to refine the mapping of the Exen loci and to examine the diet effect in F2 segregants. 19 Congenic lines Congenic lines are a powerful tool to map genes in a complex trait, and additionally, they are used to study the effects of 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 of risk each gene contributes. Congenic lines are created by repeated backcrosses to the background recipient strain. With 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). At each generation, only those offspring who have received the donor allele from the donor strain at the locus of interest were selected for the next round of backcrossing. At 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 SELH/Bc into a normal strain background, LM/Bc, and reciprocally substituting normal LM/Bc Exenl and Exen2 alleles, respectively, into the SELH/Bc background. They were created to serve multiple purposes. One of the main objectives of the congenic lines was to refine the mapping of the Exen genes. This partly involved defining the current intervals of the transferred chromosomal segments by typing various 20 Figure 1.5: A) Schematic representation of the creation of a congenic line. At each backcross generation to the recipient strain (black boxes) the residual heterozygosity (checkerboard pattern) is reduced by 50%. At each generation the donor allele/differential segment is selected for with genetic markers. B) Graphical representation of the gain of homozygosity back to the recipient strain (Y-axis) and loss of 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 of the main congenic lines to refine the mapping specifically for the Exenl locus on Chr 13. The other general objective of the congenic lines was to study the effects of the individual Exen genes. Firstly, they were used to investigate the individual effect of the Exen alleles from SELFf/Bc on exencephaly frequency, testing the multifactorial threshold model. It predicts that each congenic line with part of the liability to exencephaly from SELH/Bc will express part of the exencephaly frequency of SELH/Bc. Secondly, the congenic lines were used to test whether certain Exen loci respond to diet. Thirdly, they were used to isolate the individual effect of each Exen locus in the process of cranial neural fold elevation in early embryos. This thesis presents the work investigating the multifactorial cause of exencephaly in the SELH/Bc strain using F2 exencephaly panels and congenic lines. This work will advance the general understanding of the development of exencephaly in the SELH/Bc 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 of the studies described in the subsequent chapters. Mice Maintenance and conditions The mice were maintained in the animal unit in the Department of Medical Genetics at the University of 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 AM to 6:00 PM with a temperature of approximately 70°F. Density of mice was 2-5 per cage. They were kept specific pathogen free. Mouse stocks SELH/Bc SELH/Bc is an inbred strain at F48+. The SELH/Bc mice are non-agouti, black chinchilla (aaBBcchcch). The SELH stock was derived from a cross of a partially inbred stock of mixed genetic background (BALB/cGa, 129/-, CBA/-) to "random-bred BLU:Ha (ICR)." The mixed genetic background was homozygous for the lidgap-Gates (lgGa) mutation that was backcrossed onto the ICR background by intercross-backcross to N3. At that time a new recessive mutation, spherocytosis British Columbia, sph2Bc (Unger et al., 1983) appeared in the second intercross generation (N3 X N3). The parents of the 23 affected pups were used to begin a new stock segregating for sph c and it was maintained by brother-sister inbreeding with selection for sph2Bc carriers and against the lgGa mutation. It was in this stock, in the F5 newborns, that exencephaly was observed. The exencephaly-producing parents were selected for in subsequent generations, while sph2Bc was selected against, and it eventually became an inbred strain by sister/brother inbreeding. Most recently, the lgGa mutation in BALB/cGa mice was found to be an eight-exon deletion in the mouse Map3kl gene on distal Chr 13 (Juriloff et al., 2005). In addition to exencephaly, there have been more independent mutations observed in SELH/Bc 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 of the type 1 keratin cluster on Chr 11 (Taylor et al., 2000), one that causes a white belly spot that has been mapped to the Kit gene on Chr 5 (Kitw), and one that causes the lens of the eye to go opaque that has not been mapped (Juriloff et al., 2005). All new mutations that arose had been purged from the main SELH/Bc lineage. Ovarian teratomas have also been observed in this strain (personal communication). The apparently high spontaneous mutation rate in SELH/Bc indicates that this stock may be genetically unstable. Interestingly, the SELH/Bc strain appears to have a high level of retrotransposition of early transposons ("Etas"). One of the mutations at the albino locus mentioned above was found to be due to an ETnU insertion in exon 1 (Hofmann et al., 1998). Most recently, Baust et al. found that the SELH/Bc strain has a highly transcribed 24 variant of the ETnll retrotransposon suggesting that early retrotransposition occurs relatively frequently in the SELH/Bc strain (Baust et al., 2003). LM/Bc LM/Bc is a highly inbred strain that has been used as the control strain for SELH/Bc in previous studies as they have virtually no spontaneous production of exencephaly. They were derived by sib inbreeding of mice that carried 1/8 of their genes from the SWV/Bc strain and 7/8 of their genes from an inbred strain that had been derived from an unpedigreed stock of "C3H" mice (Juriloff et al., 1991). These mice are homozygous for lgm, a mutation that causes open eyes at birth. The animals used in 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 DNA. PCR primers designed to detect SSLPs were obtained from Research Genetics, Inc. (Huntsville, Alabama, USA), NAPS (University of British Columbia), or Sigma (Oakville, Ontario ). Some SSLP primers were designed in our lab using the website http://seq.yeastgenome.org/cgi-bin/web-primer (See below). Each PCR reaction was carried out in a 25 ul volume overlaid with mineral oil in a 650 u.1 reaction tube. Each reaction contained 100 ng of DNA and 0.14 uM of each forward and reverse primer. The rest of the reaction mixture contained "master mix", consisting of dATP, dGTP, dCTP, and dTTP (final concentration, 50 uM each; Qiagen), Taq DNA polymerase (final concentration, approximately 1.25 Units; Invitrogen), 10X PCR Buffer 25 (final concentration, 10 uM; Invitrogen), and magnesium chloride (final concentration usually 1.5 mM, range 1.5-2.5 mM; Invitrogen). PCR was usually performed in a Perkin-Elmer 4600 thermocycler, usually under the following conditions: 4.5 min. at 94°C (denaturation), followed by 35 cycles of 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 mix was added at 94°C after the DNA and primers were denatured. The marker dye bromphenol blue-xylene cyanol FF (5 pi) was added to the PCR product and approximately one third of the mixture (10 ul) was run by electrophoresis on 4% NuSieve 3:1 Agarose horizontal gels containing 0.5 ug/ml of ethidium bromide. Gels were run in IX TAE at about 150 V for 1-1.5 hours, then observed and photographed under UV light using Polaroid 667 film. SSLPs Simple sequence length polymorphisms ("SSLPs") were used to type the F2 exencephalic embryos and all other mice involved in my studies. They are highly polymorphic between inbred strains of mice and can be typed quickly and easily using the Polymerase chain reaction. The markers used for this study were designed in our lab, with the exception of the Mit markers. Generally, we used University of California Santa Cruz assemblies (http://genome.ucsc.edu/cgi-bin/hgGateway') to look up genes in the vicinity of the area we were interested in acquiring primers for. The sequence of 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 of a random distribution of 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 NAPS or Sigma. The primer sequences that were made in our lab can be found in Appendix 1. DNA preparation Phenol chloroform method DNA was prepared from stored frozen embryos by standard phenol chloroform extractions by Diana Mah for the EX-5001 F2 exencephaly panel (See Chapter 4). Approximately half of the embryo was used. It was washed in sterile saline solution before being frozen. On ice, the embryonic tissue was transferred to a 15 ml tube and broken apart. 3.5-5 ml of Lysis Buffer was added to each tube and then 200 pi of 10 mg/ml Proteinase K was added next and shaken gently to mix 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 of 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 of 100% ethanol. After gentle mixing the ethanol was poured out and the DNA was washed with 70% ethanol. The ethanol was decanted and the DNA was left to dry in 27 the fumehood for about 10 minutes. It was then transferred to a 1.5 ml tube and suspended in 100-300 pi 1 X TE depending on the size of the tissue. QIAamp DNA Mini kit (Tissue Protocol) DNA was prepared from stored frozen embryos by using the QIAamp DNA Mini Kit (Tissue Protocol) for the JHEX-5015 F2 exencephaly panel. Approximately 25 mg of 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 ml microcentrifuge tube and 180 pi of Buffer ATL was added. In addition, 20 pi of 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 pi of Buffer AL was added to the sample and mixed by vortexing and incubated at 70°C for 10 minutes. After incubation, 200 ul of 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 pi of Buffer AW1 was added and then centrifuged at 8000 rpm for 1 minute. This step was repeated with Buffer AW2 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 ml tube and 200 pi of Buffer AE 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 in the refrigerator for storage in a 1.5 ml tube. 28 QIAamp DNA Mini kit (Crude lysis protocol) This method was generally used to prepare DNA from stored frozen tail-tip tissue samples. The tail tissue was placed in a 1.5 mL microcentrifuge tube, and smashed with the "hard end" of yellow inoculating loops. 200 ul of lysis buffer ATL was added, followed by 20 ul Proteinase K and 200 ul Buffer AL. The mixture was vortexed for 15 seconds and then incubated at 56°C for 10 minutes. 400 ul of 95% ethanol was added and vortexed for 15 seconds to homogenize the lysate. Subsequently 635 ul of 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. At this point 500 ul Buffer AW1 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 ul Buffer AW2 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 L5 ml tube and 50 ul Buffer AE 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 DNA, suspended in Buffer AE, was stored in the refrigerator for future use in the 1.5 ml tube. 29 Table 2.1: Locations of the markers used in my studies in the subsequent chapters. Megabase values (Mb) were obtained from the University of California Santa Cruz (UCSC) genome browser (generally the May 2004 contig) (http://genome.ucsc.edu/cgi-bin/hgGatewav) and centimorgan (cM) values were obtained from Mouse Genome Informatics (MGI) (http://www.informatics.iax.org/mgihome/nomen/strains.shtml). Marker Chr Mb (UCSC) CM* (MGI) Other D13MU3 13 19.8 10.0 D13MU117 13 37.0 20.0 D13MU39 13 — 37.0 Not mapped in UCSC 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 — Not in MGI Nr2fl-A/B 13 74.3 45.0 G2151-A/B 13 82.0 — Not in MGI E4U3-A/B 13 85.0 — Not in MGI 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 — Not in MGI 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 30 D7MU75 7 5.9* 1.7 * Feb '03 UCSC contig Rshll-C/D 7 10.6 — Not in MGI D7MU79 7 22.8 16.0 Ccnel-A/B 7 33.5 16.0 Gnefr-C/D 7 40.5 — Not in MGI D7MU62 7 78.5 42.6 D10MU180 10 118.0 64.0 D10MU164 10 — 67.5 Not mapped in UCSC D17MU10 17 — 24.5 Not mapped in UCSC D19MU68 19 3.4 6.0 Capnl-A/B 19 5.8 3.0 * 1 cM on average relates to 1.8 Mb. I obtained this value by finding each chromosome's length (except the Y chromosome) in 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 MGI to get the cM 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 cM position in MGI. Generally, the known genes were 2 Mb within the end of the chromosome. I then divided the length of the chromosome (in Mb) by the cM position of the last known gene (with a cM position) for each chromosome and then added the 20 values up and divided that number by 20 to get the average ratio between Mb and cM, which was calculated to be 1 cM:1.8Mb. 31 Chapter 3: F2 exencephaly frequencies Introduction In the original studies of SELH/Bc, a panel of F2 exencephaly segregants from a cross between SELH/Bc to a normal non-related strain, LM/Bc, was collected as part of a study to determine the number of genes that act together to cause exencephaly in SELH/Bc and to map them. This original F2 exencephaly panel ("TG Panel") was collected by Teresa Gunn (Gunn, 1995 (thesis); Juriloff et al., 2001) in 1994 when SELH/Bc 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 SELH/Bc strain, partly to refine the mapping of the Exen loci with a larger panel, and partly to serve as a control in diet studies. This new study ("EX-5001") was collected from mice fed Purina Laboratory Rodent Diet #5001 ("PLRD #5001") by Diana Juriloff and Muriel Harris in 2003. Previous studies have showed that the exencephaly frequency in the SELH/Bc strain is affected by maternal diet (Harris and Juriloff, 2005). This observation has been consistent over the years in numerous studies since it was first observed in 1995. On PLRD #5001, the diet the mice are normally fed, the exencephaly frequency is approximately 5-10%. On Purina Mouse Diet #5015 ("PMD #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 SELH/Bc maternal genotype. Therefore, a second new F2 exencephaly study ("JHEX-5015") was collected on PMD #5015 to test the diet effect in segregants in a non-SELH/Bc mother. 32 Previous studies in the SELH/Bc strain show that there is an excess of females among exencephalic embryos. For example, in one study of SELH/Bc, 66% of exencephalic embryos were female (Juriloff et al., 1989). The sexes of the TG F2 exencephalic embryos are not known. Therefore, whether there is an excess of females among the F2 exencephalic embryos from the EX-5001 and JHEX-5015 studies was examined. Materials and Methods Mouse maintenance All mice originated from and were maintained in our animal unit in the Department of Medical Genetics at the University of British Columbia under standard conditions previously described (See General Materials and Methods). All breeding colonies that produced adult mice used in this study were maintained on Purina Laboratory Rodent Diet #5001 ("PLRD #5001"). Breeding design Reciprocal crosses between SELH/Bc and LM/Bc were made to create the Fl generation. To generate a cohort of F2 embryos, nulliparous Fl females (age 2-4 months) were bred with Fl male sibs. For both studies, each male received 2 females at a time (occasionally 1 or 3) and some males received a second set of females. Overall, the EX-5001 study derived from 143 females and 73 males and the JHEX-5015 study derived from 55 females and 27 males. Breeding for the EX-5001 study aimed to obtain 50 exencephalic embryos. Breeding for the JHEX-5015 study aimed to obtain about 30 33 exencephalic embryos. These sample sizes were expected to show a highly significant difference (p< 0.01) in exencephaly frequency if PMD #5015 caused at least a doubled rate of exencephaly. A panel of F2 exencephalic embryos from the EX-5001 study was collected by saving tissue from all exencephalic embryos from all litters. A panel of F2 exencephalic embryos from the JHEX-5015 study was collected by saving tissue from all exencephalic embryos from all litters. A panel of 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 of 39 control F2 embryos from the JHEX-5015 study was collected by saving tissue from all embryos from 3 random litters. Scoring exencephaly Pregnancies were judged by eye and palpation. Developmental stage of embryos was judged by morphological characteristics. Females were euthanized by carbon dioxide on day 14 of gestation (range E10 (1 litter) -El 6 (1 litter)) for examination of embryos. The uterus was removed, pinned to a black wax substrate, immersed in physiological saline (0.85% NaCl) solution, and cut open to reveal the conceptuses. Dead post implantation embryos ("moles") were recorded. The morphology of scoreable embryos was examined under a dissection microscope and recorded. The following defects were screened: exencephaly, spina bifida, rachischisis, major abnormalities of the limbs and tail, gastroschisis, cleft lip, and severe edema. The F2 exencephalic embryos were 34 individually washed in ice-cold sterile saline solution and stored in cryovials at -20°C for later preparation of DNA. Test of Diet Effect For the EX-5001 study, the Fl female mice were fed PLRD #5001 from weaning until autopsy. For the JHEX-5015 study, the Fl females were placed on PMD #5015 2-4 weeks before introduction of males and maintained on PMD #5015 until autopsy. Conception was always within a week of introduction of males. Cytoplasmic and X-linked effects After collecting the JHEX-5015 study, we realized we could look for "grandmother effects" (cytoplasmic and X-linked effects) on the risk of exencephaly in the F2 embryos by subdividing the data by strain of grandmother (See Fig. 3.1). In addition, we tested for a diet effect on exencephaly frequencies within the subdivided data. We also tested whether there was a "grandmother effect" on gender of exencephalic embryos. This would reflect an X-linked exencephaly risk factor in SELH/Bc or LM/Bc, if any (See Fig. 3.1). Within the grandmother groups, a diet effect on the sex of the embryos was also tested for. Following along with this type of 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 in F2 segregants from reciprocal grandmother crosses. The difference in X chromosome genotypes of the females in the F2 generation (circled) from reciprocal grandmothers would result in an impact on exencephaly if there were an X-linked recessive factor increasing risk of exencephaly. LM/Bc 9 x SELH/Bc rv F1 x F1 XLXL x XSY XLXS x XLY LM/Bc cytoplasm : XSXL : XLY : XSY ! SELH/Bc 9 x LM/Bc r? SELH/Bc cytoplasm XSXS x XLY F1 x F1 XSXL x XSY XSXS : XLXS : XSY : XLY 36 Sex genotyping For the EX-5001 exencephaly panel, DNA had been prepared from stored frozen embryos by standard phenol chloroform extraction and banked (See General Materials and Methods). For the JHEX-5015 F2 exencephaly panel and both F2 normal control panels, DNA was prepared from stored frozen embryos by using the QIAamp DNA Mini-kit (See General Materials and Methods). Sex of embryos was identified by PCR using the SmcX-1 and SmcY-1 primers (Personal Communication from David Threadgill). The primer sequences were obtained from the Smc gene and they can be found in Appendix I. The products were resolved electrophoretically on 4% NuSieve 3:1 Agarose gels, stained by ethidium bromide, and photographed over UV light. Statistical Methods To compare groups in various experiments the chi-square test of 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 TG study contained 3.7% exencephaly (43 in 1151 embryos) and the new EX-5001 study contained 2.8% exencephaly (52 in 1873 embryos) (See Table 3.1). These frequencies were not significantly different from each other (%2= 2.15;p> 0.5). In the JHEX-5015 study, the frequency of exencephaly was 4.4% (34 in 773 embryos), 37 Table 3.1: Frequency of exencephaly and post-implantation mortality rates ("moles") in F2 generations from crosses between SELH/Bc and LM/Bc in all F2 exencephaly studies. # litters # implants % moles # embryos # exens % Exen Sex ratio (F:M) TG 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 PMD #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 TG 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 (x2= 13.0;p< 0.001, x2- 5.8;p< 0.025). The exencephalic sex ratio was not significantly different between the two panels (x2= 0.20; p> 0.5), 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 EX-5001 and JHEX-5015 panels were 23:13 and 15:24. Neither control panel showed a significant deviation from a 1:1 sex ratio (x2= 2.78 and 2.08, respectively; p> 0.1). Cytoplasmic and X-linked effects on exencephaly The F2 exencephaly panels were subdivided by strain of grandmother to test for cytoplasmic and X-linked effects (See Table 3.2). In the EX-5001 study, the exencephaly frequency was significantly higher in the F2 embryos from LM/Bc grandmothers (3.5%) than from SELH/Bc grandmothers (1.8%) (x2= 4.58; p< 0.05). In the JHEX-5015 study, this "grandmother effect" followed the same trend with a higher frequency of exencephaly in F2 embryos from LM/Bc grandmothers (4.8%) than from SELH/Bc grandmothers (3.5%) but it was not statistically significant (x2= 0.56; p> 0.5). The pattern suggesting higher exencephaly frequencies in F2 embryos with LM/Bc cytoplasm could alternatively be caused by an X-linked exencephaly liability gene from LM/Bc (See Fig. 3.1 for concept on pg. 36). X-linkage is distinguishable from cytoplasmic effects because it would alter the sex ratio among F2 exencephalic embryos. All F2 males from either reciprocal cross would have the same mix of genotypes. In F2 exencephalic female embryos, theoretically half of them would be genetically the same in both reciprocal crosses, whereas the other half would be genetically different in 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 SELH/Bc grandmother group and 28:9 (3.1:1) within the LM/Bc grandmother group, fairly similar and not significantly different from each other (£= 0.03;p> 0.975) (See Table 3.2). In the JHEX-5015 panel, the exencephalic sex ratio was 40 Ind T^Mm^ThTnf ^^eqTCieS P^P1^™ mortality «** ("moles") in F2 generations from crosses between SELH/Bc and LM/Bc. The panels have been subdivided by strain of grandmother within each study on different diets. Study Grandmother # litters # implants % moles # embryos # exens % Exen Sex ratio (F:M) EX-5001 SELH/Bc 62 855 5.03 812 15 1.8 * ••»•/ 11:4 EX-5001 LM/Bc 81 1116 4.93 1061 37 3.5 28:9 JHEX-5015 SELH/Bc 17 245 7.76 226 8 3.5 7:1 JHEX-5015 LM/Bc 28 573 4.54 547 26 4.8 17:9 7:1 within the SELH/Bc grandmother group and 17:9 (1.89:1) within the LM/Bc 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 LM/Bc 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 LM/Bc grandmother groups. The response to PMD #5015 appeared to be about the same in each grandmother group; it increased the exencephaly rate in F2 embryos from LM/Bc grandmothers but it was not statistically significant (See Table 3.3). By subdividing the data in half we lost statistical power to pick up a diet effect, if 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 TG study (x2 = 2.16 and 0.53, respectively). 42 Table 3.3: Exencephaly frequencies within the "grandmother groups," subdivided by diet. Diet SELH/Bc Grandmothers LM/Bc Grandmothers PLRD #5001 1.8% 3.5% PMD #5015 3.5% 4.8% 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 SELH/Bc (5.03%) and LM/Bc (4.93%) grandmothers were essentially the same (%2= 0.01;/?= .995). In the JHEX-5015 study, 7.76% of total implants were "moles" within SELH/Bc grandmothers and 4.54% of total implants were "moles" within LM/Bc grandmothers (%2= 3.02; p> 0.1), not significantly different from each other. A diet effect on post-implantation mortality rates within grandmother groups was checked for. From SELH/Bc grandmothers, the post-implantation mortality rates from the two diet panels were not significantly different from each other (%2= 2.34; p> 0.1). A similar result was seen from LM/Bc grandmothers across the two diets (%2= 0.11; p> 0.9). This shows that the post-implantation mortality rates are not affected by strain of grandmother, diet, or an interaction between the two. Discussion The first major finding in this study was the diet effect in the F2 embryos. Given that the exencephaly rate doubles in the SELH/Bc strain on PMD #5015, we were looking for a similar increased rate of exencephaly in the F2 embryos that would double the 2.8% rate of exencephaly observed in the EX-5001 study. The exencephaly frequency from the JHEX-5015 study was significantly higher than the EX-5001 study, a 1.6 X increase. First and foremost, this suggests that the diet effect observed in the F2 studies does not require a SELH/Bc mother, a 100% SELH/Bc genetic background. This does not rule out though that the diet effect is mediated by certain maternal SELH/Bc 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 of the Exen loci. The exencephaly risk of SELH/Bc has been found to be due to a combination of 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 PMD #5015 to produce the increased exencephaly frequency. In the F2 segregants, if an embryo is homozygous or heterozygous for the SELH/Bc Exen alleles that interact with PMD #5015, they would have higher risk of becoming exencephalic than they would if the mice were fed PLRD #5001, so that more F2 embryos with that particular genotype on PMD #5015 would become exencephalic that would maybe not on PLRD #5001. Typing the DNA from the individual F2 exencephalic embryos in the EX-5001 and JHEX-5015 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 EX-5001 and JHEX-5015 studies involves general maternal effects. Firstly, we could be dealing with a general maternal effect that affects many strains of mice. One noticeable observation is that the mice become fatter faster when fed PMD #5015 compared to when they are fed PLRD #5001. Many mice might respond the same when eating PMD #5015. SELH/Bc mice also become noticeably fatter on PMD 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 PMD #5015. PMD #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 Fl 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 PMD #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 in neural tube defects to a differential embryonic growth rate but this difference was not seen in 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 of the neurulation process that increases the susceptibility of females to exencephaly (Brook et al., 1994). With the removal of the F2 exencephalic embryos (mostly female) from the samples, one might expect to see slightly skewed sex ratios in the normal F2 panels showing more males than females. When the total sample size is considered, though, removing 52 or 34 F2 exencephalic embryos (from the EX-5001 and JHEX-5015 studies, respectively) makes such a small difference to the sample that we do not expect to be able to detect an excess of males in normal embryos. The third major finding in this study was that the LM/Bc cytoplasm appeared to facilitate the expression of SELH/Bc genes that contribute to exencephaly. As previously described, we analyzed "grandmother effects" within the F2 exencephalic embryos for both new panels. In the EX-5001 study there was a significantly higher percentage of exencephaly from LM/Bc grandmothers compared to SELH/Bc grandmothers. Although this LM/Bc "grandmother effect" was not statistically significant in the JHEX-5015 study, it followed the same trend. This effect is attributed to the LM/Bc cytoplasm and not X-linkage because the sex ratio of 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 LM/Bc cytoplasmic effect is an unexpected finding as it possibly suggests an epigenetic contribution to the risk of exencephaly. 47 Chapter 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, LM/Bc. 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 PMD #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 TG studies, we expected the EX-5001 F2 exencephalic embryos' genotype distributions across the Exen loci to be comparable to those of the TG 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, if the role of one Exen locus weakened or disappeared on PMD #5015 in the F2 embryos, we would expect to see a random Mendelian genotypic ratio, whereas on PLRD #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 PMD #5015. 49 The alternative hypothesis that there is no Exen gene-diet interaction was concurrently tested. For example, if PMD #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 JHEX-5015 panels. This would suggest that the diet simply added to the genetic effect. Test crossing of segregants is a powerful way to estimate the number of loci contributing to a trait between inbred strains. Teresa Gunn's study indicated the number of major loci to contribute to the risk of exencephaly in SELH/Bc to be about three loci. The roles of the Exenl, Exen2, and Exen3 loci seem to be well established. However, the location of the Exen3 locus and possibly another locus are not well established. The previous F2 sire screen pointed towards regions on Chr 7, 10, and 17 possibly containing a locus contributing to the risk of exencephaly, but the significance was not sustained in the TG F2 exencephaly panel (Juriloff et al., 2001). For this reason they were revisited in the EX-5001 and JHEX-5015 studies, along with a new region of interest from Chr 19 that contains a MusD transposon insertion in the SELH/Bc strain (Personal communication from Dr. Dixie Mager). Therefore, the genotypes of 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 of interest to test the hypotheses stated above. 50 Materials and Methods Experimental strategy The first part of this study involved confirming the locations of Exenl (Chr 13), Exen2 (Chr 5), and Exen3 (Chr 11). This involved comparing the segregation ratios from the EX-5001 F2 exencephaly panel against the random genotype distribution of 13 SS:26 SL:13 LL at these loci (52 F2 exencephalic embryos). In addition, the segregation ratios at putative loci on Chr 7, 10, 17, and 19 were compared against the random distribution segregation ratio to test for a locus that contributed to the risk of exencephaly, if any. The second part of this study involved comparing the segregation ratios from the JHEX-5015 F2 exencephaly panel against the random genotype distribution of 8.5 SS:17 SL: 8.5 LL 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 JHEX-5015 F2 exencephaly panels could be compared to each other to test what effect, if any, PMD #5015 had on the F2 exencephalic embryos' genotypic ratios. The last part of 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 in exencephaly frequency observed in the JHEX-5015 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 of the segregation ratios would be interpreted to infer the possible mechanism of interaction present between the diet and gene(s). 51 Mice, breeding design, scoring exencephaly The descriptions of inbred strains used and their standard conditions, along with the breeding design and scoring exencephaly have been previously described in General Materials and Methods. DNA preparation DNA from the EX-5001 F2 exencephalic panel ("EX-5001") had previously been prepared from stored frozen embryos by standard phenol chloroform extractions and banked (See General Materials and Methods). DNA from the JHEX-5015 F2 exencephalic panel ("JHEX-5015") and the normal control panels was prepared from stored frozen embryos by using the QIAamp DNA Mini kit (See General Materials and Methods). Rationale for choice of markers used For the EX-5001 study, the markers used on Chr 13, 5, and 11 were chosen because the Exen loci were originally mapped there. For the JHEX-5015 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 of Exenl being located there. Further on in this study, additional markers, FgdS-C/D, D13MU193, D13MU30, and D13MU76, were typed on the JHEX-5015 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 JHEX-5015 panel. The Chr 7 markers used for the EX-5001 and JHEX-5015 studies were chosen to revisit an old result observed in the previous F2 sire screen (Juriloff et al., 2001). In that study, D7MU75 appeared to approach significance. The Chr 10 and 17 markers used for both EX-5001 and JHEX-5015 studies were also chosen to revisit regions significant in the F2 sire screen (Juriloff et al., 2001). The Chr 19 marker used in both studies, D19MU68, was chosen because our collaborator, Dr. Dixie Mager, mapped a "master" MusD element in SELH/Bc to Chr 19 at 5.7 Mb (UCSC online genome browser) (Personal communication from Dr. Dixie Mager). D19MU68 is approximately located at 3.5 Mb from the centromere, 2.2 Mb (1-2 cM) proximal to the MusD element. The hypothesis for looking at this marker was that the MusD element contributes to the risk of exencephaly and causes genetic instability in SELH/Bc. For the EX-5001 panel, another Chr 19 marker, Capnl-A/B at 5.8 Mb, was used to confirm the results of D19MU68. PCR and electrophoresis The genotypes of the F2 exencephalic embryos in both new panels were identified by PCR. The products were resolved by electrophoresis on 4% NuSieve 3:1 Agarose gels, stained by ethidium bromide, and photographed over UV light (See General Materials and Methods). 53 Statistical Methods To compare segregation ratios at genetic markers between the EX-5001 and JHEX-5015 studies the chi-square test of independence was used (http://www.georgetown.edu/faculty/ballc/webtools/web_chi.html). To compare observed segregation ratios against a 1:2:1 expected SS:SL:LL 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 IC calculator. A technique developed by R. A. Fisher was used to combine probabilities of 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 of Exenl and Exen2 were further supported by the genotypes of the F2 exencephalic embryos in the EX-5001 study. All the Chr 13 and Chr 5 markers used in this study gave segregation ratios that significantly deviated from the random distribution of 13 SS:26 SL:13 LL, showing an excess of SELH/Bc alleles across these loci. The x2 values for genotypic ratios ranged from 9.9 - 27.0 for markers on mid Chr 13 (See Table 4.1). The Chr 13 markers spanned 59.6 Mb (30 cM), from Fgd3-C/D (47.8 Mb) to D13MU76 (107.4 Mb) (See Fig. 5.2 on pg. 79 for map of Chr 13). The most significant Chr 13 markers, Fgfr4-E/F, D13MU13, and Gprk6-C/D, gave segregation ratios of 29:14:9 (x2^ 27.0; p< .001) (See Table 4.1) showing an excess of SS 54 Table 4.1: Genotype summaries of F2 exencephalic embryos and their corresponding values at various genetic markers in the EX-5001 study. Marker #SS #SL #LL X2 value P-value Fdg3-C/D 28 15 9 23.19 0.00001 Fgfr4-E/F 29 14 9 26.46 0.000002 Gprk6-C/D 29 14 9 26.46 0.000002 D13MM3 29 14 9 26.46 0.000002 Fancc-C/D 28 16 8 23.08 0.00001 Ptd-A/B 28 16 8 23.08 0.00001 D13MH193 21 26 5 9.84 0.007 D13MH30 21 27 4 11.19 0.004 D13MH76 22 24 6 10.15 0.006 D5MH95 29 20 3 28.76 0.000001 D5Mit30 28 21 3 25.96 0.000002 Gats-C/D 28 20 4 24.92 0.000004 D11Mit14 15 30 7 3.70 0.157 D11Mit10 16 28 8 2.76 0.252 Scn4a-C/D 17 27 8 3.19 0.203 D10Mit164 12 26 14 0.16 0.923 D10Mit180 11 28 13 0.46 0.795 D17MH10 13 21 18 2.88 0.237 D7MH75 22 20 10 8.30 0.016 Rshl1-C/D 20 22 10 5.08 0.079 D7MH79 18 27 7 4.73 0.094 D19MH68 3 24 25 18.92 0.00008 Capn1-A/B 4 23 25 17.66 0.0001 homozygotes. The ratio of SS:SL was approximately 2:1 from markers Fdg3-C/D -Ptchl-A/B, supporting previous data showing that SS and SL were not equivalent in effect on exencephaly and that Exenl appears to be semidominant (Juriloff et al., 2001). Interestingly, the 2:1 SS:SL ratio disappeared at D13MU193 (21:26:5) as the number of 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 (x2=9.9; p< 0.01) from D13MU193 was still highly significant showing considerable deviation from randomness. At D13MU13, the normal control panel gave a segregation ratio of 7:17:12, not straying far from Mendelian segregation (X2= 1.50; p> 0.1) (See Table 4.2). There appears to be a slight under-representation of SS homozygotes, but we would expect this as a number of SS homozygotes at the Exenl locus become exencephalic. As expected, the F2 exencephalic panel and normal panel were significantly different from each other at D13MU13 (x2= 11.64; p< 0.01). The genotype distributions from the distal Chr 5 markers significantly deviated from randomness in the EX-5001 F2 exencephaly panel providing strong evidence for the location of the Exen2 locus. The three markers spanned 9.4 Mb, from D5MU95 (122.4 Mb) to Gats-C/D (131.8 Mb). D5MU95 gave a segregation ratio of 29:20:3 (SS:SL:LL) (X2^ 29.0;p< 0.001), and D5MU30 and Gats-C/D had segregation ratios of 28:21:3 and 28:20:4, respectively (x2 = 26.0, 25.0; p < 0.001) (See Table 4.1 on pg. 55) (See Fig. 5.2 on pg. 79 for map of Chr 5). Here, at the Exen2 locus, the significance is coming from the excess of SS homozygotes and the deficiency of LL homozygotes. In the normal control panel, D5MU95 gave no significant deviation from random segregation, the ratio being 7:17:12 (x2= 1.50; p> 0.1) (See Table 4.2). Expectedly, at D5MU95, the F2 exencephaly 56 Table 4.2: Genotype summaries of control normal F2 embryos and their corresponding x2 values at selected genetic markers in the EX-5001 study. Marker SS:SL:LL ratio X2 value, p-value D13MU13 7:17:12 1.50, p> 0.1 D5MU95 7:17:12 1.50. p> 0.1 D7MU75 3:22:11 5.33, p> 0.05 DllMitlO 9:18:9 0.00, p= 1.0 D19MU68 12:18:6 2.00, p> 0.1 57 panel and the F2 normal control panel gave significantly different segregation ratios (x2= 16.73;/>< 0.001). Whether the ExenS locus on Chr 11 contributed to the risk of exencephaly was not resolved by this study. The markers spanned 7.6 Mb (7 cM) from D11MU14 at 98.4 Mb to Scn4a-C/D at 106.0 Mb. In the EX-5001 F2 exencephaly panel, D11MU14, DUMitlO, and Scn4a-C/D gave segregation ratios of 15:30:7, 16:28:8, and 17:27:8, respectively (x2 = 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 of SELH/Bc alleles at this locus (e.g. D11MU14 - 60 S:44 L). If any, the putative effect of Chr 11 was too weak to be statistically significant in this sample size. At DUMitlO, the F2 normal control panel gave a segregation ratio of 9:18:9 showing complete Mendelian segregation (x2= 0.00) (See Table 4.2 on pg. 57). The F2 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 in the risk of exencephaly (See Table 4.1 on pg. 55). In the EX-5001 exencephaly panel, D7MU75 gave a segregation ratio of 22:20:10 showing an excess of SS homozygotes at this locus (x2 = 8.3;p< 0.025). The two other Chr 7 markers, Rshll-C/D and D7MU79, gave segregation ratios of 20:22:10 and 18:27:7 (x2 = 5.1 and 4.8) showing that if there is an exencephaly-risk locus on Chr 7, it appears to be closer to D7MU75. Chr 7 markers spanned 16.9 Mb (14 cM) from D7MU75 at 5.9 Mb to D7MU79 at 22.8 Mb. In the normal control panel D7MU75 gave a segregation ratio of 3:22:11 suggesting a deficiency of SS homozygotes (x2= 5.33; p> 0.05) (See Table 4.2 on pg. 57). The F2 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 in the EX-5001 F2 exencephaly panel. The markers spanned 2.4 Mb from D19MU68 at 3.4 Mb to Capnl-A/B at 5.8 Mb. D19MU68 gave a segregation ratio of 3:24:25 (x2 = 19.0; p< 0.001) and Capnl-A/B gave a segregation ratio of 4:23:25 (x2 - 17.6; p< 0.001) showing a severe deficiency of SS homozygotes and a strong excess of LL homozygotes (See Table 4.1 on pg. 55). Surprisingly, this result suggests that this region on Chr 19 from LM/Bc may contribute to the risk of exencephaly in F2 embryos. At D19MU68, the normal control panel gave a segregation ratio of 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 of SS homozygotes (See Table 4.2 on pg. 57). As expected, the F2 exencephalic panel and normal control panel were significantly different from each other (X2=15.51;/>< 0.001). Little statistical support was present for any role of genes on Chr 10 and Chr 17 marked by D10MU164, D10MU180, and DI 7MU10. They gave segregation ratios of 12:26:14,11:28:13, and 13:21:18, respectively (x2= 0.16, 0.46, and 2.88) (See Table 4.1 on pg. 55). All markers segregated randomly in the F2 exencephalic embryos and therefore showed no evidence of involvement in the risk of exencephaly. 59 The JHEX-5015 study The expected random segregation ratio of genotypes of the 34 F2 exencephalic embryos was 8.5:17:8.5 (SS:SL:LL) for all genetic markers. The seven Chr 13 markers spanned 59.6 Mb (30 cM) from Fgd3-C/D at 47.8 Mb to D13MU76 at 107.4 Mb. In the JHEX-5015 F2 exencephaly panel, Fgd3-C/D, the most proximal Chr 13 marker used, gave a segregation ratio of 14:15:5 suggesting an excess of 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 of 13:15:6 3.36;/?> 0.1). D13MU193 gave a segregation ratio of 14:14:6 (3^= 4.83;p< 0.09), D13MU30 gave a segregation of 15:12:7 6.7; p< 0.05), and D13MU76 gave a segregation ratio of 15:11:8 (£= 7.12;p< 0.05), all showing an excess of SS homozygotes. The Chr 13 markers used in this study that did not significantly deviate from random showed a trend towards an excess of SS homozygotes and their lack of statistical significance was most likely due to the relatively small sample size. D13MU30 and D13MU76 were the only Chr 13 markers that gave segregation ratios that significantly deviated from random. The F2 normal control panel gave a random segregation ratio of 7:20:12 at Ptchl-A/B (x2= \ .2>\\p> 0.5) (See Table 4.4). The segregation ratios of 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 (x2= 4.19; /?> 0.1). Interestingly, the Exen2 locus on Chr 5 appeared to not be involved in the risk of exencephaly in the JHEX-5015 F2 exencephaly panel when mice are fed PMD #5015. D5MU95 gave a segregation ratio of 8:20:6 (3^= 1.30; p> 0.5) and D5MU30 and D5MU168 gave segregation ratios of 9:18:7 and 10:16:8, respectively (x2^ 0.35;p> 0.5), 60 Table 4.3: Genotype summaries of F2 exencephalic embryos and their corresponding x2 values at various genetic markers in the JHEX-5015 study. Marker #SS #SL # LL X2 value P-value Fgd3-C/D 14 15 5 5.24 < 0.089 Fgfr4-E/F 13 15 6 3.36 0.186 Ntrk2-C/D 13 15 6 3.36 0.186 Ptd-AJB 13 15 6 3.36 0.186 D13Mit193 14 14 6 4.83 0.089 D13Mit30 15 12 7 6.70 < 0.05 D13MH76 15 11 8 7.12 <0.05 D5MH95 8 20 6 1.30 0.522 D5MH30 9 18 7 0.35 0.839 D5MM68 10 16 8 0.35 0.839 D7Mit75 12 18 4 3.88 0.144 Rshl1-C/D 10 19 5 1.94 0.379 D7MH79 10 18 6 1.06 0.589 D11Mit10 8 20 6 1.30 0.522 Scn4a-C/D 11 18 5 2.24 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 Table 4.4: Genotype summaries of control normal F2 embryos and their corresponding x2 values at selected genetic markers in the JHEX-5015 study. Marker SS:SL:LL ratio X2 value Ptcl-A/B 7:20:12 1.31, p> 0.5 D5MU95 7:15:17 7.21, p< 0.05 D7MU75 5:23:11 3.10, p> 0.1 Scn4a-C/D 13:21:5 3.51,p>0.1 D19MU68 6:15:18 9.46, p< 0.025 61 all showing random segregation of LM/Bc and SELH/Bc alleles at this locus (See Table 4.3). This suggested that the effect from PMD #5015 replaced the effect from the Exen2 alleles from SELH/Bc in terms of contribution to the risk of exencephaly. Interestingly, the normal control panel gave a segregation ratio of 7:15:17, showing an excess of LL homozygotes (%2= 7.21; p< 0.05) (See Table 4.4). The F2 exencephalic panel and normal control panel were almost significantly different from each other (3^= 5.73;p< 0.09). In the F2 exencephaly panel, Chr 11 markers, DllMitlO and Scn4a-C/D, gave segregation ratios of 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 F2 normal control panel, Scn4a-C/D gave a segregation ratio of 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 (x2= 0.06; p> 0.9). The proximal Chr 7 markers, D7MU75, Rshll-C-D, and D7MU79, gave segregation ratios of 12:18:4,10:19:5, and 10:18:6, respectively (x2= 3.88; 1.94, and 1.06; p> 0.1) in the JHEX-5015 exencephaly panel (See Table 4.3 on pg. 61). D7MU75 suggested an excess of 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 of exencephaly in the F2 embryos. In contrast, the normal control panel gave a segregation ratio of 5:23:11 at D7MU75, suggesting a deficiency of SS homozygotes (3^= 3.10; p> 0.1) (See Table 4.4 on pg. 61). The F2 exencephalic and normal control panels were significantly different from each other at this locus (x2= 6.45; p< 0.05). 62 Chr 10 and 17 markers, D10MU180 and D17MU10, gave random segregation ratios suggesting no involvement in the risk of exencephaly in the F2 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 of 4:16:14 (j^= 6.00; p< 0.05), showing a deficiency of SS homozygotes and an excess of LL homozygotes in the F2 exencephalic embryos (See Table 4.3 on pg. 61). As found in the EX-5001 study, this result suggested that this region of Chr 19 from LM/Bc may contribute to the risk of exencephaly in F2 embryos. Surprisingly, however, the F2 normal control panel gave a similar segregation ratio of 6:15:18, also showing an excess of LL homozygotes (x = 9.46; p< .025) (See Table 4.4 on pg. 61), suggesting segregation distortion. The segregation ratios at D19MU68 from the F2 exencephalic panel and normal control panel were similar to each other (yr2= 0.59; p> 0.975). EX-5001 vs. JHEX-5015 For Chr 13, the segregation ratios from the EX-5001 and JHEX-5015 F2 exencephalic panels showed an excess of SELH/Bc alleles at the Exenl locus, most of the significance coming from a surplus of SS homozygotes. The markers used on the two F2 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 of exencephaly from Exenl is independent of diet. 63 The segregation ratios from the markers on Chr 5, on the other hand, were very different from each other between the EX-5001 and JHEX-5015 F2 exencephalic panels. For D5MU95, the genotype distribution for the EX-5001 panel was 29:20:3 (SS:SL:LL), significantly different from 8:20:6 from the JHEX-5015 panel (x2= 9.57; p< 0.01). For D5MU30, the segregation ratios from EX-5001 and JHEX-5015 were 28:21:3 and 9:18:7, respectively (x2= 8.18; p< 0.025). The JHEX-5015 segregation ratios were not significantly different from Mendelian segregation, where as the EX-5001 segregation ratios showed a great excess of SS homozygotes at the Exen2 locus. This difference between segregation ratios of the two F2 exencephalic panels suggests that the Exen2 alleles from SELH/Bc had little or no role in risk of exencephaly on PMD #5015 and had a strong role on PLRD #5001. Comparing the markers on Chr 11 between the EX-5001 and JHEX-5015 panels, the segregation ratios from DUMitlO and Scn4a-C/D were very similar to each other (X2= 0.54 and 0.01; p> 0.9). Scn4a-C/D, in both panels, approached significance and showed an excess of SELH/Bc alleles at the Exen3 locus, following the same trend as the TG Panel (Juriloff et al., 2001). This suggested that there was a small effect coming from the Exen3 alleles from SELH/Bc, but the individual sample sizes were not big enough to be statistically significant. Because all three F2 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 of the independent x2 tests to create an overall test. Although this test did not provide a statistically significant x2 value (x2== 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 (x2= 1.94, 1.52, and 0.41, respectively; p> 0.1). This provides support that a locus on Chr 7 (Exen4) contributes to the risk of exencephaly on both diets as the two panels show significant deviations from randomness, showing an excess of SS homozygotes. The Chr 19 marker, D19MU68, gave similar segregation ratios in the EX-5001 and JHEX-5015 F2 exencephalic panels (x2= 0.59; p> 0.5) showing deficiencies of SS homozygotes and excesses of LL homozygotes. However, the control F2 panel for the JHEX-5015 F2 exencephaly panel showed an excess of LL homozygotes for this region so this remains unresolved. The Chr 10 marker, D10MU180, and the Chr 17 marker, Dl 7Mill 0, gave segregation ratios that did not significantly deviate from randomness in the two F2 exencephalic panels (x2= 0.57 and 0.12, respectively; p< 0.9). This suggests that these regions on Chr 10 and 17 are not involved in the risk of exencephaly in the F2 embryos on PLRD #5001 or PMD #5015. Discussion The first major finding was that the hypothesis that one of the Exen loci interacts with PMD #5015 was supported by this study. The segregation ratios of the F2 exencephalic embryos at the Exen2 locus on Chr 5 significantly differed between the EX-5001 and JHEX-5015 studies. In the EX-5001 study, the Chr 5 markers gave non-random genotype distributions showing a huge excess of SS homozygotes suggesting 65 involvement in the risk of exencephaly in the F2 embryos. In contrast, the genotype distributions of the Chr 5 markers were quite random in the JHEX-5015 study showing no evidence of involvement in the risk of exencephaly when mice are fed PMD #5015. This difference between the two panels indicates a possible Exen gene-diet relationship and suggests that the diet effect from PMD #5015 supersedes the effect of the Exen2 alleles from SELH/Bc on Chr 5, therefore making the Chr 5 genotype virtually irrelevant. The dropping out of the Exen2 locus when mice are fed PMD #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 SELH/Bc would have a greatly elevated risk of exencephaly on PMD #5015, shifting the other loci to relatively less importance. We 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 Chr 10 or 17 with a weak effect in the EX-5001 study would interact with PMD #5015 and have a strong effect in the JHEX-5015 study. This would lead to that locus having a significantly more skewed segregation ratio on PMD #5015 than on PLRD #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 PMD #5015 adds to the genetic effect on risk of exencephaly. In this case, the segregation ratios for the EX-5001 and JHEX-5015 F2 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 JHEX-5015 study. A fourth alternative hypothesis was that heterozygosity across the Exen loci, combined with the effect from PMD #5015, would add more risk of exencephaly than heterozygosity itself on PLRD #5001. More embryos that are heterozygous across the Exen loci would become exencephalic. We would expect to see an increase in heterozygous genotypes across the Exen loci in the exencephalic embryos in the JHEX-5015 study. In this case, we did not see this pattern. In summary, the best fitting explanation is that effect from PMD #5015 superseded the effect of the Exen2 alleles from SELH/Bc. A second major finding in this study was that a locus on Chr 7, now called Exen4, appeared to contribute to the risk of exencephaly in the F2 embryos. Both F2 exencephaly panels in the EX-5001 and JHEX-5015 studies gave segregation ratios that showed an excess of SS homozygotes at this locus. In the previous F2 sire screen, this region from Chr 7 seemed to have an Exen locus as the /?-value approached statistical significance, but this finding was not supported in the original TG F2 exencephaly panel. Despite this, the same Chr 7 region was revisited in this study and showed that it did have a role in the risk of exencephaly. More interesting is that the role of the Exen4 locus appeared to be more important than the role of the Exen3 locus. This changes the genetic explanation of exencephaly in the SELH/Bc strain, as there are possibly four genes, the most important being on Chr 13, 5, and 7. The third finding in this study was that a region on Chr 19 from LM/Bc may contribute to the risk of exencephaly in F2 embryos. The two F2 exencephaly panels in 67 the EX-5001 and JHEX-5015 studies showed huge excesses of LL homozygotes and severe deficiencies of SS homozygotes at proximal Chr 19 markers. The data is hard to interpret though as the F2 normal control panel for the JHEX-5015 study showed a similar trend therefore indicating no role for this chromosomal region, but rather segregation distortion. On the other hand, the F2 normal control panel for the EX-5001 study showed a random distribution of genotypes at the same marker indicating this region may have a role when compared to the EX-5001 F2 exencephaly panel. Although the LM/Bc strain is not focused on in this study in regards to whether LM/Bc mice contribute anything to the risk of exencephaly in the F2 studies, it is interesting nonetheless. Therefore, further work is needed to determine the role, if any, of this region from Chr 19 in the risk of exencephaly in the F2 embryos. 68 Chapter 5: Congenic line exencephaly frequencies Introduction Congenic lines are a powerful tool for genetic analysis of a complex trait. They can be used to precisely map the individual contributing genes and, in addition, they isolate the individual physiological and developmental effects of the gene. As the previous genetic analyses indicated that there are about three additive genes of moderate effect that contribute to exencephaly in the SELH/Bc 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 my study were established approximately 5 years ago. One set of congenic lines had been created by transferring the chromosomal segments that contain the Exenl (Chr 13) and Exen2 alleles from SELH/Bc (Chr 5), respectively, from SELH/Bc into the normal LM/Bc strain background (See Fig. 5.1). The formal names, according to mouse c c nomenclature rules, would be LM-SELH-Tixeni and LM.SELH-/ixe«2 (http://www.informatics.jax.org/mgihome/nomen/strains.shtml). For convenience these lines are referred to as "13S/LM" and "5S/LM." The reverse congenic lines had been created by transferring the chromosomal segments that contain the normal Exenl and Exenl alleles from LM/Bc into the SELH/Bc background (See Figure 5.1). Their formal names would be SELH.LM-ExenlL and SELH.LM-Exen2L and are referred to as "13L/SELH" and "5L/SELH," respectively. For the 13S/LM and 13L/SELH 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 Mb on mid Chr 13. For the 5S/LM and 5L/SELH congenic lines, the markers used for selection to transfer the chromosomal 69 Figure 5.1: A conceptual picture of the construction of the congenic lines. Exenl (Chr 13) and Exenl (Chr 5) alleles from SELH/Bc were transferred into the LM/Bc background and vice versa. Black bars indicate the 19 SELH/Bc autosomes and the white bars indicate the 19 LM/Bc autosomes. SELH/Bc LM/Bc I OOfflooooD Chr 5 Chr 13 Exenl Exenl i i Chr 5 Chi-13 5S/LM illlVllllllll IllllllllllHflllli BDODDD 13S/LM "5L/SELH 13L/SELH 70 region of interest that contained the Exen2 locus were D5MU95, D5MU30, and D5MU168, spanning at least 12 Mb on distal Chr 5. At the time the congenic lines were being constructed one of the purposes was to confirm that the Exenl locus is on mid Chr 13 and that the Exen2 locus is on distal Chr 5. This purpose has been made redundant by the new EX-5001 F2 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 of the liability to exencephaly from SELH/Bc will express part of the exencephaly frequency of SELH/Bc. In addition, this model predicts each of the four congenic lines will differ from the background strain in exencephaly frequency, demonstrating the additivity of the Exen loci. In particular, LM/Bc has 0% exencephaly; substituting the Exenl or Exen2 alleles from SELH/Bc ("Exenl " or "Exen2 " respectively) into the LM/Bc background would make LM/Bc have some exencephaly. Reciprocally, in the reverse congenic lines, transferring the normal Exenl and Exen2 alleles from LM/Bc into the SELH/Bc background would reduce the frequency of exencephaly compared to that of SELH/Bc itself. Although additive, the Exen loci may not be equal in effect. The Exenl locus appeared to be the major locus relative to the other participating loci in previous studies (Juriloff et al., 2001). Based on this, 13S/LM would produce more exencephaly than 5S/LM and, likewise, 5L/SELH (with Exenl alleles from SELH/Bc) would produce more exencephaly than 13L/SELH. A third main use of the congenic lines was to isolate the individual effect of each Exen locus in the process of cranial neural fold elevation in early embryos and it was 71 hypothesized that these effects would be detectable in their morphological characteristics (See Chapter 7). A fourth use of the congenic lines became apparent when the diet effect was discovered (See Chapter 3 for details of the diet effect), to test whether the congenic lines respond to diet. The congenic lines are useful in examining if the diet effect requires a SELH/Bc mother (SELH/Bc maternal effect) and if certain Exen loci respond to diet. The data from the congenic lines will be helpful in interpreting the mechanism of the diet effect observed in the SELH/Bc strain and in the F2 exencephaly panels. This chapter reports on the use of the 13S/LM, 5S/LM, 13L/SELH, and 5L/SELH congenic lines to confirm the chromosomal locations of the Exenl and Exen2 loci, to test the fit of observed exencephaly frequencies against the multifactorial threshold model, and to test the genetic basis involved in elevated frequencies of exencephaly on diet PMD #5015. In addition, this chapter reports on a new congenic line, 7S/LM, that isolates the effect of the Exen4 alleles from SELH/Bc (See Chapter 4), as well as the improved definition of the chromosomal segments transferred into each congenic line by typing of additional flanking and intermediate markers. Materials and Methods Mice All mice originated from and were maintained in our animal unit in the Department of Medical Genetics at the University of British Columbia under standard conditions previously described (See General Materials and Methods). All breeding colonies that produced adult mice used in this study were maintained on Purina 72 Laboratory Rodent Diet #5001 ("PLRD #5001"). During these experiments, 13L/SELH was at N5F7-N5Fi2, 5L/SELH was at N8F5-N8F9, 13S/LM was at N6F7-N6Fio, 5S/LM was at N7F4-N7F7, 13S/LM* was at N7F4-N7F5, and 7S/LM was at N7F4-N7F5. These congenic lines were homozygous for their respective transferred chromosomal regions. According to Silver, the fraction of background loci that are still heterozygous at the Nth generation can be calculated as [(l/2)N~l], with the remaining fraction [1 - (l/2)N~ l] homozygous for the inbred strain allele (Silver, 1995). Using this calculation, 13L/SELH (N5) is identical to SELH/Bc across approximately 94% of the genome. 5L/SELH (N8) is approximately 99% identical to SELH/Bc. The 13S/LM congenic line (N6) is identical to LM/Bc across 97% of the genome and 13S/LM*, 7S/LM, and 5S/LM (all at N7) are approximately 98% identical to LM/Bc. Experimental strategy The first part of this study involved defining the differential chromosomal segments that were transferred into the SELH/Bc and LM/Bc strain backgrounds. This involved typing DNA with flanking and intermediate markers of the transferred chromosomal segments. The DNA was from mice of the same lineage the embryo collections were from. To test the multifactorial threshold model, exencephaly frequencies for each congenic line were observed on PMD #5015 to maximize the frequency of exencephaly in each line to obtain interpretable frequencies in reasonable sample sizes. This study was done in part of two sets in 2003 and 2004. In general, 6-16 males were used per line to generate embryos. 73 To test for diet effects in the congenic lines, the other part of Set 1 (see above) was collected concurrently on PLRD #5001 in 2003 to examine the genetic basis for the diet effect observed in the SELH/Bc strain and F2 exencephaly studies. Based on the data in Set 1, we hypothesized that the presence of Exenl alleles from SELH/Bc ("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 of the lines because their exencephaly frequencies had been too low to be interpretable, and to confirm the pattern of data found in Set 1. Construction of the 7S/LM and 13S/LM* congenic lines While this study was in progress, we obtained evidence for the Exen4 locus on Chr 7 from the EX-5001 F2 exencephaly panel (See Chapter 4). In the construction of the 13S/LM congenic line, proximal Chr 7 from SELH/Bc had been carried along by purposeful selection in the backcross generations and later left to chance. Therefore it was possible that proximal Chr 7 from SELH/Bc was still present. After typing the 13S/LM 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 SELH/Bc as well as the Exenl alleles from SELH/Bc. An experiment was done to separate the loci into two distinct new congenic lines, referred in our lab as "7S/LM" and "13S/LM*1." To create the 7S/LM and 13S/LM* congenic lines, the 13S/LM congenic line was backcrossed to LM/Bc. For the 7S/LM congenic line, the markers used for selection to keep the SELH/Bc chromosomal region of 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 alleles from SELH/Bc 74 sure the Chr 13 segment from SELH/Bc was not carried through. Reciprocally, to make the 13S/LM* line, the SELH/Bc alleles at Chr 13 markers were selected for and the SELH/Bc alleles at Chr 7 markers were selected against. For the 7S/LM and 13S/LM* congenic lines, the mice with the correct haplotypes for Chr 7 and 13, respectively, were intercrossed to produce homozygous progeny who maintained the lines. The 7S/LM and 13S/LM* congenic lines were made homozygous after the N7 generation. Defining the transferred chromosomal segments After the congenic lines were created, further definition of the exact segments of chromosome that were transferred in each congenic line was necessary. For 13S/LM and 13L/SELH, markers D13MU117, D13MU91, Fgd3-C/D, Fgfr4-E/F, Gprk6-C/D, Lect2-C/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 of markers). For the 5L/SELH and 5S/LM 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 Chr 7 segments in the 7S/LM and 5L/SELH congenic lines. Two to four mice were typed for each marker in each congenic line. DNA was prepared from tail tips using the QIAamp DNA Mini-kit and the genotypes of the mice were identified by PCR using the above markers as described in General Materials and Methods. Primer sequences can be found in Appendix 1 for primers that were designed in our lab. 75 Testing the multifactorial threshold model The female mice were fed PMD #5015 to maximize the exencephaly frequencies in attainable sample sizes for each congenic line. Sample sizes were limited by availability of mice and a target of 150-200 embryos per line was used. Based on the binomial distribution, this sample size would have the power to detect a frequency of 2% exencephaly or greater in 95% of samples [(1 - 0.02)150= 0.048] (Sokal and Rohlf, 1995). In x2 tests, a sample size of 150 would detect reductions in exencephaly frequency from 25% in SELH/Bc to less than 15% atp< 0.05. The females were placed on PMD #5015 at the same time as introduction of the males and maintained on this diet until autopsy for embryo collection. This experiment was done in 2003 and 2004, to make up parts of Set 1 and 2. Testing for diet effects in the congenic lines The congenic lines, 13S/LM, 5S/LM, 13L/SELH, and 5L/SELH, were used to test for diet effects in Sets 1 and 2. For embryo collection on Purina Lab Rodent Diet #5001 ("PLRD #5001), the female mice were fed PLRD #5001 from weaning until autopsy. For embryo collection on Purina Mouse Diet #5015 ("PMD #5015"), the females were placed on PMD #5015 at the same time as introduction of males and maintained on it until autopsy. In general, most (>80%) of the females were on PMD #5015 for up to one week before conception. Mice taking longer to become pregnant typically were on PMD #5015 for two weeks. 76 Scoring exencephaly Pregnancies were judged by eye and palpation. Developmental stage of embryos was judged by morphological characteristics. Females were euthanized by carbon dioxide on day 14 of gestation (range El 1-E18) for examination of embryos. The uterus was removed, pinned to a black wax substrate, immersed in physiological saline (0.85% NaCl) solution, and cut open to reveal the conceptuses. Dead post implantation embryos ("moles") were recorded. The morphology of scoreable embryos was examined under a dissection microscope and recorded. The following defects were screened: exencephaly, spina bifida, rachischisis, major abnormalities of the limbs and tail, gastroschisis, cleft lip, and severe edema. The litters were stored in glass jars, in either Bouin's fixative or 10% Buffered Formalin Acetate. Statistical methods To compare the exencephaly frequencies on the two diets in each congenic line, the chi-square test of 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 of rs near one indicates a strong relationship, 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 Chr 13 segment transferred from SELH/Bc into the LM/Bc background (13S/LM and 13S/LM*) spanned at least 87.6 Mb (50 cM) from D13MU3 (19.8 Mb) to D13MU76 (107.4 Mb). Additional markers between D13MU3 and D13MU76 confirmed that the whole segment was SELH/Bc homozygous (See Figure 5.2). The chromosomal regions that flank D13MU3 and D13MU76 were not typed and therefore their genotypes are not known. The Chr 13 segment transferred from LM/Bc into the SELH/Bc background (13L/SELH) 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 LM/Bc (See Figure 5.2). The SELH/Bc Chr 5 segment transferred into the LM/Bc background (5S/LM) used the markers D5MU95 (122.4 Mb), D5MU30 (127.1 Mb), and D5MU168 (134.5 Mb), spanning 12.1 Mb (10 cM), to select for the SELH/Bc segment. Typing of additional markers showed that the segment spanned at least 76.2 Mb (44 cM) from D5Nds2 (71.4 Mb) to D5MU122 (147.6 Mb) (See Figure 5.2). The proximal flanking marker, D5MU13 (35.9 Mb) was back to LM/Bc homozygosity. The LM/Bc Chr 5 segment transferred into SELH/Bc (5L/SELH) spanned at least 12.1 Mb (10 cM) from D5MU95 (122.4 Mb) to D5MU168 (134.5 Mb). The next closest proximal and distal flanking markers at 109.7 Mb (D5Mit24) and 147.6 Mb (D5MU122) were back to SELH/Bc homozygosity, showing a more precise area where the Exen2 locus is located, compared to the 5S/LM line. As LM/Bc markers at proximal Chr 7 had been purposely selected in early generations of creation of the 5L/SELH congenic line, this line was typed for Chr 7 78 Figure 5.2: The current chromosomal segments that are in each congenic line. White regions indicate homozygous LM/Bc regions, black regions indicate homozygous SELH/Bc regions, and the hatched regions are unknown. Length of each chromosome is indicated at bottom in Mb. Locations of the markers can be found in Table 2.1 on pg. 30. VO 13L/SELH 13S/LM' D13MH3-Fgd3-C/D\ Fgfr4-E/F \ Gprk6-C/D \ D13Mit13^ Ptc1-A/B -D13Mit193 -D13Mit30-D13Mit76 Chr 13 116.5 Mb D13MH3 D13MH13 Rd-A/B D13MH76 5L/SELH 5S/LM • LL D5MH13 D5Nds2 D5MH24 D5MH95 D5MH30 D5MM68 D5MH122 Chr 5 Chr 7 149.2 Mb • SS D7MH75 D7MH79 Ccne1-A/B Gnefr-C/D D7MH62 D5MH13 D5Nds2 D5MH24 D5MH95 D5MH30 D5MH168 D5MH122 Chr 5 • Not typed 7S/LM I D7MH75 D7MH79 Ccne1-A/B Gnefr-C/D D7MH62 Chr 7 133.1 Mb Scale __10 Mb (D7MU75). 5L/SELH was homozygous LM/Bc for markers on Chr 7; therefore the normal Exen4 alleles from LM/Bc were also transferred into this line along with the normal Exen2 alleles, complicating the interpretation of the data in this line. The SELH/Bc Chr 7 segment transferred into the LM/Bc background (7S/LM) and the LM/Bc Chr 7 segment found in the 5L/SELH line spans at least 16.9 Mb from D7MU75 at 5.9 Mb2 to D7MU79 at 22.8 Mb (See Figure 5.2). In the 7S/LM line, typing of additional markers showed that the segment spanned at least 20.6 Mb (14.3 cM) from D7MU75 to Ccnel-A/B. The next distal marker, Gnefr-C/D at 33.9 Mb, was back to LM/Bc homozygosity. In 5L/SELH, typing of additional markers showed that the LM/Bc Chr 7 segment spanned at least 16.9 Mb (14.3 cM) from D7MU75 to D7MU79. The next distal marker, Ccnel-A/B, as well as the other two, were back to SELH/Bc homozygosity. Testing the multifactorial threshold model Set 1 was the first formal examination of the frequency of exencephaly in 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 PMD #5015 to maximize the exencephaly frequency in 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 PMD #5015. There was good agreement between the values (exencephaly frequencies) obtained in Set 1 and Set 2 (See Table 5.1), with the exception of 13S/LM that had a significantly higher exencephaly frequency in Set 2 (x2= 5.00; p< 0.05). One explanation for this difference in 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. 80 Table 5.1: Exencephaly frequencies and post-implantation mortality rates ("moles") in the SELH/BC and LM/Bc strains and congenic lines on PMD #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 9.7A 5S/LM 12 145 7.6 134 5 3.7 7S/LM 12 143 7.0 133 0 0 2004 - Set 2 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 21.8A 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 A The exencephaly frequency for 13S/LM significantly differs between the 2 sets (x2= 11.5; p< 0.001) Note: Litters of <5 implants were deleted 81 c the 13S/LM congenic line was that Chr 7 (Exen4 ) was possibly segregating in Set 1 and, by chance, went to fixation in the SELH/Bc allele in the parents of 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 13S/LM* congenic line. Demonstrated on Table 5.1 (pg. 81), the congenic lines on the LM/Bc background each produced exencephaly, whereas LM/Bc itself produced 0% exencephaly. Substituting Exenl into the LM/Bc background caused LM/Bc to have approximately 3% exencephaly (13S/LM*). Substituting Exen2s into the LM/Bc background produced 1-4% exencephaly (5S/LM, Sets 1 and 2). This supported the prediction from the multifactorial threshold model that these congenic lines would express part of the liability to exencephaly from SELH/Bc. In the 7S/LM 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 of Exenl and Exen4 (13S/LM) produced more exencephaly than either Exen gene alone, but less than SELH/Bc (25%), also in agreement with the multifactorial threshold model. Reciprocally, the congenic lines on the SELH/Bc background produced less exencephaly than SELH/Bc itself. On PMD #5015, SELH/Bc had 25.1% exencephaly. The 13L/SELH congenic line had 0-1% exencephaly and 5L/SELH had approximately 10% exencephaly, both sets for both lines significantly less than SELH/Bc itself (All x2 values > 7.67; all /^-values < 0.01). These data showed that 13L/SELH and 5L/SELH 82 retained some of the risk to exencephaly from SELH/Bc, again in agreement with the multifactorial threshold model. When fed the lower risk diet, PLRD #5001, the congenic lines tested, 13S/LM, 5S/LM, 13L/SELH, and 5L/SELH, all produced some exencephaly (See Table 5.2), but some were at lower frequencies than on PMD #5015. Testing the genetic basis of the diet effect Diet Effects in Set 1 (Table 5.3, pg. 85) Part of Set 1 on PLRD #5001 was collected concurrently with the other part of Set 1 on PMD #5015 to look for diet effects. As expected, the SELH/Bc strain had a significantly higher exencephaly frequency on PMD #5015 (25%) than PLRD #5001 (12%) tf= 13.0, p< 0.001) (See Table 5.3). LM/Bc had 0% exencephaly on both diets. Comparing the exencephaly frequencies on both diets for each congenic line, Table 5.3 shows that the 13S/LM line had a significantly higher exencephaly frequency on PMD #5015 (10%) than on PLRD #5001 (4%) tf= A.5\;p< 0.05). The 13L/SELH 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 5L/SELH congenic line showed no significant difference between exencephaly frequencies on the two diets (x2= 0.07), with a slightly higher exencephaly frequency on PMD #5015 (7.5%) than on PLRD #5001 (6.4%). From Set 1, the congenic line data suggested that the lines with Exenls (13S/LM and maybe 5L/SELH) responded to PMD #5015 and that the congenic lines lacking Exenl did not (13L/SELH only). The 5S/LM line was not compared between the diets because there was no data collected on PLRD #5001 in this set. 83 Table 5.2: Exencephaly frequencies and post-implantation mortality rates ("moles") in the SELH/BC and LM/Bc strains and congenic lines on PLRD #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° 2004 - Set 2 13L/SELH 20 211 6.2 198 2 1.0 5L/SELH 20 192 16.1 161 5 3.1A 13S/LM 19 180 10.0 162 22 13.5 " 5S/LM 19 212 4.2 203 1 0.5 A One embryo had cleft face in this line B The exencephaly frequency for 13S/LM significantly differs between the 2 sets 9.17; p<0.01) Note: Litters of <5 implants were deleted 84 Table 5.3: Exencephaly frequencies in parental strains and congenic lines on PMD #5015 and PLRD #5001 for Sets 1 and 2. Exencephaly frequency PLRD #5001 PMD #5015 SELH/Bc 11.6 25.1 p< 0.001 LM/Bc 0.0 0.0 Set1 13L/SELH 0.0 0.8 5L/SELH 6.4 7.5 NS 13S/LM 4.0 9.7 p< 0.05 Set 2 13L/SELH 1.0 0.0 5L/SELH 3.1 9.8 p< 0.025 13S/LM 13.5 21.8 p< 0.05 5S/LM 0.5 1.1 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 13S/LM congenic line responded to diet, as in Set 1 (X2= 4.08; p< 0.05). On PMD #5015 the frequency was 22% and on PLRD #5001 the frequency was 14%. The exencephaly frequencies from the 5L/SELH 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 Exenls respond to diet. The exencephaly frequencies from the 5S/LM 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 13L/SELH were also 1% or less on both diets suggesting this line did not respond to diet. Thus both lines lacking Exenl from SELH/Bc did not appear to respond to diet. It is important to note that the 5S/LM and 13L/SELH congenic lines had very low exencephaly baseline frequencies on PMD #5015 making the detection of a decrease in exencephaly due to the PLRD #5001 diet effect difficult to detect. Post-implantation mortality rates ("moles ") and other defects SELH/Bc had a 10% post-implantation mortality rate on PMD #5015 and PLRD #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 SELH/Bc Exen alleles from SELH/Bc. Although, the post-implantation mortality rates from 13L/SELH mostly tended to be lower than SELH/Bc, rates from 5L/SELH typically tended to be higher than SELH/Bc. The higher rates in 5L/SELH can be attributed to effects of 86 specific litters. There appeared to be clustering where 1-3 litters (average 16% of litters) in each Set (divided by diet) contributed approximately half or more of the mole count. There was no consistent pattern of post-implantation mortality rates between the two diets, therefore suggesting this rate was independent of diet. LM/Bc had an approximate 4% post-implantation mortality rate on both diets (Tables 5.1 and 5.2; pgs. 81 and 84, respectively). With the exception of 5S/LM, all the other congenic' lines on the LM/Bc background tended to have higher post-implantation mortality rates than LM/Bc itself. This suggests that transferring Exen alleles from SELH/Bc into LM/Bc increases embryonic mortality. The post-implantation mortality rates on both diets appeared to be similar and there was no notable evidence of 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. On PMD #5015, the Spearman rank correlation coefficients were calculated to be 0.52 and 0.92 in Sets 1 and 2 respectively. On PLRD #5001, they were 0.88 and 0.81 in Sets 1 and 2 respectively. Three out of those 4 values were suggestive of a relationship between exencephaly and mole frequency, but only one ranking assessment was significant (0.92 in Set 2 on PMD #5015). This suggested that as exencephaly frequency goes up, so does post-implantation mortality, which in general happens before neural tube closure. Among the hundreds of embryos examined in 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 of 7S/LM. In addition, 1 embryo in the LM/Bc strain had severe edema. Cleft face was observed in one embryo in the SELH/Bc strain and in the 5L/SELH congenic line. In these two embryos the neural tube failed to close in the forebrain folds and therefore, the face could not develop normally. Gastroschisis was also observed in the cleft face embryo in SELH/Bc. Discussion The Chr 13 segment from SELH/Bc transferred into LM/Bc (13S/LM), and vice versa (13L/SELH), was made purposely big to ensure that the Exenl locus was transferred. The F2 sire genome screen pointed to a region on mid Chr 13 that likely contained the Exenl locus (Juriloff et al., 2001), yet this region was a zone of 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 F2 sire have to have the Exenl alleles from SELH/Bc to transmit exencephaly. More markers were typed on the original F2 exencephaly panel to help precisely map the Exenl locus, but the difficulty was that F2 recombinants in this region did not provide valuable information of the location of the Exenl locus because F2 embryos do not have to be SS or SL at this locus to have exencephaly. Taken together, these experiments showed a region on mid Chr 13 that had a high probability of containing the Exenl locus and in order to ensure that the Exenl locus was transferred into LM/Bc and vice versa, it was necessary to transfer a segment of that size. The chromosomal segments containing the Exenl locus transferred into the Chr 5 congenic lines, 5S/LM and 5L/SELH, were not as big as the Chr 13 segments transferred 88 into the strain backgrounds. The available markers for distal Chr 5 were used to transfer the Chr 5 segments into SELH/Bc and LM/Bc 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 of the differential chromosomal segment transferred into a background strain in cM can be calculated as [200(1-2"N)/N] using one marker to select for the segment in the backcrosses. For values of N greater than five, this equation can be simplified to [200/N] (Silver, 1995). It is important to note, though, that lengths of the transferred chromosomal segments can vary greatly at the same generation given the inherently random distribution of 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 of the segment so it would look like [(200/N) + length of segment], figuring that the segment would have generally half of the expected average length (200/N) flanking it on each side. The distal Chr 5 segment transferred into the 5L/SELH congenic line (at Ng) was 10 cM - 25 cM (See Figure 5.2 on pg. 79). Using the adapted equation, the expected average length at N8 was calculated to be 35 cM, larger than what was observed in 5L/SELH. The distal Chr 5 segment transferred into 5S/LM (at N7) was found to be longer, spanning 44 cM - 65 cM (See Figure 5.2 on pg. 79). At this generation (N7) the expected average length was calculated to be approximately 39 cM. This shows that the segments were expected to be quite large for 5S/LM and 5L/SELH. Therefore, because of random crossover sites and linkage, it is not surprising that the lengths of the transferred segments are what they are. 89 The first major finding in this study was that the congenic lines produced some exencephaly. One alternative expectation was that none of the lines would have exencephaly, but that each would have a mild delay in elevation of the midbrain folds in 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 in the Exenl or Exen2 into the LM/Bc background made LM/Bc have some exencephaly; taking out Exenl or Exenl and replacing it with the normal gene reduced the frequency of exencephaly, but did not make the SELH/Bc strain lose exencephaly completely. We did not make any Exen3 congenic lines, but the results of the F2 studies (See Chapter 4) suggested a minor contribution to the risk of exencephaly. The relative impact of the Exen4 locus was less easily interpretable because the only exencephaly observed in this line came from one sire in 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 of the effect of substituting normal Exen alleles from LM/Bc into SELH/Bc and transferring SELH/Bc Exen alleles into LM/Bc (See Fig. 5.3). The normal distribution curves are anchored on the threshold of SELH/Bc; 25% of SELH/Bc 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 SELH/Bc Exenl and Exen4 alleles (5L/SELH) shifted the normal distribution curve approximately 2/3 a to the left (Fig. 5.3). In both Sets, the exencephaly frequency from 5L/SELH was significantly higher than 13L/SELH (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 LM/Bc background SELH/Bc LM/Bc Embtyonic liability trait 13S/LM* 5L/SELH 5S/LM 0.08% 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 in standard deviation units and (5-x) is the mean of the distribution. The threshold is anchored to SELH/Bc. Set 1 Set 2 Combined X Mean of (% exen) (% exen) (% exen) distribution SELH/Bc - 25.1 25.1 0.674 4.326 (Exenl,2,3.4) 13L/SELH 0.8 0 0.3 2.748 2.252 (Exen2,3,4) 5L/SELH 7.5 9.8 9.3 1.311 3.689 (Exenl,3) LM/Bc _ 0 0.08A 3.156 1.844 (No Exens) 13S/LM* 2.9 2.9 1.881 3.119 (Exenl) 5S/LM 3.7 1.1 2.2 2.014 2.986 (ExenT) 13S/LM 9.7 21.8B 21.8 0.772 4.228 (Exenl,4) 7S/LM 0 2.8 1.4 2.197 2.803 (Exen4) A Estimated LM/Bc exencephaly frequency from all the combined studies in the lab, including my studies and past studies. B For 13S/LM 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) 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 in the context of SELH/Bc background and PMD #5015. The effect of transferring in SELH/Bc Exen genes into the LM/Bc background is harder to interpret. It is difficult to accurately estimate the exencephaly frequency in the LM/Bc strain because it is so low and small changes in frequency have a large effect of location of a strain relative to the threshold when the strain is at extremely low frequencies. Given this, doing the same sort of test against the multifactorial threshold model on the LM/Bc background is not going to be interpretable when determining the size of the effects of Exenl and Exen2 in the LM/Bc strain. The 13S/LM congenic line (containing the Exenl and Exen4 alleles from SELH/Bc) provided a fortuitous opportunity to test the multifactorial threshold model. In this model, if two genes act additively, the result, as measured in terms of frequency of exencephaly, may appear to be synergistic depending on the magnitude of the shifts and the relation of distribution to the threshold (Fraser, 1976). 13S/LM* had approximately 3% exencephaly; by adding the effect of the Exen4 alleles from SELH/Bc (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 13S/LM* curve shifts approximately 1.1 standard deviation units ("a") to the left from the 13S/LM curve and c that difference (1.1 a) is the predicted effect of Exen4 . This predicted value correlates to an approximate 2.3% predicted exencephaly frequency for the 7S/LM line (Exen4) (Falconer and Mackay, 1996). What was observed in the 7S/LM line was 1.4% 93 Figure 5.4: Additivity of the Exen loci. This figure demonstrates the additivity of the Exenl and Exen4 loci using the multifactorial threshold model. Indicated in 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 SELH/Bc and this figure is based on the LM/Bc mean (0% related to 3.090 a). LM/Bc mean (3.090) Threshold 13S/LM (22% - 0.772) 13S/LM* (3% - 1.881) Calculations 3.090a -0.772 a -2.318 a 3.090 a-1.881 a = 1.209 a 2.318 a- 1.209 a = 1.109 a Predicted effect of Exen4 • (1.109 sdu)« 2.3% predicted exencephaly frequency for 7S/LM 7S/LM (1.4%-2.197) 94 exencephaly, shifting the 7S/LM curve approximately 0.9 a from the LM/Bc mean value. This real value is in good general agreement with the predicted value, demonstrating very well the additivity of the Exen genes, again supporting the multifactorial threshold model. The third major finding of this study was the diet response observed in the 13S/LM and 5L/SELH congenic lines. 13S/LM had a significantly higher exencephaly frequency on PMD #5015 than on PLRD #5001 in both sets. Not only did this line show a diet effect, it also demonstrated that the diet effect does not require an SELH/Bc mother. Based on Set 2, 5L/SELH (containing Exenl ) also showed a diet effect supporting the hypothesis that congenic lines with Exenl respond to diet. Notably, mothers in both situations had Exenls, not just the embryos, leaving the possibility of a maternal effect of a semidominant Exenl from SELH/Bc as well as a direct effect in embryos. The interpretation for 13L/SELH and 5S/LM was less clear because of the low exencephaly frequencies observed in these lines. Low exencephaly frequencies on PMD #5015 indicate that there is little or no diet effect. We observed no big diet effects in these congenic lines and huge sample sizes would be needed to detect a small diet effect. It appeared that genotypes with Exenls (13S/LM and 5L/SELH) responded to diet and genotypes without Exenls (5S/LM and 13L/SELH) did not respond. These diet response results are complicated by the fact that both the 13S/LM and 5L/SELH lines are contaminated with other Exen alleles from either LM/Bc or SELH/Bc depending on the strain background. 13S/LM contains Exen4 , as well as Exenl , and 5L/SELH contains the normal Exen4 as well as the normal Exen2 from LM/Bc. Since Exen4s is not present in 5L/SELH, then the diet response in not likely due solely to Exen4. Taken together, the 95 13S/LM and 5L/SELH congenic lines indicate that the presence of Exenl , on either strain background, leads to a diet effect. Future work could test the 13S/LM* congenic line for a diet effect to further investigate the role of 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 of the diet effect, as the congenic lines with Exenl responded to diet, suggesting a gene-diet interaction. The use of the congenic lines proved valuable in further investigating the genetic basis of exencephaly and the diet effect in the SELH/Bc strain. 96 Chapter 6: Recombinant congenic lines Introduction Congenic lines are a valuable tool to precisely map the individual contributing genes of a complex trait. The previous chapter reported the improved definition of 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, 13L/SELH and 13S/LM, spanned at least 87 Mb on mid 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 in precisely mapping the Exenl locus. The 13L/SELH congenic line, as previously reported in 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 "SELH/Bc-like" frequencies if the Exenl alleles (in homozygous state) from SELH/Bc CExenF") were incorporated back into the line. This makes the presence of Exenl easy to detect and we can therefore deduce that the current LM/Bc chromosomal segment in that same line excludes the Exenl locus. If one of the recombinant congenic lines has the same low exencephaly frequency as 13L/SELH, then it can be inferred that the Exenl locus is still located in the LM/Bc chromosomal segment. Knowing whether Exenl was incorporated back into a recombinant congenic line/lines we can then compare the Chr 13 haplotypes of these lines, side by side, and conceptually deduce where the Exenl locus is excluded (LM/Bc regions) and where it is predicted to be (SELH/Bc regions). There are 97 three recombinant congenic lines that have been created in 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 of the recombinant congenic lines to attempt to more precisely map the Exenl locus on Chr 13. Materials and Methods Mice All mice originated from and were maintained in our animal unit in the Department of Medical Genetics at the University of British Columbia under standard conditions previously described (See General Materials and Methods). All breeding colonies that produced adult mice used in this study were maintained on Purina Laboratory Rodent Diet #5001 ("PLRD #5001"). During these experiments, 13L/Rec-Line 1 was at N8F2-N8F3 and 13L/Rec-Line 7 was at N8F2, and 13L/Rec-Line 6 was at N8F2. Experimental strategy Exencephaly frequencies for each recombinant congenic line were observed on PMD #5015 to maximize the frequency of exencephaly in 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 of the LM/Bc chromosomal segments, were 98 compared against each other. Using the exencephaly frequency data together with the haplotype data from each recombinant congenic line, we can conceptually deduce the location of the Exenl locus by comparing them against each other. Construction of the 13L/Rec lines Backcrossing the 13L/SELH congenic line back to SELH/Bc was the first step in creating the recombinant congenic lines (See Fig. 6.1). The "Fl" generation, being heterozygous on Chr 13, was then backcrossed to SELH/Bc to generate recombinants. All offspring from these crosses ("Fl" x SELH/Bc) were genotyped at selected Chr 13 markers to look for recombinants. When a recombinant was found, it was backcrossed again to SELH/Bc (See Fig. 6.1-C) to propagate this haplotype to obtain sibs of the same genotype since it was unlikely to obtain a male and female that had the same recombinant haplotype from the "Fl" x SELH/Bc cross. When sibs with the same haplotype were found ($ and 3) by 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 in exencephaly frequency studies (See Fig. 6.1-E). For 13L/Rec-Lines 1, 6 and 7, markers D13MU3, Fgd3-C/D, Fgfr4-E/F, Gprk6-C/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 of the creation of the recombinant congenic lines 13L/SELH X SELH/Bc B Chr 13 L ----L I L I L Chr 13 S S S S "F1" Chr 13 S-L-LL S.-.-L S....L s lis X SELH/Bc Chr 13 S-L-I-S s-is S .. In generations C, D, and E, the genotypes are selected based on markers and other segregants are not shown. SELH/Bc X "Rec Line A" Chr 13 siis s.-is S__ Chr 13 S L . L S S is ? Sib X S Sib Chr 13 S L . L Chr 13 siis L L $ Sib X $ Sib Chr 13 S L LI IS L L Chr 13 siis L LI "Rec Line B" X SELH/Bc Chr 13 L iis I S S....S Chr 13 Si s s s is $ Sib X S Sib Chr 13 L --I LS S....S Chr 13 Liis L _. S. I S ? Sib X S Sib Chr 13 LUL Ll S L IS Chr 13 LUL L _-S Homozygous Rec Line A Homozygous Rec Line B 100 Defining the chromosomal segments in the recombinant congenic lines After the recombinant congenic lines were created, further definition of the exact segments of chromosome that were transferred in each line was necessary to create sharper boundaries between the LM/Bc segments and SELH/Bc 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 in 13L/Rec-Line 7. Three to ten mice were typed for each new marker in each congenic line. The "non-Mit" markers were designed in our lab (See General Materials and Methods). Scoring exencephaly Scoring exencephaly has been described in the previous chapter (See Chapter 3). Tissue was kept from some exencephalic embryos from 13L/Rec-Line 1. They were individually washed in ice-cold sterile saline solution and stored in cryovials at -20°C. In addition, tail tissue was kept from 5 13L/Rec-Line 1 pregnant females that were part of this study. DNA was prepared by the QIAamp DNA Mini-kit (See General Materials and Methods) and was used to further define the chromosomal segments in this line. Statistical methods To compare the exencephaly frequencies of the recombinant congenic lines, the chi-square test of 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 LM/Bc chromosomal segment in 13L/Rec-Line 1 spanned at least 7.6 Mb (3 cM) from Fgfr4-E/F'to Ptchl-A/B (See Fig. 6.2). Additional markers were typed between Ptchl-A/B (at 61.8 Mb) and D13MU193 (88.4 Mb) to sharpen the boundary between the LM/Bc segment and the part that was back to the SELH/Bc background. Typing showed that the LM/Bc segment spanned at least 20.0 Mb (12 cM) from Fgfr4-E/F (54.2 Mb) to Nr2fl-A/B (74.2 Mb). The next distal marker, G2151-A/B (82.0 Mb), was back to SELH/Bc background and SELH/Bc homozygosity extended to the end of the chromosome, as confirmed by D13MU78 (115.8 Mb). 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 of 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 DNA from the 4 exencephalic embryos and typed them for Nr2fl-A/B. One of the 4 exencephalic embryos was found to be heterozygous at this marker and the other three were homozygous LM/Bc. Given that three of the exencephalic embryos were "LL" at this marker, it could be concluded that the Exenl locus is not located at Nr2fl-A/B. Therefore, it does not matter if some mice in this line are heterozygous at this marker. The LM/Bc chromosomal segment in 13L/Rec-Line 6 spanned at least 19.0 Mb (18 cM) 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 13L/Rec- Line 1 13L/Rec-Line6 13L/Rec- Line 7 D13Mit3 H Fgd3-C/D -Fgfr4-E/F \ Gprk6-C/D ^ D13Mit13 Ptd-A/B J D13MH193 -D13MH30 -D13MH76 Scale 10 Mb 0.3% D13MH3 Fgd3-C/D -Fgfr4-E/F \ Gprk6-C/D 24 D13MH13 Ptc1-A/B Adcy2-A/B Nkd2-A/B Nr2f1-A/B G2151-A/B Edil3-A/B D13Mit193 D13MH30 D13MH76 D13MH78 18.3% • LL D13MH3 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 D13MH193 -| D13MH30 D13MH76 -D13Mit78 _ 8.8% D13MH3 -I Fgd3-C/D Fgfr4-BF vl Gprk6-C/D ^ D13Mit13 Ntrk2-C/D Ftc1-A/B D13MH193 D13MH30 D13MH76-D13MH78. 20% SS § Not typed 51.2 Mb from Adcy2-A/B (at 64.6 Mb) to the bottom of the chromosome at D13MU78 (at 115.8 Mb). The part of Chr 13 that is back to the SELH/Bc background spanned at least 42.0 Mb (26 cM) from D13MU3 (19.8 Mb) to Ptcl-A/B (61.8 Mb). Some mice were heterozygous at Ptcl-A/B due to recombination in the differential segment while being brought to homozygosity and were used for exencephaly data due to timing constraints. The LM/Bc chromosomal segment in 13L/Rec-Line 7 spanned at least 35.8 Mb (25 cM) from D13MU3 (19.8 Mb) to D13MU13 (55.6 Mb). The next available marker distal of D13MU13, Ntrk2-C/D (at 57.9 Mb), was back to the SELH/Bc background. Typing at D13MU78 (115.8 Mb) showed that the homozygous SELH/Bc background extended down to the end of the chromosome. Exencephaly frequencies from the recombinant congenic lines As demonstrated in Table 6.1, 13L/Rec-Line 1 produced approximately 18% exencephaly in 13 litters, an "SELH/Bc-like" exencephaly frequency. This exencephaly frequency suggests that Exenl was integrated back into this line and that the LM/Bc segment, that spans at least 20.8 Mb, excludes the Exenl locus. From this line, we can deduce that the Exenl locus is either proximal or distal of the LM/Bc segment. 13L/Rec-Line 7 produced approximately 20% exencephaly, an "SELH/Bc-like" exencephaly frequency. This exencephaly frequency, as well, suggested that Exenl was integrated back into this line and that the LM/Bc segment, that spans at least 35.8 Mb from D13MU3 to D13MU13, excluded the Exenl locus and that it is located distal of 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") in the recombinant congenic lines on PMD #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 (x2=1.92; p>0.1). 105 the most accurate value. Importantly, though, this line did produce exencephaly in only two litters, more than would be expected if Exenl had not been incorporated back into this line, meaning more than would be expected for 13L/SELH. Surprisingly, 13L/Rec-Line 6 produced approximately 9% exencephaly, higher than the 13L/SELH congenic line, and possibly lower than the SELH/Bc strain. This exencephaly frequency, although possibly not an "SELH/Bc-like" frequency, suggested that the Exenl alleles from SELH/Bc were incorporated back into this line, as well, and that the LM/Bc segment that spans at least 41.5 Mb 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 As demonstrated in Table 6.1 on pg. 105, 13L/Rec-Line 1 had a 6.7% post-implantation 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. No other defects were observed in 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 of them produced "SELH/Bc-like" frequencies suggesting that the Exenl alleles from SELH/Bc were integrated back into these lines, so that the 106 recombinant congenic line mice were basically like SELH/Bc mice, having all the Exen alleles (Exenl, Exen2, ExenS, and Exen4) from SELH/Bc. 13L/Rec-Line 1 showed that the Exenl locus was located either proximal of FdgS-C/D at 47.8 Mb (31 cM) or distal of Nr2fl-A/B at 74.2 Mb, since the LM/Bc "island" is located between SELH/Bc background on either side (See Figure 6.2 on pg. 103). The use of the 13L/Rec-Line 7 line was to aid in figuring out whether the Exenl locus was located proximal or distal of the LM/Bc segment in 13L/Rec-Line 1. The 13L/Rec-Line 7's Chr 13 is divided nicely into the top half that is homozygous LM/Bc (from D13MU13 and up) and the bottom half that is homozygous SELH/Bc (from Ntrk2-C/D and down). By comparing the two Chr 13 haplotypes from these recombinant congenic lines that both produce SELH/Bc-like exencephaly frequencies, we were able to deduce that the Exenl locus was distal of 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 13L/SELH. 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 of 13L/Rec-Line 7 (See Fig. 6.2 on pg. 103). Given that 13L/Rec-Line 6 is homozygous LM/Bc 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 of 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 13L/SELH congenic line, but possibly not as 107 high as SELH/Bc. 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 SELH/Bc 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 SELH/Bc (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 of exencephaly in SELH/Bc. If the exencephaly frequency from 13L/Rec-Line 6 reflects an accurate value for that line (around 9%), perhaps this locus contributes less risk of exencephaly than the distal Exenl locus. Another possible explanation could be that the UCSC 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 of 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 of 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 of exencephaly in SELH/Bc, 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 of exencephaly when they appear to not share a homozygous SELH/Bc region. 108 Chapter 7: Neural 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 LM/Bc, SWV/Bc, and ICR/Be. In LM/Bc, 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 of the skin. Somite pairs first appear in the anterior portion of the trunk when the neural folds are becoming evident and the formation of new distinct somite pairs progresses caudally at regular intervals. Being visible on the dorsal side of an embryo under a microscope, the number of somite pairs ("somites") present is usually the best reflection of the overall development of an embryo and can be used as a measure of developmental state. During neural tube closure, the neural folds generally elevate and initiate contact at approximate somite stages. The initiation of neural tube closure (Closure 1) has been shown to require convergent extension (Reviewed in Copp et al., 2003). Keller at al. states that "convergent extension is a process in 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 of the prosencephalon and mesencephalon boundary (Macdonald et al., 1989; Juriloff et al., 1991; Gunn et al, 1995) and spreads bidirectionally (See Fig. 7.2). Initiation at Closure 2 is attributed to the appropriate timing of elevation of the prosencephalon and mesencephalon folds. For proper elevation of 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 in" and converge towards the midline (Copp et al., 2003). Closure 3 initiates at the most rostral end of the prosencephalon folds around the same time Closure 2 initiates, and spreads caudally to 111 Figure 7.2: Scanning electron micrographs of cranial neural tube closure in Day 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 of neural tube closure in the SELH/Bc strain. Panel e shows the characteristic "mad-cat" conformation in SELH/Bc due to splayed mesencephalon folds attributed to delayed elevation. In panel f, this SELH/Bc 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 ("ANP"). The last closure, Closure 4, is unique in that the neural folds remain separate and closure occurs by growth of a membrane that eventually covers the rhombencephalon (Geelen and Langman, 1977) (See Fig. 7.1). Cranial neural tube closure is typically completed by the 20-somite stage. The pattern of normal "classic" mouse neural tube closure discussed above has not been observed in the SELH/Bc mouse strain. An early developmental study found that SELH/Bc 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 of Closure 2 is thought to be due to the delayed elevation of 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 of mesencephalon neural fold elevation is considered to be the liability trait. If the timing of elevation of the mesencephalon neural folds in SELH/Bc embryos is too late, then the embryo will become exencephalic. Figure 7.2 shows the characteristic splayed mesencephalon folds that are seen in the SELH/Bc strain. SELH/Bc embryos close their neural tubes by extension of Closure 3, seen earliest at the 15-somite stage in the previous developmental study, zipping all the way around to meet Closure 4 in the rhombencephalon (See Fig. 7.2 on previous page) (Macdonald et al., 1989). All SELH/Bc embryos have this abnormal mechanism of neural tube closure and most embryos complete it successfully and live healthy lives. However, in approximately 10-30% of SELH/Bc embryos (depending on maternal diet) the mesencephalon folds remain unelevated or not elevated enough and subsequently they become exencephalic. For this study, the pattern of neural tube closure is again revisited 113 in SELH/Bc and LM/Bc, 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 LM/Bc background (13S/LM and 5S/LM) would delay mesencephalon fold elevation relative to LM/Bc but would be less delayed than in SELH/Bc. Conversely, introduction of the normal LM/Bc 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 LM/Bc. 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 in the congenic lines will give insight into the individual developmental effects of the Exen loci on cranial neural tube closure and will investigate whether mesencephalon fold elevation could in fact be considered a continuous liability trait under the multifactorial threshold model. This chapter reports on the first formal examination of the process of cranial neural tube closure in the 13S/LM, 5S/LM, 13L/SELH, and 5L/SELH congenic lines. Materials and Methods Mice and breeding design All mice originated from and were maintained in our animal unit in the Department of Medical Genetics at the University of British Columbia under standard conditions previously described (See General Materials and Methods). All breeding colonies that produced adult mice used in this study were maintained on Purina Laboratory Rodent Diet #5001 ("PLRD #5001"). During these experiments, SELH/Bc was at F46-F48, LM/Bc was at F84-F86, 13L/SELH was at N5F7-N5F12, 5L/SELH was at N8F5-N8F9, 13S/LM was at N6F7-N6Fio, and 5S/LM was at N7F4-N7F7. 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 of the mice was switched to Purina Mouse Diet #5015 ("PMD #5015") upon introduction of the males and they were fed PMD #5015 until autopsy. Females were checked for vaginal plugs by 10:00 a.m. each morning up to a week after introduction of the male. Those that had plugs were separated into new cages and kept there until sacrifice. Ovulation generally occurs in relation to the midpoint of the dark cycle in 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, PMD #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, LM/Bc 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 LM/Bc and SELH/Bc. In addition 116 the rates were compared between LM/Bc and the congenic lines on the LM/Bc background (13S/LM and 5S/LM, respectively), and between SELH/Bc and the congenic lines on the SELH/Bc background (13L/SELH and 5L/SELH, respectively). Collection of Day 8/9 embryos Upon autopsy, the uterus was removed and pinned to a black wax substrate, immersed in physiological saline (0.85% NaCl) solution, and cut open to reveal the conceptuses. Embryos were immediately collected, intact in 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 in a separate jar. Upon scoring, the embryos were dissected out of their deciduas and chorions (membranes) under 70% ethanol, given an ID number, and stored individually in cryovials in 70% ethanol. The number of embryos collected from each major congenic line and parental strain ranged from 86 to 141. The classification system The system of classification used here is based from the MacDonald et al. study (1989). It was developed from examination of ICR/Be (normal), SWV/Bc (normal), and SELH/Bc embryos at various phases of cranial neural tube formation. The stages are as follows: 1) "Folds evident" - The anterior neural folds are visible as symmetrical thickened bulges in the anterior half of the embryo. 117 2) "Folds bent" -The 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 of the mesencephalon/prosencephalon boundary. 6) "Closure 3 begun without Closure 2" - Closure 3 has obviously begun without Closure 2. This stage is unique to SELH/Bc embryos, where Closure 3 has spread caudally a notable distance through the prosencephalon (Closure 2 and Closure 3 initiate concurrently). 7) "Only prosencephalon fused" - self-explanatory. This is the subsequent stage for Stage 6, when Closure 3 has "zipped" the length of the prosencephalon. 8) "Fused to mid-mesencephalon with ANP open" - A later stage of Closure 2. Closure 2 has spread caudally and rostrally but has not met Closure 3 yet, leaving an open region of the rostral prosencephalon, the anterior neuropore (ANP).. 9) "Fused to mid-mesencephalon with ANP closed" - The prosencephalon up to the mid-mesencephalon is fused. This is due either to the caudal spreading of Closure 3 towards Closure 4 from the rhombencephalon like in SELH/Bc, or spreading from the site of Closure 2 as well as from Closure 3, like in normal strains. 118 10) "Fused to apex with ANP open" - Closure 2 has spread caudally to the top of the head (apex) and rostrally but has not met Closure 3 yet. 11) "Fused to apex with ANP closed" - The prosencephalon folds have fused as well as the mesencephalon folds through to the top of the head (apex) so that fusion of the cranial neural tube in 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 of the cranial neural folds, especially the mesencephalon folds, between LM/Bc and SELH/Bc in 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 of Stage 4 and they are as follows (See Fig. 7.3): 4A') "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 sub-stage 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 sub-stages 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 of the mesencephalon folds is approximately 140°-180° looking "face-on" at the embryo. 4A) "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 of the mesencephalon folds (looking face-on at the embryo) is approximately 110°-130°. 4B') "Prosencephalon folds are parallel or "pear-shaped," small to medium gap, mesencephalon folds >90°" - 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 of the prosencephalon folds is closer together than the rostral part of 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 <90°" - The conformation of the prosencephalon folds are the same as the above sub-stage but angle of the mesencephalon folds is <90°. 4C) "Prosencephalon folds V-shaped, mesencephalon folds <90°" - The prosencephalon folds have a small to wide gap between them and are V-shaped, where the most rostral part of the prosencephalon folds are very close together and get wider as you move caudally up the prosencephalon folds. The angle of the mesencephalon folds is <90° looking face-on. 4C) "Prosencephalon folds close and parallel or "pear-shaped," mesencephalon folds <90°" - 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 of the mesencephalon folds is <90° looking face-on. 4D) "Verging upon Closure 2, mesencephalon folds <90°" - The neural folds are so close together at the Closure 2 site that they are certainly going to initiate contact. The angle of 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 of 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 if in Stage 4. Graphs were used to show the distribution of embryos at various somite counts among the stages of 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 in 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 of mesencephalic folds, therefore more or less quantifying how much they had elevated. This was done by imagining tangent lines on the surface of the neural folds and estimating the angle they made when they crossed each other (See Fig. 7.3 on pg. 118). 122 Data Analysis The mean somite counts for embryos in 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 in the figures that represent the patterns of neural tube closure in LM/Bc, SELH/Bc, and the congenic lines are not in 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 of the sub-stages of 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 in 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 in each stage, as well, to know how much emphasis to put on it because embryos can vary in somite count at any given stage. LM/Bc and SELH/Bc (See Figs 7.4 and 7.5) Neural tube closure in LM/Bc and SELH/Bc followed the patterns previously described in MacDonald et al. and Juriloff et al. (1989 and 1991, respectively). LM/Bc 123 and SELH/Bc embryos went through the early stages of 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. LM/Bc embryos had a mean somite count of 11.5, whereas SELH/Bc embryos had a mean somite count of 13.2, approximately 2 somites difference. Looking at Stage 4 more closely (See Fig. 7.5 on pg. 126), the difference between LM/Bc and SELH/Bc was quite noticeable. Both strains had embryos that went through 4A' and 4A, but SELH/Bc embryos were approximately 1.5 somites older than the LM/Bc embryos going through these sub-stages. LM/Bc proceeded quickly to sub-stage 4B ("Prosencephalon folds parallel or "pear-shaped" and small to medium gap, mesencephalon folds < 90°) around the 11-12-somite stage, then to 4C ("Prosencephalon folds close and parallel or "pear-shaped," mesencephalon folds < 90°) around the 12-somite stage, and finally to 4D ("Verging upon Closure 2, mesencephalon folds < 90°), the only "LM/Bc- specific" sub-stage, around the 12-13-somite stage. SELH/Bc embryos, on the other hand, went through all the sub-stages except for 4D, and went through two "SELH/Bc-specific" sub-stages that LM/Bc did not go through, 4B' and 4C (See Fig. 7.5 on pg. 126). This is not surprising given the morphologies of these two sub-stages are characterized and caused by delayed mesencephalon fold elevation. Some SELH/Bc embryos go through 4B' at about 13-somite stage, some go through 4B at about the 14-15-somite stage (approximately 3 somites older than LM/Bc at this stage), and some SELH/Bc embryos go through 4C at about the 15-16-somite stage. At 4C, the "oldest" sub-stage LM/Bc and SELH/Bc share, the SELH/Bc 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 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 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 1. Folds evident Figure 7.4: Comparison between LM/Bc and SELH/Bc mean number of somite pairs present at different stages of closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. SELH/Bc 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, SELH/Bc had two outliers indicated by the stars. ct) .4. (3) SELH/Bc LM/Bc 4D. Verging upon Closure 2, mesencephalon folds < 90° 4C. Prosencephalon folds close and parallel or "pear shaped''. mesencephalon folds <90° 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 4A'. Prosencephalon folds gap wide, mesencephalon folds flat Figure 7.5: Comparison between LM/Bc and SELH/Bc mean number of 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. (12) SELH/Bc LM/Bc (31) H H o 1 -I h H 1 h -I 1 10 11 12 13 14 15 16 17 18 19 20 21 22 23 approximately 4 somites difference from LM/Bc. This striking difference demonstrates how far behind SELH/Bc is from LM/Bc in regards to mesencephalon fold elevation. In addition, SELH/Bc and LM/Bc show signs that they proceed through closure in a different way past this point as LM/Bc goes through 4D and SELH/Bc does not. At this point SELH/Bc and LM/Bc proceeded through closure in a different way, as previously documented (Macdonald et al., 1989). By the time LM/Bc initiated Closure 2 (Stage 5) (See Fig. 7.4 on pg. 125), at around the 13-somite stage, SELH/Bc was still going through Stage 4. Subsequently, LM/Bc proceeded through the late closure stages between 14 and 19 somites, and their neural tubes were closed on average by the 19-somite stage (Figure 7.4 on pg. 125). In contrast, SELH/Bc 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 in Stages 7, 9, and 12 that supported this type of closure. Given that virtually all embryos SELH/Bc do not do Closure 2, there is no anterior neuropore (ANP) to close and therefore most SELH/Bc embryos do not go through Stages 8 and 10. One embryo, however, was a bit ambiguous and had a tiny slit ANP at the "chin", and therefore was put in Stage 8. In a previous study, one SELH/Bc embryo was also put in this stage (Macdonald et al., 1989), so perhaps SELH/Bc does go through this stage on a rare occasion. No SELH/Bc embryos have been observed initiating Closure 2 though (Stage 5), they have only been observed in Stage 8. Although no embryos were observed with closed neural tubes in this study, previous work had shown that SELH/Bc embryos close their neural tubes on average by the 22-somite stage, if they are going to close. 127 In summary, LM/Bc embryos close their neural tubes quicker than the SELH/Bc embryos. LM/Bc embryos demonstrate classic normal cranial neural tube closure, whereas SELH/Bc demonstrate an abnormal mechanism for neural tube closure. LM/Bc vs. 13S/LM (See Figs. 7.6 and 7.7) The 13S/LM congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure of substituting the Exenl alleles (as well as the Exen4 alleles) from SELH/Bc into LM/Bc. Demonstrated in Figure 7.6, 13S/LM consistently went through the neural tube closure stages a little later in developmental age than LM/Bc, but not by much. Some embryos appeared to close at the same rate as LM/Bc, whereas others were delayed 1-2 somites in the closure stages. For all the stages, including the sub-stages of Stage 4, LM/Bc embryos go through, some 13S/LM embryos do so as well. In addition, they gained some "SELH/Bc-specific" stages. This first became apparent when some 13S/LM embryos, around the 11-somite stage, were observed in sub-stage 4B' (See Fig. 7.7 on pg. 130). Furthermore, the majority of embryos in this sub-stage demonstrated a new conformation of 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 in either LM/Bc or SELH/Bc (This conformation was not particular to this line though, as it was observed in all the congenic lines). In addition, one embryo was observed in 4C, the other "SELH/Bc-specific" sub-stage. The 13S/LM line maintained Closure 2 in some embryos (Stage 5), but initiated contact approximately 1 somite later than LM/Bc. 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 2 Folds bent + 1. Folds evident T to 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° 4C. Prosencephalon folds "V-shaped."' mesencephalon folds < 90° Figure 7.7: Comparison between LM/Bc and 13S/LM mean number of 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. 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 -p folds gap wide, mesencephalon folds O flat 13S/LM LM/Bc H 1 1 1 1 1 1 1-II 12 13 14 15 16 17 18 + + + 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 13S/LM 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 5S/LM in sub-stage 4D, which is verging upon Closure 2. Photograph D shows an embryo from 5S/LM, in sub-stage 4B', 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 5L/SELH demonstrating the rostral Closure 2 in the prosencephalon region that was observed in both the 5L/SELH and 13L/SELH lines. Photographs were taken by Dr. Diana M. Juriloff. ABC D E observed going through Stages 6 and 7, "SELH/Bc-specific" stages. Given that this line produces a considerable amount of exencephaly though, this is not surprising. In conclusion, 13S/LM tends to suggest that substituting in the Exenl (and Exen4) alleles from SELH/Bc into LM/Bc 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 5S/LM congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure of substituting the Exen2 alleles from SELH/Bc into LM/Bc. As demonstrated in Figure 7.9, 5S/LM consistently went through the main neural tube closure stages slightly later than LM/Bc. No clear pattern was evident for the sub-stages of Stage 4 (See Fig. 7.10 on pg. 134). For all the stages, including the sub-stages of Stage 4, LM/Bc embryos go through (with the exception of Stage 10), 5S/LM embryos do so as well. In addition, this line gained some "SELH/Bc-specific" stages. Demonstrated in Figure 7.9, some 5S/LM embryos go through sub-stage 4B' at around the 11-12-somite stage and, as discussed previously, the majority of embryos in this sub-stage demonstrated a new conformation of neural tube closure (See Fig. 7.8(D) on pg. 131). The other observed "SELH/Bc- specific" stage was Stage 6, where one embryo was observed doing Closure 3 without Closure 2 at the 17-somite stage. In conclusion, 5S/LM tends to suggest that introducing the Exen2 alleles from SELH/Bc into LM/Bc contributes to a little delay of mesencephalic fold elevation. The majority of 5S/LM embryos appeared to maintain Closure 2 and close their neural tubes 132 13 Fused completely 12. Fused to rhombencephalon 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 2 Folds bent 1. Folds evident Figure 7.9: Comparison between LM/Bc and 5S/LM mean number of somite pairs present at different stages of closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. 5S/LM LM/Bc 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° 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 4A'. Prosencephalon folds gap wide, mesencephalon folds flat Figure 7.10: Comparison between LM/Bc and 5S/LM mean number of 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. (5) % 5S/LM • LM/Bc 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 like LM/Bc, 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 13L/SELH congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure of substituting the normal Exenl alleles from LM/Bc into SELH/Bc. SELH/Bc and 13L/SELH 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 13L/SELH went through it one somite ahead of SELH/Bc. In the sub-stages of stage 4, 13L/SELH embryos not only went through all the sub-stages SELH/Bc did, but did so at an earlier developmental age then SELH/Bc, about 1-2 somites difference (See Fig. 7.12 on pg. 137). In addition, 13L/SELH gained stage 4D, an "LM/Bc-specific" sub-stage, going through at about the 13-14-somite stage. The effect of substituting the normal Exenl alleles from LM/Bc into SELH/Bc appeared to speed up the elevation of the mesencephalic folds compared to those of SELH/Bc. Some 13L/SELH embryos were observed initiating Closure 2, another "LM/Bc-specific" stage, at around the 14-somite stage. Interestingly, though, contact was initiated more rostrally than the LM/Bc Closure 2 site in some embryos in Stage 5 (See Fig. 7.8(E)). 13L/SELH continued to show evidence for Closure 2 as more embryos were found in Stage 8, a stage SELH/Bc 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 in Stage 7. It is likely that neural tube closure is completed earlier in 135 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 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 4-Figure 7.11: Comparison between SELH/Bc and 13L/SELH mean number of somite pairs present at different stages of closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. SELH/Bc 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, SELH/Bc had two outliers indicated by the stars. !(1) 10 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° 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 4A\ Prosencephalon folds gap wide, mesencephalon folds flat Figure 7.12: Comparison between SELH/Bc and 13L/SELH mean number of 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. (6), 13L/SELH SELH/Bc io 12 13 14 15 16 17 19 20 21 22 23 13L/SELH than in SELH/Bc, although no embryos were observed with completed neural tube closure. In conclusion, 13L/SELH suggested that substituting the normal Exenl alleles from LM/Bc into SELH/Bc accelerated mesencephalon fold elevation and restored Closure 2 in some, but not all embryos. 13L/SELH gained some "LM/Bc-specific" stages, but still maintained "SELH/Bc-specific" stages. SELH/Bc vs. 5L/SELH (See Figs. 7.13 and 7.14) The 5L/SELH congenic line provided the opportunity to observe the morphological effect on cranial neural tube closure of substituting the normal Exen2 alleles from LM/Bc into SELH/Bc. SELH/Bc and 5L/SELH proceeded through the first three stages of neural tube closure at the same developmental age and began to diverge at Stage 4 (See Fig. 7.13). At Stage 4, 5L/SELH embryos were ahead of SELH/Bc embryos by about 1 somite. When Stage 4 is blown up into its sub-stages (See Fig. 7.14 on pg. 140), 5L/SELH was ahead of SELH/Bc at all the sub-stages by about 1-2 somites and gained 4D, an "LM/Bc-specific" stage, suggesting that introducing the normal Exen2 alleles into SELH/Bc speeds up the elevation of the mesencephalic folds. As in the other congenic lines, some 5L/SELH embryos in 4B' displayed the new neural tube closure conformation with closely apposed prosencephalon folds and widely splayed mesencephalon folds (See Fig. 7.8(D) on pg. 131). Figure 7.13 demonstrated that Closure 2 was restored in some embryos (Stage 5) and this happened around the 15 to 16-somite stage. Interestingly, contact initiated more rostrally than in LM/Bc (as in 13L/SELH) (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 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 2. Folds bent 1. Folds evident Figure 7.13: Comparison between SELH/Bc and 5L/SELH mean number of somite pairs present at different stages of closure. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. SELH/Bc 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, SELH/Bc had two outliers indicated by the stars. • 5L/SELH • SELH/Bc —I 1 1 1 1 1 1 1 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 o 4D. Verging upon Closure 2. mesencephalon folds <90° 4C. Prosencephalon folds close and parallel or "pear shaped". mesencephalon folds < 90° 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 4A\ Prosencephalon folds gap wide, mesencephalon folds flat Figure 7.14: Comparison between SELH/Bc and 5L/SELH mean number of somite pairs present at the sub-stages of Stage 4. The horizontal bars indicate somite range and the number offset from mean somite count indicates the sample size. 5L/SELH SELH/Bc io 11 12 13 14 15 16 17 IX 19 20 21 22 23 embryo was observed in Stage 8, providing evidence for Closure 2. Some embryos began Closure 3 without Closure 2 (Stage 6), like SELH/Bc, and this occurred at approximately the same time that SELH/Bc embryos went through Stage 6, at the 17-somite stage. 5L/SELH embryos appeared to complete closure of their neural tubes a bit ahead of SELH/Bc, although their mean somite count of 22.4 does not give an accurate value as there was one old litter (D9/10) in this sample where the majority of embryos had had their neural tubes closed for quite some time. In conclusion, 5L/SELH suggested that substituting the normal Exen2 alleles from LM/Bc into SELH/Bc accelerated mesencephalon fold elevation and restored Closure 2 in some, but not all embryos. In addition, 5L/SELH gained some "LM/Bc-specific" stages, but still maintained the "SELH/Bc-specific" stages. Summary All four of the congenic lines had distinct heterogeneity of mechanisms of neural tube closure where some embryos had Closure 2 like LM/Bc and some embryos closed without Closure 2 like SELH/Bc. Introducing SELH/Bc Exen genes into LM/Bc tended to delay cranial neural tube closure a bit. In addition, Exenl and Exen2 from SELH/Bc appeared to have similar size effects on cranial neural tube closure when introduced into LM/Bc. Conversely, inserting normal LM/Bc Exen genes into SELH/Bc seemed to accelerate cranial neural tube closure by about 1-2 somites. Here, inserting the normal Exenl into SELH/Bc appeared to have a slightly bigger effect than inserting the normal Exen2 into SELH/Bc. All four congenic lines demonstrated a conformation of 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 in some embryos in 13L/SELH and 5L/SELH. Discussion Given that the multifactorial threshold model explains the risk of exencephaly in SELH/Bc, the hypothesis coming into this study was that the continuous variable underlying the threshold scale was timing of 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 in the degree of delay of mesencephalon fold elevation compared to SELH/Bc and LM/Bc. Had there been no detectable delay in the congenic lines compared to the parental strains, this would have suggested that the threshold mechanism in exencephaly, the continuous variable along the liability scale, would have been a different kind of threshold. The neural tube patterns observed from the congenic lines supported the hypothesis that the liability scale for exencephaly is timing of the mesencephalic fold elevation, that there is quantitative variation in mesencephalon fold elevation that is influenced by the Exen loci. Introduction of the SELH/Bc alleles into the LM/Bc background generally caused more delay of the mesencephalon folds compared to LM/Bc but not as much as SELH/Bc as measured by the mean somite counts in the stages that they shared (Figs. 7.6, 7.7, 7.9, and 7.10). Introduction of the LM/Bc alleles into the SELH/Bc background typically accelerated the mesencephalon fold elevation compared to SELH/Bc as measured by the mean somite counts in the stages that they shared (Figs. 142 7.11, 7.12, 7.13, and 7.14). The observation that Closure 2 occurred in some embryos and Closure 3 without Closure 2 occurred in other embryos within the same congenic line supports this hypothesis as well. This was the second finding of the study, that all the congenic lines had both types of neural tube closure; some embryos closed their neural tubes like SELH/Bc while other embryos closed their neural tubes like LM/Bc. 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 of elevation of 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 of neural fold elevation, as well. A question raised regarding the distinct heterogeneity of mechanisms of neural tube closure is whether the threshold for exencephaly would be the same threshold for "Closure 3 beginning without Closure 2" in the congenic lines, but this is not the case given that nearly all of SELH/Bc embryos close their neural tubes with Closure 3 (without Closure 2) and only 25% become exencephalic. Most likely they are related in part to each other in the congenic lines, but it does not necessarily mean that for embryos closing their neural tubes like SELH/Bc that they will be exencephalic. More support for "Closure 3 without Closure 2" being considered a threshold trait comes from the observation of one rare SELH/Bc embryo in Stage 8, where the ANP is open, providing evidence for Closure 2 occurring. One SELH/Bc embryo was observed in Stage 8 in the previous developmental study as well (Macdonald et al., 1989). In a normally distributed SELH/Bc population, these rare embryos would fall in the other 143 extreme (the left tail of 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 in the 13L/SELH and 5L/SELH congenic lines was more rostrally-offset than the Closure 2 in LM/Bc, 13S/LM, and 5S/LM, occurring in the prosencephalon folds (See Fig. 7.8(E) on pg. 131). Transferring in the normal Exenl or Exen2 alleles from LM/Bc into the SELH/Bc background, respectively, accelerated the mesencephalon fold elevation so that Closure 2 was able to initiate. In addition, transferring in the Exenl alleles from LM/Bc appeared to make the congenic line (13L/SELH) more "normal," like LM/Bc, than transferring in the Exen2 alleles from LM/Bc (5L/SELH). In 13L/SELH embryos undergoing Closure 2, some initiated at the prosencephalon/mesencephalon boundary like in LM/Bc, and some initiated more rostrally-offset, whereas only rostrally-offset Closure 2 was observed in embryos initiating Closure 2 in 5L/SELH. It would be interesting if SELH/Bc were to have Closure 2 if it would be more rostrally-offset as in the 13L/SELH and 5L/SELH congenic lines. The third finding of this developmental study was a new conformation of morphology observed in the congenic lines that paired a more advanced prosencephalon fold elevation with a less advanced mesencephalon fold elevation (See Fig. 7.8(D) on pg. 131). From the earlier developmental study of SELH/Bc (Macdonald et al., 1989) it seemed that the completion of prosencephalon fold elevation was in part assisted by mesencephalon fold elevation to bring the caudal end of the prosencephalon folds into apposition and for initiation of Closure 2 in normal strains. This was based on the "V-shape" that characterized the gap of the prosencephalon folds in some SELH/Bc embryos 144 with very delayed mesencephalon fold elevation, and assuming this "V-shaped" gap was due to the delay of elevation of the contiguous mesencephalon folds inhibiting the prosencephalon folds from completing elevation in the caudal end. The new variation observed in the congenic lines contradicts this and suggests that prosencephalon fold elevation is (or can be) independent of mesencephalon fold elevation as the entire prosencephalon region can be completely elevated and the folds be in apposition with each other, yet the mesencephalon folds are still relatively splayed or just beginning to elevate. One explanation for the new conformation observed in the congenic lines is that the transferring in of other genes into a different strain background has disrupted the genetic buffering, called canalization, of the recipient genetic buffering system so that there is more variation in the developmental process of neural tube closure. The idea, proposed by Waddington (1975), is that "developmental reactions are in general canalized, meaning that they are adjusted so as to bring about one definite end-result regardless of minor variations in conditions during the course of the reaction." Waddington states that, "the genotype can absorb certain amount of its own variation without exhibiting any alteration in development and that development is canalized in the naturally selected animal." This canalization, or genetic buffering, of the genotype is evidenced most clearly by the constancy of the wild type. In this case, the LM/Bc strain would be considered the wild type. There appears to be very little variation in neural tube closure among LM/Bc embryos; it is amazingly uniform. The SELH/Bc 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 of the embryos close their neural tubes. Described by Waddington, "canalization is a feature of the system, which is built up by natural selection, ensuring the production of the normal, optimal type in the face of unavoidable hazards of existence and morphological regulation may fail if the abnormalities are too great or occur too late in development" (Waddington, 1975). When this canalization is broken down, there is more variation. This is best evidenced in mutants of Drosophila where there is scarcely a mutant which is comparable in constancy with the wild type. In this case, the SELH/Bc and LM/Bc genetic buffering systems have been disrupted with the introduction of the LM/Bc and SELH/Bc Exen alleles, respectively, and perhaps the other passenger loci due to linkage. This disruption is seen more clearly in the congenic lines on the LM/Bc background, where normally the embryos are very uniform in the way they close their neural tubes. The introduction of the Exen alleles from SELH/Bc not only introduced the mutant alleles that delay mesencephalon fold elevation, they disrupted the genetic buffering of the LM/Bc system so that perhaps the prosencephalon folds and mesencephalon folds are not quite so in 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 of SELH/Bc could have been disrupted as well, despite becoming more normal like LM/Bc with the introduction of 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 SELH/Bc, some were observed closing their neural tubes like LM/Bc, some were observed with the new conformation, and some were observed with variations in 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 LM/Bc appeared to delay mesencephalon fold elevation a bit and introducing the normal Exenl and Exen2 alleles, respectively, into SELH/Bc background appeared to accelerate mesencephalon fold elevation. The nature of the relationship between the mesencephalon and prosencephalon folds was hard to interpret with the findings of this study. It is unknown whether they are separate entities with different elevation mechanisms, whether the same genes that cause elevation of the mesencephalon folds cause elevation of the prosencephalon folds, or whether the genes have similar roles in both tissues but maybe one gene is more important in one tissue than the other. It is clear that completion of mesencephalon fold elevation is dependent on prosencephalon fold elevation, but it is not clear whether completion of prosencephalon fold elevation is dependent on mesencephalon fold elevation. The new conformation observed in the congenic line embryos suggests that prosencephalon fold elevation can be independent of mesencephalon fold elevation. The preliminary generally accepted view of neural tube closure in mice was that neural tube closure was a continuous bidirectional process where the neural folds elevated and made contact in the cervical region and then fusion proceeded continuously in both the caudal and rostral directions, forming the neural tube. More work provided insight that the pattern of 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, in the cervical and rhombencephalon regions, respectively, have been shown to initiate closure using separate mechanisms other than the typical mechanism characterized by "rolling up of 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 in that the neural folds remain separate and closure occurs by growth of a membrane that eventually covers the rhombencephalon (Geelen and Langman, 1977). Variations of the cranial neural tube patterns in the congenic line embryos suggest that the prosencephalon and mesencephalon folds could have slightly different mechanisms of 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 of the number of litters collected in the parental strains and congenic lines. # of 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 of litters that were collected at various gestational ages . 8/18 8/21 8/22 8/23 9/0 9/1 9/9 9/10 Range of somites obtained LM/Bc 1 3 2 4 1 0-20 13S/LM 2 1 3 2 1 1-27 5S/LM 4A 4 1 2-19A SELH/Bc 5 2 4 2 1 0-23 13L/SELH 4 4 3 1 3-18 5L/SELH 1 1 2 4 1 1 5-25 A 1 litter with > 27 somites excluded from further analysis 9/9 = 9 A.M. on day 9 of gestation 149 Table 7.3: Mean somite count in parental strains and congenic lines at all the stages. Number in 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 6 7 8 9 10 11 12 13 LM/Bc 0.0 (1) 5.7 (10) 8.3 (12) 11.5 (31) 13.0 (3) 14.2 (10) 14.8 (4) 14.0 (1) 16.2 (10) 17.0 (5) 19.0 (5) 13S/LM 1.0 (1) 6.5 (4) 9.0 (5) 11.8 (35) 14.1 (7) 17.7,14.5+ (3) 20.4 (5) 15.5 (2) 15.3 (3) 16.0 (2) 16.5 (4) 17.4 (5) 24.7 (12) 5S/LM 2.0 (1) 7.0 (1) 8.3 (6) 11.7 (33) 14.3 (3) 17.0 (1) 14.8 (4) 15.0 (3) 16.4 (14) 18.0 (1) 19.0 (1) SELH/Bc 2.0 (2) 5.6 (5) 8.5 (17) 13.2 (99) 17.0 (3) 17.0 (4) 15.0 (1) 19.0,17.5+ (3) 22.0 (1) 13L/SELH 3.5 (2) 6.0 (5) 8.3 (15) 12.1 (53) 14.3* (4) 17.0 (1) 14.5 (6) 15.7 (3) 16.0 (1) 17.5 (2) 5L/SELH 6.0 (8) 8.4 (10) 12.2 (38) 15.7* (3) 17.0 (2) 16.0 (1) 16.0 (1) 16.0 (1) 18.5 (6) 19.2 (5) 22.4 (12) *rostral closure 2 +The mean somite count after the outlier at 24 somites was taken out. o Table 7.4: Mean somite count in the parental strains and congenic lines in the sub-stages of Stage 4 of cranial neural tube closure. Numbers in parentheses indicates the sample size in each group. Stage descriptions can be found in Materials and Methods. Sub-stages of stage 4 cranial neural tube closure 4A' 4A 4B' 4B 4C 4C 4D LM/Bc 9.6 (8) 11.0(1) — 11.4(5) — 11.8(5) 12.7(12) 13S/LM 10.7 (6) 10.8 (5) 10.9 (7) 12.8 (4) 15.0(1) 12.4 (5) 13.1 (7) 5S/LM 9.6 (5) 10.3 (3) 11.5 (7) 11.5(2) — 12.5 (4) 12.8 (12) SELH/Bc 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 of non-syndromic, multifactorial human NTDs 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 of interest and risk for NTDs. There is clear potential for genetic heterogeneity of NTD. There are over 90 single-gene mutations that cause NTDs in mice, most of the genes having human homologues (Harris, 2001), suggesting that misregulation of any of the human homologues could possibly increase risk for NTDs. The non-Mendelian pattern of increased recurrence risk with number affected in a family has pointed to a genetically multifactorial basis of NTD, meaning that individual NTD cases are due to combinations of risk alleles across several gene loci. Environmental factors that have been shown to influence risk of NTDs 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 SELH/Bc 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 NTDs. As the molecular defects leading to the delay of cranial neural fold elevation in SELH/Bc embryos are not evident, a future step would be to uncover the cause by identifying the loci that contribute to the risk of exencephaly in SELH/Bc (Juriloff et al., 2001). Not only would identifying the loci provide insight into the molecular defects in SELH/Bc, it would provide insight into the type of genes that are involved in additivity, as the Exen loci that contribute to the risk of exencephaly in SELH/Bc act in 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 mid Chr 13, distal Chr 5, and Chr 11, respectively. One of the goals of this work was to attempt to refine the mapping of the Exen loci in the regions that had been previously established. The EX-5001 F2 exencephaly panel, collected from mice fed the regular diet PLRD #5001, and congenic lines were used to do so. The EX-5001 study confirmed the locations of Exenl and Exen2, on Chrs 13 and 5, respectively, and there was a slight suggestion for Exen3 on Chr 11. It was in the EX-5001 study that proximal Chr 7 was first revealed in my studies to potentially containing a locus that contributed to the risk of exencephaly. This region was revisited in the first place because it approached significance in the previous F2 sire genome screen, yet was not supported in the original F2 exencephaly panel (Juriloff et al., 2001). The JHEX-5015 study, collected from mice fed the high-risk diet PMD #5015, was collected, for the most part, to test for a diet effect in F2 segregants. Given that the same markers were used to genotype the JHEX-5015 F2 exencephalic embryos across the Exen loci to test for gene-diet interactions, their genotypes confirmed the roles of the Exenl locus and the new locus on Chr 7 that was named Exen4. There has always been the question, when analyzing the F2 exencephalic embryo genotype summaries, of whether the genotype summaries that have the greatest deviation from random segregation, hence the highest x2 value/lowest /?-value, are of greater significance pointing to where the Exen locus most likely is located, or the genotype summaries with the least LM/Bc homozygotes or alleles in general (including the heterozygous genotypes). For example, in the EX-5001 study, the markers in 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 of SELH/Bc homozygotes and heterozygous genotypes, almost double the amount. The markers more distal showed less LM/Bc homozygotes, and less SELH/Bc homozygotes as well, with the number of SELH/Bc 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 in good agreement with the results from the F2 sire screen and original F2 exencephaly panel that indicated that the location of the Exenl locus was on Chr 13 near D13MU13 (p< 0.001) (Juriloff et al., 2001). The JHEX-5015 panel, on the other hand, tells a slightly different story. The same Chr 13 markers that showed twice the number of SELH/Bc homozygotes than heterozygotes in the EX-5001 panel showed equal amounts of the two genotypes and the least number of LM/Bc 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 in regards to mapping because of the limited sample size, it appeared that this panel suggests that the location of the Exenl locus is more distal, between D13MU193 and D13MU76. Given that the F2 exencephaly panels suggest different regions where the Exenl locus is most likely located according to ^-values, this shows they may not be so reliable in determining the exact location. However, the region most significant in the JHEX-5015 panel was significant in the EX-5001 panel as well, just not the most significant region. F2 exencephaly panels use probability to determine the location of the Exen loci so that there is always the chance of false positives. They create a zone of probability that peaks at the most significant region, determined by the goodness of fit x2 test for segregation ratios at 154 markers typed on the F2 exencephalic embryos, that teeters off bidirectionally creating blurry boundaries. For this reason, the congenic lines were used to help refine the mapping of the Exenl and Exen2 loci. The advantage of 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 of the Exen locus, or the region it is included in, is more definite. Chapter 5 showed that the transferred Chr 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 13L/SELH congenic line, to break up this transferred chromosomal segment into smaller sized intervals. Three recombinant congenic lines were created that had differing haplotypes. All three of them produced higher exencephaly frequencies than 13L/SELH, two of them producing "SELH/Bc-like" 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 of them are homozygous for SELH/Bc alleles. One possible explanation is that there are two genes on Chr 13 that contribute to the risk of exencephaly in SELH/Bc. Its interesting that the F2 exencephaly panels could indeed support this, as the markers that span the length of 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 of the gene that would be below Nr2fl-A/B is in better agreement with the JHEX-5015 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 PMD #5015. One possible candidate gene is Dhfr (at 88.5 Mb), because of its role in the folate pathway. Candidate genes for the other locus that is possibly above Fgd3-C/D include Fgd3 itself, which is a Rho GEF protein and a member of the FGD1 family that has a role in regulating the actin cytoskeleton and activating the c-Jun N-terminal kinase (INK) 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 in mice, though as part of 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 of NTD is unknown (Takeuchi et al., 1995), and Msx2 which is a homeodomain transcription factor that is expressed in the developing craniofacial region (Winograd et al., 1997) (Reviewed in Harris and Juriloff, 1999; Juriloff and Harris, 2000; Copp et al., 2003). Refining the mapping of the Exen2 locus on Chr 5, as well as the Exen3 and Exen4 loci, was less of a priority in my studies. The suggestions of the general region likely to contain the Exen2 locus from the EX-5001 F2 exencephaly panel and the Chr 5 congenic lines (5S/LM and 5L/SELH) agreed with each other. To date, no previously known NTD mutations appear to have been mapped to distal Chr 5, but there are some potential candidate genes around D5MU168 involved in the actin cytoskeleton that could have implications for a role of loss of actin function in 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 As the identities of the Exen loci remain unresolved, the nature of the genetic variants of the Exen loci that contribute to the risk of exencephaly in SELH/Bc remains unresolved as well. They could be polymorphisms that are maintained by selection that slightly impede the function of an Exen protein, loss-of function mutations of the Exen loci, or perhaps gain-of-function mutations of 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 of a sequence containing the 5' region of the p23 gene into the first intron of TbxlO, which caused ectopic expression (gain-of-function mutation) of a p23-TbxlO chimeric transcript, driven by the p23 gene promoter, that encoded a protein product identical to the normal variant of 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 in the developing lip, where normally it is not expressed. It is a good demonstration of how the normal expression domain of a gene does not necessarily predict its role in a defect because gain-of-function mutations have ectopic expression, which is their mechanism of disrupting development. Of significance as well, this study has shown that it is important to consider all genes in the candidate gene region and not to limit the genes to those known to be expressed in the tissue of interest. Another alternative could be that the Exen loci are disrupted by insertions of EtnlJ or MusD elements that disrupt the function of the Exen loci. SELH/Bc appears to be more prone to early transposon element (EtnU) insertions than other strains. Two out of 19 reported mutations in mouse cell lines caused by insertion of Eta elements into genes had been found in the SELH/Bc mice suggesting that Etn retrotransposition occurs relatively 157 frequently in this strain (Baust et al., 2003). Baust et al. identified a young, coding-competent MusD element in SELH/Bc that contributed to the majority of MusD transcripts suggesting that this element was overexpressed in SELH/Bc embryos. It is possible a MusD/EtnTJ transcript has inserted into an Exen locus and has disrupted the function of it. Once the Exen loci have been identified, analysis of the mutations will help unravel the mystery of the type of genetic variants that are involved in multifactorial birth defects. One other goal of this work was to determine if the response observed in the SELH/Bc strain to PMD #5015 was due to a gene-diet interaction. Taken together, the F2 exencephaly segregation studies and diet studies in congenic lines presented in this thesis indicated that there is a gene-diet interaction with the Exenl alleles from SELH/Bc. The JHEX-5015 F2 exencephaly study showed that the Exen2 locus had little or no role in risk of exencephaly when mice are fed PMD #5015, as demonstrated by the Mendelian (random) segregation ratios at the selected markers from the F2 exencephalic embryos. When mice are fed PLRD #5001, the diet the mice are regularly fed, the Exen2 locus appeared to strongly contribute to risk of exencephaly as shown by the "significantly deviated from random" segregation ratio. The conclusion from these F2 exencephaly panels was that the effect from PMD #5015 replaced the effect of the Exen2 alleles from SELH/Bc. In the EX-5001 and JHEX-5015 studies, both Exenl and Exen4 appeared to contribute to the risk of exencephaly on either diet. If only one Exen locus appeared to be involved in the risk of exencephaly in the JHEX-5015 study the explanation would have been a gene-diet interaction with that specific Exen locus so that no other Exen locus was important on PMD #5015. 158 The conclusion from Chapter 4 was that the effect from PMD #5015 superceded the effect of the Exen2 alleles from SELH/Bc so that it had little or no role in risk of exencephaly in the F2 segregants. From the congenic line studies (Chapter 5), however, it appeared that the congenic lines with Exenl alleles from SELH/Bc (13S/LM and 5L/SELH) responded to diet and the congenic lines without Exenl alleles from SELH/Bc (5S/LM and 13L/SELH) did not respond. The 13S/LM congenic line provided the strongest support for the Exen gene-diet interaction because it showed that the diet effect does not require a SELH/Bc mother. This suggested that there was a gene-diet interaction between the Exenl alleles from SELH/Bc and PMD #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 of the risk of exencephaly because of the strong gene-diet interaction between the Exenl alleles from SELH/Bc and PMD #5015, not that the effect from Exenl was directly replaced by the effect of PMD #5015. Why the Exen4 alleles from SELH/Bc showed that they still contributed to the risk of exencephaly in the JHEX-5015 study is uncertain. Perhaps the Exen4 alleles from SELH/Bc are involved in the diet response to PMD #5015, as well, if they are present. Uncovering the mechanism of the diet effect is difficult because the Exenl locus, the gene indicated in the diet response, has not yet been identified. Identification of the Exenl locus could possibly help in identifying the aspect in PMD #5015 that strongly influences exencephaly frequency in SELH/Bc. Interestingly, PMD #5015 is not the only diet to increase the risk of exencephaly in SELH/Bc (Harris and Juriloff, 2005). By chemical analysis, PLRD #5001 and PMD #5015 differ in level of almost every nutrient 159 and PLRD #5001 contains 3 natural ingredients that are not present in PMD #5015, as well as other diets shown to increase exencephaly frequency in the SELH/Bc 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 if 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 NTDs, suggesting some opportunities for further investigation of the gene-diet interaction (Anderson et al., 2005; Becerra et al., 1990) observed in my studies. This diet effect, attributed to a gene-diet interaction with the Exenl locus, on NTD frequency may point to new nutritional approaches to prevention of human NTDs. The last main goal of this study was to determine the individual effects of the Exen loci on cranial neural tube closure. The congenic lines showed that elevation of the mesencephalon folds could indeed be considered a liability trait as varying amounts of delay of elevation of the mesencephalon folds were observed as well as both types of closure (LM/Bc and SELH/Bc) in each congenic line. Though the hypothesis that the liability trait for SELH/Bc is likely timing of mesencephalon fold elevation was generally supported, new variations of 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 in humans who are a heterogeneous mix of NTD risk genes. In summary, the work presented here demonstrates the complexity of the risk of exencephaly in SELH/Bc mice and that exencephaly can be considered a multifactorial 160 threshold defect. In fact, this was the first empirical test of the multifactorial threshold model in vertebrates. Given that the risk of NTDs appears to be very complex in the human population, the SELH/Bc mouse strain has proved to be a valuable animal model, and more investigation into the genetic basis of exencephaly and the diet effect in the SELH/Bc strain will provide insights into the mechanisms underlying human multifactorial neural tube defects. Identifying the Exen loci, especially the Exenl locus, will be a next important step to unraveling the molecular defects in SELH/Bc. The interacting gene pathways that participate in the multifactorial SELH/Bc risk can then be identified which are of potentially broader significance (Juriloff et al., 2001). Uncovering the molecular role of the Exen loci in neural fold elevation hopefully will contribute to the understanding of neural tube closure and the mechanisms by which NTDs arise in 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 of mobile EtnTJ retroelements and their coding-competent MusD relatives in 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 case-control study. Pediatrics, 85:1-9. Boyles, A.L., P. Hammock, and M.C. Speer (2005) Candidate gene analysis in human neural tube defects. Am J Med Genet C Semin Med Genet., 135C:9-23. Bronson, F.H., CP. Dagg, and G.D. Snell (1966) Reproduction. In: Biology of the laboratory mouse, 2nd Edition, E.L. Green (ed.), McGraw-Hill Inc., New York, pp. 194-196. Brook, F.A., J.P. Estibeiro, and A.J. Copp (1994) Female predisposition to cranial neural tube defects is not because of a difference between the sexes in the rate of embryonic growth or development during neurulation. JMed Genet., 31:383-387. Bush, J.O., Y. Lan, and R. Jiang (2004) The cleft lip and palate defects in Dancer mutant mice result from gain of function of the TbxlO gene. PNAS, 101:7022-7027. Chen, Z. and R.R. Behringer (1995) twist is required in 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 of neurulation. DevDyn. 221:117-145. Copp, A.J., N.D. Greene, and J.N. Murdoch (2003) The genetic basis of mammalian neurulation. Nat Rev Genet., 4:784-793. Curtin, J.A., E. Quint, V. Tsipouri, R.M. Arkell, 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 of Celsrl disrupts planar polarity of inner ear hair cells and causes severe neural tube defects in the mouse. Curr Biol. 13:1129-1133. Darvasi, A. (1998) Experimental strategies for the genetic dissection of 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 NTD 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. 6th Edition. Sinaurer Associates, Inc, Sunderland, MA, online edition. 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 of abnormal neural tube closure in SELH/Bc mice (Thesis). University of British Columbia. Gunther, T., M. Struwe, A. Aguzzi, and K. Schughart (1994) Open brain, a new mouse mutant with severe neural tube defects, shows altered gene expression patterns in the developing spinal cord. Development 120:3119-3130. Hall, J.G., J.M. Friedman, B.A. Kenna, J. Popkin, M. Jawanda, and W. Arnold (1988) Clinical, genetic, and epidemiological factors in neural tube defects. Am JHum Genet., 43:827-837. Hamblet, N.S., N. Lijam, P. Ruiz-Lozano, J. Wang, Y. Yang, Z. Luo, L. Mei, 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. Ullrich, S. Cook, R.T. Bronson, and M.T. Davisson (1997) Forebrain overgrowth (fog): a new mutation in the mouse affecting neural tube development. Teratology, 55:231-240. Harris, M.J. and D.M. Juriloff (1999) Mini-review: Toward understanding mechanisms of genetic neural tube defects in mice. Teratology, 60:292-305. Harris, M.J. (2001) Why are the genes that cause risk of 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 in SELH/Bc strain mouse embryos. Birth Defects Res A Clin Mol Teratol., 73:532-540. Hofmann, M., M.J. Harris, D.M. Juriloff, and T. Boehm (1998) Spontaneous mutations in SELH/Bc mice due to insertions of early transposons: molecular characterization of 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 in mice with targeted modification of 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 in fog mice. Proc. Natl. Acad. Sci. USA, 98:9683-9687. Hui, CC. and A.L. Joyner (1993) A mouse model of Greig cephalopolysyndactyly syndrome: the extra-toesJ mutation contains an intragenic deletion of the GH3 gene. Nat Genet. 3:241-246. 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, 64:189-200. Juriloff, D.M., M.J. Harris, and D.G. Mah (2005) The open-eyelid mutation, lidgap-Gates, is an eight-exon deletion in the mouse Map3kl gene. Genomics 85:139-142. 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 Abl 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: An 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 Krtl 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 CH. (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.Xu, L. Hu, T. Asakura, D. vander Kolk, J.D. Strandberg, and G.L. Semenza (1997) Perinatal lethality and multiple craniofacial malformations in MSX2 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 in our lab. Refer to the SSLPs section in 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 A A AC 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' -TCTATGTTC ACCGG ATCTG A 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' Smcx-1, y-1 X,Y 5 '-CCGCTGCCAAATTCTTTGG-3' 5 '-TGAAGCTTTTGGCTTTGAG-3' 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