I Med Genet 1992; 29: 145-151

REVIEW ARTICLE

Splotch locus mouse mutants: models for neuraltube defects and Waardenburg syndrome type Iin humans

Connie E Moase, Daphne G Trasler

The majority of data that contribute to ourunderstanding of mammalian embryonic de-velopment are obtained from the study ofanimal models. An increasing availability ofmolecular tools to dissect these model systemshas enabled us to establish parallels withhuman development that range from morpho-genetic similarities to homologous DNA se-quences. The mouse mutant splotch (Sp) haslong been recognised as a model for humanneural tube defects (NTDs) and, morerecently, has become a candidate model forWaardenburg syndrome type I (WSI) inhumans."2 This latter circumstance is basedon similarities in some of the defects in neuralcrest cell (NCC) derivatives that are shared bySp mutants and Waardenburg patients, as wellas the possibility that WSI may be localised tohuman chromosome 2q37, a region known toshare homologies with mouse chromosome 1where the Sp locus is found.

Neurulation occurs between gestation day8-5 and late day 9 in the mouse (plug day= day0) and this is comparable to days 20 to 26 ofhuman embryonic development.3 Neural crestcells originate from the neural tube and beginto emigrate just before cephalic neural tubeclosure, or shortly after trunk neurulation inmammalian embryos. Thus, the two eventsoccur almost simultaneously in development.NCCs then migrate to various regions of thebody and differentiate into a variety of struc-tures including neuronal cells of the sensoryand autonomic ganglia, nerve supportingstructures, mesoectodermal structures, certaincells of the endocrine system, and melano-cytes.45 The fact that both neurulation andNCC emigration are disrupted in the Sp mu-tant makes this a useful model for understand-ing the basis of two fundamental develop-mental processes.

This article reviews the results of investiga-tions that have examined splotch locus mu-tants. These studies include histological ana-lyses at the structural and ultrastructural levels,gene-teratogen interactions, as well as out-comes from immunohistochemical, biochemi-cal, cellular, and molecular approaches. Infor-mation gathered from the analyses of thesemutants has contributed to our understandingof neurulation and neural crest cell emigration,and has provided clues as to how these twofundamental processes may be development-ally related.

Splotch locus allelesORIGIN OF SPLOTCH LOCUS MUTANTSSplotch locus mutants include several allelicvariants that serve as mouse models for bothNTDs and deficiencies in NCC derivatives.The original splotch mutation arose sponta-neously on a C57BL inbred background andwas first described by Russell.6 Since then,Dickie7 reported other de novo splotch muta-tions (Spj, Sp2-, and Sp37), as well as theappearance of an allele splotch delayed (Spd).More recently, Beechey and Searle8 describedthree radiation induced mutations at thislocus, two of which appeared to be the same asthe original Sp mutation (Sp)H and Sp2H), anda third was referred to as splotch retarded(Spr), an allele of Sp and Spd.

Early linkage analysis showed Sp belongedto linkage group XIII and was positionedbetween the coat colour mutation leaden (In)and the hair follicle mutation fuzzy (fz).910Recombination frequencies between Spd andeither In orfz were found to be similar to thoseof Sp,7 supporting the idea that Spd was indeedallelic to Sp. Since then, the splotch locus hasbeen mapped more specifically to band C4 ofchromosome 111 and investigations to definethis locus further using DNA probes are cur-rently in progress.

HOMOZYGOUS FEATURESMutations at the Sp locus are classified assemidominant lethal,7 with differences inhomozygous phenotype being the factor thatdistinguishes between splotch and splotchdelayed. Most Sp homozygotes develop spinabifida (lumbosacral rachischisis) and over halfexhibit exencephaly (cranioschisis) as well,owing to lack of closure in the hindbrainregion.12 On occasion, these mutants may de-velop only a tail flexion defect that results in acurly tail.'3 However, all mutants die in uteroat approximately 13 or 14 days of gestation.'2This is in contrast to the Spd mutant whichdevelops only spina bifida and survives untilbirth, hence undergoing delayed death.7

Unlike Sp and Spd, splotch retarded homo-zygotes are presumed to die before implan-tation.8 This mutation is more severe than Spor Spd as it involves a cytogenetically detect-able deletion of the 1C4 band, which consti-tutes approximately 2% of the total physicallength of chromosome 1.11

Department ofBiology and Centrefor Human Genetics,McGill University,1205 Avenue DrPenfield, Montreal,Quebec, Canada H3AlBi.C E Moase, D G Trasler

Correspondence toDr Trasler.

145 on June 15, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.3.145 on 1 March 1992. D

ownloaded from

Moase, Trasler

HETEROZYGOUS PHENOTYPESp, Spd, and Spr heterozygotes display a simi-lar phenotype that is characterised by a whitebelly patch, feet, and tail tip, although Sprheterozygotes occasionally lack the ventralspotting. This pigmentation defect resultsfrom the failure of neural crest cells to popu-late these regions sufficiently during develop-ment.'2 Other NCC derived structures havebeen shown to be deficient or absent in both Spand Spd mutants as well. These include spinalganglia,2 14 Schwann cells,'5 and NCC derivedstructures of the heart.'6 17 What distinguishesthe Spr heterozygote from Sp and Spd genecarriers is the fact that it experiences an overallgrowth retardation which persists throughoutadulthood.8 However, because this particularmutant has only recently become available, ithas not been characterised as extensively as Spor Spd.

PHENOTYPIC VARIATIONSVariations in expressivity have been obtainedby outcrossing Sp or Spd gene carriers to micewith different genetic backgrounds. Using Spdon a heterogeneous BALB/c stock, Kalter'8showed that each of three teratogens known tocause exencephaly could do so at a higherfrequency in Spd homozygotes than in non-mutant litter mates. Thus, even though Spdhomozygotes do not normally express exence-phaly, they appear to be liable to such anoccurrence. This was further supported by theobservation that exencephaly occurred at afrequency of 6-1% in day 16 Spd homozygousembryos from Fl mice that had originated byoutcrossing Spd inbreds to an In(1)lRk stock.'9

Outcrossing Sp inbreds to In(1)lRk miceresulted in an altered distribution frequency ofNTDs in Sp homozygous embryos as well.The proportions ofNTDs changed from 56%of mutants having exencephaly in addition tospina bifida'2 to 100% of mutant embryoshaving exencephaly, with spina bifida occur-ring only about 25% of the time.'9 In addition,Sp mutants on this background can survivelonger than the usual 13 to 14 days of gestationin an inbred line, with viable embryos beingpresent up to day 18 of gestation.'9

Splotch as a model for human NTDsIn humans, NTDs are primarily the result ofinteractions between genetic and environmen-tal factors. The cellular and molecularmechanisms of neurulation have yet to be fullyunderstood. However, one approach towardsthis end is to examine abnormal neural tubeclosure in genetic models where environmentalfactors can be manipulated. Various NTDmouse mutants such as loop-tail,202' crooked,22rib-fusion,23 and extra-toes,2425 express addi-tional gross anomalies in unrelated organ sys-tems, thus complicating the analysis of NTDpathogenesis. Another model, the curly-tail

2627-mutant, is a recessive mutation with incom-plete penetrance that has yet to be assigned to aparticular chromosome. This, in addition to

the fact that only about 60% of all curly-tailembryos develop an NTD, makes it difficult tocarry out investigations on abnormal neuraltube closure until the development of an NTDis well under way.

In the case of Sp, however, only the skulland vertebral malformations that are associ-ated with an open neural tube are present inconjunction with the defects in certain NCCderivatives, which also originate from theneural tube. Therefore, it is possible that amechanism directly involved in neurulation isthe primary target of the Sp mutation. This,combined with the fact that Sp mutants can beidentified before the manifestation of anNTD,1928 make it a good model for under-standing abnormal neural tube closure.

Pathogenesis of splotch mutant defectsSome of the first studies involving the splotchmutant examined the pathology of the abnor-mally developed neural epithelium. Hsu andVan Dyke,29 as well as Auerbach,'2 reported anextensive overgrowth of neural tissue in openregions of the neural tube. Hsu and VanDyke29 attributed this to an increase in mitoticactivity between 13-5 and 14 5 days of ges-tation. Using tritiated thymidine incorpora-tion, Wilson30 showed that mesencephalic cellsof day 10 and 1 1 Sp embryos undergo a longercell cycle than that found in heterozygous orwild type embryos, and that proliferation isactually decreased in these mutants. However,through analysis of even younger embryos,Kapron-Bras and Trasler3' found that themitotic index was similar between all regionsof day 9 Sp and control embryos. Thus, anydifference in the mitotic index between mu-tants and non-mutants probably reflectssecondary changes following the initial disrup-tion of neural tube closure.

Other histological features of abnormallydeveloped Sp or Spd mutants include neuraltube irregularities throughout the embryo, ab-normal tail morphology, distortions of thebrain lumen'2 with significantly reduced lumensize,32 abnormal otic vesicle differentiation inthose with exencephaly,33 disorganised neuro-epithelial tissue with more intercellular space,34significant reductions in the area of the neuro-epithelium as well as the forebrain,32 andreduced or absent neural crest cell deriva-tives.'2 14 However, because these observationswere from embryos that already exhibited thedefect, it was impossible to determine whetherthey were primary or secondary effects of themutation. In order to understand the aetiologyof Sp locus mutations, it is necessary to ex-amine embryos before the appearance of thedefect.

Aetiology of splotch NTDsNeuropore measurements of embryos derivedfrom intercrosses of Sp or Spd heterozygotesshow that presumptive homozygous mutantshave longer posterior neuropores than theirnon-mutant heterozygous and wild type litter

146

on June 15, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.3.145 on 1 M

arch 1992. Dow

nloaded from

Splotch locus mouse mutants

mates. In addition, presumptive Sp mutantshave longer anterior neuropores.3536 This delayin neuropore closure was postulated to be theunderlying factor that predisposed Sp genecarriers (that is, heterozygotes), which aremorphologically normal, to develop NTDsupon exposure to a teratogenic agent'637 (seesection on Splotch and retinoic acid, below).These results were later corroborated by usingembryos that were positively identified (seesection on Chromosomal markers, below) asSp gene carriers.'8 Similarly, this delay inposterior neuropore closure has been shown tooccur in the curly-tail mutant discussed pre-viously.39 However, the mechanism respon-sible for this delay in closure in any of thesemutants has yet to be elucidated.

Markers for splotch locus mutantsGRAFTED EMBRYONIC ECTODERMAs mentioned above, differentiation betweenmutant and non-mutant embryos before theexpression of a malformation is necessary inorder to identify potential primary causal fac-tors of abnormal development. One method ofdistinguishing Sp mutants from their hetero-zygous and wild type litter mates involvesisolating dorsal ectodermal tissue togetherwith its underlying mesoderm from day 9embryos, and implanting it into eitherembryonic chick coelom, or the anterior eyechamber of adult albino mice.'2 Grafted ecto-derm from Sp homozygotes fails to generatepigment after 16 days, whereas tissues fromheterozygous and wild type embryos producepigmentation. Using this technique to identifyday 9 Sp mutant embryos, Wilson and Finta44showed the frequent occurrence of gap junc-tional vesicles in the lumbosacral region of theneural groove that subsequently fails to close.Gap junctional vesicles, which are normallyabundant in C57B1/6J embryos a day earlier ingestation,4' are suggested to represent a break-down of gap junctions, and thus may be anindicator of decreased cell-cell communicationin the Sp mutant at this stage of development.

CHROMOSOMAL MARKERSRobertsonian translocation Rb (1.3)1BnrAlthough Sp mutants can be identified byeither skin grafts or the length of their neuro-pores relative to non-mutant embryos at thesame developmental stage, it is not possible todistinguish between heterozygous and wildtype embryos by these methods. Thus, it wasnecessary to develop a more specific marker.Kapron-Bras et at42 devised a breeding schemethat introduced a 1.3 Robertsonian transloca-tion (a fusion of chromosomes 1 and 3) into theSp line. This could then be used as a markerfor the chromosome carrying the wild typeallele. However, using this system to predictthe genotype provided 80% reliability inheterozygotes, but only 60% accuracy inhomozygotes owing to cross over. Therefore, amore dependable genotyping strategy wasrequired.

Inversion In(1) IRkA more reliable marker was established usingthe In(l)lRk mouse line which is homozygousfor a paracentric inversion that spans the Splocus. Using an appropriate breeding design(fig 1), Moase and Trasler'9 showed that theaccuracy of this marker was greater than 98%owing to recombination suppression by theinversion. Because this inverted segmentencompasses certain biochemical loci that dif-fer in isotype from the splotch line,4' and alsoincludes 42% of the total length of chromo-some 1,4" the genotype of each individual em-bryo can be established by one oftwo methods.Individual embryos at day 11 of gestation orolder can be genotyped on the basis of theirisocitrate dehydrogenase profile'9 (fig 2), whileyounger embryos can be identified using cyto-genetic analysis28 (fig 3). Until the Sp locus orclosely linked markers are identified, thismarking scheme involving the In(l)lRk inver-sion is the most reliable method to date fordifferentiating between mutant, heterozygous,and wild type embryos.One aspect that is revealed with the use of

this marker is the fact that 4% of day 16 Spdmice, and 9% of day 9 Sp mice which exhibitan NTD are actually heterozygotes'9 (unpub-lished observations). Using this system, how-ever, it is not possible to determine whetherNTDs are found in inbred heterozygotes aswell, or only when these mice are outcrossed tothe In(1)1Rk stock.

Gene-teratogen interactionsNumerous studies have examined the effect ofteratogens on mice that are carriers of a mutantgene. Embryos heterozygous for a recessive orsemidominant mutant gene often express thehomozygous phenotype when exposed to aspecific teratogen. This positive effect may beindicative of possible shared mechanisms

Parents:(Chromosome 1) Sp/+

SPI1l

F1

F2

In(1)l Rk

IX+ t+

Spt; +11(Discard)

Intercross

Sp jSp Sp l

Mutant Heterozygote Wild type

Figure 1 Breeding scheme to obtain biochemically andcytogenetically marked splotch mice (chromosome 1).The hatched region contains the cytogeneticallydetectable In (1)IRK inversion, and also carries theisocitrate dehydrogenase b (Idh-lb) isozyme. These aremarkers for the wild type allele in F2 mice, while solidchromosomes carry the Idh-la isozyme, and indicate thepresence of the Sp allele in F2 mice. (AdaptedfromMoase and Trasler.'9)

147

on June 15, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.3.145 on 1 M

arch 1992. Dow

nloaded from

Moase, Trasler

Figure 2 Isocitrate dehydrogenase profiles fromindividual day 1 1 Sp/Sp, Spl + , and + / + embryos.The slow migrating band (aa) in lane I is from a Sphomozygote, lane 2 is heterozygous for Idh-P andIdh-1' (ab) isotypes and is from a Sp heterozygote, lane3 contains the fast migrating variant (bb) and is from a+ / + embryo. (O= origin; from Moase and Trasler.'9)

through which mutant genes and teratogensexert their effects. Cole and Trasler45 showedthat progeny obtained from a cross of eithercrooked or rib-fusion heterozygotes to non-

mutant strains were more susceptible to NTDsafter in utero exposure to teratogenic doses ofinsulin than were progeny from intercrossesbetween non-mutants. Other examples of pos-itive interactions include 5-fluorouracil withthe limb mutants luxoid and luxate,46 trypanblue and NTDs in brachyphalangy carriers,47trypan blue or actinomycin D with tail malfor-mations in brachyury gene carriers,4849 and 6-aminonicotinamide with cleft lip in dancerheterozygotes .50

SPLOTCH AND RETINOIC ACID (RA)RA induction of NTDsRetinoic acid (RA), a vitamin A analogue, hasbeen extensively studied with respect to itsinteraction with splotch and splotch-delayed.A positive gene-teratogen interaction has beenshown for carriers of the Sp gene in vivo3637 as

well as in vitro using the whole embryo culture

Figure 3 A chromosomal spread Jrom a cytogeneticallymarked day 9 Spd heterozygous embryo. G bandingshows the pattern associated with the chromosome I thatcarries the Spd (or Sp) allele (a), and the chromosome1 derivedfrom the In(l)lRk line which carries theinversion and, hence, the wild type allele (b). (FromMoase and Trasler.28)

system.38 However, no positive interactionwith retinoic acid was observed for carriers ofSpd.35 In an extensive histological analysis todetermine whether Sp and retinoic acid wereacting through the same mechanism to pro-duce NTDs, Kapron-Bras and Trasler3'showed that the Sp gene is associated withreducing both the number of NCCs releasedfrom the neural tube, and the amount of extra-cellular space surrounding the neural tube,including the area between the neural tube andsurface ectoderm. On the other hand, retinoicacid caused disruptions in the spatial relation-ship between the notochord and neural tube,as well as in the shape of the neural tube. Thus,retinoic acid and the Sp gene affect differentaspects of neurulation, but combine to induceNTDs in a greater degree than either factor byitself.

RA reduction of NTDsAlthough treatment of non-mutant embryoswith 30 to 60 mg/kg of retinoic acid on day 8 ofgestation causes NTDs,37 it was also notedthat, as with the curly-tail mutant,5' low reti-noic acid doses of 5 mg/kg administered oneday later in gestation (day 9) significantlyreduces the incidence of NTDs in litters fromSp heterozygote intercrosses, without increas-ing the resorption frequency.52 When thisstudy was repeated using genotypically identi-fiable Spd embryos, it was found that thereduction in NTD frequency was actually at-tributable to retinoic acid induced selectivemortality of mutant embryos, and that thisinduced mortality was not enough to increasethe resorption frequency significantly.'9 Ingenotypically identifiable Sp embryos, theeffect was somewhat different. Rather thanreducing the NTD frequency, these low RAdoses appeared to induce NTDs in day 11heterozygotes. However, there was a higherincidence of developmentally retarded, mal-formed embryos in the treated group, suggest-ing that even a low dose of RA was indeeddetrimental to embryonic development andviability.

Studies involving the curly-tail mutant haveshown that administration of the DNA inhibi-tor mitomycin C,53 or maternal food depriva-tion during embryonic neurulation,54 signific-antly reduces the NTD frequency withoutincreasing the resorption frequency. Althoughthe effect of DNA inhibitors has not beentested in Sp locus mutants, a pilot studyby Mehin and Trasler69 indicates that fooddeprivation does not reduce the overallNTD frequency in litters obtained fromSpd/ + intercrosses. This, in conjunction withother differences, indicates that Sp and curly-tail mutants do not share a common aetiologi-cal basis for NTDs other than the fact thatboth experience delays in neural tube closure.

Splotch extracellular matrixcomponentsAccording to histological observations,3' theclosed neural tube in the trunk region of Sp/Sp

148 on June 15, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.3.145 on 1 March 1992. D

ownloaded from

Splotch locus mouse mutants

embryos is in close apposition to the overlyingectoderm, and, unlike control sections, hasvery little extracellular space between the twotissues. Because fewer NCCs were observed inthese regions compared to controls, it wassuggested that the lack of extracellular spacecould be due to alterations in extracellularmatrix components and that this, in turn, mayinhibit NCC emigration from the neural tube.

Several immunohistochemical and ultra-structural analyses have focused on the tem-poral and spatial localisation of specific extra-cellular matrix constituents in Sp and Spdmutants. Using late day 9 or day 11 Spdembryos that exhibited the mutant phenotype,O'Shea and Liu55 found that the basal laminaof the neuroepithelium was disorganised, andthat there was less laminin and collagen IV,and more fibronectin and heparan sulphateproteoglycan (HSPG) in the dorsolateral re-gion of the neural tube compared to controls.Fibronectin and HSPG were also found to bedisplaced in the mutants during secondaryneurulation. McLone and Knepper56 alsoobserved alterations in certain glycoconjugatesand glycosaminoglycans in day 10 Sp mutants.Using enzymatic digestion to quantitate par-ticular extracellular matrix components, theseauthors reported that the distribution of hya-luronic acid (HA) and chondroitin sulphate(CSPG) was approximately equal in mutantneuroepithelium, whereas in normal embryosHA is predominant in the open neural tubeand CSPG is predominant in the closed neuraltube. McLone and Knepper56 also found dif-ferences in lectin staining patterns between Spmutants and controls, with a persistent andintense concanavalin-A staining on the abnor-mal luminal surface.

Other studies have examined Sp and Spdmutant embryos at slightly earlier stages ofdevelopment in order to assess extracellularmatrix aspects before the abnormal phenotypeis apparent. These embryos were identified asmutant by the In(1)lRk marker. Trasler andMorriss-Kay57 observed greater amounts ofboth CSPG and HSPG antibody fluorescencein the neuroepithelial basement membraneof cranial and caudal sections from 5 to 15somite stage Sp and Spd mutant embryoscompared to non-mutants. In an extensiveultrastructural and morphometric analysis ofcaudal Spd mutant sections at 15 to 18 somitestage of development, Yang and Trasler58observed that the basal lamina of the overlyingsurface ectoderm was not as well formed asthat of controls, that there were fewer meso-dermal cells in mutant embryos, and that theneuroepithelium was disorganised and con-tained a greater amount of intercellular space.This latter observation corroborated similarfindings in slightly older Sp mutants.34 Fur-thermore, it showed that these morphogeneticdifferences were present even before the de-fects become apparent, suggesting that one ofthe primary effects of the mutant gene is toelicit these changes in the neuroepithelium. Incontrast, Yang and Trasler58 did not observethe significant differences in gap junctionalvesicle formation between Spd mutants and

controls that were reported by Wilson andFinta4O from their examination of early day 9Sp mutants. Sp mutants have been shown tobe more severely affected than Spd mutants inother respects, such as viability, type ofNTD,and reductions in spinal ganglia and emigrat-ing NCCs. Thus, it is not surprising thatdifferences in numbers of gap junctional ves-icles are observed between Sp and Spd as well.

In vitro analysis of splotch NCCsAlthough NTDs occur in most embryos thatare homozygous for Sp or Spd, genotypicanalysis has shown that 11% of mutant em-bryos do not exhibit such a defect. However,histological examination showed that theseparticular embryos did have severe reductionsor absences of NCC derived spinal ganglia.14This indicated that defects observed in NCCderivatives are not the result ofNTDs in thesemutants. In fact, morphometric analysisshowed that even some heterozygous embryoshave significantly reduced volumes of spinalganglia compared to wild type embryos, butpresumably this deficiency is not detrimentalto the survival of heterozygotes. Thus, sinceone defect was not contingent on the others'occurrence, it was possible that neural tubeclosure and the release of NCCs from theneural tube may instead share a regulatoryevent that, if disrupted, could result in thedefects seen in both developmental pathways.Kapron-Bras and Trasler3' have shown that

few NCCs are released from mutant neuraltubes. However, it was unclear as to whetherthis occurred as a result of an abnormality inthe neuroepithelium, the ECM into whichNCCs emigrate, or whether the NCCs them-selves were at fault. Examination of NCCemigration from explanted neural tubes28showed that there was approximately a 24 hourdelay in the release of NCCs from mutantneural tubes compared to non-mutants, withSp being more severely affected than Spd.Provision of an enriched three dimensionalextracellular matrix containing laminin, colla-gen IV, HSPG, and entactin failed to enhanceNCC release significantly, and assessment ofmitotic indices showed no difference in NCCproliferation between mutant and non-mutantgenotypes. Therefore, these findings indicatedthat the neuroepithelium from which NCCsarise may be faulty in the Sp mutant withrespect to the mechanism involved in the re-lease of NCCs.28

Cell adhesion molecules (CAMs) insplotchSince specific cell adhesion molecules areknown to be involved in neural development,it is of interest to examine some of these withrespect to Sp. Using immunofluorescent tech-niques, Moase and Trasler59 found that neuro-epithelial tissue in sections from mutant em-bryos fluoresced with greater intensity inresponse to antibodies against the neural celladhesion molecule N-CAM than did controlsections. Further analysis by immunoblotting

149

on June 15, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.3.145 on 1 M

arch 1992. Dow

nloaded from

150

showed that both Sp and Spd mutants, as wellas day 9 heterozygous embryos, exhibitedaltered N-CAM profiles compared to wildtype embryos. These alterations may involve achange in the conversion of high molecularweight N-CAM to lower molecular weightforms, similar to the observation in anotherneurological mutant, staggerer.60 However, inthe case of splotch, this difference is apparentearly in development rather than postnatally.N-CAM is coded by a gene localised to chro-mosome 9 in mice. Therefore, this alteration insplotch probably relates to regulation orpost-translational processing of the N-CAMprotein, which may or may not be directlyinfluenced by the Sp locus on chromosome 1.In view of the fact that N-CAM copurifieswith certain ECM components, includingHSPG and CSPG,6162 which appear necessaryfor N-CAM mediated cell adhesion, it is pos-sible that an altered N-CAM species couldbring about the cascade of abnormalitiesobserved in Sp locus mutants. However,further investigation is necessary to substan-tiate this idea.

Present and future directionsIn addition to investigations that analyse Splocus mutations at the cellular level, progress isalso being made towards defining this locus atthe molecular level. Carriers of the Spr alleleare useful in this respect owing to the largedeleted segment at this locus. Four geneswhich fall within this deleted region have beenidentified63; however, these markers are pre-sent in the other allelic mutants. Thus, it isnecessary to find more markers and to establisha multilocus linkage map of this region in orderto ascertain homologous sequences betweenmouse and man. Foy et all have tentativelyestablished linkage between Waardenburgsyndrome type I and placental alkaline phos-phatase, which is localised to human chromo-some 2q37. These findings have been corro-borated by another group that examined asingle large WSI pedigree.2 This is interestingin view of the fact that chromosome 2q37 isknown to share homologies with the regioncontaining the splotch locus on mouse chro-mosome 1. The fact that phenotypic similari-ties occur in splotch mice and WSI patients(that is, pigmentation disturbances and occa-sional NTDs in WSI subjects6468) suggeststhat Waardenburg syndrome may be thehuman counterpart to the mouse mutantsplotch. In addition, continuing studiesinvolving molecular analysis of the Sp locusmay eventually show mouse-human homolo-gies in DNA sequences that are associated withthe development of NTDs.

Note added in proofRecently Pax-3, a murine DNA bindingprotein expressed during early neurogenesiswas mapped to mouse chromosome 1.70 Pax-3was found to be deleted in Spr/+ mice and theSp2H allele was shown to have a 32 nucleotide

deletion within the paired homeo domaincoding portion of Pax-3.71 Therefore, Pax-3must have a key role in neurulation. Further,Pax-3 is the mouse homologue of the humanHuP2 gene,72 which suggests the latter maymap near 2q37 and may be altered in WSI.

This research was supported by NaturalSciences and Engineering Research Council(NSERC) and Medical Research Council(MRC) postgraduate studentships (CEM) aswell as an MRC (Canada) research grant(DGT).

1 Foy C, Newton V, Wellesley D, Harris R, Read A. Assign-ment of the locus for Waardenburg syndrome type I tohuman chromosome 2q37 and possible homology to thesplotch mouse. Am J Hum Genet 1990;46:1017-23.

2 Asher JH Jr, Morell R, Friedman TB. Waardenburg syn-drome (WS): the analysis of a single family with a WSImutation showing linkage to RFLP markers on humanchromosome 2q. Am J Hum Genet 1991;48:43-52.

3 Rugh R. The mouse. Its reproduction and development. Min-neapolis: Burgess Publishing, 1968:304-5.

4 Weston JA. The migration and differentiation of neuralcrest cells. Adv Morphogenet 1970;8:41-114.

5 LeDouarin N. Migration and differentiation of neural crestcells. Curr Top Dev Biol 1980;16:31-85.

6 Russell WL. Splotch, a new mutation in the house mouseMus musculus. Genetics 1947;32:107A.

7 Dickie MM. New splotch alleles in the mouse. 7 Hered1964;55:97-101.

8 Beechey CV, Searle AG. Mutations at the Sp locus. MouseNews Letter 1986;75:28.

9 Snell GD, Dickie MM, Smith P, Kelton DE. Linkage ofloop-tail, leaden, splotch and fuzzy in the mouse. Heredity1954;8:271-3.

10 Parsons PA. A balanced four-point linkage experiment forlinkage group XIII of the house mouse. Heredity1958;12:77-95.

11 Evans EP, Burtenshaw MD, Beechey CV, Searle AG. Asplotch locus deletion visible by Giemsa banding. MouseNews Letter 1988;81:66.

12 Auerbach R. Analysis of the developmental effects of alethal mutation in the house mouse. 7 Exp Zool1954;127:305-29.

13 Russell WL, Gower JS. Offspring from transplanted ovar-ies of fetal mice homozygous for a lethal gene (Sp) thatkills before birth. Genetics 1950;35:133.

14 Moase CE, Trasler DG. Spinal ganglia reduction in thesplotch-delayed mouse neural tube defect mutant. Tera-tology 1989;40:67-75.

15 Franz T. Defective ensheathment of motoric nerves in thesplotch mutant mouse. Acta Anat (Basel) 1990;138:246-53.

16 Effmann EL, Whitman SA. Cardiovascular defects in mur-ine homozygous splotch embryos. Teratology1986;33:36C.

17 Franz T. Persistent truncus arteriosus in the splotch mutantmouse. Anat Embryol 1989;180:457-64.

18 Kalter H. Experimental teratological studies with themouse CNS mutations cranioschisis and delayed splotch.7Craniofac Genet Dev Biol 1985;1(suppl):339-42.

19 Moase CE, Trasler DG. Retinoic acid-induced selectivemortality of splotch-delayed mouse neural tube defectmutants. Teratology 1987;36:335-43.

20 Strong LC, Hollander WF. Hereditary loop-tail in thehouse mouse accompanied by imperforate vagina andwith lethal craniorachischisis when homozygous. J Hered1949;40:329-34.

21 Stein KF, MacKensen JA. Abnormal development of thethoracic skeleton in mice homozygous for the gene forloop-tail. Am J Anat 1957;100:205-23.

22 Morgan W. A new crooked-tail mutation involving distinctpleiotropism. 7Genet 1954;52:354-73.

23 Theiler K, Stevens LC. The development of rib-fusion, amutation in the house mouse. AmJ Anat 1960;106:171-83.

24 Johnson DR. Extra-toes: a new mutant gene causing multi-ple abnormalities in the mouse. J Embryol Exp Morphol1967;17:543-81.

25 Johnson DR. Brachyphalangy, an allele of extra-toes in themouse. Genet Res 1969;13:275-80.

26 Gruneberg H. Genetical studies on the skeleton of themouse. VIII. Curly-tail. J Genet 1954;52:52-67.

27 Seller MJ, Embury S, Polani PE, Adinolfi M. Neural tubedefects in curly-tail mice. II. Effect of maternal adminis-tration of vitamin A. Proc R Soc Lond fBiolJ1979;206:95-107.

28 Moase CE, Trasler DG. Delayed neural crest cell emig-ration from Sp and Spd mouse neural tube explants.Teratology 1990;42:171-82.

29 Hsu CY, Van Dyke JH. An analysis of growth rates inneural epithelium of normal and spina bifidous (myelo-schisis) mouse embryos. Anat Rec 1948;100:745.

30 Wilson DB. Proliferation in the neural tube of the splotch(Sp) mutant mouse. I Comp Neurol 1974;154:249-56.

31 Kapron-Bras CM, Trasler DG. Histological comparison ofthe effects of the splotch gene and retinoic acid on the

Moase, Trasler

on June 15, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.3.145 on 1 M

arch 1992. Dow

nloaded from

151Splotch locus mouse mutants

closure of the mouse neural tube. Teratology 1988;37:389-99.

32 O'Shea KS, Rheinheimer JST, O'Shea JM. Morphometricanalysis of the forebrain anomalies in the delayed splotchmutant embryo. J7 Craniofac Genet Dev Biol 1987;7:357-69.

33 Deol MS. Influence of the neural tube on the differentiationof the inner ear in the mammalian embryo. Nature1966;209:219-20.

34 Morris GL, O'Shea KS. Anomalies of neuroepithelial cellassociations in the splotch mutant embryo. Dev Brain Res1983;9:408-10.

35 Dempsey EE. Neural tube defects - pathogenesis and gene-teratogen interaction in the mouse. MSc thesis, McGillUniversity, Montreal, 1981.

36 Dempsey EE, Trasler DG. Early morphological abnormal-ities in splotch mouse embryos and predisposition togene- and retinoic acid-induced neural tube defects. Ter-atology 1983;28:461-72.

37 Kapron-Bras CM, Trasler DG. Gene-teratogen interactionand its morphological basis in retinoic acid-inducedmouse spina bifida. Teratology 1984;30:143-50.

38 Kapron-Bras CM, Trasler DG. Interaction between thesplotch mutation and retinoic acid in mouse neural tubedefects in vitro. Teratology 1988;38:165-73.

39 Copp AJ, Seller MJ, Polani PE. Neural tube developmentin mutant (curly-tail) and normal mouse embryos: thetiming of posterior neuropore closure in vivo and in vitro.Embryol Exp Morphol 1982;69:151-67.

40 Wilson DB, Finta LA. Gap junctional vesicles in the neuraltube of the splotch (Sp) mutant mouse. Teratology1979;19:337-40.

41 Wilson DB, Finta LA. Early development of the brain andspinal cord in dysraphic mice: a transmission electronmicroscopic study. Comp Neurol 1980;190:363-71.

42 Kapron-Bras CM, Vekemans MJJ, Trasler DG. A chromo-some marker for the early detection of mouse embryoscarrying the neural tube defect mutation splotch. CanGenet Cytol 1986;28:862-6.

43 Roderick TH. Strain distributions of genetic polymor-phisms in the mouse. In: Heiniger HJ, Dorey JJ eds.Handbook on genetically standardized JAX mice. 3rd ed.Bar Harbor, Maine: The Jackson Laboratory, 1980:2.22-2.30.

44 Davisson MT, Roderick TH. Chromosomal banding pat-terns oftwo paracentric inversions in mice. Cytogenet CellGenet 1973;12:398-403.

45 Cole WA, Trasler DG. Gene-teratogen interaction in insu-lin-induced mouse exencephaly. Teratology 1980;22:125-39.

46 Dagg CP. Combined action of fluorouracil and two mutantgenes on limb development in the mouse. I Exp Zool1967;164:479-89.

47 Johnson DR. Trypan blue and the extra-toes locus in themouse. Teratology 1970;3:105-10.

48 Hamburgh M, Herz R, Landa G. The effect of trypan blueon expressivity of the brachyury gene "T" in mice.Teratology 1970;3:111-8.

49 Winfield JB, Bennett D. Gene-teratogen interaction: poten-tiation of actinomycin D teratogenesis in the house mouseby the lethal gene brachyury. Teratology 1971;4:157-71.

50 Trasler DG, Kemp D, Trasler TA. Increased susceptibilityto 6-amino-nicotinamide-induced cleft lip of heterozy-gote dancer mice. Teratology 1984;29:101-4.

51 Seller MJ, Perkins KJ. Prevention of neural tube defects incurly-tail mice by maternal administration of vitamin A.Prenat Diagn 1982;2:297-300.

52 Kapron-Bras CM, Trasler DG. Reduction in the frequencyof neural tube defects in splotch mice by retinoic acid.Teratology 1985;32:87-92.

53 Seller MJ, Perkins KJ. Effect of mitomycin C on the neuraltube defects of the curly-tail mouse. Teratology1986;33:305-9.

54 Copp AJ, Crolla JA, Brook FA. Prevention of spinal neuraltube defects in the mouse embryo by growth retardationduring neurulation. Development 1988;104:297-303.

55 O'Shea KS, Liu LH. Basal lamina and extracellular matrixalterations in the caudal neural tube of the delayed splotchembryo. Dev Brain Res 1987;37:1 1-26.

56 McLone DG, Knepper PA. Role of complex carbohydratesand neurulation. Pediatr Neurosci 1986;12:2-9.

57 Trasler DG, Morriss-Kay GM. Immunohistochemicallocalization of chondroitin and heparan sulfate proteogly-cans in pre-spina bifida splotch mouse embryos. Terato-logy 1991;44:571-9.

58 Yang XM, Trasler DG. Abnormalities of neural tubeformation in pre-spina bifida splotch-delayed mouse em-bryos. Teratology 1991;43:643-57.

59 Moase CE, Trasler DG. N-CAM alterations in splotchneural tube defect mouse embryos. Development (inpress).

60 Edelman GM, Chuong CM. Embryonic to adult conversionof neural cell adhesion molecules in normal and staggerermice. Proc Natl Acad Sci USA 1982;79:7036-40.

61 Cole GJ, Schubert D, Glaser L. Cell-substratum adhesionin chick neural retina depends upon protein-heparansulfate interactions. J Cell Biol 1985;100: 1192-9.

62 Hoffmann S, Edelman GM. A proteoglycan with HNK-1antigenic determinants as a neuron-associated ligand forcytotactin. Proc Narl Acad Sci USA 1987;84:2523-7.

63 Epstein DJ, Malo D, Vekemans M, Gros P. Molecularcharacterization of a deletion encompassing the splotchmutation on mouse chromosome 1. Genomics 1991;10:89-93.

64 Arnvig J. The syndrome of Waardenburg. Acta Genet StatMed (Basel) 1959;9:41-6.

65 Lavergne MG. Probleme d'eugenisme pose par une familleatteinte du syndrome de Waardenburg-Klein. J GenetHum 1959;10:80-5.

66 Bwibo NO, Mkono MD. Waardenburg's syndrome in anAfrican child. Hum Hered 1970;20:19-22.

67 Roux CH, Baheux G, Gaulier M, Caldera R, Soepardan L.Une observation familiale portent sur quatre generationset vingt trois sujets. Ann Genet (Paris) 1970;13:125-8.

68 David TJ. Waardenburg's syndrome in two siblings andtheir parents. Hum Genet 1971;14:81-2.

69 Mehin N, Trasler DG. Possible prevention of neural tubedefects in splotch-delayed (Spd/Spd) by food deprivation.Mouse Genome 1991;89:556.

70 Goulding MD, Chalepakis G, Deutsch U, Erselius JR,Gruss P. Pax-3, a novel murine DNA binding proteinexpressed during early neurogenesis. EMBO J7 1991;10:1135-47.

71 Epstein DJ, Vekemans M, Gros P. Splotch (Sp"H), a muta-tion affecting development of the mouse neural tube,shows a deletion within the paired homeo domain ofPax-3. Cell 1991;67:Nov 15.

72 Burri M, Tromvoukis Y, Bopp D, Frigerio G, Noll M.Conservation of the paired domain in metazoans and itsstructure in three isolated human genes. EMBO J1989;8:1 183-90.

on June 15, 2020 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.29.3.145 on 1 M

arch 1992. Dow

nloaded from