specification and morphogenesis of the zebrafish larval head skeleton

Developmental Biology 233, 239–257 (2001)doi:10.1006/dbio.2001.0201, available online at http://www.idealibrary.com on

REVIEW

Specification and Morphogenesis of the ZebrafishLarval Head Skeleton

Charles B. Kimmel,*,1 Craig T. Miller,* and Cecilia B. Moens†*Institute of Neuroscience, 1254 University of Oregon, Eugene, Oregon 97403-1254; and†Division of Basic Science and Program in Developmental Biology, Fred HutchinsonCancer Research Center, 1100 Fairview Avenue North, Seattle, Washington 98109

Forward genetic analyses can reveal important developmental regulatory genes and how they function to patternmorphology. This is because a mutated gene can produce a novel, sometimes beautiful, phenotype that, like the normalphenotype, immediately seems worth understanding. Generally the loss-of-function mutant phenotype is simplified fromthe wild-type one, and often the nature of the pattern simplification allows one to deduce how the wild-type genecontributes to patterning the normal, more complex, morphology. This truism seems no less valid for the vertebrate headskeleton than for other and simpler cases of patterning in multicellular plants and animals. To show this, we review selectedzebrafish craniofacial mutants. “Midline group” mutations, in genes functioning in one of at least three signal transductionpathways, lead to neurocranial pattern truncations that are primarily along the mediolateral axis. Mutation of lazarus/pbx4,encoding a hox gene partner, and mutation of valentino/kreisler, a hox gene regulator, produce anterior–posterior axisdisruptions of pharyngeal cartilages. Dorsoventral axis patterning of the same cartilages is disrupted in sucker/endothelin-1mutants. We infer that different signal transduction pathways pattern cartilage development along these three separate axes.Patterning of at least the anterior–posterior and dorsoventral axes have been broadly conserved, e.g., reduced Endothelin-1signaling similarly perturbs cartilage specification in chick, mouse, and zebrafish. We hypothesize that Endothelin-1 also isan upstream organizer of the patterns of cellular interactions during cartilage morphogenesis. © 2001 Academic Press

Key Words: craniofacial patterning; skeleton; head; specification; morphogenesis; Hox genes; Endothelin-1; zebrafish.

tbcgtmapfagiAp(fc

SPECIFICATIONS AND MORPHOGENESISGO HAND IN HAND

Development in multicellular animals includes assign-ments of fates to naive embryonic cells and the arrange-ment of these cells into higher level functional assem-blages. How cascades of developmental regulatory genesfunction to mediate these two kinds of processes, specifi-cation and morphogenesis, is currently under intense studyin a variety of experimental systems. A likely generaliza-tion is that controls of specification and morphogenesis areintimately interconnected. It would seem sensible that, atthe very least, fate specification must include instructionsabout morphogenesis along with instructions toward how

1 To whom correspondence and reprint requests should be ad-dressed. Fax: (541) 346-4548. E-mail: [email protected].

medu.

0012-1606/01 $35.00Copyright © 2001 by Academic PressAll rights of reproduction in any form reserved.

he cell should undergo functional specialization. Hence,oth differentiation and morphogenesis may ultimately beontrolled by the same upstream developmental regulatoryenes. A clear example in zebrafish is the T-box transcrip-ion factor Spadetail (Griffin et al., 1998). Loss-of-functionutation of the spadetail gene results in prominent cell-

utonomous defects in somitogenesis (a morphogeneticrocess) and myogenesis (muscle differentiation). spadetailunctions very early in development; expression is presentnd maintained specifically in presomitic mesoderm duringastrulation. Genetic targets under its positive regulationnclude myoD, a myogenic gene (Weinberg et al., 1996;macher and Kimmel, 1998), and a protocadherin (paraxialrotocadherin), a putative morphoregulatory geneYamamoto et al., 1998). Loss of these separate downstreamunctions correlates with, and possibly explains, the twolasses of phenotypic disturbances observed in spadetail
utants.
239

fhlstaebaaeaigw(tecse

240 Kimmel, Miller, and Moens

A SIMPLE SYSTEM APPROACH TO HEADSKELETAL PATTERNING

Below we explore development of the early head skeletonin zebrafish as a system for understanding hierarchicalgenetic control of organogenesis. Skeletal organs, bones andcartilages, have several attractive attributes for such analy-sis. In the skeletal system, perhaps more than any other,organ morphology is intimately connected with its func-tion. Because of this fact, the shapes and sizes of skeletalelements can be presumed to be under stringent selection,and the development of organ form exquisitely regulated.Indeed, because mineralized bone fossilizes readily, weunderstand much of the history of how skeletal elementshave evolved along various vertebrate lineages. Addingdevelopmental genetic analyses to this understanding iscurrently revealing how skeletal development itself mighthave evolved.

Facilitating a mechanistic analysis, the skeletal systemhas relatively few types of specialized cells. In youngzebrafish the system is further simplified: A functionallarval skeleton develops over the course of only a few days.The early skeletal elements are very small and mostly madeof cartilage, on which we focus here. Few cells of only twotypes, chondrocytes and perichondrial cells, comprise thesecartilages, with simple monolayered arrangements dis-cussed further below. Most of the cartilages are located notfar beneath the organism’s surface, where they can be easilyand directly visualized in the intact preparation as theyform. As in other vertebrates the cartilaginous head skel-eton has two prominent subdivisions, the neurocranium,protecting the brain and sensory organs, and the pharyngealskeleton, supporting the feeding and gill-breathing struc-tures (Fig. 1). Mutational analyses to be discussed below(see also Schilling, 1997) include both regions, and we alsodescribe a hypothesis of cartilage morphogenesis, the“joints build stacks” hypothesis, motivated from descrip-tive studies of chondrogenesis in wild-type and mutantembryos.

MIDLINE GROUP MUTATIONS PERTURBDEVELOPMENT OF THE BASICRANIALCARTILAGES

In his monumental 1894 study of monsters and thelessons we can learn from them about developmentalpatterning, Bateson included the “bulldog trout” (Fig. 2). Inthis deformed fish the anterior head is dramatically com-pressed. In contrast, the form and size of the lower jaw isnormal. Because of the disparity the lower jaw protrudeswell forward of the cranium, just as can be the case in agenuine bulldog: upper and lower facial parts are not wellmatched. The example reveals some local autonomy in thedevelopment of these two facial regions.

Bateson does not go on to examine the skeletal deformi-
ties of the bulldog trout directly, nor does the bulldog trout,
Copyright © 2001 by Academic Press. All right

ound by chance in a Scottish lake, tell us anything aboutow or why development went wrong. Just over 100 yearsater, “bulldog zebrafish,” i.e., developmental mutants,uch as the silberblick (slb) mutant shown in Fig. 3, providehe beginnings of an explanation. The bulldog phenotypeppears at the early larval stage (i.e., shortly after thembryo hatches, roughly at 3 days postfertilization, andegins postembryonic development). Similar phenotypesre observed in larval fish homozygous for loss-of-functionlleles at any of several separate chromosomal loci. Inter-stingly, the mutants were all first identified by defectsrising earlier, during embryogenesis, and were all placednto a common embryonic phenotypic class, the “midline”roup (Brand et al., 1996). The mutants were classified thisay because, irrespective of the genetic pathway affected

see below), they all shared embryonic phenotypes knowno result from perturbed development at the midline. Forxample, mutant embryos in the midline group all exhibityclopia: the eyes are closer to the midline than normal,ometimes fused together at the midline. The strong infer-nce, recognized by Brand et al. (1996), is that defective

midline signaling during embryonic development underliesthe later craniofacial bulldog phenotype as well as theearlier, embryonic phenotypes.

Although the bulldog facial phenotype might not itselfhint at involvement of the midline in patterning, theskeletal phenotype immediately does so. Staining the larvalcartilages with Alcian blue reveals a profound neurocranialdeficiency that underlies the bulldog face in slb mutants(Piotrowski et al., 1996). The defects show most clearly in adissected out and flat-mounted preparation (Figs. 3C and3D). In the wild-type early larva, the basicranium anteriorto the region known as the polar cartilages consists of abilateral pair of rods, the trabeculae, that are separated inthe midline by a prominent hypophyseal fenestra. Thetrabeculae come together more anteriorly and they join oneanother, fusing together in the midline in a region termedthe trabeculae communis. More anterior still this pairedmidline cartilage flares outwards laterally as the ethmoidplate. In slb mutants this pattern is simplified. Ahead of thepolar cartilage region there is only a single cartilaginous rodpresent just at the midline and just about the diameter of asingle trabecula.

In cellular terms, a complex and precisely shaped two-dimensional array of cartilage is converted into a muchsimpler one-dimensional array when a functional geneproduct (a Wnt protein; see below) is absent. The change inthe basicranium appears to be very indirect, due ultimatelyto an early mesodermal signaling defect that perturbs for-mation of the prechordal plate, i.e., the midline mesodermof the head rudiment (see below). That head skeletal defectsmight be due to defective development of the prechordalplate would not be news to embryologists experimentingwith salamander embryos nearly 70 years ago. They showedthat defects of the same nature, in particular a fusion of thepair of trabeculae within the midline, resulted from treat-
ment of early embryos with teratogenic chemicals (e.g.,
s of reproduction in any form reserved.

241Zebrafish Larval Head Skeleton

FIG. 1. The young larval zebrafish (A, 5 days postfertilization, left side view) and the layout of its cartilaginous head skeleton. For a furtherdescription see Schilling and Kimmel (1997) and Cubbage and Mabee (1996). Anterior is to left in this and other figures in the paper exceptwhere indicated. B shows a ventral view of an Alcian blue-labeled, whole-mounted preparation. Cartilages are indicated in the first ormandibular arch (pq, m) and in the second or hyoid arch (hs, ch). C–E show drawings of such preparations. (C) The neurocranial (orbasicranial) cartilages and notochord from the dorsal aspect. The eyes fit into the shallow grooves along the sides of the ethmoid plate andtrabeculae. The otic vesicles fit into the prominent cavities to either side of the notochord and parachordal cartilages . The brain’s posteriorpituitary fits into the prominent midline cavity ahead of the notochord, the hypophysial fenestra. (D). The neurocranium (diagrammaticallyelevated dorsalward for the sake of clarity) and the pharyngeal skeleton in side view. (E). The pharyngeal skeleton in ventral view.Abbreviations used: A, anterior; bb, basibranchial; bh, basihyal; cb, ceratobranchial; ch, ceratohyal; D, dorsal; ep, ethmoid plate; hb,hypobranchial; hs, hyosymplectic; ih, interhyal; L, lateral; M, medial; m, Meckel’s; not, notochord; P, posterior; pch, parachordal; pq,palatoquadrate; tr, trabecula; V, ventral.
FIG. 2. The “bulldog-headed trout,” a craniofacial monster found in nature. From Bateson (1894).
Copyright © 2001 by Academic Press. All rights of reproduction in any form reserved.


LiCl) that perturbed early midline development and alsoproduced cyclopia (review: De Beer, 1937, p. 447).

Considering the midline group of mutants collectively,one observes a range of severity of the phenotypic defects.Figure 4 shows this range for a subset of the midline group

FIG. 3. The “bulldog-headed zebrafish,” comparing the wild-ty(unpublished photographs, courtesy of Dr. Corinne Houart). (C, D)as a flat mount, viewed from the dorsal aspect (unpublished, B. Ullmbetween the parachordals and trabeculae. The polar cartilages foGoodrich, 1930), but probably not in zebrafish. The abbreviationsFIG. 4. A phenotypic series of a subset of midline group mutants, c(igu), arranged in approximate order of severity. Dorsal views, as in Fare at separate genetic loci. The abbreviations are as in Fig. 1 (mod

mutants, representing the genes schmalspur, detour, cha-


melion, you-too, and iguana. The presentation followsBrand et al. (1996), who arranged the mutants into aphenotypic series by ordering them according to the sever-ity of skeletal reduction. This arrangement revealed quite acurious sequence with respect to the pattern change as well

T, A) and slb mutant (B) embryos at 2 days of developmentcranial cartilages labeled with Alcian blue, dissected, and laid outand C.B.K.); pc, polar cartilage region; a distinctive region of joinings separate elements in some organisms (e.g., see De Beer, 1937;s in Fig. 1.eleon (con), detour (dtr), schmalspur (sur), you-too (yot), and iguanaC. Two examples of con mutants are included; the other mutations

from Brand et al., 1996, with sur added).

pe (WBasiannrm aare ahamig. 1ified

(Fig. 4). Along the series, except for schmalspur, there


onpcl(Fpuptib

lcedbaddttppmntnialcscbs

d

ptc


seems to be progressive loss of cartilage from both anteriorand lateral positions. Furthermore, the loss seems not torespect the names assigned to the cartilage regions. That is,the ethmoid plate shrinks in both its anterior–posterior andmediolateral aspects, and in the same mutants the trabec-ulae shorten along the anterior–posterior axis as well andare progressively displaced from lateral to a median posi-tion.

The somewhat coordinated anterior–posterior andmedial–lateral pattern truncation seems worth further ex-ploration. Not all mutants might obey the same rule. Brandet al. (1996) report that the schmalspur mutant phenotypeis distinctive in that more posterior defects (midline carti-lage fusion in the region of the polar cartilages) withoutcorresponding anterior changes, i.e., the ethmoid plateappears normal.

According to Brand et al., the defects in the midline groupf mutants show up along the entire length of the ventraleurocranium. However, we are impressed that significantattern changes they show are all anterior to the para-hordals: the transition from unaffected to affected carti-ages occurs rather abruptly at the polar cartilage regionFig. 3C). In even their most severe example, iguana (igu,ig. 4) the cartilages posterior to this transition look ap-roximately the same as those in the wild type. Yet onlynpatterned islands of cartilage are present anterior to theolar cartilage region in iguana mutants. We conclude thathe mutants in this midline group are revealing genesnvolved in pattern regulation of particularly the anteriorasicranium.If the skeletal disturbances to the basicranium are indeed

imited in anterior–posterior extent, as we just argued, howan we understand why this might be so? Following Brandt al. (1996), suppose that the cartilage defects follow fromefects in the embryonic midline. Hence the differenceetween anterior and posterior basicranium could be due ton anterior–posterior difference in the midline, due to aifference in the cells responding to the midline signal, orue to a difference in both signaling and responding cellypes together. Indeed there is evidence for changes in bothypes: The putative developmental boundary region, theolar cartilage region (Fig. 3), develops just ahead of arominent and famous transitional zone in the embryonicidline mesoderm, a known signaling center. Here the

otochord ends and the prechordal plate begins. Further,he developmental origin of the cells forming the neurocra-ial cartilages might change in the same location. Accord-ng to fate mapping studies in avians (Le Lievre, 1978) andmphibians (Chibon, 1967), and extirpation experiments inamprey (Languille and Hall, 1986, 1988), cranial neuralrest forms the trabeculae but not the parachordals, whicheemingly come from paraxial mesoderm. Hence, the polarartilage region may mark an important transitional zone inoth the midline signaling cells and the responding pre-
umptive cartilage cells. h

MIDLINE GROUP GENES FUNCTION INTHREE DIFFERENT SIGNALTRANSDUCTION PATHWAYS

Subsequent to their discoveries in mutagenesis screens, anumber of the genes in the midline group have beenmolecularly identified. silberblick encodes a Wnt-11 or-tholog (Heisenberg et al., 2000). cyclops encodes a Nodalortholog (TGFb superfamily, Rebagliati et al., 1998; Sam-path et al., 1998; review Schier and Shen, 2000). one-eyedpinhead encodes a Crypto ortholog (EGF-CFC gene family;Zhang et al., 1998; review: Shen and Shier, 2000) thatfunctions downstream to Nodal signaling. schmalspur en-codes a FoxH1/FAST1 homolog (winged-helix transcriptionfactor; Pogoda et al., 2000; Sirotkin et al., 2000) and also inthe Nodal signal-transduction pathway. syu is an orthologof sonic hedgehog (Schauerte et al., 1998), and yot encodesa Gli-2 ortholog (Karlstrom et al., 1999), a transcriptionfactor regulated by Hh signaling.

Hence, the mutations, where known, identify genes thatfunction in cell–cell signaling pathways, and at least threesignal transduction pathways are involved. From expressionanalysis it seems likely that two, possibly all three, of thesesignals do not act directly on the cartilage-forming cells. Ifsuch is the case, this is an important lesson for understand-ing craniofacial mutant phenotypes: the genes responsiblefor severe craniofacial phenotypes might be remote tocartilage development itself.

Both cyc/nodal and syu/shh are expressed by the anteriormidline mesoderm, but slb/wnt11 has a complex expres-sion pattern, including expression in head neural crest thatmight eventually form the anterior basicranium. However,the perhaps more important expression domain is withinearly paraxial mesoderm, before mesoderm has migrated toreach where the head will form (Heisenberg et al., 2000).Here it acts to turn on a “noncanonical” Wnt pathwayinvolving intracellular Ca21 release and a G-protein-ependent activation of kinases (review: Kuhl et al., 2000)

to mediate cell polarity and polarized cell movement. Thesemovements underlie the early extension of the embryonicaxis and the initial formation of the midline head meso-derm in the late gastrula (Heisenberg and Nusslein-Volhard,1997; Heisenberg et al., 2000). Such movements go wrongin slb/wnt11 mutants, and the prechordal plate is severelydefective. Such a function, at ca. 8 h postfertilization (h), isremote indeed from head cartilage formation, chondrogen-esis occurring near the end of embryogenesis, 2 days later.

Similarly, the role of cyc/nodal signaling must be indirecton the cartilage-forming cells. Here expression is present inprechordal plate, i.e., just at the region adjacent to wherethe anterior basicranium forms, but the time of expressionseems wrong. The cyc/nodal gene is expressed as theresumptive mesodermal cells move anteriorward alonghe axis during gastrula, but shortly after this migration isompleted expression is downregulated. This occurs several
ours before neural crest migration begins, such that it

ntsm

tlFcshacaslgpt


seems unlikely that the crest cells ever encounter theCyc/Nodal signal when they arrive at their destinationbeneath the forebrain. This timing might mean that cyc/

odal signaling serves to set up a subsequent midline signalhat in turn affects neural crest migration and/or cartilagepecification at the midline. Supporting this scenario, aarker for premigratory neural crest, crestin, is downregu-

lated in cyc mutants (Rubinstein et al., 2000).Syu/shh expression begins in mesoderm in the early

midgastrula, and like cyc/nodal expression, expression ofsyu/shh is downregulated after the prechordal plate meso-derm has migrated to underlie forebrain neurectoderm.Hence it also seems unlikely that a Syu/Shh signal fromprechordal plate acts on cartilage-forming neural crest.However, we propose that in fact defective Syu/Shh signal-ing is a part of the proximate cause of the cartilage dysmor-phogenesis in all of these mutants. In response to inductionby the prechordal plate (and that clearly depends on cyc/nodal functioning in the prechordal plate; Hatta et al.,1994) the responding neurectoderm begins a course ofdevelopment that includes its own upregulation of syu/shh.This expression persists through the time (pharyngulastages) when cartilage-forming cells move into the neigh-borhood of the ventral neurectoderm and are potentiallyready to respond. Hence, in this possible scenario the Shhsignal that acts on postmigratory cartilage-forming neuralcrest comes from the ventral ectoderm of the primordialforebrain, not the prechordal plate. The distinctive surmutant phenotype mentioned above is partly explained bythis model: In most of the midline group mutants expres-sion of syu/shh is perturbed throughout the anterior ventralforebrain. In sur mutants specifically, syu/shh expression ismissing just in a gap that corresponds in anterior–posteriorlevel to the basicranial defect (Brand et al., 1996).

That signaling from neurepithelium can promote neuro-

FIG. 5. Cartilage phenotypes and dlx2 expression in WT andflat-mounted cartilages of the first and second pharyngeal segmenanterior to the left. In the WT (A) the elements are separated byelements as well as the ventral elements in the two segments are fuDV joints are missing, as is the interhyal cartilage (ih) in the secolabeled by RNA in situ hybridization for expression of dlx2. The to the three streams of postmigratory neural crest that have populatzr mutant.IG. 6. Cartilage phenotypes and hox gene expression in WT andartilages in 7-day-old larvae (from Moens et al., 1998). (A) The Wegment) uniquely contains a small interhyal (ih) cartilage. At thisypobranchials. The ceratohyal (ch) is distinctively larger in size thls are tapering elements, as shown in B for cb1 (third pharyngeal sartilages in the same (third) segment in three individual val mutannd thickened, now more like a ceratohyal (C). Separate hypobranchometimes present (arrowheads in D and E). (F–I) hox gene expressioeft (from Prince et al., 1998). (F, G) hoxb2 at 19 h (20-somite stagene hoxb2 (F, G) is ectopically expressed in neural crest that wiosterior gene hoxb3 (H, I) normally has a distinctive domain of str
he mutant (arrowheads).

cranial cartilage formation and participate in its patterningis well known from transplantation studies (review: Thoro-good, 1983). Furthermore, in craniofacial cartilage develop-ment there has been demonstrated an essential role of Shh(Hu and Helms, 1999). The model predicts that one mightbe able to rescue the basicranial defects of any of themutants by locally supplying a source of Shh to the post-migratory cranial crest.

The phenotypes that we have been discussing are largelylimited to the neurocranium. In particular, they do notinclude the pharyngeal cartilages, and in some respects thisis surprising. For example, an extended process of the dorsalcartilage in the first pharyngeal arch (the palatoquadrate’spterygoid process) articulates with the anterior lateral re-gion of the ethmoid, a region missing in slb/wnt11 mu-tants. Hence we expected to see a corresponding deforma-tion of the palatoquadrate in these mutants, but their studydid not reveal any change (B. Ullmann and C.B.K., unpub-lished findings). Perhaps only the severest of the midlinegroup mutants have pharyngeal cartilage phenotypes: cyc/nodal mutants have extensive deletion of the anteriorbasicranial cartilage (like the igu phenotype shown in Fig.4), and here there are extensive cartilage fusions in themidline of the mandibular arch as well (C.T.M., unpub-lished findings).

REGULATION OF hox GENE FUNCTIONAND AP PATTERNING

The segmentally organized pharyngeal cartilages of thelarval zebrafish have a primitive anterior–posterior (AP)organization, with paired dorsal and ventral elements in thefirst and second arches and paired ventral elements in thefive more posterior gill arches (Fig. 1). The elements in the

utants, (from Popperl et al., 2000). (A, B) Alcian blue-stained,pared from day 4 embryos. Left side views, dorsal to the top and

s. The abbreviations are as in Fig. 1. In the mutant (B) the dorsalto one another. DV fusions within each segment occur as well; thegment. (C, D) Dorsal views (anterior to the left) of 20-h embryosseparate bilateral patches of expression in the WT (C) corresponde pharyngeal arch primordia. The streams are fused together in the

mutants. (A–E) Ventral views, anterior to the top, of flat-mountedharyngeal segments 2–7. The hyoid segment (second pharyngeal, there is no separate hypohyal cartilage, the serial homolog of thesegmental homologs, the ceratobranchials (cb1–5). Ceratobranchi-nt) in another WT preparation at higher magnification. (C–E) The

he principal element, normally a ceratobranchial is often truncatedmay be missing (arrow in D), and small interhyal-like elements areWT (F, H) and val (G, I) mutants, dorsal views with anterior to the, I) hoxb3 expression at 14 h (10-somite stage). The more anteriorpulate the third pharyngeal segment (arrowhead in G). The moreexpression in rhombomeres 5 and 6, and this domain is missing in

lzr mts prejointsed

nd sehreeed th

valT p

stagean itsegmets. Tialsn in

e). (Hll poong



first and second arches are morphologically unique and thuscan be used as differentiated markers of AP identity withinthe head periphery.

Anterior–posterior patterning in vertebrate embryos, asin Drosophila, depends on the Hox genes, whose overlap-ping expression domains demarcate regions of positionalidentity (reviewed in Lumsden and Krumlauf, 1996). Hoxgene expression is regulated in a complex and interdepen-dent manner (reviewed in Nonchev et al., 1997; Studer etal., 1998). At the top of the hierarchy lies an essentialresponsiveness, on the part of particular Hox genes, toretinoic acid (RA; reviewed in Marshall et al., 1996; Gouldet al., 1998). Although a gradient of RA along the AP axishas been difficult to demonstrate, the phenotypes of em-bryos in which RA synthesis or responsiveness is blocked,either genetically or pharmacologically, strongly suggests arequirement for RA in vertebrate AP patterning (Niederrei-ther et al., 2000; White et al., 2000; Gale et al., 1999; vander Wees et al., 1998).

Fusions and duplications of pharyngeal cartilages of quitedifferent sorts than those described above are observed inmutants in which Hox gene function is altered, due tomutations in either Hox partners or in upstream regulators.The cartilage fusions in lazarus (lzr) mutants are among themore spectacular (Figs. 5A and 5B; Popperl et al., 2000). Inwild types the cartilages in the adjacent pharyngeal seg-ments (adjacent arches) are separate elements from oneanother, although as shown in Fig. 5A, the dorsal cartilagesin the hyoid and mandibular arches are close neighbors. Thedorsal and ventral cartilages in each segment are separate aswell, and they articulate at a prominent joint (this jointregion including the small interhyal cartilage in the hyoidarch). In lzr mutants the hyoid and mandibular cartilagesare quite thoroughly fused to one another, dorsal withdorsal, ventral with ventral, and dorsal with ventral.Clearly the processes that underlie formation of isolatedchondrogenic islands in wild types are defective in lzrmutants, implying that the function of the wild-type gene isto pattern cartilage islanding. The islanding mechanism isnot lost altogether: there are none of the fusions at themidline in lzr mutants, such as are present in mutants ofthe midline group.

What is this islanding function (or functions)? lazarusencodes a protein of the Pbx homeodomain family (Pbx4;Popperl et al., 2000), expressed ubiquitously. Such proteinsare known to interact directly with Hox proteins (reviewedin Mann and Chan, 1996). Hox and Pbx proteins form adimeric DNA-bound complex in the regulation of transcrip-tional targets (Passner et al., 1999; Piper et al., 1999). Henceone might expect pbx genes to function in the same generalmanner as hox genes. However, we caution that some ofpbx gene-mediated functions in Drosophila do not dependon hox gene interaction (Casares and Mann, 1998; Abu-Shaar and Mann, 1998; Gonzalez-Crespo et al., 1998). Infact we do not know, but only suppose, that in patterningcartilage development the wild-type Lzr/Pbx4 protein
works with a Hox protein partner.

The dorsal to dorsal and ventral to ventral cartilagefusions occur along the anterior–posterior axis, and theycan be related to defects in segmental patterning and inanterior–posterior specification of identity in a straightfor-ward manner. In wild-type zebrafish, as in other verte-brates, at least some of the neural crest that will laterdevelop as cartilage migrates from the dorsal lateral marginof the neural plate in three streams present along the APaxis. These streams can be visualized by expression of theneural crest marker dlx2 (Figs. 5C and 5D) and are alsoapparent in time lapse recordings (C.B.K. and R. Keynes,unpublished observations). In lzr/pbx4 mutants, at least asassayed by dlx2 expression, the streams are fused together;the crest appears to migrate as “a single, uninterruptedsheet, . . . with neural crest cells populating the normallycrest-free zones lateral to the otic vesicle and to r3” (Popperlet al., 2000).

This fusion of the streams of migrating crest that willform the pharyngeal cartilages might directly underlie thecartilage fusions that occur in the AP axis. We suppose thatthe anterior two streams in wild types give rise, in arestricted manner, to the postmigratory arch neural crest(the ectomesenchyme) of the corresponding anterior twopharyngeal segments, the mandibular and hyoid. Evidencefor this supposition in zebrafish comes from clonal analysis:single premigratory neural crest cells contribute progeny toone or the other of these pharyngeal segments but not toboth together. The segments, by this analysis, are celllineage compartments (Schilling and Kimmel, 1994). Dis-ruption of compartmentation, as would seem to be the casein lzr/pbx4 mutants, might be at least partly responsible forthe AP fusions of cartilage elements in the first and secondarches. Time lapse and cell lineage analyses in the mutantare required to ascertain whether individual neural crestcells contribute to both first and second arch cartilages, aswould be predicted by this model.

A further issue is whether the fusion of crest streams, andtherefore, as we propose, of cartilage elements, is itself dueto an intrinsic defect in the specification of AP identity inpremigratory neural crest populations or to the absence oflzr-dependent signals that would otherwise serve to sepa-rate the crest into streams. Hox genes, and by inference,lzr/pbx4, are clearly implicated in the former mechanism,such that an AP “hox code” established within the hind-brain is carried into the periphery by the neural crest (Huntet al., 1991). It is easy to imagine how the downstreameffectors of Hox/Pbx4 complexes could result in the mutualrepulsion of crest streams based on their AP identity.However there is also support for a model by which cranialcrest is separated into streams by extrinsic influences. Anumber of groups have demonstrated the existence ofnon-cell-autonomous signals, acting either within the hind-brain or within the mesoderm of the head periphery, thatinfluence the organization of neural crest into streams(Farlie et al., 1999; Graham et al., 1993; Sechrist et al.,1993). Recent chimeric analysis of the ErbB4 mutant mouse
demonstrated that the mismigration of a late population of

(

p(w

(bTbv(mFL5iw


neural crest cells in that mutant results from the absence ofsignals that normally function to keep these streams sepa-rate (Golding et al., 2000). The primary role of ErbB4, areceptor expressed in hindbrain rhombomeres 3 and 5, maytherefore be to establish a crest-inhibitory domain adjacentto those segments. Although lzr/pbx4 mutants have arelated phenotype, the lzr/pbx4 expression pattern offerslittle in the way of an indication of where it might primarilyfunction to control crest migration. Whether lzr/pbx4 func-tions autonomously within the neural crest or nonautono-mously in its environment remains to be determined bygenetic mosaic analysis. In either case, dysfunction of hoxgenes is implicated by the lzr/pbx4 mutant phenotypePopperl et al., 2000).

While lzr/pbx4 mutants exhibit a simplified cartilageattern that may result from loss of hox function, valentino

val) mutants undergo a cartilage duplication that correlatesith gain of hox function. val encodes a bZIP transcription

factor, homologous to the mouse kreisler (kr) gene, and aregulator of hox gene expression within hindbrain rhom-bomeres 5 and 6 (Moens et al., 1998; Cordes et al., 1994;Manzanares et al., 1997, 1999a, b; McKay et al., 1994). Inval mutants, an ectopic cartilage resembling the smallinterhyal cartilage characteristic of the hyoid arch is asso-ciated with the dorsalmost aspect of the ceratobranchialcartilage in the third pharyngeal arch (Figs. 6A–6E; Moenset al., 1998). Normally the third and more posterior archesare simplified versions of the first two arches, containingonly ventral elements homologous to Meckel’s and theceratohyal. The ectopic putative interhyal in the third archis reminiscent of ectopic cartilage elements on the hyoidbone of kreisler mutant mice (Frohman et al., 1993).

The cartilage duplication in val/kr mutants correlatesclosely with changes in hox expression. In both the fish andthe mouse, loss of val/kr function results in the ectopicexpression in the third arch of hox genes normally in thesecond arch, and the loss of expression of hox genes nor-mally in the third arch. Thus hoxb2, which is normallyexpressed in the second arch at 19 h, is ectopically ex-pressed in val/kr mutants in a population of crest cellsmigrating into the third arch, while hoxb3 expression in the

FIG. 7. Phenotype and gene expression in WT and suc mutants (frA, B) The craniofacial appearance in left side view of WT and alue-labeled, flat-mounted elements of the mandibular and hyoid ahe asterisks in D show regions of ventral cartilage that may correar, 100 mm. (E) Whole mount RNA in situ preparation (36 hentralpharyngeal arches. The first four segments are indicated (1–F–K) Targets of suc signaling, as revealed by comparing the RNA

ounted embryos. (F, G) goosecoid (gsc), 30 h. (H, I) dHAND, 28 hIG. 8. Chondrification of the dorsal (pq) and ventral (m) cartilageeft side views of Alcian blue-labeled whole-mounted embryos, pho3 h. The eye is at the top of the field and the rudiment of the addus visible. Other abbreviations are as in Fig. 1. The dorsal cartilage, t
ithin the same mandibular condensation that will also form Meckel

region of the hindbrain that gives rise to third arch crest isreduced (Figs. 6H and 6I; Prince et al., 1998). The loss ofhoxb3 expression is a direct effect of loss of val/kr function,as kr has been shown to be a direct regulator of hoxb3expression in the mouse (Manzanares et al., 1997). Theectopic hoxb2 expression may result from the loss, in val/krmutants, of almost all of r5. As a result, r4-derived crest,which normally migrates anterior to the otic vesicle andinto the second arch, migrates caudal to the otic vesicle andinto the third arch as well, carrying with it hoxb2 expres-sion (Figs. 6F and 6G) and, by inference, the wherewithal tospecify an interhyal cartilage. Alternatively, the r6-derivedthird crest stream could be homeotically transformed intosecond arch identity.

Mutations in hox genes themselves have not been de-scribed so far in zebrafish. Targeted mutagenesis of Hoxa2in the mouse shows a crucial role for this gene in patterningalong the anterior–posterior axis (Gendron-Maguire et al.,1993; Riijli et al., 1993). The gene is expressed in the hyoidbut not the mandibular pharyngeal segment. Hoxa2 mu-tants have a homeotic, anterior transformation phenotype:skeletal derivatives of the mandibular segment are dupli-cated and derivatives of the hyoid segment are deleted,revealing its role as a selector gene specifying segmentalidentity of these anterior pharyngeal segments. Gain-of-function analyses have been carried for Hoxa2 in the chickand Xenopus (Grammatopoulos et al., 2000; Pasqualetti etal., 2000), and the results of these studies are entirelyconsistent with it’s functioning as a homeotic selectorgene, for the phenotypes are essentially the opposite ofthose of the loss-of-function mutants. Thus, when trans-genic techniques are used to express Hoxa2 in the mandib-ular segment, the cartilages are homeotically transformedto a hyoid phenotype, i.e., there is posterior transformation.

We cannot so easily reconcile dorsal–ventral patterningdefects such as we observe in lzr mutants with hox genepatterning. Recently Hoxa2 and Hoxb2 have been impli-cated in dorsoventral patterning in the mouse hindbrain(Davenne et al., 1999), but it is not known if they have acorresponding role in the pharyngeal arches. As we come to

iller et al., 2000). WT are to the left and suc mutants to the right.utant at day 4. (C, D) The cartilage phenotype at day 4. Alcian(pharyngeal segments 1 and 2). The abbreviations are as in Fig. 1.

d to m and ch in the first and second segments, respectively. Scalet side view) showing segmental expression of suc/et-1 in thed labeling is also present in endodermal pouches 2 and 3 (p2, p3).

ession patterns in WT and suc mutants. Left side views of wholeK) dlx2 (in red) and EphA3 (in blue), 32 h. Scale bar (K), 50 mm.hin a single precartilage condensation in the mandibular segment.phed with Nomarski optics (from Schilling and Kimmel, 1997). (A)mandibular muscle (am, mediating movement between pq and m)latoquadrate (pq), has begun to chondrify (become Alcian-positive)

om Msuc mrchesspon, lef4), anexpr. (J,

s wittogractorhe pa

’s cartilage (m). An unlabeled bridge of condensed mesenchyme,



present just beneath the muscle rudiment, connects these two labeled regions and shows there is but a single condensation. See also Fig.10, for the appearance of the condendensation in a live embryo. This bridging region will eventually develop as the joint between pq andm. (B) Another embryo at about the same stage in which m is also lightly labeled, and the joint-forming bridge remains unlabled. Thearrowhead indicates the hyomandibular pouch, separating the mandibular and hyoid arches. (C) 60 h. Both cartilages, but not the jointregion between them, are strongly labeled. Scale bar, 50 mm.



FIG. 9. A phenotypic series of anterior arch mutants, arranged in approximate order of severity (unpublished examples, C.T.M; see alsoKimmel et al., 1998). (A) Alcian-stained flat mounts of mandibular and hyoid cartilages. Abbreviations and orientation are as in Fig. 1D.
The three mutations are at three separate loci; sucker (suc) encodes Endothelin-1; sturgeon (stu) and schmerle (she) have not been identified

wam

tvt1amTcekchhgp

gps

1apaaoEsabpcNaSo(

ifnsgttamgoa

t


next, however, it is clear that dorsal and ventral cartilageshave separate identities.

A CONSERVED ENDOTHELIN-1 SIGNALSPECIFIES VENTRAL CARTILAGEIDENTITY IN THE PHARYNGEAL ARCHES

One hundred and nine mutations were isolated in alarge-scale genetic screen in Tubingen by their prominentcraniofacial phenotypes (Schilling et al., 1996; Piotrowskiet al., 1996; e.g., Figs. 7A and 7B). Staining the mutants withAlcian blue revealed the underlying defects in the pharyn-geal cartilages (e.g., Figs. 7C and 7D, see also below, Fig. 9).Among them, 6 mutations representing four genes namedsucker (suc), sturgeon (stu), schmerle (she), and hoover (hoo)

ere placed together into a single phenotypic “anteriorrch” class because they primarily affected cartilages in theandibular and hyoid arches (Piotrowski et al., 1996).Anterior arch mutants share three related defects in both

he mandibular and hyoid arches. Ventral cartilages areariably reduced in size, changed in orientation, and fusedo the dorsal ones (Piotrowski et al., 1996; Kimmel et al.,998), i. e., the joint normally present between the dorsalnd ventral cartilages is missing. Mutation of suc causes theost severe loss of first and second arch ventral cartilages.his phenotype suggested an Endothelin-1 (Et-1) gene as aandidate for harboring the suc mutation, because in mousembryos homozygous for targeted disruption of Et-1, Mec-el’s cartilage in the mandibular arch and the ventralartilage in the hyoid arch are both severely reduced (Kuri-ara et al., 1994). That is, the craniofacial phenotypes ofomozygous mouse Et-1 mutants and zebrafish homozy-ous suc mutants are essentially identical. Genetic map-ing of suc revealed conserved synteny of the relevant

molecularly. The mutants are arranged by severity of the phenotyphe least severe and suc2 the most severe. (B) Drawings illustrating

the two segments; green, ventral cartilages; blue, the DV joint regiventral cartilages in the mutants (see text). By this hypothesis, ipolarities of the ventral cartilages (green arrows), which we sucondensations, become progressively more like that of the dorsal eFIG. 10. Morphogenesis (chondrogenesis) in the anterior pharypreparations (left side view). (A) 52 h, (B) 58 h (unpublished confallowing the various regions of developing cartilages to be pseudocosame orientation in Fig. 1D. Here the two distinctive regions oseparately, the strut-like symplectic region (sy) and the plate-like hyfor more detail see Kimmel et al., 1998) is left uncolored. Both ofsegment, the ceratohyal (ch), are developing out of a common primoMeckel’s cartilage (only partly included here) can be seen to be deveas described previously from analysis of fixed material (Schilling aFIG. 11. Abnormally curving chondrocyte stacks in the ventral rpattern in a WT (to the left). Arrows indicate proposed polarities (seecartilages when Suc/Et-1 signaling is defective is underlain by abnoret al., 1998). hm and sy, hyomandibular and symplectic regions of
bar, 100 mm.

enomic regions between zebrafish and humans, and map-ing, sequencing, and rescue analyses demonstrated thatuc encodes a zebrafish Et-1 ortholog (Miller et al., 2000).Endothelin-1 is a secreted peptide (Yanagisawa et al.,

988), and marker analyses suggest that in the pharyngealrches it is secreted by cells that are close neighbors ofostmigratory neural crest cells in the ventral pharyngealrches (Miller et al., 2000). Here the peptide could serve aspositional signal to specify the crest toward developmentf ventral cartilages (review: Francis-West et al., 1998).xpression of suc/et-1 begins during late segmentationtages (ca. 16 h), several hours after crest migration beginsnd many hours before cartilage formation (chondrogenesis)egins. suc/et-1 expression is localized to at least threeharyngeal arch tissues—the paraxial mesodermal archores, the surface ectoderm, and pharyngeal endoderm.eural crest itself does not appear to express the gene. In

greement with this finding, mosaic analysis suggests thatuc/Et-1 signaling function is required in the environmentf the neural crest cells, not autonomously in the crestMiller et al., 2000).

At the time (pharyngula stages; ca. 22–28 h) when signal-ng and response seems to occur, the crest cells that willorm the cartilages shown in Fig. 7C have completed (orearly completed) their migration. Observed gene expres-ion defects in suc/et-1 mutants begin at this later, postmi-ratory stage and ET-1 can rescue ventral cartilage forma-ion in suc/et-1 mutants as late as 28 h, well after most ofhe cranial neural crest has completed migration (Miller etl., 2000). Hence the signal seems unimportant for crestigration; rather, it may act in specification at a postmi-

ratory stage. The same conclusion was reached in studiesf the Et-1 receptor, EdnrA, function in mice (Clouthier etl., 2000).Correlating with the later defects being restricted to more

s judged by the amount of cartilage present in the mutants; stu2 isinterpretation of the mutant phenotypes: Red, dorsal cartilages inissing in mutants. We propose there is a change in polarity of the

T the dorsal and ventral cartilages have opposite polarities. Thee indicates their direction of outgrowth from the precartilagents (red arrows) with increasing severity of the phenotypes.l wall, as revealed by BODIPY ceramide labeling in live WTmicrographs, M. Jones and C.B.K.). Negative images are shown,for clarity of presentation. The mature cartilages are shown in thedorsal hyoid cartilage (the hyosymplectic, Fig. 1) are indicated

ndibular region (hm). The more dorsal part (hyomandibular region;cartilage-forming regions, as well as the ventral cartilage in this

m, the hyoid condensation. In contrast, the palatoquadrate (pq) andg out of another shared primordium, the mandibular condensation,immel, 1997; see also Fig. 8).

of the hyoid cartilage of a suc mutant embryo, compared to the), as in Fig. 9. We propose that the abnormal polarities of the ventralorientation of stacking of the chondrocytes (modified from Kimmelyosymplectic cartilage. Other abbreviations are as in Fig. 1. Scale

es, aour

on, mn Wpposlemengeaocalloredf theomatheserdiu

lopinnd Kegiontextmalthe h


YG(d

aeul


ventral cartilages, suc/et-1 is expressed only ventrally in theembryonic pharyngeal arches (Fig. 7E). There, as we pro-pose, it acts to specify the postmigratory neural crest cellsto their ventral identities. By this model a high level of theSuc/Et-1 signal is a necessary requirement in the specifica-tion of ventral cartilage identity. Whether this signal issufficient for ventral specification could be tested in gain-of-function experiments, but so far no such experimentshave been reported.

MOLECULAR TARGETS ANDDEVELOPMENTAL ROLES OFENDOTHELIN-1 SIGNALING

How is the Suc/Et-1 signal transduced into a “ventralidentity” response? The Et-1 signal transduction pathwayincludes a membrane-bound metalloprotease endothelin-converting enzyme-1 (ECE-1) (Xu et al., 1994). This en-zyme, present in both cells that make Et-1 and cells thatreceive Et-1, cleaves a longer precursor form of big Et-1 intoits secreted mature 21-amino-acid form (Xu et al., 1994;

anagisawa et al., 1998). A seven-transmembrane,-protein-coupled receptor, endothelin type A receptor

EdnrA), is expressed by the responding cells. Targetedisruption in mice of either ECE-1 (Yanagisawa et al., 1998)

or of EdnrA (Clouthier et al., 1998) produces a craniofacialphenotype indistinguishable from that of Et-1 mutants.Furthermore, pharmacological inhibition of EDNRA inavian embryos inhibits ventral arch cartilage formation,suggesting that a role for Et-1/EDNRA signaling has beenhighly conserved in gnathostome evolution (Kempf et al.,1998). As yet, genes homologous to Ece-1 and EdnrA havenot been described in zebrafish; they are, of course, candi-dates for other, as yet molecularly unknown, anterior archgenes identified in the mutant screens (see below).

The signal transduction pathway following EdnrA recep-tor activation by Et-1 has been well characterized in phys-iological contexts (see Huggins and Pelton, 1997; Pollockand Highsmith, 1998, for review). However, signal trans-duction downstream of EdnrA during embryonic develop-ment is largely unknown. Regardless of how the pathwayworks, a prominent downstream response is well under-stood, a transcriptional activation of target genes. Like theSuc/Et-1 signal itself, this response to the signal by neuralcrest-derived cartilage progenitors has been evolutionarilyconserved to a large extent between mouse and zebrafish.Targets in both species include homeobox genes of the Dlxand Msx families, the homeobox gene goosecoid (gsc), andthe bHLH gene dHAND (Fig. 7; Clouthier et al., 1998, 2000;Thomas et al., 1998; Miller et al., 2000).

In zebrafish, there clearly are several distinct styles oftarget regulation. Some genes, including dHAND, msxE,and dlx3, begin expression specifically in the ventral archmesenchyme of the pharyngula (Figs. 7H and 7J). Thedomains of expression are all similar and correlate both
temporally and spatially with suc/et-1 expression itself.

Further, the expression of these genes is severely reduced insuc/et-1 mutants (Figs. 7I and 7K). Hence, as we suppose,these genes are activated in these ventral arch domainsspecifically by the Suc/Et-1 signal. In contrast, regulation ofthe gene dlx2, which in combination with dlx1 is criticalfor arch patterning (Qui et al., 1995, 1997), seems morecomplex. Expression of dlx2 begins before suc/et-1 expres-sion is detected and may be present in all migrating crest,including that destined for the dorsal domain as well as theventral domain. Later, the ventral but not dorsal expressionof dlx2 comes under the positive control of suc/et-1, asrevealed by downregulation of ventral but not dorsal dlx2expression in suc/et-1 mutants. A third variation on thetheme is of gsc. This gene comes on well after the Suc/Et-1signal (at about 28–30 h), first in the ventral domain andthen in the dorsal domain (Fig. 7F). Ventral but not dorsalexpression depends on the signal, as revealed by the missingventral expression domain in suc/et-1 mutants (Fig. 7G).

Pharyngeal arch expression of mammalian orthologs ofmany of these genes also requires Et-1/EdnrA signaling. InEt-1 and EdnrA mutant mice, ventral (distal) arch dHANDexpression is dramatically affected (Thomas et al., 1998;Clouthier et al., 2000). EdnrA mutant mice also lack Gscand Dlx3 expression (Clouthier et al., 1998, 2000). Severaldifferences exist, however; for example, Et-1 mutant micehave normal Msx1 expression (Thomas et al., 1998),whereas an msx gene in fish seems to more directly requireEt-1 signaling (Miller et al., 2000).

A wonderfully interesting issue for future studies is tolearn how the upregulation of these transcription factors bysuc/et-1 relates to patterning the later cartilage morphogen-esis that the mutant phenotype shows us is controlled bythe signal. Mutagenesis of some of these target genes in themouse (e.g., Gsc, Dlx2, dHAND; Yamada et al., 1995;Rivera-Perez et al., 1995; Qiu et al., 1995; Thomas et al.,1998) and fish (dHAND, C.T.M., D. Yelon, D. Stainier, andC.B.K, unpublished results) reveals that their functions arealso essential for properly patterned craniofacial chondro-genesis. What is missing at this point for better understand-ing is a careful comparative analysis of the phenotypes ofembryos bearing mutations in the upstream and down-stream genes. In zebrafish, an interesting beginning of suchpathway analysis can be made by differences in the pheno-types of suc/et-1 mutants and other mutants in the anteriorarch class. The latter are generally less severe than suc/et-12 with respect to the loss of ventral cartilage (Kimmel etl., 1998; see below). Where the suc/et-1 targets have beenxamined in these other mutants (e.g., in schmerle, C.T.M.,npublished work) they are also observed to be downregu-ated, but not to such a great extent as in suc/et-12. Hence

these studies support the proposals that the genes allfunction in the same pathway (Piotrowski et al., 1996;Kimmel et al., 1998) and further that transcriptional regu-lation of target genes in the cells responding to the Suc/Et-1signal is quantitatively correlated with ventral cartilagedevelopment.

Another suc/et-1 target in zebrafish that immediately


d

ibtv

wstsasicggtTAiet

ilmsctsdvt


connects to morphogenesis is the gene EphA3, expressedlike dHAND in the ventral mesenchyme specifically andownregulated in suc/et-1 mutants (Figs. 7J and 7K). Eph

genes encode receptor tyrosine kinases. They are activatedby cell-bound ligands, ephrins, and generally Eph activationresults in an avoidance response of the Eph-expressing celltoward the ephrin-expressing neighbor. Such avoidanceshave been shown to be important for establishing polarity,e.g., in the polarized spread of Eph-expressing retinal axonsacross an ephrin-expressing midbrain tectum, the synaptictarget tissue of these axons (review: Holder and Klein,1999). Eph–ephrin-mediated repulsion can also function indevelopmental boundary formation and segmental restric-tion (Robinson et al., 1997). The best known case is in thehindbrain. This tissue is made from a segmental series ofrhombomeres (review; Guthrie, 1995), and mixing of cellsof adjacent rhombomeres is restricted (Fraser et al., 1990;Birgbauer and Fraser, 1994). Correlated with the restriction,cells on one side of a developing rhombomere boundaryexpress an Eph and cells on the other side the correspondingephrin (Xu et al., 1999). When this situation is perturbed, asin overexpression studies (Xu et al., 1995; Mellitzer et al.,1999), the restrictions toward mixing are lifted. Further-more, in the valentino mutant discussed earlier, loss ofrhombomere boundaries correlates with the loss of normalephrin–Eph boundaries in the r4-r7 region of the hindbrain(Cooke et al., 2001).

We can readily suppose that ephrin–Eph mediation ofcellular patterning, including avoidance interactions at de-velopmental boundaries, is also a feature of cartilage devel-opment and is under control of Et-1 signaling. As notedabove, a prominent phenotype of all of the anterior archmutants is the loss of the joint region that normallyseparates the dorsal and ventral cartilages in each segment.Imagine that the joint is the site of a developmentalboundary present in wild-type embryos but missing inmutants. The position where the joint appears correspondswith the border of the EphA3 expression domain. We knowthat the dorsal and ventral cartilages in a single segmentarise as two separate sites of chondrification within a singleprecartilage condensation (Fig. 8; Schilling and Kimmel,1997). Further, the joint develops between these sites as aspecial region of cells that do not chondrify; rather, thedorsal and ventral zones remain separate as chondrificationspreads through the condensation (Fig. 8). Hence the jointregion behaves as a morphogenetic dorsal–ventral boundarywithin the precartilage condensation. The presence ofEphA3 on the cells on the ventral side of this putativedevelopmental boundary predicts an appropriate ligand(e.g., ephrinA2 or ephrinA3) on the dorsal side, if indeed theEph–ephrin system is included in the molecular machinerythat patterns the joint.

Finally, downregulation of EphA3 could also at leastpartly underlie what we propose is a polarity change in theanterior arch mutants. A prominent aspect of the suc/et-1mutant phenotype is the orientation of the remnant of
persisting ventral cartilage (Fig. 9; Kimmel et al., 1998). As r

we indicate by the direction of the green arrows in Fig. 9,the regions of cartilage we interpret as a reduced Meckel’scartilage in the mandibular arch and ceratohyal in the hyoidarch both point in a posterior direction rather than ananterior direction, as in the wild type. Correlating with theless severe loss of ventral cartilage in she and stu mutantsthe orientations of the same cartilage regions in thesemutants are more normal, stu2 more so than she2. Hence,n a way that matches severity of the phenotype measuredy the amount of ventral cartilage present in the mutantshere is a progressively more severe polarity change of theentral cartilage in the series stu2 , she2 , suc/et-12. As

explained next, the polarity change could be revealing areversal in the direction of outgrowth of the ventral carti-lage during morphogenesis. The reversal could well be aconsequence of a missing repulsive interaction betweencells that normally have distinct ventral and dorsal identi-ties.

MORPHOGENESIS: JOINTS MAKE STACKS

In the suc/et-1-dependent regulation of EphA3 we envis-age a particular mode of cellular behavior that producesspecific cartilage form. Namely, repulsive cellular interac-tions occur at the incipient dorsal–ventral joint region andare ultimately responsible, in part anyway, for morphoge-netic behaviors carving out separate dorsal and ventralelements. As we propose to be the case in suc/et-1 mutants,

hen the behaviors fail, so too does morphogenesis. Repul-ion between cells may be part of the suite of interactionshat occur during the stages of “chondrogenesis,” whenpecified but still functionally immature (or “undifferenti-ted”) mesenchymal cells in so-called precartilage conden-ations form specifically shaped elements of now special-zed chondrocytes and cartilage matrix. Past authors haveonceptually separated cartilage cytodifferentiation (histo-enesis) and cartilage shaping (morphogenesis; e.g., Thoro-ood, 1983; Noden et al., 1999). Features of both can occurogether in the development of the zebrafish head skeleton.he embryonic cartilage matrix begins to be labeled withlcian blue just as the embryo enters the third day follow-

ng fertilization (Fig. 8). Nearly simultaneously, we also seearly signs of cells taking on regular lined-up arrangementshat could underlie cartilage shaping.

Patterning can be complex from the outset and exquis-tely precise. For example, we have shown, studying Alcian-abeled material, that in the mandibular and hyoid seg-

ents on each side of the pharynx separate chondrogenicites develop stereotypically out of a single precartilageondensation (Schilling and Kimmel, 1997). The pattern forhe mandibular segment was discussed above, where wehow two separate chondrogenic sites present in the con-ensation represent the future dorsal palatoquadrate andentral Meckel’s cartilage (Fig. 8). In the hyoid segmenthere are three sites, two of them corresponding different
egions of the dorsal hyosymplectic cartilage and the third

iiwd

ston“g

wtamd(acopdinlidoteA

fcagbaslldfpa

ectbnupa


site representing the ventral cartilage in this segment (theceratohyal). In newer, unpublished studies we have recentlyconfirmed this pattern by using confocal microscopy toimage the developing condensations in live embryosstained vitally with BODIPY ceramide. Figure 10 shows anexample of this imaging at two times separated by about6 h. So far we have studied one of these chondrogenic sitesin the most detail, a region of the dorsal cartilage termedthe symplectic (Kimmel et al., 1998). It has the simpleststructure, for the symplectic develops as a single file ofcartilage cells (as shown for a stage about 2 days later in Fig.11, sy). We first recognize the symplectic rudiment when ithas as few as two or three cells that emerge out of theprecartilage condensation; about five are present in Fig.10A. These cells appear at a specific location of the conden-sation, just where the dorsal–ventral joint will eventuallydevelop, and more cells add to the symplectic row asdevelopment continues (Fig. 10B). Using lineage tracer dyeto label individual cells in the developing symplectic (Kim-mel et al., 1998), we found that lengthening of the cell rownvolves progressive addition of cells, not randomly alongts length, but preferentially at or near the joint region, fromhich the symplectic thus appears to grow outward. Theirection of outgrowth is indicated by the arrow in Fig. 11.This pattern can be understood by proposing that the

pecific neighborhood of the condensation that later formshe dorsal–ventral joint acts, at the earlier stage, as anrganizing center for chondrogenesis. Cells in the jointeighborhood are recruited to join into cell rows, orstacks,” perhaps by intercalating with one another, sug-ested from the lineage studies (Kimmel et al., 1998). By

this hypothesis the symplectic develops as a single stackelongating in a ventral direction out of the joint neighbor-hood. This “joints make stacks” hypothesis can readilyaccomodate the polarity change discussed above in suc/et-1mutants: We imagine that the polarized stacking of cellsmaking the ventral cartilage remnants is disrupted. Stacksnow emerge from the mutant joint region with reversedorientation (Fig. 11). Confocal time-lapse analyses of chon-drogenesis in wild-type and mutant embryos could directlytest the hypothesis.

By the joints makes stacks hypothesis the patterning ofcartilage shape is proximately controlled, at least in part,within the precartilage condensation itself, rather than bythe surrounding environment. If so, environmental signalswould act to organize the primordium at an earlier stage.For the Suc/Et-1 signal, such timing nicely fits the earlyexpression observed of suc/et-1—expression is presentmore than a day before chondrogenesis is initiated. We asyet have no evidence for later primordium-intrinsic versus-extrinsic control of patterning, but older experiments inother species support intrinsic regulation. As studied par-ticularly in avian embryos, precartilage condensations putinto organ cultures develop into cartilages of recognizablydifferent shapes according to the specific rudiment isolated(Jacobson and Fell, 1941; Weiss and Amprino, 1940; re-
viewed in Thorogood, 1983).

The mode of cartilage development just described,herein separate cartilages arise within a single condensa-

ion, also occurs in the mandibular arch and during cranialnd pharyngeal development of other species of fish (Bert-ar, 1959). However, a different pattern is present in limb

evelopment of a variety of tetrapods. Shubin and Alberch1986) found that separate cartilages within the limb eachrise from separate Alcian blue-positive condensations. Theondensations themselves arise by a characteristic patternf branching and budding from an earlier single Alcian-ositive condensation. So the relative timing of matrixeposition versus element formation seems quite differentn the two modes. However, we think that the difference isot really a fundamental one, for we see an example of whatooks like the limb bud pattern in the hyoid arch. Thenterhyal cartilage, positioned between the hs and the ch,evelops about a day later than its neighbors, and it devel-ps by budding out of the young hs. In fact it comes fromhe same dorsal–ventral joint region we have been consid-ring, but only after this region has also developed anlcian-positive matrix.

PROSPECTUS

For each of the kinds of mutants we considered, thephenotypes are pattern deformations, simplified from thewild-type pattern in particular ways rather than just beingrandom losses of pattern elements. The mutations are allloss-of-function alleles. Connecting these two observationssuggests directly how the wild-type functions of the par-ticular genes are contributing to pattern complexity.

The midline group mutants (represented by slb/wnt112,or example) deform a bilateral two-dimensional sheet ofomplex morphology into a one-dimensional array justlong the midline. The implied wild-type function of theenes of this group is to position the initial normallyilateral sites of chondrogenesis of the anterior basicraniumnd perhaps by specifying where migrating neural crest willettle down, as further studies could reveal. Further, mo-ecular studies reveal that these genes function in intracel-ular signaling pathways crucial for embryonic midlineevelopment and suggest a particular scenario in which Shhrom the ventral neural tube is a likely proximate influenceatterning cartilage development along the mediolateralxis.Cartilage fusions in lzr/pbx4 mutants correlate with

arlier fusions of segmental streams of migrating neuralrest that will form these cartilages. Hence we suggest thathis gene functions in establishment of developmentaloundaries along the anterior–posterior axis that restrictseural crest migration and that restricted migration in partnderlies the formation of separate cartilages in adjacentharyngeal segments. Because Pbx and Hox proteins inter-ct, hox gene function is implicated. The same is the case

for val/kr function; we can interpret the cartilage pheno-
type in val/kr mutants as being due to either a mismigra-

B

B

B

B

C

C

C

C

C

C

C

D

D

F

F

F

F


tion of neural crest along the AP axis or to a change in APidentity; either interpretation is consistent with defectivehox gene-dependent patterning.

The anterior arch mutant (e.g., suc/et-1) phenotypes sim-plify pattern by reducing development of ventral pharyngealcartilages and fusing them into the dorsal ones. Along withthe molecular work establishing where and when the Suc/Et-1 signal must act, the studies show a crucial signalingthat occurs relatively late in the embryo: The local envi-ronment acts upon a primordium of postmigratory neuralcrest. Genes that could directly affect cartilage morphogen-esis (EphA3) are controlled by this signal.

The mutational analyses have revealed genes that func-tion in several different genetic pathways. Further, by thephenotypic changes we see that each genetic pathwayconnects to patterning along only one of the three coordi-nate axes of the embryo—medial–lateral (Wnt, Nodal,Hedgehog), anterior–posterior (Pbx, Val, Hox, and henceperhaps involving retinoic acid), and dorsal–ventral (Et-1).These findings motivate the hypothesis that head cartilagemorphology arises through largely separate control of pat-tern formation along these three separate axes. Character-izing new mutants is a way to learn whether this proposalhas any validity.

All of the genes identified so far act in pattern specifica-tion, apparently rather distantly upstream of cartilage mor-phogenesis itself. Hence the mutants now available maynot be very useful for examining details of control ofmorphogenesis: e.g., what underlies the assembly of cellrows postulated by the joints make stacks hypothesis?Upstream functions might be expected for some of themutants, like ones reviewed above with pleiotropic pheno-types that include disruption of embryogenesis before car-tilage progenitors are even present. Hence, a way to findlater acting genes is to screen later and screen more specifi-cally for very discrete cartilage phenotypes. Screens byAlcian blue labeling of larval cartilages are in progress, andit is clear from initial results that mutants with very subtlephenotypes can be identified and recovered and that theprevious screens are far from saturation. Hence the prospectfor continued progress is bright indeed.

ACKNOWLEDGMENTS

We are very grateful to Bonnie Ullmann and Martha Jones, whocontributed to new studies described here. Unpublished work wassupported by NIH Grants DE13834, NS17963, and HD22486.

REFERENCES

Abu-Shaar, M., and Mann, R. (1998). Generation of multipleantagonistic domains along the proximodistal axis during Dro-sophila leg development. Development 125, 3821–3830.

Amacher, S. L., and Kimmel, C. B. (1998). Promoting notochord fateand repressing muscle development in zebrafish axial mesoderm.
Development 125, 1397–1406.

ateson, W. (1894). “Materials for the Study of Variation.” Mac-millan, London.

ertmar, G. (1959). On the ontogeny of chondral skull inCharacidae, with a discussion on the chondroncranial baseand the visceral chondrocranium in fishes. Acta Zool. 40,203–364.

irgbaur, E., and Fraser, S. E. (1994). Violation of cell lineagerestriction compartments in the chick hindbrain. Development120, 1347–1356.

rand, M., Heisenberg, C. P., Warga, R. M., Pelegri, F., Karlstrom,R. O., Beuchle, D., Picker, A., Jiang, Y. J., Furutani-Seiki, M., VanEeden, F. J. M., Granato, M., Haffter, P., Hammerschmidt, M.,Kane, D. A., Kelsh, R. N., Mullins, M. C., Odenthal, J., andNuesslein-Volhard, C. (1996). Mutations affecting developmentof the midline and general body shape during zebrafish embryo-genesis. Development 123, 129–142.asares, F., and Mann, R. (1998). Control of antennal versus legdevelopment in Drosophila. Nature 392, 723–726.hibon, P. (1967). Marquage nucleaire par la thymidine tritiee desderives de la crete neurale chez l’amphibien urodele Pluerodeleswaltlii Macihah. J. Embryol. Exp. Morphol. 18, 343–358.louthier, D. E., Hosoda, K., Richardson, J. A., Williams, S. C.,Yanagisawa, H., Kuwaki, T., Kumada, M., Hammer, R. E., andYanagisawa, M. (1998). Cranial and cardiac neural crest defects inendothelin-A receptor-deficient mice. Development 125, 813–824.louthier, D. E., Williams, S. C., Yanagisawa, H., Wieduwilt, M.,Richardson, J. A., and Yanagisawa, M. (2000). Signaling pathwayscrucial for craniofacial development revealed by Endothelin-Areceptor-deficient mice. Dev. Biol. 217, 10–24.ooke, J. E., Moens, C. B., Roth, L. W. A., Durbin, L., Shiomi, K.,Brennan, C., Kimmel, C. B., Wilson, S. W., and Holder, N. (2001).Eph signalling functions downstream of Val to regulate cellsorting and boundary formation in the caudal hindbrain. Devel-opment 128, 571–580.ordes, S. P., and Barsh, G. S. (1994). The mouse segmentation genekr encodes a novel basic domain-leucine zipper transcriptionfactor. Cell 79, 1025–1034.ubbage, C. C., and Mabee, P. M. (1996). Development of thecranium and paired fins in the zebrafish Dano rerio (Ostari-ophysi, Cyprinidae). J. Morphol. 229, 121–160.avenne, M., Maconochie, M. K., Neun, R., Pattyn, A., Chambon,P., Krumlauf, R., and Rijli, F. M. (1999). Hoxa2 and Hoxb2 controldorsoventral patterns of neuronal development in the rostralhindbrain. Neuron 22, 677–691.e Beer, G. R. (1937). “The Development of the Vertebrate Skull.”Oxford Univ. Press, Oxford. Reprinted 1985, Chicago Univ.Press, Chicago.

arlie, P. G., Kerr, R., Thomas, P., Symes, T., Minichiello, J., Hearn,C. J., and Newgreen, D. (1999). A paraxial exclusion zone createspatterned cranial neural crest cell outgrowth adjacent to rhom-bomeres 3 and 5. Dev. Biol. 213, 70–84.

rancis-West, P., Ladher, R., Barlow, A., and Graveson, A. (1998).Signalling interactions during facial development. Mech. Dev.75, 3–28.

raser, S., Keynes, R., and Lumsden, A. (1990). Segmentation in thechick embryo hindbrain is defined by cell lineage restrictions.Nature 344, 431–435.

rohman, M. A, Martin, G. R., Cordes, S. P., Halamek, L. P., andBarsch, G. S. (1993). Altered rhombomere-specific gene expres-sion and hyoid bone differentiation in the mouse segmentation
mutant, kreisler (kr). Development 117, 925–936.

G

G

G

G

G

G

G

G

G

H

H

H

H

H

H

H

H

J

K

K

K

K

K

L

L

L

M

M

M

M

M

M

M

M

M


Gale, E., Zile, M., and Maden, M. (1999). Hindbrain respecificationin the retinoid-deficient quail. Mech. Dev. 89, 43–54.endron-Maguire, M., Mallo, M., Zhang, M., and Gridley, T.(1993). Hoxa-2 mutant mant mice exhibit homeotic transforma-tion of skeletal elements derived from the cranial neural crest.Cell 75, 1317–1331.olding, J. P., Trainor, P., Krumlauf, R., and Gassmann, M. (2000).Defects in pathfinding by cranial neural crest cells in micelacking the neuregulin receptor ErbB4. Nature Cell Biol. 2,103–109.onzalez-Crespo, S., Abu-Shaar, M., Torres, M., Martinez, A. C.,Mann, R. S., and Morata, G. (1998). Antagonism between extra-denticle function and Hedgehog signalling in the developinglimb. Nature 394, 196–200.oodrich, E. S. (1930). “Studies on the Structure and Developmentof Vertebrates.” Macmillan. London.ould, A., Itasaki, N., and Krumlauf, R. (1998). Initiation ofrhombomeric Hoxb4 expression requires induction by somitesand a retinoid pathway. Neuron 21, 39–51.raham, A., Heyman, I., and Lumsden, A. (1993). Even-numberedrhombomeres control the apoptotic elimination of neural crestcells from odd-numbered rhombomeres in the chick hindbrain.Development 119, 233–245.rammatopoulos, G. A., Bell, E., Toole, L., Lumsden, A., andTucker, A. S. (2000). Homeotic transformation of branchial archidentity after Hoxa2 overexpression. Development 127, 5355–5365.riffin, K. J. P., Amacher, S. L., Kimmel, C. B., and Kimelman, D.(1998). Molecular identification of spadetail: regulation of ze-brafish trunk and tail mesoderm formation by T-box genes.Development 125, 3379–3388.uthrie, S. (1995). The status of the neural segment. TrendsNeurosci. 18, 74–79.atta, K., Puschel, A. W., and Kimmel, C. B. (1994). Midlinesignaling in the primordium of the zebrafish anterior centralnervous system. Proc. Natl. Acad. Sci. USA 91, 2061–2065.eisenberg, C.-P., and Nusslein-Volhard, C. (1997). The function ofsilberblick in the positioning of the eye anlage in the zebrafishembryo. Dev. Biol. 184, 85–95.eisenberg, C.-P., Tata, M., Rauch, G.-J., Saude, L., Concha, M. L.,Geisler, R., Stemple, D. L., Smith, J. C., and Wilson, S. W. (2000).Silberblick/Wnt11 mediates convergent extension movementsduring zebrafish gastrulation. Nature 405, 76–81.older, N., and Klein, R. (1999). Eph receptors and ephrins: Effectormolecules of morphogenesis. Development 126, 2033–2044.u, D., and Helms, J. A. (1999). The role of sonic hedgehog innormal and abnormal craniofacial morphogenesis. Development126, 4873–4884.uggins, J. P., and Pelton, J. T. (Eds.) (1997). “Endothelins inBiology and Medicine.” CRC Press, Boca Raton.ughes, M. K., and Hughes, A. L. (1993). Evolution of duplicategenes in a tetraploid animal, Xenopus laevis. Mol. Biol. Evol. 10,1360–1369.unt, P., Gulisano, M., Cook, M., Sham, M. H., Faiella, A.,Wilkinson, D., Boncinelli, E., and Krumlauf, R. (1991). A distinctHox code for the branchial region of the vertebrate head. Nature353, 861–864.

acobson, W., amd Fell, H. B. (1941). The developmental mechanicsand potencies of the undifferentiated mesenchyme of the man-dible. Q. J. Microsc. Sci. 82, 563–586.arlstrom, R. O., Talbot, W. S., and Shier, A. F. (1999). Comparative
synteny cloning of zebrafish you-too: Mutations in the Hedgehog

target gli2 affect ventral forebrain patterning. Genes Dev. 13,388–393.empf, H., Linares, C., Corvol, P., and Gasc, J.-M. (1998). Pharma-cological inactivation of the endothelin type A receptor in theearly chick embryo: a model of mispatterning of the branchialarch derivatives. Development 125, 4931–4941.immel, C. B., Miller, C. T., Kruse, G., Ullmann, B., BreMiller,R. A., Larison, K. D., and Snyder, H. C. (1998). The shaping ofpharyngeal cartilages during early development of the zebrafish.Dev. Biol. 203, 245–263.uhl, M., Sheldahl, L. C., Park, M., Miller, J. R., and Moon, R. T.(2000). The Wnt/Ca21 pathway. Trends Genet. 16, 279–283.urihara, Y., Kurihara, H., Suzuki, H., Kodama, T., Maemura, K.,Nagai, R., Oda, H., Kuwaki, T., Cao, W.-H., Kamada, N., Jishage,K., Ouchi, Y. U., Azuma, S., Toyoda, Y., Ishikawa, T., Kumada,M., and Yazaki, Y. (1994). Elevated blood pressure and craniofa-cial abnormalities in mice deficient in endothelin-1. Nature 368,703–710.

angille, R. M., and Hall, B. K. (1986). Evidence of cranial neuralcrest contribution to the skeleton of the sea lamprey, Petromy-zon marinus. In “Progress in Developmental Biology,” Part B, pp.263–266. A. R. Liss, New York.

angille, R. M., and Hall, B. K. (1988). Role of the neural crest indevelopment of the cartilaginous cranial and visceral skeleton ofthe medaka, Oryzias latipes (Teleostei). Anat. Embryol. 177,297–305.

Le Lievre, C. S. (1978). Participation of neural crest-derived cells inthe genesis of the skull in birds. J. Embryol. Exp. Morphol. 47,17–37.

umsden, A., and Krumlauf, R. (1996). Patterning the vertebrateneuraxis. Science 274, 1109–1115.ann, R., and Chan, S.-K. (1996). Extra specificity from extraden-ticle: the partnership between HOX and PBX/EXD homeodomainproteins. Trends Genet. 13, 772–774.anzanares, M., Cordes, S., Kwan, C. T., Sham, M. H,, Barsh, G. S.,and Krumlauf, R. (1997). Segmental regulation of Hoxb-3 bykreisler. Nature 387, 191–195.anzanares, M., Cordes, S., Ariza-McNaughton, L., Sadl, V., Ma-ruthainar, K., Barsh, G., and Krumlauf, R. (1999a). Conserved anddistinct roles of kreisler in regulation of the paralogous Hoxa3and Hoxb3 genes. Development 126, 759–769.anzanares, M., Trainor, P. A., Nonchev, S., Ariza-McNaughton,L., Brodie, J., Gould, A., Marshall, H., Morrison, A., Kwan, C. T.,Sham, M. H., Wilkinson, D. G., and Krumlauf, R. (1999b). Therole of kreisler in segmentation during hindbrain development.Dev. Biol. 211, 220–237.arshall, H., Morrison, A., Studer, M., Popperl, H., and Krumlauf,R. (1996). Retinoids and Hox genes. FASEB J. 10, 969–978.ellitzer, G., Xu, Q., and Wilkinson, D. G. (1999). Eph receptor andephrins restrict cell intermingling and communication. Nature400, 77–81.cKay, I. J., Muchamore, I., Krumlauf, R., Maden, M., Lumsden,A., and Lewis, J. (1994). The kreisler mouse: a hindbrain segmen-tation mutant that lacks two rhombomeres. Development 120,2199–2211.iller, C. T., Schilling, T. F., Lee, K.-H., Parker, J., and Kimmel,C. B. (2000). sucker encodes a zebrafish Endothelin-1 required forventral pharyngeal arch development. Development 127, 3815–3828.oens, C. B., Cordes, S. P., Giorgianni, M. W., Barsch, G. S., and
Kimmel, C. B. (1998). Equivalence in the genetic control of

N

N

P

P

P

P

P

P

P

P

Q

Q

R

R

R

R

R

S

S

S

S

S

S

S

S

S

S

S

S

T


hindbrain segmentation in fish and mouse. Development 125,381–391.

Niederreither, K., Vermot, J., Schuhbaur, B., Chambon, P., andDolle, P. (2000). Retinoic acid synthesis and hindbrain patterningin the mouse embryo. Development 127, 75–85.oden, D. M., Marcucio, R., Borycki, A.-G., and Emerson, C. H., Jr.(1999). Differentiation of avian craniofacial muscles. I. Patternsof early regulatory gene expression and myosin heavy chainsynthesis. Dev. Dyn. 216, 96–112.onchev, S., Maconochie, M., Gould, A., Morrison, A., and Krum-lauf, R. (1997). Cross-regulatory interactions between Hox genesand the control of segmental expression in the vertebrate centralnervous system. Cold Spring Harbor Symp. Quant. Biol. 62,313–323.

asqualetti, M., Ori, M., Nardi, I., and Rijli, F. M. (2000). EctopicHoxa2 induction after neural crest migration results in ho-meosis of jaw elements in Xenopus. Development 127, 5367–5378.

assner, J. M., Ryoo, H. D., Shen, L., Mann, R. S., and Aggarwal,A. K. (1999). Structure of a DNA-bound Ultrabithorax-Extradenticle homeodomain complex. Nature 397, 714 –719.

iper, D. E., Batchelor, A. H., Chang, C. P., Cleary, M. L., andWolberger, C. (1999). Structure of a HoxB1-Pbx1 heterodimerbound to DNA: Role of the hexapeptide and a fourth homeodo-main helix in complex formation. Cell 96, 587–597.

iotrowski, T., Schilling, T. F., Brand, M., Jiang, Y. J., Heisenberg,C. P., Beuchle, D., Grandel, H., Van Eeden, F. J. M., Furutani-Seiki, M., Granato, M., Haffter, P., Hammerschmidt, M., Kane,D. A., Kelsh, R. N., Mullins, M. C., Odenthal, J., Warga, R. M.,and Nuesslein-Volhard, C. (1996). Jaw and branchial arch mu-tants in zebrafish. II: anterior arches and cartilage differentiation.Development 123, 345–356.

ollack, D. M., and Highsmith, R. F. (1998). “Endothelin Receptorsand Signaling Mechanisms.” Springer-Verlag and R.G. Landes,Berlin.

ogoda, H., Solnica-Krezel, L., Driever, W., and Meyer, D. (2000).The zebrafish forkhead transcription factor FoxH1/Fast1 is amodulator of nodal signaling required for organizer formation.Curr. Biol. 10, 1041–1049.

opperl, H., Rikhof, H., Chang, H., Haffter, P., Kimmel, C. B.,and Moens, C. B. (2000). lazarus is a novel pbx gene thatglobally mediates hox gene function in zebrafish. Mol. Cell 6,255–267.

rince, V. E., Moens, C. B., Kimmel, C. B., and Ho, R. K. (1998).Zebrafish hox genes: Expression in the hindbrain region ofwild-type and mutants of the segmentation gene valentino.Development 125, 393–406.iu, M., Bulfone, A., Martinez, S., Meneses, J. J., Shimamura, K.,Pedersen, R. A., and Rubenstein, J. L. R. (1995). Null mutation ofDlx-2 results in abnormal morphogenesis of proximal first andsecond branchial arch derivatives and abnormal differentiation inthe forebrain. Genes Dev. 9, 2523–2538.iu, M., Bulfone, A., Ghattas, I., Meneses, J. J., Christensen, L.,Sharpe, P. T., Presley, R. Pedersen, R. A., and Rubenstein, J. L. R.(1997). Role of the Dlx homeobox genes in proximodistal pat-terning of the branchial arches: Mutations of Dlx-1, Dlx-2, andDlx-1 and -2 alter morphogenesis of proximal skeletal and softtissue structures dervived from the first and second arches. Dev.Biol. 185, 165–184.ebagliati, M. R., Toyama, R., Fricke, C., Haffter, P., and Dawid,
I. B. (1998). Cyclops encodes a nodal-related factor involved

in midline signaling. Proc. Natl. Acad. Sci. USA 95, 9932–9937.ijli, F. M., Mark, M., Lakkaraju, S., Dierich, A., Dolle, P., andChambon, P. (1993). A homeotic transformation is generated inthe rostral branchial region of the head by disruption of Hoxa-2,which acts as a selector gene. Cell 75, 1333–1349.ivera-Perez, J. A., Mallo, M., Gendron-Maguire, M., and Gridley,T. (1995). goosecoid is not an essential component of the mousegastrula organizer but is required for craniofacial and rib devel-opment. Development 121, 3005–3012.obinson, V., Smith, A., Flenniken, A. M., and Wilkinson, D. G.(1997). Roles of Eph receptors and ephrins in neural crest path-finding. Cell Tissue Res. 290, 265–274.ubinstein, A. L., Lee, D., Luo, R., Henion, P. D., and Halpern,M. E. (2000). Genes dependent on zebrafish cyclops functionidentified by AFLP differential gene expression screen. Genesis 1,86–97.

ampath, K., Rubinstein, A. L., Cheng, A. M. S., Liang, J. O.,Fekany, K., Solnica-Krezel, Korzh, V., Halpern, M. E., and Wright,C. V. E. (1998). Induction of the zebrafish ventral brain andfloorplate requires cyclops/nodal signalling. Nature 395, 185–189.

chauerete, H. E., van Eden, F. J., Fricke, C., Odenthal, J., Strahle,U., and Haffter, P. (1998). Sonic hedgehog is not required for theinduction of medial floor plate cells in the zebrafish. Develop-ment 125, 2983–2893.

chier, A. F., and Shen, M. M. (2000). Nodal signalling in vertebratedevelopment. Nature 403, 385–389.

chilling, T. (1997). Genetic analysis of craniofacial developmentin the vertebrate embryo. Bioessays 19, 459–468.

chilling, T. F., and Kimmel, C. B. (1994). Segment and cell typelineage restrictions during pharyngeal arch development in thezebrafish embryo. Development 120, 483–494.

chilling, T., and Kimmel, C. (1997). Musculoskeletal patterning inthe pharyngeal segments of the zebrafish embryo. Development124, 2945–2960.

chilling, T. F., Piotrowski, T., Grandel, H., Brand, M., Heisenberg,C. P., Jiang, Y. J., Beuchle, D., Hammerschmidt, M., Kane, D. A.,Mullins, M. C., Van Eeden, F. J. M., Kelsh, R. N., Furutani-Seiki,M., Granato, M., Haffter, P., Odenthal, J., Warga, R. M., Trowe,T., and Nuesslein-Volhard, C. (1996). Jaw and branchial archmutants in zebrafish. I: Branchial arches. Development 123,329–344.

echrist, J., Serbedzija, G. N., Scherson, T., Fraser, S. E., andBronner-Fraser, M. (1993). Segmental migration of the hindbrainneural crest does not arise from its segmental generation. Devel-opment 118, 691–703.

hen, M. M., and Shier, A. F. (2000). The EGF-CFC gene family invertebrate development. Trends Genet. 16, 303–309.

hubin, N. H., and Alberch, P. (1986). A morphogenetic approach tothe origin and basic organization of the tetrapod limb. Evol. Biol.20, 318–390.

irotkin, H. I., Gates, M. A., Kelly, P. D., Schier, A. F., and Talbot,W. S. (2000). fast1 is required for the development of dorsal axialstructures in zebrafish. Curr. Biol. 10, 1051–1054.

tuder, M., Gavalas, A., Marshall, H., Ariza-McNaughton, L.,Rijli, F. M., Chambon, P., and Krumlauf, R. (1998). Geneticinteractions between Hoxa1 and Hoxb1 reveal new roles inregulation of early hindbrain patterning. Development 125,1025–1036.homas, T., Kurihara, H., Yamagishi, H., Kurihara, Y., Yazaki, Y.,
Olson, E. N., and Srivastava, D. (1998). A signaling cascade

v

W

W

W

X

X

X

Y

Y

Y

Y

Z


involving endothelin-1, dHAND and Msx1 regulates develop-ment of neural-crest-derived branchial arch mesenchyme. Devel-opment 125, 3005–3014.

Thorogood, P. V. (1983). Morphogenesis of cartilage. In “Cartilage,Vol. 2. Development, Differentiation and Growth” (B. K. Hall,Ed.), pp. 223–254. Academic Press, New York.

an der Wees, J., Schilthuis, J. G., Koster, C. H., Diesveld-Schipper,H., Folkers, G. E., van der Saag, P. T., Dawson, M. I., Shudo, K.,van der Burg, B., and Durston, A. J. (1998). Inhibition of retinoicacid receptor-mediated signalling alters positional identity in thedeveloping hindbrain. Development 125, 545–556.einberg, E. S., Allende, M. L., Kelly, C. S., Abdelhamid, A.,Murakami, T., Andermann, P., Doerre, O. G., Grunwald, D. J.,and Riggleman, B. (1996). Developmental regulation of zebrafishMyoD in wild-type, no tail, and spadetail embryos. Development122, 271–280.eiss, P., and Amprino, R. (1940). The effect of mechanical stresson the differentiation of scleral cartilage in vitro and in theembryo. Growth 4, 245–258.hite, J .C., Highland, M., Kaiser, M., and Clagett-Dame, M.(2000). Vitamin A deficiency results in the dose-dependentacquisition of anterior character and shortening of the caudalhindbrain of the rat embryo. Dev. Biol. 220, 263–284.u, D., Emoto, N., Giaid, A., Slaughter, C., Kaw, S., deWit, D., andYanagisawa, M. (1994). ECE-1: A membrane-bound metallopro-tease that catalyzes the proteolytic activation of bigendothelin-1. Cell 78, 473–485.

u, Q., Alldus, G., Holder, N., and Wilkinson, D. (1995). Expressionof truncated Sek-1 receptor tyrosine kinase disrupts the segmen-
tal restriction of gene expression in the Xenopus and zebrafishhindbrain. Development 121, 4005–4016.

u, Q., Mellitzer, Robinseon, V., and Wilkinson, D. G. (1999). Invivo cell sorting in complementary segmental domains mediatedby Eph receptors and ephrins. Nature 399, 267–271.amada, G., Mansouri, A., Torres, M., Stuart, E. T., Blum, M.,Schultz, M., De Robertis, E. M., and Gruss, P. (1995). Tar-geted mutation of the murine goosecoid gene results incraniofacial defects and neonatal death. Development 121,2917–2922.amamoto, A., Amacher, S. L., Kim, S.-H., Geissert, D., Kim-mel, C. B., and De Robertis, E. M. (1998). Zebrafish paraxialprotocadherin is a downstream target of spadetail involved inmorphogenesis of gastrula mesoderm. Development 125,3389 –3397.anagisawa, M., Kurihara, H., Kimura, S., Tomobe, Y., Kobayashi,M., Mitsui, Y., Yazaki, Y., Goto, K., and Masaki, T. (1988). Anovel potent vasoconstrictor peptide produced by vascular endo-thelial cells. Nature 332, 411–415.anagisawa, H., Yanagisawa, M., Kapur, R. P., Richardson, J. A.,Williams, S. C., Clouthier, D. E., de Wit, D., Emoto, N., andHammer, R. E. (1998). Dual genetic pathways of endothelin-mediated intercellular signaling revealed by targeted disruptionof endothelin converting enzyme-1 gene. Development 125,825–836.hang, J., Talbot, W. S., and Sachier, A. F. (1998). Positionalcloning identifies zebrafish one-eyed pinhead as a permissiveEGF-related ligand required during gastrulation. Cell 92, 241–251.

Received for publication October 23, 2000
Accepted January 22, 2001
Published online April 16, 2001


specification and morphogenesis of the zebrafish larval head skeleton

Documents