XATH-1,a Vertebrate Homolog ofDrosophila atonal,Induces Neuronal Differentiation within Ectodermal...

DEVELOPMENTAL BIOLOGY 187, 1–12 (1997)ARTICLE NO. DB978572

XATH-1, a Vertebrate Homolog of Drosophilaatonal, Induces Neuronal Differentiationwithin Ectodermal Progenitors

Peter Kim, Amy W. Helms,* Jane E. Johnson,*and Kathryn Zimmerman1

Department of Developmental Neurobiology, The Rockefeller University, 1230 York Avenue,New York, New York 10021; and *Department of Cell Biology and Neuroscience, Universityof Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75235

XATH-1, a basic/helix-loop-helix transcription factor and a homolog of Drosophila atonal and mammalian MATH-1, isexpressed specifically in the dorsal hindbrain during Xenopus neural development. In order to investigate the role ofXATH-1 in the neuronal differentiation process, we have examined the effects of XATH-1 overexpression during Xenopusdevelopment. XATH-1 induces the expression of neuronal differentiation markers, such as N-tubulin, within the neuralplate as well as within nonneural ectodermal progenitor populations, resulting in the appearance of process-bearing neuronswithin the epidermis. The related basic/helix-loop-helix genes neurogenin-related-1 and neuroD are not induced in responseto XATH-1 overexpression within the embryo, suggesting that XATH-1 may activate an alternate pathway of neuronaldifferentiation. In further contrast to neurogenin-related-1 and neuroD, high-level expression of general neural markersexpressed earlier in development, such as N-CAM, is not induced by XATH-1 overexpression. Competent ectodermalprogenitors therefore respond to ectopic XATH-1 expression by initiating a distinct program of neuronal differentiation.q 1997 Academic Press

INTRODUCTION neural territory. These patterns of neuronal differentiationappear to be driven in part by a family of basic/helix-loop-helix (bHLH) transcription factors related to the DrosophilaThe formation of the vertebrate nervous system is initi-proneural genes achaete-scute and atonal (for review, seeated by signaling events that occur at the time of gastrula-Jan and Jan, 1993). These molecules, which include thetion. Classical studies in amphibians have established thatachaete-scute-related vertebrate genes MASH-1 and XASH-signals from the dorsal lip of the blastopore, termed Spem-3 (Johnson et al., 1990; Zimmerman et al., 1993; Turner andann’s organizer, induce neural specification within the ecto-Weintraub, 1994) and the atonal-related vertebrate genesderm (for review, Spemann, 1938; Hamburger, 1988). Sev-MATH-1, MATH-2/nex-1, neuroD, and neurogenin (Aka-eral molecular candidates for the organizer-derived neu-zawa et al., 1995; Shimizu et al., 1995; Bartholoma andralizing signal have been isolated, including noggin,Nave, 1994; Lee et al., 1995; Ma et al., 1996), have providedfollistatin, and chordin (Lamb et al., 1993; Hemmati-Bri-insight into the molecular dynamics of the neuronal differ-vanlou et al., 1994; Sasai et al., 1995). These moleculesentiation pathway.induce the expression of general early neural markers, such

The vertebrate bHLH proneural homologs are expressedas N-CAM (Kintner and Melton, 1987), within ectodermalin distinct spatial and temporal patterns during develop-progenitors. In contrast to early markers which are ex-ment that coincide with patterns of neurogenesis in thepressed throughout the neural plate, markers specific forembryo. A better understanding of the function of bHLHdifferentiating neurons, such as N-tubulin, are expressedmolecules during neurogenesis has emerged in part fromin spatially and temporally restricted patterns within theoverexpression studies in Xenopus embryos. In the devel-oping Xenopus nervous system, several classes of primaryneurons express differentiated neuronal markers, including1 To whom correspondence should be addressed. Fax: (212)327-

7140. E-mail: [email protected]. N-tubulin, at the neural plate stage of development. A num-

1

0012-1606/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID DB 8572 / 6x24$$$161 06-12-97 15:18:26 dba

2 Kim et al.

ber of criteria suggest that Xenopus neurogenin-related-1 neurogenesis in the embryonic nervous system (Harten-stein, 1993). Overexpression of XATH-1 in Xenopus em-(X-NGNR-1) and neuroD function in the differentiation of

these neurons. X-NGNR-1 expression precedes and then bryos results in the ectopic expression of neuron-specificmarkers, such as N-tubulin, within the neural plate as welloverlaps the domains of primary neurogenesis within the

neuroectoderm, while neuroD expression is coincident with as within ectodermal progenitors normally destined for anonneural fate, including epidermis. Neither neuroD nor X-the expression of differentiated neuronal markers within

these progenitors (Lee et al., 1995; Ma et al., 1996). The NGNR-1 is induced in response to XATH-1 overexpression,suggesting that XATH-1 function is not dependent uponoverexpression of either X-NGNR-1 or neuroD within Xen-

opus embryos results in precocious as well as ectopic neu- these molecules. Further distinguishing XATH-1 from neu-roD and X-NGNR-1 is the finding that early general neuralronal differentiation within the neuroectoderm (Lee et al.,

1995; Ma et al., 1996). In addition, these molecules function markers, including N-CAM, are not ectopically expressedin response to XATH-1 overexpression. This finding sug-in a cascade, with X-NGNR-1 activating the expression of

neuroD within competent progenitors (Ma et al., 1996). gests that XATH-1 preferentially regulates a subset of genesnormally expressed at late stages in the neuronal differentia-The overexpression of X-NGNR-1 and neuroD also alters

the specification of ectodermal progenitor populations oth- tion pathway, providing molecular evidence that neuronaldifferentiation can be uncoupled from earlier stages of neu-erwise destined for a nonneural fate. In Xenopus embryos

overexpressing either gene, the mesectodermal cranial neural induction. Furthermore, our findings suggest that spe-cific bHLH containing transcription factors control distinctral crest and a subset of epidermal cells express neuronal

markers including N-CAM and N-tubulin and display a typ- mechanisms of neurogenesis in the embryo.ical process-bearing neuronal morphology (Lee et al., 1995;Ma et al., 1996). Similarly, animal caps normally fated forepidermal differentiation stably express the neural-specific

MATERIALS AND METHODSmarker N-CAM when isolated from embryos overex-pressing neuroD or X-NGNR-1 (Lee et al., 1995; Ma et al.,1996). These findings suggest that the neuronal differentia- Molecular cloning of XATH-1 gene. A probe which containedtion pathway contains multiple entry points and that ‘‘dif- the protein coding region of the MATH-1 gene (Helms and Johnson,

unpublished data) was used at low stringency to screen a genomicferentiation’’ genes such as X-NGNR-1 and neuroD can acti-library prepared from Xenopus liver DNA (Stratagene). A singlevate this pathway in competent ectodermal progenitors,clone of approximately 15 kb was isolated in a screen of 106 colo-acting independently of earlier organizer-derived neural in-nies. PCR amplification with degenerate primers based on se-ducers (Lee et al., 1995).quences within the bHLH region of the Drosophila atonal geneIn contrast to X-NGNR-1 and neuroD, a number of bHLH(Jarman et al., 1993) indicated that the bHLH region was containedcontaining genes are expressed in patterns that are non-within a 2.2-kb NotI/EcoRI fragment. The sequences of these prim-

overlapping with the initial domains of neurogenesis within ers were U: 5* ACGGATCCGCTGCTGA(C/T)GCI(C/A)GIGAthe embryo, suggesting a role in the differentiation of later (G/A)(A/C)G 3*; and D: 5* AGGAATTCCAT(C/T)TGIA(G/A)IGTdeveloping progenitor populations within the neuroectod- (C/T)TC(G/A)TG(C/T)TT 3*. The putative protein coding regionerm. Within this group of bHLH genes, Xenopus XASH- was sequenced in its entirety on both strands using a subclone

which linked the 2.2-kb fragment above to a downstream 6-kb3 and murine MATH-1 are expressed in highly restrictedEcoRI fragment.patterns within the developing nervous system. XASH-3 is

In situ hybridization analysis and antibody staining. The ge-expressed at the midpoint of the mediolateral axis of thenomic subclone which contained the bHLH region of the XATH-neuroectoderm in mid gastrula stage embryos and is main-1 gene was linearized with BsaHI and transcribed with T3 polymer-tained in progenitors at the midpoint of the dorsoventralase to generate an approximately 250-bp antisense probe for XATH-axis of the neural tube, well after primary neurogenesis is1. The N-CAM, N-tubulin, engrailed-2, twist, snail, and slug probes

complete in the embryo (Zimmerman et al., 1993; Turner have been described previously (Krieg et al., 1989; Richter et al.,and Weintraub, 1994). The effects of XASH-3 overex- 1988; Hemmati-Brivanlou et al., 1991; Hopwood et al., 1989;pression are dose-dependent, driving ectopic neuronal dif- Sargent and Bennett, 1990; Nieto et al., 1994). A probe for nrp-1ferentiation within the neuroectoderm at low doses but de- was obtained by PCR amplification of Stage 35 cDNA with primers

based on published sequences (U: 5* GGGTTTCTTGGAACAAGClaying neuronal differentiation at high doses, possibly as a3* and D: 5* ACTGTGCAGGAACACAAG 3*; Richter et al., 1990),result of activated signaling through the notch/delta lateralyielding a fragment of approximately 300 bp. The Xenopus PAX-3inhibition pathway (Ferreiro et al., 1994; Chitnis and Kint-gene was isolated from a Stage 17 neurula cDNA library (Kintnerner, 1996). The unique dose-dependent response to XASH-and Melton, 1987) using a probe which contained a portion of the3 overexpression points to the possibility that this group ofhomeo- and paired domains of chick PAX-3 gene (Goulding et al.,bHLH genes might differ from X-NGNR-1 and neuroD in1993). Xenopus LH-2 homologs were isolated from a Stage 27 head

their mechanism of action during neuronal differentiation. library (Hemmati-Brivanlou et al., 1991) using a probe containingIn order to further investigate these mechanisms, we have the LIM and homeodomains of the chick LH-2B gene (T. Jessell,

isolated a Xenopus homolog of the MATH-1 gene, called unpublished results). A probe for Wnt-3A was isolated by PCR ofXATH-1. XATH-1 is expressed within neural progenitors in a Stage 17 cDNA library using primers based on published se-

quences and yielding an approximately 700-bp fragment (U: 5*dorsal regions of the developing hindbrain, a region of late

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID DB 8572 / 6x24$$$162 06-12-97 15:18:26 dba

3XATH-1 Induces Neuronal Differentiation

GAGATGGGCTGCTTTGG 3* and D: 5* CCCTGTTCGGGC- NGNR-1, and neuroD but is not present in the MATH-1ATCTGG 3*; Wolda et al., 1993). protein coding domain (Fig. 1A, bold). With the exception

Albino embryos were staged according to Nieuwkoop and Faber of a few short stretches of amino acid homology, including(1967). Whole mount in situ hybridization analysis was performed a five-amino-acid identity at the N-terminus of the proteinusing previously described techniques with only minor modifica- coding domain, no major homologies are detected betweentions, including the substitution of BM Purple (Boehringer Mann-

XATH-1 and MATH-1 outside of the bHLH region.heim) for chromogenic reactions (Hemmati-Brivanlou et al., 1990;Characterization of XATH-1 expression during XenopusHarland, 1991). Selected embryos were dehydrated and mounted in

development. The expression of XATH-1 during develop-paraplast for sectioning.ment was characterized using whole mount in situ hybrid-For whole mount antibody staining, embryos were fixed for 1

hr in MEMFA and stored in methanol at 0207C. Embryos were ization techniques (Hemmati-Brivanlou et al., 1990; Har-rehydrated prior to antibody incubation in PBT (PBS, 2 mg/ml BSA, land et al., 1991). For these studies we employed a probe0.1% Tween 20). The HNK-1 antibody (Abo and Balch, 1981) was synthesized from the C-terminus of the XATH-1 proteinused at a 1:1 dilution; the Islet-1 antibody (Ericson et al., 1992) was coding domain. This probe does not contain the bHLH do-used at a 1:10 dilution. Peroxidase-conjugated secondary antibodies main and shares no homology with other members of the(Boehringer Mannheim) were used at a 1:100 dilution. The embryos bHLH family.were developed using DAB substrate (Vectastain).

The expression of XATH-1 was neural specific in thesePreparation and injection of mRNA for injection. X-NGNR-1assays. Low levels of XATH-1 expression were detected atmRNA was prepared as previously described (Ma et al., 1996).early neurula stages of development in regions of the pre-MATH-1 mRNA was synthesized from a plasmid (pMATH1c) con-sumptive hindbrain (Stage 17, data not shown). Higher lev-taining a 2-kb genomic fragment. Within this genomic fragment,

the protein coding sequence was flanked by noncoding regions (200 els of expression were detected following neural tube clo-bp 5* and 800 bp 3* ). The protein coding region of the XATH-1 gene sure in anterior regions of the hindbrain (Fig. 2A). Doublewas amplified from the partial genomic clone described above by labeling with engrailed-2 (Hemmati-Brivanlou et al., 1991)PCR using primers which contained the identified start and stop suggested that the anterior boundary of XATH-1 expressioncodons U: 5* CCGGATCCTCGCACTTACCTGTCATGGC 3*; and coincided with the midbrain/hindbrain junction, while dou-D: 5* CCGGATCCTTAATACGACTCACTATAGG 3*. The re- ble labeling with Krox-20 (Bradley et al., 1992) suggestedsulting product was subcloned into the BamHI site of the SP64T

that the posterior boundary of XATH-1 expression wasexpression vector (Krieg and Melton, 1984). The nuclear b-gal con-within the rhombomere 3 domain (data not shown). At latertaining plasmid has been described previously (Vize et al., 1991). Allstages (úStage 25), XATH-1 was expressed along the entiremRNAs were synthesized using recommended protocols (Ambion).anteroposterior extent of the hindbrain (Figs. 2B and 2C) andYields were calculated based on incorporation of radioactive tracer

nucleotide. a sharp anterior boundary at the morphologically distinctEggs from albino Xenopus laevis were in vitro fertilized and incu- midbrain/hindbrain junction was readily apparent (Fig. 2B,

bated in 0.11 MMR. Embryos were dejellied using 3% cysteine, arrowhead). XATH-1 expression was maintained within thestained with nile blue, and transferred to 0.51 MMR, 0.3% Ficoll. hindbrain at Stage 35, the latest stage examined (data notEmbryos were injected in a single blastomere at either the two- or shown). XATH-1 was also expressed in the developing tri-the four-cell stage. Three hundred picograms of MATH-1 mRNA, geminal ganglion and the otic vesicle (Figs. 2A and 2C; otic10 pg of X-NGNR-1, or 10 or 100 pg of XATH-1 mRNA was injected

vesicle, asterisk; trigeminal, arrow).in a volume of 5 nl. All embryos were co-injected with 100 pg ofCross sections of XATH-1 hybridized embryos reveal thatb-gal mRNA. Injected embryos were transferred to 0.11 MMR.

XATH-1 expression is specific to dorsal regions of the neuralStaged embryos were then fixed for 1 hr in MEMFA. For in situtube (Fig. 2D). Within these regions, XATH-1 is expressedhybridization, embryos were first analyzed for expression of co-

injected b-gal before transfer to methanol and storage at 0207C. both in progenitors at the ventricular surface and in cellslocated in lateral regions of the neural tube (Fig. 2D). Thesefindings suggest that XATH-1 expression initiates withinproliferative progenitors and is maintained in cells that haveRESULTSinitiated early stages of differentiation. Taken together,these data indicate that the sites of XATH-1 expression dur-Isolation of the Xenopus XATH-1 gene. A Xenopus ho-

molog of the MATH-1 gene, termed XATH-1, was isolated ing Xenopus development correlate with sites of MATH-1 expression during murine development (Akazawa et al.,in a low-stringency screen of a Xenopus genomic library

(Fig. 1A and Materials and Methods). XATH-1 is highly ho- 1995). However, XATH-1 expression appears more specificas it is not expressed in the developing spinal cord.mologous to MATH-1 within its bHLH domain, sharing

homology of greater than 98% (52/53 amino acids; Fig. 1B). Effects of XATH-1 overexpression during embryonic de-velopment. The function of XATH-1 during neural devel-Homology within the bHLH domain with Drosophila

atonal is 70%, while homology to both MATH-2 and neu- opment was investigated by overexpression in Xenopus em-bryos. Embryos were injected at the two- or four-cell stageroD is approximately 60% (Fig. 1B). Homology to other

members of the bHLH family is less than 40% within the in the animal pole of a single blastomere with either 10 or100 pg of in vitro transcribed XATH-1 mRNA. Both dosesbHLH domain (Fig. 1B). An acidic rich region of amino acid

homology located upstream of the bHLH domain, a putative of XATH-1 mRNA yielded similar results. In addition, theinjection of MATH-1 mRNA induced an identical pheno-regulatory motif, is shared among XATH-1, atonal, X-


AID DB 8572 / 6x24$$$162 06-12-97 15:18:26 dba

4 Kim et al.

FIG. 1. Sequence analysis of the XATH-1 protein coding domain. (A) The putative protein coding region of the XATH-1 gene as deducedfrom sequence analysis of a XATH-1 genomic clone. The bHLH region is underlined. An upstream acidic rich region with homology toneuroD is indicated by bold print. (B) Comparison of the XATH-1 bHLH domain to previously characterized genes including MATH-1(Akazawa et al., 1995), MATH-2 (Shimizu et al., 1995), atonal (Jarman et al., 1993), neuroD (Lee et al., 1995), and MASH-1 (Johnson etal., 1990). Asterisks represent conserved residues within the bHLH domain.


AID DB 8572 / 6x24$$8572 06-12-97 15:18:26 dba


FIG. 2. XATH-1 expression during Xenopus development. Whole mount in situ hybridization analysis was performed during early stagesof Xenopus development. The whole mount in situ hybridization pattern of XATH-1 expression at Stage 22 (A) and Stage 27 (B, C) isshown. Anterior is to the right. The midbrain/hindbrain boundary is indicated by an arrowhead (B). Expression in the otic vesicle isindicated by an asterisk and expression in the trigeminal ganglia is indicated by an arrow (A and C). A cross section of a Stage 27 embryo(D) shows expression within posterior regions of the developing hindbrain. Note that XATH-1-positive cells have exited the ventricularzone and are now found at the lateral edge of the neural tube.

type to XATH-1 overexpression within developing embryos. opment (0/9 embryos, data not shown), suggesting thatXATH-1 expression is not sufficient to drive neuronal differ-All embryos were co-injected with b-gal mRNA, which

served as a lineage tracer. Embryos positive for b-gal within entiation within gastrula stage neuroectoderm. At the neu-ral plate stage, N-tubulin is normally expressed in threethe neural tube and/or the epidermis were subjected to fur-

ther analysis. In agreement with previous observations, in- columns of primary neurons within the presumptive spinalcord (asterisks, Fig. 3A). In XATH-1-injected embryos, thejection of b-gal mRNA alone led to no alterations in the

normal pattern of marker gene expression (data not shown). domain of N-tubulin staining was expanded (Fig. 3A). Theectopic N-tubulin-positive cells were most prominent inXATH-1-injected embryos were first analyzed for expres-

sion of neuron-specific N-tubulin, an early marker of differ- intermediate and lateral regions of the neural plate and ex-tended into the epidermis (8/11 embryos). Following neuru-entiating neurons (Richter et al., 1988). In XATH-1-injected

embryos, no N-tubulin expression was detected prior to its lation, ectopic N-tubulin was apparent in the epidermis ofinjected embryos, often reaching the ventral extreme of thenormal onset of expression at the neural plate stage of devel-


AID DB 8572 / 6x24$$$163 06-12-97 15:18:26 dba

6 Kim et al.

06-12-97 15:18:26 dba


embryo (30/32 embryos; Fig. 3B). Within this region, a punc- XATH-1 bHLH domain, a region of nonhomology betweenXATH-1 and MATH-1. Ectopic XATH-1 expression wastate pattern of ectopic N-tubulin expression was observed

(Fig. 3B). This pattern is similar to that previously reported not observed in neural plate stage embryos (0/8 embryos;data not shown). In neural tube stage embryos injectedfor neuroD and X-NGNR-1 (Lee et al., 1995; Ma et al., 1996)

and as previously discussed (Lee et al., 1995) may result with MATH-1, ectopic XATH-1 expression was detectedspecifically within the epidermis (9/11 embryos) in a pat-either from the specific conversion of a subset of progenitors

(e.g., ciliated epidermal cells) or from the activation of lat- tern similar to, although more limited than that observedfor N-tubulin (Fig. 4A).eral inhibitory pathways within epidermal progenitors.

Cross sections of XATH-1-injected embryos further illus- The finding that MATH-1 can ‘‘autoregulate’’ XATH-1expression suggested the possibility that these genes mighttrate the high levels of ectopic N-tubulin expression ob-

served in epidermal regions of the embryo following neural induce regional neural markers within the epidermis spe-cific for the XATH-1 expression domain. We therefore ex-tube closure (Fig. 3C). In contrast, cross sections failed to

reveal any consistent ectopic, high-level expression of N- amined other genes normally expressed in the dorsal hind-brain, including PAX-3, engrailed-2, LH-2, and Wnt-3Atubulin within the neural tube of XATH-1-injected em-

bryos, although several embryos showed marginally in- (Goulding et al., 1991; Hemmati-Brivanlou et al., 1990;Wolda et al., 1993; Xu et al., 1993), in XATH-1-injectedcreased levels of N-tubulin in dorsal regions (Fig. 3C). In

embryos overexpressing XATH-1, a lateral expansion of the embryos. Engrailed-2 initiates expression at the neural platestage and is maintained in neural progenitors at the mid-neural tube was also noted (Fig. 3C).

The expression of two additional markers of differentiat- brain/hindbrain boundary following neural tube closure, in-cluding progenitors in the dorsal hindbrain that will giveing neurons, HNK-1 and Islet-1, was examined in XATH-1-

injected embryos. In Xenopus, HNK-1 is expressed on all rise to cerebellar granule neurons (Hemmati-Brivanlou andHarland, 1989; Hemmati-Brivanlou et al., 1990; Fig. 4B).neurons as they initiate process outgrowth (Nordlander,

1989, 1993) and Islet-1 is a LIM/homeodomain protein ex- PAX-3 expression initiates at gastrula stages within lateralregions of the neuroectoderm (data not shown) and persistspressed in a number of cell types, including Xenopus pri-

mary motor and sensory neurons (Ericson et al., 1992). In following neural tube closure within dorsal regions of theneural tube (Fig. 4C). Xenopus LH-2A is expressed in ante-neural tube stage embryos injected with XATH-1, ectopic

HNK-1 (11/12 embryos) and Islet-1 (15/19 embryos) staining rior regions of the neural tube and eye as well as in dorsolat-eral regions of the hindbrain, where expression is specificwas observed in the epidermis (Figs. 3D–3F). Furthermore,

a subset of these cells displayed a typical neuronal process- to differentiating neurons (Fig. 4D). Wnt-3A initiates ex-pression at the neurula stage and is localized to dorsal re-bearing morphology (Fig. 3G). We have observed similar

effects upon nonneural ectodermal progenitors fated to give gions of the neural tube (Wolda et al., 1993, data not shown).In XATH-1-injected embryos, none of these markers wasrise to cranial neural crest. Mesectodermal cranial crest

markers including twist, snail, and slug are down-regulated expressed ectopically, suggesting that XATH-1 is not suffi-cient to induce region-specific neural properties within non-and neuronal differentiation markers are ectopically ex-

pressed within these progenitors in embryos overexpressing neural ectodermal progenitors (Fig. 4 and data not shown).XATH-1 function is not dependent on X-NGNR-1 andXATH-1 (data not shown). Taken together, the ectopic ex-

pression of N-tubulin, Islet-1, and HNK-1 suggests that neuroD. The altered fate of nonneural ectodermal progeni-tors in embryos overexpressing XATH-1 to a large extentXATH-1 overexpression drives the appearance of markers of

the differentiated neuronal phenotype within cells normally parallels the effects of X-NGNR-1 and neuroD overex-pression in Xenopus embryos (Lee et al., 1995; Ma et al.,fated to give rise to nonneural ectodermal dervatives.

Expression of regionalized markers in XATH-1-in- 1996). As X-NGNR-1 activates the expression of neuroD(Ma et al., 1996), we investigated the expression of thesejected embryos. In order to examine the type of neuron

produced in response to ectopic XATH-1 expression, the genes in XATH-1-injected embryos. At the neural platestage of development, X-NGNR-1 and neuroD are expressedexpression of the endogenous XATH-1 gene was examined

in MATH-1-injected embryos. For these analyses, we used in domains of primary neurogenesis, coinciding with do-mains of N-tubulin expression in the embryo. Although thea probe prepared from sequences downstream of the

FIG. 3. Effects of XATH-1 and MATH-1 overexpression on Xenopus development. Either whole amount in situ hybridization (N-tubulin,X-NGNR-1, neuroD) or antibody staining (HNK-1, Islet-1) was used to analyze embryos injected with XATH-1 (A, C-I) or MATH-1 (B).Light blue staining within lateral epithelium represents b-gal-positive cells, dark purple is in situ reaction product (A–C, H–I), brown isthe product of antibody detection (D–G). In all panels, anterior is to the left and the injected side of the embryo is indicated by an arrow.Asterisks (A, H, and I) indicate columns of primary neurons. (A) N-tubulin, Stage 14. (B) N-tubulin, Stage 22. (C) Cross section throughtrunk showing N-tubulin staining within the neural tube and epidermis. (D) HNK-1, Stage 22. (E) Cross section through trunk region ofStage 27 embryo showing HNK-1-positive cells within the epidermis. (F) Islet-1, Stage 25. (G) Detail of process-bearing neurons withinthe epidermis of Stage 27 embryo. Neurons are stained for HNK-1. (H) X-NGNR-1, Stage 14. (I) neuroD, stage 16.


AID DB 8572 / 6x24$$$163 06-12-97 15:18:26 dba

8 Kim et al.

FIG. 4. Expression of cell type-specific markers in XATH-1- and MATH-1-injected embryos. Gene expression was characterized by insitu hybridization in MATH-1 (A)- and XATH-1 (B–D)-injected embryos. Anterior is to the left. Arrows point to injected side of theembryo. (A) XATH-1. (B) engrailed-2. (C) PAX-3. (D) LH-2A.

N-tubulin domain is expanded in XATH-1-injected embryos velopment (Kintner and Melton, 1987). XATH-1-injectedembryos were examined following neural tube closure, aat the neural plate stage of development (Fig. 3A), no expan-

sion of either X-NGNR-1 (7/7 embryos) or neuroD (13/14 time when N-tubulin-positive cells are seen within the epi-dermis of nearly all injected embryos (30/32 embryos). Inembryos) is observed in XATH-1-injected embryos (Figs. 3H

and 3I). Although neuroD expression was disorganized contrast to N-tubulin, high-level ectopic N-CAM expres-sion was not observed in the majority of XATH-1-injectedwithin cranial ganglia in anterior regions of XATH-1-in-

jected embryos following neural tube closure, ectopic ex- embryos (12/16 embryos; Fig. 5A). In these embryos, ectopicN-CAM was not visible within the epidermis even whenpression of X-NGNR-1 (7/7 embryos) or neuroD (6/8 em-

bryos) was not apparent within the epidermis of the major- normal expression within the neural tube and eye wereclearly visible (Fig. 5A). Although low levels of N-CAMity of embryos (data not shown). We also examined the

effects of X-NGNR-1 overexpression on XATH-1. In em- were detected within the epidermis after prolonged chromo-genic reactions (7/12 embryos), this expression contrastedbryos injected with X-NGNR-1, no ectopic expression of

XATH-1 was detected (6/6 embryos). As X-NGNR-1 induces with the ectopic expression of N-tubulin and HNK-1, asectopic N-CAM expression levels were barely detectableectopic neuroD expression in overexpression assays (Ma et

al., 1996), these results also suggest that XATH-1 does not over background and always well below that of the endoge-nous gene.function downstream of neuroD. Taken together, these

findings suggest that XATH-1 may function to activate a XATH-1 also failed to induce the expression of a secondearly neural marker, nrp-1. During normal development,novel pathway of neuronal differentiation within compe-

tent ectodermal progenitors. nrp-1 is expressed throughout the neural plate and is main-tained within the neural tube (Fig. 5D, Richter et al., 1990;Expression of early neural markers in XATH-1-injected

embryos. XATH-1 induces the expression of downstream Knecht et al., 1995). In addition, nrp-1 is transiently ex-pressed in the cranial ganglia (Knecht et al., 1995). In XATH-markers specific to differentiated neurons. We next exam-

ined the expression of early neural markers, including N- 1-injected embryos, high-level ectopic expression of nrp-1 was not detected within the majority of embryos (7/10CAM and nrp-1, that initiate expression earlier than XATH-

1 during normal development. N-CAM is a neural-specific embryos, Fig. 5C).In order to directly compare the effects of XATH-1 andmarker that initiates expression at the gastrula stage of de-


AID DB 8572 / 6x24$$$163 06-12-97 15:18:26 dba


FIG. 5. XATH-1 does not induce the expression of early neural markers. Expression of N-CAM and nrp-1 was analyzed by in situhybridization in XATH-1 (A, C)- and X-NGNR-1 (B)-injected embryos. Anterior is to the left. Light blue staining is specific to lacZ-positivecells; purple staining represents cells positive for in situ probes. Arrows indicate injected side of the embryo. (A and B) N-CAM. (C) nrp-1. In (D) the nrp-1 expression domain is shown at Stage 22 in a cross section from the hindbrain region of a normal embryo.

related bHLH genes, we examined the expression of N-CAM Xenopus homolog of atonal, termed XATH-1 based on itsin X-NGNR-1-injected embryos. In contrast to XATH-1, X- homology to the mammalian MATH-1 gene (Akazawa etNGNR-1 induced ectopic N-CAM expression within the al., 1995), that is expressed specifically in dorsal regions ofepidermis of a majority of embryos (4/5 embryos; Fig. 5B), the developing hindbrain as well as in the trigeminal gan-a finding similar to that previously reported for neuroD (Lee glion and the otic vesicle. In the dorsal hindbrain, neuronalet al., 1995). The differing induction of N-CAM expression progenitors initiate late stages of differentiation only afterin response to XATH-1, compared to X-NGNR-1 and neu- neural tube closure (Hartenstein, 1993). The time frame ofroD, further distinguishes the actions of this molecule from XATH-1 expression is therefore consistent with a role inthose previously described for bHLH family members. the neuronal differentiation process. Functional character-When coupled with our previous results, these findings sug- ization of XATH-1 in Xenopus embryos suggests that thisgest that XATH-1 preferentially regulates the expression of gene functions as a positive regulator of a subset of neuron-a subset of genes specific to differentiating neurons. specific targets. Furthermore, the ectopic expression of this

gene is sufficient to specify neural fate in ectodermal pro-genitors normally destined for nonneural fates, includingepidermis.DISCUSSION

Response to XATH-1 overexpression within the neuroec-toderm. XATH-1 function is apparently modified by cofac-A number of vertebrate homologs of the Drosophila neu-tors within the neuroectodermal territory. In contrast toral determination genes achaete-scute and atonal have beenneuroD and X-NGNR-1, XATH-1 does not induce preco-identified which appear to play varying roles in neural de-cious neuronal differentiation in gastrula stage neuroecto-velopment based on expression patterns and functional

characterizations. Our current studies have identified a derm, suggesting that necessary cofactors may be absent


AID DB 8572 / 6x24$$$163 06-12-97 15:18:26 dba

10 Kim et al.

at this stage of development. The induction of ectopic N- 1 transcriptional regulation (e.g., induction of high-level N-CAM expression), the correlation between the temporal pat-tubulin expression in neural plate stage embryos suggests

that neural progenitors gain competence to respond to tern of XATH-1 expression during normal development andthe specific induction of differentiated markers in responseXATH-1 overexpression at subsequent stages of develop-

ment. As X-NGNR-1 and neuroD, both markers of primary to XATH-1 overexpression suggests that XATH-1 may nor-mally function as a transcriptional regulator of a distinctneuron domains, are not expanded in the neuroectoderm of

XATH-1-injected embryos, the progenitors responding to group of downstream differentiation targets.Based on previous studies of neuroD, a stepwise modelXATH-1 are apparently precursors of secondary neuron pop-

ulations. of neuronal differentiation has been proposed in which ‘‘dif-ferentiation’’ genes can bypass early stages of neural speci-XATH-1 is not sufficient to induce region-specific prop-

erties within isolated neurons. The Drosophila achaete- fication to drive differentiation in competent progenitors(Lee et al., 1995). This model is based on the finding thatscute and atonal genes encode information for the specifi-

cation of particular classes of peripheral neurons (for review, neuroD, a gene not expressed at the neural specificationstage, can drive neuronal differentiation within nonneuralsee Jan and Jan, 1993). The neurons induced by XATH-1

within the epidermis provide a means of assessing its ability ectodermal derivatives (Lee et al., 1995). In contrast toXATH-1, neuroD induces the ectopic expression of bothto direct the differentiation of a particular subtype of neuron

in the absence of patterning signals within the neural tube. early neural specification markers, such as N-CAM, anddifferentiated neuronal markers, such as N-tubulin. ThisWe have examined a number of markers which overlap the

XATH-1 expression domain during normal development. model is then strengthened by our data, which show thatthe normal expression of early neural specification markersOur choice of markers was informed by fate mapping stud-

ies in mice which link MATH-1 expressing progenitors to can in fact be uncoupled from the expression of late neu-ronal differentiation markers in response to XATH-1.both LH-2-positive commissural neurons in the spinal cord

and engrailed-2-positive granule cell neurons in the cerebel- XATH-1 induces neuronal differentiation through a dis-tinct molecular mechanism. Our results indicate thatlum (Akazawa et al., 1995; A. Helms and J. Johnson, unpub-

lished data). Neither of these markers is ectopically ex- XATH-1 function is apparently independent of X-NGNR-1and neuroD. Moreover, XATH-1 induces low N-CAM, highpressed in response to XATH-1 overexpression. Although

these findings do not completely exclude the possibility N-tubulin-positive neurons distinct from those induced byX-NGNR-1 and neuroD (Lee et al., 1995; Ma et al., 1996).that XATH-1 induces a cell type for which we do not have

a specific marker, they suggest that ectopic XATH-1 expres- What do these molecular distinctions mean in the contextof normal development? One obvious difference betweension is not sufficient to induce the differentiation of several

neuronal subtypes with which it is associated. The spatially these molecules is that X-NGNR-1 and neuroD are ex-pressed during both primary and secondary neurogenesis inrestricted pattern of XATH-1 expression during normal de-

velopment may therefore reflect a role in specifying the the embryo, while XATH-1 is expressed only during latewaves of neurogenesis in the embryo. In contrast to primarytemporal pattern of neuronal differentiation within the

hindbrain rather than a role in the specification of a unique neuron populations which differentiate almost immediatelyafter neural induction, secondary neuron populations differ-neuronal identity.

XATH-1 overexpression does not induce high-level ex- entiate for prolonged periods during embryogenesis. XATH-1 may therefore control an aspect of neurogenesis uniquepression of early neural markers. In contrast to the ec-

topic expression of N-tubulin and other differentiated neu- to these later differentiating progenitor populations.ronal markers, high levels of N-CAM and nrp-1 are not in-duced within the epidermis in response to XATH-1overexpression. Although we cannot fully exclude the pos-

ACKNOWLEDGMENTSsibility that XATH-1 induces a neuronal subtype that nor-mally expresses low levels of N-CAM and nrp-1, we feelthat this is unlikely as both genes normally overlap the We gratefully acknowledge Drs. Mary Beth Hatten, Ali Hem-XATH-1 expression domain within the dorsal hindbrain (see mati-Brivanlou, David Anderson, Gordon Fishell, and Peter Fisher

for their valuable insights during the course of this work as wellFigs. 5A, 5C, and 5D). Furthermore, we have recently notedas for their editorial advice during the preparation of the manu-a similar discordance between ectopic N-tubulin and N-script. We are particularly grateful to the Hemmati-Brivanlou labCAM expression in response to XASH-3 overexpression,for providing numerous reagents used in these studies and for theirsuggesting that XASH-3 and XATH-1 may share commoncontinued expert technical assistance. In addition, we thank Dr.mechanistic elements (P. Kim and K. Zimmerman, unpub-M. Sargent for providing the Xenopus slug probe, Dr. Tom Jesselllished observations).for providing PAX-3 and LH-2 probes and Islet-1 antibody, Dr. Jane

During normal development, N-CAM and nrp-1 expres- Dodd for providing HNK-1 antibody, Drs. Qiufu Ma and Davidsion initiates at gastrula stages, earlier than the expression Anderson for providing X-NGNR-1 clones and Ari Gershon for pro-of XATH-1 and differentiated neuronal markers (Kintner viding the neuroD probe. K. Zimmerman also acknowledges Maryand Melton, 1987; Knecht et al., 1995). Although it is possi- Beth Hatten for her continued support and encouragement. A.W.H.

was supported by NIH Training Grant T32GM07062. This workble that epidermal factors inhibit specific aspects of XATH-


AID DB 8572 / 6x24$$$164 06-12-97 15:18:26 dba


was supported by an NIH First Award (R29 HD32968, K.Z.) and by Hopwood, N. D., Pluck, A., and Gurdon, J. B. (1989). A Xenopusthe Council for Tobacco Research, U.S.A. (J.E.J.). mRNA related to Drosophila twist is expressed in response to

induction in the mesoderm and the neural crest. Cell 59, 893–903.

Jan, Y., and Jan, L. (1993). HLH proteins, fly neurogenesis and verte-REFERENCESbrate myogenesis. Cell 75, 827–830.

Jarman, A. P., Grau, Y., Jan, L. Y., and Jan, Y. N. (1993). atonal isAbo, T., and Balch, C. (1981). A differentiation antigen of human the proneural gene that directs chondrotonal organ formation in

NK and K cells identified by a monoclonal antibody (HNK-1). J. the Drosophila peripheral nervous system. Cell 73, 1307–1321.Immun. 127, 1024–1029.

Johnson, J. E., Birren, S. J., and Anderson, D. J. (1990). Two rat ho-Akazawa, C., Ishibashi, M., Shimizu, C., Nakanishi, S., and Kage-

mologues of Drosophila achaete-scute specifically expressed inyama, R. (1995). A mammalian helix-loop-helix factor structur-

neuronal precursors. Nature 346, 858–861.ally related to the product of Drosophila proneural gene atonalKintner, C. R., and Melton, D. A. (1987). Expression of Xenopus N-is a positive transcriptional regulator expressed in the developing

CAM RNA in ectoderm is an early response to neural induction.nervous system. J. Biol. Chem. 270, 8730–8738.Development 99, 311–325.Bartholoma, A., and Nave, K.-A. (1994). NEX-1: A novel brain-

Knecht, A. K., Good, P. J., Dawid, I. B., and Harland, R. M. (1995).specific helix-loop-helix protein with autoregulation and sus-Dorsal–ventral patterning and differentiation of noggin-inducedtained expression in mature cortical neurons. Mech. Dev. 48,neural tissue in the absence of mesoderm. Development 121,217–228.1927–1936.Bradley, L. C., Snape, A., Bhatt, S., and Wilkinson, D. G. (1992). The

Krieg, P. A., and Melton, D. A. (1984). Functional messenger RNAsstructure and expression of the Xenopus Krox-20 gene: Conservedare produced by SP6 in vitro transcription of cloned cDNAs.and divergent patterns of expression in rhombomeres and neuralNucleic Acids Res. 12, 7057–7070.crest. Mech. Dev. 40, 73–84.

Krieg, P. A., Sakaguchi, D. S., and Kintner, C. (1989). Primary struc-Chitnis, A., and Kintner, C. (1996). Sensitivity of proneural genesture and developmental expression of a large cytoplasmic domainto lateral inhibition affects the pattern of primary neurons inform of Xenopus laevis neural cell adhesion molecule (NCAM)Xenopus embryos. Development 122, 2295–2301.Nucleic Acids Res. 17, 10321–10335.Ericson, J., Thor, S., Edlund, T., Jessell, T. M., and Yamada, T.

Lamb, T. M., Knecht, A. K., Smith, W. C., Stachel, S. E., Econo-(1992). Early stages of motor neuron differentiation revealed byexpression of homeobox gene Islet-1. Science 256, 1555–1560. mides, A. N., Stahl, N., Yancopolous, G. D., and Harland, R. M.

Ferreiro, B., Kintner, C., Zimmerman, K., Anderson, D., and Harris, (1993). Neural induction by the secreted polypeptide noggin. Sci-W. A. (1994). XASH genes promote neurogenesis in Xenopus em- ence 262, 713–718.bryos. Development 120, 3649–3655. Lee, J. E., Hollenberg, S. M., Snider, L., Turner, D. L., Lipnick, N.,

Goulding, M. D., Chalepakis, G., Deutsch, U., Erselius, J. R., and and Weintraub, H. (1995). Conversion of Xenopus ectoderm intoGruss, P. (1991). Pax-3, a novel murine DNA binding protein neurons by NeuroD, a basic helix-loop-helix protein. Science 268,expressed during early neurogenesis. EMBO J. 10, 1135–1147. 836–843.

Goulding, M. D., Lumsden, A., and Gruss, P. (1993). Signals from Ma, Q., Kintner, C., and Anderson, D. (1996). Identification of neuro-the notochord and floor plate regulate the region-specific expres- genin, a vertebrate neuronal determination gene. Cell 87, 43–52.sion of two Pax genes in the developing spinal cord. Development Nieto, M. A., Sargent, M. G., Wilkinson, D. G., and Cooke, J. (1994).117, 1001–1016. Control of cell behavior during vertebrate development by Slug,

Hamburger, V. (1988). ‘‘The Heritage of Experimental Embryology: a zinc finger gene. Science 264, 835–839.Hans Spemann and the Organizer.’’ Oxford Univ. Press, New Nieuwkoop, P., and Faber, J. (1967). ‘‘Normal Table of XenopusYork. laevis.’’ North–Holland, Amsterdam.

Harland, R. M. (1991). In situ hybridization: An improved whole- Nordlander, R. H. (1989). HNK-1 marks earliest axonal outgrowthmount method for Xenopus embryos. In ‘‘Methods in Cell Biol- in Xenopus. Dev. Brain Res. 50, 147–153.ogy’’ (B. Kay and H. Peng, Eds.), Vol. 36, pp. 685–695. Academic

Nordlander, R. H. (1993). Cellular and subcellular distribution ofPress, London.

HNK-1 immunoreactivity in the neural tube of Xenopus. J.Hartenstein, V. (1993). Early pattern of neuronal differentiation inComp. Neurol. 335, 538–551.the Xenopus embryonic brainstem and spinal cord. J. Comp. Neu-

Richter, K., Grunz, H., and Dawid, I. (1988). Gene expression inrol. 328, 213–231.the embryonic nervous system of Xenopus laevis. Proc. Natl.Hemmati-Brivanlou, A., and Harland, R. (1989). Expression of en-Acad. Sci. USA 85, 8086–8090.grailed-related protein is induced in the anterior neural ectoderm

Richter, K., Good, P. J., and Dawid, I. B. (1990). A developmentallyof early Xenopus embryos. Development 106, 611–617.regulated, nervous system-specific gene expressed in XenopusHemmati-Brivanlou, A., Frank, D., Bolce, M., Brown, B., Sive, H.,encodes a putative RNA-binding protein. New Biol. 2, 556–565.and Harland, R. (1990). Localization of specific mRNAs in Xeno-

Sargent, M. G., and Bennett, M. F. (1990). Identification in Xenopuspus embryos by whole-mount in situ hybridization. Develop-of a structural homologue of the Drosophila gene Snail. Develop-ment 110, 325–330.ment 109, 967–973.Hemmati-Brivanlou, A., de la Torre, J. R., Holt, C., and Harland,

Sasai, Y., Lu, B., Steinbeisser, H., and De Robertis, E. M. (1995).R. M. (1991). Cephalic expression and molecular characterizationRegulation of neural induction by the Chd and Bmp-4 antagonis-of Xenopus En-2. Development 111, 715–724.tic patterning signals in Xenopus. Nature 376, 333–336.Hemmati-Brivanlou, A., Kelly, O. G., and Melton, D. A. (1994). Fol-

Shimizu, C., Akazawa, C., Nakanishi, S., and Kageyama, R. (1995).listatin, an antagonist of activin, is expressed in the SpemannMATH-2, a mammalian helix-loop-helix factor structurally re-organizer and displays direct neuralizing activity. Cell 77, 283–

295. lated to the product of Drosophila proneural gene atonal, is spe-


AID DB 8572 / 6x24$$$165 06-12-97 15:18:26 dba

12 Kim et al.

cifically expressed in the nervous system. Eur. J. Biochem. 229, pression of Xwnt-3A and Xwnt-1 in neural tissue of Xenopuslaevis embryos. Dev. Biol. 155, 46–57.239–248.

Xu, Y., Baldassare, M., Fisher, P., Rathbun, G., Oltz, E. M., Yanco-Spemann, H. (1938). ‘‘Embryonic Development and Induction.’’poulos, G. D., Jessell, T. M., and Alt, F. W. (1993). LH-2: A LIM/Yale Univ. Press, New York.homeodomain gene expressed in developing lymphocytes andTurner, D. L., and Weintraub, H. (1994). Expression of achaete-neural cells. Proc. Natl. Acad. Sci. USA 90, 227–231.scute homolog 3 in Xenopus embryos converts ectodermal cells

Zimmerman, K., Shih, J., Bars, J., Collazo, A., and Anderson, D. J.to a neural fate. Genes Dev. 8, 1434–1447.(1993). XASH-3, a novel Xenopus achaete-scute homolog, pro-

Vize, P., Hemmati-Brivanlou, A., Harland, R., and Melton, D. vides an early marker of the planar neural induction and position(1991). Assays for gene function in developing Xenopus embryos. along the mediolateral axis of the neural plate. Development 119,In ‘‘Methods in Cell Biology’’ (B. Kay and H. Peng, Eds.), Vol. 36, 221–232.pp. 685–695. Academic Press, London.

Received for publication January 7, 1997Wolda, S. L., Moody, C. J., and Moon, B. T. (1993). Overlapping ex- Accepted March 21, 1997


AID DB 8572 / 6x24$$$165 06-12-97 15:18:26 dba

Date post:	17-Oct-2016
Category:	Documents
Upload:	peter-kim
View:	213 times
Download:	1 times

XATH-1,a Vertebrate Homolog ofDrosophila atonal,Induces Neuronal Differentiation within Ectodermal...

Documents