PDF hosted at the Radboud Repository of the Radboud University Nijmegen

The following full text is a publisher's version.

For additional information about this publication click this link. http://hdl.handle.net/2066/145592

Please be advised that this information was generated on 2021-10-11 and may be subject to change. Genetic Factors in Humar Neural Tube Defects

ИМ

I; Л lini

Genetic Factors in Human Neural Tube Defects ISBN: 90-9012589-2 Druk: Printpartners Ipskamp Enschede

The research described in this thesis was carried out at the Department of Human Genetics of the University Hospital Nijmegen and supported by grants from the Dutch Prinses Beatrix Fonds.

Cover photos: Cross-sections through a chick embryo during successive stages of neurulation. Modified from Schoenwolf, G.C. (1982) Scanning Electron Microsc. I: 289-308 Genetic Factors in Human Neural Tube Defects

een wetenschappelijke proeve op het gebied van de Medische Wetenschappen

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Katholieke Universiteit Nijmegen, volgens besluit van het College van Decanen in het openbaar te verdedigen op woensdag 12 mei 1999 des namiddags om 1.30 uur precies

door

Franciscus Antonius Hol

geboren op 21 juli 1964 teOss Promotor: Prof. Dr. V.C.H.H. Ropers (Max Planck Institute for Molecular Genetics, Berlin)

Co-promotor: Dr. E.C.M. Mariman

Manuscript Commissie: Prof. Dr. G.W.A.M. Padberg Prof. Dr. C.H.C.M. Buys (RUG) Prof. Dr. Ir. F.J.M. Trijbels voor mijn ouders Contents

Abbreviations 8

Chapter 1: General Introduction 9 1.1 Neural Tube Defects: General introduction 11 1.2 Neural Tube Embryology 11 1.3 Classification of Neural Tube Defects 14 1.4 Neural Tube Defects: A multifactorial trait 16 1.5 Etiology of human Neural Tube Defects 20 1.6 Mouse genetic models and Neural Tube Defect pathogenesis 25 1.7 Evidence for multifactorial inheritance of Neural Tube Defects in mice 27 1.8 PAX genes and Neural Tube Defects 29 1.9 Folic acid and Neural Tube Defects 31 1.10 Objectives of the present study 35 References 37

Chapter 2: Exclusion Mapping of the Gene for X-lmked Neural Tube Defects in an Icelandic Family 47

Chapter 3: Absence of Linkage between Familial Neural Tube Defects and the PAX3 Gene 55

Chapter 4: A Frameshift Mutation in the Gene for PAX3 in a Girl with Spina Bifida and Mild Signs of Waardenburg Syndrome 63

Chapter 5: PAX Genes and Human Neural Tube Defects: An Amino Acid Substitution in PAX1 in a Patient with Spina Bifida 71

Chapter 6: Altered Regulation of PDGFRa Transcription in vitro by Spina Bifida Associated Mutant Paxl Proteins 79

6 Chapter 7: Molecular Genetic Analysis of the Gene Encoding the Tnfunctional Enzyme MTHFD (methylenetetrahydrofolate-dehydrogenase, methenyltetrahydro- folate-cyclohydrolase, formyltetrahydrofolate synthetase) in Patients with Neural Tube Defects 87

Chapter 8: General Discussion 97 8 1 X-hnked genes in Neural Tube Defects 99 8 2 PAX genes and Neural Tube Defects 99 8 3 Folic acid and Neural Tube Defects 103 8 4 Conclusion and future prospects 104 References 106

Chapter 9: Summary / Samenvatting 111 Summary 113 Samenvatting 115

Dankwoord 117 Curriculum vitae 118 List of publications 119

7 Abbreviations bp base pair BSA bovine serum albumine CH 5,10-methenyltetrahydro folate cyclohydrolase CI confidence interval cM centi-Morgan DNA deoxyribonucleic acid dNTP deoxynuclcotide triphosphates ЕС embryonal carcinoma FISH fluorescent in situ hybridization FS 10-formyltetrahydrofolate synthetase GDB Genome Data Base IBD identity by descent kb kilo base pairs MS methionine synthase MTHFR 5,10-methylenetetrahydrofolate reductase MTHFD 5,10-methylenetetrahydrofolate dehydrogenase NCC neural crest cells nt nucleotide NTD neural tube defect OMIM Online Mendelian Inheritance in Man OR odds ratio PCR polymerase chain reaction PDGFRa platelet derived growthfactor receptor alpha RA retinole acid RFLP restriction fragment length polymorphism RNA ribonucleic acid RT reverse transcriptase RT-PCR reverse transcription polymerase chain reaction SBA spina bifida aperta SBC spina binda cystica SBO spina binda occulta SHMT serinehydroxylmethyltransferase SSCP single strand conformation polymorphism TDT transmission / disequilibrium test UTR untranslated region VPA valproic acid WS Waardenburg syndrome

8 Chapter 1

General Introduction

9

General introduction

1.1 Neural Tube Defects: General introduction The term neural tube defects (NTD) refers to a wide spectrum of common congenital malformations with an average incidence of 1-2 per 1000 births, involving defective formation of the brain and/or spinal cord. Anencephaly and spina bifida are the best recognised examples. They are the result of incomplete formation of the neural tube during early embryonic deve­ lopment, which in humans takes place in the fourth week of gestation. Striking variation in incidence is known to exist between populations, varying from 1 in 3000 in the low risk Finnish population to more than 1 in 300 in high risk areas in Ireland and the UK (Dolk et al., 1991). In the Netherlands, NTDs occur with an incidence of approximately 1 in 700 (Eurocat Working group, 1991). Evidence has mounted over the years showing that approximately 85 percent of NTD cases are the result of a combined influence of environmental and genetic factors, indicating a multifactorial etiology (Copp & Bemfield, 1994). Marked geographical variation, and association with dietary folate deficiency or specific teratogens point to environmental factors, whereas the description of a growing number of major genes in the mouse and the significant excess in close relatives provide clear evidence for genetic components (Campbell et al. 1986). A small proportion of NTD cases, however, can be attributed to specific causes like a single gene defect in rare autosomal conditions such as Meckel syndrome and sacral agenesis (Paavola et al., 1995; Ross et al., 1998), whereas others are associated with chromosome abnormalities such as trisomy 13, trisomy 18 and triploidy (van Maldergem et al., 1989; Rodriquez et al., 1990; Seller et al., 1995a). Over the past decades there has been a world-wide decline in the number of NTD births. Two important developments in the field of NTDs have contributed to this phenomenon, although they can not account for the entire decrease in NTD prevalence (Seller, 1994a). As a consequence of prenatal diagnostic tests, like screening of maternal serum for alpha-fetoprotein (AFP) and, more recently, elaborate ultrasound screening as part of routine prenatal care, therapeutic termination of pregnancy has obscured the true incidence of NTDs. More recently, primary prevention through folate suppletion of women at risk has further contributed to the decline of incidence rates (MRC Vitamin Study Research Group, 1991). Review of the literature of human NTDs and animal models reveals that there is much information but little conclusive evidence about the cause of NTDs. One of the challenges of modem molecular genetics is the identification of predisposing genetic factors in complex diseases like NTD. Knowing these genetic risk factors will lead to understanding the relevant pathogenetic mechanisms and eventually, to their prevention.

1.2 Neural Tube Embryology To understand the pathogenesis of neural tube defects insight into the normal process of neuro-embryologic development is required. The neural tube is formed during a process called 11 Chapter 1 neurulation, which in humans takes place during the fourth week after fertilization. Much of the knowledge on mammalian neurulation comes from studies in the mouse (Morris-Kay et al., 1994) but studies on human embryos have also been performed (O'Rahilly & Müller, 1994).

Neurulation Upon gastrulation of vertebrate embryos, the internal endodermal, the intermediate mesodermal, and the external ectodermal layers are formed. During the third week of human development, a subset of the ectodermal cell population will give rise to the neural tube, which will eventually differentiate into the central nervous system, including the brain and spinal cord. The process by which the flat layer of ectodermal cells is transformed into a hollow tube is called neurulation and has been the subject of intensive study over the last two decades. It has become increasingly clear that neurulation involves a complex of morphogenetic interactions both within the neural plate and between the neural plate and the surrounding non-neural tissues and that it is driven by many genes.

A Neural plate D Neural crest

Figure 1: Formation of the neural tube: primary neurulation (Adapted from Browder et al, 1991)

12 General introduction

Neurulation can be divided into two subsequent developmental stages; primary neundation, during which the brain and the larger part of the spinal cord is formed, and secondary neurulation which accounts for those parts of the neural tube that will differentiate into the lower sacral and caudal regions of the spine.

Primary neurulation In the normally developing human embryo the neural tube is formed from a single sheet of ectodermal cells. The first indication that a region of ectoderm is destined to become neural tissue is a change in cell shape. Midline ectodermal cells become elongated while the cells destined to form the epidermis become more flattened.Th e elongation of dorsal ectodermal cells causes these future neural regions to rise above the surrounding ectoderm, thereby creating the neural plate (Fig 1). This neural plate can be regarded as the actual primordium of the central nervous system. Very little is known about mechanisms involved in this neural induction process, but, it is generally believed that signals transmitted by the underlying notochord play a key role in this process (Schoenwolf, 1994). As soon as the neural plate has formed, its edges become elevated relative to its midline, converting the plate into a groove. The neural groove deepens and folds develop on either side. In humans, the neural groove and folds are first seen around postovulatory day 18. Once the neural folds have developed, around day 22, the tips of the folds fuse with each other at the dorsal midline thereby forming the first part of the neural tube. The two open ends of the neural tube are called the anterior (cephalic) neuropore and the posterior (caudal) neuropore, respectively.

Figure 2: Different initiation sites that have been proposed for neural tube closure in human embryos during neurulation (Adapted from van Allen et al, 1993).

Closure of the neural tube proceeds in a bidirectional fashion and, in mice, is initiated at at least four sites along the anterior-posterior axis (Golden & Chemoff, 1993). In humans, five

13 Chapter 1 closure sites have been proposed based on the spatial distribution of neural tube defects (Van Allen, 1993; Seller 1995b)(Fig 2). The anterior neuropore closes within a few hours around day 24. The final event in primary neurulation is closure of the posterior neuropore. Closure of this neuropore requires about one day to completion and takes place around day 26.

Secondary neurulation The lower sacral and caudal regions of the spine account for a relatively minor proportion of the human body axis. Secondary neurulation begins around day 26 once the caudal neuropore has closed. The process occurs through a canalisation event at the most caudal end of the embryo without the prior formation of neural folds. In human embryos, the transition zone between primary and secondary neurulation is in the region of the second sacral vertebra (Muller & O'Rahilly, 1987). Secondary neurulation is more or less complete around week 8.

1.3 Classification of Neural Tube Defects Neural tube defects (NTDs) usually arise when some parts of the neural folds fail to close or fuse during the process of neurulation. As a result, the neural tube remains open at this position along the anterior-posterior axis. At the site of the lesion, the neural tissue will be exposed to the in utero environment and will give rise to defects such as anencephaly and spina bifida. Spina bifidas with an open neural tube result from failure of the neural folds to fuse in the caudal region. Lack of fusion at the cephalic end often results in anencephaly. More rarely encephalocele, another cranial NTD occurs. The terminology used to distinguish different neural tube defects is often confusing. Different authors may use different criteria and nomenclatures. This is probably due to the fact that neural tube defects, and spina bifida in particular, show wide spectrum of clinically overlapping disease phcnotypes. This section will describe the clinical conditions that can be considered as the most common neural lube defects.

Spina bifida The term spina bifida is widely used to refer to incomplete formation of the vertebral arches with or without involvement of the meninges or the spinal cord. The incidence in the Netherlands is about 1 per 1700 births in (Eurocat Working group, 1991). It is now generally accepted to divide this disorder into two major classes, i.e. (a) spina bifida occulta, which implies that the lesion is covered with skin, and (b) spina bifida aperta, referring to open lesions. The latter is further subdivided into two main conditions, (i) meningocele and (ii) myelomeningocele

14 General introduction

Spina bifida occulta In spina bifida occulta the vertebral arches are not fused There is not, however, any distension of the meninges and generally the spinal cord and its membranes are normal The site of the spinal defect is skm-covcred and sometimes marked by a slight swelling, a dimple m the skin, a capillary nevus or a hairy patch but often there is no external evidence of the defect This condition rarely has any functionally consequences Non-closure of the fifth lumbar vertebral arch and/or the first sacral arch is seen in at least 5% of the adult population (Boone et al, 1985, Fidas et al, 1987) Although many consider this condition solely as a vertebral anomaly that does not belong to the spectrum of NTDs, one could argue that it may reflect a mild form of spina bifida

Spina bifida aperta Meningocele Meningocele is the less severe and less common type affecting between approximately 10 and 20 percent of all cases with spina bifida aperta In this condition, the meninges herniates through the gap in the spine to form a smooth cystic sac filled with cerebrospinal fluid, without involvement of the spinal cord Therefore, there is usually no significant neurological impairment although some weakness of legs, bladder or bowel may be present Meningocele most commonly involves the lower lumbar or sacral region and, in general the bony defect is less extensive than in myelomeningocele

Spina bifida aperta Myelomeningocele Myelomeningocele is the second major type of spina bifida aperta and is seen m approximately 80 percent of cases Again, there is failure of the vertebrae to fuse and distension of the meninges, but this condition differs from meningocele in that the spinal cord itself protrudes into the sac and is abnormal, the result usually being a permanent and irreversible neurological disability In this condition, the structural defects of the spinal cord are very variable and different terms are used to desenbe them In the most severe form, the nervous tissue is directly exposed due to complete failing of closure of the neural tube The seventy of the handicap depends on the position of the defect and on the number of vertebral segments involved, which may vary considerably from only a few to many bifid spines In general, high (cervical) or low (sacral) defects produce less severe handicap in survivors than do lesions in the middle (lumbar or thoracic) parts of the spinal column Spina bifida with significant neural disability is frequently associated with hydrocephalus and an Amold-Chian malformation, a condition in which the cerebellum herniates through the foramen magnum into the spinal column

15 Chapter 1

Anencephaly Anencephaly represents a relatively common but severe form of NTD due to impaired fusion of the anterior neuropore In the Netherlands, the incidence is approximately 1 per 1500 births (Eurocat Working Group, 1991) In this congenital abnormality forebrain and cranial vault are virtually absent Anencephalics are commonly stillborn and those bom alive die shortly after birth The condition is more common in females (Seller, 1987) In a small proportion of these infants there is also exposure of the complete spinal cord This condition is referred to as cramorachischisis

Encephalocele The term encephalocele refers to a condition which is due to a defect in the fusion of the bones of the skull, usually in the occipital region of the head, and which allows soft tissue to herniate The incidence in the Dutch population is about 1 per 5000 births (Eurocat Working Group, 1991) Some of the lesions contain only meninges and some actual brain tissue Rarely, like in Meckel syndrome, the condition is inherited as a recessive trait (Paavola et al, 1995)

1.4 Neural Tube Defects: A multifactorial trait A large number of common disorders which show familial aggregation but a complex pattern of inheritance are thought to result from the interaction of genetic factors and environment The genetic mechanisms underlying these multifactorial disorders are poorly understood Neural tube defects represent a classical example of a complex disease with multifactorial etiology (Hall et al, 1988)

The liability/threshold model One very useful model to explain multifactorial inheritance is the liability/threshold model (Falconer, 1981) According to this model, all of the factors which influence the development of a multifactorial disorder, whether genetic or environmental, add up to form a single entity known as liability The liabilities of all individuals in a population form a continuous variable, which has a normal (Gaussian) distribution in both the general population and in relatives of affected individuals However, the liability distribution for family members will be shifted to the right, ι e to higher values, depending on the degree of their relationship to the affected index case (Fig 3) Accordingly, discontinuous phenotypes can be explained by assuming that the sum of adverse genetic and/or environmental factors must exceed a threshold before the abnormal phenotype results In the general population the proportion of individuals with liabilities exceeding the threshold is the incidence of the disease

16 General introduction

General Population

Population incidence

Relatives

Familial incidence

Liability

Figure 3: Liability curves in the general population and in relatives of affected people for a multifactorial disorder (Adaptedfrom Mueller & Young, 1995)

Liability can not be measured but the mean liability of a group can be determined from the incidence of the disease in that group using statistics of the normal distribution The attraction of this model is that it provides a simple explanation for the observed familial risks in conditions like spma bifida

1) The incidence of the condition is highest among relatives of the most severely affected patients, presumably because they represent individuals with relatively high liabilities (1 e they map to the extreme nght of the curve) n) The risk is highest amongst close relatives of the index case and decreases rapidly in more distant relatives For example in spina bifida the risks to first, second and third degree relatives of the index case are approximately 4%, 1% and less than 0 5%, respectively (Carter, 1974, Khoury et al, 1982) in) If there is more than one affected close relative then the risk for other relatives is increased In spina bifida, if one child is affected the risk for a subsequent sibling to be affected is approximately 3%, for newborns with two affected sibs it is approximately 10%, and for a child with three affected sibs it is approximately 20 % (Seller, 1994a)

17 Chapter 1

Tools for mapping the genetic component(s) of multifactorial diseases In recent years there seems to be an increasing interest among molecular geneticist to study complex disorders The mam reason for this is that tools are now becoming available to map gene loci involved in this category of medically relevant diseases (reviewed by Lander & Schork, 1994) A particularly important development in this field has been the construction of dense genetic linkage maps containing a large number of highly polymorphic markers (Dib et al, 1996) Availability of these maps permits mapping of loci involved in complex disorders The ultimate goal in disease mapping is to isolate and clone the disease gene itself, which would in turn lead to an understanding of the underlying biological mechanisms, and contribute significantly to the development of molecular diagnostics and preventive strategies One resource that will be extremely important in the near future for cloning disease-susceptibility genes is the genome-wide map of expressed sequence tags (EST) currently under development (Schuier et al, 1996) The isolation of mapped susceptibiltity genes will then be earned out via the 'positional candidate' approach where all the expressed sequences in the region of interest need to be evaluated through combinations of association studies and mutation assays until the disease gene is identified Nonetheless, identification of genetic factors underlying multifactorial traits is still a difficult task One phenomenon that particularly complicates identification of susceptibility genes is the fact that genetic heterogeneity is very likely, meaning that mutations in different genes may result in identical phenotypes Several techniques can be applied to map genetic factors in complex disorders and will be discussed below However, succeeding in detecting true disease loci using any of these techniques, requires that a significant part of the patients have descended from a common ancestor In this way, the problem of genetic heterogeneity will be reduced to a minimum

Linkage analysis Classical linkage analysis is applicable if a genetic model detailing the mode of inheritance can be proposed for a particular pedigree (Lathrop et al, 1985). It is the method of choice for the genomic localisation of simple Mendehan traits Applications to complex traits, like NTD, can be more problematic because it may be hard to find a precise model that adequately explains the inheritance pattern Further, in NTD genetic heterogeneity is likely, decreasing the ability to detect linkage by pooling lod scores from different kindreds Moreover, informative families with large numbers of affected individuals are scarce Therefore, alternative approaches to conventional linkage analysis are needed.

18 General introduction

Affected sib pair analysis If affected siblings share a particular allele more often than would be expected by chance, this indicates that that allele or its locus is involved in some way in causing the disease. This approach does not depend on a specific mode of inheritance and uses nuclear families with two affected sibs. However, many sib pairs are required in order to obtain useful information on the location of a predisposing gene. In fact, mapping a disease gene to a 1 cM interval will require around 600 families if the gene causes a twofold increase in risk to offspring (Lander & Schork, 1994). Affected sib pairs with neural tube defects are simply not available in large numbers, which makes this method less appropriate.

'Identity-by-descent ' mapping in founder populations In a founder population patients with a genetic disease are likely to share predisposing genes from a common ancestor (Puffenberger et al., 1994; te Meerman et al, 1995). Depending on their relationship, patients are expected to share extended segments of DNA around the disease gene. Such shared segments can be identified through a genomic search using highly polymorphic markers. The method allows defining the map position of the relevant genes and is applicable to many genetic diseases, because it does not presuppose a genetic model.

Allelic association studies Association studies are not depending on familial inheritance patterns at all. Rather, they are case-control studies based on comparison of unrelated affected and unaffected individuals from a population. Positive allelic association means that across the whole population, people who have a certain allele at some locus (A) have a statistically more than random chance of having some particular (disease-) allele (D) at a second locus. Association between alleles A and D can arise for three reasons i) Positive association can occur if allele A is actually a cause of disease D, either as major gene or as modifier gene, ii) It can also occur if allele A does not cause the trait but is in linkage disequilibrium with the actual cause which means that allele A is in close vicinity to the disease-causing allele D. iii) Positive association may also be an artifact due to population stratification (see following paragraph).

Testing for allelic association will always be restricted to candidate loci because scanning the entire genome for association is not feasible on both practical and statistical grounds. Bassically,

19 Chapter 1 two methods for demonstrating allelic association are commonly applied i.e., the classical case- control studies and the 'transmission / disequilibrium test'. a) Case-control study of disease-marker association Disease-marker associations are found by comparing the frequencies of a particular marker allele in a series of patients and a series of healthy controls (Weeks & Lathrop, 1995). The biggest potential pitfall of this method is the choice of a control group. The frequency of the allele of interest in the control scries may be obscured by population stratification effects, which means that the control group may be drawn from genetically (ethnic) distinct subpopulations, each having its own prevalence rate. This problem can be overcome by applying an association study that uses an 'internal control' for allele frequencies,lik e the 'transmission / disequilibrium test'. b) Transmission / Disequilibrium Test The transmission test for linkage disequilibrium or, in short, the transmission / disequilibrium test (TDT), considers families with one or more affected offspring, where at least one parent is heterozygous for an allele which is suspected of being associated with the disease, and evaluates the frequency with which that allele is transmitted to affected offspring (Spielman et al, 1993). A parent heterozygous for an associated allele Al and a nonassociated allele A2 should more often transmit Al than A2 to an affected child. Compared to conventional tests for linkage and association, the TDT has the advantage that it can be applied to small families and does not require series of matched controls. A drawback of this method is that it requires large numbers of heterozygotes in order to reach significance. Therefore, when evaluating the transmission of rare alleles, the conventional population based case-control study is more appropriate.

1.5 Etiology of human Neural Tube Defects Very little is known about the actual cause of neural tube defects. Epidemiologic studies have indicated a multifactorial etiology for most of the cases in which a genetic predisposition plays a part, as do environmental factors (Copp et al, 1990). It appears that the genetic predisposition is polygenic and there are indications that there are also multiple environmental factors.

Evidence in favour of a genetic basis in NTD Several lines of evidence argue for a genetic basis in the majority of human NTDs. The heritability in human neural tube defects, which can be defined as the relative contribution of genetic factors, has been estimated to be about 60 percent (Copp & Bernfield, 1994).

20 General introduction

Familial occurrence Family studies reveal an increased risk for NTD not only for parents of an affected child but also for their first-, second-, and third-degree relatives (Carter, 1974, Khoury et al, 1982) A significant number multiple-case families has been described (Drainer et al, 1991, Manman & Hamel, 1992, Chatkupt et al, 1995, Byrne et al, 1996) In most of these families, the transmission of NTD does not seem to follow a Mendehan pattern of inheritance However, a few reports in literature concern families with apparent monogenic inheritance of NTD suggesting that here, the disease may be caused by single genes (Lynch et al, 1995, Ross et al, 1998) In particular, the X-hnked pattern of inheritance in гаге NTD families argues for the presence of predisposing genes on the X chromosome (Tonello et al, 1980, Baraitser & Bum, 1984, Tonello, 1984, Jensson et al, 1988) Figure 4 shows an example of a kindred with apparent X-hnked recessive spina bifida (Oman-Ganes & Shokeir, 1984) However, this particular family was not available for molecular analysis

0-Γ-0

D-rO jZbr—О 0 0Ю00 ¿Η—0"

ітут^

Figure 4: Pedigree of a Canadian family with apparent X-hnked recessive spina bifida, originally reported bv Oman-Ganes & Shokeir (1984) Affected members display lumbosacral myelomeningocele

Several lines of evidence point to the existence of a gene for spina bifida in Xq27 Goerss et al (1993) reported on a US family with a duplication in Xq26-q27 in two brothers, their mother and maternal grandmother The two brothers carrying the duplication displayed spina bifida whereas a third healthy brother mhented the normal X chromosome The NTD in the two brothers may be due to a double dose or to disruption of a cntical gene In addition, a de novo X/autosomal translocation (t(X,22)(q27,ql21)) associated with lumbosacral spina bifida aperta

21 Chapter I

has recently been reported, again providing strong evidence for a causative gene in Xq27 (Fryns et al, 1996) It is noteworthy that this part of the human X chromosome is likely to carry the human counterpart of the Bent tail gene (Bn), which gives rise to X-hnked NTD in the mouse (Lyon & Searle, 1990)

Ethnic variation For common NTDs, including anencephaly and spina bifida, the incidence is in the order of 1 in 1000 births However, there is considerable variation between populations of different ethnic origin suggesting that the genetic background is of importance for the NTD susceptibility of an individual Although one could argue that these differences between ethnic groups are not necessarily genetic but could also reflect cultural (environmental) differences, for instance dietary habits, it is now generally assumed that genetic factors play a larger part in this phenomenon This is supported by the observation that ethnic differences m the incidence of of NTD are persistent for at least several generations after migration and subsequent adaption to the new social-cultural environment For instance, the worldwide low birth-frequency of neural tube malformations among Negro populations with a common origin in West and Central Africa is well-documented (Leek 1984, Mynanthopoulos & Melmck 1987) This low incidence of NTDs is preserved under a variety of different environmental conditions so that one must assume that their lower relative nsk is due to genetic influences

Sex differences Another feature of NTD epidemiology is the distorted sex ratio Females predominate among individuals with upper NTDs, especially anencephaly, whereas the sex ratio is more equal, or slightly male predominant, in lower spina bifida (Cuckle, 1993, Seller, 1995c) One hypothesis that has been proposed to explain for this sex bias is the fact that female embryos may have a slower growth rate than males which results in a longer time period necessary to complete neurulation and the correspondingly greater nsk for intervening adverse environmental influences (Seller, 1987) Another possible explanation for the differences in sex ratio has been proposed by Kallen et al (1994) who found a male excess in NTD fetuses delivered spontaneously before week 20, indicating selective prenatal death risks for NTD fetuses of different sexes Both proposed mechanisms, however, were shown not to be responsible for the observed female preponderance seen in exencephahc curly tail mice (Brook et al, 1994). Several studies report on the excess of affected individuals among maternal relatives of the patient compared to relatives of the father (Byrne et al, 1996) In particular, a significant female preponderance m the group of normal transmitting parents in multiple-case families was found in two independent studies (Chatkupt et al, 1992, Manman & Hamel, 1992) Both genomic

22 General introduction imprinting and abnormal X ìnactivation have been proposed as the underlying mechanisms by which inherited factors may account for this phenomenon Sex influences in NTD may also be ascribed to genes on the X-chromosome (Jensson et al, 1988, this thesis)

Mouse mutants Strong evidence for the hentabihty of neural tube defects comes from studies m mice At least 10 spontaneous mutations causing NTDs of various types have been described (Copp et al, 1990, Hams & Junioff, 1997)(Table 1) The existence of so many neural tube defect-promoting genes comes as no surprise since neurulation represents a dynamic multistep developmental process in which many different genes interact Therefore, numerous distinct gene products may be required for proper neurulation

Table 1: Mutant mouse genes with NTD phenotypes (Modifiedfrom Copp & Bernfield 1994) Mouse mutation Gene symbol Chromosome NTD phenotype Loop-tail LP 1 Craniorachischisis Cranioschisis Cm Nd Exencephaly Crooked tail Cd 6 Exencephaly Rib fusions Rf Nd Exencephaly SELH" - Nd Exencephaly Extra toes Xt 13 Exencephaly Exencephaly Xn Nd Exencephaly Fused Fu 17 Exencephaly and spina bifida Curly tail Ct 4 Exencephaly and spina bifida Splotch Sp 1 Exencephaly and spina bifida Bent-tail Bn X Exencephaly and spina bifida Vacuolated lens VI 1 Spina bifida Patch Ph 5 Spina bifida Axial defects Axd Nd Spina bifida Snubnose Sno 4 Spina bifida 'Multiple genes are implicated in the SELH strain predisposition to neural tube defects nd no data

Only three of the murine genes listed in table 1 have been cloned to date namely Splotch (Sp), Extra toes (Xt) and Patch (Ph) The Sp locus encodes the РахЗ transcription factor, which is a member of the family of Pax genes The corresponding human gene is mutated in Waardenburg syndrome type 1 and 3 which is known to be associated with spina bifida (Read & Newton, 1997, this thesis) The Xt mutation disrupts the Gli3 gene (Hui & Joyncr, 1993) which was identified as a member of the КгирреІ-теЫса zinc finger protein gene family The human GLI3 locus is disrupted in some patients with Greig's cephalopolysyndactyly (Vortkamp et al, 1991) and recently, GLI3 frameshift mutations were identified in patients with the autosomal dominant

23 Chapter I

Pallister-Hall syndrome (Kang et al., 1997). The Ph mutation involves an extensive deletion encompassing the PDGFRa gene (Stephenson et al., 1991). To date, no human disorder has been associated with mutations in this latter gene.

Twin studies Twin studies have traditionally been used for elucidating the relative contributions of genetic and environmental factors to the development of a given condition. These studies arc based on the assumption that all twins share similar prenatal environments, and that monozygotic twins share identical genotypes whereas dizygotic twins are no more similar genetically than are other siblings. Differences in the concordance rates between monozygotic and dizygotic twin pairs will reflect the contribution of genetic factors. With respect to neural tube defects the studies have been difficult and somewhat inconclusive (Elwood & Elwood, 1980). They are complicated by the low incidence of twins with spina bifida or anencephaly, which makes it difficult to collect a sufficiently large sample (Carter, 1974). One factor that may be responsible for a reduced number of affected twin pairs is their premature loss due to spontaneous abortion. To date twin studies have provided little useful information on the genetic predisposition of neural tube defects. On the other hand, studies in twins remain of interest because the twinning phenomenon itself might be related to the development of neural tube malformations (Garabedian & Fraser, 1994; Kallen et al, 1994)

Environmental factors in NTD etiology Many environmental factors have been suggested to play a role in NTD etiology and have been reviewed in several papers (Carter, 1974; Seller, 1987; Hall et al., 1988). Marked variation in the incidence of NTDs exists between closely related but geographically distant populations suggesting the involvement of specific environmental factors. The most striking illustration for this are the British isles, where the incidence is much higher in Northern Ireland, South Wales and Scotland than in the London area (Dolk et al., 1991). Socioeconomic status is frequently reported as an important factor with an inverse correlation between the incidence of NTD and the height of the social class (Carter, 1974). Nutritional deficiency may play a role in this relation since certain vitamins, folic acid in particular, have been proposed as important factors, as will be discussed in more detail in paragraph 1.9. Another reported association is that of maternal zinc status, although different studies report conflicting results (Zimmerman, 1984; Buamah et al., 1984; Milunsky et al., 1992a). The potentially teratogenic action of certain therapeutic drugs is well known, for example, from the increased risk for NTD offspring in pregnant epileptic women receiving the anti-convulsant sodium valproate (Lindhout & Schmidt, 1986). Maternal insulin dependent diabetes mellitus is another known risk factor

24 General introduction

(Mynanthopoulos & Melnick, 1987) Other proposed environmental factors are hyperthermia, seasonal trends and maternal age (Carter, 1974, Layde et al, 1980, Dallaire et al, 1984, Milunskyetal,1992b) The distinction between environmental and genetic factors may not always be that clear Although classified as environmental, some of the factors listed above may only have an effect in individuals that are genetically susceptible For instance, low nutritional folate intake may only be a risk factor for neural tube defects in individuals with a mutation in a gene encoding one of the many folate-mediated enzymes leading to a less efficient metabolic handling of the available folate This brings us back to the concept of multifactorial inheritance

1.6 Mouse genetic models and Neural Tube Defect pathogenesis Although neural tube defects are quite frequent in humans, the cellular and molecular mechanisms that give rise to these developmental abnormalities are poorly understood In humans, examination of embryos around the time of neurulation is often not feasible on both ethical and practical grounds In the mouse, access to early embryogenesis is not restricted and, therefore, this animal has become an important in vivo system for studying NTD pathogenesis The study of mouse mutants with NTDs will also provide important insights into the etiology and pathogenesis of NTDs m humans

Naturally occurring mouse mutants Many mouse systems have been utilized to study NTD pathogenesis, including spontaneously arisen strains such as curly tail (ci), Splotch (Sp) and Loop tail (Lp) (Copp & Bernfield, 1994)

Curly tail In the curly tail (ci) mutant, first described by Gruneberg in 1954, postenor neuropore closure is delayed resulting from a reduced rate of cell proliferation in the notochord and hindgut just before the normal time of neuropore closure, whereas the overlying neuroepithehum is proliferating normally (Copp et al, 1988) This uncoordinated proliferation of cell layers results m a ventral curvature of the entire caudal region of the embryo and causes the delay or prevention of closure of the neuropore The ct mutation represents a single recessive mutation with incomplete penetrance Around 60 percent of homozygotes are affected to a variable degree, with the affected mice displaying either an isolated tail flexion defect (a curled or kinked tail) or lumbosacral spina bifida together with a tail defect A small percentage of embryos (1 to 5 percent) also exhibits exencephaly, a cranial NTD related to anencephaly in humans The molecular basis of the ct defect, to date, is unclear Interestingly, the NTD phenotype of the ct

25 Chapter 1 mutant displays apparently similar features to NTDs seen in humans (Neumann et al, 1994) As in humans, the predominant defects are lumbosacral, are slightly more common in males, and frequently associated with hydrocephalus Moreover, exencephaly is occasionally present and is more common in female animals

Splotch The original Splotch (Sp) mutant was first described by Russell and Roscoe (1947) and is classified as semidominant lethal In contrast to the ct mutation the basic defect in the Splotch (Sp) mouse appears to reside in the neuroepitheleum itself, most likely in the neural crest cells, resulting in reduced pigmentation in heterozygotes and small or absent dorsal root ganglia together with lumbosacral spina bifida m homozygous animals of which over half also exhibit exencephaly (Moase & Trasler, 1992) Mutations in РахЗ, a developmental control gene, turned out to be responsible for the Splotch phenotype (Epstein et al, 1991, Epstein et al, 1993)

Loop tail In the mutant Loop tail (Lp), the homozygote develops a lethal condition resembling craniorachischisis in humans, in which the neural tube is open throughout the brain and the spine Heterozygotes develop tail defects and occasionally spina bifida The embryonic defect in Loop-tail appears to involve either a general inability of the spinal neural folds to become apposed along the spinal region, or a defect in the process of neural fold fusion (Gerrelli & Copp, 1997) To date, the responsible gene has not been cloned

Bent tail Another interesting, but poorly documented mouse mutant is the Bent-tail (Bn) mouse for which the responsible gene has been localized on the X chromosome, but has not been isolated yet (Lyon & Searle, 1990) Both homozygous and heterozygous animals may have tail defects and open neural tube defects in the sacral region and/or cranial region The penetrance is incomplete This mouse mutant has been proposed as a model for X-lmked neural tube defects in humans

NTDs as a result of targeted inactivation of specific genes In addition to the natural mutant mouse strains, several artificial mutants displaying neural tube defects have recently been generated via gene targeting techniques (reviewed by Hams and Junloff, 1997) Disruption of the Csk gene (C-Src kinase) which encodes a negative regulator of the Src family tyrosine kinases, leads to a complex phenotype including neural tube defects (Imamoto & Sonano, 1993) Targeted modification of the Apo В gene results m incomplete

26 General introduction closure of the neural tube in a subset of homozygous embryos, and gives rise to hydrocephalus in some of the heterozygous animals (Homanics et al., 1993; Huang et al., 1995). Two recent reports identify the p53 gene with a potential role during neural embryonic development since a significant proportion of p53 homozygous knockout embryos failed to show closure of the neural tube resulting in exencephaly and subsequent anencephaly (Armstrong et al., 1995; Sah et al., 1995). Mice deficient for the breast and ovarian cancer susceptibility gene BRCA1 have been generated via targeted deletion. Most of these animals exhibit lethal abnormalities of the neural tube (Gowen et al., 1996). In another study, disruption of the BRCA1 gene gave rise to an even more severe phenotype (Hakem et al., 1996). Homozygous disruption of the AP2 transcription factor results in multiple congenital defects, among which cranial neural tube defects (Zhang et al, 1996; Schorle et al., 1996). Disruption of the Marcks gene results in exencephaly (Stumpo et al., 1995). Like Marcks the MacMarcks gene (also called F52) is a substrate for protein kinase С and disruption of this gene also leads to isolated neural tube defects, in particular anencephaly (Wu et al., 1996; Chen et al., 1996). The Cartl gene represents a paired-class homeobox gene that acts as a transcription factor and, when homozygously inactivated, leads to lethal cranial neural tube defects (Zhao et al., 1996). Inactivation of the c- Ski proto-oncogene results in lethal exencephaly in homozygous animals (Berk et al., 1997). The observation that mutations in a wide range of genes can lead to NTDs strongly suggests that proteins with numerous functions are required for neural tube formation. These include transcription factors, tumor supressors, signal transduction molecules and protein transport molecules. Some of them, in particular the ones associated with isolated NTD in mice, may be considered as likely candidates for neural tube defects in humans.

1.7 Evidence for multifactorial inheritance of Neural Tube Defects in mice Gene-gene interaction Several lines of evidence indicate that neural tube defects in the mouse, even in strains that display apparently monogenic inheritance of the trait, have a polygenic nature. It is well documented that for many mouse mutations manifestation of the phenotype is greatly influenced by the genetic background of the animal (Harris and Juriloff, 1997). This has led to the idea that the mouse genome contains modifying genes that are capable of modulating the incidence and severity of mutant phenotypes. For example, there are indications that at least three modifier loci are involved in the severity of the phenotype observed in homozygous curly tail mice, one of which has already been mapped (Letts et al., 1995). Another report has provided evidence for the existence of modifier loci in the Splotch mouse (Asher et al., 1996). Only a few studies have addressed the effect of the combined occurrence of distinct neural tube defect-promoting mutations in double mutant animals. By performing crosses between

27 Chapter 1 mice carrying independently segregating mutations, potential interactions of candidate genes can be evaluated. Preliminary findings suggest that interactions between non-allelic mutations may be responsible for the occurrence of specific NTDs. Estibeiro and colleagues (1993) described an interaction between the Splotch (Sp) and curly tail (ct) mutation in double mutants. They showed that defects in the mice carrying two mutant genes exceeded those of the corresponding single mutants in both incidence and severity. Juriloff et al. (1989) described the genetic analysis of the cause of exencephaly in the SELH mouse strain. About 17 percent of SELH fetuses are exencephalic and the liability to exencephaly in these mice appeared to be a multifactorial threshold trait, in which cranial neural tube defects result from additive interaction of two or three genes of minor effect. These loci are currently being mapped (Gunn et al., 1996). The authors state that the SELH mouse strain resembles human neural tube defects in type of genetic etiology and may be a valuable model in the study of neural tube defects. Helwig et al. (1995) crossed two independent natural mouse mutants, undulated (un) and Patch (Ph), respectively, and found that in the undulated-Patch double-mutant mice, (un/un Ph/+), a phenotype emerged resembling spina bifida occulta in humans. Moreover, similar results were obtained when crossing the heterozygous Patch mouse with tail kinks (tk), another recessive mutation responsible for vertebral anomalies (Imai et al., 1993). Again double mutants, tk/tk Ph/*, exhibit spina bifida occulta (R. Balling, personal communication) which illustrates the (poly)genetic complexity involved in the etiology of neural tube defects. The defective genes in undulated and Patch are Paxl and PDGFRa, respectively (Balling et al., 1988; Stephenson et al., 1991). The gene responsible for tail kinks has not been cloned yet. The results suggest that Paxl and PDGFRa act in the same functional pathway and that mutations in these genes may be involved in specific forms of NTD.

Gene-environment interactions A great many teratogenic agents have been identified as causing neural tube defects in rodent embryos (Copp et al., 1990), but in only a few cases have such teratogens been administered to embryos carrying neural tube defect-promoting mutations. These studies show that gene- teratogen interactions can either increase or decrease the frequency of defects. This has been examined in detail for retinole acid which has an interesting spectrum of interactions with the mutants Axial deformity (Axd), curly tail (ct), Loop tail (Lp), Splotch (Sp) and with strain SELH. Depending on the type of predisposed strain, the type of lesion, and the timing of administration, either an increase, a decrease, or no effect on the incidence and severity of the defects could be observed (Copp, 1994). For example, this teratogen promotes the occurence of spina bifida in Sp

28 General introduction mice where the mortality of homozygotes is increased (Moase & Trasler, 1987), whereas in ct mice spina bifida is prevented (Chen et al., 1994). In addition, nutritional supplements have been reported to affect the incidence and severity of NTDs in predisposed mouse strains. Methionine, an essential amino acid, has been shown to reduce the incidence of NTDs in the Axd mouse while folate has no effect (Essien & Wannberg, 1993). NTDs resulting from the action of the mutant gene ct may be prevented by administration of myo-inositol, a water-soluble B-complex vitamin (Greene & Copp, 1997), whereas folate and methionine have no effect on the incidence or the severity of the defect in ct mice (Seller, 1994b; Van Straaten et al., 1995). The anencephaly phenotype in the Cartl knock-out mouse can be suppressed by prenatal treatment with folic acid although the animals still die shortly after birth (Zhao et al., 1996). Further, it was shown recently that both folic acid and thymidine can prevent NTDs in Sp embryos, whereas methionine increases the number of fetal defects (Fleming & Copp, 1998).

In conclusion, a picture is emerging from studies on mouse mutants of a complex interacting network of neural tube defect-promoting genes, whose expression is influenced by genetic modifyers or environmental factors. It is likely that a similar etiological concept can also explain the incidence and transmission patterns of NTDs in human outbred populations. Here, the chance association of one or more mutated genes, together with a suitable combination of modifying genes and environmental factors, will give rise to apparently 'sporadic' cases of neural tube defects.

1.8 Pax genes and Neural Tube Defects Pax genes are developmental control genes that are expressed in the neuroepithelium before and during neural tube closure. Therefore, they are of particular interest since they may regulate cellular events associated with neurulation. Pax genes encode a family of transcription factors which have in common that they all contain a paired box motif that was originally identified in the Drosophila segmentation gene paired (reviewed by Strachan & Read, 1994). The paired box spans about 384 nucleotides and specifies a highly conserved DNA-binding protein element of about 128 amino acids called the paired domain. In addition to the paired domain, some Pax genes also possess a homeodomain, which is another DNA-binding motif. As expected for genes with a critically important function, they show very strong sequence conservation between distantly related species (Noll, 1993). At present, nine different members of this family are known in mouse (Paxl to Pax9) and man (PAX1 to PAX9), which can be subgrouped in four classes based on the location of introns, the presence of a homeobox and of an octapeptide coding region, and especially on degree of identity between the paired box sequences (Fig 5).

29 Chapter J

Pax genes in the same structural class tend to show similar expression patterns, although there are specific differences. In the mouse, all Pax genes, except Paxl and Pax9, are expressed in the neuroepithelium of the developing neural tube around the time of fusion. Paxl and Pax9 are expressed in mesodermal derived tissues from which the vertebral column develops.

Chromosomal Location

Paired Domain Oc Homeodomain Mouse Human PAX1 -0- 2 20p11 PAX9 -Ш—амии—D- 12 14q12-q13

PAX3 1 2q35

PAX7 4 1p36

PAX2 -C 19 10q25 PAX5 -C 4 9p13 PAX8 -C 2 2q12-q14

PAX4 -GSS^S 6 7 PAX6 -Ш^^^^^^ШЬ 2 11p13

Figure 5: The Pax gene family (Modified from Strachan & Read 1994).

Mutant phenotypes of Pax genes correlate very well with the observed expression patterns. For example, Рахб is expressed during eye formation and is mutated in small eye (sey) mice and in human aniridia (Hill et al., 1991; Jordan et al., 1992). Pax2 is expressed during eye and kidney development and is mutated in a human family with kidney and eye abnormalities (Sanyanusin et al., 1995). Paxl is expressed in the sclerotome and is required for normal development of ventral vertebral structures. The gene is mutated in undulated (un) mice suffering from axial skeleton abnormalities (Balling et al., 1988). Analysis of undulated mice demonstrates that Paxl is essential for the condensation of the mesenchymal sclerotome cells and for the initiation of chondrogenesis (Wallin et al., 1994). As already outlined, mutations in Paxl have also been implicated in the development of a specific form of spina bifida in the mouse (Helwig et al., 1995). In the context of the present study, the РахЗ gene is of particular interest. The gene is expresssed in the limb muscle, neural tube and neural crest, and is inactivated by deletions or point mutations in various murine Splotch (Sp) alleles (Goulding et al., 1991; Epstein et al.,

30 General introduction

1991; Epstein et al, 1993). It has already been mentioned that, the mouse mutant Splotch has long been recognised as a good model for human neural tube defects (Moase & Trasler, 1992). In humans, mutations in the PAX3 gene cause Waardenburg syndrome (type 1 & 3). Waardenburg syndrome is an autosomal dominant condition characterized by varying combinations of deafness, dystopia canthorum and pigmentary disturbances of the eyes, hair and skin caused by impaired migration of embryonic neural crest cells (reviewed by Read & Newton, 1997). Occasionaly, spina bifida has been reported as part of the phenotype in patients with Waardenburg syndrome type 1 (Chatkupt et al, 1993; Moline & Sandlin, 1993; Baldwin et al., 1995; this thesis). Recently, in a specific strain of diabetic mice displaying a high incidence of NTD, it was shown that РахЗ expression is significantly reduced in the developing neural tube (Phelan et al., 1997). These findings suggest a role for РахЗ in diabetic embryopathy which -also in humans- is associated with an increased risk for NTD (Myrianthopoulos & Melnick, 1987). Accordingly, the human РАХЗ gene presents itself as an excellent candidate for neural tube defects in man.

1.9 Folic acid and Neural Tube Defects Primary prevention of neural tube defects through folate supplementation One of the most exciting findings of the last decades is the fact that folic acid, a compound belonging to the group of В vitamins, can prevent serious birth defects such as spina bifida and anencephaly when administered to the mother, from about about one month before to eight- twelve weeks after conception. Already back in 1965 Hibbard & Smithells suggested that folate might have a role in the etiology of neural tube defects. The group of Smithells (1980) was the first to supplement women who have had a previous NTD pregnancy with a multivitamin tablet that included folate, which led to a marked reduction in recurrences of NTDs. The Medical Research Council trial (MRC, 1991) and the study of Czeizel & Dudas (1992) convincingly showed that folic acid alone is sufficient to reduce both recurrences and occurrences of NTD. It has been estimated that approximately 50 percent of all NTD cases can be prevented through folate supplementation. Although, the preventive effect of folate is well-documented by now, very little is known about the underlying mechanism through which folate exerts its effect. The folate status is reflected in the serum and red blood cells levels. Serum folate reflects recent dietary intake and falls rapidly upon folate-depleted dietary regimes, whereas red blood cell folate levels are not subject to daily fluctuations reflecting instead the average folate status over the previous 120 days when the circulating populations of red cells were synthesized. Studies on the folate status of women who had NTD offspring were sometimes conflicting, but overall there is a trend for such women to have slightly lower than normal levels, particularly in their red blood cells

31 Chapter 1 although almost always, levels are still m the normal range (Yates et al, 1987, Wald, 1994) This suggests that the cause of the defect is a metabolic block or increased requirement for one of the many folate-dependent enzymes rather than simple folate deficiency Additional evidence for this hypothesis comes from studies in the mouse where it was shown that folate deficiency is not by itself sufficient to produce neural tube defects (Heid et al, 1992) A study by Seller (1995d), m which the type of defect was evaluated in a group of 13 children who displayed NTDs despite the fact that their mother had taken folate around the time of conception, suggested that folic acid deficiency does not act specifically on one particular closure site of the neural tube Hook and Czeizel (1997) reported on a possible association between folic acid supplementation and increased prevalence of spontaneous abortion suggesting that folic acid diminishes the rate of birth defects by selectively inducing abortions, a phenomenon referred to as 'terathanasia' However, others have disputed this hypothesis because of the borderline statistic significance of the data (Bum & Fisk, 1997) Moreover, the effect can be explained very well by the fact that folic acid could permit the survival of otherwise non-vtable embryos to a point when they are recognised as abortions (Hall, 1997, Schorah et al, 1997) Recently, Fleming and Copp (1998) have shown that the Splotch mutation is associated with impaired folic acid metabolism and that folic acid can prevent the development of NTDs in these genetically predisposed embryos This strengthens the argument for primary prevention of human NTDs by folic acid and suggests that it is unlikely that folic acid promotes the abortion of affected fetuses

Folic acid-homocysteine metabolism Folate acts as cofactor for enzymes involved in de novo DNA and RNA biosynthesis and is also involved in the supply of methyl groups to the so-called methylation cycle (Fig 6), which uses methionine and produces homocysteine (reviewed by Fowler, 1997) The major form of folate in serum is 5-methyltetrahydrofolate This is not only the major transport form of folate, it is the principal methyl donor for methionine synthase, a vitamin B12-requinng enzyme Methionine, an essential amino acid, loses its methyl group via reactions involving S-adenosyl methionine The resulting homocysteine can be re-methylated in the presence of adequate amounts of 5-methyltetrahydrofolate and vitamin B12 to regenerate methionine, thus conserving the supply of this essential nutrient Failure of regeneration may result m the accumulation of potentially toxic homocysteine, which produces NTDs in chick embryos but not m the rat (Rosenquist et al, 1996, van Aerts et al, 1994)

32 от

от _ η V 2 £* Ι­ л ο

чч 0) о м M £ ОТ >· η о С F с 3 >l > ОТ>· ОТ от ε С «< ОТ ι- ρ С «fe, о af "S о IО. 5 от о о 1 ν \ о ε с \ _1 0) m \ о _І •о \4 a о>% ш ш о ω E о о ε g >(Л. *" о о E с о Ш X о δ J о < υ с/э "i* 0) о s s •а го к 5 о E го гоЕ >.£ CL ve £ о ÍS < 3 » -о = П) «> E >· го о со ь •"гас и. о о і= о Е. 53 E h. Chapter 1

Recent reports on increased mean homocysteine levels in mothers of NTD offspring argue for a role of one or more folate related enzymes in the etiology of folate preventable neural tube defects (Steegers-Theunissen et al., 1994; Mills et al., 1995; Lucock et al., 1997). Mutations affecting the genes encoding such enzymes may result in reduced activity, leading to mildly increased homocysteine levels that might lead to a higher risk for NTD offspring. The metabolic effects of some of these mutations may be overcome by folic acid supplementation, explaining the protective effect of this vitamin. Accordingly, one of the enzymes that has been studied is the 5,10-methylenetetrahydrofolate reductase (MTHFR) which catalyzes the reduction of 5,10-methylenetetrahydrofolate to 5- methyltetrahydrofolate (Fig 6). Frosst et al. (1995) identified a common thermolabile polymorphism (677C->T) in the gene encoding this protein, which is associated with decreased enzyme activity and elevated mean homocysteine levels. Interestingly, van der Put et al. (1995) found that homozygosity for this MTHFR variant is associated with spina bifida in a Dutch population, suggesting that this mutation is a genetic risk factor in NTD. Similar observations were made in Irish and US populations, respectively (Whitehead et al., 1995; Kirke et al., 1996; Ou et al., 1996). Conversely, studies among a British, French and German cohorts, respectively, failed to detect any association (Papapctrou et al., 1996; Morrison et al., 1998; Momet et al., 1997; Koch et al., 1998). Both the British and the German investigators applied the transmission / disequilbrium test (TDT), which is a 'within family' test in which parents of affected individuals are used as internal controls, and hence requires no separate series of matched controls (Spiclman et al., 1993). This may be of importance since the observed homozygote frequency for the thermolabile MTHFR allele, seems to vary considerably between different populations (Wilcken & Wang, 1996). Accordingly, the ethnic background of study participants should be documented whenever the prevalence of the thermolabile mutation is evaluated through traditional population based case-control association studies. More definite conclusions regarding the role of the thermolabile MTHFR allele in neural tube defect etiology have to await the results of 'within family' allele association studies. Nonetheless, it has been estimated that homozygosity for the thermolabile MTHFR allele, presuming it represents a genuine risk factor, can explain only about one-quarter of the folate preventable neural tube defects, meaning that additional genes are likely to be involved (van der Put et al., 1996; Bjorke-Onsen et al., 1997).

Methionine synthase (MS) represents a folate- and vitamin B,2 dependent key enzyme in the folate-homocysteine metabolism, which is responsible for the re-methylation of homocysteine to methionine (Fig 6). The mild hyperhomocysteinemia observed in mothers of an NTD child can be explained by mutations in the MS gene which affect the activity of the corresponding protein (Lucock et al., 1997). Recently, the cDNA of the MS gene has been cloned and mutations causing the inborn error of metabolism cblG have been described (Ledere et al., 1996).

34 General introduction

However, association studies, applying the TDT test, revealed no significant correlation between NTD susceptibility and some specific alleles at the MS locus in a combined British and Dutch cohort (Morrison et al., 1997; Morrison et al., 1998). Moreover, no relevant mutations were found in the coding region of the MS gene in a small sample of hyperhomocysteinaemic NTD patients (van der Put et al., 1997). Although the MS gene appears to be a good candidate for NTD, no genetic evidence supporting this hypothesis has been reported thus far. The folate receptor alpha (FR-a) is another important folate related gene. The corresponding protein is involved in transportation of folate into the cell. The gene has recently been tested as a candidate for NTD in a large sample of patients however, no genetic variation was found in the coding region of the gene (Barber et al., 1998).

1.10 Obj ectives of the present study Primary prevention is the ultimate aim of all those working in the field of congenital abnormalities. By primary prevention we mean preventing the abnormality to occur in the first place. This is of course more desirable than secondary prevention, which refers to the diagnosis of NTDs in utero and the subsequent termination of affected pregnancies. This latter form of 'prevention' places a lifelong psychological burden on parents faced with the choice between ending a desired pregnancy or giving birth to a (severely) handicapped child. Therefore, any form of primary prevention is welcome for both medical and ethical reasons. The discovery of the protective effect of maternal folic acid supplementation has opened the way to preconceptional intervention. It has become clear, however, that folate administration can only prevent about 50 percent of NTD cases, which means that additional intervention strategies are needed. Identification of genes involved in the etiology of NTD will lead to a better understanding of the pathogenic mechanism(s) underlying these congenital abnormalities and will be of importance for the development of novel therapies preventing them to occur. Moreover, an understanding of the molecular basis of predisposition will open the way to more effective targeting of preventive strategies towards individuals at increased genetic risk. As yet, very little is known about the genetic factors involved in the etiology of NTD. Accordingly, the main objective of the present study is the identification of genetic factors predisposing to NTD in humans. Chapter 1 of this thesis serves as an introduction and a review on NTD in man and mouse. Chapter 2 describes the pedigree analysis that has been performed in an Icelandic family with apparent X-linked NTD in order localize the responsible gene in this family. Through review of the literature, several members of the PAX gene family were identified as excellent candidates for neural tube defects in man. In chapter 3 we excluded linkage between PAX3 and neural tube defects in a cohort of Dutch and US families. Subsequently, we conducted mutation

35 Chapter I analysis of the complete coding region of the PAX3 gene as well as the paired domain regions of PAX1, PAX7 and PAX9 in a large panel of patients. Chapter 4 reports on a mutation in the gene for PAX3 in a patient with spina bifida and Waardenburg syndrome whereas chapter 5 deals with a mutation identified in the paired domain of the PAX1 gene of a fetus with spina bifida. In chapter 6 we present evidence that PAX1 and PDGFRa may act in the same functional pathway. Moreover, we show that the PAX1 mutation, as identified in one of our patients with spina bifida, affects the functionality of the corresponding PAX I protein. Finally, in chapter 7 we set out to investigate the gene of an important folate related enzyme involved in homocysteine metabolism, i.e. the methylenetetrahydrofolate dehydrogenase gene (MTHFD).

36 General introduction

References Armstrong-JF; Kaufman-MH; Hamson-DJ; Clarke-AR. High-frequency developmental abnormalities in p53-deficient mice. Curr-Biol (1995) 5: 931-6

Asher-JH Jr; Harnson-RW; Morell-R; Carey-ML; Fnedman-TB. Effects of РахЗ modifier genes on craniofacial morphology, pigmentation, and viability: a murine model of Waardenburg syndrome variation. Genomics (1996)34: 285-98

Baldwin CT; Hoth CF; Macina RA; Milunsky A. Mutations in PAX3 that cause Waardenburg syndrome type I: ten new mutations and review of the literature. Am-J-Med-Genet (1995)58: 115-22

Balhng-R; Deutsch-U; Gruss-P. undulated, a mutation affecting the development of the mouse skeleton, has a point mutation in the paired box of Pax 1. Cell (1988)55: 531-5

Baraitser-M; Bum-J. Neural tube defects as an X-hnked condition. Am-J-Med-Genet (1984)17: 383-5

Barber-RC; Shaw-GM; Lammer-EJ; Greer-ΚΑ; Biela-TA; Ілсеу-SW; Wasserman-CR; Finnell-RH. Lack of association between mutations in the folate receptor-alpha gene and spina bifida. Am-J-Med- Genet(1998)76: 310-317

Berk-M; Desai-SY; Heyman-HC; Colmenares-C. Mice lacking the ski proto-oncogene have defects m neurulation, craniofacial, patterning, and skeletal muscle development. Genes-Dcv (1997)11: 2029-39

Bjorke-Monsen-AI; Ueland-PM; Schneede-J; Vollset-SE; Refsum-H. Elevated plasma total homocysteine and C677T mutation of the methylenetetrahydrofolate reductase gene in patients with spina bifida. Q-J-Med (1997) 90593-596

Boone-D; Parsons-D; Lachmarm-SM; Sherwood-T. Spina bifida occulta: Lesions or anomaly? Chn- Radiol (1985)36:159-161

Brook-FA; Estibeiro-JP; Copp-AJ. 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. J- Med-Genet (1994) 31: 383-7

Browder-LW; Enckson-CA; Jeffery-WR. Developmental Biology. (1991) 3rd ed. Saunders College Publishing, USA

Buamah-PK; Russell-M; Bates-G; Ward-AM; Skillen-AW. Maternal zinc status: a determination of central nervous system malformation. Br-J-Obstet-Gynaecol (1984)91: 788-90

Bum-J; Fisk-NM. Terathanasia, folic acid, and birth defects. Lancet (1997)350:1322

Byme-J; Cama-Α; Reilly-M; Vigharolo-M; Levato-L; Boni-L, Lavia-N; Andreussi-L. Multigeneration maternal transmission in Italian families with neural tube defects. Am-J-Mcd-Gcnct (1996)66: 303-10

Campbell-LR, Dayton-DH, Sohal-GS. Neural tube defects: A review of human and animal studies on the etiology of neural tube defects. Teratology (1986)34:171-187

Carter-CO. Clues to the aetiology of neural tube malformations. Dev-Med-Child-Neurol (1974) 16 (Supp 32): 3-15

37 Chapter 1

Chatkupt-S; Lucek-PR; Koenigsberger-MR; Johnson-WG. Parental sex effect in spina bifida: a role for genomic imprinting? Am-J-Med-Genet (1992)44: 508-12

Chatkupt-S; Chatkupt-S; Johnson-WG. Waardenburg syndrome and myelomeningocele in a family. J- Med-Genet (1993) 30:83-84

Chatkupt-S; Hol-FA; Shugart-YY; Geurds-MP; Stenroos-ES; Koenigsberger-MR; Hamel-ВС; Johnson- WG; Manman-ЕС Absence of linkage between familial neural tube defects and PAX3 gene. J-Med- Genet( 1995) 32: 200-4

Chen-WH; Momss-Kay-GM; Copp-AJ. Prevention of spinal neural tube defects in the curly tail mouse mutant by a specific effect of retinole acid. Dcv-Dyn (1994) 199: 93-102

Chen-J; Chang-S; Duncan-SA; Okano-HJ; Fishell-G; Aderem-A Disruption of the MacMARCKS gene prevents cranial neural tube closure and results in anencephaly. Proc-Natl-Acad-Sci-USA (1996) 93: 6275-9

Copp-AJ; Brook-FA; Roberts-Ш. A cell-type-specific abnormality of cell proliferation m mutant Curly tail) mouse embryos developing spinal neural tube defects. Development (1988)104: 285-95

Copp-AJ; Brook-FA; Estibeiro-JP; Shum-AS; Cockroft-DL. The embryonic development of mammalian neural tube defects. Prog-Neurobiol (1990)35: 363^03

Copp-AJ; Bernfield-M. Etiology and pathogenesis of human neural tube defects: insights from mouse models. Сшт-Opin-Ped (1994) 6: 624-631

Copp-AJ. Genetic models of mammalian neural tube defects. Oba-Found-Symp (1994)181: 118-34

Cuckle-HS; Wald-NJ; Althouse-R. Sex differences in the location of a spina bifida lesion. J-Med-Genet. (1993)30:262-3

Czeizel-AE; Dudas-I. Prevention of the first occurrence of neural-tube defects by penconceptional vitamin supplementation. N-Engl-J-Med (1992)32: 1832-5

Dallaire-L; Melancon-SB; Potier-M; Mathieu-JP; Ducharme-G. Date of conception and prevention of neural tube defects. Chn-Genet (1984) 26: 304-7

Dib-C; Faure-S; Fizames-C; Samson-D; Drouot-N; Vignal-A; Millasseau-P; Marc-S; Hazan-J; Seboun- E; Lathrop-M, Gyapay-G; Monssette-J; Weissenbach-J. A comprehensive genetic map of the human genome based on 5,264 microsatellites. Nature (1996)380: 152-4

Dolk-Η; De Wals-P; Gillerot-Y; Lechat-MF; Ayme-S; Comel-M; Cuschien-A; Game-Ε; Goujard-J; Laurence-KM; Lilhs-D; Lys-F; Nevm-N; Owens-J; Radic-A; Stoll-C; Stone-D; Ten Kate-L. Heterogeneity of neural tube defects in Europe: The significance of site of defect and presence of other major anomalies in relation to geographic differences m prevalence. Teratology (1991)44: 547-559

Drainer-Ε; May-HM; Tolmie-JL. Do familial neural tube defects breed true? J-Med-Genet. (1991)28: 605-8

Elwood-JM; Elwood-JH. Epidemiology of anencephalus and spina bifida. (1980) Oxford: Oxford university press.

38 General introduction

Epstein-DJ; Vekemans-M; Gros-P. Splotch (Sp2H), a mutation affecting development of the mouse neural tube, shows a deletion within the paired homeodomain of Pax-3. Cell (1991)67: 767-74

Epstein-DJ; Vogan-KJ; Trasler-DG; Gros-P. A mutation within intron 3 of the Pax-3 gene produces aberrantly spliced mRNA transcripts in the Splotch (Sp) mouse mutant. Proc-Natl-Acad-Sci-USA (1993) 90: 532-6

Epstein-JA; Shapiro-DN; Cheng-J; Lam-PY; Maas-RL. РахЗ modulates expression of the c-Met receptor during limb muscle development. Proc-Natl-Acad-Sci-USA (1996)93: 4213-8

Essien-FB; Wannberg-SL. Methionine but not fohnic acid or vitamin B-12 alters the frequency of neural tube defects in Axd mutant mice. J-Nutr (1993) 123: 27-34

Estibeiro-JP; Brook-FA; Copp-AJ. Interaction between Splotch (Sp) and curly tail (ct) mouse mutants in the embryonic development of neural tube defects. Development (1993)119: 113-21

EUROCAT Working Group. Prevalence of neural tube defects in 20 regions of Europe and the impact of prenatal diagnosis, 1980-1986. J-Epidemiol-Commumty-Health (1991)45: 52-8

Falconer-DS. Introduction to quantitative genetics. (1981) 2nd ed. Longman, London

Fidas-A; MacDonald-HL; Elton-RA; Wild-SR; Chisholm-GD; Scott-R. Prevalence and patterns of spina bifida occulta in 2,707 normal adults. Clin-Radiol (1987)38: 537-42

Flermng-A; Copp-AJ. Embryonic folate metabolism and mouse neural tube defects. Science (1998)280: 2107-09

Fowler-B. Disorders of homocysteine metabolism. J-Inher-Metab-Dis (1997)20: 270-85

Frosst-P; Blom-HJ; Milos-R; Goyctte-P; Sheppard-CA; Matthews-RG; Boers-GJ; den-Heijer-M; Kluijtmans-LA; van-den-Heuvel-LP; et-al. A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat-Genet (1995) 10: 111 -3

Fryns-JP; Devnendt-K; Moerman-Ph. Lumbosacral spina bifida and myeloschizis m a female foetus with de novo X/autosomaltranslocation (t(X;22)(q27;ql21)). Genet-Couns (1996)7:159-60

Garabedian-BH; Fraser-FC. A familial association between twinning and upper-neural tube defects. Am- J-IIum-Genet(1994)S5: 1050-3

Gerrelh-D; Copp-AJ. Failure of neural tube closure in the loop-tail (Lp) mutant mouse: analysis of the embryonic mechanism. Bram-Res-Dev-Brain-Res (1997)102: 217-24

Gocrss-JB; Kames-PS; Thibodeau-SN; Johnson-DD; Zimmerman-D et al. Cytogenetic and molecular studies of a duplication of Xq26 and Xq27 in two brothers with neural tube defects. Am-J-Hum-Genet (1993) 53(Suppl) 440(abstract)

Golden-J A; Chemoff-GF. Intermittent pattern of neural tube closure in two strains of mice. Teratology (1993)47:73-80

Gouldmg-MD; Chalepakis-G; Deutsch-U; Ersehus-JR; Gruss-P. Pax-3, a novel murine DNA binding protein expressed during early neurogenesis. EMBO-J (1991) 10: 1135-47

39 Chapter 1

Gowen-LC; Johnson-BL; Latour-AM, Suhk-KK, Koller-BH Brcal deficiency results in early embryonic lethality characterized by neuroepithelial abnormalities Nat-Genet (1996)12: 191-4

Greene-ND; Copp-AJ. Inositol prevents folate-resistant neural tube defects in the mouse. Nat-Med (1997)3:60-6

Gruhneberg-H. Genetical studies on the skeleton of the mouse. Ш. Curly tail. J-Genet 1954 52. 52-54

Gunn-TM; Junloff-DM; Hams-MJ. Exencephaly and cleft cerebellum in SELH/Bc mouse embryos are alternative developmental consequences of the same underlying genetic defect. Teratology (1996)54: 230-6

Hakem-R; de-la-Pompa-JL; Sirard-C; Mo-R; Woo-M, Hakem-A; Wakeham-A; Potter-J; Reitmair-A; Bilha-F; Firpo-Ε; Hui-CC, Roberts-J; Rossant-J; Мак-TW. The tumor suppressor gene Brcal is required for embryonic cellular proliferation in the mouse. Cell. (1996)85: 1009-23

Hall-JG; Fnedman-JM; Kenna-BA; Popkin-J; Jawanda-M; Amold-W. Clinical, genetic, and epidemiological factors in neural tube defects. Am-J-Hum-Gcnet (1988)43: 827-37

Hall-JG. Terathanasia, folic acid, and birth defects. Lancet (1997)350 1322

Hams-MJ; Junloff-DM. Genetic landmarks for defects in mouse neural tube closure. Teratology (1997) 56 177-187

Heid-MK; Bills-ND; Hinnchs-SH; Chfford-AJ. Folate deficiency alone does not produce neural tube defects in mice. J-Nutr (1992) 122: 888-94

Helwig-U; Imai-K; Schmahl-W; Thomas-BE; Vamum-DS; Nadeau-JH; Ballmg-R. Interaction between undulated and Patch leads to an extreme form of spina bifida in double-mutant mice. Nat Genet Nat Genet (1995) 11. 60-63

Hibbard-ED; Smithells-RW Folic acid metabolism and human embryopathy. Lancet (1965)1. 1254-56

Hill-RE, Favor-J; Hogan-BL; Ton-CC; Saunders-GF; Hanson-IM, Presser-J; Jordan-T; Hastie-ND; van- Heyningen-V. Mouse small eye results from mutations in a paired-like homeobox-containing gene Nature (1991) 354-522-5

Homanics-GE; Smith-TJ; Zhang-SH; Lee-D; Young-SG; Maeda-N. Targeted modification of the apohpoprotem В gene results in hypobetalipoprotememia and developmental abnormalities in mice. Proc-Natl-Acad-Sci-USA (1993) 90: 2389-93

Hook-EB; Czeizel-AE. Can terathanasia explain the protective effect of folic-acid supplementation on birth-defects? Lancet (1997) 350. 513-15

Huang-LS; Voyiaziakis-E; Markenson-DF; Sokol-KA; Hayek-T; Breslow-JL. apo В gene knockout in mice results in embryonic lethality in homozygotes and neural tube defects, male infertility, and reduced HDL cholesterol ester and apo Α-I transport rates in heterozygotes J-Clin-Invest ( 1995) 96. 2152-61

Hui-CC; Joyner-AL. A mouse model of greig cephalopolysyndactyly syndrome: tins extra-toes mutation contains an intragenic deletion of the Gh3 gene. Nat-Genet (1993)3. 241-6

40 General introduction

Hum-DW; Bell-AW; Rozen-R; MacKenzie-RE. Pnmaiy structure of a human tnfiinctional enzyme. Isolation of a cDNA encoding methylenetetrahydrofolate-dehydrogenase methenyltetrahydrofolate- cyclohydrolase formyltetrahydrofolate-synthetase. J-Biol-Chem (1988)263: 15946-50

Imai-K; Nass-SJ; Olowson-M; Balling-R. The genetic map around the tail kinks (tk) locus on mouse chromosome 9. Mamm-Genome (1993)4: 560-4

Imamoto-A; Sonano-P. Disruption of the csk gene, encoding a negative regulator of Src family tyrosine kinases, leads to neural tube defects and embryonic lethality in mice. Cell (1993)73: 1117-24

Jensson-O; Amason-A; Gunnarsdottir-H; Petursdottir-I; Fossdal-R; Hreidarsson-S. A family showing apparent X linked inheritance of both anencephaly and spina bifida. J-Med-Genet (1988)25: 227-9

Jordan-T; Hanson-I; Zaletayev-D; Hodgson-S; Prosser-J; Seawnght-A; Hastie-N; van-Heynmgen-V. The human PAX6 gene is mutated in two patients with aniridia. Nat-Genet (1992)1: 328-32

Kallen-B; Cocchi-G; Knudsen-LB; Castilla-EE; Robcrt-E; Daltveit-AK, Lancaster-PL, Mastroiacovo-P. International study of sex ratio and twinning of neural tube defects. Teratology (1994)50: 322-31

Kang-S; Graham-JM Jr; Olney-AH; Biesecker-LG. GLI3 frameshift mutations cause autosomal dominant Palhster-Hall syndrome. Nat-Genet (1997) 15: 266-68

Khoury-MJ; Enckson-JD; James-LM. Etiologic heterogeneity of neural tube defects. II. Clues from family studies. Am-J-Hum-Genet (1982)34: 980-7

Kirke-PN; Mills-JL; Whitehead-AS; Molloy-A; Scott-JM. Methylenetetrahydrofolate reductase mutation and neural tube defects. Lancet (1996)348: 1037-8

Koch-MC; Stegmann-K; Ziegler-A; Schröter-B; Ermert-A. Evaluation of the MTHFR C677T allele and the MTHFR gene locus in a German spina bifida population. Eur J Pediatr (1998)157: 487-92

Lander-ES; Schork-NJ. Genetic dissection of complex traits. Science (1994)265: 2037-48

Lathrop-GM; Lalouel-JM; Juher-C; Ott-J. Multilocus linkage analysis in humans: detection of linkage and estimation of recombination. Am-J-Hum-Genet (1985)37: 482-98

Layde-PM; Edmonds-LD; Enckson-JD. Maternal fever and neural tube defects. Teratology (1980) 21: 105-8

Leck-I. The geographical distribution of neural tube defects and oral clefts. Br-Med-Bull (1984) 40. 390- 5

Leclerc-D; Campeau-E; Goyette-P; Adjalla-CE; Chnstensen-B; Ross-M; Eydoux-P; Rosenblatt-DS; Rozen-R; Gravel-RA. Human methionine synthase: cDNA cloning and identification of mutations m patients of the cblG complementation group of folate/cobalamin disorders. Hum-Mol-Genet (1996)5: 1867-74

Letts-VA; Schork-NJ; Copp-AJ; Bemfield-M; Frankcl-WN. A curly tail modifier locus, mctl, on mouse chromosome 17. Genomics (1995) 29. 719-24

Lindhout-D; Schmidt-D. In-utero exposure to valproate and neural tube defects. Lancet (1986) 1: 1392- 93 41 Chapter 1

Lucock-MD; Wild-J; Lumb-CH; Oliver-M; Kendall-R; Daskalakis-I; Schorah-CJ; Levene-MI. Risk of neural tube defect-affected pregnancy is associated with a block in maternal one-carbon metabolism at the level of N-5-methyltetrahydrofolate:homocysteine methyltransferase. Biochem-Mol-Med (1997) 61: 28^0

Lynch-SA; Bond-PM; Copp-AJ; Kirwan-WO; Nour-S, Balling-R, Manman-Ε; Burn-J; Strachan-T. A gene for autosomal dominant sacral agenesis maps to the holoprosencephaly region at 7q36. Nat-Genet (1995)11:93-5

Lyon-MF; Searle-AG. Genetic variants and strains of the laboratory mouse. (1990) 2nd ed. New York: Oxford University Press

Manman-ECM; Hamel-BCJ. Sex ratios of affected and transmitting members of multiple case families with neural tube defects. J-Med-Genet (1992) 29: 695-8

Mills-JL; McPartlm-JM; Kirke-PN; Lee-YJ; Conley-MR; Weir-DG; Scott-JM. Homocysteine metabolism m pregnancies complicated by neural-tube defects. Lancet (1995)345: 149-51

Milunsky-A; Moms-JS; Jick-H; Rothman-KJ; Ulcickas-M; Jick-SS; Shoukimas-P; Willett-W. Maternal zinc and fetal neural tube defects. Teratology (1992a) 46: 341-8

Milunsky-A; Ulcickas-M; Rothman-KJ; Willett-W; Jick-SS; Jick-H. Maternal heat exposure and neural tube defects. JAMA (1992b) 268: 882-5

Moase-CE; Trasler-DG. Retinole acid-induced selective mortality of splotch-delayed mouse neural tube defect mutants. Teratology (1987) 36: 335-43

Moase-CE; Trasler-DG. N-CAM alterations in Splotch neural tube defect mouse embryos. Development (1991)113: 1049-1058

Moase-CE; Trasler-DG Splotch locus mouse mutants: models for neural tube defects and Waardenburg syndrome type I in humans. J Med Genet (1992) 29: 145-51

Molme-ML; Sandhn-C. Waardenburg syndrome and meningomyelocele. Am-J-Med-Genet (1993)47: 126

Mornet-E; Muller-F; Lenvoise-Furet-A; Delezoide-AL; Col-JY; Simon-Bouy-B; Serre-JL. Screening of the C677T mutation on the methylenetetrahydrofolate reductase gene in French patients with neural tube defects. Hum-Genet (1997) 100: 512^t

Morrison-K; Edwards-YH; Lynch-SA; Bum-J; Hol-FA; Manman-ECM. Methionine synthase and neural tube defects. J-Med-Genet (1997) 34: 958

Momsson-K; Papapetrou-C; Hol-FA; Manman-ECM; Lynch-SA; Bum-J; Edwards-YH. Susceptibility to spina bifida; an association study of five candidate genes. Ann Hum Genet (1998)62: 379-396

Momss-Kay-G; Wood-Η; Chen-WH. Normal neurulation in mammals. Ciba-Found-Symp (1994)181: 51-63

MRC Vitamin Study Research Group. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet (1991)338: 131-7

42 General introduction

Mueller-RF; Young-ID. Emery's elements of medical genetics. (1995) 9th ed. London; Churchill Livingstone

Muller-F; O'Rahilly-R. The development of the human brain, the closure of the caudal neuropore, and the beginning of secondary neurulation at stage 12. Anat-Embryol (1987) 176: 413-30

Mynanthopoulos-NC; Melmck-M. Studies in neural tube defects. I. Epidemiologic and enologie aspects. Am-J-Med-Genet (1987) 26: 783-96

Neumann-PE; Frankel-WN; Letts-VA; Coffin-JM; Copp-AJ; Bemfield-M. Multi factorial inheritance of neural tube defects: localization of the major gene and recognition of modifiers met mutant mice. Nat- Genet. (1994) 6: 357-62

Noll-M. Evolution and role of Pax genes. Curr-Opin-Genet-Dev (1993)3: 595-605

Oman-Ganes-L; Shokeir-MHK. Neural tube defects: possible X-lmked inheritance. Am-J-Hum-Genet (1984)36: 67(S) 195(abstract)

O'Rahilly-R; Muller-F. Neurulation in the normal human embryo. Ciba-Found-Symp (1994)181: 70-82

Ou-CY; Stevenson-RE; Brown-VK; Schwartz-CE; Allen-WP; Khoury-MJ; Rozen-R; Oakley-GP Jr; Adams-MJ Jr. 5,10 Methylenetetrahydrofolate reductase genetic polymorphism as a risk factor for neural tube defects. Am-J-Med-Genet (1996)63: 610-4

Paavola-P; Salonen-R; Weissenbach-J; Peltonen-L. The locus for Meckel syndrome with multiple congenital anomalies maps to chromosome 17q21-q24. Nat-Genet (1995)11: 213-5

Papapetrou-C; Lynch-SA; Bum-J; Edwards-YH. Methylenetetrahydrofolate reductase and neural tube defects. Lancet (1996) 348: 58

Payne-J; Shibasaki-F; Mercola-M. Spina bifida occulta m homozygous Patch mouse embryos. Dev-Dyn (1997)209: 105-16

Phelan-SA; Ito-M; Loeken-MR. Neural tube defects m embryos of diabetic mice: role of the Pax-3 gene and apoptosis. Diabetes (1997) 46: 1189-97

Puffenberger-EG; Kauffman-ER; Bolk-S; Matise-TC; Washington-SS; Angnst-M; Weissenbach-J; Garver-KL; Mascan-M; Ladda-R; et-al. Identity-by-descent and association mapping of a recessive gene for Hirschsprung disease on human chromosome 13q22. Hum-Mol-Genet (1994)3: 1217-25

Read-AP; Newton-VE. Waardenburg syndrome. J-Med-Genet (1997) 34: 656-665

Rodriguez JI, Garcia M, Morales C, Morillo A, Delicado A. Trisomy 13 syndrome and neural tube defects. Am J Med Genet ( 1990) 36: 513-6

Rosenquist-TH; Ratashak-SA; Selhub-J. Homocysteine induces congenital defects of the heart and neural tube: effect of folic acid. Proc-Natl-Acad-Sci-USA (1996)93: 15227-32

Ross-AJ; Ruiz Perez-V; Wang-YM; Hagan-DM, Scherer-S; Lynch-SA; Lindsay-S; Custard-E; Belloni- E, Wilson-DI; Wadey-R; Goodman-F; Orstavik-KH; Monclair-T; Robson-S, Reardon-W; Bum-J; Scambler-P; Strachan-T. A homeobox gene, HLXB9, is the major locus for dominantly inherited sacral agenesis. Nat-Genet (1998) 20- 358-361 43 Chapter I

Rusell WL, Roscoe B. Splotch, a new mutation in the house mouse Mus Musculus. Genetics (1947) 32: 102

Sah-VP; Attardi-LD; Mulligan-GJ; Williams-BO; Bronson-RT; Jacks-T. A subset of p53-deficient embryos exhibit exencephaly. Nat-Genet (1995) 10: 175-80

Sanyanusm-P; Schimmenti-LA; McNoe-LA; Ward-TA; Pierpont-ME; Sulhvan-MJ; Dobyns-WB; Eccles-MR Mutation of the PAX2 gene in a family with optic nerve colobomas, renal anomalies and vesicoureteral reflux. Nat-Genet (1995) 9. 358-63

Schoenwolf-GC. Formation and patterning of the avian neuraxis: one dozen hypotheses. Oba-Found- Symp (1994) 181: 25-38

Schorah-CJ, Smithells-RW; Seller-MJ. Terathanasia, folic acid, and birth defects. Lancet (1997)350: 1323

Schorle-H; Meier-P; Buchert-M; Jaenisch-R; Mitchell-PJ Transcription factor AP-2 essential for cranial closure and craniofacial development. Nature (1996)381· 235-8

Schuier GD, Boguski MS, Stewart EA, Stein LD, Gyapay G, Rice K, White RE, Rodriguez-Tome P, Aggarwal A, Bajorck E, Bentohla S, Birrcn BB, Butler A, Castle AB, Chiannilkulchai N, Chu A, Clee C, Cowles S, Day PJ, Diblmg T, Drouot N, Dunham I, et al. A gene map of the human genome. Science (1996)274:540-546

Scott-JM; Weir-DG; Molloy-A; McPartlm-J; Daly-L; Kirke-P. Folic acid metabolism and mechanisms of neural tube defects. Ciba-Found-Symp (1994) 181: 180-7

Seller-MJ. Neural tube defects and sex ratios. Am-J-Med-Genet (1987)26: 699-707

Seller-MJ. Risks m spina bifida. Dev-Med-Child-Neurol (1994a)36: 1021-5

Seller-MJ. Vitamins, folic acid and the cause and prevention of neural tube defects. Ciba-Found-Symp (1994b) 181: 161-73

Seller MJ. Neural tube defects, chromosome abnormalities and multiple closure sites for the human neural tube. Clin Dysmorphol (1995a)4: 202-7

Seller-MJ. Further evidence for an intermittent pattern of neural tube closure in humans. J-Med-Genet (1995b) 32: 205-7

Seller-MJ. Sex, neural tube defects, and mulhsite closure of the human neural tube. Am-J-Med-Genet (1995c) 58: 332-6

Seller-MJ. Multi-site neural tube closure in humans and maternal folate supplementation. Am-J-Med- Genet (1995d) 58: 222-4

Smithells-RW; Sheppard-S; Schorah-CJ; Seller-MJ; Nevin-NC; Hams-R; Read-AP; Fielding-DW. Possible prevention of neural-tube defects by penconceptional vitamin supplementation. Lancet (1980) 1: 339-40

Sonano-P. The PDGF alpha receptor is required for neural crest cell development and for normal patterning of the somites. Development (1997)124: 2691-700 44 General introduction

Spielman-RS; McGinnis-RE; Ewens-WJ. Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM). Am-J-Hum-Genet (1993)52: 506-16

Steegers-Theunissen-RP; Boers-GH; Tnjbels-FJ; Finkclstem-JD; Blom-HJ; Thomas-CM; Borm-GF; Wouters-MG; Eskes-TK. Maternal hyperhomocysteinemia: a nsk factor for neural-tube defects? Metabolism (1994) 43: 1475-80

Stephenson-DA; Mercola-M; Anderson-Ε; Wang-CY; Stiles-CD; Bowen-Pope-DF; Chapman-VM. Platelet-den ved growth factor receptor alpha-subumt gene (Pdgfra) is deleted in the mouse Patch (Ph) mutation. Proc-Natl-Acad-Sci-USA (1991)88: 6-10

Strachan-T; Read-AP. PAX genes. Curr-Opin-Genet-Dev (1994) 4: 427-38

Stumpo-DJ; Bock-CB; Tuttle-JS; Blackshear-PJ. MARCKS deficiency in mice leads to abnormal brain development and perinatal death. Proc-Natl-Acad-Sci USA (1995)92: 944-48

Te-Meerman-GJ; Van-der-Meulen-MA; Sandkuijl-LA. Perspectives of identity by descent (IBD) mapping in founder populations. Clm-Exp-AUergy (1995)25 Suppl 2· 97-102

Tonello-HV; Warren-ST; Lmdstrom-JA. Possible X-hnked anencephaly and spina bifida—report of a kindred. Am-J-Med-Genet (1980) 6: 119-21

Tonello-HV. Report of a third kindred with X-hnked anencephaly/spina bifida. Am-J-Med-Genet (1984) 19:411-2 van-Aerts-LAGJM; Blom-HJ; de-Abreu-RA; Tnjbels-FJM; Eskes-TKAB; Copius-Peereboom- Steegeman-JHJ; Noordhoek-J. Prevention of neural tube defects by and toxicity of L-homocysteine in cultured postimplantation rat embryos. Teratology (1994)50: 348-360 van-Allen-MI; Kalousek-DK; Chemoff-GF; Junloff-D; Hams-M; McGillivray-BC; Yong-SL; Langlois- S; MacLeod-PM; Chitayat-D; et-al Evidence for multi-site closure of the neural tube in humans. Am-J- Med-Genet (1993) 47: 723-43 van-der-Put-NM; Steegers-Theunissen-RP; Frosst-P; Tnjbels-FJ, Eskes-TK; van-den-Heuvel-LP; Manman-ЕС; den-Heyer-M; Rozen-R; Blom-HJ. Mutated methylenetetrahydrofolate reductase as a risk factor for spina bifida. Lancet (1995) 346: 1070-1 van-der-Put-NM; van-den-Heuvel-LP; Steegers-Theunissen-RP; Tnjbels-FJ; Eskes-TK; Manman-EC; den-Heyer-M; Blom-HJ. Decreased methylene tctrahydrofolate reductase activity due to the 677C~>T mutation in families with spina bifida offspnng. J-Mol-Med (1996)74: 691-4 van-der-Put-NM; van-der-Molen-EF; Kluijtmans-LA; Heil-SG; Tnjbels-JM; Eskes-TK; Van- Oppenraaij-Emmerzaal-D; Bancrjce-R; Blom-HJ. Sequence analysis of the coding region of human methionine synthase: relevance to hyperhomocystemaemia m neural-tube defects and vascular disease. Q-J-Med(1997)90:511-7 van-Maldergem-L; Gillerot-Y; Kouhscher-L. Neural tube defects and omphalocele in tnsomy 18. Clin- Genet (1989) 35: 77-8 van-Straaten-HW; Blom-Η; Peeters-MC; Rousseau-AM; Cole-KJ; Seller-MJ. Dietary methionine does not reduce penetrance in curly tail mice but causes a phenotype-specific decrease in embryonic growth. J-Nutr (1995) 125: 2733-40 45 Chapter 1

Vortkamp-A; Gessler-M; Grzeschik-KH. GLI3 zinc-finger gene interrupted by translocations in Greig syndrome families. Nature (1991)352. 539-40

Wald-NJ Folic acid and neural tube defects: the current evidence and implications for prevention. Ciba- Found-Symp (1994) 181: 192-208

Wallin-J; Wilting-J; Koseki-H; Fntsch-R; Chnst-B; Balling-R. The role of Pax-1 in axial skeleton development. Development (1994) 120: 1109-21

Weeks-DF; Lathrop-GM. Polygenic disease: methods for mapping complex disease traits. Trends-Genet (1995)11:513-19

Whitehead-AS; Gallagher-P; Mills-JL; Kirke-PN; Burke-H; Molloy-AM; Weir-DG; Shields-DC; Scott- JM A genetic defect in 5,10 methylenetetrahydrofolate reductase in neural tube defects. Q-J-Med (1995) 88. 763-6

Wilcken-DE; Wang-XL. Relevance to spina bifida of mutated methylenetetrahydrofolate reductase. Lancet (1996) 347: 340

Wu-M; Chen-DF; Sasaoka-T; Tonegawa-S. Neural tube defects and abnormal brain development in F52- deficient mice. Proc-Natl-Acad-Sci-USA (1996)93: 2110-5

Yates-JR; Ferguson-Smith-MA; Shenkin-A; Guzman-Rodnguez-R; White-M; Clark-BJ. Is disordered folate metabolism the basis for the genetic predisposition to neural tube defects? Chn-Genet (1987)31: 279-87

Zhang-J; Hagopian-Donaldson-S; Serbedzija-G; Elsemore-J; Plehn-Dujowich-D; McMahon-AP; Flavell- RA; Wilhams-T. Neural tube, skeletal and body wall defects in mice lacking transcription factor AP-2. Nature (1996) 381: 238^1

Zhao-Q; Behnnger-RR; de-Crombrugghe-B. Prenatal folic acid treatment suppresses acrania and meroanencephaly m mice mutant for the Cartl homeobox gene. Nat-Genet (1996) 13: 275-83

Zimmerman-AW. Hyperzincemia in anencephaly and spina bifida: a clue to the pathogenesis of neural tube defects? Neurology (1984) 34. 443-50

46 Chapter 2

Exclusion Mapping of the Gene for X-linked Neural Tube Defects in an Icelandic Family

Hoi FA, Geurds MP A, Jensson O, Hamel BCJ, Moore GE, Newton R, Manman ECM.

Hum Genet 93: 452-456 (1994)

47

Exclusion mapping of the gene for X-linked neural tube defects in an Icelandic family

F. A. Hoi1, M. P. A. Geurds1, O. Jensson2, B. C. J. Hamel', G. E. Moore1, R. Newton1, E. C. M. Ma riman1 1 Department of Human Genetics University Hospital Nijmegen PO Box 9101 6500 HB Nijmegen The Netherlands 2The Blood Bank Department of Medical Genetics University Hospital Reykjavik PO Box 1408 IS 121 Reykjavik Iceland 3 Action Research Laboratory for the Molecular Biologv of Fetal Development RPMS Institute of Obstetncs and Gynaecology Queen Charlotte s and Chelsea Hospital Goldhawk Road London W6 OXG LK

Received 20 July 1993 / Revised 17 September 1993

Abstract. Various polymorphic markers with a random of the underlying factors has been hampered by the fact distribution along the X chromosome were used in a link­ that most of these families do not display a Mendelian age analysis performed on a family with apparently X pattern of inheritance (Chafkupt et al 1992 Manman and linked recessive inheritance of neural tube detects (NTD) Hamel 1992) This has precluded the use of conventional The lod score values were used to generate an exclusion genetic methods that are routinely employed for the eluci map of the X chromosome, this showed that the responsi­ dation of monogenic disorders Fortunately a few fami ble gene was probably not located in the middle part of hes have been reported that show a Mendelian, usually X Xp or in the distal region of Xq A further refining of linked, pattern of inheritance of NTD (Tonello et al 1980, these results was achieved by tuplotype analysis, which Fellousetal 1982 Fincmanctal 1982 Oman Ganes and indicated that the gene for X linked NTD was located ei Shokcir 1984, Baraitser and Burn 1984, Tonello 1984, ther within Xp21 1-pter, distal from the DMD locus, or in Jensson et al 1988) the region Xql2-q24 between DXS106 and DXS424 Here we report the analysis of an Icelandic family Multipoint linkage analysis revealed that the likelihood with apparently X linked recessive inheritance of NTD for gene location is highest for the region on Xp The re (Jensson et al 1988) By employing exclusion mapping gion Xq26-q28, which has syntenic homology with the haplotype analysis and multipoint linkage analysis, we segment of the murine X chromosome carrying the locus have been able to exclude large regions of the X chromo for 'bent tail (Bn), a mouse model for X-linked NTD is some as the location for the responsible gene excluded as the location for the gene underlying X linked NTD in the present family Thus, the human homologue of the Bn gene and the present defective gene arc not Materials and methods identical, suggesting that more than one gene on the X chromosome plays a role in the development of the neural Description of the family tube The present family (Fig 1 ) has prev lously been described by Jens son et al (1988) I he following types of NTD were observed among the affected members II 1 a cystic tumour at ihe lumbar region and club foot III 5 lumbosacral meningomyelocele IV 1 Introduction and IV 2 anencephaly IV 3 occipital encephalocele and ihoraco lumbar meningomyelocele At the age of 28 individual II 6 had a neurosarcoma excised from the upper part of the gluteal sulcus With a birth prevalence of about 1/1000, neural lube de about 2 cm on the left side Furthermore he had a dark brown fects (NTD) are among the most prevalent congenital hairy nevus of about 2 cm in diameter in Ihe uppermost part of the malformations m Western Europe Despite this high fre­ lumbar region slightly left of the midline of the spine this has also quency, essential information on their etiology is lacking been surgically removed No closure defects were seen upon X ray In general, it is accepted that NTD are attributable to an analysis of Ihe lower vertebral column One of his three daughters developed hydrocephaly because of veninculo cisternal sienosis incomplete closure of the neural tube at the fourth week of Blood was sampled from relevant members of this family and gestation, this results from a combination of environmen DNA was isolated by the procedure of Kunkel et al (1985) tal factors and a predisposing genetic background The in­ fluence of genetic factors is indicated by the familial clus lenng of NTD tn about 5% of cases, but the identification Marker anahsis In the present study 26 highly polvmorphic repeat markers have Correspondence to E С M Manman been analysed of which 18 appeared to be informative [Table 1

49 the 5 DYS I polymorphism reponed by leerer et al (1991) was Linkage analysis and exclusion mapping used for the analysis ot the DMD locus] The analysis involved polymerase chain reaction (PCR) amplification of genomic DNA Linkage analyses were performed on the basis of X-lmked reces­ with the appropriale oligonucleotide primers (see Table 1) under sive inheritance with reduced penetrance for males and a disease the conditions described by Weber and May (1989) VP dCTPwas gene frequency of 0 0001 The mutation rate was estimated to be included in the reaction mixture in order to label the amplified of the same magnitude Pairwise lod scores were calculated for fragments After PCR the amplification products were separated every marker using the MLINK option of the program LINKAGE, by electrophoresis on standard 6 Ш sequencing gels Subsequent­ version 5 04, (Lathrop et al 1985) and the values obtained tor dif­ ly, the gels were dried and exposed overnight to Kodak X ray film ferent recombination fractions ( = 0 05 0 10 0 20 and 0 10) were for visualization of the allelic bands For Southern analysis of lo­ used as input data for the program EXCLUDE (Edwards 1987) cus DXSI06 K^g genomic DNA was digested with fl^Ill accord­ The processing of these data resulted in a graphical image of the ing to the manufacturer s instructions (Gibco BRL) Digested DNA probability of gene location along the X chromosome Multipoint was then subjected to electrophoresis on Ι9Ό agarose gels in TAE· linkage analysis was performed using the L1NKMAP option of the (40mMTRIS HAc I mM LDTA pH7 5)and thereafter the DNA program LINKAGE version 5 10 Data from consecutive five- was blotted onto a nylon membrane (GeneScreen Plus) in 0 4 N point linkage analyses were combined to obtain the results as given NaOH To visualize the allelic bands blots were hybndi7cd over­ in Tig 4 Information concerning the relative order of marker loci night with the appropriate probe (cpX203, Kidd et al 1989) and, and the genetic distances between them was obtained from the after being washed lor 2 χ 15 min in 40mM phosphate buffer 1% Genome DataBase and from the literature (Lafreniere et al 1991, SDS at 6TC, the blot was exposed overnight to Kodak X ray film Huang el al 1992. Murray et al 1992, Schlessinger et al 1993)

&T0 llro ¿Ι ό ù ό ή см5Ъ lia 2 3 4 5 6 7a78

Fig. 1. Pedigree of the Icelandic family with ap­ ±^> parently X linked NTD This family has been described by Jensson et al (1988) The follow­ ing types of NTD were present spina bifida IV • (II 1, III 5 and IV 3). anencephaly (IV 1 and 1 2 IV 2), encephalocelc (IV 3)

Table 1. Results from pairwise linkage analysis

Locus Location Lod score (θ) References

000 0 05 0 10 0 20 0 30

DXS443 p22 1 0 68 060 0 52 0 36 0 19 Browne et al 1992a DXS451 p22 3-p21 2 0 75 0 70 064 0 49 0 34 Browne et al 1992a DMD p21 3-p2l 1 » -0 28 -0 02 0 17 0 20 Feeneretal 1991 DXS538 pll 2I-p2l 1 ^» -0 28 -0 02 0 17 0 20 Browne et al 1991a MAOA pll 4—pi 1 3 -~ -0 83 -0 53 -0 24 -0 10 Black et al 1991 MAOB pll 4-pll 3 0 54 0 50 046 0 38 0 26 Konradietal 1992 PGKIP qll 2-ql2 0 54 0 50 0 46 0 38 0 26 Browne et al 1992b DXS106 qll 2-ql2 ^»37 -0 84 -0 53 -0 24 -ΟΙΟ Kiddetal 1989 DXYS1 q21 31 0 15 0 14 0 13 009 004 Browne et al 1991Ь DXS3 q21 3 0 54 0 50 0 46 0 38 0 26 Stanieretal 1991 DXS454 q2l l-q23 0 75 0 70 064 0 49 034 Weber et al 1990 DXS178 q21 33-q22 0 15 0 14 0 13 009 004 Allen and Belmont 1992 DXS456 q2l-q22 -0 37 -031 -0 25 -0 14 -0 06 Lutyetal 1990 COL4A5 q22-q23 0 49 0 45 042 034 0 25 Barker et al 1992 DXS424 q24-q25 ^= -0 07 0 14 0 26 0 24 Lutyetal 1990 DXS425 q26-q27 1 ^= -I 44 -0 89 -0 39 -0 15 Lutyetal 1990 HPRT q26 1 ^, -0 77 -0 49 -0 22 -0 09 He-ame and Todd 1991 FRAXA q27 3 -^х. -0 83 -0 53 -0 24 -0 10 Richards et al 1991 F8C q28 -3 55 163 -1 07 -0 55 -0 28 Lallozetal 1991

50 Pter

Fig. 2. Exclusion mapping of the gene for X linked NTD DNA was isolated from all persons of the family shown in Fig I and analysed with various poly morphic markers To pinpoint the posi Hon of the individual markers on the X Centromere -н chromosome we have used a diagram CGK1P1 OXS106 matic representation of a GTG banded chromosome as a frame The middle of the chromosomal band(s) to which a marker has been assigned (Table 1) was taken as its location For some markers mapping at the same interval the relative order was obtained from Schlessinger et al ( 1994) Lod scores were calculated and processed by the program EX CLUDb (b-dwards 1987) Піе areas ил HPRT - dirthc resulting curve represent the rela FRAXA- live probabilities of gene location at dif FBC -. ferent regions along the X chromosome Qter RELATIVF PROBABILITY OF GENE LOCATION

I ι I , ι I •ι ó Ù α I • I л І 1 1 2 Il 1 7а II 7 и э ι β Il d DXS443 "D BD" D DB в В D DDl1 DDlI D DXS451 С ED D CE D L E CDl EDI D DWD A CD D AD С D В AC ВС С DXS538 С BA A CA В А В CB BB в M АО A С CD D CD D С CC CC С с Flg. 3. Haplotype analysis in (he family with X ΝΛΑΟΒ В AA A BA А А D BA DA А linked NTD Haplotypes were constructed trom PGK1P1 В AA A BA А А Α BA AA А the alleles of informative markers distributed DXS106 A AB В AA А В A AA AA А along the entire X chromosome (Thompson A BA A А A AA AA А DXYS1 AB В 1987) The number undercut) person corre­ DXS3 A BB В В В AB BB AB В в sponds to her/his position in the pedigree of DXS454 A AB В AA А В С AB CB В Fig 1 Haplot) pes of deceased persons (I 1 12 DXS178 В AB В ЗА А В С BB CB в and II 1 ) were deduced from the haplotypes of DXS456 В AB В BA 3 В A BB AB в COL4A5 В AA А BA А А С BA CA А their offspring It should be noted that by this DXS424 D AC А DA С С В DC BC С procedure the deduced homo?ygo/itv for indi DXS4¿í В ED E BE - D Ρ 3D f-B D vidual I 2 contains an extra degree of uncer HPRT С AA А CA А А А СЛ AC А tainty Chromosomal segments that co segre FRAXA Α BA В AB А В A AA AA А gate with the disorder arc indicated by & filled F8C С EA E CA А А E CA С bar

Results middle region of Xp or at the distal pari of Xq Although Fig 2 provides a fair indication of gene location, EX Exclusion mapping of the gene for X linked NTD CI UDE is not an adequate method to determine the most likely location of the responsible gene on the X chromo Linkage analysis was performed in a family with appar­ some Therefore, other genetic methods have been em­ ently X-hnked NTD (Fig 1) by using highly polymorphic ployed markers, randomly distributed along the X chromosome (Table 1 ) In order to perform lod score calculations, the Haplot\pe and multipoint linkage analysis penetrance of the disorder was estimated from the ex­ tended pedigree (Jensson et al 1988) and was set at 70% In order to define the X chromosomal regions in which Lod scores were determined for various values of Θ the gene for X-hnked NTD could be located, haplotypes (Table 1) These data were further processed by the pro­ were constructed according to the principles outlined by gram EXCLUDE (Edwards 1987) The graphical output Thompson (1987), for the marker loci that were used for of this evaluation is shown in Fig 2 Our results indicate exclusion mapping (Table I ) In this way two segments of that the responsible gene is probably not located at the the X chromosome could be defined that co-segregate

51 SJ3B S454 COL4A5 S ?B / 424 S451 OMD I »G« Ρ' S Λ ! i Vi I г^ ^ нрат ^ f

Fig. 4. MuHipoint linkage analysis with markers along the entire X chromosome Por Ihe relative location of [he various marker loti, the linkage maps of Donnely el al (1993) and bchlessinger et al (1993) were used as a frame supplemented with the mapping data of Murray et al (1992) The total length ot the X chromosome is esti­ mated at 220cM

with NTD in the present family Xp21 1-pter, distal to the cause of the probability of a reduced penetrance for the DMD locus, and the region Xql2-q24 between DXS106 disorder Interestingly, a girl has recently been reported and DXS424 (big 3) Attempts to narrow these regions by with anencephaly and a terminal deletion of Xp in one of the analysis of additional repeat markers were unsuccess­ her X chromosomes [46, X, del(X) (p22 1), Plaja et al ful because of non intormativity Multipoint linkage 1993), thereby providing further evidence for the presence analysis was perfomed lo obtain an indication of the rela­ of a gene for neural tube formation in the distal part of tive importance of gene location in different areas of the Xp Although, in the present family, obligate female earn­ X chromosome As can be seen in Fig 4, the highest loca­ ers of the genetic defect show no sign of NTD, it is tempt­ tion scores were obtained for the two regions previously ing to speculate that a mutation in the same gene might act identified by haplotype analysis maximal location scores as the predisposing factor By focussing on the proximal of 5 8 for the Xp-region and of 3 5 for the Xq region The region of Xq, deletions in this area have been found in pa­ relative odds for location of the gene in the Xp-region, the tients with various disorders other than NTD (Cremers et Xq-region and the next most likely arca (MAOA/B al 1990) Indeed, the band Xq21 can be entirely lost in -DXS106, maximal location score -2 9) are 80 25 1 males, without giving rise to NTD Although these obser­ showing that, apart from the co-segregating segments, the vations suggest that the gene for X-linkcd NTD is not lo­ remainder of the X chromosome seems to be excluded cated in Xq21, the possibility remains that even the com­ plete absence of the respective gene in males would not necessarily give rise to a disease phenotype In order to Discussion provide more substantial data on gene location, we have recently begun to sample additional members of the pedi­ We have been able to define by exclusion mapping and gree In this way, we hope to raise the informativity of the haplotype/multipoint linkage analysis, two segments of family, in order to allow a more accurate localization of the X chromosome, Xp21 1-pter and Xql2-q24, which the responsible gene by extended linkage analysis co-segregate with NTD in an Icelandic family (Jensson et An interesting result of our present study is the exclu­ al 1988) Although rare events of double recombination sion of the region Xq25-q28 Part of the excluded chro­ cannot be excluded, our data indicate that the gene for X- mosomal segment, ι e the region overlapping the genes linked NTD is located in one of these regions Multipoint for HPRT, L1CAM and f-8C, has syntenic homology with analysis indicates that the distal region of Xp is three the region of the murine X chromosome (Lyon and Kirby times more likely to harbour the gene than the Xq-seg- 1992) encompassing the locus for "bent tail" (Bn), a ment Information concerning the clinical status of male mouse model for X-linked NTD (Garber 1952) Our re­ II 6 may allow us to distinguish further between these two sults therefore indicate that the human homologue of the locations This individual shares the haplotype of the sus­ murine Bn gene is different from the gene underlying pected region on Xq with his affected brother II I, but has NTD in the present family It suggests that more than one inherited a different chromosomal segment for the distal gene on the X chromosome plays a role in the formation part of Xp Clinical examination of this person, including of Ihe neural tube investigation of the lumbar vertebral column by X-ray, re­ Acknowledgement* We thank S D van der Velde Visser and vealed no sign of NTD, a result that suggests that the gene E M A Boender-van Rossum for establishing cell lines and for is more likely to be located on Xp than on Xq However, cell culture This study was financially supported by the Dutch the latter region cannot be firmly excluded in this way be­ Pnnses Bealnx Fonds grant no 90-3154

52 References committee and catalogs of cloned and mapped genes and DNA polymorphisms Cytogenet Cell Genet 51 622-947 Allen RC, Belmont JW (1992) Dinucleotide repeal polymorphism Konradi C, Олеішч L, Breakefield XO (1992) Highly polymorphic al the locus DXS178 Hum Mol Genet 1216 (GT)n repeat sequence in intron II of the human MAOB gene BaraiLser M, Burn J (1984) Neural tube defects as an X-linked con­ Genomics 12 176-177 dition Am J Med Genet 17 181 385 Kunkel LM. Monaco ΑΡ. Middlcsworth W. Ochs HD, Lati SA Barker DF, Cleverly J, Fain PR (1992) Two СЛ-dinucleotide poly­ (1985) Specific cloning of DNA fragments absent from the morphisms at the COL4A5 (Alport syndrome) gene in Xq22 DNA of a patient with an X chromosome deletion Proc Natl Nucleic Acids Res 20 929 Acad Sci USA 82 4778^1782 Black GCM, Chen ZY. Craig IW, Powell JF (1991) Dinucleotide Lafrenière RG, Brown CJ. Powers VE, Carrel L, Davies KE. repeal polymorphism at the MAOA locus Nucleic Acids Res Barker DF, Willard HF (1991) Physical mapping of 60 DNA 19 689 markers in the p21 1—»q21 3 region of the human X chromo­ Browne DL. Luty JA, Lilt M (1991a) Dinucleotide repeat poly­ some Genomics 11 352-363 morphism at the DXS538 locus Nucleic Acids Res 19 1161 Lallo7 MRA, McVey JH. Pattinson JK, Tuddenham EGD (1991) Browne DL, Zonana J. Litt M (1991b) Dinucleotide repeat poly Haemophilia A diagnosis by analysis of a h> perv anablc dinu­ morphism at the DXYS1X locus Nucleic Acids Res 19 1721 cleotide repeat within the factor VIII gene Lancet 338 207-211 Browne D, Barker D, Litt M (1992a) Dinucleotide repeat poly­ Lathrop GM, Lalouel JM. Julier С Ott J (1985) Multilocus linkage morphisms at the DXS365, DXS443 and DXS451 loci Hum analysis in humans, detection of linkage and estimation of re­ MolGenel I 211 combination Am J Hum Genet 37 482 498 Browne DL, Zonana J. Litt M (1992b) Dinucleotide repeal poly­ Luty JA, Guo Z, Willard HF Lcdbetter DH, Ledbetler S, Lilt M morphism at the PGK1 PI locus Nucleic Acids Res 20 1169 (1990) Five polymorphic microsalellitc VNTRs on the human Chatkupl S, Lucek PR, Koenigsberger MR. Johnson WG (1992) X chromosome Am J Hum Genet 46 776-783 Parental sex effect in spina bifida a role for genomic imprint­ Lyon MF, Kirby MC (1992) Mouse chromosome alias Mouse ing' Am J Med Genet 44 508-512 Genome 90 22-44 Cremers FPM, Sankila EM, Brunsman F, Jay M, Jay В, Wright A Mariman ECM, Hamel BCJ (1992) Sex ratios of affected and Pmckers AJLG, Schwartz M Pol DJR van de Wieringa B, transmitting members of multiple case families with neural Chapelle A de la. Pawlowitzki IH, Ropers HH (1990) Dele­ tube delects J Med Genet 29 695-698 tions in patients with classical choroideremia vary in size trom Murray JC, Ludwigsen SJ, Buetow KH (1992) A comprehensive 45 kb to several megabases Am J Hum Genet 47 622-628 genetic linkage map of the human genome (X chromosome) Donncly A, Kozman H, Gedeon A, Webb S, Lynch M. Richards Science 258 67-86 R, Mulley J (1993) A genetic map of microsalellne markers on Oman-Ganes L, Shokeir M (1984) Neural tube defects possible the X chromosome (abstract) 6th International Workshop on X-hnkcd recessive inheritance Am J Hum Genel 36 67(S) the Fragile X and X linked Mental Retardation, ρ 10 Plaja Л, Vendrell T, Sarret E, Torran N Mediano С (1993) Termi­ Edwards JH (1987) Exclusion mapping J Med Genet 24 539-543 nal deletion of Xp and anencephaly (abstract 327) 25lh Annual FeenerCA, Boyce FM, Kunkel LM (1991) Rapid detection ol CA Meeting of the European Society ol Human Genetics. Barce­ polymorphisms in cloned DNA application to the 5' region of lona, Spain the dystrophin gene Am J Hum Genet 48 621-627 Richards RI, Holman К, Kozman H, Krcmer E, I ynch M, Fellous M, Boue J, Malbrunot CWollman E, Sasportes M, Van Pritchard M. Yu S, Mulley J. Sutherland GR (1991) Fragile X Cong N, Marcelli A, Rebourcel R, Hubert C, Demenais F. Li­ syndrome genetic localisation by linkage mapping of two mi- ston RC, Namboodin KK, Kaplan EB, Fellous M (1982) A crosalellile repeats FRAXAC1 and FRAXAC2 which immedi­ five-generation family with sacral agenesis and spina bifida ately flank the fragile sue J Med Genet 28 818-823 possible similarities with the mouse T-locus Am J Med Genet Schlessinger D, Mandel JL. Monaco AP, Nelson DL, Willard HF 12 465^87 (1993) Report ot the Fourth International Workshop on human Гіпетап RM, Jorde LB, Manin RA, Hassledl SJ, Wing SD, X chromosome mapping 1993 Cytogenet Cell Genet (in press) Walker ML (1982) Spinal dysraphia as an autosomal dominant Stanier P, Newton R, Forbes SA, Ivens A. Moore Gfc ( 1991 ) Poly­ defect in four families Am J Med Genet 12 457-464 morphic dinucleotide repeal at the DXS3 locus Nucleic Acids Garber ED (1952) "Bent-tail", a dominant sex-linked mutation in Res 19 4793 the mouse Proc Natl Acad Sci USA 38 876-879 Thompson EA (1987) Crossover counts and likelihood in multi­ Hearne CM, Todd JA (1991) Tetranucleotide repeat polymor­ point linkage analysis IMA J Math Appi Med Biol 4 93-108 phism at the HPRT locus Nucleic Acids Res 19 5450 Tonello H (1984) Report ol a third kindred with X-linked anen- Huang THM, Cottingham RW, Lcdbetter DH. /oghbi H Y (1992) ccphaly/spina bifida Am J Med Genet 19 411-412 Genetic mapping of four dinucleotide repeal loci, DXS453. Tonello H, Warren S. Lindstrom J (1980) Possible X-linked anen­ DXS458, DXS454, and DXS424, on the X-chromosomc using cephaly and spina bifida - report of a kindred Am J Med Ge­ multiplex polymerase chain reaction Genomics 13 375-180 net 6 119-121 Jensson O, Arnason A. Gunnarsdomr H, Pctursdotlir I. Fossdal R, Weber JL, May PE (1989) Abundant class of human DNA poly Hrcidarsson S (1988) A family showing apparent X-linked in­ morphisms which can be typed using the polymerase chain re­ heritance of both anenccphaly and spina bifida J Med Genet action Am J Hum Genet 44 388-396 25 227-229 Weber JL, Kwitck AE. May Pb, Polymeropoulos MH, Lcdbetter S Kidd KK, Bowcock AM, Schmidtke J, Track R, Riccuiti F, Hulch- (1990) Dinucleotide repeal polymorphisms al the DXS354, ings G, Bale A, Pearson P, Willard HF (1989) Report of the DXS454 and DXS458 loci Nucleic Acids Res 18 4037

53

Chapter 3

Absence of Linkage between Familial Neural Tube Defects and the PAX3 Gene

Chatkupt S, Hoi FA, Shugart YY, Geurds MP A, Stenroos ES, Koenigsberger MR, Hamel BCJ, Johnson WG, Manman ECM.

J Med Genet 32: 200-204 (1995)

55

Absence of linkage between familial neural tube defects and PAX3 gene

S Chatkupt, F A Hoi, Y Y Shugan, Μ Ρ A Geurds, E S Stenroos, M R Koenigsberger, В С J Hamel, W G Johnson, E С M Manman

Abstract been recognised as a model for human NTD Neural tube defects (NTD) are among the Splotch homozygotes develop spina binda, most common and disabling birth defects. meningocele, and exencephaly ' ' Most mutants The aetiology of NTD is unknown and their die in utero Splotch heterozygotes have pig­ genetics are complex. The majority of mentation defects resulting in white feet, tail NTD cases are sporadic, isolated, non- tip, and belly patch These pigmentation de­ syndromic, and generally considered to be fects as well as deficiencies in neural crest multifactorial in origin. Recently, PAX3 derived tissues and cells (NCC), that is, spinal (formerly HuP2, the human homologue of ganglia and Schwann cells, arc caused by the mouse Pax-J), on chromosome 2q3S-37, failure of NCC to populate these regions suf­ was suggested as a candidate gene for NTD ficiently during development n" Mutations in because mutations of Pax-3 cause the the Pax-Ì gene result in the Sp phenotype " mouse mutant Splotch (Sp), an animal The paired box containing genes, the Pax model for human NTD. Mutations in PAX3 genes, encode for sequence specific DNA bind­ were also identified in patients with Waar­ ing transcription factors that play a role in denburg syndrome type 1 (WS1). At least embryonic development To date, nine pax eight patients with both WS1 and NTD genes, Paxl-9, have been isolated ",6 Pax-3 is have been described suggesting pleiotropy expressed in the neural tube, in the NCC, in or a contiguous gene syndrome. the dermomyotome of the developing somites, in limb buds, and in the developing brain '4 " Seventeen US families and 14 Dutch The Pax-ì gene, located on mouse chro­ families with more than one affected per­ mosome 1, is homologous to the human PAX3 son with NTD were collected and 194 or formerly HuP2 gene at 2q35-37 ' ' Mutations people (SO affected) from both data sets in PAX3 have been described in patients with were genotyped using the PAX3 poly­ Waardenburg syndrome type 1 (WS1), a syn­ morphic marker. The data were analysed drome consisting of pigmentary disturbances Departments of using affecteds only linkage analysis. The Ncurosciences and resulting from abnormalities related to NCC lod scores were -7-30 (US), -3-74 Pediatrics, UMDNJ- emigration, a pathogenesis similar to that of Sp New Jersey Medical (Dutch), and -11 -04 (combined) at 0 = 0-0, mice '" '" Reports of at least eight patients with School, MSB-H-506, under the assumption of the autosomal 185 South Orange both WS1 and NTD raise the possibility of dominant model. For the recessive model, Avenue, Newark, pleiotropy or a contiguous gene syndrome 2' " the lod scores were -3·30 (US), 1-46 New Jersey 07103, To test this hypothesis, we conducted linkage USA (Dutch), and 4-76 (combined) at θ = 0·0. S Chatkupt analysis on 31 NTD families using the PAX3 M R Kocnigsbergcr Linkage between PAX3 and familial polymorphic marker NTD was excluded to 9-9 cM on either side Department of Human of the gene for the dominant model and to Genetics, University Hospital Nijmegen, 3-63 cM on either side of the gene for the Nijmegen, recessive model in the families studied. Patients and methods The Netherlands No evidence of heterogeneity was detected PATIENTS, FAMILY, AND CLINICAL EVALUATION F A Hoi using the HOMOG program. Our data Μ Ρ A Geurds Families from the United States were as­ 8 Г J Hamel indicate that PAX3 is not a major gene for certained by referrals from spina bifida clinics, Ь С M Manman NTD. and by responses of patients to notices in Department of patient newsletters Our specific request was Genetics and (J Med Genet 1995,32 200-204) for families with more than one case of spina Development, bifida cystica (SB) or other NTD Syndromic or Columbia University, chromosome abnormality cases were excluded New York, USA Y Y Shugart Neural tube defects (NTD) are among the Diagnoses were based on detailed clinical in­ most common birth defects and have been formation from interviews by us, from direct Department of review of the medical records (31 records), or Neurology, UMDNJ- associated with certain syndromes and chro­ Robert Wood Johnson mosome abnormalities (including trisomies 13, from medical record review by physicians and Medical School, 18, and 21)' and an X linked condition - ' The nurses in the referring SB clinics Information New Brunswick, majority of NTD cases are sporadic, isolated, obtained for index and other cases included New Jersey, USA family pedigree, number of affected cases in Г S Stenroos non-syndromic, and generally considered to be W G Johnson multifactorial in origin with a hentabiliry each family, sex, ethnic background, and birth dates of the cases, their mothers, and their of about 70 to 80% " However, a number Correspondence to fathers Dr Chatkupt of familial cases have been documented im­ Received 17 July 1994 plicating genetic susceptibility factors in fa­ For the Dutch families, entena for selection Revised version accepted 7 2 for publication milial NTD ' of cases and information obtained were similar 9 November 1994 The mouse mutant Splotch (Sp) has long These families were selected in collaboration 57 cFSl· ff* I* iFll· SB SB SB SB ^η-jZf JZ^-j-jZf 0-y&

fra dv¿^, 1· DY

Cf-HZÍ ^-pjzf ir i· да foi ^тіъ cn cSoB SB ¿È %в AC АС •f «== AC

Cf-f-D' ,- >гі* 1* D1* D-j-¿0 SBO SB "Τ ircri ώ irò àvcf rf^k (Tal· Obro Λ AC AC SB I—L7V SB

¿¿-г* D-pÓ-i-r-D CfyCf^í ¿f¿f

SB f¿l· £ra~ïïirai f

l*-rizf ^-г^

СР-рСТ CP-pCT αγο· D-r-Ο ¿γα*

SB SB SBO SB SB SrtB SB 1SB [filSB S·B

Pedigrees of aü informative NTD families Shaded 0-^-0 squares and areles represent ¿-ri* affected subjects A dût indicates a person who was gi-j-ér J) ¿P-j-çf genotyped AC= anencephalyt ИС= encephalocele, SB = spina tf—tfo. fib f4 obtained and their transformed were calculated from disease incidence and cell lines we e established DNA was prepared penetrances from each transformed cell line by standard Linkage analysis was performed using the methods '4 The short tandem repeat poly­ LINKAGE computer package (version 5 l)2" morphism (STRP) located on the 5' side of Because of the unclear mode of inheritance of exon 1 of the PAX3 gene was used 2! We de­ SB and NTD, we performed analyses under the signed a new set of primers flanking the same assumption of dominant and recessive models repeat in order to reduce the size of the PCR Under the dominant assumption, the disease products This made it easier to separate the allele frequencies were estimated to be 0 00025 diflerent alleles on acrylamide gel Forward and 0 00088 for the US and the Dutch data pnmer 5'-AGTTGCTGAGGGCGGAGAAG- respectively, and 0 022 and 0 042 for the US 3' and reverse pnmer 5'-GAAATCACAAGA- and the Dutch respectively under the recessive GGATAGAGGCT-3' Product sizes were 192 model The affecteds only method was used in to 218 bp DNA was amplified by PCR using the linkage analysis by making the phenotype published conditions25 in a 25 μΐ reaction of unaffected family members unknown In order to assess the possible linkage of NTD to the PAX3 gene, the marker PAX3 was selected for genotyping Allele frequencies for Table 1 Lod scores under dominant model, penetrance -0 45 this marker used in the US data set were as 5 1) published' (using control US population) For the Dutch data set, allele frequencies were 00 0 01 0 05 0 1 02 03 04 calculated from 37 unrelated people (Dutch) IS 7 3 -5 14 2 94 1 6B -0 59 -0 20 0 06 selected from the same geographical areas as Duich 3 74 2 10 -0 58 0 04 0 24 0 20 0 10 Dutch NTD families The allele frequencies -11 04 7 24 -3 52 0 00 Sum -1 72 0 35 0 04 from the two populations were not significantly different

Table 2 Lod scores under recessive model, penetrance = 0 45 0 Results Two point lod scores were calculated for vari­ oo 0 01 0 05 0 1 02 03 04 ous recombination fractions assuming Qm — % LS -3 30 -257 -1 16 0 45 0 02 0 08 0 04 penetrance of 0 45 and autosomal dominant Dutch -1 46 1 09 0 31 0 09 0 28 021 0 09 Sum 4 76 3 66 - 1 47 0 36 0 30 0 29 0 13 (table 1) or autosomal recessive models (table 2) Results for affecteds only analysis showed exclusion of linkage between familial N"I D and the PAX3 gene to 9 9 cM on either side of the Table 3 Lod scores under dominant model, penetrance = 0 27 gene for the dominant model and to 3 63 cM on either side of the gene for the recessive 0 model in the present families Two point lod 00 0 01 0 05 0 1 02 03 04 scores were also· calculated assuming pene­ trance of 0 27 (tables 3 and 4) The two point US 6 21 4 57 2 68 -1 57 0 56 0 19 -0 05 Dutch -301 -1 88 0 54 0 02 0 24 0 20 0 09 lod score assuming the phenocopy rate of zero Sum -9 22 6 45 3 22 1 59 0 32 001 0 04 showed similar results (data not shown)

59 Table 4 Lod scares under recessive modtï, pen ctrance=--0 2? reasons Therefore the analysis was performed il including SBO However, because there were only four SBO in the daia set, excluding them 00 001 0 05 0 1 02 03 0 I did not have significant influence on lod scores LS 3 15 2 49 1 16 0 47 0 00 0 07 0 03 There has been concern regarding genetic Dunn 1 II 0 84 0 22 0 11 0 27 0 20 0 08 Sum 126 3 33 1 38 0 36 0 27 0 27 0 11 heterogeneity based on the level of the defects (high level SB above Til ι low level SB) perhaps representing defects in neurulation ver­ sus canalisation" However, this question re­ No evidence for genetic heterogeneity was mains controversial " It w as also suggested that 15 detected using the HOMOG program " (data AC may be distinct from SB although both not shown) may occur in the same family and the oc­ currence rate of each type is increased in sibs of children affected by the other4 In this re­ Discussion spect, the same major gene is likely to contribute The application of linkage analysis to disease to different phenotypes Nevertheless, if in fact which display complex traits carries several different phenotypes such as SB and AC are difficulties because of (1) the uncertainty of caused by different loci, combining data may mode of inheritance and penetrance, (2) the result in a false negative lod score This is not small family sizes, (3) the unclear phenotypes, the case in our senes All affected subjects in and (4) genetic heterogeneity '°J' However, the the US families had SB and SBO whereas the complex traits are among the most common affected subjects in the Dutch data had vaneties human disorders and efforts should be made of NTD, SB, AC, and FC However, only the to unra\ el such problems Linkage analysis has patients with SB, SBO, and EC from the Dutch been used successfully to locate predisposing families were genotyped Linkage was not de­ genes for complex traits such as familial Alz­ tected in the US families with only SB, a result heimer's disease " similar to that from the Dutch subset Although NTDs have been considered to Such "exclusion' results must be interpreted have multifactorial threshold inheritance with cautiously, especially since the phenocopv rate the phenotype depending upon an interaction is only approximate However, our data exclude between genetic and environmental factors, PAX3 as a major gene in the present families monogenic inheritance with a major con­ with familial NTD and familial SB under the tribution of environmental factors has been assumption of the above parameters suggested '4 Another study" in which SBO was included with SB and sacral agenesis supported The cooperation of the families of the staff of spina binda clinics in the Lnucd States and of the Duich Patient Or autosomal dominant inheritance with se­ ganisation BOSK ITie Metherlands is gratefully acknowledged gregation distortion similar in type to that seen The support of NIH grants 2 S07 KR05393 (SC) R29 iò NS29893 (SC WGD and HGO(100HtY>. S) the foundation with alleles at the Τ locus in the mouse of I V1DNJ SC) the March of Dimes Birth Delects Foundation However, it is possible that a gene for NTD #8 1Y91 0939 (SC) and the Prinses Beatrix Tonds И 005 segregates in an autosomal recessive manner '" (LOMM) arc gratefully acknowledged \Xe also thank Prol 13r H H Ropers for helpful discussions and Dr John Ht ran and In addition, autosomal recessive inheritance Anindita Sarangi for technical assistance has been suggested for some families with anencephaly '"" If there is a major gene se­ 1 Khoury MJ Fnckson JD James LM Fnological hetero gregating in NTD families either in an auto­ gcneity of neural tube defects clues from epidemiology Am 7 bptdcim Ι 19Θ2 115 53Й 4B somal dominant or autosomal recessive man­ 2 Barauser M Bum J Brief clinical repon neural lube defects ner, linkage analysis is likely to be able to locate as an X linked condition AmJMcdGenel 1984 17 383 5 3 Tonello HV Report of akmdrcd wiihX linked anencephaly such a gene spina bifida Am J Med Genet 1984 19 111 12 4 ГопсІІо HV Higgins JV X linked midline delects Am J Because the mode of inheritance and pene­ Med Genet 198^21 143 6 5 Jenssen О Amason A Gunnarsdottir Η Petursdottir I trance are uncertain, we analysed the data un­ Fossdal R Hreidarsson S Λ family showing apparent X der the assumption of both autosomal linked inheritance of both anencephaly and spina bifida dominant and autosomal recessive models J Mtd Gena 1968 25 227 9 6 Bodmcr WF ( avalli Sforza LI Genette* evolution and One of the worst potential model mis- man San Francisco Freeman 1976 specifications is the misspecifìcation of dom­ 7 Carter CO David PA Laurence KM A family study of D major central nervous system malformations in South inance ' and therefore the recessive model was Wales J Mad Genet 1968 5 81 106 also applied m our data Assumption of a high 8 Nevin NC Johnston VÎT* A family study of spina bihda and anencephalus m Belfast Northern Ireland (1964 to 1968) penetrance may falsely generate exclusion of J Vied Crenel 1980 17 203 11 the linkage whereas a low penetrance approach 9 Williamson EM Incidence and family aggregation of maior congenital malformations of central nervous system J may reduce the power to detect linkage We MedGenet 1965 2 161 72 earned out the analysis using a variety of para­ 10 Manman LCM Hamel BCJ Sex ratio of affected and transmitting members of multiple cases ol neural tube meters with different levels of penetrance 0 45 defects J \led bena 1992 29 695 8 and 0 27 Our result did not show e\ idence of 11 Chatkupt S Lucek PR, Koemgsbcrgcr MR Johnson WG Parental sex effect in spina binda a role for genomic linkage between familial NTD and this marker 5 impnnting Am J Med (renet 1992 44 508 12 which is located within the PAX3 region itself 12 Chatkupt S Skumick JH Jaggi M Mitruka К Koenigs beiger MR Johnson WG Study of genetics epidemiology SBO has been considered a variable ex­ and vitamin usage in familial spina binda in the Lnitcd 4 States in the 1990s Seurohey 1994 44 65 70 pression of the SB gene ""^ Radiography of 13 Moase Cb I rasier DG Splotch locus mouse mutants the spine was not performed in the unaffected models for neural tube defects and Waardenburg syndrome subjects in our senes apart from two subjects type 1 in humans J Med Genet 1992 29 145 51 14 Chalepakis G Stoykova A ViiinholdsJ Tramblav Ρ Gnjss from the US families and two from the Dutch Ρ Pax gene regulators in the developing nervous system family who had SBO detected on radiography У Seumbtol 1993 24 I 367-84 15 Epstein DJ \ekemans M Gros Ρ splotch (Sp ") a mutation of the spine performed for other medical affecting development of the mouse neural tube shows α 60 deletion within the paired homeodomain οΓ Pax-3 Cell 30 lender ES Mapping complex genetic traits in humans In 1941,67 767 74 Da\ics КС, ed Genome anahsis a practical approach 16 Stapleton P, Weith A, Urbanek P, Когтік /, Busslingcr Oxford IRL Press, 1989 171-91 M Chromosomal localization of several PAX genes and 31 On J Cutting a Gordian knot in the linkage analysis of cloning of a novel family member, PAX-9 \aturc Genet complex human traits Am J Med Genet 1990,4*219 21 1993,3 292-7 32 Risch N Genetic linkage and complex diseases with special 17 Gouldmg МП, Chalepakis G, Deutsch V, brselius J, Gruss reference to psychiatric disorders Genet Epidemiol 1990, P, Pax-3, a novel murine DNA binding protein expressed 7 3 16 during earlv neurogenesis EMHO J 1991,10 Π 35 47 33 Pericak-Vance MA, Bebout JL, Gaskell PC Jr, e/u/ Linkage IS Baldwin CT, Hoth CF, Amos JA, О da-SiKa E, Milunskv studies in familial Alzheimer's disease evidence for chro A An exontL mutation in the HuP2 paired domain gene mosome 19 linkage Am J Hum Gtnet 1991,48 1031 50 causes Waardenburg's syndrome Samrv 1992,355 637 8 34 Demcmas M, I eMerrer M, Bnard ML, Elston RC Neural 19 Morell R, Friedman ТВ, МоеЬораічго S, Hartono, Seow- tube defects in France segregation analysis Am J Mid ito, Asherjr JH A frameshift mutation in the HuP2 paired Іп .і 1982,11 287-98 domain of the probable human homolog of murine Pax- 35 Fellous M, Boue J, Malbrunot C, et al A fi ve- generot ion 3 is responsible for Waardenburg ssndrome tvpe 1 in an farmi} with sacral agenesis and spina binda possible sim­ Indonesian family Hum Mol Ganci 1992,1 243-7 ilarity with the mouse T-locus Am J Med Genet 1982,12 20 Iassabehji M, Read AP, Newton VE, et al Waardcnburg's 465 87 syndrome patients have mutations in ihe human homologue 36 Bennett D The T-locus in the mouse Cell 1975,6 44 1 54 ofthePax-J paired box gene Saturn 1992,355 635 6 37 LorberJ The family history of spina bifida cystica Pediatrics 21 Care/am-Gavin M, (Marren SK, Steege Τ Waardenburg 1965,35 589 95 syndrome associated with myelomeningocele Am J Med 38 Fdiag П, ГееЬі AS, Al Aw adi SA Konsyndromal anen- Genet 1992,42 135-6 cephalv possible autosomal recessive variant Am J Mid 22 Chatkupt S, Chatkupt S, Johnson WG Waardenburg svn- Genet 1986,24 461-1 dromc and myelomeningocele in a familv J XUd Genet 39 Shaffer LG. Mara/ita ML, Bodunha J, New lin A, Nance 1993,30 83-4 WF Evidence for a major gene in familial ancntcphaly 23 Moline ML, Sandlin С Waardenburg syndrome and men- Am J Med Gtnet 1990.36 97101 ingomvelocele Am J Med Gtnet 1993.47 126 40 Clerget-Darpoux F, Bonam-Pclhe C, Hoche/ J Effect* of 24 Sambrook J, Fnisch EF, Mamaus Τ Molecular ¡.Urning А misspeurving genettL parameters in lod store analysis laboratory manual 2nd ed Cold Spring Harbor, NY Cold Вюпшп^ І9Я6 42 393 9 Spring Harbor laboratory Press, 1989 41 Carter CO, fc\ans KA, Till К Spinal dvsraphism generic 25 Wilcox ER, Rivolta MN, Ploplis В, Potterf SB, Fex J The relation to neurol tube malformations У Med G%.m.t 1976, PAX3 gene is mapped to human chromosome 2 together 13 313 50 with a highly informative CA dmucleotide repeat Hum 12 Fincmun RM, Jorde I B, Martin RA, Hasstcdt SJ, Wing SD, Mol Genet 1992,1 215 Walker ML Spinal dvsraphia as an autosomal dominant 26 Edmonds ID, James LM Temporal trends in the prevalente delect in tour families Am J Med Gena 1982,12 457 64 of congenital malformations at birth based on the birth 43 Campbell LR, Dauon DH, Sohal GS Neural tube defects defects monitoring program, United States, 1979 1987 a review of human and animal studies on the etiology of MMVCR 1990.39CSS-4) 19 23 neural tube defects teratology 1986,34 171 87 27 Verheij JBGM, hdens M, Cornel MC, Groodioff JW, Ten 44 Sever Lh Spinal anomalies and neural tube defects Am У Kate LP Incidentie en prevalentie van genetisch bepaalde MtdGtntt 1983,15 313-5 aandoeningen in Nederland, een literatuuronderzoek Sed 45 Tonello HV, Higgms JV Possible causal heterogeneity in Tjtdsi.hr Geneeskd 1994,138 71 7 spina bihdatvsiica Amy MedGaictSuppl 1985,21 13 20 28 Lathrop GM, Іліоцеі JM, Julier C, Ou J Strategies for 46 Seller MJ Neural tube defects are neurulation and can­ mutilocus linkage analysis in humans Proc S'ati Atad Vu alization forms causally distinct Am J Mid Genet 1990, USA 1984,81 3413-6 35 394 6 29 Ott J Analysis of human genetu, linkage Baltimore Johns 47 Nora JJ, Fraser FC .Medital gtnenct principle and practia. Hopkin«. University Press, 1991 2nd ed Philadephia Ixa & Febiger, 1981 308

61

Chapter 4

A Frameshift Mutation in the Gene for PAX3 in a Girl with Spina Bifida and Mild Signs of Waardenburg Syndrome

Hoi FA, Hamel BCJ, Geurds MP A, Mullaart RA, Ватт FG, Macina RA, Manman ECU.

J Med Genet 32:52-56 (1995)

63

A frameshift mutation in the gene for PAX3 in a girl with spina bifida and mild signs of Waardenburg syndrome

Frans A Hoi, Ben С J Hamel, Monique Ρ A Geurds, Reinier A Mullaart, Frederic G Barr, Roberto A Macina, Edwin С M Manman

Abstract which is expressed in defined regions of the Neural tube defects (NTD) are among the developing neural tube and in various neural most prevalent congenital malformations crest derived tissues, can cause NTD in ho­ In man. Based on the molecular defect of mozygous embryos " In the heterozygous state, Splotch, an established mouse model for РахЗ mutations do not cause but seem to pre­ NTD, and on the clinical association be­ dispose to NTD in a strain specific manner ' " tween NTD and Waardenburg syndrome A similar situation ma> exist in humans, where (WS), mutations in the PAX3 gene can be mutations in the PAX3 gene are known to cause expected to act as factors predisposing to Waardenburg syndrome (WS)," ' a condition human NTD. To test this hypothesis, 39 which is occasionally associated with NTD '* '' patients with familial NTD were screened Therefore, it is tempting to speculate that in by SSC analysis for mutations in exons 2 man, too, mutations in the gene for PAX3 (also to 6 of the human PAX3 gene. One patient referred to as HuP2) constitute genetic nsk with lumbosacral meningomyelocele was factors for NTD If so, their frequency should identified with a 5 bp deletion in exon be increased in patients with this disorder S approximately SS bp upstream of the conserved homeodomain. The deletion causes a frameshift with a stop codon al­ Materials and methods most immediately after the mutated site. ASCERTAINMENT OF PATIENTS AND DNA Clinical investigation of the index patient ISOLATION indicated mild signs of WS type I. Varying Patients were selected from the Dutch popu­ signs of this syndrome were found to co- lation in collaboration with the patient or­ segregate with the mutation in the family. ganisation BOSK and from the records of the Our results support the hypothesis that Nijmegen hospital departments Thirty nine mutations in the gene for PAX3 can pre­ Department of Human families were selected with more than one pa­ Genetics, University dispose to NTD, but also show that, in tient who had an affected third degree or closer Hospital Nijmegen, general, mutations within or near the con­ relative (first cousin, great aunt, or great uncle PO Box 9101, 6500 HB Nijmegen, The served domains of the PAX3 protein are of the proband) ' Genomic DNA was isolated Netherlands only very infrequently involved in familial from one patient from every family according F A Hol NTD. to the procedure of Miller et al ъ The types of В С J Hamel Μ Ρ A Geurds N1D in the test patients were spina bifida (37), E С M Manman (J Med Genet 1995,32 52 5b) encephalocele (1), and cramorachischisis (1) Department of Child Neurology, University Hospital Nijmegen, Neural tube defects (NTD) are congenital mal­ SSC ANALYSIS PO Box 9101, 6500 HB Nijmegen, The formations resulting from incomplete closure DNA fragments overlapping exons 2 to 6 of Netherlands of the neural tube dunng early embryonic de­ the human PAX3 gene were amplified bv the R A Mullaart velopment In man, their prevalence at birth is polymerase chain reaction (PCR) from gen­ about 1/1000 NTD are thought to result from Department of omic DNA together with 5' and 3' flanking Pathology and an interaction between environmental and pre­ intron sequences Amplification was carried Laboratory Medicine, disposing geneuc factors which interfere with out in a total volume of 25 μ| containing 50 ng University of Pennsylvania School the normal neurulation process ' The in­ of genomic DNA, 0 45 mmol/1 of each primer, of Medicine, volvement of genetic factors is reflected by the 0 1 mmol/1 dCTP, 0 4 mmol/1 dATP, 0 4 mmol/ Philadelphia, increased recurrence nsk for close relatives of 1 dG'I P, 0 4 mmol 1 dTTP, 0 1 μΐ |V'P]dCTP Pennsylvania 19104, USA patients Only about 3% of all cases are familial (Amersham) in PCR buffer (50 mmol/1 KCl, F G Ban and large families with multiple cases arc ex­ lOmmol'l Tns-HCl, pH 8 3, 1 mmol 1 DTF, tremely rare 2 ' Therefore, it is practically im­ 0 001% gelatine, 1 5-6 mmol/1 MgCl ) with The Wistar Institute, possible to identify the underlying genetic 0 5 U Taq DNA polymerase (Boehnnger 3610 Spruce Street, 4 0 Philadelphia, factors by linkage studies Flucidation of these Mannheim) Samples were denatured at 92 С Pennsylvania 19104, factors is essential to understanding the patho­ for five minutes and then subjected to 35 cycles USA genesis of NTD and for the identification of of amplification 92°C for 50 seconds, 55°C R A Macina persons at nsk of having affected offspring for 50 seconds, 72°C for one minute 30 sec­ Correspondence to An alternative approach to shed more light on onds Fxon 2 was analysed as two partly over­ Dr Manman these genetic factors is the analysis of suitable lapping fragments The following pnmers were Rtceivcd 31 May 1991 Revised \eruon accepted for animal models In one of the models for NTD, used for amplification (fig 1), some of which publication 1 August 1994 Splotch,'"' mutations in the gene for Paxi,6 are identical to those reported by Tassabehji a

65 Exon 2 Εχοπ 3 Exon 4 Exon 5 Exon 6

t E_3 1 t 1 ! 3 ί 1 t 1

Figure 1 Schematic rctprescniation r)/í/ít par/

a!" exon 2, 5' fragment (266 bp) 5J-GAA- session of the conserved paired domain The GACTGCGAAATTACGTGCTGC-3' and paired domain of the PAX3 gene is encoded 5'-ACAGGATCTTGGAGACGCAGCC-3', by (part of) exons 2, 3, and 4 ;6 In addition exon 2, 3' fragment (208 bp) 5'-AAC- the PAX3 gene contains two other conserved CACATCCGCCACAAGATCG-3' and 5'- domains an octapeptide motif encoded by a GACCACAGTCTGGGAGCCAGGAGG-3', segment of exon 4,21' and a homeodomain en­ exon 3 (237 bp) 5'-CACCTGGCCC- coded by the 3' and 5' part of exons 5 and AGGGTACCGGGTAC-3' and 5'-CGGGG- 6, respectively " To test the hypothesis that TAATAGCGACTGACTGTC-3', exon 4 mutations in the PAX3 gene might predispose (242 bp) 5'-AGCCCTGCTTGTCTCAAC- to the development of NTD, genomic DNA CATGTG-3' and 5'-TGCCCTCCAAGT- was isolated from 39 patients of multiple case CACCCAGCAAGT-3', exon 5 (304 bp) 5'- NTD families and the exons were screened GACTTGGATCAATCTCAGTTTT-3' and for mutations by SSC analysis (Materials and 5'-TAGGACACGGAGG'riTGG-3', exon 6 methods, fig 1) When exon 5 was analysed, (250 bp) 5'-'ITCATCAGTGAAATCCTT- not only the normal band pattern, but several AAATT-3' and 5'-CGCCTGGAAGTTACT- additional bands were observed in the DNA of TTCTA-3' Aliquots of the amplified DNA one patient (fig 2A) To evaluate this further, were mixed with one volume formamide dye the amplification products were subjected to denaturing gel electrophoresis, which showed buffer, denatured at 95°C for five minutes, and the presence of a heterozygous deletion (fig placed on ice, 4 μΐ samples were loaded on a 5% 2B) The location and size of the deletion were non-denaturing Polyacrylamide gel containing determined by direct sequencing of the eluted 10% glvcerol and on a similar gel without allelic DNA fragments (Materials and meth­ glycerol Electrophoresis was for three to six C ods) A 5 bp deletion was detected in exon 5 hours at 35W and 5 C The gels, were dried approximately 55 bp upstream of the ho­ and exposed overnight to Kodak X-omat S film meodomain (fig ЗА). This causes a shift in in order to visualise the separate bands the normal reading frame for translation with premature termination of polypeptide synthesis almost immediately downstream of the mut­ DIRECT SEQUENCING OF NORMAL MUTANT ated site (fig 3B) ALLEI ES To determine the nature of the shifted bands in the SSC analysis, 4 μΐ of amplification prod­ uct was loaded on a 6 6% denaturing Poly­ CLINICAL EXAMINATION OF THE PATIENT AND HER RELATIVES acrylamide gel The gel was electrophoresed at 2 60 W for three hours at room temperature, Knowing that PAX3 mutations can cause WS ' dried, and exposed overnight to Kodak X-omat (MIM 193500), signs of this disorder could be S film to visualise the bands Bands rep­ present in the patient and some of her relatives resenting wild type and mutant alleles were cut Therefore, the family (fig 4) was clinically (re)- out of the gel DNA was eluted from each of examined The major signs of WS are a typical the gel slices in 50 μΐ distilled water for one hour facies with dystopia canthorum as the most at 37 'C and reamplificd under the conditions frequently observed characteristic, pigmentary described above Subsequently, the amplified disturbances like a frontal blaze of white hair, DNA fragments were purified by electro­ heterochromia indes, white eyelashes and phoresis on a 1% agarose gel (one hour, 10 V/ leucoderma, and partial or complete cochlear cm), allowed to migrate into ultra low gelling deafness WS follows an autosomal dominant temperature agarose (Sigma), and sliced out of pattern of inheritance with a wide variability of the gel This material served as substrate for expressivity "" direct sequencing using the Cycle Sequence The index patient (III 5) was seen at the age Kit (BRL) according to the protocol of the of Щ years She was bom with a lumbosacral manufacturer Sequences were determined in meningomyelocele for which she was operated two directions with the forward and reverse on shortly after birth Because of developing amplification primers after 5' end labelling with hydrocephalus, a ventnculopentoneal shunt "P was inserted She is mentally retarded. Her height is 128 5 cm (<3rd cernile), she weighs 26 kg (50th cernile for height), and has an Results occipitofrontal circumference of 53 7 cm A PAX3 GENE MUTATION IN A PATIENT WITH (50th-90th centile) She has dystopia can­ SPINA BIHDA thorum (ICD 43 mm, >97th cernile, OCD PAX3 belongs to a family of embryonic tran­ 85 mm, 50th centile), leading to blepharo- scription factors, which arc related by pos­ phimosis, broad and high nasal root, hy-

66 hypoplastic nasal alae, and a round nasal tip. She has vitiligo of the left hand and wrist. She has no heterochromia irides and no hearing loss. The maternal grandfather of the index pa­ tient (1.2) has heterochromia irides and dys­ topia canthorum, but no pigmentary abnormalities and no long standing hearing loss. No abnormalities were seen on a pho­ tograph of the maternal grandmother (1.1). The maternal aunt of the index patient (II. 1) has no signs of WS. Another sister of the mother (II. 3) was born with a lumbar men­ ingomyelocele and hydrocephalus, but died at the age of 6 months without having left the Figure 2 Molecular analysa of exon 5 of the PAX3 gene. Auwradiographs show the allelic band patterm obtained hospital. It is unknown whether she had any with (A) SSC analysis and (B) denaturing gel analysis sign of WS. No material was saved for genetic of genomic DNA from Two control persons (lanes 1 and 2) analysis. and from a patient with spina bifida (lane 3). With DNA from the patient, SSC analysis shows bands with Several sibs of the index patient III.4, III.8, abnormal mobility in addition to the wild type bands, III.9, and III. 10, show the facial characteristics indicating the presence of a heterozygous mutation in exon of WS. III.9 was born with a white forelock, 5. On denaturing gel electrophoresis the aberrant allele appears to be of reduced length owing lo a deletion. which subsequently disappeared, and has uni­ lateral hearing loss. III.8 had poliosis. These observations show that WS is indeed poplastic nasal alac, a round nasal tip, and segregating in this family and that the index smooth philtrum. There is a naevus above the patient has a mild expression of this syndrome right eye. The palate is high arched and there in combination with spina bifida. Based on the is dental crowding. Below the spina bifida she presence/absence of specific symptoms, three has a deep sacral pit. She has no heterochromia subtypes of Waardenburg syndrome are dis­ irides, no pigmentary disturbances, and no tinguished. WS-I (MIM 193500) and WS-II hearing loss. (MIM 193510) are characterised by the pres­ The mother of the index patient (II.4) has a ence or absence of dystopia canthorum, re­ similar appearance with dystopia canthorum spectively, whereas the disorder is diagnosed (ICD 41mm, >97th cernile; OCD 85 mm, as WS-III (MIM 148820) if limb deformities 25th-50th cernile), leading to blepharo- are among the symptoms. Accordingly, the phimosis, brushy eyebrows, a high nasal root, present family can be categorised as having WS-I. So far, WS with NTD patients have only been reported in families with WS type I.

CORRELATION BETWEEN MUTATION AND PHENOTYPE The pattern of inheritance of WS is compatible with that of an autosomal dominant disorder. To investigate further the relationship between the clinical signs and the mutation discovered in the index patient, exon 5 was amplified from the DNA of all available persons and analysed by denaturing gel electrophoresis. As can be seen in fig 5, there is an exact correlation between the presence/absence of the abnormal allele and the phenotype (Z=+2-40 at θ = 00).

Discussion The association between NTD and WS is well documented. Interestingly, of the 11 patients ., AGC GAG CGA GCC TCA GCA CCC ІСАА TfjA GAT GAA GGC TCT GAT with NTD and WS reported since 1988, eight N Ser Glu Arg Ala Ser Ala Pro Gin Ser Asp Glu Gly Ser Asp represent familial cases of NTD.""4 This in­ cludes the index patient of the present study, who had a maternal aunt with spina bifida. AGC GAG CGA GCC TCA GCA CCC XXX xxAGA TGA AGG CTC TGA Apparently, there is an increased recurrence M Ser Glu Arg Ala Ser Ala Pro Arg *** risk of NTD in families with WS, which cor­ roborates the common aetiology of both dis­ Figure 3 (A) DNA sequence of the normal (Ν) and mutant (M) allele of exon 5 of a patient with spina bifida as shown by cycle sequencing. The boxed sequence in the normal orders. The molecular defect in two other allele is deleted in the mutant allele. The arrow marks the site of the deletion in the patients with WS and NTD has previously mutant allele. (B) Partial cDNA and protein sequence of the region containing the been reported.12 " Both cases concern missense deletion as deduced from the cycle sequencing results. The mutant gene contains a premature stop codon shortly after the site of the deletion. The boundary between exons 4 mutations in exon 2 changing an amino acid and 5 is indicated by a vertical bar. within the paired domain of the PAX3 protein.

67 Figure 4 Pedigree of the family with two closely related patients with lumbar meningomyelocele and hydrocephalus: the index patient III. 5 and her maternal aunt II. 3. All members zvere clinically examined for symptoms of WS. Those with a positive diagnosis of WS in addinoti to the index patient are indicated by shaded symbols.

determining factors. A similar situation may exist in humans, where additional factors may ЪЪ^а modify the phenotypic expression of the same PAX3 mutation in different persons. Spina bifida is not the only malformation of ònèòòòή homozygous Splotch embryos. In 50% ex- encephaly is observed and congenital heart de­ іМШІ ••—•:'-•>:•:• •.- ·;--;ν Κ ·χ&&υ& ШЁШ&г jflQrfnfefrr Γ Γ fects also occur, which are regarded as the major cause of death. In humans, anenccphaly 1.2 111 11.4 11.5 III 3 III.4 III.S III.6 III.7 III.8 III.9 and congenital heart defects do not seem to Figure 5 Cosegregation of the exon 5 deletion with be associated with WS but, considering the symptoms of WS. All available members of the family influence of other genetic factors on the pheno- were analysed for the presence of the mutant allele by IXJR amplification of exon 5 and subsequent denaturing gel type, it may be worth looking for PAX3 mut­ separatum of amplified fragments. All members diagnosed ations in patients with NTD and congenital as having WS symptoms appear to carry the mutant heart defects. allele. The pathophysiological processes leading to NTD in Splotch have not yet been elucidated. Suggested mechanisms include delayed mi­ Here we show that mutations disrupting the gration of neural crest cells and an abnormal open reading frame of the PAX3 gene may also curvature of the caudal region. More likely, be found in patients with WS and NTD. these phenomena are secondary to a defect of Despite the fact that carriers of a PAX3 the neuroepithelium, where the РахЗ gene is mutation probably have an increased risk for expressed before neural tube closure.10 The NTD, in the present study only one of 39 detection and functional characterisation of patients with familial NTD was found to have РАХЗ gene mutations in patients with NTD such a mutation indicating that, in general, may help to clarify the pathogenesis of NTD PAX3 mutations are an infrequent cause of further. familial NTD. However, SSC analysis is not completely sensitive, leaving the possibility that We thank the working group "Hydrocephalus en Spina Bifida" of the Dutch patient organisation BOSK for their assistance in some mutations have not been detected by this contacting families with NTD. We also thank Professor Dr H H Ropers for critically reading the manuscript, H Egtbens for method. Further, mutations could be present blood sampling, and S van der Veldc-Visser and E Boender­ in exons 1, 7, or 8, which have not yet been van Rossum tor cell culmnng and KBV transformation. This study was financially supported by the Dutch Prinses Beatrix examined in detail. Nevertheless, mutations Fonds, grants no 90-3154 and 93-005. RAM acknowledges within or near the conserved domains of the the support of Dr Harold Riethman, in whose laboratory his PAX3 protein are not likely to play a major experiments were carried out (NIH grant CA47983). role in familial NTD. 1 Copp AJ, Brook FA, Estibeiro P, Shum ASW, Cockrofl Di- Because of the findings in Splotch mice, it is The embryonic development of mammalian neural tube defects Prog Neumbiol 1990;35:363-403. not surprising that NTD may be present in 2 Chalkupt S, Lucek PR, Koenigsberger R, Johnson WG. humans carrying a mutation in the РАХЗ gene. Parental sex effect in spina bifida: a role for genomic imprinting? Am J Med Genet 1992;44:508-12. Homozygous Splotch embryos die on day 13 3 Manman ECM, Hamel BCJ. Sex ratios of affected and of gestation and 50% have lumbosacral spina transmitting members of multiple case families with neural tube defects. J Med Genet 1992;29:695-8. bifida. Heterozygous animals display pig­ 4 Hoi FA, Geurds MPA, Jensson O, et al. Exclusion mapping mentary disturbances, but have a normally of the gene for X-hnked neural tube defects in an Icelandic Tamils Hum Genet 1994;93:452-6. closed neural tube, yet breeding experiments 5 Lyon MF, Searle AG. Genetic variants and strams of the have shown that a heterozygous РахЗ mutation laboratory mouse. 2nd ed. New York: Oxford University Press, 1990. influences the incidence of NTD in animals 6 Epstein DJ, Vekemans M, Gros Ρ Splotch (Sp-[1), a mutation already committed to NTD development." " affecting development of the mouse neural rube, shows a deletion within the paired homeodomain of Pax-3. Cell Apparently, in those animals the occurrence of 1991;67:767 74. NTD depends on a combination of pre­ 7 Goulding MD, Chalepakis G, Deutch U, Ersehus JR, Gruss

68 Ρ Pax-3, a novel murine DNA binding protein expressed an Indonesian familv Яши Mol*. Genet 1992,4:243 7 during early neurogenes* bXÍBO J 1991,10:1135-47 17 Bun J, Greenberg J, Winship I, Beightnn P, Ramesar R A 8 Moase CE, Trasler DO Splotch lotus mouse mutants splice junction mutation in PAX3 causes Waardenburg models for neural tube defects and VX aardenburg syndrome svndrome in a South African family Hum Molte Genet type 1 in humans J Med Gena 1992,29:115 51 1994,3:197 8 9 Konyukhov BV, Mironova V Interaction of the mutant 18 De Saxe M, Kromberg JGR, Jenkins Τ Waardenburg svn­ genes splotch and fidget in mite Savia Genet 1979,15: drome m South Africa Part I \n évaluation

69

Chapter 5

PAX Genes and Human Neural Tube Defects: An Amino Acid Substitution in PAX1 in a Patient with Spina Bifida

Hoi FA, Geurds MP A, Chatkupt S, Shugart YY, Balling R, Schrander-Stumpel CTRM, Johnson WG, Hamel BCJ, Manman ECM.

J Med Genet 33:655-660 (1996)

71

PAX genes and human neural tube defects: an amino acid substitution in PAX1 in a patient with spina bifida

F A Hoi, Μ Ρ A Geurds, S Chatkupt, Y Y Shugart, R Balling, С Τ R M Schrander-Stumpel, W G Johnson, В С ƒ Hamel, Ь С M Manman

Abstract role Pedigree analysis and linkage data suggest From studies in the mouse and from the that the human X chromosome contains one or clinical and molecular analysis of patients more contributing genetic factors ' More with type 1 Waardenburg syndrome, par­ specifically, several members of the PAX gene ticular members of the PAX gene family family, encoding a class of related embryonic are suspected factors in the aetiology of transcription factors,1 have been proposed as human neural tube defects (NTD). To candidate genes for NTD At present nine investigate the role of PAX1, PAX3, PAX7, members have been identified and studies in and PAX9, allelic association studies were the mouse have shown that they are all performed in 79 sporadic and 38 familial expressed in the developing brain, the neural NTD patients from the Dutch population. tube, or the paraxial mesoderm Department of Human Sequence variation was studied by SSC Mutations in the gene for Pax-1 are respon­ Genetics, University analysis of the paired domain regions of sible for the phenotype of the mouse strain Hospital Nijmegen, the PAX1, PAX7, and PAX9 genes and of undulated with various vertebral anomalies PO Box 9101, 6500 HB Nijmegen, the complete PAX3 gene. In one patient Interestingly, litters resulting from a cross The Netherlands with spina binda, a mutation in the ΡΛΧ1 between undulated (un) and Patch (Ph) mice F A Hol gene was detected changing the conserved have a high incidence of lumbar spina bifida Μ Ρ A Geurds В С J Hamel amino acid Gin to His at position 42 in the occulta' indicating that Pax-1 can be involved E С M Manman paired domain of the protein. The muta­ in specific forms of NTD tion was inherited through the maternal Prom the association between spina bifida Department of line from the unaffected grandmother and and Waardenburg syndrome type 1 (WS1),° " Neurosciences and Pediatrics, was not detected in 300 controls. In the the PAX3 gene can be regarded as a risk factor UMDNJ-New Jersey PAX3 gene, variation was detected at sev­ for human NID Mutations in this gene have Medical School, eral sites including a Thr/Lys amino acid Newark, New Jersey, been detected in various patients with both USA substitution in exon 6. All alleles were spina bifida and WS I ' Moreover, mutations S Chatkupt present among patients and controls in in the murine homologue of this gene cause the about the same frequencies. However, an Splotch phenotvpe Homozygous Splotch mice Department of increased frequency of the rare allele of a exhibit severe NTD whereas heterozygous Genetics and Development, silent polymorphism in exon 2 was found Pax-3 mutations seem to increase the inci­ Colombia University, in NTD patients, but no significant associ­ dence of NTD in other predisposed strains New York, NY, USA ation was observed (p=0.06). No sequence Y Y Shugarl Linkage analysis has not provided genetic variation was observed in the paired evidence for a major role of PAX3 in human Institut fur domain of the PAX7 and PAX9 genes. NTD ' Saugetiergenetik, Our findings so far do not support a major The above findings prompted us to investi­ GSb-Forschungszentrum role of the PAX genes examined in the Neuherberg, gate further the role of PAX 1 and PAX3, and Oberschleissheim, aetiology of NTD. However, the detection their paralogues PAX9 and PAX7, respectively, Germany of a mutation in PAX1 suggests that, in bv searching for mutations and sequence van- R Balling principle, this gene can act as a risk factor ants in association with non-svndromic NTD for human NTD. Department of Clinical Genetics, University (JMulOim.1 1946 33:655-660) Hospital Maastricht, Materials and methods Maastricht, Key words neural tube defects, spina bthdj IH\ The Netherlands VSv hRTAIWH-Vl OH PVTIFSTS CTRM genes Patients with non-svndromic NTD were se­ Schrander Stumpel lected from the Dutch population in collabora­ tion with the Dutch patient organisation Department of Neurology, Neural tube defects (NTD) constitute a major BOSK and from the records of our institute LVtDNJ-Robert Wood group of congenital malformations with an Multiple case families were selected according Johnson Medical incidence of approximately 1 1000 pregnan­ to the criteria described previously In short, School, New Brunswick, New Jersey, cies It is generally accepted that they represent there had to be two or more affected members USA multifactorial traits with genetic and em iron- in each tamilv with a close degree of relation­ VX G Johnson mental factors contributing to the aetiology ship (€3) In this way 38 multiple case families Information about the identity of the genetic were selected One affected member of each of Corn.spondi.nc lo Dr Hol factors involved is scarce Induction of NTD these families was included in the present by anti-epileptic drugs' and prevention of studv The types of NTD in this group were Reeuved 27 October 1045 NTD h\ folic acid suggest that genes for drug Revised version avwptvd for spina bifida (36), encephalocele (1), and publication 22 March 110b receptors or metabolic enzymes ma\ plav a craniorachischisis (1) In addition, a group of

73 w: >

PAX7

PAX1

higure 1 St.ht.mauc r pnstiitatioii of parts of the different PAX series that wert subjected lo S&C analysis Arrowheads uilh lonneclinz bars nprttint thi amplification primers and amplified fragments The filled region the hatched region and the double hatched region rtpnst.nl the paired dt mam the Octapepltde sequenCt. and the homeodoinatn respectively

79 sporadic patients was also analysed, which for five minutes and then subjected to 35 cycles included patients with spina bifida (75), with of amplification 92°C for 50 seconds, 55°C for anencephaly (2), and with encephalocele (2) 50 seconds, 72°C for one minute 30 seconds Lnaffected and unrelated subjects were ran­ Aliquots of the amplified DNA were mixed domly chosen from the Dutch population and with 1 volume formamide dye buffer, dena­ used as a control group in the present studv tured at 95°C for five minutes, and placed on Blood was sampled and DNA was extracted ice Samples (4 μΐ) were loaded on a 5% according to the procedure of Miller tí al non-denaturing Polyacrylamide gel containing 10% glycerol and on a similar gel without glyc­

SbC ANA! i sis erol Electrophoresis was for 3-6 hours at 40 W PCR amplification was performed to produce and 4 С The gels were dried and exposed the appropriate DNA fragments (fig 1) Ampli­ overnight on Kodak X-omat S film fication was carried out in a total volume of 25 μΙ containing 50 ng of genomic DNA, 0 45 SCQLl ΝΠΝΟ OI· NORMAL AND VARIANT ALLELES mmol 1 of each primer (table 1, Isogen To determine the molecular nature of the Bioscicnce, The Netherlands), 0 1 mmol 1 shifted bands observed by SSC analysis, bands dCTP, 0 4 mmol/1 dATP, 0 4 mmoll dGTP, representing wild type and variant alleles were 0 4 mmol 1 dTTP 0 1 μΐ fa[32]P]dCTP cut out of the gel DNA was elutcd from each (Amersham) in PCR buffer (50 mmol 1 KCl, of the gel slices in 50 μΐ aquadest for one hour 10 mmol 1 TRIS HCl, pH 8 3, 1 mmol 1 DTE, at 37°C and reamplified under the conditions 0 001% gelatine, 1 5 6 mmol 1 MgCl,) with described above Subsequently, the amplified 05 U of Taq DNA polymerase (Boehringer, DNA fragments were purified by electrophore­ Mannheim) Samples were denatured at 92 С sis on a 1% agarose gel (one hour, 10 V/cm),

Table I Primen used for SSf mal\sis

S t (bp) Η π irti pro яг Hever e primer

PAXJ I 13S С t Ι GGATATAÄlTl С CGAGC G 3 •> CGCIGAGGCCCTCCCT1A 3 2A 266 GAAGAC IGCGAAATTACGIGCIGC 3 5 ACAGGATCΠGGAGACGCAGCC 3 2D 20В ААГСАСА1СС Ci( ( At AACATC С 3 5 GAC CACACTCTGGGAGCCACGAGG 3 3 237 С « ( TGCC CC AGOG 1 AC CGGG I AC 3 5 CGGGGIAArAGCGACTGAC ГОТС 3 4 'з' AG( ( ( 1 (•( I 1GI С ТГААССА1G ГС 3 5 TGCCCTCCAAG1С ACCCAGCAAGT 3 5 304 CACTTGGATC ЛАК К ЧОТТП 3 5 IAGGACACGGAGGrriGG 3 2э0 I I ( AI С AGTGAAATC С I I \AAI I 3 5 CGCC1GGAACTTACTTTCTA 3 3">0 CJAAC ПК К IGCTGGCCTA 3 5 TGGTTCTGGIAIAtAGCAAATC 3 313 ССТСТ"! 1111 IAGGIAAIGGGA 3 5 IGAGTTTATCTCCC ITC С AGO 3 PAX·* 2Α IHK -, К ( ATCCTCACC ( ГСС A( С I 3 •> CGGGAGATGACACAGGGCCG 3 2Β 214 -, GC СіЛС С С С I СГГТАЛС С AC A 3 3 AAGCCAGCIGC С AGCCTCTGTG 3 3 200 3 ССССАТГС(АК I I К CACTC 3 5 IGC CGGCTCAGCIGC С TTCTCA Э 4 189 -, I IGCTC1I ITGCC1 I TGAAI 1 ICT3 -, GCC 1С GC AGCCCAGGGAA 3 PAXl 1 194 SKC С.С.ГТС ACTCTTGTC I GG 3 5 ATCTTGCIС ACGCAGCCGTG 3 2 201 :> ACGCC А К С G( ITGCGCATT 3 3 AGTCCCGGArGlGCTTGACC 3 3 200 -, ( ( TGGC GC G( I At АЛ( GAG\ 3 5 TGGAGC ICACCGAAGGCACA 3 4 200 1 С !GG( AT( TTTGCC TGGGAG 3 η AGGGGTACTGGTAGATGTGG 3 PAX9 1 200 -, AI I [ IGC \GAGCCAGCCTTC 3 5 С GAGCCCGTC TC GTTGTATC 3 188 S GGCCCAACIGGGC А К ( САГ 3 5 GGK rCTCTGCTTGlAGCTC 3 202 5 \l( I IGC ( AGCAGCC AKGG 3 5 TGCCGA1C I I CTTGCGCAGAAT 3 200 •> АІГ 1 It GC ( К GCAGATC CCG 3 3 GGGTACGAG1AGA rGTGGTl GT 3 222 -, AITACGACTCATAC AAGC AGC АСГ 3 5 С ГС I AC С TTGGTCGGTGATGG 3 allowed to migrate into ultra low gelling from 300 unaffected controls suggesting that temperature agarose (Sigma), and sliced out of this alteration is aetiologically related to the the gel. This material served as substrate for NTD of this person. Moreover, the functional direct sequencing using a cycle sequence kit importance of the Gin residue at this position according to the protocol of the manufacturer is corroborated by the fact that it has been con­ (BRL). Sequences were determined in two served between several human PAX genes as directions with the forward and reverse ampli­ well as the murine Pax-1 gene (fig 5). Finally, fication primers after 5' end labelling with the predicted structure of the mutant peptide [32] P. Whenever the resolution was too poor to showed the loss of a beta sheet conformation cut the allelic bands out of the SSC gel surrounding the displaced amino acid (fig 6). separately, non-radioactive PCR was per­ Together, these results argue for a contribution formed on genomic DNA. Amplification prod­ of the PAX1 mutation to the development of ucts were purified on an agarose gel and cloned spina bifida in one patient. However, the same into the Smal site of plasmid vector Bluescript mutation was also found in the unaffected SK* (Stratagene). Dideoxy DNA sequencing mother and grandmother showing that this was conducted on double stranded DNA from factor alone is not sufficient to induce a NTD positive clones using the T7 sequencing kit during embryogenesis. (Pharmacia) with T7 and T3 primers. Se­ In addition, we have examined all patients quences were obtained from at least three for sequence variation in the paired domain of independent clones. the PAX9 gene (fig 1), which is highly homolo­ gous to the PAX1 gene. No band shifts were detected. Results SSC ANALYSIS Ob' THE PAX 1 AND PAX9 GENES Double mutant mice with the genotype ¡un ss< WAI.YSIS Ob THE РАХЗ AND PAX7 GENES un;Ph/+¡ exhibit an occult form of spina Although linkage studies have not provided any bifida' indicating that mutations in Pax-1 can indication for PAX3 being a major aetiological influence the development of the vertebral factor for familial NTD,"1 a less prominent role arches. In order to assess the relevance of cannot be excluded in this way, in particular for PAX1 for human NTD, the paired domain sporadic cases. Therefore, a similar strategy- region, which is the only part of the gene that was applied as performed for the PAX1 gene, has been cloned to date, was subjected to SSC that is, using SSC analysis to search for muta­ analysis. DNA from 79 sporadic and 38 famil­ tions or rare sequence variants primarily ial NTD patients from the Dutch population occurring among patients. The complete cod­ was screened for the presence of sequence vari­ ing sequence of the PAX3 gene contained in ation (fig 1). This resulted in the detection of a exons 1 to 8, together with their flanking intron shifted band in a single sporadic patient with sequences, were analysed (fig 1). The most spina bifida (fig 2A). Direct sequencing of the dramatic change in the banding patterns was shifted fragment showed a nucleotide substitu­ observed for exon 5 in a familial patient. The tion (G—>C, fig ЗА) leading to the exchange of detailed analysis of this case, which turned out glutamme at position 42 of the paired domain to be a 5 bp deletion causing spina bifida and by histidine. This nucleotide substitution WS1, has been described elsewhere. Besides disrupts an Alu\ restriction site in the gene. the normal band pattern, shifted bands were Therefore, the presence of the base change detected for exons 2, 6, and 7 (fig 2). The could be confirmed by Alu\ restriction diges­ nature of the observed shifts was determined as tion of the PCR product leaving the DNA from described in Materials and methods. One of this patient intact (fig 4). In the same way it the shifts observed for the exon 2 fragment (fig was shown that the base change was absent 2B) and the shift observed for the exon 7 frag-

PAX1 PAX3

Ρ С С Ρ С С Ρ С Ρ С С Ρ С С

щт

Paired Exon 2 Exon 2 Exon 6 Exon 7 domain fragment fragment fragment fragment exon

Figure 2 Allelic band pattern obtained through SSC analysis of different fragments of the PAX I gene (A) and the PAX3 gene (B-Ε). The first lane shows the altered pattern (P) which teas occasionally observed in patients or controls or both, whereas the more common SSC pattern (C) is shown in lanes 2 and .?. Arrowheads mark the observed shifted allelic bands.

75 Tabls 2 iii.í/ií<.f/L/i< oí ritt drtfst di! SS( Imi irr flu ƒ' I \ î e Ut fi/ Durali /• ÍMtíír and labk ? hr^üitsriss o( thsT alkL slironiosomt of the sxon soritiols 2C 1 polymorphism nr Dull h paru.nts and íOntmls

Piru ir ( ,m Í Pjrunr tsrirrols

SSC Im&iisni \ і-ч il \ N ir \ \ 1 ηφι ìlC\ \ Fri ivilitt t-\ m 2 0 intr η pi Kmi rphism •> 01 ООН S 41 0 114 1 ilkli Id 22d (120 li -ISS 0 15 0 06 E\on 2 sileni ρ Kim rphism p in 0 1 (¡0 224 0 2(1 E\ in (ι \A pohim rphism 1 (il OOi ( 41 Olid t-NOti ~ 3 intr in p( hnuirphism SD

mem (fig 2E) were shown to represent single (Thr315Lys, fig 3D) According to the struc­ nucleotide substitutions in the 5 and 3 ture prediction analysis (not shown) this amino flanking intronic sequences of exon 2 (T—»C, acid substitution does not seem to be influenc­ fig 3B) and 7 (G »A, fig 3b), res>pectivel\ ing the protein structure Genotyping of the Another shift ol the same fragment of exon 2 control group showed that all the sequence (hg 2C) turned out to be the result ot a single \anants were present both in patients and con­ nucleotide substitution (C—>T) at the third trols with approximated equal frequencies in Τ allele position of codon 43 representing a silent base both groups (table 2) Interestingly, the of the exon 2 silent polymorphism seemed to change (G1\43G1\, fig 3C), which has previ- be present more often in patients than in con­ ouslv been reported bv Tassabehji a al ' In the trols However, a thorough analysis of the data 3 half of exon 6, а С ->Л change was identified (fig 2D) downstream of the homeodomam (hetero7vgosuy1iomozygosit> scoring, Τ allele frequency determination) applying chi-square statistics, did not show a significant association between the Τ allele and NTD (p=0 06, odds Τ ratio = l 49, 95 Я confidence interval 0 96- Ρ 2 30,table 3) A Hnallv, for PAX7, which closely resembles X PAX3 in structure and expression, the same 1 groups of patients and controls were screened for the presence of sequence variation in the 1 paired domain (exons 2 to 4, fig 1 ), which is the onl\ part of this gene cloned so far No band shifts were detected in the SSC analysis с t 5'-actgcgaaattacgtgctgctgttctttgctttt Discussion tattttcctccagtgacttttcccttgcttctct PAX genes encode a class of highly conserved transcription factors with a characteristic DMA Intron 1 ι Exon 2 binding paired domain, which play important roles in embrvonic development ' To deter­ ttttcaccttcccacagTGTCCACTCCCCTCGGC-3' mine more accurately the extent to which genetic variation in the PAX1 and PAX3 genes, and their paralogues PAX9 and PAX7, might predispose to NTD we have performed SSC analysis of both familial and sporadic patients Exon 2 Analysis of the paired domain of PAX 1 showed 5'-CAGGGCCGCGTCAACCAGCTCGGCGGÇGTTTTTA-3' a missense mutation in one sporadic NTD GlnGlyArgValAsnGlnLeuGlyGlyValPhe patient The NTD was detected by amniocen­ I tesis and biochemical analysis of amniotic Gly fluid, indicated by the fact that the mother was a known carrier ot a balanced translocation t(7,20)(q22,ql3 2) After pregnancy termina­ tion at 19 weeks of gestation, clinical examin­ ation showed an open lumbar spina bifida of A Exon 6 Intron 6 about 1 5 cm in size Uniparental disomy of chromosome 7 or 20 in the fetus was excluded 5' -TACCAGCCCAÇATCTATTCCACAAGgtaccgagg-t 3 ' by the analysis of genetic markers Cytogenetic TvrGlnProThrSerlleProGln analysis showed the presence of the same I balanced translocation in the fetus and the Lys maternal grandmother Apparently, the muta­ tion in the PAX1 gene, which is located at 20pll, cosegregates with the translocation chromosomes Although there is no indication Exon 7 Intron 7 for a maior pathological effect of the balanced t translocation, an influence on the phenotype of 5'-CCTCTCACCTCAG gtcagtcccgtgtttctagac - 3' the fetus cannot be excluded In fact, under a multifactorial threshold model both genetic Flours 3 Partial D\A and prouin ssqiurut of ι lit paired domani i\on of 1ЧХ1 (A) as abnormalities, that is, the PAX1 mutation and stell as PAX3 s\on 2 (B, C), LXOII 6 (D) and exon ? il) ¡he nusLotids substitutions thai the translocation, may have contributed to the art rtsponsiblt for tlu SSC shifts in hS 2 art. slioun ll'liui the rnn-Londs substitution un SS rise to an amnio acid substitution in tlu dsdiusdpsptids. this is also shoz*n appearance of the NTD

76 ation (p>0 1) was observed with any of the -o alleles of the dinucleotide polymorphism and similar results were obtained using patients and controls from the US population (not shown) Moreover, no significant association £ was observed between NTD and the nevvlv detected pohmorphisms described in this study, although suggestive results were ob­ tained with the Τ allele of а С Τ silent polvmorphism in exon 2 (p=0 06) The Τ at f the pohmorphic site bv itself does not • Mutant allele necessanlv have to be pathogenic to explain a • Wild type allele possible association, however, one could imag­ ine that a difference in codon usage or rigare 4 Pedigree of'thefantih in ichich the PAXl mutation и segregating l*nscilcc oj pre-mRNA processing might lead to a slight the mutation could be confirmed bs Aliti restriction digeritili oj lile anipli/icd PAXl parted domani fragment Heterozygous earners shoze a mutant alleile band m addition to the iclld inequality in expression efficiency between rvpe band both allelic forms of the PAX3 gene In this respect, embryos with a Τ allele would have a PAXl HUMAN ELAQLGIRPCDISR 0 LRVSHGCVSKILARY minor disadvantage during neurulation How­ Paxl MOUSE ELAQLGIRPCDISR Q LRVSHGCVSKILARY ever, at this point the evidence for association is РАХЭ HUMAN EMAHHGIRPCVISR Q LRVSHGCVSKILCRY not convincing and additional groups of PAX7 HUMAN EMAHHGIRPCVISR Q LRVSHGCVSKILCRY PAX9 HUMAN ELAQLGIRPCDISR Q LRVSHGCVSKILARY patients need to be studied to determine PAXl PATIENT ELAQLGIRPCDISR H LRVSHGCVSKILARY whether the increased frequencv of the Τ allele in our patients has relevance for non- blgure 5 Partial protein sequence of different human and minine PAX getter shin tug that svndromal NTD the gluiamitic (Q> residue at this position in the paired domain of the PAXl gc ne h higlih eonserted The patient carruiif; the PAXl mutation Ішч a him line (H) nudile at this In this studv we have investigated whether position particular members of the PAX gene familv Based on data from the Splotch mouse and could plav a role in the aetiologv of human from the analysis of families and patients with NTD No indications for an involvement of Waardenburg s>ndrome, the PAX3 gene is a PAX7 and PAX9 have been obtained so far No likely candidate for neural tube defects Link­ significant association was detected between age analysis of multiple case families with the PAX3 and non-svndromal NT D On the other hand, when considering the increased fre­ dinucleotide repeat marker ' immediatelv up­ 1 stream of the PAX3 gene did not provide anv quencv of NID in WS1 families, PAX3 indication for a mj|or involvement of this mutations do seem to predispose to certain gene ' However, situations of negative linkage syndromic forms of NTD Our present results but positive allelic association have been argue for a role of the PAXl gene in the aetiol­ reported for other genes and disorders, such as ogv of NTD as shown bv the detection of a the TGrOt gene m cleft Up and cleft palate missense mutation in the paired domain At With respect to NTD, no significant associ- the same time our data show that the detected mutation in PAXl is not sufficient to cause the development of the disorder Other factors, environmental or genetic or both, would have Turns ^ У\ ν ^ to exert a negative influence on the neurulation process to increase the risk further brora animal studies the gene for PDGFRa underlv- ing the Patch phenotvpc is a suspected Alpha helices —' \ candidate

We ill ml. the Dutch Working Croup on H\droceph-tlus and Beta sheets / \ / 4 Spina bVi i ι ol the ρ и ent organisation liOSk lor their hulp in eon'aeling the píllenle nid 1 mulles We also thank Prol Dr H H Ropers and Protcssor I Strichan tor hulplui discussions and S PAXl Wild type LPNAIHLRI YELAOLGIRPCDISRO.LHVSHGCVSKIIARYNETGSI \αη der Velde Visser and I \ m Rossun Boenders lor eell cuilure md t-BV translormations I lus stuch w is supported h\ t the Dutch Prinses Beatrix fonds grint \o <Н 00} and 'И 1Ъ21 I he cooperation of the lamilles and the stufi ol spina bihda clinics in the ISA is grjtetulh acknowledged I he support of\]l I grants R24 SS248Q1 S( and IK,011008 Л SI ind the \lireh ol Dimes Birth Defects houndanon VVJ SC is ЛХ. ^^. tiratcljlh acknowledged We liso think Q Li С Tongian Л Sarangi and 1 S Sunroos lor technical assistance HI RB BH and Г VI are members ol I\ I ï Cil R the Internant nal \eurll lube Lmhruiloev Genetics tnd I pidcnuolotrv Rese trch Consortili TI to idertlh genes Which predispose tO ПеиГЗІ lube Alpha helices delects

1 llgcr Romani Dc-iehl \ lakoh S .ι il I et il growth milor mallorm liions ind minor anomalies in infants born to Women re-ecwing e ilproic acid 7 Pelili 14£θ 108 907 100 1 2 MRs \ itlmin Studc Rese ireh Croup Pretention ol neural LPNAIRLRIVELAQLGIRPCDISRHLRVSHGCVSKILARYNETGSI tube defects results ol the Vlcdicil Research с ouncil t Vititnin Stud\ I oi.et 1ЧО| }1S Щ " ΐ Hol I V Gsurds VIP\ ícnssinO ./ ι I \clusion mapping Ol the gctlc tor \ linked neural t ibe detects in an Icelandic figure 6 RLSUIÌS Of the teeOlldar\ proleltl itrilLtltrc prediction Of tile Z*lld t\pe (A) Olid ttmih Him („.и ι mo ι 91 ι-,: η mutant (B) PAXl peptide, ntpnttich С omputer anals^i·,, nunc the ( hou 1 aunan 1 Sirachin I Re id -VP PAX genes ( in Орт Осп, ί\ί algorithm, shozes the loss of α β sheet conformation at the position of the initiation I'M I 4 U" "SK

77 5 Helwig l Imai К Sehmahl W ι j! Inii.nii.tum Kwnn 20 CstiKin» JP Brook FA Copp AJ Interaction bttwecn iinJiilat l and Pai h каоЧ κ» an ιχιηπκ ft rm 1 spina plihi Spi and Lt LÌ al Absence of linkagi i\pi 1 in humans 7 \LJO и ι 1402 29 Ui ->1 bitween familial mural lubi defects and PAX3 gent J 7 Hailing R Deutsch L Gruss Ρ ni ini tu t a imu inon ìlliet \l /CF it 19QT 32 200 4 ing the development ol thi IIKUSL skt.lt.ion has a point 21 Manman Ft M Hamil ВС J Six ranos o| afftcted and mutation in tht paind h >x ol Pax l < ПЧЬНЪэЗІ •> transmuting mimbt rs ol multiple tase familiis i\ ith neural θ WallinJ WiltmgJ koseki M I riisth R Christ В Balling R tubLdikiis 7 lícJí«.«tf ]092 2» оОэ S Thi rok ol PJ\ 1 in IM il skeleton dt.itiopmt.ni ƒ) / p- 21 Mill Lr SA DvktsDD Pokski HI A simple salting out pro niuit 1044 120 1 UH) 21 9 Di Saxi M krombirg ICTR Jenkins Ί Waardenburg ci dun ftir ixiracting DNA trom human nucleated cells sindrome in South Africa Part I An ¿valúan in of the SniLitAudiltLt 19HH 16 1215 clinical findings in 11 families S í/r \Uä J 198 ( 66 Ъб 24 Iassabih]i M Wiimn \I Ltserton К tí al PAX3 gene 61 strutturi and mutations close analogies Ixmun IO NirodSA Sugli RiriLlt J Hnflmun HJ С спІчНаг miare Waardenburg svndrome and tht Sphti.fi mouse Hum tum in a patient with Waaidenhurg sindrome -im J \LJ lío/« Gttft-f 1994 3 1069 74 (,uiu 1988 31 903 - 2r> Wilcox LR Ri\olta MN Ploplis В PotttrfSB F-ex J Tht Π Da Silva CO Waardtnburg 1 sindromi a clinica! and PAX3 gene is mapptd to human thromosomt 2 together genttit studi oftwolargi Вплііап kmdrids and literature nidi a highli informativi С A d mu ι koude repeat Hum mim Am 7 Wi i э turiinthrtt isolated human gtnts hKiHOJiVW 8 11НЭ 59 26 3 5 90 17 Balduin С Г Hoth С Г Amos JA da Sili a FO Milunski A 29 Tassabihji M Riad AP Ntw-ton Vb et ai Waardenburg s An exonii mutation in thi HuP2 paind ili main gem ssndromi patients haie mutations in the human homo tausts Wair knburgs smdrt nu Satini. 1902 355 637 H logue of the Ран 3 paired box gene \atun. 1992 18 H« ili íl· Milunski A 1 ipski \ Shiifir R Clarren SK. 355 635 6 Balduin ( 7 Mutations in the paired domain of the 30 Macina R Barr F Galili S Reithman H Genomic organi human PA\3 pene (.ausi kkin Waardenburg sindnmi zation of the human PAX3 gene DNA sequence análisis W S III as well as Waardinburg sindromi tipi I (VX S I of the region disrupted in aliLolar rhabdomjosarcom ι •im J Hum buut 1993 52 455 62 GiwwtJkt 1995 26 1 8 19 Hoi hA Hamil B( J Cnurds MPA ι al A frame shift 31 Stapleton Ρ Weith A Lrbanek Ρ Kozmik 7 Busslingtr M mutali η in thi gim lor PA\1 in a girl with spina binda Chromosomal localization of sum PAX gtnts and and mild svmptorrs of Waardenburg sindromi J Wirf cloning of a novel lamils member PAX9 Saturi Guut Guiu 199j 32 э2 6 1993 3 292 7

78 Chapter 6

Altered Regulation of PDGFRa Transcription in vitro by Spina Bifida Associated Mutant Paxl Proteins

Joosten PHLJ, Hoi FA, van Beersum SEC, Peters H, Hamel BCJ Afink GB, van Zoelen EJ, Manman ECM

Proc Natl Acad Sci USA 95:14459-14463 (1998)

79

Altered regulation of platelet-derived growth factor receptor-a gene-transcription in vitro by spina bifida-associated mutant Paxl proteins (neural (übe defects/embryonal carcinoma/osteosarcoma)

PAUI H L. J JoosTbN*t, FRANS A. Hoit, SYIVIA E С. VAN BEERSUMÍ, НЫКО PFTFRS§, BFN С J. HAMEL* , Gijs В. ΑΓΙΝΚ', EVERARDUS J J. VAN ZOELEN*, AND EDWIN С M MARIMANÍ 'Department of Cell Biology University of Nijmegen I oernooivcld 1, 6S2S Ы) Nijmegen The Netherlands ^Institute Mammalian Cicneltcs 05Γ National Research. Center for Emironmcnt and Health HS7o4 Neuherberg Germanv ^Department of Hunan Genetics Uni\crsil\ Ночрпаі Nijmegen PO Box 4101 6S00 HB Nijmegen The Netherlands and'Department of Pathology University Hospital Uppsala University S 751KS I ppsala Sweden

Edited by N M Le Douarin. College de Frunce. Nogent-sur-Marne Cedex France and approved Juh 27, 1998 (received for reiten V/rn //, I99HÌ

ABSTRACT Mouse models show thai congenital neural tion in the paired domain of the Paxl gene, which results in a tube defect·; (NTDs) can occur as a result of mutations in the protein with altered DNA-bindmg affinity and transcriptional platelet-derived growth factor receptor-α gene (PDGFRa). activity (5, 6) Analysis of the un phenotype has shown that Mice heterozygous for the PDGFRa-mutation Patch, and at Paxl is essential for normal vertebral development (7, 8) the same time homozygous for the undulated mutation in the Although these mice do not display spina bifida (sb), a high Paxl gene, exhibit a high incidence of lumbar spina bifida incidence of lumbar sb was observed in double mutants occulta, suggesting a functional relation between PDGFRa resulting from a cross between un and Patch (Pit) indicating and Paxl. Using the human PDGFRa promoter linked to a that Paxl can be involved in NTD aetiology (У) Moreover, we luciferase reporter, we show in the present paper that Paxl recently identified a missense mutation in the paired domain acts as a transcriptional activator of the PDGFRa gene in of the human PAXI gene in a patient with sb, suggesting that differentiated Tera-2 human embryonal carcinoma cells. Two mutations in this gene may also play a role in human NID ( 10) mutant Paxl proteins carrying either the tmdulated-mulation The PDGFRa gene is located within a 50-400-kb region that or the Gin —> His mutation previously identified by us in the has been deleted in the mouse mutant Patch (11) lleterozy PAX1 gene of a patient with spina bifida, were not or less gotcs. are characterized by patches of white fur and have an effective, respectively. Surprisingly, Paxl mutant proteins intact axial skeleton, whereas homozygous embryos display appear to have opposing transcriptional activities in undif­ ferentiated Tera-2 cells as well as in the U-2 OS osteosarcoma occult sb involving the entire spinal column and die during cell line. In these cells, the mutant Paxl proteins enhance early embryogencsis (12) Targeted inactivation of the murine PDGFRa-promoter activity whereas the wild-type protein PDGFRa gene results in mice with sb at the thoracic level ( П) does not. The apparent up-regulation of PDGFRa expression Recently, we have cloned and characterized the human in these cells clearly demonstrates a gain-of-function phenom­ PDGFRa promoter and shown that the 5'-flanking region enon associated with mutations in Pax genes. The altered together with the noncoding exon-1 acts as a functional transcriptional activation properties correlate with altered promoter for the fi 4-kb tuli length receptor transcript (14, 15) protein-DNA interaction in band-shift assays. Our data pro­ The observed NTD-phenotype of the double-mutant mice vide additional evidence that mutations in Paxl can act as a with the (un/un, Ph/ + ) genotype, suggests an interaction risk factor for NTDs and suggest that the PDGFRa gene is a between Paxl and PDGI Ra This suggestion is supported by direct target of Paxl. In addition, the results support the the observation thai PDGFRa-cxprcssion overlaps with Paxl hypothesis that deregulated PDGFRa expression may be expression during embryonic development, particularly in the causally related to NTDs. sclerotome, which forms the vertebral column (9) PDGrRa- mutatcd mouse strains show no effect on Paxl-cxprcssion in the segmented paraxial mesoderm (13) Therefore, Paxl may Neural tube defects (NTDs) form a major group of congenital act as an upstream regulator of the PDGFRa Consequently, malformations with an average incidence of =Ί per 1,000 both Paxl and PDGI Ra are considered to be candidate genes pregnancies NTDs are generally accepted to represent mul­ for sb tifactorial traits with genetic and environmental factors con­ tributing to their aetiology To date, very little is known about Chalepakis et al (6) have shown Paxl to bind to a specific the identity of the genetic factors involved, although several DNA sequence recognition sequence 4 (RS4) In the present candidate genes have emerged from studies of mutant mouse study, we show that the PAXI mutation previously identified strains which display NTD or related malformations (1) in a patient with sb, affects the RS4-binding properties of the protein In addition, we demonstrate that wild-type (wt) and Paxl is a member of the Pax gene family of developmental mutated Paxl proteins have different effects on PDGFRa control genes, which encode transcription factors that contain transcription in Tera-2 embryonal carcinoma cells in a differ­ a DNA-binding "paired domain" (2) The gene is highly entiation-dependent manner and in the U-2 (OS) osteosar­ conserved between species and there is a 100% conservation between the paired domains of murine Paxl and its human coma cell line The data obtained are discussed within the counterpart PAXI at the ammo acid level (3, 4) The mouse hypothesis that deregulated PDGI-Ra transcription may be mutant undulated (un) represents a recessive missensc muta- causally related to NTD, including sb

The publication cosls of this article were defrayed tn part bv page charge This paper was submitted directlv (Track II) lo Ihe Proceedings office Abbreviations NTDs, neural tube defects PDGFRa. platclcl-denvcd payment This amele must therefore be hereby marked advertisement in growth factor receptor α ГС, embryonal carcinoma. RA, retinóte accordance with 18 U S С §1734 solely lo indicate this fact acid wt, wild type un. undulated sb spina bifida. Patch Ph © 1948 tivThe National Academy of Sciences 0Щ7 842Ί '98 '9514459 5$2 00/0 +To whom reprint requests should be addressed e-mail Ρ Joosicn(a PNAS is available online at www pnas org sci kun ni

81 MATERIALS AND METHODS for 7 days before further use U-2 OS cells were seeded at high density (5 0 X 104/cm2) and grown in DMEM/nutncnt mix Paxl Constructs. Expression constructs containing the full F12 (1 1) with supplements and conditions as mentioned length murine Paxl cDNA as well as the full-length Paxl above undulated construct (6) were kindly provided by R Balling Transfection, Luciferase, and ß-Galactosidase Assays. The (Munich. Germany) To generate the sb-Pa\l mutant, the Gin -944/ +118 human PDGrRa-promoler construct (14, 15) at position 42 of the paired domain was replaced by a His and expression vectors containing no insert, wt-Paxl-, sb- corresponding to the mutation previously found in a patient Paxl-, and un-Paxl-encoding inserts, respectively, were tran­ w ith sb (9) Mutagenesis experiments were carried out by using siently Iransfected into Tera-2 cells or U-2 OS cells by using the the Altered Sites System kit (Promcga) according to the calcium phosphate coprccipitation method (19) Luciferase manufacturers protocol by using oligonucleotide primer 5'- activity was detected 48 h after transfection (Luciferase assay GAGACCCGCAGGTGCCTACTGAT-3 Presence of the kit, Promcga) The luciferase activity was corrected for trans­ mutation (bold) was confirmed by dyedeoxy termination cycle fection efficiency by measuring the /3-galactosidase activity due sequencing (ABI) of the constructs on an ΑΒΠ70Α automated to a cotransfccled cytomegalovirus promoter-driven lacZ gene sequencer construct Transfections were done in duplicate with different In Varo Transcription/Translation. The synthesis of wt batches of DNA Paxl, sb-Paxl, and un-Paxl proteins was performed by using the 1 η r-T7-Couplcd Promcga Rabbit Reticulocyte Lysate System (Promcga) according to the manufacturer's protocol For this purpose a T7-promoter was introduced by PCR on RESULTS Paxl expression constructs which encode Paxl (wt-Paxl), Production of wt and Mutant Paxl. Го test the functionality un-Paxl, or sb-denved Paxl (sb-Paxl), respectively Forward of the different Paxl proteins, lull-length murine Paxl proteins primer (including T7 promoter and start codon) V-CGCTA- were generated in an m vitro transcription/translation system ATACGACTCACTATAGGAACAGACCACCATGGAGC- In addition to the wt-Paxl protein (wt-Paxl), two NTD- AGACGTACCGAAGTGAAC-V and reverse primer 5 -G- associatcd mutant proteins were synthesized also The first one GCTGTGGCTCTGTGAGAG-3' (located in 3' untranslated (sb-Paxl) contained the Gin -» His substitution that was region, ref 6) were used to generate the wt-Paxl-, un-Paxl-, previously identified by us at position 42 of the paired domain and sb-Paxl-encoding templates These templates were used in a patient with sb (10) The second mutant protein (un-Paxl) subsequentlv for m \iiro transcription/translation contains the Gly —> Ser replacement at position 15 of the Generation uf Paxl-Specific Antibodies. Paxl-specific poly­ paired domain that is known to be responsible for the un clonal antibodies were obtained by immunizing two rabbits phenotvpe in mice (5) with a Paxl-specific peptide coupled to malcimidc-activated To show that full-length proteins are synthesized, we per­ kcyholc-limpet-hemocyanin (Pierce) according lo the manu­ formed a Western blot analysis For this purpose polyclonal facturer s instructions and established protocols (16) The antibodies were generated recognizing the C-terminal part of Paxl-specific peptide (NH2)-FKHREGTDRKPPSPG- Paxl which is not affected in the mutated proteins These (COOH) corresponds to amino acid number 127-141 of the antibodies detect a protein of the predicted size of 40 kDa in Pax 1 -protein (ft) A total of 1 mg of coupled protein was mixed protein extracts of the developing vertebral column of 13 5- with incomplete Freund s adjuvans and used for initial immu­ day-old mouse embryos (Fig 1) In contrast, no Paxl protein nization Booster injections with 100 μg of coupled protein w as detectable in the extract of the [/ns mouse embryo in which were given in 4-week intervals 1 otal serum was collected 11 the Paxl gene is deleted (7), showing that the antibodies are days after the fourth boost, designated as 289-IV, and used in Paxl-specific By using these antibodies, the three in vitro- these studies generated Paxl proteins arc delectable at the same level as the Western Blotting. Protein extracts were separated on a I0"r wt-Paxl protein (Fig 1), thereby demonstrating that full- SDS-polyaciylamide gel and subsequently transferred to a length proteins have been synthesized nitro-cellulose membrane, which then was prctrcatcd for 20 Binding Properties of wt and Mutant Paxl. To study the min with blocking buffer (0 49r gelatine/i^ BSA/10 mM Tris, DNA-binding ability of the different Pax proteins, electro- pH 7 5/150 mM NaCI) The membrane is incubated subse­ phorctic mobility shift assays were carried out by using the quently with the Paxl antibody (diluted 1 200 in RIA buffer PRS4 consensus-binding sequence for Paxl (6) Fig 2 shows 10 mM Tris, pH 7 5/160 mM NaCI/Ki Triton/0 І7г SDS/ that with wt-Paxl the formation of a specific DNA-protein 0 5^ sodiumdcsoxycholatc) for 1 hr at room temperature and complex causes a clearly detectable band-shift Using the treated with goat anti rabbit alkaline phosphatase conjugate antibody, Paxl was identified within this complex (not shown) for I hr at room temperature (Bio-Rad, dilution 1 5,000 in In contrast, no bandshift was detected with un-Paxl, which is RIA buffer) The membrane was then treated with A PB buffer in accordance with the earlier work of Chalepakis et al (6) In (100 mM Tris pH 9 5, 100 mM NaCI, 5 mM MgCl2) and this context one should realize, however, that un-Paxl was subsequently stained by using bromochloromdolylphosphate shown to be able to bind to other sequences Remarkably, and nitrobluctelrazolin as substrates sb-Paxl induces a shift but with a different mobility from that Elect rophorctic Mobility Shirt Assay. In these assays, a induced by wt-Paxl Because both proteins have an identical 12P-3' end-labeled PRS4 oligonucleotide was used as described molecular size (Fig 1). sb-Paxl apparently binds to the DNA- (6) (5 •-TGGGCTCACCGTTCCGCTCTAGATATCTCGA- sequence in a slightly different conformation compared with 3 ) The assays were performed essentially as described (17, wt-Paxl These altered binding characteristics may have con­ 18) sequences for the proper functioning of the Paxl transcription Cell Culture. Tera-2 embryonal carcinoma (EC) cells factor (Clone 13) were grown in a modification of minimal essential Effect of wt and Mutant Paxl on PDGFRor Transcription. medium lacking nucleosides and deoxynucleosides, supple­ As already mentioned, the phenotype of (un/un Ph/ + ) mented with W7t (vol/vol) fetal calf scrum and 44 mM double mutants suggests that Paxl and PDGFRa participate

NallCO, in a 7 57t C02 atmosphere at 37°C Undifferentiated in the same functional cascade One possibility is that Paxl Tera-2 EC cells were seeded at high density (5 0 X 104/cm2), ads as an upstream regulator of PDGFRa To assess the and differentiation of cells was induced bv the addition of effect of the various Paxl proteins on the transcriptional retinole acid (RA) (5 μΜ) 12 h after the cells were seeded at activity of the PDGFRa promoter, plasmids encoding these low density (5 0 X M'/cm2), and maintained in this medium proteins were cotransfccled with the -944/ +118 human

82 pc ne wt sb un

49.5 —

32.5 —

Flu. 1. In vitro transcription and translation of Paxl and derived mutants. The in vitra transcription/translation products of wt-Paxl (wt), sb-dcrivcd-Paxl (sb). and un-Paxl (un) were detected (in Western blot by a spécifie anti-Paxl antibody. As a positive (pc) and negative control (nc) total protein extract from the developing vertebral column (if 1 3..S-day-old wt and a homozygous undulated shorttail (i/ns) embryos were used. respectively. In Wls mutant mice, the Paxl gene has been deleted completely. The antibody detects a predicted band of = 40 kDa. which is not present in the negative control. No Paxl was detected in the transcription/translation lysate (not shown).

PDGFRa promoter linked to the luciferase gene. Transfec- efficiently promote I'Diil Ra transcription in Tcra-2 RA tion studies were carried out in Tera-2 cells. In their undif­ cells. Interestingly, in Tcra-2 EC cells a significant increase ferentiated state, these HC cells resemble cells in early st ages in activity of the PDGFRa promoter was observed with of human development. Only after ш vitro differentiation by sb-Paxl whereas only little effect was seen with un-Paxl (Fig. RA these cells express a functional PDGFR« (20). No 3B). Fig. 3C shows that in U-2 OS sb-Paxl and un-Pax both endogenous expression of Paxl was observed in both Tera-2 induce promoter activity, the latter being the most potent EC and Tera-2 RA cells by reverse transcription—PC" R activator under these conditions, whereas wt-Pax is again analysis (not shown). Because Paxl is one of the earliest ineffective. These results show that wt and mutated Paxl markers for sclerotome differentiation, additional transfec- proteins can have opposing transcriptional activities, de­ tion experiments were carried out in the U-2 OS osteosar­ pending on the specific transcriptional background in dif­ coma cell line. These cells have endogenous PDGFRa- ferent cell types. Moreover, it shows that mutations in Paxl cxpression (21). Fig. .VI shows that wt-Paxl induces up to can result in a gain-of-function with respect to transcrip­ threefold stimulation of promoter activity in Tera-2 RA tional activity of the protein. cells. Un-Paxl is ineffective in increasing promoter activity in these cells, whereas sb-Paxl has intermediate activity. These results show that an intact Paxl protein is essential to DISCUSSION

Ly un wt sb SC О In the present study, we have shown that the Gin —» His mutation in ΡΛΧ1, previously identified by us in a patient with Slots sb, is associated with altered in vitro DNA-binding properties of the full length protein. Our results suggest that this mutation * affects the functionality of the PAX-1 protein and. therefore. may have contributed to the phenotype observed in the patient. In addition, we investigated the influence of Paxl on PDCiFRu transcription. We showed that wt-Paxl protein enhances PDGFRa transcription in differentiated Tcra-2 hu­ * man embryonal carcinoma cells (Fig. 3.4). whereas two mutant Paxl proteins carrying either the undulated mutation (un- Paxl ) or the Gin —> His mutation (sb-Paxl). were not or less effective, respectively. Surprisingly. Paxl mutant proteins ap­ pear to have opposing transcriptional activities, depending on ·* the cell differentiation status in the Tcra-2 cells, as well as in f ' н f ψ ι the U-2 OS osteosarcoma cell line. (Fig. 3 В and C). These activities of the mutated Pax-1 proteins are the first examples Fio. 2. Electrophoretic mobility shift assays with Paxl and derived of mutations in a Pax gene that can be associated with a mutants. In vitro transcription/translation products of wt-Paxl, sb- gain-of-function phenomenon. This could have yet unknown derived Paxl. and un-Paxl were assayed for binding activity on the consequences during embryonic development and may play a PRS4 probe derived from the e5 site in the Dmsophila even-skipped role in the pathogenic mechanism underlying specific types of promoter. The lanes Ly. un. wt. sb. SC. and 0. correspond to incubation NTD. of the probe with reticulocyte lysate. un-Paxl. wt-Paxl. sb-Paxl. and wt-Paxl + unlabeled PRS4 probe in a 500-fold excess, and H^O. The above observations support the hypothesis that Paxl is respectively. The Paxl bands arc indicated with arrows. Nonspecific an upstream regulator of PDGFRn and that deregulated bands are indicated with asterisks. PDGFR« gene transcription may be causally related to sb. In

S3 4.0 3 5 3.0 0 J 2 5 ü л I 2 0 И 1 5 ¿, •tí OIO . JL ' •M X 0 S . / 0.0 _ ib JJt X Pl -»-wfc +eb +un Pl +wt +áb +un

HG 1 Regulation of transcriptional activity of Ihc POOF'Rn promoter by Pax-1 and derived mutants The activity of the 944/-*-118 promoter lucilcrasc construct (Pl) is regulated differently by wt-Paxl. sb-PaxI, and un Paxl in RA differentiated Tcra-2 RA cells (A, hatched bars), undifferentiated Tcra 2 ЬС tells (B open bars), and U-2 OS (<". black bars) Values arc presented as mean promoter activity relative to the activity of the -944/-* 118 clone cotransfectcd with the expression construct without a Paxl insert which was set to 1 Error bars indicate the sample SD after four lucifcrase//3-galactosidasc measurements this way, it can be explained that sb is not only observed in the gain-of-function of the protein The different activities of the targeted PDGFRa null mutation (13) and in homozygous wt and mutant Paxl proteins correlate with an altered con­ Patch (Ph/Ph) mice (12) but also in Patch (Ph/ + ) heterozy­ formation of the corresponding protein-DNA complex in the gotes with an additional Paxl mutation (un/un) (9) Accord­ bandshift assays The present data corroborate our earlier ingly the sb phenotype in the patient carrying the sb-Paxl hypothesis that mutations in PAXl can act as risk factors in mutation may be the result of impaired regulation of PDGFRa NTD but also present the PDGFRa gene as the central gene transcription In this context, the gain-of-function phenome­ involved Additional studies of this gene have to be performed non could be very important for effectively deregulating to investigate whether mutations in PDGhRa contribute to PDGFRa-expression and thus could contribute as well to the human NID, particularly to sb sb phenotvpe It is well realized, however, that NTDs including sb are a multifactorial trait and that other genes and/or We thank Necltje Aris and Jose Hendriks for technical assistance environmental influences are likely to be involved F И and Ε M arc members of ÌNTTGCR the International Neural We have previously shown that the human PDGFRa gene Tube Embryology Genetics and Epidemiology Research consortium is able to give rise to multiple mRNA products as a conse­ to identify genes that predispose to NTDs This study was supported quence of alternative promoter use and alternative splicing by Ihe Dutch Prinses Beatrix Fonds Grants % 0212 and 97-0107 (18, 22) In Iera-2 EC cells, a 1 S-kb and a 5 O-kb transcript have been identified, which are transcribed from a promoter in Harris, M J & Junloff, J M (1997) Tautology 56, 177-187 intron 12 of the gene (18) In contrast, in Tera-2 RA cells a Strachan Τ & Read A Ρ (1994) Curr Opm Cene! Dei 4, 6 4-kb and a 3 O-kb transcript are present, for which transcrip­ 427 418 tion is driven by the promoter upstream of exon 1 Because the Burn M Tromvoukis, Y Bopp D, Frigcno, G & Noll M (1989) EMBOJ *, 1181-1190 ή 4-kb transcript generates the functional, full length en-recep­ Schnitlgcr S , Rao, V V , Deutsch, U , Gruss Ρ, Balling, R & tor, this promoter has been used tor the current activation Hansmann I (1992) Ce noma s 14, 740-744 studies with Paxl The -994/+118 region of the human Balling, R Deutsch, U & Gruss, Ρ (1988) Cell 55, 511-535 PDGFRa promoter contains the larger part of the information Chalcpakis G bntsch, R Fickenschcr, H Deutsch, IJ, Goul- required for tissue-specific regulation in transgenic mice, as ding, M & Gruss, Ρ (1991) Cell 66, 871-884 previously shown by us (15.) This promoter fragment contains Wallin J Wilting J Koscki, Η , Fritsch R Christ, В & Balling, multiple putative Paxl-binding site core elements (5'-GTTCC- R (1994) Development (Cambridge, UK.) 120, 1109-1121 3'. rcf 6) Balling R Ncubuser, A & Christ, В (19%) Semin. Cell De\ Additional studies will have to indicate whether Paxl actu­ Biol 7, 129-136 ally binds to the PDGf Ra promoter directly, or rather medi­ Helwig, U , Imai К Schmahl, W , Thomas, В E.Varnum D S, ates its effect via interaction with other components of the Nadcau J Η & Balling, R (1995) Nat Genet 11, 60-63 transcription machinery In the case of direct binding of the 10 Hoi, F A Geurds, Μ Ρ A , Chatkupl, S. Shugart, Y Y . proteins to the PDGFRa promoter, the different activities Balling R Schrander-Stumpcl С Τ R M , Johnson, W G , could be explained hy an altered cofactor dependency of the Hamel В С J & Manman. E С M (1996)7 Med Genet 8, mutated proteins In the case of an indirect influence of Paxl 655-660 proteins on PDGFRa-promoter activity, at least tor undu­ Stephenson, D A , Mercola, M Anderson, fc , Wang. С Y , Stiles, С D, Bowen-Popc, D r & Chapman V M (1991) Лис lated, it also might be possible that other genes arc activated Natl Acad Su USA 88, 6-10 because of its altered DNA-bindmg specificity (6). 12 Pasnc, J Slubasaki F & Mercóla M (1997) Dev Dyn 209, In summary, the present results emphasize the relevance of 105-116 PDGFRa expression regulation during vertebral column de­ 13 Sonano Ρ (1997) Development (Cambridge, UH) 124, 2691- velopment and support the model that impaired expression of 27(H) this gene may be causallv related to sb Evidence from the Afink (, В Nistcr, M , Stassen, В H G J.Joosten, Ρ H L J, transfection studies suggests that Paxl is indeed an upstream Rademakers Ρ J H, Bongcam Rudioff, fc & Van Zoelen, regulator of PDGFRa, and mutations in Paxl may result in a fc J J (1995) Oncogene 10, 1667-1672

84 15 Zhang, X О, Afink, G В, Svcnsson, К. Jacobs, J J L, 19 Sambrook J rntsh, E F & Manutis Τ (1989) Molecular Guenthcr, Τ, Forsberg-Nilsson, К, Van Zoelcn, F J J , Wesl- Clonmfî A t aboraton Manual (Cold Spring Harbor Lab Press crmark, В & Nislér, M (1998) Mec /ι Dei 70,167-180 Plainview, NY) 2nd Ed 16 Harlow, E & Lane, D (1988) Antibodies A laboratory Manual 20 Mosselman. S, Claesson-Wclsh, L, Kamphuis, J S & van (Cold Spring Harbor Lab Press Plainview. NY) Zoclen. Ь J J (1994) Cancer Res 54,220-225 17 Schreiber, E, Matthias, Ρ. Mueller, M M & Schaffner, W 21 Bctsholtz С Westermark, В. Ek. В & Heldin С -H (1984) Celt (1989) Nucleic Acids Res 17,6419 39, 447-457 18 Kraft, H J , Mosselman, S , Smits, Η Α, Hohcnstcin, Ρ Piek, 22 Mosselman, S , Looijenga L Η J , Gillis ALM, van Rooijcn, Ε , Chen, Q , Artel. Κ &vanZoclcn,E J J (1996)./Bio/ Chem M A, Kraft, H J. van Zoelcn E J J & Oostcrhuis J W (1996) 22, 12873-12878 Proc Natl Acad Su ISA 93,2884-2888

85

Chapter 7

Molecular Genetic Analysis of the Gene Encoding the Trifunctional Enzyme MTHFD (methylenetetrahydrofolate-dehydrogenase, methenyltetrahydrofolate-cyclohydrolase, formyltetrahydrofolate synthetase) in Patients with Neural Tube Defects

Hoi FA, van der Put NMJ, Geurds MP A, Heil SG, Hamel BCJ, Mariman ECM, Blom HJ.

Clin Genet 53:119-125 (1998)

87

Molecular genetic analysis of the gene encoding the trifunctional enzyme MTHFD (methylenetetrahydrofolate-dehydrogenase, methenyltetrahydrofolate-cyclohydrolase, formyltetrahydrofolate synthetase) in patients with neural tube defects

Hoi ΓΑ, van der Put NMJ Geurds MPA Heil SG, Tnjbels P-JM Frans A Hoi", Nathalie MJ van Hamel BCJ Manman ЬСМ Blom HJ Molecular genetic analysis of der Put6, Monique PA Geurds3, the gene encoding the trifunctional enzyme MTHH) (methyleneletrahy- Sandra G Heil", Frans drofolate-dehydrogenase, mcthcnyltetrahydrofolate-cyclohvdrolase JM Trubels", Ben CJ Hamel', formyltctrahydrololate-synlhetase) in patients with neural tube defects 9 1 Edwin CM Manman and Henk Clin Genet 1998 52 119 I2 ) © Munksgaard, 1998 J Blom" Depar*ment o( Human Genptirs It is now well recognized that penconceptional folic acid or folic acid Department oí paedialr es Univesily containing multivitamin supplementation reduces the risk of neural Hospital Ni|megen Nijmegen The tube defects (NTDs) Recently we were able to show that homozygosity Ne'nerands tor a thcrmolabile variant of the enzyme methylenetetrahydroiolatc re­ ductase is associated with an increased risk for spina binda in patients recruited from the Dutch population However this genetic risk factor Key words Iole aed - could not account for all folic acid preventable NTDs In an attempt me'hyleneletrahydro'olato dehydrogenase - MTHFD neural lube defp^ts - spina to identify additional folate related enzymes that contribute to NTD Dilida etiology we now studied the melhylenetetrahydrofolate dehydrogenase gene on chromosome 14q24 which encodes a single protein with three Correspond ng author FA Hol Univers ty catalytic properties important in the folate metabolism The cDNA se Hospital N.jmeger Department of Hu-man quence of 48 familial and 79 sporadic patients was screened for the Genetics PO Box 9101 6500 MB Ni,™ presence of mutations by single strand conformation polymorphism gen The Netherlands Fax - 31 24 (SSCP) analysis followed by sequencing Two amino acid substitutions 3540488 e mail f hol®an'rg azn ri were identified The first one (R293H) was detected in a patient with familial spina bifida and not in 400 control individuals The mutation Rece ved ir- revised versior 3 Noverrber was inherited from the unaffected maternal grandmother and was also 1997 accepted for pubi cation / Noven present in two younger brothers of the index patient, one of them ber 1997 displaying spina bifida occulta and the other being unaffected The second change turned out to be an amino acid polymorphism (R6^1Q) that was present in both patients and controls with similar frequencies Our results so far provide no evidence for a major role of the mcthvlenctetrahydrofolate-dehydrogenase (M THFD) gene in NTD eti­ ology However, the identification of a mutation in one family suggests that this gene can act as a risk factor for human NTD

Neural tube defects (NTDs) in humans, which mental factors Among the many possible include severe malformations such as spina bifida contributory factors in human N TDs. maternal and anencephaly, make up a significant proportion vitamin status has been implicated as a key factor of today's birth defects The average incidence of Some years ago it was convincingly shown that NTD is approximately 1-2 per 1000 but varies folic acid or folic acid containing multivitamin between different countries and socio-economic supplementation in early pregnancy prevents a sig­ groups The aetiology of NTDs is considered to be nificant percentage of NTDs (1. 2) holte acid multifactorial which means that they can be as­ derivatives, in particular the active forms, are im­ cribed to a combination of genetic and environ­ portant in several critical single-carbon transfer

89 Homocysteine Methionine

5-methyltelrahydrofolate Tetrahydrofolate Senn«

iL il Forniate

Glycine MTHFR

5 10-methylenetetrahydrofolate 10-formyltetrahydrofolate

^/[ciï] к 5 10-methenyltetrahydrofolate Fig 1 Simplified representation of Ihc homoc>steine folate metabolism The enzymes shown in this scheme arc boxed MTHFR= 5 10 methylcnelclrahydrofolaic reductase MTHFD 5 10 mcthvlenctctrahydrofolatc dehydrogenase CH 5 10-methcnyltetrahvdro folate c\cloh>drolasc FS 10 formyltetrahydrofolate synthase SHMT serinehydroxylmcthyltransferasc MS methionine synthase MTHFD CH and FS are catalytic properties of a single protein reactions including the biosynthesis of purines and for this observation could be that 5-methvltetrahy- thymidine and the remethylation of homocysteine drofolate, the methyl donor in the remethylation of to methionine Recently, mothers of infants with homocysteine to methionine (Fig 1), is not avail­ NTD were shown to have increased homocysteine able in adequate amounts Therefore, enzymes levels (3, 4) The enzyme 5,10-methylenctetrahy- such as MTHFR, which play a role in the produc­ drofolate reductase (MTHFR) plays an important tion of 5-methyltetrahydrofolate are good candi­ role m folate mediated homocysteine metabolism dates tor NTD etiology Another en/yme involved by reducing 5 10-methylenetetrahydrofolate to 5- in this pathway, the mcthylcnctctrahydrofolate de- methyltetrahydrofolate which acts as the principal h)drogcnase (MTHFD) possesses three enzymatic methyl-group donor in homocysteine remethyla­ properties 5 10-methylenetetrahydrofolate dehy­ tion (big 1) Ьrosst et al (5) identified a common drogenase, 5,10-methenyltetrahydrofolate cyclohy- polymorphism (677C-»T) in the gene encoding drolase and 10-formyltctrahydrofolatc synthetase, this protein Homozygosity for this MTHFR vari­ respectively, and catalyses three sequential reac­ ant is associated with decreased enzyme activity, tions in the interconversion of one-carbon deriva­ elevated mean plasma homocysteine levels, and a tives of tetrahydrofolate which are substrates for redistribution of folates (6 7) We found that ho­ methionine, thymidilate, and de пою purine syn­ mozygosity for this mutation is associated with thesis (11) Here we report on the mutation analy­ spina bifida in the Netherlands thereby identifying sis of the MTHFD cDNA in patients with NTD the first genetic risk tactor for spina bifida in humans (6) Nevertheless, not more than 25% of Material and methods the folic acid preventable NTD cases can be ex­ Clinical material plained by presence of this mutation (8, 9) This clearly indicates that, beside MTHFR other en­ With approval of the local ethics committee pa­ zymes in the folate mediated homocysteine tients with non-syndromic NTD have been col­ metabolism are likely to be involved as well More­ lected from the Dutch population The patient over, analysis ot spina bifida cases and their par­ group consisted of 38 unrelated familial and 79 ents excluding individuals that were homozygous sporadic NTD cases The criteria applied for selec­ lor the 677C-»T mutation in the MTHFR gene, tion of the familial patients have been described in showed decreased mean plasma folate levels and detail elsewhere (12) The types of NTD in the increased levels oí total plasma homocysteine in familial group were spina bifida (n = 36), en- both patients and parents (10) One explanation cephalocele (n = 1) and craniorachischisis (n= 1)

90 Table 1. Primers used for SSCP-analysis of the MTHFD сША (11)

Fragment Size (Ър) Forward primer Reverse primer

1 228 5'-TGGGCAGCGGACTAATAAAG-3' S-GAGTGCCrriGATCCCAATC-S· 2 214 5-TGTGAAGCTGAAGGCTGCTG-3· 5-ATCCACATCCnCTCGGGTG-3· 3 229 S'-CACTGAAGAAGTGATCMTGC-S1 5' -CAGAAGCAAGTCATGCATCG-3' 4 250 5'-GGTTGGGCGCAGTAMATAG-3· 5,-CACAACTTΠΊCTCCCATTTGG•3, 5 250 5'-GGAATCAAnATGTCCCAGATG-3· 5'-TGTCACnGGAACAGGTGTC'3 6 243 5-CCnAACCTCMGACACCTG-3' 5--GTAGnGTGCTTTTCCCnCT-3' 7 232 5'-GAATAACTCCAACACCCCTG-3' 5'-nAGCTGCAGTGATGGCATG-3· 8 225 5-TAATCTCCACCTCACAGGTG-3· 5'-TCATCTGTCAGTGTGGTAGG-3 9 238 S-CTAGGCAnGAAAAGACTGAC-S1 5-GAGAAGTGGTGAGAGCCAG-3' 10 244 5'-CTCTGTGGCCAGTGAMrTATG-3· 5'-ATGTGCGATGπGGCAAACG•3• 11 233 5-CACTCCAGTGTTTGTCCATG-3' 5--GCATCTTGAGAGCCCTGAC-3' 12 245 5·-ΑθΰΤΰΰτοΰτοϋττοποο·3' 5-GGAAAGGCGGCTGATGAGG-3' 13 253 54ATGCATTCAAGACGGATACAG-3' 5'-AAITCMTGTCATCTGCTCC-3' 14 226 5'-ATCAGGATCAnGCACAGAAG-3· 5-ACAGAAAACCAGCCCCAACG-3' 15 198 5'-TGTCCCTACAGGCTTCÄTTC-3' δ'-ΟΤΤϋπΟΜΟΑΤΰΟΑΤυΟΤΟ^1

The sporadic group included cases with spina (Table 1), 10 mM Tris. 50 mM KCl, 1.5-3 mM binda (η = 75), anencephaly (η = 2), and MgCK, 0.01% (w/v) gelatin, 0.2 mM dNTP, 0.1 μΐ encephalocele (n = 2). It should be noted that the [x,2PldCTP (Amersham), and 0.5 U of Taq DNA proportion of anencephalics in the patient group is polymerase (Boehringer, Mannheim). Samples under-represented with respect to the normal were subjected to an initial denaturation at 95°C population. Genomic DNA was extracted from for 5 min followed by 35 cycles of amplification: whole blood according to the procedure of Miller 95°C for 50 s, 53°C for 50 s and 72°C for 90 s. For et al. (13) and lymphoblastoid cell lines from each SSCP analysis the PCR products were run on a 5% patient were established using Epstein-Barr virus. non-denaturing Polyacrylamide gel with 10% glyc­ Unaffected and unrelated subjects were randomly erol and on a similar gel without glycerol at 4°C. recruited from the Dutch population and used as a The gels were dried and exposed overnight to Ko­ control group. dak X-omat films. Direct sequencing of aberrant PCR fragments was performed by dyedcoxy termi­ nation cycle sequencing (ABI) on an ABI370A RNA isolation and cDNA synthesis automated sequencer. Total RNA was isolated from cultured lymphoblasts using the RNAzol™B kit according to the protocol of the manufacturer. Isolated RNA Allele frequencies of the observed aberrant alleles in was subsequently DNasel treated (Life Technolo­ control individuals gies) to eliminate genomic DNA contamination Only genomic DNA was available from control and to preclude amplification of MTHFD pseudo- individuals. In order to find out whether the gene sequences (14). First strand synthesis was G878A (R293H) mutation could also be detected performed in a 20 μΐ volume containing 250 ng in genomic DNA of controls, fragments surround­ total RNA, 100 ng random hexamer primers ing this site were amplified using MTHFD specific (Pharmacia), 10 mM Tris, 50 mM KCl, 5 mM primer fh200 (5-AGAGAGGGCGAGCTTCAT-

MgCl2, 0.01% (w/v) gelatin, 1 mM dNTPs (Perkin C-3') in combination with the fragment 5 reverse Elmer), 50 U RNasin (Pharmacia) and 1 U primer (Table 1). Primer fh200 corresponds to a MMLV-RT (Life Technologies). Incubation was small region of the cDNA that seems to be deleted for 10 min at room temperature, 60 min at 37°C in the pseudogene (personal observations). A and 6 min at 95°C. As a control for DNA contam­ product of roughly 1100 bp was obtained which ination reverse transcriptase (RT) was omitted points to the presence of an intron of around 900 from one sample in each reaction. bp in size. Sequencing of the fragment revealed the intron/exon boundary between position 855 and 856 of the cDNA (nucleotides are numbered from RT-PCR and SSCP analysis the initiation codon). PCR amplification using in­ PCR amplification of the MTHFD fragments was tron primer fh208 (5-TGATTGAAATGGAGTG- performed in a 25 μΐ reaction volume containing ACCTG-3') together with primer fh209 (5'CCTG- 4.5 μΐ cDNA sample, 15 pmol of each primer GCTTAAATTTCTCCAGGACA-3') produces a

91 fh208 ь. 5 ' -tttgtgattgaaatggagtgacctggagcttgaaaatgaggtgaggatta 3 ' -aaacactaactttacctcactggacctcgaacttttactccactcctaat

intron ι exon aaataaccaccttttttaagtttnatttatttnattttgcttagAGCACA tttattggtggaaaaaattcaaantaaataaantaaaacgaatcTCGTGT

λ t GTAGAGAGTGCCAAGCGTTTCCTGGAGAAATTTAAGCCAGGAAAGTGOAT-3 · CATCTCTCACGGTTCGCAAAGGACCTCTTTAAATTCGGTCCTTTCACCTA - 5 · C V fh209

Fm 2 Strategy lh il was followed to confirm Ihe pásente of the G878A (R291H) muUlion in genomic DNA by PC R followed by restriction digestion Primer fh209 contains a mismatch at position 21 which introduces a Malli (5 CATG 1 ) in the product only when the A allele is amplified The \ciIleal arrow marks the position of the mutation The intron exon boundary is indicated by a \citical bar Intron and exon sequences art in lower and uppercase respecuvck fragment of 137 bp (Tig 2) Primer fh209 contains tivities and a C-tcrminal domain corresponding to a mismatch at nucleotide position 23 which intro­ about two-thirds of the original protein in which duces a Nlalll site (5-CATG-3) in the product the ATP-dependent synthetase activity resides (Fig onlv when the A allele is amplified (Fig 2) С onse- 3) In order to look for mutations in this gene we qucntlv presence of the mutation can be tested conducted SSCP analysis using 15 overlapping through Nlalll restriction digestion of the PCR primer sets encompassing the complete coding se­ products quence (Fig 3, Table 1) RT-PCR from total Presence of the G1958A (R6530) substitution in lymphoblastoid RNA yielded products of the ex­ genomic DNA of control individuals was investi­ pected size indicating that the gene is expressed in gated by PCR amplification using the fragment 11 cultured lymphoblasts Two different aberrant primers (Table 1 ) This resulted in a product that bands were detected in the SSCP analysis The first was 98 bp larger in size than one would predict one was found in fragment 5 in a male familial from the cDNA sequence indicating the presence patient with spina bifida Sequencing of the shifted of a small intron Sequencing showed that the band re\ealed that the patient was heterozygous intron was located between position 1996 and 1997 for the nucleotide substitution G878A leading to of the cDNA Presence of the nitron allowed us to the exchange of arginine for histidine at position distinguish between the gene and the pseudogene 294 of the protein (Fig 4A) Presence of the muta­ product which lacks introns The G->A change tion was confirmed in genomic DNA via PCR abolishes an Hpall restriction site (5 -CCGG-3 ) in amplification followed by Nlalll digestion as de­ the DNA and therefore digestion ol amplification scribed in the materials and methods section Test products with Hpall was conducted in order to ing of additional lamily members revealed that the determine the genotype of the subjects (Table 2) mutation was inherited from the healthy mother and maternal grandmother and showed that the mutation was also present in two older brothers of Results the index patient the oldest one displaying spina SSCP analysis of 'he MTHFD cDNA bifida occulta the other one being healthy (Fig 5) It is not known whether presence of this mutation The human Ml HPD gene which encodes a single is reflected in the plasma homocysteine levels of protein with three specific en/ymatic activities that carrier individuals since blood samples from mem­ are important in lolate metabolism was considered bers of this family were no longer available The to be a candidate for human N1 D The MTHhD mutation could not be detected in 300 unrelated cDNA contains an open reading frame of 2805 bp control individuals which encodes a protein of 945 amino acids (11) Proteolysis experiments with mammalian tnfunc The second shifted band was observed for frag­ tional en7vmcs (15) dcmonstialed that the protein ment 11 in several unrelated patients Sequencing can be separated in two functional domains an of the altered band revealed a G1958A substitution N-termina! domain of approximately one third of which causes the replacement of the arginine the native polypeptide which contains the cyclohv residue at position 653 by a glulamine in the de­ drolase and NADP-dependent dehydrogenase ac­ duced protein (Fig 4B) The observed nucleotide

92 Table 2 Prevalence of the MTHFD 1958G-A(-^) polymorphism m the Dutch population

Geno'yre frequencies % (n) Allele frequencies

Familial patients 37 1Γ31 51 4(19) 1 1 4(4) О 63 0 37 Sporadic patients 28 4i'9) 53 7(36) 179(12) О 55 0 45 Cor trois 29 ЭТОТ 51 3(172) 18 8(63) О 56 О 44 substitution disrupts an НраІІ restriction site as the synthethase activity resides in the C-tcrminal which allowed us to confirm the presence of the half of the protein. Hum and MacKen/ie (19) base-change through restriction digestion of the showed via independent expression of the two ma­ PCR fragment (Materials and methods). Using this jor domains in Escherichia coli that the linker-re­ test for the analysis of unrelated controls showed gion, which joins the two domains in the native that the alteration is present in both patients and trifunctional protein, is likely lo be localized be­ controls with similar frequencies (Table 2). No tween residue 292 and 310. The first alteration significant differences in the plasma homocysteine detected in the present study is located within this levels could be observed when comparing patients putative inter-domain region and concerns an homozygous for the G-allele to patients that were argine to histidine substitution at position 293 either homozygous or heterozygous for the A-allele (Л293Н). The fact that this mutation is located (not shown) No further bandshifts could be within this part of the gene makes it less likely that detected in any of the analysed fragments. it has any direct implications for the different enzymatic activities of the protein. On the other hand, the linker-region is likely to Discussion play a role in maintaining the structural integrity Recent reports on hyperhomocysteinemia in of the multifunctional protein which in eukaryotes families with NTD argue for a role of one or more is a horrrodimer. Therefore, mutations in this part folate mediated enzymes in the etiology of folic of the gene still might have consequences for the acid preventable NTD 0. 4) Previously, we found proper functioning of the enzyme. The mutation that homozygosity for the MTHFR 677T variant was only found to be present in certain members of is a risk factor for NTD in the Dutch population one single NTD family. Although both amino (6, 9) Since then, this finding has been corrobo­ acids are chemically similar, thus, representing a rated by similar observations in other populations conservative substitution, the amino acid exchange (16-18) Here we report on two alterations de­ seems to represent a mutation rather than a poly­ tected in the MTHFD gene by SSCP analysis of 38 morphism because the H-allcle could not be de­ unrelated familial and 79 sporadic NTD cases. The tected in 300 control individuals. Moreover, the MTHFD gene encodes a afunctional enzyme arginine residue at position 293 is conserved in catalysing three sequential reactions in the inter- both the rat and the yeast homologue which pro­ conversion of one-carbon derivatives of tetrahy- vides additional evidence that this residue is of drofolate. The protein consists of two major importance (11, 20). Both affected members of the domains' the N-tcrminal region which harbours family inherited the mutant allele (Fig 5). How­ the dehydrogenase/cyclohydrolase domain, where­ ever, assuming that this mutation has a causal role,

ATG TAA MTHFD

1 .—. 3. ,11.- 2 4

200 bp

Fig .i Schematic representation of (he MTHFD ct>N/\ according to Hum et al (11). The gene encodes a single protein with three catalytic activities thai are important m folate metabolism The N-terminal region (I) of the corresponding protein harbours the dehydrogenase cyclohydrolase domain whereas the synthethase activity resides in the C-terminal half (II) I he hatched region represents Ihe putative connector region Arrowheads with connecting bars represent the amplification primers and amplified fragments used in Ihe SbCP analysis

93 A codone 290 291 292 293 294 295 OjO Ser АЛ. at Lya Arg Ph« Leu Wildtype allele 5 * - AGT GCC AAG CGT TTC CTG -

Mutant allele ACT GCC AAG CAT TTC CTG - Ser Ala Lya Hia Phe Leu SBO SBA

Mutant allele -^ ·"* " «M» — «— ШШ

fig 5 Pedigree of the family in wich the R293H mutation is segregating Members of the family were analyzed for the presence of the mutant allele by PC R amplification and subse­ quent Nlalll digestion The (wo types of NTD rccogni/ed in codoni» 650 651 652 653 654 655 this family were spina bilida occulta (SBO) and spina bifida Ils Ala Asp Arg xle Als aperta (SBA) respectively The mutant allele was present in Common allele 5 · — ATT GCA GAC CGG АТС GCA two brothers with spina bifida but also in some unaffected members of Ihc Tamil) The mutation was absent in 100 control individuals Occasional allele 5 * - ATT GCA GAC CAG АТС GCA - 3 • He Ala Asp Girt Xle Ala f-ii; 4 Paiual cDNA and protein sequence оГ the MTHFD In conclusion, our data provide evidence that gene showing the R293H (Л) and the R653Q (B) substitutions mutations in the MTHFD gene can act as a risk idenlilied in the present study The first substitution was ob­ factor in human NTD, however, it is unlikely that served in one single NTD family and was absent in 300 control this gene plays a major role in the etiology of individuals whereas the second one was shown to be an amino acid polymorphism Arrowheads mark the altered nucleotides human NTD clearly additional factors, genetic and/or environ­ Acknowledgements mental, have to be involved since the mutation is We thank the Dutch patient organisation BOSK for their assistance in contacting the families We also greatly ac­ also present in some of the unaffected family mem­ knowledge S van der Velde Visser and E van Rossum- bers Boenders for cell culture and LBV transformations The second alteration identified, R653Q. is likely Finally, we thank Professor H H Ropers for helpful discus­ to represent an amino acid polymorphism since it sions This study was supported by the Dutch Pnnses Beatrix Tonds grant no 94 1104 and 95 0521 Ж BH and occurred in both patients and controls and no ЬМ are members of INTEGER the International Neural significant allelic association between this polymor­ Tube Fmbryology Genetics and Epidemiology Research phism and NTD could be observed (Table 2). consortium to identify genes which predispose to neural Moreover, this alteration is not associated with a tube defects significant change in plasma homocysteine levels Therefore, our data provide no evidence for in­ volvement of the R653Q alteration in NTD etiol­ References ogy 1 MRC Vitamin Study Research Group Prevention of neu­ ral tube delects results of the Medical Research Council According to the threshold theory for complex Vitamin Study Lancet 1991 338 131 П7 disorders, like NTD. a critical threshold value 2 Czeizcl AK Dudas I Prevention of the first occurrence of needs to be exceeded before the defect actually neural-tube delects bv pcnconceptional vitamin supple­ develops Exceeding such a critical value might be mentation N Engl J Med 1992 327 1832 1835 accomplished by the additive effect of several mi­ 3 Steegcrs Theunissen RPM Boers GHJ Tnjbels FJM Finkelstem JD Blom HJ Thomas С MG Borm GF nor negative influences caused by particular allelic Wouters MGAJ bskes TKAB Maternal hyperhomocys- variants of different proteins that participate in the tctncmia a risk factor for neural-lube defects' Metabolism same functional cascade or metabolic pathway In 1994 43 1475 1480 this light, we compared the occurrence of the alter­ 4 Mills JI McPartlm JM Knke PN 1 ее YJ Conley MR ations identified in the present study with the dis­ Weir DG Scott JM Homocysteine metabolism in preg­ nancies complicated b\ neural tube detects Lancet 1995 tribution of the 677C Τ alleles of the MTU PR 345 149 151 gene None of the two changes identified in the 5 l·rosst Ρ Blom HJ Milos R Goyette Ρ Sheppard CA, present study showed an apparent correlation with Matthews RG Boers GJ den Heijer M Kluijtmans I A, the 677C—>T mutation in the MTHFR gene (not van den Heuvel LP et al Λ candidate genetic risk factor shown) for vascular disease a common mutation in mcthvlenele trahydrofolate reductase Nat Genet 1995 10 111 113

94 6 van dor Pul NMJ. Sleegers-Theunissen RPM Frossl P. tubedeiecls J Med Genet 1992 29 695 698 Tnjbds FJM. Eskes ΤΚΛΒ, van den Heuvel LP. Manman 11 Miller SA, Dykes DD Polcsky HF A simple salting out FCM den Heyer M. Rozen R Blom HJ Mulaled procedure for extracting DNA from human nucleated melhylenetetrahydrofolate reductase as a risk factor for cells Nucleic Acids Resi988 16 1215 spina binda Lancet 1995 346 1070 1071 14 Italiano С John SW Hum DW Mackenzie Rb. Rozcn 7 Kluijtmans LA, van den Heuvel LP. Boers GH. F-rossl Ρ R A pscudogenc on the X chromosome lor the human Stevens EM, van Oost BA, den Heijcr M Trubels f-J, tnfunctional enzyme MTHFD (methyleiieletrahydrofolale Rozen R. Blom HJ Molecular genetic analysis in mild dehydrogenase melhenylletrahydrofolate cyclohydrolase hyperhomocysleincmia a common mutation in the formyltctrahydrofolale synthetase) Genomics 1991 10 mcthylcnclclrahydrofolalc reductase gene is a genetic risk 1071 1074 factor for cardiovascular disease Am J Hum Genet 1996 15 Villar F. Schuster В Peterson D. Schirch V Cl-Tetrahy- 58 35 41 drofolatc synthase from rabbit liver Structural and kinetic 8 Pose) DL. Khoury MJ Mulinare J Adams Jr MJ, Ou CY properties of the enzyme and its two domains J Biol Chem Is mutated МГНг-R a risk factor for neural tube defects'' 1985 260 2245 2252 Lancet 1996 147 686 687 16 Whitehead AS Gallagher P. Mills JL. Kirke PN. Burke H. 9 van der Put NMJ. van den Heuvel LP. Stcegers-Theums- Molloy AM Weir DO. Shields DC" Scott JM A genetic sen RPM. Trubels FJM. Eskes TKAB, Manman FCM, defect in 5.10-methylcnetctrahydrofolaic reductase in neu Denheyer M, Blom HJ Decreased methylene lelrahydro- ral lube delects Q J Med 1995 88 761 766 folate reductase activity due to the 677C-*T mutation in 17 Ou CY Stevenson R F Broun VK Schwanz I F. Allen families with spina bifida offspring J Mol Med 1996 74 WP. Khoury MJ Rozcn R Oakley GI' Adams MJ 691 694 5,10-Mcthylcncletrahydrofolate reductase genetic poly­ 10 van der Put NMJ. Thomas С MG. Fskes ΤΚΛΒ. Trubels morphism as a risk factor for neural lube delects Am J HM Stccgcrs-Theunisscn Rl'M Manman ECM, de Med Genet 1996 61 610 614 Graaf-Hcss A. Smcitink JAM, Blom HJ Altered folate 18 van der Put NMJ. Eskes I KAB. Blom W Is the common and vitamin B12 metabolism in families with spina bihda 677C-*T mutation in the melhylcnetelrahydrololalc re­ offspring Q J Med 1997a 90 505 510 ductase gene a risk factor for neural tube delects' A mete 11 Hum DW Bell AW, Rozen R MacKen/ic RF Primary analysis Q J Med 1997b 90 111 115 structure of a human tnlunctional enzyme Isolation of a 19 Hum DW Mackenzie RE Expression of active domains cDNA encoding melhylenctetrahydrofolate dehydroge- of a human lolate- dependent tnlunctional cn/yme in nase-melhcnyltetrahydrololate cyclohydrolase-formyltc- Eschenihta <»/f Protein Fng 1991 4 493-500 trahydrofolale synthetase J Biol Chem 1988 263 20 Thigpen AF Wesl MG Appling DR Rat Cl-tetrahydro- 15946 15950 lolate synthase cDNA isolation tissuc-specilic levels of 12 Manman ECM. Hamel BCJ Sex ratios of affected and the mRNA. and expression of the protein in yeast J Biol transmitting members of multiple case families with neural Chem 1990 265 7907 7913

95

Chapter 8

General Discussion

97

General discussion

8.1 X-linked genes in Neural Tube Defects There is convincing evidence that the X chromosome harbours at least one predisposing gene that plays a role m the etiology of NTD Several families displaying apparent X-linked inheritance have been reported (Tonello et al, 1980, Baraitser & Bum 1984, Tonello 1984, Jensson et al, 1988) Recently, a de novo X/autosomal translocation involving Xq27 has been reported in a patient with spina bifida (Fryns et al, 1996) At this moment, we are in the process of cloning the breakpoints of a small duplication in Xq26-q27, which co-segregates with spina bifida m an American family (Goerss et al, 1993) In addition, an X-hnked gene seems to be responsible for the NTD phenotype in the Bent-tad mouse (Lyon & Searle, 1990) Haplotype and multipoint linkage analyses were performed in the Icelandic family reported by Jensson et al (1988), in a relatively large family with X-linked anencephaly and spina bifida (Chapter 2) While initially, the relevant gene could be excluded from approximately half of the X chromosome (Hoi et al, 1994), subsequent studies in the same family by Newton et al (1994) excluded the entire X-chromosome This puzzling result may be due to the fact that m the latter studies the penetrance of NTD in males carrying the defective allele was assumed to be complete Considering the complex pattern of inheritance in many multiple-case families and the incomplete penetrance in seemingly monogenic mutant mouse strains displaying NTD, like the curly tail and the X-linked Bent-tail mouse, this assumption is not very likely The NTD phenotype in the Icelandic family is probably not the human homologue of the Bent-tail mutation since the Xq27 region, which is syntenic to the Bent tail locus has been excluded in this family However, Bent tail has been mapped to a region of the X chromosome carrying several evolutionary rearrangements Evidently, a more definite conclusion regarding the location of the human counterpart of this gene has to await more accurate mapping of Bent tail in the mouse

8.2 PAX genes and Neural Tube Defects Studies in the mouse present the PAX3 gene as a very strong candidate for neural tube defects (Moase & Trasler, 1992) In the present study, the role of the PAX3 gene in neural tube defects in humans has been investigated Linkage analysis provided no evidence tor a major role of PAX3 (Chatkupt et al, 1995) (Chapter 3) Also, no mutations were found within the coding region of the gene in non-syndromic NTD patients, and no significant association could be demonstrated between the disease and particular allelic variants of the PAX3 gene (Chapter 4 & 5) Therefore, it was concluded that, in humans, PAX3 is not involved as a major gene in non-syndromic NTDs On the other hand, there is evidence that the gene can act as a risk factor for syndromic NTDs since mutations in this gene cause the dominant Waardenburg syndrome (WS, type 1 & 3) Patients with WS1 have a moderately increased

99 Chapter 8

risk for spina bifida and PAX3, mutations have been detected in patients with both spina bifida and WS1 (Baldwin et al., 1995; Hol et al., 1995a; Read & Newton, 1997)(Chapter 4). Heterozygous Splotch mice show white belly spots, whereas homozygous embryos display both cranial and caudal NTDs as a result of mutations in the murine РахЗ gene. Accordingly, it was expected that, in humans, homozygotes are much more severely affected than heterozygotes and that they have a high probability to develop neural tube defects. To date, two cases of homozygosity for WS1 have been documented. Ayme and Philips (1995) reported on a fetus that was the product of a brother-sister incest who both had typical WS1. The fetus had major abnormalities very reminiscent of homozygous Splotch mouse embryos, including ancncephaly. The second case concerned a child bom to consanguineous parents, with severe hypopigmentation, profound congenital deafness and contractures of the forearm, but without neural tube defect (Zlotogora et al., 1995). Therefore in humans, homozygosity for mutations in PAX3 may be compatible with life, at least in early infancy, as demonstrated by this latter case, and does not necessarily cause neural tube defects. In general, symptoms in patients with WS1 are highly variable, even within families, without a clear correlation between genotype and phenotype (Farrer et al., 1994). Several studies have indicated that the clinical variability of WS1 may be due to the involvement of modifier loci (Reynolds et al., 1996). The striking concordance seen in monozygotic twins with Waardenburg syndrome strongly suggests that the clinical variation is genetically controlled (Pandya et al., 1996). Thus, the spina bifida phenotype occasionally seen in patients with WS1 may be the result of genetic variation at loci that modulate expression of the PAX3 gene. Since РАХЗ encodes a transcription factor with specific DNA-binding properties, it is reasonable to suppose that genetic variation in downstream targets (or upstream regulators) may influence the expression of the disease. From expression studies in Splotch mice several potential downstream target genes for РахЗ have been proposed, including Msx2, cMet, MyoD, Myf-5 and versican (Takahashi, 1996; Henderson et al., 1997; Maroto et al., 1997). In addition, the gene for the neural cell adhesion molecule (NCAM) has been proposed as a downstream target of РахЗ (Chalepakis et al., 1994). The promoter region of NCAM harbours a paired domain binding site, and several Pax proteins, including РахЗ, have been shown to influence NCAM expression in vitro (Hoist et al., 1997). Moreover, in Splotch mice, expression and post-translational modification of the gene for NCAM is altered which is probably a direct effect of the РахЗ mutation (Moase & Trasler, 1991; Neale & Trasler, 1994). Therefore, the NCAM gene can be regarded as an excellent candidate to play a role in neural tube formation. Alternatively, variation in genes from completely different molecular pathways may affect the biological expression of РахЗ mutations. РахЗ and its paralogue Pax7, which belongs to the same sub-class of the Pax gene family, share a very

100 General discussion similar structure and similar expression pattern (Strachan & Read, 1994) Homozygous Pax7 knockout mice display dysgenesis of cephalic neural crest derivatives and die shortly after birth, but have no neural tube defects (Mansoun et al, 1996) From the phenotype in these mice it has been suggested that the function of РахЗ and Pax7 may be partly redundant Therefore, the combined presence of adverse genetic information in both РахЗ and Pax7 may result in neural tube defects In the present study, the paired domain region of the PAX7 gene, which was the only sequence cloned at the time (Bum et al, 1989), was screened for mutations in NTD patients, but without success Recently, the full length coding region, as well as the genomic organisation of PAX7 has been elucidated (Vorobyov et al, 1997) It will be of interest to look for sequence variation m the remaining part of this gene PAX1 and its paralogue PAX9 are expressed in the developing vertebral column rather than m the developing central nervous system The PAX1 gene has been proposed as a candidate for several hereditary vertebral malformation syndromes, like Alagille syndrome, Khppel-Feil syndrome and Jarcho-Levin syndrome which are associated with spina bifida (Gardner 1979, Lendon et al, 1981, David et al, 1996, Giacoia & Say, 1991) However, no mutations have been found in the paired domain region of the gene in patients with any of these syndromes (Balling, 1994, Hoi et al, 1995b, Smith & Tuan 1994) Studies in the mouse have presented the PAX1 gene as a candidate for spina bifida (Helwig et al, 1995) Recently, it was shown that Paxl expression was decreased m valproic acid (VPA) treated chicken embryos suggesting that Paxl is a molecular target in VPA axial skeletal teratogenicity (Barnes et al, 1996) Since VPA is a well-known NTD-causing teratogen in humans, these findings link Paxl to neural tube development (Lindhoul & Schmidt, 1986) We identified a missense mutation (Gln42His) in the paired domain of PAX1 in a patient with spina bifida (Hoi et al, 1996) (Chapter 5) The mutation, however, was also present in unaffected relatives indicating that on its own it is not sufficient to cause spina bifida and that additional factors must be involved Nevertheless, this is the first PAX1 mutation identified that may be related to human disease Moreover, the intriguing results of the transfection studies indicate that wildtype and mutant Paxl proteins could have opposing activities depending on the differentiation status of the expressing cells (Joosten et al, 1998) (Chapter 6) If this actually happens during embryogenesis this may have enormous consequenses for the developing embryo Therefore, our results may highlight a new molecular mechanism responsible for specific mammalian diseases Furthermore, our data suggest that the PDGFRa gene is a downstream target of the PAX1 transcription factor Consequently, PDGFRa itself makes up a good candidate for NTD, particularly since deletion or targeted disruption of this gene causes spina bifida in mice (Payne et al, 1997, Sonano 1997) However, no mutations have been

101 Chapter 8

identified so far in the coding region of the PDGFRa gene in a large panel of patients with spina bifida (unpublished results). Two major categories of mechanisms are postulated to cause spina bifida: (i) Failure to form and/or close the neural tube caused by defects residing in either neural or mesodermal cells, as exemplified by the defects seen in Splotch and curly tail, respectively, (ii) Alternatively, abnormalities in the adjacent mesoderm from which the supportive and protective tissues are derived have been proposed to cause vertebral spina bifida occulta. Examples of the latter category are the defects resulting from mutations in the PDGFRa gene in the mouse (Payne et al., 1997). It seems likely that NTDs resulting from mutations in Paxl are also caused by this latter mechanism, since Paxl and PDGFRa seem to participate in the same functional cascade (Chapter 6) and mutations in Paxl are known to disturb vertebral development (Wallin et al. 1994; Helwig et al. 1995). However, the defect seen in the fetus carrying the PAX1 mutation concerned a lumbar spina bifida aperta. The mutation had been inherited from the seemingly unaffected mother and maternal grandmother. It would be of interest, in the light of the present discussion, to see whether the transmitting members possibly display spina bifida occulta. None of these individuals, however, had any neurological or other clinical signs, which is why they were not subjected to radiological examination. These findings raise the question whether spina bifida occulta (SBO) and spina bifida aperta (SBA) may be etiologically related. SBO are vertebral anomalies in which the neural arches have not fused together and therefore, rather represent a vertebral anomaly with little or no clinical manifestations. A very mild form involving non-fusion of the fifth lumbar and/or the first sacral vertebrae is commonly (5-20%) seen in humans (Boone et al. 1985; Fidas et al. 1987). On the other hand, there is evidence that SBO and SBA may share a common (genetic) etiology. Lendon et al. (1981) provided evidence that congenital vertebral abnormalities and SBA are etiologically related. In addition, family studies revealed that in parents of SBA patients, the frequency of SBO is significantly elevated (Lorber & Levick, 1967; Gardner et al., 1974), although another study could not confirm this finding (Schweitzer et al., 1993). SBA and SBO occasionally occur in the same family as exemplified by the present study in which we encountered several families with both types of defects (personal observations). Studies on mutant mouse strains displaying NTD suggest that the mechanism of SBO- development may be heterogeneous and that only some forms of SBO may be related to SBA (Harris & Juriloff, 1997). In this light, it is interesting that in mice, VPA treatment can results in cither exencephaly, SBA or SBO suggesting similar causative mechanisms for each of these defects (Nau, 1994). In conclusion, it seems likely that in certain families the occurrence of SBO and SBA results from a common genetic predisposition.

102 General discussion

8.3 Folic acid and Neural Tube Defects Environmental factors influencing the incidence of NTD provide another major lead to the identification of the genetic pathways and specific genetic risk factors Clinical trials demonstrate that folate can prevent a considerable proportion of NTD in humans (MRC Vitamin study group, 1991, Czeizel & Dudas, 1992) suggesting that neural tube closure is sensitive to the overall efficiency of the folate metabolism, but little is known about the underlying molecular mechanism Elevated homocysteine levels have been reported in mothers who had a previous NTD pregnancy and mutations in genes encoding enzymes from the folate mediated homocysteine pathway have been proposed to play a role in NTD etiology (Steegers-Theunissen et al, 1994, Lucock et al, 1997) On the other hand, elevated levels of homocysteine are also seen in inbom errors of folate metabolism involving deficiencies of specific enzymes like MTHFR or MS, for which no association with NTDs has been reported (Fowler, 1997) We identified a missense mutation (Arg293His) m the MTHFD gene, which is a crucial component of the folate mediated homocysteine pathway (Hoi et al, 1998)(Chapter 7) The mutation was present in one spina bifida family with two affected boys, one displaying SBA and the other SBO, who both inherited the mutation from their mother Also here, family analysis revealed that this mutation alone is not sufficient to cause NTD since not all carriers are affected One obvious factor that may have played an additional role in this family could have been the folate status of the mother during early pregnancy It can be easily imagined that a combination of low maternal folate together with inefficient handling of the available folate due to the mutation in the MTHFD gene, either in the mother, the fetus or in both, has caused the defect For the moment this remains highly speculative because no data are available on the folate status nor on the homocysteine levels of the mother Moreover, it is not yet known whether the observed mutation actually affects the catalytic activity of the protein Functional assays should shed more light on this matter Fleming and Copp (1998) have recently shown that Splotch mice have a defect in their folate metabolism There appears to be a metabolic deficiency in the supply of folate for the biosynthesis of pynmidine Exogenous folic acid and thymidine both correct the biosynthetic defect and prevent some NTDs in homozygous Splotch embryos Interestingly, this links the PAX3 gene, which is mutated in the Splotch mouse, to the folic acid metabolism and brings together two major topics of the present study However, at this point it remains unclear exactly how the PAX3 gene is connected to the folic acid metabolism One obvious explanation might be that specific genes encoding enzymes involved in folic acid metabolism are downstream targets of the PAX3 transcription factor

103 Chapter 8

To date, periconceptional folate supplementation is the sole form of primary prevention for neural tube defects. However, it has been estimated that at least one-third of NTD cases cannot be prevented in this way even if every woman would be supplemented. Recently, animal experiments have shown that inositol, a water-soluble B-complex vitamin, can reduce the incidence of folate-resistant NTD in genetically predisposed curly tail mice (Greene & Copp, 1997). This ameliorating effect is brought about via the inositol-mediated signal transduction pathway. Also in humans there are indications for a link between inositol metabolism and NTD since diabetic women, who have an increased risk for NTD offspring (Myrianthopoulis & Melnick, 1987), show increased myo-inositol levels in plasma, urinary and cerebrospinal fluid (Greene et al., 1987). Inositol was not included in the multivitamin preparations used in previous intervention studies. Therefore, its effectiveness in preventing human NTDs has not yet been tested. It would be of great interest to see whether, in analogy to the folic acid story, evidence can be found for a protective role of this vitamin in human NTD.

8.4 Conclusion and future prospects Evidence from studies in the mouse indicates that many genes may play a role in impaired neural tube formation. Accordingly, it can be expected that also in humans the causal mechanisms leading to neural tube defects will be very heterogeneous. In the present study we have set out to identify genetic factors involved in human neural tube defects. Although we managed to identify several genetic risk factors that may influence the genetic susceptibility for NTDs, no single factor nor a combination of factors has as yet been identified that can adequately explain the etiology and pathology of this disorder. It becomes increasingly clear that susceptibility to neural tube defect results from the additive effect of two or more independently segregating genetic risk factors in combination with specific adverse environmental triggers. Therefore, exploring gene-gene interactions will be a promising approach towards the elucidation of NTD etiology. The identification of genetic factors that predispose to NTD tumes out to be laborious. Therefore, extensive automation of the analytical procedures will significantly contribute to the elucidation of the molecular basis of these disorders. One of the most promising strategies to successfully identify predisposing genes in neural tube defects will be the candidate gene approach. Right now, the Human Genome Project is providing data on a growing number of genes. This information, together with a better understanding of the process of mammalian neurulation, including detailed information on gene expression profiles during this developmental process, will eventually point to new candidate genes. Analysis of these candidates in large cohorts of patients, either by association studies, but preferably by automated mutation analysis, will undoubtedly lead to the identification of genes involved in

104 General discussion these congenital defects However, with respect to neural tube defects, and probably many other multifactorial traits, it seems that neither environmental nor genetic studies alone can generate the complete information necessary to understand the etiology and pathogenesis of these complex disorders For this reason, a multidisciphnary approach, which integrates biochemistry, molecular biology, epidemiology and human/mouse genetics will be required in future attempts to unravel the etiology and pathogenesis of NTD

105 Chapter 8

References Ayme-S; Philip-N. Possible homozygous Waardenburg syndrome m a fetus with exencephaly Am-J- Med-Genet (1995) 59: 263-5

Baldwin CT; Hoth CF; Macina RA; Milunsky A. Mutations in PAX3 that cause Waardenburg syndrome type I: ten new mutations and review of the literature. Am-J-Med-Genet (1995)58: 115-22

Balhng-R. The undulated mouse and the development of the vertebral column. Is there a human PAX-1 homologue? Clin-Dysmorph ( 1994) 3:185-191

Baraitser-M; Bum-J. Neural tube defects as an X-hnked condition. Am-J-Med-Genet (1984)17: 383-5

Bames-GL Jr; Manani-BD; Tuan-RS. Valproic acid-induced somite teratogenesis m the chick embryo: relationship with Pax-1 gene expression. Teratology (1996)54· 93-102

Boone-D; Parsons-D; Lachmann-SM; Sherwood-T. Spina bifida occulta: Lesions or anomaly? Clm- Radiol (1985)36:159-161

Bum-M; Tromvoukis-Y; Bopp-D; Fngeno-G; Noll-M. Conservation of the paired domain in metazoans and its structure m three isolated human genes. EMBO-J (1989)8: 1183-90

Chalepakis-G; Jones-FS; Edelman-GM; Gruss-P. Pax-3 contains domains for transcription activation and transcription inhibition. Proc-Natl-Acad-Sci-USA (1994)91: 12745-9

Chatkupt-S; Hol-FA; Shugart-YY; Geurds-MP; Stenroos-ES; Koemgsberger-MR; Hamel-ВС; Johnson- WG, Manman-ЕС Absence of linkage between familial neural tube defects and PAX3 gene. J-Med- Genet (1995) 32: 200-4

Czeizel-AE; Dudas-I. Prevention of the first occurrence of neural-tube defects by penconceptional vitamin supplementation. N-Engl-J-Med (1992)32: 1832-5

David-KM; Copp-AJ; Stevens-JM; Hayward-RD; Crockard-HA. Split cervical spinal cord with Khppel- Feil syndrome: seven cases. Brain (1996) 119: 1859-72

Farrer-LA; Amos-KS; Asher-JH Jr; Baldwin-CT; Diehl-SR; Fnedman-TB; Greenberg-J; Grundfast-KM; Hoth-C; Lalwani-AK; et-al. Locus heterogeneity for Waardenburg syndrome is predictive of clinical subtypes Am-J-Hum-Genct(1994)55: 728-37

Fidas-A; MacDonald-HL; Elton-RA; Wild-SR; Chisholm-GD; Scott-R. Prevalence and patterns of spina bifida occulta in 2,707 normal adults Clin-Radiol (1987)38: 537-42

Flcming-A; Copp-AJ. Embryonic folate metabolism and mouse neural tube defects. Science (1998)280: 2107-09

Fowler-B. Disorders of homocysteine metabolism. J-Inher-Metab-Dis (1997)20: 270-85

Fryns-JP; Devnendt-K; Moerman-Ph Lumbosacral spina bifida and myeloschizis m a female foetus with de novo X/autosomaltranslocation (t(X;22)(q27;ql21)). Genet-Couns (1996)7:159-60

Gardner-RJM; Alexander-C; Veale-AMO. Spina bifida occulta in the parents of offspring with neural tube defects. J-Genet-Hum (1974) 22: 389-95 106 General discussion

Gardner-WJ. Klippel-Feil syndrome, îniencephalus, anencephalus, hindbram hernia and mirror movements: overdistention of the neural tube. Childs-Brain (1979)5: 361-79

Giacoia-GP; Say-B. Spondylocostal dysplasia and neural tube defects. J-Med-Genet (1991) 28: 51-3

Goerss-JB; Kames-PS; Thibodeau-SN, Johnson-DD; Zimmerman-D et al. Cytogenetic and molecular studies of a duplication of Xq26 and Xq27 in two brothers with neural tube defects. Am-J-Hum-Genet (1993) 53(Suppl) 440(abstract)

Greene-ND; et al. N-Engl-J-Med (1987) 316: 599

Greene-ND; Copp-AJ. Inositol prevents folate-resistant neural tube defects in the mouse. Nat-Med (1997)3:60-6

Hams-MJ; Junloff-DM. Genetic landmarks for defects in mouse neural tube closure. Teratology (1997) 56: 177-187

Helwig-U; Imai-K; Schmahl-W; Thomas-BE; Vamum-DS; Nadeau-JH; Balling-R Interaction between undulated and Patch leads to an extreme form of spina bifida in double-mutant mice. Nat Genet Nat Genet (1995) 11: 60-63

Henderson-DJ; Ybot-Gonzalez-P; Copp-AJ. Over-expression of the chondroitin sulphate proteoglycan versican is associated with defective neural crest migration in the РахЗ mutant mouse (fylotch) Mech- Dev (1997) 69: 39-51

Hol-FA; Geurds-MPA; Jensson-O; Hamel-BCJ; Moore-GE; Newton-R; Manman-ECM. Exclusion mapping of the gene for X-hnked neural tube defects in an Icelandic family. Hum-Genet (1994) 93: 452- 456

Hol-FA; Hamel-BCJ; Geurds-MPA; Mullaart-RA; Barr-FG; Macina-RA; Manman-ECM. A frameshift mutation m the gene for PAX3 in a girl with spina bifida and mild signs of Waardenburg syndrome. J- Med-Genet (1995a) 32: 52-56

IIol-FA; Hamel-BCJ; Geurds-MPA; Hansmann-I; Nabben-FA; Damels-O; Manman-ECM Localization of Alagille syndrome to 20pll.2-pl2 by linkage analysis of a three-generation family. Hum-Genet (1995b) 95: 687-90

Hol-FA; Geurds-MPA; Chatkupt-S; Shugart-YY; Balling-R; Schrander-stumpel-CTRM; Johnson-WG; Hamel-BCJ; Manman-ECM. PAX genes and neural tube defects: an amino acid substitution in PAX1 in a patient with spina bifida. J-Med-Genet (1996)33: 655-660

Hol-FA; van der Put-NMJ; Geurds-MPA; Heil-SG; Hamel-BCJ, Manman-ECM; Blom-HJ Molecular analysis of the human tnfunctional enzyme MTH.FD (Methylenetetrahydrofolate Dehydrogenase- Methenyltetrahydrofolatc Cyclohydrolase- Formyltetrahydrofolate synthetase) in patients with neural tube defects. Clinical-Genet (1998) 53:-119-125

Holst-BD; Wang-Y; Jones-FS; Edelman GM. A binding site for Pax proteins regulates expression of the gene for the neural cell adhesion molecule in the embryonic spinal cord. Proc-Natl-Acad-Sci-USA (1997) 94: 1465-70

107 Chapter 8

Jensson-O; Amason-A, Gunnarsdottir-H; Petursdottir-I; Fossdal-R; Hreidarsson-S. A family showing apparent X linked inheritance of both anencephaly and spina bifida. J-Med-Genet (1988)25: 227-9

Joosten-PHLJ; Hol-FA; van Beersum-SEC; Peters-Η; Hamel-BCJ; Afink-GB, van Zoelen-EJJ; Manman-ECM. Altered regulation of PDGFRa transcnption m vitro by spina bifida associated mutant Paxl proteins Proc-Natl-Acad-Sci-USA(1998)95:14459-63

Lendon-RG; Wynne-Davies-R; Lendon-M. Are congenital vertebral anomalies and spina bifida cystica related? J-Med-Genet (1981) 18: 424-427

Lmdhout-D; Schmidt-D. In-utero exposure to valproate and neural tube defects. Lancet (1986)1: 1392-3

Lorber-J; Levick-K. Spina bifida cystica. Arch-Dis-Child (1967)42: 171-73

Lucock-MD; Wild-J; Lumb-CH; Oliver-M; Kendall-R; Daskalakis-I; Schorah-CJ; Levene-MI. Risk of neural tube defect-affected pregnancy is associated with a block in maternal one-carbon metabolism at the level of N-5-methyltetrahydrofolate:homocysteine methyltransferase. Biochem-Mol-Med (1997) 61: 28-40

Lyon-MF; Searle-AG. Genetic variants and strains of the laboratory mouse. (1990) 2nd ed. New York: Oxford University Press

Mansoun-A; Stoykova-A; Torres-M; Gruss-P. Dysgenesis of cephalic neural crest derivatives m Pax7-/- mutant mice. Development (1996) 122: 831-8

Maroto-M; Reshef-R; Munsterberg-AE; Koester-S; Gouldmg-M; Lassar-AB. Ectopic Pax-3 activates MyoD and Myf-5 expression in embryonic mesoderm and neural tissue. Cell (1997)89: 139-48

Moase-CE; Trasler-DG N-CAM alterations in Splotch neural tube defect mouse embryos. Development (1991)113: 1049-1058

Moase-CE; Trasler-DG Splotch locus mouse mutants- models for neural tube defects and Waardenburg syndrome type I in humans. J Med Genet (1992) 29: 145-51

MRC Vitamin Study Research Group. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet (1991)338' 131-7

Myrianthopoulos-NC; Melnick-M. Studies in neural tube defects. I. Epidemiologic and etiologic aspects. Am-J-Med-Genet (1987) 26: 783-96

Nau-H. Valproic acid-induced neural tube defects. Ciba-Found-Symp (1994) 181: 144-52

Neale-SA; Trasler-DG. Early sialylation on N-CAM in splotch neural tube defect mouse embryos. Teratology ( 1994) 50: 118-24

Newton-R; Stanier-P; Loughna-S; Henderson-DJ; Forbes-SA; Farrall-M; Jensson-O; Moore-GE. Linkage analysis of 62 X-chromosomal loci excludes the X chromosome in an Icelandic family showing apparent X-hnkcd recessive inheritance of neural tube defects. Clin-Genet (1994) 45: 241-9

Pandya-A; Xia-XJ; Landa-BL; Amos-KS; Israel-J; Lloyd-J; James-AL; Diehl-SR; Blanton-SH; Nance- WE. Phenotypic variation in Waardenburg syndrome: mutational heterogeneity, modifier genes or polygenic background? Hum-Mol-Genet (1996) 5: 497-502 108 General discussion

Payne-J; Shibasaki-F; Mercola-M. Spina bifida occulta in homozygous Patch mouse embryos. Dev-Dyn (1997)209: 105-16

Read-AP; Newton-VE. Waardenburg syndrome. J-Med-Genet (1997)34: 656-665

Reynolds-JE; Marazita-ML; Meyer-JM; Stevens-CA; Eaves-LJ; Amos-KS; Ploughman-LM; MacLean- C; Nance-WE; Diehl-SR. Major-locus contnbutions to variability of the craniofacial feature dystopia canthorum in Waardenburg syndrome. Am-J-Hum-Genet (1996)58: 384-92

Schweitzer ME, Balsam D, Weiss R, Spina bifida occulta. Incidence in parents of offspring with spina bifida cystica. Spine 18: 785-86 (1993)

Smith-CA; Tuan-RS. Human PAX gene expression and development of the vertebral column. Clin- Orthop (1994) 302: 241-50

Sonano-P. The PDGF alpha receptor is required for neural crest cell development and for normal patterning of the somites. Development (1997)124: 2691-700

Steegers-Theunissen-RP; Boers-GH; Tnjbels-FJ; Finkelstein-JD; Blom-HJ; Thomas-CM; Borm-GF; Wouters-MG; Eskes-TK. Maternal hyperhomocysteinemia: a risk factor for neural-tube defects'? Metabolism (1994) 43: 1475-80

Strachan-T, Read-AP. PAX genes. Curr-Opin-Genet-Dev (1994)4: 427-38

Takahashi Y. Organization of the spina bifida neural tube in Splotch (Pax-3 defective) mouse embryos. Dev-Growth-Dif ( 1996) 38: 23-31

Tonello-HV; Warren-ST; Lindstrom-JA Possible X-linked anencephaly and spina bifida—report of a kindred. Am-J-Med-Genet (1980) 6: 119-21

Tonello-HV. Report of a third kindred with X-linked ancncephaly/spma bifida. Am-J-Med-Genet (1984) 19:411-2

Vorobyov E; Mertsalov I; Dockhom-Dwomiczak B; Dwomiczak B; Horst J. The genomic organization and the full coding region of the human PAX7 gene. Genomics (1997)45: 168-74

Wallm-J; Wilting-J; Koseki-H; Fntsch-R; Christ-B; Ballmg-R. The role of Pax-1 in axial skeleton development. Development (1994) 120: 1109-21

Zlotogora-J; Lerer-I; Bar-David-S; Ergaz-Z; Abehovich-D. Homozygosity for Waardenburg syndrome Am-J-Hum-Genet. (1995) 56: 1173-8

109

Chapter 9

Summary / Samenvatting

πι

Summary Neural tube defects (NTD) are congenital malformations of the central nervous system including defective formation of the brain (anencephaly) and spinal cord (spina bifida). They occur with an average incidence of about of 1 per 1000 newborns. NTD represent a classical example of multifactorial disorders and are generally ascribed to a combination genetic predisposition and unfavourable environmental influences. From studies in the mouse it appears that multiple genes can be involved in NTD, and a similar situation seems to exist for humans. However, very little is known about these genetic factors. Knowing the genetic factors will eventually give insight in the pathogenic processes causing this disorder. Accordingly, the main objective of the present study is the identification of genetic factors that are involved in the etiology of NTD. To actually identify genetic risk factors for NTD two strategies have been followed: (1) genetic analysis of families with apparently Mendelian inheritance of NTD, and (2) analysis of candidate genes selected from animal models or epidemiologic data on humans. Linkage analysis in a relatively large Icelandic family with apparent X-linked NTD led to the exclusion of about half of the X chromosome and provided evidence for the existence of more than one NTD-gene on the X-chromosome (Chapter 2). There are now over 20 different mouse strains with Mendelian inheritance of NTD. In this respect two genes have been under investigation: The PAX3 gene underlying the Splotch phenotype and the PAX1 gene for undulated. Linkage between PAX3 and familial NTD was excluded indicating that this gene is not a major gene for non-syndromic NTD (Chapter 3). However, in one patient a small deletion in the PAX3 gene was found, which was shown to cause type 1 Waardenburg syndrome, a dominant disorder clinically associated with NTD, suggesting that mutations in this gene predispose to NTD (Chapter 4). No significant allelic association between protein polymorphisms in PAX3 and human NTD was detected using a sufficient number of healthy controls. In another patient a missense mutation in the paired domain of PAX1 was detected (Chapter 5). The corresponding mutant PAX1 protein displays altered DNA-binding properties in bandshift assays. Using the human PDGFRa promoter linked to the luciferase gene we were able to demonstrate that PAX1 can act as an upstream regulator of PDGFRa expression (Chapter 6). Mutations in PAX1, including the one identified in our patient, alter this activity. Moreover, the effect of each individual mutation on PDGFRa expression appears to depend on the cell type and differentiation status of the cell that was used for the transfection studies. Our results suggests that mutations in PAX1 can act as a risk factor for NTD. Furthermore, PAX1 and PDGFRa may act in the same functional cascade, and deregulated PDGFRa expression may be causally related to spina bifida.

113 Chapter 9

Environmental factors influencing the incidence of NTD provide another major lead to the identification of the genetic pathways and specific genetic risk factors. Clinical trials demonstrate that folic acid can prevent a considerable proportion of NTD in suggesting that neural tube closure is sensitive to the overall efficiency of the folate metabolism. Accordingly, genes involved in the folate pathway make up good candidates for NTD. The MTHFD gene encodes a protein with three catalytic activities important in folate metabolism and has been investigated. A missensc mutation was identified in one family suggesting that mutations in this gene may act as risk factors for NTD (Chapter 7).

114 Samenvatting Neurale buis defecten (NBD) zijn ernstige aangeboren afwijkingen van het centrale zenuwstelsel en komen gemiddeld voor bij 1 op de 1000 geboorten Ze worden veroorzaakt doordat vroeg in de zwangerschap (vierde week) de neurale buis niet volledig sluit De meest voorkomende vormen zijn spina bifida (open ruggetje) en anencefalie (kattekopje) Het wordt algemeen aangenomen dat erfelijkheid een rol speelt bij NBD Naast erfelijke factoren kunnen er ook omgevingsfactoren in het spel zijn Tegenwoordig denkt men dat een combinatie van genetische aanleg en negatieve omgevingsinvloeden verantwoordelijk is voor het ontstaan van deze aandoening Wat er precies mis gaat tijdens het sluitingsproces, en welke genen hierbij betrokken zijn is niet of nauwelijks bekend Wanneer we meer zouden weten van de genetische factoren die een rol spelen bij dit proces, zouden we wellicht ook meer begrijpen van de mechanismen die eraan ten grondslag liggen Dit laatste zou dan weer van belang kunnen zijn voor de ontwikkeling van nieuwe methoden om NBD te voorkomen De huidige studie beoogt het opsporen van genetische factoren die een rol spelen bij het ontstaan van NBD Voor het opsporen van genetische risico factoren hebben we een tweetal strategieën gevolgd (1) genetische analyse van families die een Mendehaanse overerving van NBD vertonen, en (2) analyse van kandidaatgenen die naar voren zijn gekomen uit dierexperimenteel onderzoek of uit epidemiologische studies bij de mens Koppehngsonderzoek in een relatief grote IJslandse familie met een ogenschijnlijk X- gebonden vorm van NBD stelde ons in staat om ongeveer de helft van het X chromosoom uit te sluiten Een groot deel van de huidige kennis over het ontstaan van NBD is afkomstig van studies bij de muis Inmiddels zijn er meer dan 20 muizenstammen bekend die een Mendehaanse overerving van NBD vertonen Een tweetal genen, PAX3 en PAX1, waarvan bekend is dat ze betrokken zijn bij het ontstaan van NBD m de muis, zijn door ons nader onderzocht bij de mens Koppehngsonderzoek in families met NBD leverde geen aanwijzingen op dat PAX3 betrokken is bij familiaire non-syndromische NBD Vervolgens is door ons het gehele humane PAX3 gen gescreend op de aanwezigheid van mutaties m een groep van circa 120 NBD patiënten In het gen werden een aantal polymorfismen geïdentificeerd, echter geen van hen vertoonde significante associatie met NBD De resultaten geven aan dat PAX3 geen 'major gene' is voor non-syndromische NBD Echter, in een patient met spina bifida werd een kleine deletie in het PAX3 gen gevonden Nader klinisch onderzoek gaf aan dat deze patient leed aan type 1 Waardenburg syndroom, een dominante aandoening die klinisch geassocieerd is met NBD Deze laatste bevinding suggereert dat mutaties in het PAX3 gen predisponeren voor een syndromische vorm van NBD Analyse van het PAX1 gen bracht een mutatie aan het licht in een andere patient met spina bifida De mutatie is gelokaliseerd in het DNA-bindende 'paired domain' Het 115 Chapter 9 corresponderende mutante eiwit vertoond veranderde DNA-bindende eigenschappen. Tevens hebben we aan kunnen tonen dat PAX1 kan fungeren als 'upstream' regulator van het PDGFRa gen. Inactivatie van dit laatste gen in de muis resulteert in een ernstige vorm van spina bifida. Mutaties in PAX1, inclusief de door ons gevonden mutatie, blijken van invloed op de expressie van PDGFRa. Het effect van afzonderlijke mutaties blijkt afhankelijk van cel type en differentiatie status van de cellen. Onze bevindingen duiden er op dat PAX1 en PDGFRa participeren dezelfde functionele route en dat mogelijk deregulatie van PDGFRa expressie de oorzaak is van sommige vormen van spina bifida. Omgevingsfactoren die van invloed zijn op de incidentie van NBD vormen een andere benadering waarlangs kandidaatgenen kunnen worden gedefinieerd. Uit klinische studies is gebleken dat foliumzuur een aanzienlijk percentage van het aantal NBDs kan voorkomen. Het mechanisme van deze beschermende werking is onbekend. Waarschijnlijk is de sluiting van de neurale buis gevoelig is voor de efficiëntie van het folaat metabolisme. Vandaar dat genen die betrokken zijn bij het folaat metabolisme in principe goede kandidaten zijn voor NBD. Het MTHFD gen is zo'n kandidaatgen en codeert voor een eiwit met drie verschillende katalytische eigenschappen die van belang zijn in het folaat metabolisme. Ook dit gen is door ons onderzocht en in een familie werd een missense mutatie aangetroffen. Waarschijnlijk vormen mutaties in dit gen een risico factor voor NBD.

116 Dankwoord Het heeft even geduurd maar nu is dan toch het proefschrift gereed' Dit was nooit gelukt zonder de hulp van velen Vanaf deze plaats wil ik iedereen bedanken die op de een of andere manier een bijdrage heeft geleverd aan de totstandkoming van dit boekje In het bijzonder de labgenoten van de sectie moleculaire genetica, het lab waar het grootste deel van het onderzoek werd uitgevoerd Ik denk met veel plezier terug aan de jaren dat ik binnen jullie lab mijn werkplek had Een aantal mensen wil ik graag even persoonlijk bedanken Allereerst professor Ropers, mijn promotor Beste Hilger, bedankt voor de mogelijkheid om te werken en mezelf te ontplooien binnen de afdeling Anthropogenetica Ik was verheugd dat je, ondanks je drukke functie dis hoofd van het MPI m Berlijn, toch als mijn promotor wilde optreden Edwin Manman, co-promotor en initiator van het NBD onderzoek Bedankt voor de goede- en vooral ook prettige begeleiding Ik heb veel van je geleerd Ook m de toekomst hoop ik nog regelmatig onaangekondigd je kamer binnen te mogen vallen, om een minuutje van je tijd te vragen (en er vervolgens vijftien te nemen) Professor Brunner, beste Han Door steeds op het juiste moment de vraag te stellen -Hoc staat het met het proefschrift'7- wist je op een pnma manier de vereiste druk op de ketel te houden Beste Monique, tja wat moet ik over jou vertellen Ik had geluk Van het begin af aan kreeg ik ondersteuning van een analist, JIJ dus Het zal iedereen duidelijk zijn dat zonder jouw inspanningen dit proefschrift er nooit gekomen was Veel van het praktische werk is door jouw handen uitgevoerd Je enthousiasme hierbij was voor mij vaak een enorme stimulans Onze samenwerking was -denk ik- uniek Bedankt' Beste Ben, bedankt voor de klinische ondersteuning van het NBD onderzoek Sylvia, Marga en Ronald, bedankt voor de praktische ondersteuning, goede adviezen en vooral ook voor de gezelligheid Jullie zijn meer dan collega's Sandra en Maurice, bedankt voor jullie stage bijdragen aan dit onderzoek Betsy, hooikoorts-maatje, bedankt voor de mentale steun tijdens de jaarlijkse pollen uitbarsting Saskia en Liesbeth, mijn waardering voor het perfect uitvoeren en organiseren van het vele kweekwerk voor het NBD onderzoek Alle medewerkers van de afdeling DNA-diagnostiek, bedankt voor de getoonde gastvrijheid op het moment dat het lab moleculaire genetica even te klein werd voor oudgedienden Thanks to all the members of the INTEGER consortium for overall support, good ideas and enjoyable meetings Naast het werk is er natuurlijk ook het thuisfront Allereerst mijn familie Zonder jullie onvoorwaardelijke steun en vertrouwen was het me echt nooit gelukt Bedankt voor alles' Dan zijn er nog de mensen die ik voor het gemak even zal samenvatten als de 'vriendenclub' Door het voortdurend tonen van belangstelling, en te zorgen voor de broodnodige ontspanning en gezelligheid, hebben jullie wezenlijk bijgedragen aan de totstandkoming van dit proefschrift

117 Curriculum vitae Franciscus Antonius Hol werd op 21 juli 1964 geboren te Oss. Na het behalen van het eindexamen MAVO (1980, de Pelgrim, Oss) en HAVO (1982, Vincent van Gogh, Oss) begon hij 1982 met een opleiding aan de Hogere Laboratorium School te Oss. In het kader van deze opleiding doorliep hij een stage bij de firma Organon International te Oss, op de afdeling General Pharmacology (Dr. D. Meuleman). In 1986 behaalde hij zijn HLO diploma in de richting biochemie waarna hij startte met de studie scheikunde aan de Katholieke Universiteit Nijmegen. In 1987 onderbrak hij deze studie in verband met het vervullen van de dienstplicht (1987-1988) waarna hij de studie hervatte en in 1990 slaagde voor het verkort doctoraal examen scheikunde. Gedurende zijn doctoraal deed hij een hoofdvak biochemie waarbij hij moleculair biologisch onderzoek deed naar specifieke membraan eiwitten en beta- crystallines in de ooglens (Prof. Dr. Bloemendal). Sinds 1991 is hij als wetenschappelijk medewerker werkzaam op de afdeling Anthropogenetica van het Academisch Ziekenhuis "St. Radboud" te Nijmegen waar het in dit proefschrift beschreven onderzoek is uitgevoerd.

118 List of publications

Silahtaroglu-A; Hol-FA; Jensen-РКА; Erdel-M; Duba-НС; Geurds-MPA; Knoers-NVAM; Manman- ECM; Tiimer-Z; Utermann-G Wirth-J; Bugge-M;Tommerup-N. Molecular cytogenetic detection of 9q34-breakpoints associated with Nail-Patella syndrome. Eur J Hum Genet, in press

Momsson-K; Papapetrou-C; Hol-FA; Manman-ECM; Lynch-SA; Bum-J; Edwards-YH. Susceptibility to spina bifida; an association study of five candidate genes. Ann Hum Genet (1998)62:379-396

Redolfi-E; Montagna-C; Mumm-S; Affer-M; Susani-L; Reinbold-R, Hol-FA; Vezzoni-P; Cimino-M; Zucchi-I. Identification of CXorfl, a novel intronless gene in Xq27.3, expressed in human hippocampus. DNA Cell Biol (1998) 17:1009-1016

Joosten-PHLJ; Hol-FA; van Beersum-SEC; Peters-Η; Hamel-BCJ; Afink-GB; van Zoelen-EJJ; Manman-ECM; Altered regulation of platelet derived growth factor receptor-alpa-gene-transcription in vitro by spina bifida- associated mutant Paxl proteins. Proc Natl Acad Sci USA (1998)95: 14459- 14463

Momson-K; Edwards-YH; Lynch-SA; Bum-J; Hol-FA; Manman-ECM Methionine synthase and neural tube defects. J Med Genet (1997) 34:958

Hol-FA; van der Put-NMJ; Geurds-MPA; Heil-SG; Hamel-BCJ; Manman-ECM; Blom-HJ. Molecular analysis of the human tnfunctional enzyme MTHFD (Methylenetetrahydrofolate Dehydrogenase- Methenyltetrahydrofolate Cyclohydrolase- Formyltetrahydrofolate synthetase) in patients with neural tube defects. Clinical Genet (1998) 53:119-125

Hol-FA; Geurds-MPA; Cremers-C; Hamel-BCJ; Manman-ECM Identification of two PAX3 mutations causing Waardenburg syndrome, one within the paired domain (M62V) and the other downstream of the homeodomain (Q282X). Hum Mutat (1998) Suppl:S145-S147

Momson-K; Papapetrou-C; Attwood-J; Hol-FA; Lynch-S; Sampath-A; Hamel-BCJ; Bum-J; Sowden- J; Stott-D; Manman-ECM; Edwards-Y. Genetic mapping of the human homologue (Τ) of mouse T(Brachyury) and a search for allele association between human Τ and spina bifida. Hum Molec Genet (1996) 5:669-674

Hol-FA; Geurds-MPA; Chatkupt-S; Shugart-YY; Balhng-R; Schrander-stumpel-CTRM; Johnson- WG; Hamel-BCJ; Manman-ECM. PAX genes and neural tube defects: an amino acid substitution in PAX1 in a patient with spina bifida. J Med Genet (1996) 33:655-660

119 Hol-FA; Hamel-BCJ; Geurds-MPA; Mullaart-RA; Barr-FG; Macina-RA; Manman-ECM. A frames- hift mutation in the gene for PAX3 in a girl with spina bifida and mild signs of Waardenburg syndrome. J Med Genet (1995) 3252-56

Hol-FA; Hamel-BCJ; Geurds-MPA; Hansmann-I; Nabben-FAE; Daniels-O; Manman-ECM. Localization of Alagille syndrome to 20pll.2-pl2 by linkage analysis of a three generation family. Hum Genet (1995) 95687-690

Chatkupt-S; Hol-FA; Shugart-YY; Geurds-MPA; Stenroos-ES; Koenigsberger-MR; Hamel-BCJ; Johnson-WG, Manman-ECM. Absence of linkage between familial neural tube defects and PAX3 gene. J Med Genet (1995) 32:200-204

Hol-FA; Geurds-MPA; Jensson-O; Hamel-BCJ; Moore-GE; Newton-R; Manman-ECM. Exclusion mapping of the gene for X-linked neural tube defects in an Icelandic family. Hum Genet (1994)93: 452-456

Jamieson-CR; van-der-Burgt-I; Brady-AF; van-Reen-M; Elsawi-MM; Hol-FA; Jeffery-S; Patton-MA; Manman-ECM. Mapping a gene for Noonan syndrome to the long arm of chromosome 12. Nat Genet (1994)8:357-360

Hol-FA; Geurds-MPA; Hamel-BCJ; Manman-ECM. Improving the polymorphism content of the 3' UTR of the human IGFR gene. Hum Molcc Genet (1993) 1:347 van-Rens-GL; Hol-FA; de-Jong-WW; Bloemendal-Η. Presence of hybndizing DNA sequences homologous to bovine acidic and basic beta-crystallms in all classes of vertebrates. J Mol Evol (1991) 33. 457-63

120