INTRODUCTION

Regulation of axial polarity in the vertebrate central nervoussystem (CNS), somite and limb appears to depend on intercel-lular signals. For example, in the CNS, contact-mediatedsignals emanating from the underlying notochord are thoughtto initiate the development of a specialized population ofventral midline cells, the floor plate, in overlying neuralectoderm (Yamada et al., 1991; Placzek et al., 1993). Dif-fusible signals from the floor plate and notochord then inducethe formation of motor neurons on either side of the floor plate(Yamada et al., 1991, 1993). Similarly, the notochord and floorplate appear to be the source of signals responsible for sclero-tomal development in the somite (Porquie et al., 1993; Brand-Saberi et al., 1993). Thus, ventral polarity in the CNS andsomite is regulated by these two signaling centres. In the limb,a signaling centre localized in posterior mesenchyme, the zoneof polarizing activity (ZPA), controls anterior-posteriorpolarity (Saunders and Gasseling, 1968). Several lines ofevidence suggest that the ZPA is the source of a diffusiblefactor or factors whose concentration determines polarity: high

concentrations for posterior structures, low concentrations formore anterior limb elements (Tickle, 1981). Interestingly, ZPAactivity is found in the notochord and floor plate (Wagner etal., 1990), suggesting that common signals may regulate pat-terning of the CNS, somites and limbs.

Recent studies have identified a family of vertebrate genesencoding putative signaling molecules related to the Drosophilasegment polarity gene, hedgehog (hh). One member of thisfamily, Sonic hedgehog (Shh), is expressed in the notochord,floor plate and ZPA of the limb (Echelard et al., 1993; Krausset al., 1993; Riddle et al., 1993; Chang et al., 1994; Roelink etal., 1994). Moreover, functional studies in zebrafish, chick, ratsand mice indicate that Shh may encode a common signal regu-lating polarity in the CNS, somite and limbs. Ectopic expressionof Shh in the CNS in vivo, or application of Shh-expressing cellsor purified peptides to neural explants in vitro, leads toinduction of floor plate and motor neurons (Echelard et al.,1993, Krauss et al., 1993; Roelink et al., 1994; Martí et al.,1995). These results indicate that Shh peptides may mediateboth short-(floor plate) and long-(motor neuron) rangeinduction in the vertebrate CNS. In the somite, similar

2537Development 121, 2537-2547 (1995)Printed in Great Britain © The Company of Biologists Limited 1995

Sonic hedgehog (Shh) encodes a signal that is implicated inboth short- and long-range interactions that pattern thevertebrate central nervous system (CNS), somite and limb.Studies in vitro indicate that Shh protein undergoes aninternal cleavage to generate two secreted peptides. Wehave investigated the distribution of Shh peptides withrespect to these patterning events using peptide-specificantibodies. Immunostaining of chick and mouse embryosindicates that Shh peptides are expressed in the notochord,floor plate and posterior mesenchyme of the limb at theappropriate times for their postulated patterningfunctions. The amino peptide that is implicated in intercel-lular signaling is secreted but remains tightly associatedwith expressing cells. The distribution of peptides in theventral CNS is polarized with the highest levels of proteinaccumulating towards the luminal surface. Interestingly,Shh expression extends beyond the floor plate, into ventro-lateral regions from which some motor neuron precursors

are emerging. In the limb bud, peptides are restricted to asmall region of posterior-distal mesenchyme in close asso-ciation with the apical ectodermal ridge; a region thatextends 50-75 µm along the anterior-posterior axis.Temporal expression of Shh peptides is consistent withinduction of sclerotome in somites and floor plate andmotor neurons in the CNS, as well as the regulation ofanterior-posterior polarity in the limb. However, we canfind no direct evidence for long-range diffusion of the19×103 Mr peptide which is thought to mediate both short-and long-range cell interactions. Thus, either long-rangesignaling is mediated indirectly by the activation of othersignals, or alternatively the low levels of diffusing peptideare undetectable using available techniques.

Key words: Shh peptides, CNS patterning, limb patterning, chick,mouse

SUMMARY

Distribution of Sonic hedgehog peptides in the developing chick and mouse

embryo

Elisa Martí1, Ritsuko Takada1,†, David A. Bumcrot1, Hiroshi Sasaki2 and Andrew P. McMahon1,*1Department of Molecular and Cellular Biology, The BioLabs, Harvard University, 16 Divinity Avenue, Cambridge, MA 02138, USA2Laboratory of Developmental Biology, Institute for Molecular and Cellular Biology, Osaka University, 1-3 Yamada-oka, Osaka565, Japan

*Author for correspondence†Present address: Center for Molecular and Developmental Biology, Kyoto University, Kitashirakawa, Sakyo-Ku, Kyoto 606-01, Japan

2538

approaches have shown that Shh can induce the expression ofPax-1, a sclerotomal marker (Fan and Tessier-Lavigne, 1994;Johnson et al., 1994), whilst in the limb, Shh-expressing cellsplaced at the anterior margin result in mirror-image duplicationsof digits, similar to those produced in response to grafts of ZPAcells (Riddle et al., 1993; Chang et al., 1994).

These experimental studies are consistent with Shh playinga major role in patterning, acting as a short -(floor plate) andlong-(motor neuron, limb and sclerotome) range signal.Similarly, Drosophila Hh has been proposed to act locally, inlimb development, as well as at a distance, in cellular pattern-ing of the epidermis and in driving morphogenesis of the eye(Ingham, 1993; Basler and Struhl, 1994; Heemskerk andDiNardo, 1994; Tabata and Kornberg, 1994). These results aresupported by antibody studies which indicate that Hh issecreted and can be detected a few cell diameters beyond cellstranscribing the hh gene (Taylor et al. 1993; Lee et al. 1994;Tabata and Kornberg, 1994).

In the vertebrate, processing and secretion of Shh has beenaddressed in a variety of cultured cells. The primary Shh trans-lation product enters the secretory pathway, undergoes signalpeptide cleavage and glycosylation at a single conserved N-linked glycosylation site and is cleaved into two peptides: anamino 19×103 Mr species and a carboxyl 27×103 Mr glycosylatedform (Chang et al., 1994; Lee et al., 1994; Bumcrot et al., 1995).Interestingly, processing is autocatalysed, the catalytic sitemapping to the carboxyl peptide (Lee et al., 1994). Both peptidesare secreted, although only the larger form is readily detected inthe supernatant when the mature protein is processed (Lee et al.,1994; Bumcrot et al., 1995). Initially it was suggested that thetwo secreted peptides may independently mediate short- or long-range signaling functions (Lee et al., 1994). However, recentresults indicate that all signaling activity resides in the 19×103

Mr peptide in Drosophila Hh (Fietz et al., 1995; Porter et al.,1995) and mouse Shh (Martí et al., 1995).

We have generated peptide-specific polyclonal antibodiesand have used these to investigate the expression of Shh in thenotochord, CNS and limb of mouse and chick embryos. Ingeneral, the temporal expression of Shh is consistent with itspostulated signaling functions. However, the 19×103 Mrpeptide is detected in close association with cells transcribingthe Shh gene. Thus, we are unable to detect significantdiffusion of the Shh signal in the vertebrate embryo.

MATERIALS AND METHODS

Whole-mount in situ hybridizationWhole-mount in situ hybridization was performed according to Parret al. (1993). Digoxigenin-labelled RNA probes were preparedaccording to Wilkinson (1992). Mouse and chick Sonic hedgehogprobes have been described in Echelard et al. (1993) and Riddle et al.(1993).

Shh antibodies Generation of Ab80 has been described elsewhere (Bumcrot et al.,1995). Ab79 was raised against a bacterially expressed, hexa-histidinetagged protein corresponding to approximately the amino terminaltwo thirds of mouse Shh. The antiserum was affinity purified asdescribed elsewhere (Bumcrot et al., 1995). Ab77 was raised againsta glutathione-S-transferase fusion protein containing amino acids 369to 425 of chick Shh (Riddle et al., 1993). For all three antisera, inoc-

ulation of New Zealand White rabbits, as well as test and productionbleeding, was carried out at Hazleton Products, Inc.

Western blot analysis of antibody specificityCOS cells were transfected with mouse and chick Shh expression con-structs as described (Bumcrot et al., 1995). Cell lysates were preparedand analyzed by immunoblotting as described (Bumcrot et al., 1995).Affinity-purified Ab79 and Ab80 were each used at a dilution of1:200.

Whole-mount immunostainingEmbryos were fixed in 4% paraformaldehyde in PBS (pH 7.4) for 16hours at 4ºC, rinsed in PBS pH 7.4, dehydrated in an ethanol seriesand stored in 100% ethanol at −20ºC. Before immunostaining,embryos were rehydrated through an ethanol series. Endogenous per-oxidase activity was inactivated with 1% H2O2 in PBS. Embryos werethen rinsed in PBT (PBS containing 0.5% Triton X-100) and blockedfor 2 hours at room temperature, using 10% heat-inactivated sheepserum and 1% bovine serum albumin (BSA) in PBT. Embryos wereincubated overnight at 4ºC with antisera diluted 1:500 in PBS con-taining 1% sheep serum and 0.1% BSA and 0.5% Triton X-100. Shhantibodies were detected with goat anti-rabbit-HRP (1:250) (JacksonImmunochemicals), followed by DAB color detection, or withbiotinylated goat anti-rabbit antibodies (1:500) (Jackson Immuno-chemicals) and a streptavidin-conjugated β-gal (1:500) (Boehringer)followed by β-galactosidase color detection.

Embryos were mounted in 85% glycerol in PBT, or were clearedin benzyl alcohol-benzyl benzoate (1:2) before photographing with anOlympus SZH stereomicroscope.

Immunostaining on sectionsAntibody staining on cryostat (10 µm) or paraffin (10 µm) sectionsgave similar results and thus were used interchangeably. Sonichedgehog antibodies were used at a dilution of 1:500.

A rabbit polyclonal antibody to the LIM homeodomain proteinislet-1 (kindly provided by T. Edlund) was used in order to detectearly differentiating motor neurons. A mouse monoclonal antibody tothe IgG-like glycoprotein SC-1/BEN (Developmental StudiesHybridoma Bank) was used in order to stain floor plate and early dif-ferentiating motor neurons. HNF-3β was detected using a rabbit poly-clonal anti-HNF-3β antibody (H. Sasaki, unpublished data). Compar-ison of immunostaining in mouse embryos with previous reports ofexpression of related family members, including HNF-3α, togetherwith western blot analysis of cross reactivity, suggest that thisantibody is HNF-3β specific. Rabbit primary antibodies were detectedwith either HRP-conjugated goat anti-rabbit or AP-conjugated goatanti-rabbit. Mouse antibody was detected with HRP-conjugated goatanti-mouse.

Double labellingDouble labelling for RNA and protein was performed following thewhole-mount in situ hybridization (Parr et al., 1993) with severalmodifications. The proteinase K and RNAse treatments were elimi-nated and postfixation was performed in 4% paraformaldehyde. Afterpostfixation, embryos were photographed, dehydrated and paraffinembedded. Antibody staining was performed on 10 µm paraffinsections as described above.

Double labelling for two different antigens was performed simul-taneously on either cryostat or paraffin sections. Antibody detectionwas performed by incubating with the relevant secondary antibodies(anti-rabbit AP, and anti-mouse HRP), followed by sequential colordetection.

RESULTS

Polyclonal antibodies were raised in rabbits against mouse and

E. Martí and others

2539Shh in chick and mouse embryo

chick Shh proteins (Fig. 1A). A summary ofthese activities is presented (Fig. 1B). Ab77was raised against a C-terminal region of the26×103 Mr carboxyl peptide (C-peptide) ofchick Shh. We have been unable to affinitypurify this antiserum effectively due to theinsolubility of the immunogen, thus westernblot analysis of the antibody specificity hasnot been performed. However, on the basisof immunostaining of COS cells transfectedwith full-length and truncated mouse andchick Shh proteins, it is clear that thisantiserum only cross-reacts with theunprocessed form of chick Shh and its C-peptide cleavage product. There is no crossreactivity with mouse Shh which is quitedivergent in this region (Echelard et al.1993). Ab79 was raised against a 28×103 Mrmouse Shh peptide which encompasses theentire N-peptide and a short region of the C-peptide. Western analysis of COS cellsexpressing chick and mouse Shh peptidesindicates that Ab79 recognizes the full-length and amino 19×103 Mr Shh peptides ofchick and mouse Shh (Fig. 1C), as well astheir zebrafish and human counterparts (datanot shown). In addition, there is weak crossreactivity against both the carboxyl 26×103

Mr chick and 27×103 Mr mouse peptides(Fig. 1C). Ab80, which was raised against asmall region of the N-peptide, has beendescribed in detail elsewhere (Bumcrot etal., 1995). This antibody only recognizes thefull-length and amino 19×103 Mr peptide ofchick and mouse Shh (Fig. 1C). Cell cultureand embryonic studies indicate that the vastmajority of Shh protein is rapidly processedto generate the amino and carboxyl peptides(Bumcrot et al., 1995). Thus, it is likely thatthe various antisera only detect the terminalcleavage products in this study. Moreover,despite the extensive similarity amongst ver-tebrate Hh proteins in the N-peptide(Echelard et al., 1993), it is likely that ourstudies only detect Shh peptides. A detailedanalysis of the expression of Shh, Dhh andIhh shows no overlap in their respectiveexpression domains in the notochord, CNSor limbs (M. Bitgood and A. P. M., unpub-lished data)

Mesodermal and CNS expression ofShh peptides(A) Mesoderm expressionWhole-mount immunohistochemistry wasperformed to mouse and chick embryosfrom early head fold stages. Identical resultswere obtained with Ab79 and Ab80 in themouse and with all three antisera in thechick. Results are presented for Ab80(mouse) and Ab79 (chick) (Figs 2-4). At

Cross reactivity of Shh antisera

chick mouse

Antisera full length 26kDa 19kDa full length 27kDa 19kDa

Ab77 +++ +++ - - - -

Ab79 +++ + +++ +++ + +++

Ab80 +++ - +++ +++ - +++

1 437

sig. pep. cleavage

6 His 40-437: (Ab 79)

HHHH

HH

gel isolated

GSTGST 44-143: (Ab 80)

1 437

sig. pep. cleavage

1 425

sig. pep. cleavage

GSTGST 369-425: (Ab 77)

CHICK Shh:

MOUSE Shh:

27 201

25 199

A

B

C Ab79

Mo Ch

F

C

N

15.4

18.3

29.5

45.5

Ab80

Mo Chmock

Fig. 1. Antisera raised for this study: summary of peptide specificity and species cross-reactivity. (A) Bacterially expressed peptides used as antigens for polyclonal antibodyproduction. The indicated regions of mouse and chick Shh which were fused in frame toglutathione-S-transferase (GST) or six consecutive histidine residues (H). The regionlabeled ‘gel isolated’ is the electrophoretically purified breakdown product that was usedto raise antiserum 79. The region predicted to encode the signal peptide (filled box), aswell as the proposed proteolytic cleavage site is indicated for mouse and chick Shh. (B) Immunoreactivity of the three antibodies used in this study. (C) Western blotshowing cross-reactivity of Ab79 and Ab80 with chick (Ch) and mouse (Mo) Shhpeptides expressed in cos cells. F, mature protein; C, carboxyl peptide; N, amino peptide.

2540

early head fold stages in the mouse (8.0 days postcoitum [dpc])prior to formation of the 1st somite and in the stage 5-6 chick(Hamburger and Hamilton, 1951), Shh-expressing cells extendfrom the node into the node-derived cells of the head processwhich underlie the neural plate (Fig. 2A,B; data not shown).Expression in the head process is mosaic, with Shh proteinoutlining single cells at the midline (Fig. 2B). By early somitestages, Shh-expressing cells extend more anteriorly, into theprechordal plate mesoderm underlying the forebrain of thechick (Fig. 3A,B), and caudally along the entire length of thehead process and notochord into the node in both the chick andmouse (Figs 2C, 3A-E). Expression is lost in the prechordalplate of the chick by stage 15 (Fig. 4A,B), so that the chickand mouse at this stage (mouse 9.5 dpc; 20-25 somites) showsimilar mesodermal expression, along the entire length of thehead process and notochord (Figs 2D, 4A-D). In the chick atstage 18 and in the mouse at 11.5 dpc, Shh expression becomesrestricted to the notochord caudal to the forelimb buds (datanot shown).

(B) CNS expressionShh is detected in the CNS of the mouse in the ventral midbrainat the 8-somite stage (Fig. 2C) and slightly earlier in the chick.Expression then extends rostrally and caudally such that, bystage 10 in the chick, Shh peptides are present in the ventralCNS from the forebrain to the spinal cord up to the level ofthe fifth somite (Fig. 3A-E). By 9.5 dpc in the mouse, approx-imately stage 15 in the chick, Shh is restricted to the ventralmidline in most of the brain, with the exception of the rostraldiencephalon where expression skirts the presumptive hypo-thalamic region, the paramedian plate (Figs 2D, 4B). In the

spinal cord, ventral midline expression extends caudally to theposition of the 4th and 5th most recently formed somites (Figs2D, 4A,D). Thus, there is a continuous ventral strip of Shhexpression at, or close to, the ventral midline of the CNSextending from the rostral limit of the CNS in the forebrain tocaudal regions of the spinal cord. This expression persists inthe mouse until at least 15.5 dpc (data not shown). In addition,there is a dorsal expansion of Shh expression at the zonalimitans intrathalamica, a major pathway of axonal migrationthat lies at the boundary between prosomeres 2 and 3 in thediencephalon (Figdor and Stern, 1993), which is visible in thestage 20 chick (Fig. 4E) and 10.5 dpc mouse brain (data notshown).

Shh expression and CNS patterningWhole-mount analysis of Shh expression at the midlinesuggests that Shh peptides are restricted to a domain verysimilar to that previously reported for Shh transcripts (Echelardet al., 1993; Riddle et al., 1993). To examine the distributionof Shh peptides more closely, particularly with relation toventral patterning of the CNS, we performed a number ofsingle- and double-labelling studies on sections of chick andmouse embryos.

As with earlier studies, similar results were obtained withthe various antisera, with one exception. Ab77 and Ab79, butnot Ab80, detect a halo of Shh peptide surrounding thenotochord of the stage 15 chick embryo at the brachial level(Fig. 5A-C), presumably reflecting the distribution of thecarboxyl 26×103 Mr peptide, which has been shown to be morefreely diffusible in cell culture (Bumcrot et al., 1995). In allother respects immunostaining is identical and similar to that

E. Martí and others

Fig. 2. Whole-mount immunostaining of mouse embryos. (A,B) Lateral and frontal views of a presomitic stage mouse embryo (Ab80).Extraembryonic membranes have nonspecific β-gal activity. (C) Lateral view of an 8-somite-stage embryo showing the expression of Shh inthe notochord (not) and node (nd). Initial expression of Shh in the midbrain is also shown (arrowhead) (Ab80). (D) 9.5 dpc mouse embryoimmunostained with Ab79. Shh is expressed in the notochord, ventral midline of the CNS (except in the presumptive hypothalamic areas) andin endoderm derivatives, branchial arches, foregut (g) and hindgut (hg).

2541Shh in chick and mouse embryo

obtained at brachial levels of the mouse using either Ab79 orAb80 (Fig. 5D,E). In both the notochord and ventral CNS, Shhpeptides are pericellular, indicating secretion of Shh protein.Moreover, in the ventral spinal cord, most of the protein accu-mulates at the luminal cell surface in a region corresponding,approximately, to the limits of the morphologically definedfloor plate (Kingsbury, 1930).

Double labelling for expression of Shh RNA and protein inthe mouse (Fig. 5E) and chick (data not shown) indicates thatpeptides are only detected on, or very close to, the surface ofcells transcribing the Shh gene. Thus, we are unable to detectany diffusion of the 19×103 Mr signaling peptide. At a laterstage, in the spinal cord of the 9.5 dpc mouse embryo at thebrachial level (Fig. 5F,G), we observe a similar tight rela-tionship between the distribution of Shh protein and Shh-tran-scribing cells. However, both now appear to extend ventro-lateral to the morphologically defined floor plate. Thisexpansion is also apparent in stage 18 chick embryos (data notshown). Interestingly, at 10.5 dpc, when commissural axontracts project contralaterally across the floor plate, Shh

peptides become widely dispersed beyond the floor plate cellsthat are transcribing the Shh gene (Fig. 5I), suggesting that the19×103 Mr peptide may adhere to the surface of thesemigrating axons. It is clear that this, and all other aspects ofantibody immunoreactivity at each of the stages examined, arespecific since all immunostaining is blocked by preincubationof antisera in recombinant Shh peptide (Fig. 5H and data notshown).

It has been suggested that Shh expression at the midlinemight be activated by HNF-3β, a member of the winged-helixclass of transcriptional regulators, and that HNF-3β and Shhmay subsequently regulate each other’s expression by apositive feedback mechanism (Echelard et al., 1993). Toexplore the relationship between these two factors, mouseembryo sections were double-stained for Shh RNA, which isspatially and temporally equivalent to the domain of Shhprotein expression (see above), and for HNF-3β protein, usingan HNF-3β-specific antibody (H. Sasaki, unpublished data). Asexpected from earlier studies (Echelard et al., 1993), HNF-3βexpression in the ventral CNS precedes that of Shh (data not

Fig. 3. Shh peptides in astage 10 (10 somites) chickembryo. (A) Whole-mountimmunostaining showingthe planes of sections in B-E. All immunostaining wasperformed using Ab79. (B) Section through thelevel of the forebrainshowing Shh in theprechordal plate mesoderm(pcp) and in the ventralforebrain (bp, basal plate).(C) Section through thelevel of the first somiteshowing the expression ofShh in the notochord (not)and in the floor plate (fp).(D) Section through the fifthsomite. Shh is expressed inthe notochord, but only veryweakly in the floor plate.(E) Section through thecaudal unsegmentedmesoderm and open neuralplate showing that Shh isexpressed in the notochordbut not in neural tissue.

2542

shown), consistent with HNF-3β playing a role in the activa-tion of Shh. By 9.5 dpc, Shh expression is also detected in theventral spinal cord at brachial levels. Interestingly, there is agraded distribution of HNF-3β expression; highest ventrally inthe floor plate cells that also express Shh, and weaker laterallyin cells outside of the Shh expression domain (Fig. 6A). Thus,the subsequent maintenance of high levels of HNF-3β corre-lates with the presence of high levels of Shh peptides. Theseresults raise the interesting possibility that cells expressinghigh versus low levels of HNF-3β may have different fates,floor plate and motor neurons, respectively. That there aredosage requirements for HNF-3β for normal development ofthe nervous system is suggested by the neurological phenotypeobserved in mice lacking a single copy of the gene (Ang andRossant, 1994; Weinstein et al., 1994).

In vitro studies have demonstrated that the 19×103 Mr Shhpeptide is both necessary and sufficient for the induction ofmotor neurons in vitro (Martí et al., 1995), as judged by theactivation of an early motor neuron marker, Islet-1, a LIMdomain transcription factor (Thor et al., 1991; Yamada et al.,1993). Thus, Shh produced by the notochord and floor platemay play a role in motor neuron induction in vivo. This con-clusion is supported by double labelling studies in stage 18chick embryos. Clearly, Islet-1 and Shh-expressing cellspartially overlap in the ventral CNS (Fig. 6B). Thus at leastsome motor neuron precursors appear to arise from, or are incontact with, Shh-expressing cells. Shh expression was alsocompared to that of the IgG-related glycoprotein SC-1/BEN(Tanaka and Obata, 1984; Yamada et al., 1991), a floor plate

and motor neuron marker. At the ventral midline, SC-1/BENis restricted to floor plate cells, whilst Shh expression extendsventrolaterally, abutting the most ventral SC-1/BEN-express-ing motor neurons (Fig. 6C). Thus, even in the absence ofappreciable diffusion of Shh protein, Shh expression is con-sistent with the direct induction of at least the ventralmostmotor neurons.

Shh expression and limb patterningNormal expression of Shh in the ZPA (Riddle et al., 1993;Chang et al., 1994), together with in ovo studies in whichectopic expression in the anterior mesenchyme of the limbleads to limb duplications (Riddle et al., 1993), support thehypothesis that Shh mediates the function of the ZPA. As theZPA has been hypothesized to be the source of a long-rangemorphogen (Tickle et al., 1975; Tickle, 1981), the distributionof Shh peptides in the limb mesenchyme is of considerableinterest.

Chick and mouse limbs express Shh peptides throughout theperiod in which Shh is transcribed in a region approximatingto the operationally defined ZPA (Riddle et al., 1993). Similarresults are obtained with antisera recognizing either Shhpeptide (Fig. 7A,B). Immunoreactivity remains tightlylocalized to the posterior-distal limb mesenchyme, extendingapproximately 50 to 75 µm across the anterior-posterior axisbetween stages 20 to 25 (Fig. 7A-C). Expression of Shh RNAand protein in the posterior mesenchyme is not uniform. Thehighest levels of both are observed in cells immediatelybeneath the AER (Fig. 7C), although no immunoreactivity is

E. Martí and others

Fig. 4. Whole-mountimmunostaining of stage 15and 18 chick embryos usingAb79. (A) Stage 15 chickembryo immunostained forShh peptides. (B) Expressionof Shh in the ventral midlineof the brain. Expression isseen in the floor plate athindbrain (hb) and midbrain(mb) levels and in the basalplate (bp) at rostral midbrainand forebrain (fb) levels. Shhis not expressed in the ventralmidline at the level of theparamedian plate (pmp), theregion of the presumptivehypothalamus. (C) At trunklevels Shh is expressed in thenotochord (not) and in thefloor plate of the spinal cord(fp). (D) At caudal levels Shhis expressed in the notochordbut not in the floor plate(arrow). (E) Stage 20 chickembryo immunostained forShh peptides showing theexpression of Shh protein inthe basal plate, preoptic areas(poa), and in the zonalimitans intrathalamica at thejunction between prosomeres(P) 2 and 3.

2543Shh in chick and mouse embryo

seen in the AER itself. Further, isolated patches of smallgroups of posterior mesenchyme cells (3 to 5) are observedwhich clearly do not immunostain (arrows, Fig. 7C) despite thefact that these clusters are surrounded by cells transcribing Shh,which have Shh protein on their surface. Thus, although wecan see differences in Shh expression in the posterior limb budmesenchyme, we cannot detect any long-range diffusion ofShh protein.

DISCUSSION

Shh and midline developmentOur results provide evidence that Shh peptides are expressedin midline cell types, from gastrulation stages, in the chick andmouse embryo. In the mesoderm, Shh expression extends fromthe most rostral midline mesoderm of the prechordal plate,underling presumptive forebrain regions, to the notochord at

Fig. 5. Expression of Shh RNA andpeptides in the spinal cord andnotochord of chick and mouseembryos. (A-C) Sections through thebrachial level of a stage 15 chickembryo stained with antibodiesspecific to the carboxy-terminalpeptide (Ab77, A), both peptides(Ab79, B) and the amino-terminalpeptide (Ab80, C) showing that thecarboxy-terminal peptide, but not theamino peptide, diffuses from thenotochord (not). (D,E) Sectionsthrough the brachial level of a 9.0dpc mouse embryo immunostainedfor Shh (Ab79, D) and doublelabelled for Shh RNA (blue color)and peptides (brown color, E).Protein accumulates at the luminalsurface of floor plate cells (arrow)and decorates the surface ofnotochord cells. (F,G) Sectionsthrough the brachial level of a 9.5dpc mouse embryonic spinal cordimmunostained for Shh (Ab79) anddouble labelled for RNA andpeptides. In the floor plate, it isapparent that peptide decorates thesurface of expressing cells andaccumulates mostly at the ventricular(arrow) and, to a lesser extent, basalsurface. (H,I) Sections through thebrachial level of a 10.5 dpc mouseembryonic spinal cord. (H)Consecutive section to I, showingabsence of immunostaining followingpreincubation of antisera with 10µg/ml of recombinant peptide. (I) Peptide accumulation is seen atthe apical (luminal) surface of ventralmidline cells (arrow in I) and liningthe tracts of the commissural axons(com, ventrolateral arrows)

2544

spinal cord levels. By early somite stages Shh peptides arepresent at the ventral midline of the presumptive CNS,extending rostrally and caudally, to occupy most of the ventralmidline. Later an exception is seen in the rostral diencephalonwhere Shh is not expressed in the midline of the presumptivehypothalamic region. Whilst Shh expression overlaps the clas-sically defined wedge-shaped region termed the floor plate, itis clear that Shh-expressing cells also extend laterally, beyondthe floor plate. Thus, at spinal cord levels, the domain of Shhexpression is considerably broader than that of the floor platemarker, SC-1/BEN. In the most rostral regions of the brain,which do not exhibit a floor plate morphology, Shh is alsoexpressed in midline cells. Therefore, expression of Shh does

not demarcate a particular cell type, but rather a position withinthe ventral CNS.

Several processes have been associated with the signalingproperties of the notochord and floor plate region; the inductionor self induction of floor plate, the specification of neuronalidentity (motor neurons) and the induction of sclerotome devel-opment in the ventral somite. Recent studies, both in vitro andin vivo, indicate that all these activities may be attributed toShh. Thus, either ectopic expression of Shh in the embryo(Echelard et al., 1993; Krauss et al., 1993; Roelink et al., 1994)or application of Shh-expressing cells (Roelink et al., 1994) orShh peptides (Martí et al., 1995) to neural explants in vitroleads to floor plate induction. Moreover, purified Shh at the

E. Martí and others

Fig. 6. Expression of Shh RNA and peptides in the spinal cord, in relation to ventral markers. (A) Section through 10.5 dpc mouse embryospinal cord, showing the restriction of high levels of expression of HNF-3β protein (brown, nuclear staining) to the ventral midline where itcolocalizes with Shh RNA (blue color). Low levels of expression of HNF-3β are present in presumptive motor neurons which are notexpressing Shh. (B) Section through a caudal stage 18 chick embryonic spinal cord showing Shh RNA labelling (blue color) at the ventralmidline. Shh is expressed in a broader domain than the floor plate. Some cells expressing the motor neuron marker Islet-1 (brown, nuclear)overlap the Shh expression domain (arrows). (C) Section through a stage 18 chick spinal cord at the brachial level immunostained for Shhpeptides (Ab79, purple color) and for SC-1/BEN (brown color). SC-1/BEN is localized on the surface of floor plate cells and differentiatingmotor neurons. Shh expression at the ventral midline, extends lateral to SC-1/BEN in the floor plate and abuts SC-1/BEN expression in motorneurons (arrow).

Fig. 7. Expression of Shh in the limb bud of stage 25 chick embryos. (A,B) Whole-mount immunostained chick embryos using antisera specificto the amino peptide (Ab80, A) and to the carboxy-terminal peptide (Ab77, B), showing expression of Shh (arrow) restricted to the posterior-distal mesenchyme of the limb bud. Distal is right and posterior down. (C) Section through a similar stage limb bud double labelled for ShhRNA (blue color) and peptides (Ab79, brown color). Shh peptides are detected on the surface of mesenchyme cells that are expressing Shh. Thearrow denotes a patch of cells which do not immunostain despite the close association with Shh-expressing mesenchyme. AER, apicalectodermal ridge.

2545Shh in chick and mouse embryo

appropriate concentration can induce motor neurons in theabsence of floor plate induction (Martí et al., 1995). Finally,Shh can activate the ventral sclerotomal marker Pax-1 (Fan andTessier-Lavigne, 1994; Johnson et al., 1994). These inductiveevents, which apparently involve both short- and long-rangeinteractions, all appear to be mediated by the 19×103 Mr Shhpeptide (Martí et al., 1995; unpublished data). Similarly, allDrosophila Hh signaling can be ascribed to a homologous19×103 Mr peptide (Fietz et al., 1995; Porter et al., 1995). Thus,a key question is how does the expression and distribution ofthis peptide correlate with these signaling activities?

Detailed analysis of the timing of floor-plate- and motor-neuron-inducing activity by the notochord and floor plate hascome from the studies of Placzek et al. (1993) and Yamada etal. (1993), which are summarized and compared to theexpression of Shh peptides (Table 1). Data are based on theinducing properties of notochord and floor plate (excised atparticular axial levels) on chick and rat neural explants in vitro.Shh expression in the stage 6 chick notochord is consistentwith the floor-plate-inducing properties of the notochord at thistime. However, strong Shh expression is also detected at stage10, when floor-plate-inducing activity is apparently alreadylost from the notochord. In these experiments, floor plateinduction is assayed after 24-48 hours of culture, and it is notclear for what proportion of this period the neural explant mustreceive a floor-plate-inducing signal to initiate floor platedevelopment. As our data indicates that Shh expression in thenotochord decreases after stage 15, there may be insufficientexpression, either quantitatively or temporally, to effect floorplate induction. In contrast, both floor-plate-inducing activityand Shh expression are present at late stages (Table 1).

However, floor-plate-inducing activity appears to be present inthe ventral neural plate prior to expression of Shh, although itis possible that explanted floor plate tissue may activate Shhduring the course of the culture period, as would occur in vivo.In summary, our study is consistent with the conclusion thatthe 19×103 Mr Shh peptide may mediate the local induction offloor plate by the notochord and/or floor plate in vivo. Theinconsistencies can be accounted for by the experimentalapproaches which do not assay inducing activity at the time ofexplant, but rather at some indeterminate time during theculture period.

A similar analysis of motor neuron induction indicates astriking correlation with the expression of Shh in the inducingtissues (Table 1). Interestingly, notochord of stage 15/20embryos is unable to induce floor plate but does induce motorneurons. As Shh levels in the notochord are decreasing afterstage 15, the changing inductive capacity of the notochord mayreflect decreasing levels of Shh activity (see later). A detailedanalysis of sclerotomal-inducing activity in the notochord andfloor plate has not been performed, although it is clear that Shhis expressed in the inducing tissue at the time of grafting (Fanand Tessier-Lavigne, 1994). Thus, the temporal expression ofShh in midline tissues is generally consistent with its mediatinginductive events in the spinal cord and somites.

In contrast to floor plate induction, motor neuron and scle-rotome induction do not appear to require contact with theinducing tissue, suggesting that long-range inductions aremediated by a diffusible signal. However, we have been unableto detect appreciable diffusion of the 19×103 Mr Shh peptidein vivo. There are several possible explanations that mayreconcile the various data.

For example, whereas in vitro a diffusible signal frommidline tissues may induce motor neurons and sclerotome, thismay not be the case in vivo. Our results indicate that someventral motor neurons may develop in direct contact with cellbound Shh; however, this is unlikely to be true of all motorneurons. Alternatively, Shh may induce the expression of asecond diffusible signal which mediates long-range signaling.This does not appear to be the case in the somite at least (Fanand Tessier-Lavigne, 1994). The simplest explanation, and theone we favor, is that long-range signaling is mediated by a dif-fusible form of the 19×103 Mr peptide, but this cannot bedetected using current procedures. If correct this model wouldpredict that floor plate induction may require high concentra-tions or a cell surface bound form (or both), whilst motorneuron and sclerotome induction may be initiated in responseto low amounts of diffusing Shh protein. Indeed, in vitrostudies indicate that motor neuron induction can occur over a100-fold concentration range of the 19×103 Mr peptide, whichis unable to induce floor plate (Martí et al., 1995).

Interestingly, our studies demonstrate a distinct polarity tothe distribution of Shh peptides on the surface of ventralmidline cells in the CNS. Shh accumulates towards theluminal, and to a lesser extent, the basal surface, of the cell. Itis tempting to speculate that if there is diffusion of the 19×103

Mr peptide in the spinal cord, it is through the lumen acting atthe luminal surfaces of responding cells.

It is clear from the work reported here that Shh peptides arepresent in the midline mesoderm and ventral CNS at all axiallevels. Interestingly, grafts of trunk notochord induce sero-toninergic cells in hindbrain explants (Yamada et al., 1991),

Table 1. Ventral midline tissue inducing activitiescompared to Shh expression

(A) Summary of motor neuron inducing activity and Shhexpression

Stage mn induction* Shh expression

notochord 10 +++ +++(brachial) 15 +++ +++

20 +++ +25 − −

floor plate 10 − +(brachial) 16 +++ +++

20 +++ +++25 +++ +++

*Data from Yamada et al., 1993.

(B) Summary of floor plate inducing activity and Shhexpression

Stage fp induction* Shh expression

notochord 6 +++ +++(1st somite) 10 − +++

20 − +

floor plate 6 +++ −(1st somite) 10 +++ +

20 +++ +++25 +++ +++35 +++ nd

*Data from Placzek et al., 1993.

2546

whilst floor plate from the spinal cord induces dopaminergicneurons in midbrain explants (Hynes et al., 1995). Theinduction of the appropriate ventral cell type for each axiallevel suggests that different responses may be mediated bycommon midline signals, the specificity of the response beingdetermined by the receiving tissue. Indeed, studies on thezebrafish embryo indicate that the ventral midline of the CNShas common properties along its entire length (Hatta et al.,1991, 1994). The expression of Shh makes it an attractivecandidate for this common ventralizing activity. Thus, theresponse to Shh signaling in the CNS may be governed by boththe distance from the source of Shh-expressing cells and theposition of responding cells along the anterior-posterior axis.

Although there is evidence that Shh may play an inductiverole in the early CNS, its expression at later stages suggestsadditional functions. At the ventral midline, it appears that Shhpeptides move laterally on the surface of migrating axons. Inthe brain, we observe strong expression of both Shh peptidesat the zona limitans intrathalamica, separating prosomeres (P)2 and 3 in the rostral diencephalon (Figdor and Stern, 1993;Puelles and Rubinstein, 1993). Expression here is associatedwith the axon tracts of a major dorsal projection of ventralneurons, the mamillothalamic tract. Thus, there is some sug-gestion that Shh may play a role in axonal outgrowth orguidance. Alternatively, such boundaries may initiate ormaintain an underlying segmental organization in the forebrain(Puelles and Rubinsten, 1993). Interestingly, Drosophila Hhparticipates in a reciprocal signaling pathway, which maintainsthe parasegmental boundary. A further parallel with theDrosophila embryo can be drawn from the fact that this stripof Shh-expressing cells in the diencephalon separates cells inP2 that express Wnt-3 from those in P3 expressing Dlx-1/2.(Puelles and Rubinstein, 1993). Drosophila counterparts ofthese genes, wingless and distalless, respectively, are regulatedby Hh (Ingham 1993; Diaz-Benjumea et al., 1994).

Shh and limb developmentThe results of elegant in vivo surgeries in the chick limbsuggest that the cells of the ZPA produce a signal which actsover 200 µm, or 20 cell diameters, to pattern the anterior-posterior axis of the limb (Honig, 1981). High levels of thissignal are thought to be required for the specification ofposterior limb elements, whilst lower levels are sufficient foranterior structures (Tickle et al., 1975; Tickle, 1981). Thus, thelimb offers one of the best examples in vertebrate embryoge-nesis where a long-range morphogen may play a significantrole. Shh is expressed in, and mediates the activity of the ZPAof the limb (Riddle et al., 1993; Chang et al., 1994), and is thusa good candidate for this postulated morphogen. We show thatShh peptides are expressed in the posterior distal limb budmesenchyme at the appropriate stages. There is a graded dis-tribution of both RNA and protein, which are highest in thevicinity of the AER, but detectable peptide extends no morethan 50 to 75 µm along the anterior-posterior axis. Thus, weare unable to demonstrate any long-range diffusion of Shh inthe limb. As argued in the previous section, this most likelyreflects the technical difficulty of detecting low amounts ofprotein, although the alternative explanation of a second signalhas to be treated seriously.

Recently, BMP-2, a member of the TGF-β family of secretedpeptide factors, has been shown to be expressed in the ZPA

and appears to extend anterior to the domain of Shh expression(Vogel and Tickle, 1993). Interestingly, Dpp, a close relativeof BMP-2, is activated in response to Hh signaling inDrosophila (Heberlin et al., 1993; Ma et al., 1993). Moreover,ectopic activation of Dpp, in the absence of Hh, gives similarpattern alterations to ectopic expression of Hh (Capdevila andGuerrero, 1994; Ingham and Fietz, 1995). Thus, Dpp may havelong-range activity and may mediate the dose-dependenteffects of Hh. Whether BMP-2, or another related factor,mediates the apparent long-range action of Shh in the ver-tebrate limb remains to be determined. Experiments implant-ing BMP-2 containing beads into the anterior margin of thechick limb have not resulted in alteration of limb pattern(Vogel and Tickle, 1993).

This work was supported by grants from the HFSP and NIH to A.P. M., and by the Spanish Ministry of Education (E. M.) and theMuscular Dystrophy Association (D. A. B.). We wish to thank BiancaKlumpar for her technical assistance. We would like to thank DrThomas Edlund for the gift of the anti-islet-1 antiserum and Dr CliffTabin for the chick Shh cDNA clone.

REFERENCES

Ang, S-L. and Rossant, J. (1994) HNF-3β is essential for node and notochordformation in mouse development. Cell 78, 561-574.

Basler, K. and Struhl, G. (1994) Compartment boundaries and the control ofDrosophila limb pattern by hedgehog protein. Nature 368, 208-214.

Brand-Saberi, B., Ebensperger, C., Wilting, J., Balling, R. and Christ, B.(1993) The ventralizing effect of the notochord on somite differentiation inchick embryos. Anat. Embryol. 188, 239-245.

Bumcrot, D. A., Takada, R. and McMahon, A. P. (1995) Proteolyticprocessing yields two secreted forms of Sonic hedgehog. Molec. Cell Biol.15, 2294-2303.

Capdevila, J. and Guerrero, I. (1994) Targeted expression of the signallingmolecule decapentaplegic induces pattern duplications and growthalterations in Drosophila wing. EMBO J. 13, 4459-4468.

Chang, D. T., López A., von Kessler, D. P., Chiang, C., Simandl, B. K.,Renbin, Z., Seldin, M. F., Fallon, J. F. and Beachy, P. A. (1994) Products,genetic linkage, and limb patterning activity of a murine hedgehog gene.Development 120, 3339-3353.

Diaz-Benjumea, F., Cohen, B. and Cohen S. M. (1994) Cell interactionsbetween compartments establishes the proximal-distal axis of Drosophilaleg. Nature 372, 175-179.

Echelard, Y., Epstein, D. J., St-Jacques, B., Shen, L., Mohler, J.,McMahon, J. A. and McMahon A. P. (1993) Sonic hedgehog, a member ofa family of putative signalling molecules, is implicated in the regulation ofthe CNS polarity. Cell 75, 1411-1430.

Fan, C-M. and Tessier-Lavigne, M. (1994) Patterning of mammalian somitesby surface ectoderm and notochord: Evidence for sclerotome induction by ahedgehog homolog. Cell 79, 1175-1186.

Fietz, M., Concordet, J-P., Barbosa, R., Johnson, R., Krauss, S.,McMahon, A. P., Tabin, C. and Ingham, P. W. (1994) The hedgehog genefamily in Drosophila and vertebrate development. Development, 1994Supplement, 43-51.

Fietz, M. J., Jacinto, A., Taylor, A. M., Alexandre, C. and Ingham, P. M.(1995). Secretion of the N-terminal fragment of the hedgehog protein isnecessary and sufficient for hedgehog signalling in Drosophila. CurrentBiol. (in press).

Figdor, M. C. and Stern, C. D. (1993) Segmental organization of theembryonic diencephalon. Nature 363, 630-634.

Hamburger, V. and Hamilton, H. L. (1951) A series of normal stages in thedevelopment of the chick. J. Morphol. 88, 49-92.

Hatta, K., Kimmel, C. B., Ho, R. K. and Walker, C. (1991) The cyclopsmutation blocks specification of the floor plate of the zebrafish centralnervous system. Nature 352, 339-341.

Hatta, K., Puschel, A. W. and Kimmel, C. B. (1994) Midline signaling in theprimordium of the zebrafish anterior central nervous system. Proc. Natl.Acad. Sci. USA 91, 2061-2065.

E. Martí and others

2547Shh in chick and mouse embryo

Heberlein, U., Wolff, T. and Rubin, G. M. (1993) The TGFβ homolog dppand the segment polarity gene hedgehog are required for propagation of amorphogenetic wave in the Drosophila retina. Cell 75, 913-926.

Heemskerk, J. and DiNardo, S. (1994) Drosophila hedgehog acts as amorphogen in cellular patterning. Cell 76, 449-460.

Honig, L. S. (1981) Positional signal transmission in the developing chicklimb. Nature 291, 72-73.

Hynes, M., Poulsen K., Tessier-Lavigne, M. and Rosenthal, A. (1995)Control of neuronal diversity by floor plate: contact-mediated induction ofmidbrain dopaminergic neurons. Cell 80, 95-101.

Ingham, P. W (1993) Localized hedgehog activity controls spatial limits ofwingless transcription in the Drosophila embryo. Nature 366, 560-562.

Ingham, P. W. and Fietz, M. J. (1995) Dosage dependent effects of hedgehogand decapentaplegic activity on patterning of the Drosophila wing. CurrentBiol. 5, 432-440.

Johnson, R. L., Laufer, E., Riddle, R. D. and Tabin, C. J. (1994) Ectopicexpression of Sonic hedgehog alters dorso-ventral patterning of somites. Cell79, 1165-1173.

Kingsbury, B. F. (1930) The developmental significance of the floor plate ofthe brain and spinal cord. J. Comp. Neurol. 50, 177-207.

Krauss, S., Concordet, J. P. and Ingham, P. W. (1993) A functionallyconserved homolog of the Drosophila segment polarity gene hh is expressedin tissues with polarizing activity in the zebrafish embryo. Cell 75, 1431-1444.

Lee, J. J., Ekker, S. C., von Kessler, D. P., Porter, J. A., Sun, B. I. andBeachy, P. A. (1994) Autoproteolysis in hedgehog protein biogenesis.Science 266, 1528-1537.

Ma, C., Zhou, Y., Beachy, P. A. and Moses, K. (1993) The segment polaritygene hedgehog is required for progression of the morphogenetic furrow inthe developing Drosophila eye. Cell 75, 927-938.

Martí, E., Bumcrot, D. A., Takada, R. and McMahon, A. P. (1995)Requirement of 19K sonic hedgehog for induction of distinct ventral celltypes in CNS explants. Nature 375, 322-325.

Parr, B. A., Shea, M. J., Vassileva, G. and McMahon A. P. (1993) MouseWnt genes exhibit discrete domains of expression in the early embryonicCNS and limb buds. Development 119, 247-261.

Placzek, M., Jessell, T. M. and Dodd, J. (1993) Induction of floor plate bycontact-dependent, homeogenetic signals. Development 117, 205-218

Porquie, O., Coltey, M., Teillet, M. A., Ordahl, C. and Le Douarin, N. M.(1993) Control of dorso-ventral patterning of somitic derivatives bynotochord and floor plate. Proc. Natl. Acad. Sci. USA 90, 5242-5246.

Porter, J. E., von Kessler, D. P., Ekker, S. C., Young, K. E., Lee, J. J.,Moses, K. and Beachy, P. A. (1995) The product of hedgehogautoproteolitic cleavage active in local and long-range signalling. Nature374, 363-366

Puelles, L. and Rubinstein, L. R. (1993) Expression patterns of homeobox andother putative regulatory genes in the embryonic mouse forebrain suggest aneuromeric organization. Trends in Neurosc. 16, 472-479.

Riddle, R. D., Johnson, R. L., Laufer, E. and Tabin, C. (1993) Sonichedgehog mediates the polarizing activity of the ZPA. Cell 75, 1401-1416.

Roelink, H., Augsburger, A., Heemskerk, J., Korzh, V., Norlin, S., Ruiz iAltaba, A., Tanabe, Y., Placzek, M., Edlund, T., Jessell, T. M. and Dodd.J. (1994) Floor plate induction by vhh-1 a vertebrate homolog of hedgehogexpressed by the notochord. Cell 76, 761-775.

Saunders, J. W. and Gasseling, M. (1968) Ectodermal-Mesenchymalinteraction in the origin of limb symmetry. In Epithelial-MesenchymalInteraction, (ed. R. Fleischmayer and R. E. Billingham). pp 78-97.Baltimore: Williams and Wilkins).

Tabata, T. and Kornberg, T. B. (1994) Hedgehog is a signalling protein witha key role in patterning Drosophila imaginal discs. Cell 76, 89-102.

Tanaka, H. and Obata, K. (1984) Developmental changes in unique cellsurface antigens of chick embryo spinal motor neurons and ganglion cells.Dev. Biol. 106, 26-37.

Taylor, A. M., Nakano, Y., Mohler, J. and Ingham, P. W. (1993) Contrastingdistributions of patched and hedgehog proteins in the Drosophila embryo.Mech. Dev. 42, 89-96.

Thor, S., Ericson, J., Brannstrom, T., Edlund, T. (1991) The homeodomainLIM protein isl-1 expressed in subsets of neurons and endocrine cells in theadult rat. Neuron 7, 881-889.

Tickle, C. (1981) The number of polarizing region cells required to specifyadditional digits in the developing chick wing. Nature 289, 295-298.

Tickle, C., Summerbell, D. and Wolpert, L. (1975) Positional signaling andspecification of digits in chick limb morphogenesis. Nature 254, 199-202.

Vogel, A. and Tickle, C. (1993) FGF-4 maintains polarizing activity ofposterior limb bud cells in vivo and in vitro. Development 119, 199-206.

Wagner, M., Thaller, C., Jessell, T. and Eichele, G. (1990) Polarizingactivity and retinoids synthesis in the floor plate of the neural tube. Nature345, 819-822.

Weinstein, D. C., Ruiz i Altaba, A., Chen, W. S., Hoodless, P., Prezioso, V.R., Jessell, T. M. and Darnell, J. E. Jr. (1994) The winged helixtranscription factor HNF-3β is required for notochord development in themouse embryo. Cell 78, 575-588.

Wilkinson, D. G. (1992) Whole mount in situ hybridization of vertebrateembryos. In In situ Hybridization: A Practical Approach (ed. D. G.Wilkinson), pp 75-83. Oxford: IRL Press.

Yamada, T., Pfaff, S. L., Edlund, T. and Jessell, T. M. (1993) Control of cellpattern in the neural tube: motor neuron induction by diffusible factors fromnotochord and floor plate. Cell 73, 673-686.

Yamada, T., Placzek, M., Tanaka, H., Dodd, J. and Jessell, T. M. (1991)Control of cell pattern in the developing nervous system: polarizing activityof the floor plate and notochord. Cell 64, 635-647.

(Accepted 12 May 1995)