TERATOLOGY 46:599-604 (1992)

Neural Tube Defects Without Neural Crest Defects in Splotch Mice THOMAS FRANZ Anatomisches Institut, Abteilung fur Neuroanatom ie: Universitats-KrankenhausEppendo$, ZOO0 Hamburg 20, Germany

ABSTRACT Homozygous Splotch mutant mice (SpiSp) die on day 14 of gestation with neural tube defects, curly tail, and malformations of neural crest derivatives. SplH mice, which have a radiation-induced allele of Splotch with a similar phenotype, were used for this study. The neural tube defects are always located in the lumbosacral region and in 50% of the cases also in the region of the hindbrain. In this report, rare cases of neural tube defects and tail defects among the offspring of crosses between Splotch (SplH)heterozygotes are presented, which are not associated with a neural crest defect. This suggests that the development of the neural tube and neural crest defects in this mutant is caused by independent mechanisms o r is dependent on the dosage of the mutant gene, with different thresholds being pathogenetic in the neural tube and neural crest, respectively. 8 1992 Wiley-Liss, Inc. All Splotch (Sp) homozygous mutant embryos show neural tube defects in the lumbosacral region and, in 50% of the cases, also in the hindbrain (Auerbach, '54; Franz, '89). The phenotype of the Splotch homozygous mutant embryo is usually a combination of a neural tube defect with a neural crest defect. The neural crest defect in homozygous mutant embryos (Sp/Sz) and a radiation-induced allele (SplH/Spl ) is characterized by the agenesis of melanocytes, defective formation of spinal ganglia (Auerbach, '54; Moase and Trader, '89), and Schwann cells (Franz, 'go), as well a s malformations of the pharyngeal arch-derived glands and blood vessels, and the septation of the truncus arteriosus (Franz, '89). The latter malformations mimic the phenotype after cranial neural crest ablation in birds (Besson et al., '89; Bockman and Kirby, '84; Nishibatake et al., '87). Homozygotes of the splotch-delayed (Spd) allele show less pronounced defects of the neural crest and a lumbosacral neural tube defect (Dickie, '64; Moase and Trasler, '89). It has thus far been reported that homozygous mutants of all Splotch alleles show a combination of tail deformation, neural tube defect, and neural crest defects. Recently, i t has been emphasized that normal neural tube closure involved changes in 0

1992 WILEY-LISS, INC.

both the neural plate and surrounding tissue (reviewed by Schoenwolf, '90). Disturbances of tissue growth have been demonstrated in the curly-tail mutant, where the uncoordinated growth of neurectoderm, hindgut endoderm, and notochord has been proposed to cause the lumbosacral neural tube defect (Copp et al., '88). Lengthening of the cell cycle (Wilson, '74), reduced growth of the mesoderm and notochord, and disorganisation of the neuroepithelium have been demonstrated in both Splotch and Splotch-delayed (Spd) mutants (Yang and Trasler, '91). As in the Splotch homozygotes both neural tube closure and neural crest formation are affected, one might assume that the Splotch mutation disrupts a common mechanism regulating both events. Most recently, it has been found that Splotch mutants show mutations in the homeodomain of the Pax-3 gene (Epstein e t al., '91) and Waardenburg Syndrome patients, who have neural crest cell-derived defects (Moase and Trasler, '92), also have mutations in the human homologue of Pax3 (Tassabehji et al., '92; Baldwin et al., '92). Pax-3 is expressed in the neural folds, the

Received March 16, 1992; accepted July 27, 1992

600

T.FRANZ

dorsal half of the neural tube, the somites, the developing spinal ganglia, and the mesenchyme of the limb buds (Goulding et al., '91). It is not clear, however, how mutations of the Pax-3 gene could interfere with normal neural tube closure and neural crest differentiation and whether Pax-3 mutations affect both neural tube and neural crest development by the same mechanism. In this report, I resent deviations of the ) mutant phenotype that usual Splotch (Spl€F have occurred albeit rarely in my mouse colony. These findings demonstrate that tail deformations and neural tube defects can occur independently of neural crest defects among the offspring of Splotch (SplH)heterozygote crosses. MATERIALS AND METHODS

Mice SplH is a radiation-induced allele of the Splotch locus that reproduces the phenotype of the spontaneous1 arisen Splotch mutation (Franz, '89). SpXI i + breeding pairs in a (C3H x 101)F, background were obtained from MRC Radiobiology Unit, Harwell, Didcot, England. Heterozygotes have been crossed once with C57BL/6 mice and the nonagouti heterozygote offspring intercrossed for the past 3 years. Estrous females were caged with males overnight and examined for a vaginal plug the morning after. The day on which the plug was detected was termed day 0.5 of gestation. The deviant phenotypes described in this report have been collected over the past 3 years from well over 300 embryos derived from Splotch heterozygote intercrosses, which have been examined for overt alterations of the phenotype. Preparation of specimens Pregnant females were sacrificed by cervical dislocation on the days indicated and the embryos removed. The embryos were dissected from the yolk sac and amnion and fixed either in 4% formaldehyde or in Carnoy's solution overnight. Embryos fixed in Carnoy's solution were dehydrated in absolute ethanol and immersed in methylbenzoate before embedding in paraffin. Sections of 5 pm thickness were obtained on a rotorary microtome (Reichert & Jung, FRG). Sections were stained with hematoxylideosin, dehydrated, and the slides mounted in Eukitt. Pictures were taken on

a Zeiss photomicroscope using Ilford PanF film. RESULTS

Tail defect without neural tube or neural crest defects During 3 years of breeding Splotch mice, eight mice have been born that showed a curly tail. All of these were born from different parents and were, without exception, SplH heterozygotes. The tail was formed like a corkscrew with one complete turn near its base. X-Ray pictures of some of these mice did not show signs of spina bifida occulta (not shown), nor were there any neurological indications of defects in the peripheral nervous system, such as insensitivity t o touch or pain, or motor deficits. One embr 0,which was found in a litter from an Sp' Bheterozygote intercross on day 13.5 of gestation, also showed a curly tail without spina bifida and was serially sectioned. In serial sections, this embryo did not reveal the typical neural crest defect of Splotch homozygotes, but showed normally sized spinal ganglia and Schwann cells in the lumbosacral region (Fig. 1A). Two pairs of SplHheterozygotes with curly tail defects were intercrossed and gave four litters with the expected litter size, but none of the heterozygous live offspring ( n= 28) inherited the tail defect. The tail defect was also not passed onto the next generation by another SplH heterozygous female with a curly tail that had been crossed with an SplH male with a normal tail ( n= 36). Spina bifida without neural crest defect One embryo was found that was macroscopically classified as homozygous, because it showed a small sacral neural tube defect and a bent tail (Fig. 1B). This embryo was the same size as the phenotypically normal littermates and was bigger than homozygous littermates, which showed both a lumbosacral and cranial neural tube defect. Serial sections of this embryo revealed that it did not have the typical neural crest defects observed in Splotch homozygotes. In particular, spinal ganglia, the paravertebra1 autonomic ganglia, and Schwann cells were formed even in the region of the sacral neural tube defect (Fig. 1C). Two other embryos of the same litter, that showed neural tube defects, also exhibited the characteristic neural crest defects.

NEURAL, CREST DEFECT IN SPLOTCH MOUSE

F$. 1. Unusual phenotypes among the offspring of Sp' heterozygote intercrosses. (A) Parasagittal section of an embryo on day 13.5 of gestation. Paraffin, hematoxylideosin, x 38. At the bottom right of the picture, the tail is seen t o bend dorsad, as is usually found in Splotch homozygotes. The arrowheads indicate the presence of spinal ganglia in lumbosacral segments. (B)Two embryos of the same litter on day 13.5 of gestation. One shows a bent tail and sacral neural tube defect (arrowhead). This embryo was tentatively classified as SplH homozygous, because of the neural tube defect. The embryo on the right is phenotypically normal. See also C. (Cf Cross section at a sacral segment of the embryo with

601

neural tube defect, shown in B. Paraffin, hematoxylin: eosin, x 52. Arrowhead indicates spinal ganglia, small arrow indicates paravertebral autonomic ganglia. The neural tube is not closed. tD1 Splotch homozygous embryo on the left, showing both cranial and lumbosacral neural tube defect. This embryo also showed the typical neural crest defects in serial sections. Thc lktermate embryo on the right exhibits only a cranial neural tube defect (arrowhead) and did not reveal a neural crest defect upon histological examination. Note the similar size and neural tube malformation ofthe embryos. Both embryos were removed on day 13.5 of gestation.

602

T.FRANZ

Cranial neural tube defect without lumbosacral neural tube defect or neural crest defect One embryo derived from a cross between SplHheterozygotes was found that showed a cranial neural tube defect, but no lumbosacral neural tube defect. The tail was also formed normally. The embryo had the same size as littermates with both cranial and lumbosacral neural tube defects, but was smaller than the wild-type littermates (Fig. 1D).The exencephaly in this embryo resembled that normally seen in 50% of the Splotch homozygotes. Serial sections of this embryo showed that all spinal ganglia and Schwann cells were formed normally. Cranial nerve ganglia, which are derived from both the neural crest and the ectoderm, were developed normally, as they are in SplHhomozygotes with exencephaly as well. There was no defect in the formation of the pharyngeal archderived glands and the truncus arteriosus was septated normally (not shown).

rying the Splotch mutation, but not in another one, carrying the Extra-toes mutation, although the genetic background was the same. This would indicate that the phenotypic deviations observed among the Splotch offspring in this report were not caused by genes in the genetic background alone, but rather by a synergistic effect of the Splotch mutation with either background genes or epigenetic factors. Deformations of the tail have been observed in many mouse mutants, e.g., curlytail, meander-tail, or flexed-tail (Griineberg, '54; Hollander and Waggie, '77; Hunt et al., '33). These tail malformations resemble the tail abnormality found among some Splotch heterozygotes in my mouse colony. It is improbable, however, that in eight separate cases the tail deformation was caused by independent mutations a t one of the aforementioned loci. Moreover, the tail deformation was not passed onto the next generation. Thus, it may have been caused by a combination of Splotch heterozygosity and unknown epigenetic factors. The observation of curly tails among Splotch offspring DISCUSSION that show no abnormality in neural crestderived structures demonstrates that synerMutations at the Splotch locus usually in- gistic effects, which cause tail deformations terfere with both the normal closure of the in Splotch heterozygotes, do not cause neuneural tube and the formation of neural ral crest defects. crest-derived organs and cell populations. In While the neural tube defect in Spd hothis report, isolated cases are presented, in mozygotes is exclusively found in the lumwhich these two aspects of the Splotch phe- bosacral region (Dickie, ,641, 50% of the Sp notype are segregated. The reported segre- and SplH homozygotes also show a neural gations are, however, unidirectional in that tube defect in the hindbrain region and exthey are neural tube defects without neural encephaly (Auerbach, '54; Franz, '89). A crest defects, but not vice versa. Neural very high frequency of cranial neural tube crest defects alone, however, might have defect without associated caudal neural easily been missed, because they would not tube defect has also been observed in produce externally visible effects in midges- Splotch mutants derived from Splotch hettation embryos. erozygotes that had been crossed with a Thus far, the neural tube defect has been mouse line carrying the In(1)Rkl inversion used as the principal criterion to determine (Moase and Trasler, '87). Among these, even Splotch homozygosity. The In(1)lRK inver- the longevity of homozygotes was prolonged sion on chromosome 1, which includes the to day 18 of gestation. Such hybrid vigour Splotch locus, has been used as a cytogenet- has never been observed in my stock of ical marker for the Splotch mutation (Moase Splotch mice, even though they carry genes and Trasler, '87). The validity of this of three strains, i.e., C3H, 101, and C57BL/ marker is, however, only 98% (Roderick, 6, in their backgrounds. '83).The isolated cases presented in this reIn this report, an isolated case of cranial port have occurred so rarely that a statisti- neural tube defect without lumbar neural cal verification of their genotype using the tube defect is presented. It is interesting to inversion marker would not have been pos- note that this malformation was not associated with a defect in the corresponding neusible. The observed neural tube defects reported ral crest segments. This observation corrobhere occurred only in the mouse colony car- orates the assumption that in Splotch

NEURAL CREST DEFECT IN SPLOTCH MOUSE

mutant embryos the neural tube defects and the neural crest defect are not caused by the same mechanism at the corresponding segmental levels. It is not clear whether the rare segregation of the neural tube defect and the neural crest defect among Splotch offspring presented here reflects a heterogeneity of the Splotch locus. Homozygotes of the Splotchdelayed (Spd)mutation, which unlike SpiSp embryos do not develop cranial neural tube defects on the original inbred background (Dickie, ’64),also show less severe neural crest defects (Moase and Trasler, ’89). On the other hand, “cardiac neural crest” defects in SplHhomozygotes occur irrespective of their showing hindbrain neural tube defects (Franz, ’89). The morphological examination of homozygotes of different Splotch alleles alone thus does not allow speculation about the interdependence of neural tube closure and neural crest formation, and the role of the Splotch locus therein. However, it has been shown that 3.1% of heterozygous embryos (Table 1, Moase and Trasler, ’87) have neural tube defects and also that some heterozygotes without neural tube defects show a significant reduction of lumbosacral spinal ganglia volume at day 15 of gestation (Moase and Trasler, ’89). Thus it is entirely probable that the two exceptional embryos reported in the present study were heterozygotes. The observation that spinal ganglia are reduced in size in Splotch heterozygotes and absent in homozygotes suggests a gene dosage effect of Pax-3 in the developing spinal ganglia. The expression of Pax-3, which is mutated in Splotch mice, is limited to the neural folds and later t o the dorsal half of the neural tube and the adjacent somites, but only some differentiated derivatives of the neural crest, e.g., the developing spinal ganglia (Epstein et al., 1991; Goulding et al., 1991). Given that Pax-3 expression is not detected in neural crest cell populations which are absent in Splotch mutant mice (e.g., melanoblasts), it may be speculated that Pax-3 is involved earlier during ontogeny in delineating the border zone between the surface ectoderm and the neural ectoderm, from where the neural crest originates. Not knowing which genes are regulated by the gene product of the Pax-3 gene, and the expression of which genes is required for the normal closure of the neural tube, it is difficult to speculate how a mod-

603

ulation of the gene dosage of Pax-3 could cause a neural tube defect. The only aspect of the Splotch phenotype that is compatible with postnatal life is the neural crest defect-derived white spotting (as also seen in Waardenburg heterozygotes) and the tail defect, which in this report was found only in Splotch heterozygotes. If the other mutant embryos described in this report were also Splotch heterozygotes, that were phenotypically altered by a combination of epigenetic factors and a genetic bias toward a specific phenotype, one may conclude that this combination favours the production of neural tube defects without concomitant major neural crest defects. In that case, it follows that homozygosity for the Splotch mutation is required in order to produce major defects of the neural crest such as absence of spinal ganglia and melanocytes, but not of the neural tube. ACKNOWLEDGMENTS

I am grateful for the technical support provided by Mrs. S. Schwartz. Special thanks go to Prof. Dr. Richter, who took some X-ray pictures of Splotch heterozygotes. LITERATURE CITED Auerbach, R. (1954) Analysis of the developmental effects of a lethal mutation in the house mouse. J . Exp. Zool., 127t305-329. Baldwin, C.T., C.F. Hoth, J.A. Amos, F.O. daSilva, and A. Milunsky (19921 An exonic mutation in the HUP2 paired domain gene causes Waardenburg’s syndrome. Nature (London),355,637-638, Hesson, W.T., 111, M.L. Kirby, L.H.S. Van Mierop, and J.R. Teabeaut (19861 Effects ofthe size of lesions of the cardiac neural crest at various embryonic ages on incidence and type of cardiac defects. Circulation, 73: 360-364. Hockman, D.E., and M.L. Kirby (19841 Dependence of thymus development on derivatives of the neural crest. Science, 223t498-500, Copp, A.J., F.A. Brook, and H.J. Roberts (1988) A celltype-specific abnormality of cell proliferation in mutant (curly tail) mouse embyros developing spinal neural tube defects. Development, 104:285-295. Epstein, D.J., M. Vekemans, and P. Gros I 1991) Splotch (Sp”H), a mutation affecting development of the mouse neural tube, shows a deletion within the paired homeodomain of Pax-3. Cell, 67t767-774. Franz, T. (1989) Persistent truncus arteriosus in the Splotch mutant mouse. Anat. Embryol., 18Ot457-464. Franz, T. (1990) Defective ensheathment of motoric nerves in the Splotch mutant mouse. Acta Anat., 138: 246-253. Goulding, M.D., G. Chalepakis, U. Deutsch. J.R. Erselius, and P. Gruss (19911 Pax-3, a novel murine DNA binding protein expressed during early neurogenesis. EMBO J., IOt1135-1147.

604

T. FRANZ

Griineberg, H. (1954) Genetical studles on the skeleton of the mouse. VIII Curly-tail. J. Genet., 5252-67. Hollander, W.F., and K.S. Waggie (1977) Meander tail: A recessive mutant located in chromsome 4 of the mouse. J. Hered., 68r403-406. Hunt, H.R., R. Mixter, and D. Permar (1933)Flexed tail in the mouse, Mus musculus. Genetics, 18:335-366. Moase, C.E., and D.G. Trasler (1987) Retinoic acid-induced selective mortality of Splotch-delayed mouse neural tube defect mutants. Teratology, 36t335-343. Moase, C.E., and D.G. Trasler (1989) Spinal ganglia reduction in the Splotch-delayed mouse neural tube defect mutant. Teratology, 40:67-75. Moase, C.E., and D.G. Trasler (1990) Delayed neural crest cell emigration from Sp and Spd mouse neural tube explants. Teratology, 42:171-182. Moase, C.E., and D.G. Trasler (1992) Splotch locus mouse mutants: models for neural tube defects and Waardenburg syndrome type I in humans. J. Med. Genet., 29:145-151. Nishibatake, M., M.L. Kirby, and L.H.S. Van Mierop

(1987) Pathogenesis of persistent truncus arteriosus and dextroposed aorta in the chick embryo after neural crest ablation. Circulation, 75t255-264. Roderick, T.H. (1983) Using inversions to detect and study recessive lethals and detrimentals in mice. In: Utilization of Mammalian Specific Locus Studies in Hazard Evaluation and Estimation of Gentic Risk. F.J. de Serres and W. Sheridan, eds. Plenum, New York, pp. 135-167. Schoenwolf, G.C. (1990) Mechanisms of neurulation: Traditional viewpoints and recent advances. Development, 109.243-270. Tassabehji, M., A.P. Read, V.E. Newton, R. Harris, R. Balling, P. Gruss, and T. Strachan (1992) Nature (London), 3553535-636. Wilson, D. (1974) Proliferation in the neural tube of the Splotch (Sp) mutant mouse. J. Comp. Neurol., 154: 249 -256. Yang, X.-M., and D.G. Trasler (1991) Abnormalities of neural tube formation in pre-spina bifida Splotch-delayed mouse embryos. Teratology, 435434357,