A targeted sequencing panel identifies rare damaging variants in multiple genes in the cranial neural tube defect, anencephaly.

Neural tube defects (NTDs) affecting the brain (anencephaly) are lethal before or at birth, whereas lower spinal defects (spina bifida) may lead to lifelong neurological handicap. Collectively, NTDs rank among the most common birth defects worldwide. This study focuses on anencephaly, which despite having a similar frequency to spina bifida and being the most common type of NTD observed in mouse models, has had more limited inclusion in genetic studies. A genetic influence is strongly implicated in determining risk of NTDs and a molecular diagnosis is of fundamental importance to families both in terms of understanding the origin of the condition and for managing future pregnancies. Here we used a custom panel of 191 NTD candidate genes to screen 90 patients with cranial NTDs (n = 85 anencephaly and n = 5 craniorachischisis) with a targeted exome sequencing platform. After filtering and comparing to our in-house control exome database (N = 509), we identified 397 rare variants (minor allele frequency, MAF < 1%), 21 of which were previously unreported and predicted damaging. This included 1 frameshift (PDGFRA), 2 stop-gained (MAT1A; NOS2) and 18 missense variations. Together with evidence for oligogenic inheritance, this study provides new information on the possible genetic causation of anencephaly.

Multifactorial inheritance of neural tube defects: localization of the major gene and recognition of modifiers in ct mutant mice.

Neural tube defects (NTD) in humans have been considered to have a multifactorial aetiology, however the participating genes have not been identified. The curly-tail (ct) mutant mouse develops NTD that resemble the human malformations in location, pathology and associated abnormalities. Moreover, there appears to be multifactorial influence on the incidence of NTD in offspring of curly-tail mice. We now describe a linkage analysis that localizes the ct gene to distal chromosome 4 in mice. Further analysis using recombinant inbred strains demonstrates the presence of at least three modifier loci that influence the incidence of NTD. This study provides definitive evidence for multifactorial inheritance in a mouse model of human NTD.

A gene for autosomal dominant sacral agenesis maps to the holoprosencephaly region at 7q36.

Sacral agenesis is a rare disorder of uncertain incidence that has been reported in diverse populations. Although usually sporadic and most commonly associated with maternal diabetes, there is a hereditary form which may occur in isolation or with a presacral mass (anterior meningocele and/or presacral teratoma) and anorectal abnormalities, which constitute the Currarino triad (MIM 176450). The radiological hallmark of hereditary sacral agenesis is a hemi-sacrum (sickle-shaped sacrum) with intact first sacral vertebra. Bowel obstruction is the usual neonatal presentation, but, unlike other neural tube defects, adult presentation is not uncommon. The major pathology is confined to the pelvic cavity and may present as a space-occupying lesion or meningitis due to ascending infection. All recurrences in families have been compatible with autosomal dominant inheritance except for those associated with the isomerism gene at Xq24-q27.1 (ref. 3). Several associated cytogenetic defects have been reported, including 7q deletions. Previous studies failed to detect linkage to HLA markers, but we now present evidence for a location on 7q36. The same region also contains a gene for holoprosencephaly, an early malformation of the extreme rostral end of the neural tube.

Inositol prevents folate-resistant neural tube defects in the mouse.

Clinical trials demonstrate that up to 70% of neural tube defects (NTDs) can be prevented by folic acid supplementation in early pregnancy, whereas the remaining NTDs are resistant to folate. Here, we show that a second vitamin, myo-inositol, is capable of significantly reducing the incidence of spinal NTDs in curly tail mice, a genetic model of folate-resistant NTDs. Inositol increases flux through the inositol/lipid cycle, stimulating protein kinase C activity and upregulating expression of retinoic acid receptor beta, specifically in the caudal portion of the embryonic hindgut. This reduces the delay in closure of the posterior neuropore, the embryonic defect that is known to lead directly to spina bifida in curly tail embryos. Our findings reveal a molecular pathway of NTD prevention and suggest the possible efficacy of combined treatment with folate and inositol in overcoming the majority of human NTDs.

Embryonic folate metabolism and mouse neural tube defects.

Folic acid prevents 70 percent of human neural tube defects (NTDs) but its mode of action is unclear. The deoxyuridine suppression test detects disturbance of folate metabolism in homozygous splotch (Pax3) mouse embryos that are developing NTDs in vitro. Excessive incorporation of [3H]thymidine in splotch embryos indicates a metabolic deficiency in the supply of folate for the biosynthesis of pyrimidine. Exogenous folic acid and thymidine both correct the biosynthetic defect and prevent some NTDs in splotch homozygotes, whereas methionine has an exacerbating effect. These data support a direct normalization of neurulation by folic acid in humans and suggest a metabolic basis for folate action.

Death before birth: clues from gene knockouts and mutations.

A survey of mouse gene knockouts, transgene insertions and spontaneous mutations that are lethal prenatally reveals that surprisingly few developmental disturbances lead to death of the embryo and early foetus. These disturbances include failure to establish and maintain a vascular circulation, and failure to make the transition from yolk-sac-based to liver-based haematopoiesis. The embryo must also establish gestation-dependent routes of nutritional interaction with the mother, including implantation, formation of a yolk-sac vascular circulation, and formation of a chorioallantoic placenta. A number of embryonic organ and body systems, including the central nervous system, gut, lungs, urogenital system and musculoskeletal system, appear to have little or no survival value in utero.

Cardiovascular defects associated with abnormalities in midline development in the Loop-tail mouse mutant.

Loop-tail (Lp) is a naturally occurring mouse mutant that develops severe neural tube defects. In this study, we describe complex cardiovascular defects in Lp homozygotes, which include double-outlet right ventricle, with obligatory perimembranous ventricular septal defects, and double-sided aortic arch, with associated abnormalities in the aortic arch arteries. Outflow tract and aortic arch defects are often related to abnormalities in the cardiac neural crest, but using molecular and anatomic markers, we show that neural crest migration is normal in Lp/Lp embryos. On the other hand, the heart fails to loop normally in Lp/Lp embryos, in association with incomplete axial rotation and reduced cervical flexion. As a consequence, the ventricular loop is shifted posteromedially relative to its position in wild-type embryos. This suggests that the observed cardiac alignment defects in the Lp mutant may be secondary to failure of neural tube closure and incomplete axial rotation. Double-sided aortic arch is a rare finding among mouse models. In humans, it is usually an isolated malformation, only rarely occurring in combination with other cardiac defects. We suggest that the double-sided arch arises as a primary defect in the Lp mutant, unrelated to the alignment defects, perhaps reflecting a role for the (as-yet-unknown) Lp gene in maintenance/regression of the aortic arch system.

RhoB is expressed in migrating neural crest and endocardial cushions of the developing mouse embryo.

RhoB mRNA expression was examined in the developing mouse embryo between E8.5 and E11.5. Specific expression was found in migrating neural crest (NC) cells, from the first stages of their migration at E9.5, throughout the migration period. Expression is maintained in NC derivatives for at least one embryonic day after they reach their final destinations, but is then down-regulated. RhoB is also expressed in non NC-derived neural tissues, including motor neurones and the floor plate of the neural tube. RhoB mRNA expression is also found in the developing endocardial cushions of the atrioventricular and outflow regions of the developing heart.

Over-expression of the chondroitin sulphate proteoglycan versican is associated with defective neural crest migration in the Pax3 mutant mouse (splotch).

Splotch mice, which harbour mutations in the Pax3 gene, exhibit neural crest-related abnormalities including pigmentation defects, reduced or absent dorsal root ganglia and failure of cardiac outflow tract septation in homozygotes. Although splotch neural crest cells fail to colonise target tissues, they initiate migration in vivo and appear to migrate as well as wild type neural crest cells in vitro, suggesting that the neural crest abnormality in splotch may reside not in the neural crest cells themselves, but rather in the extracellular environment through which they migrate. We have examined the expression of genes encoding extracellular matrix molecules in Sp2H homozygous embryos and find a marked over-expression of transcripts for the chondroitin sulphate proteoglycan versican in the pathways of neural crest cell migration. Use of cadherin-6 expression as a marker for neural crest demonstrates a striking correlation between up-regulation of versican expression and absence of migrating neural crest cells, both in the mesenchyme lateral to the neural tube and in the lower branchial arches of Sp2H homozygotes. Pax3 and versican have mutually exclusive expression patterns in normal embryos whereas, in Sp2H homozygotes, versican is generally over-expressed with 'infilling' in regions that would normally express functional Pax3. Versican, like other chondroitin sulphate proteoglycans, is non-permissive for migration of neural crest cells in vitro, and we suggest that over-expression of this molecule leads to the arrest of neural crest cell migration in splotch embryos. Pax3 may serve to negatively regulate versican expression during normal development, thereby guiding neural crest cells into their pathways of migration.

Abnormalities of floor plate, notochord and somite differentiation in the loop-tail (Lp) mouse: a model of severe neural tube defects.

Mouse embryos homozygous for the loop-tail (Lp) mutation fail to initiate neural tube closure at E8.5, leading to a severe malformation in which the neural tube remains open from midbrain to tail. During initiation of closure, the normal mouse neural plate bends sharply in the midline, at the site of the future floor plate. In contrast, Lp/Lp embryos exhibit a broad region of flat neural plate in the midline, displacing the sites of neuroepithelial bending to more lateral positions. Sonic hedgehog (Shh) and Netrin1 are expressed in abnormally broad domains in the ventral midline of the E9.5 Lp/Lp neural tube, suggesting over-abundant differentiation of the floor plate. The notochord is also abnormally broad in Lp/Lp embryos with enlarged domains of Shh and Brachyury expression. The paraxial mesoderm shows evidence of ventralisation, with increased expression of the sclerotomal marker Pax1, and diminished expression of the dermomyotomal marker Pax3. While the expression domain of Pax3 does not differ markedly from wild-type, there is a dorsal shift in the domain of Pax6 expression in the neural tube at caudal levels of Lp/Lp embryos. We suggest that the Lp mutation causes excessive differentiation of floor-plate and notochord, with over-production of Shh from these midline structures causing ventralisation of the paraxial mesoderm and, to a lesser extent, the neural tube. Comparison with other mouse mutants suggests that the enlarged floor plate may be responsible for the failure of neural tube closure in Lp/Lp embryos.

Retinoic acid induces down-regulation of Wnt-3a, apoptosis and diversion of tail bud cells to a neural fate in the mouse embryo.

The tail bud comprises the caudal extremity of the vertebrate embryo, containing a pool of pluripotent mesenchymal stem cells that gives rise to almost all the tissues of the sacro-caudal region. Treatment of pregnant mice with 100 mg/kg all-trans retinoic acid at 9.5 days post coitum induces severe truncation of the body axis, providing a model system for studying the mechanisms underlying development of caudal agenesis. In the present study, we find that retinoic acid treatment causes extensive apoptosis of tail bud cells 24 h after treatment. Once the apoptotic cells have been removed, the remaining mesenchymal cells differentiate into an extensive network of ectopic tubules, radially arranged around the notochord. These tubules express Pax-3 and Pax-6 in a regionally-restricted pattern that closely resembles expression in the definitive neural tube. Neurofilament-positive neurons subsequently grow out from the ectopic tubules. Thus, the tail bud cells remaining after retinoic acid-induced apoptosis appear to adopt a neural fate. Wnt-3a, a gene that has been shown to be essential for tail bud formation, is specifically down-regulated in the tail bud of retinoic acid-treated embryos, as early as 2 h after retinoic acid treatment and Wnt-3a transcripts become undetectable by 10 h. In contrast, Wnt-5a and RAR-gamma are still detectable in the tail bud at that time. Extensive cell death also occurs in the tail bud of embryos homozygous for the vestigial tail mutation, in which there is a marked reduction in Wnt-3a expression. These embryos go on to develop multiple neural tubes in their truncated caudal region. These results suggest that retinoic acid induces down-regulation of Wnt-3a which may play an important role in the pathogenesis of axial truncation, involving induction of widespread apoptosis, followed by an alteration of tail bud cell fate to form multiple ectopic neural tubes.

Development of a lethal congenital heart defect in the splotch (Pax3) mutant mouse.

OBJECTIVE: The splotch (Sp2h) mutation disrupts the Pax3 gene and is lethal in homozygotes. The aim of the present study was to investigate the cause of lethality. METHODS AND RESULTS: Using the splotch (Sp2H) mouse mutant, we demonstrated that approximately 60% of Sp2H homozygotes die in utero at 13.5-14.5 days of gestation. All these embryos have cardiac malformations involving partial or complete failure of septation of the outflow tract. Although the cause of death in utero is unknown, the dying embryos are edematous, their superior caval veins are over-expanded, and the fetal liver is enlarged and engorged with blood, all signs of cardiac failure. The remaining Sp2H homozygotes die around the time of birth, and these embryos have grossly normal hearts. All Sp2H homozygotes have neural tube defects, either spina bifida, exencephaly, or both. Although these defects clearly do not cause death in utero, they are very likely responsible for the perinatal death of homozygotes that survive to late gestation. There is no correlation between the presence or absence of a cardiac defect and the type of neural tube defect. On the other hand, there is a striking correlation between presence of a cardiac defect and reduction or absence of dorsal root ganglia, which are derivatives of the neural crest. CONCLUSIONS: In this paper, we show that the lethality has a biphasic pattern, and the data strongly suggests that mid-gestation lethality is due to cardiac defects and not the associated neural tube defects. This finding supports the idea that 'conotruncal' cardiac defects involving the ventricular outflow tracts develop as a result of failure of the 'cardiac' neural crest to colonise the developing heart in the mid-gestation embryo, and that the resulting heart defects are solely responsible for the observed mortality.

Mechanisms of normal and abnormal neurulation: evidence from embryo culture studies.

The method of whole embryo culture has been used extensively in analyzing the mechanisms underlying formation of the mammalian neural tube. These studies have provided insight into the cell lineage of the various tissues that comprise the neurulation stage embryo, the role of microfilaments, extracellular matrix and cell proliferation in the morphogenetic events of neural tube closure and the action of specific genes and gene products in establishment of the nervous system. This information is of considerable importance not only as a means of elucidating the processes of normal embryogenesis but also to shed light on the pathogenesis of important human birth defects.

The development of field vole (Microtus agrestis) and mouse blastocysts in vitro: a study of trophoblast cell migration.

Outgrowth of field vole blastocysts and their constituent tissues in vitro results in the appearance of trophoblast giant cell monolayers plus migratory cells of two types. The large migratory cells, which are never seen in mouse outgrowths, resemble trophoblast giant cells in morphology and DNA content. They probably correspond to giant cells which have been observed to migrate throughout the endometrium of vole implantation sites in vivo. It is suggested that their appearance in vitro may depend upon an interaction between the trophectoderm and inner cell mass (ICM) of the blastocyst. Small migratory cells also emerge from vole explants and are occasionally seen in mouse outgrowths. They are probably of ICM origin, and in some cases have given rise to long-term vole cell lines.

Role of the polar trophectoderm in determining the pattern of early post-implantation morphogenesis in mammals: evidence from development of the short-tailed field vole, Microtus agrestis.

The peri-implantational embryogenesis in the field vole, Microtus agrestis, is described. Implantation is interstitial, as it is in the mouse, but egg cylinder formation occurs by invagination of the blastocyst's embryonic pole and not (as in the mouse) by formation of a multilayered extra-embryonic ectoderm. This difference can be attributed to loss in the field vole of the central portion of the polar trophectoderm at the time of blastocyst attachment. In comparing early postimplantation development of mammalian species, three morphogenetic variables should be considered: (i) continued proliferation of polar trophectoderm; (ii) mechanical constraints on the direction of its growth; (iii) variations in the degree to which polar trophectoderm is maintained intact after implantation.

Failure of neural tube closure in the loop-tail (Lp) mutant mouse: analysis of the embryonic mechanism.

Loop-tail (Lp) is unique among mouse mutants in failing to initiate neural tube closure at the cervical/hindbrain boundary (so-called 'Closure 1'), at the 5-7 somite stage. Lp/Lp embryos go on to develop a malformation that closely resembles cranio-rachischisis, the most severe neural tube defect found in humans. We investigated several possible embryological mechanisms that may underlie this failure of neural tube closure in Lp. The genotypes of Lp/Lp, Lp/+ and +/+ embryos from mixed litters were identified using the polymerase chain reaction to amplify a polymorphic microsatellite sequence that is very closely linked to Lp. At post-neurulation stages of development, Lp/Lp embryos have a shortened body axis, which could suggest a defect of axial elongation as the primary anomaly in Lp. However, we found that axial elongation is normal in Lp homozygotes prior to the stage of defective Closure 1, indicating that the shortened body axis of later embryos is a secondary effect of the neurulation anomaly, or an independent effect of the Lp mutation. Some workers have reported cell proliferation rates to be abnormal in later stage Lp/Lp embryos. We observed variations in [3H]thymidine labelling index, and mitotic index, between embryonic tissues, and between embryos at different somite stages. However, Lp/Lp, Lp/+ and +/+ embryos had closely similar cell proliferation parameters, arguing against a mechanism based on faulty embryonic growth. Thirdly, we tested the hypothesis that the defect in loop-tail results from an inability of the neural folds to become apposed, specifically at the site of Closure 1. By tying a silk suture around the embryonic axis, at the future site of Closure 1, we were able to effect convergence of the neural folds at this site. Neural fold closure failed to progress along the body axis in sutured Lp/Lp embryos, however, in contrast to operated Lp/+ and +/+ embryos which exhibited normal progression of neural tube closure. The embryonic defect in loop-tail appears, therefore, 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.

Expression of PACAP, and PACAP type 1 (PAC1) receptor mRNA during development of the mouse embryo.

Pituitary adenylate cyclase activating polypeptide (PACAP) and vasoactive intestinal polypeptide (VIP) have been reported to have a number of neurotrophic effects. We have examined the expression of mRNA for PACAP and PACAP type 1 (PAC1) receptor in the mouse embryo by in situ hybridization and the effects of PACAP and VIP on the growth of mouse embryos in vitro. Although we were unable to detect gross effects of either peptide on the growth rates of embryos maintained in culture, mRNAs for both PAC1 receptor and PACAP peptide were present in the nervous system from day 9.5 of embryonic development. PAC1 receptor mRNA was most abundant in the neural tube and the rhombencephalon and was present also in the dorsal root and trigeminal ganglia and the sympathetic chain. The distribution of mRNA for the PACAP peptide overlapped in part with that of the receptor, but was more extensively distributed in the rhombencephalon and in the developing hypothalamus. Within the neural tube, PAC1 receptor mRNA was located in the roof and floor plates, while the distribution of PACAP peptide mRNA was more complex, being located in two columns of cells in the ventromedial neural tube (consistent with the position of developing autonomic motor neurons) and in cells in the dorsolateral neural tube. These data are concordant with a role for PACAP or a related peptide in neural development.