Signalling interactions during facial development

Review article

Signalling interactions during facial development

Philippa Francis-West*, Raj Ladher, Amanda Barlow, Ann Graveson

Department of Craniofacial Development, UMDS, Guy’s Tower, Floor 28, London Bridge, London SE1 9RT, UK

Received 2 February 1998; revised version received 12 May 1998; accepted 12 May 1998

Abstract

The development of the vertebrate face is a dynamic multi-step process which starts with the formation of neural crest cells in thedeveloping brain and their subsequent migration to form, together with mesodermal cells, the facial primordia. Signalling interactions co-ordinate the outgrowth of the facial primordia from buds of undifferentiated mesenchyme into the intricate series of bones and cartilagestructures that, together with muscle and other tissues, form the adult face. Some of the molecules that are thought to be involved have beenidentified through the use of mouse mutants, data from human craniofacial syndromes and by expression studies of signalling moleculesduring facial development. However, the way that these molecules control the epithelial-mesenchymal interactions which mediate facialoutgrowth and morphogenesis is unclear. The role of neural crest cells in these processes has also not yet been well defined. In this reviewwe discuss the complex interaction of all these processes during face development and describe the candidate signalling molecules and theirpossible target genes. 1998 Elsevier Science Ireland Ltd. All rights reserved

Keywords:Facial primordia; Homeobox genes; Signalling molecules; Patterning; Neural crest; Wnt; Hh; Bone morphogenetic proteins;Fibroblast growth factors; Mouse mutants; Human craniofacial syndromes

1. Introduction

Whilst much attention has been paid to the embryonicorigin of the facial structures and pathways of neural crestmigration (Johnston, 1966; Le Lie`vre and Le Douarin,1975; Le Lievre, 1978; Noden, 1975, 1978, 1988; Coulyand Le Douarin, 1985, 1987, 1990; Lumsden et al., 1991;Couly et al., 1993a,b, 1995; Le Douarin et al., 1993; Scher-son et al., 1993; Ko¨ntges and Lumsden, 1996) very little isknown about the mechanisms which control outgrowth andpatterning of the facial primordia. Thus, regions within thefacial primordia responsible for co-ordinating the develop-ment of surrounding tissue have not been identified. Suchsignalling centres and interactions have been identified forother vertebrate structures such as the limb, somites and thenotochord/neural tube from results of experiments in whichectopically transplanted tissues modified the developmentof neighbouring tissues. The majority of these experimentshave utilized the developing chick embryo as a model sys-

tem, partly because it is accessible to manipulation in ovo.Signalling interactions that control patterning of the facialprimordia have received far less attention because the laterdeveloping face is less accessible in ovo.

Facial structures are derived from the facial primordia,which are initially a number of discrete buds surroundingthe primitive oral cavity, consisting of a neural crest- andmesodermally-derived mesenchymal core covered by anepithelial layer of ectoderm and endoderm (Fig. 1A). Ske-letal structures are derived from the neural crest, whilst themesoderm gives rise to myoblasts and angioblasts (Noden,1978, 1983 1988; Couly et al., 1995). The neural crest cellsarise within the peripheral neural ectoderm early in devel-opment and subsequently migrate into the presumptivefacial primordia. In the chick, neural crest cells from regionsanterior to rhomobomere 3 are the source of the developingfacial primordia (Johnston, 1966; Le Lie`vre and LeDouarin, 1975; Le Lie`vre, 1978; Noden, 1975, 1978; Lums-den et al., 1991; Scherson et al., 1993; Ko¨ntges and Lums-den, 1996).

The lower jaw is formed from the paired mandibular

Mechanisms of Development 75 (1998) 3–28

0925-4773/98/$19.00 1998 Elsevier Science Ireland Ltd. All rights reservedPII S0925-4773(98)00082-3

* Corresponding author. E-mail: [email protected]

https://www.researchgate.net/publication/17215888_A_Radio-autographic_Study_of_the_Migration_and_Fate_of_Cranial_Neural_Crest_Cells_in_Chick_Embryo?el=1_x_8&enrichId=rgreq-f7d239a5-d3e7-4e87-b94b-0d16c11c6644&enrichSource=Y292ZXJQYWdlOzEzNTQ2OTUzO0FTOjEwMzczNDY2MzE4ODQ5MUAxNDAxNzQzNjc5Nzc5

primordia whilst the upper jaw is formed from the fronto-nasal mass, lateral nasal process and maxillary primordia(Fig. 1A). In the upper jaw, analysis of the ratPax-6mutant

has given us some insight into which skeletal elements areformed from each primordia but a definitive study in a nor-mal embryo has not been done (see Fig. 1B and Osumi-

Fig. 1. (a) Diagram showing the primordia of the chick face at stage 24. f, Frontonasal mass; l, lateral nasal process; md, mandibular primordia; mx, maxillaryprimordia; o, primitive oral cavity. (b) Diagram showing the skeletal structure of the embryonic chicken head (after Lillie, 1952). Cartilaginous elements aremarked in blue whilst membrane bones are red. Proximal elements are on the left. The maxillary primordia give rise to the pterygoid, palatine, quadratojugaland jugal. Based on data from the ratSeymutant (Osumi-Yamashita et al., 1997), the maxillary primordia probably also gives rise to maxilla and thepremaxilla whilst the lateral nasal primordium gives rise to the lateral nasal wall, nasal and lachrymal bones. The frontonasal mass gives rise to thenasalseptum. The mandibular primordia give rise to the skeletal structures of the lower jaw. In the chick, distal facial structures are formed from mid-brain crestwhereas the more proximal elements such as the pterygoid and quadratojugal are derived from a mixture of midbrain and rhombomeric crest (Couly et al.,1996; Kontges and Lumsden, 1996).

a

4 P. Francis-West et al. / Mechanisms of Development 75 (1998) 3–28

Yamashita et al., 1997). The skeletal structures from each ofthe facial primordia are connected with those formed fromneighbouring primordia. For example, the maxillary primor-dium forms several bones in a ‘proximo-distal’ sequence.The most proximal structures articulate with the Meckels’cartilage derived from the mandibular primordia, whilst thedistal structures are joined to the skeletal structures derivedfrom the frontonasal mass (Fig. 1B). Thus, growth and mor-phogenesis of the primordia must be tightly co-ordinated toobtain a normal and functional facial skeleton. Abnormaldevelopment of the facial primordia can result in manyhuman syndromes such as Treacher–Collins and Alagillessyndrome, but more commonly cleft lip and palate, the latterof which affect one in 750 of live births (reviewed by Win-ter, 1996; see also Johnston and Bronsky, 1995; Thorogood,1998).

Both ectodermal and mesenchymal factors control pat-terning and growth during many developmental processes.For example, during limb development,Wnt-7a andEngrailed-1(En-1), which are expressed in the dorsal andventral ectoderm, respectively, control patterning along thed-v axis, whilst patterning along the a-p axis can be deter-mined by mesenchymal expression ofShh (reviewed byJohnson and Tabin, 1997). Outgrowth of the developinglimb bud is maintained by the expression of members ofthe FGF family (Fgf-2, -4 and -8) in the ectoderm at thedistal tip. Signalling interactions along all three axes arealso co-ordinated; Wnt-7a and Fgfs can maintain the expres-sion ofShhin the mesenchyme, whilst Shh can maintain theexpression ofFgf-4 in the AER (reviewed by Johnson andTabin, 1997). Although some of the molecules involved inthese patterning events have been identified, these signallingregions were first elucidated by classical embryological stu-dies in which pieces of limb mesenchyme were graftedectopically into different regions of the developing limbbud and in experiments in which the limb ectoderm wasremoved or rotated (reviewed by Tickle and Eichele,1994). Some of the equivalent experiments have beendone with the facial primordia and these have shown thatthe outgrowth of the facial primordia requires epithelial-mesenchymal interactions (Wedden, 1987). However, todate, such experiments have not implicated any specificregion(s) of mesenchyme or ectoderm which control pat-terning of the developing facial primordia. As in the devel-oping limb, the facial primordia have three axes andsignalling information must be co-ordinated to controlgrowth and patterning. Thus, it would seem likely thatthere must be specific control regions of the developingfacial primordia.

Our understanding of facial development is also hinderedas it is a multi-step process involving the formation andmigration of neural crest followed by the correct outgrowthof the facial primordia and the ultimate morphogenesis anddifferentiation of the skeletal elements. Many factorsinvolved in facial development are expressed during a num-ber, if not all, of these processes. Therefore, when facial

abnormalities occur due to mutation of a gene, it is oftennot known which stages of facial development have beenaffected. Furthermore, as brain and facial development areinterlinked, either directly through inductive interactions, ordue to secondary effects of the changing shape of the devel-oping brain and skull, it is not always apparent if facialabnormalities due to mutation of a gene are a direct effecton facial development, or are secondary due to abnormalbrain development (reviewed by Opitz and Gilbert, 1982;see also Couly and Le Douarin, 1985, 1987).

Much previous work has focused on the roles of retinoicacid during facial development. Retinoic acid plays a keyrole in the developing facial primordia, as shown by knock-out studies of retinoic acid receptors in mice (Lohnes et al.,1994). Retinoic acid can also be teratogenic in excess andcan modulate the expression of genes during facial devel-opment (reviewed by Morriss-Kay, 1993; Morriss-Kay andSokolova, 1996; see also Lee et al., 1995; Brown et al.,1997; Helms et al., 1997; Iulianella and Lohnes, 1997).Although important, since the function and effect of retinoicacid during facial development has been well reviewed else-where (Morriss-Kay, 1993; Morriss-Kay and Sokolova,1996; Brickell and Thorogood, 1997), in this review wewill focus on the current knowledge of signalling factorsand homeobox-containing genes that are expressed in thedeveloping facial primordia. We first present the currentknowledge of these factors during facial development andthen present current thinking and knowledge within the fieldof tissue interactions during facial development. Finally, wepresent areas that are still not understood during facialdevelopment which are important areas for future research.

2. Signalling factors during facial development

Secreted factors that appear to be involved in facial devel-opment include members of the fibroblast growth factor(FGF), transforming growth factorb (TGFb), hedgehog(hh) and wingless (Wnt) gene families. Signalling mole-cules such as endothelin 1 (ET-1), platelet-derived growthfactor A (PDGFa), epidermal growth factor (EGF), TGFa,stem cell factor, and membrane bound ligands such asjagged 1 and 2 (serrate 1 and 2) are also involved in facialdevelopment (Table 1; Fig. 2; see below for references andLee et al., 1985; Pratt, 1987; Wilcox and Derynck, 1988;Coffin-Collins and Hall, 1989; Matsui et al., 1990; Orr-Urtreger et al., 1990; Kronmiller et al., 1991; Morrison-Graham et al., 1992; Orr-Urtreger and Lonai, 1992; Shumet al., 1993).

Some of these factors are expressed at several stages offace development and it is not always clear which stages ofdevelopment are affected by mutation of a gene. A summaryof the expression of these genes is presented in Table 1,while the factors that have been shown to be involved inspecific stages of face development are summarized in Fig.2.

5P. Francis-West et al. / Mechanisms of Development 75 (1998) 3–28

https://www.researchgate.net/publication/13903401_Molecular_Models_for_Vertebrate_Limb_Development?el=1_x_8&enrichId=rgreq-f7d239a5-d3e7-4e87-b94b-0d16c11c6644&enrichSource=Y292ZXJQYWdlOzEzNTQ2OTUzO0FTOjEwMzczNDY2MzE4ODQ5MUAxNDAxNzQzNjc5Nzc5

https://www.researchgate.net/publication/19508940_Epithelial-mesenchymal_interactions_in_the_development_of_chick_facial_primordia_and_the_target_of_reinoid_action?el=1_x_8&enrichId=rgreq-f7d239a5-d3e7-4e87-b94b-0d16c11c6644&enrichSource=Y292ZXJQYWdlOzEzNTQ2OTUzO0FTOjEwMzczNDY2MzE4ODQ5MUAxNDAxNzQzNjc5Nzc5

2.1. Fibroblast growth factor family

The FGF family consists of at least 18 members. They areproliferative factors and in vitro studies have suggested thatthey may also have roles controlling cell differentiation,survival and motility (reviewed by Basilico and Moscatelli,1992). At least six members of the FGF family,Fgf-1, -2, -4,-5, -8 and-12are expressed in the developing facial primor-dia. Of these,Fgf-2, -4 and -8 are expressed in theepithelium.Fgf-2 is expressed throughout the epitheliumof the chick facial primordia, whileFgf-4 and -8 areexpressed in restricted regions (Drucker and Goldfarb,1993; Crossley and Martin, 1995; Wall and Hogan, 1995;Vogel et al., 1996; Barlow and Francis-West, 1997; Helmset al., 1997; Richman et al., 1997). FGFs can partially sub-stitute for epithelial signals in their ability to support out-growth of the facial primordia and FGF-2 can increaseproliferation in micromass cultures of the frontonasalmass primordia (Richman and Crosby, 1990; Richman etal., 1997). This suggests that members of the FGF familymay control outgrowth of the developing facial primordia,analogous to the role of FGFs during development of thelimb bud. Furthermore, transgenic expression of the extra-cellular ligand-binding domain of FGFR2b, which acts as adominant-negative receptor, results in the development ofcleft palate and a smaller face and skull (Celli et al., 1998).FGFR2b normally binds FGF2, -3, -7 and -10, suggestingthat some of these ligands are involved in facial develop-ment.

Expression of other members of the FGF family has notbeen fully characterized. Expression ofFgf-1 and -5have

been detected by Northern analysis of E13.5 developingmouse faces (He´bert et al., 1990). In situ analysis at thisstage has shown thatFgf-5 is expressed in regions of muscleformation in the developing face, suggesting a role in myo-genesis, as has been proposed for FGF-5 in other regions ofthe developing embryo (He´bert et al., 1990; Haub and Gold-farb, 1991).Fgf-12 is expressed in the mouse craniofacialmesenchyme between E10.5–14.5 but a detailed expressionpattern for this gene has not yet been reported (Hartung etal., 1997).

The fibroblast growth factor receptors,Fgfr-1, -2 and-3are expressed in the mesenchyme of the facial primordiaand are later associated with some regions of chondrogen-esis (Wanaka et al., 1991; Peters et al., 1992; Yamaguchi etal., 1992; Patstone et al., 1993; Matovinovic and Richman,1997; Wilke et al., 1997). In humans, there are several cra-niofacial syndromes that are due to mutation of the FGFreceptors (reviewed by Muenke and Schell, 1995; Yamagu-chi and Rossant, 1995; Wilkie et al., 1995; Wilkie, 1997).For example, Crouzon’s syndrome and achondroplasia, bothin part characterized by a smaller midface, are due to con-stitutive activation of FGFR2 and FGFR3, respectively(Jabs et al., 1994; Reardon et al., 1994; Meyers et al.,1995). Apert’s syndrome, caused by a missense mutationin the FGFR2 gene, is characterized by craniosynostosis,the premature fusing of the skull and the formation of hypo-plastic zygomatic arch, maxillary and nasal bones (Cohenand Kreiborg, 1996). While the facial defects in these syn-dromes are in part a secondary consequence to changes inskull development, they may also be due to direct changesin facial outgrowth and facial skeletal development.

Table 1

Signalling factors and transcription factors expressed in the developing face

Signalling factors Homeobox containing genes Other transcription factors

Migrating neural crest: Wnt-5a En protein AP-2a

Dlx-1a, 2a Gli-2a and 3a

Msx-1a,b and 2b twista,b

Otx-2a

Pax-3a,b and 7a

Facial primordia: Activinbaa Barx-1 and 2 AP-2a

Bmp-2, 4a, 5a and 7a Dlx-1–6 (1a2a) Gli-1, 2aand 3a

EGF En-2 Pax-1a

ET-1a GH-6 twista,b

FGF1, 2,4, 5, 8 and 12 Gsc-1a

PDGF-a Mhoxa (Prx-1)Jagged 1b and 2a Msx-1a,b and 2b

Shha,b Otx-2a

TGFa Pax-3a,b, 7a, 6a,b

TGFb1, 2a and 3a Ptx-1Wnt-5a, 10a,10b, 11 Reigb

S8Tlx-1 (Hox11)Uncx-4.1

The genes listed under the ‘other transcription factors’ only include those in which mutation or knockout of function results in facial abnormalities.aMutation or loss of function by knockout studies in mice results in some facial abnormalities.bThere is a human syndrome associated with mutation of this gene.


2.2. TGFb superfamily

The TGFb signalling family is a multi-gene family whichincludes the TGFb-1, -2 and -3 molecules themselves, andmore distantly related members such as the bone morpho-genetic proteins (BMPs), growth and differentiation factors(GDFs) and activins/inhibin family. Many members of thisfamily and their receptors are expressed in the developingfacial primordia (see references below and Dewulf et al.,1995; Roelen et al., 1997). During development, membersof this family have been shown to be involved in patterning,epithelial-mesenchymal interactions, cell proliferation,apoptosis and chondrogenesis (reviews by Kingsley, 1994;Hogan, 1996; Moses and Serra, 1996). They are secretedmolecules that act as dimers. Although there is some func-

tional evidence to suggest that they act as heterodimers invivo (Thomas et al., 1997), it is not yet clear if they alwaysact as heterodimers and which combinations of heterodi-mers they form.Tgfb-1, -2 and -3 are all expressed in theearly facial mesenchyme, with later expression being asso-ciated with some regions of skeletal development (Pelton etal., 1989; Pelton et al., 1991; Millan et al., 1991). TheTgfb-1 knockout dies before 10.5 days of embryonic develop-ment, so its role in facial development cannot be evaluatedby this approach (Dickson et al., 1995). In theTgfb-2 knock-out, there are defects in maxillary and mandibular develop-ment, occasionally resulting in cleft palate (Sanford et al.,1997) while in theTgfb-3 knockout, there is always cleftpalate. However, this is not due to abnormal development ofthe facial primordia or to the failure of the palatal shelves toform, as they do meet but fail to fuse (Kaartinen et al., 1995;Proetzel et al., 1995). Thus, TGFb3 is specifically requiredfor the fusion of the palatal shelves, which is consistent withits high level of expression at the palatal shelf interface.Recent work, has shown that TGFb3 promotes epithelial-mesenchymal transformations of the touching epithelialseams in the palatal shelves (Kaartinen et al., 1997; Sun etal., 1998). Indeed, even the palatal shelves of developingchick embryos, which do not meet in vivo and undergopartial transformation when placed together in vitro, canbe induced to fuse in the presence of TGFb3, but notTGFb-1 or -2 (Sun et al., 1998). AlthoughTgfb-3 isexpressed in many other regions of the developing face,compensation by other members of the TGFb family mayaccount for the normality of the remainder of the knockoutface.

The expression of a few members of the BMP subdivisionof the TGFb family have been characterized in detail (Fran-cis-West et al., 1994; Bennett et al., 1995; Wall and Hogan,1995; Barlow and Francis-West, 1997). The function ofthree members of the Bmp (-2, -4 and -7) family has beenknocked out in mice and there is a naturally occurring muta-tion in another member of the family, Bmp-5, resulting inthe short-ear mutant (Kingsley et al., 1992). The homozy-gous knockouts ofBmp-2 and Bmp-4 are not useful forfacial studies as they die well before the formation of thefacial skeletal structures (Winnier et al., 1995; Zhang andBradley, 1996). However, the chick model has given insightinto the roles of BMP-2 and BMP-4 during facial develop-ment. In the early developing chick face,Bmp-4 isexpressed in restricted domains in the epithelium in associa-tion with underlying regions ofBmp-2, Bmp-7, Msx-1andMsx-2expression in the mesenchyme (Fig. 3A,B; Francis-West et al., 1994; Wall and Hogan, 1995; Barlow and Fran-cis-West, 1997 and our unpublished data). Ectopic applica-tion of BMP-2 or BMP-4 can activate the expression ofMsx-1andMsx-2, which can also result in the bifurcationof skeletal structures (Fig. 3C,E; Barlow and Francis-West,1997). Furthermore, ectopic application of BMP-2 canincrease proliferation in the mandibular primordia (Barlowand Francis-West, 1997), whilst haplo-insufficiency of

Fig. 2. Flow chart showing the processes involved during facial develop-ment. We have classed these processes as early patterning events thatspecify head mesoderm or control neural tube formation, control neuralcrest migration or regulate patterning and outgrowth in the facial primor-dia. Also listed are examples of genes that appear to be involved in theseprocesses and their possible targets of action (see text for references). Theassignation of genes to a particular process does not necessarily imply thatit does not act at other stages of development. One example isOtx-2whichis also expressed in the migrating neural crest, so although it has a role inearly development, this does not discount it from acting during later devel-opment. It should be emphasized that these genes do not affect all cranio-facial structures, and that their effects are regionalized. One example isPax-6, which controls development of anterior structures of the face.Another isET-1, which controls the expression ofGoosecoidonly in thefirst and second arches. A list of other genes expressed during facialdevelopment is presented in Table 1.


Bmp-4in C57Bl/6 mice results in the formation of shorterfrontal and nasal bones in 12% of heterozygous mice (Dunnet al., 1997). Together, these data suggest that BMP-2 andBMP-4 may control outgrowth of the facial primordia.

Knockout ofBmp-7 in mice affects the development ofsome cranial bones and can result in the formation of amuch smaller pterygoid bone (Dudley et al., 1995; Luo etal., 1995). The lack of effect on most of the facial skeletal

structures shows that Bmp-7 alone is not crucial for facialdevelopment and suggests that there is a functional redun-dancy between the BMPs. In the developing chick face,expression ofBmp-7 co-localizes withBmp-2 expression(our unpublished data) as occurs in several regions of thedeveloping embryo (Lyons et al., 1995) and it is possiblethat expression of Bmp-2 compensates for the lack of Bmp-7.

Fig. 3. Whole mount in situ hybridisation showing the expression of (A)Bmp-4(B) Msx-1at stages 25 and 26, respectively, in the developing chick face. (C)and (D) show the expression ofMsx-1andShh, 48 and 72 h, respectively, after a bead soaked in BMP-2 has been placed in the posterior region of thedeveloping maxillary primordia at stage 20. This shows that BMP-2 can activate the expression ofMsx-1 in chick developing facial primordia and that theextension ofMsx-1expression (arrowed) is associated with the extension ofShhexpression in the overlying epithelium (arrowed). In (C) the bead is markedby a white arrowhead. In (C) and (D), the normal domains ofMsx-1andShhexpression are shown on the right hand side of the embryo. These manipulationscan result in the bifurcation or apparent duplication of skeletal structures. (E) shows the effect of placing a BMP-2 bead at the posterior edge of the maxillaryprimordia at stage 19. This would result in the extension ofMsx-1andShhexpression. On the treated side of the face, there is an additional skeletal structurewhich resembles the palatine because of its shape and contact with the pterygoid. For further details see Barlow and Francis-West, 1997. md, Mandibularprimordia; mx, maxillary primordia; p, palatine: pe, ‘ectopic’ palatine; pt, pterygoid.


TheBmp-5mutant is characterized by shortened ears anda smaller skeletal structure (Green, 1951; Kingsley et al.,1992).Bmp-5is expressed in the developing cartilage ele-ments and it is thought that the defects seen in the mutant aredue to abnormal skeletal growth (Kingsley et al., 1992; Kinget al., 1994). There is also a human syndrome, characterizedby a similar phenotype such as ear defects and short stature,which has been proposed to be caused by mutation in theBMP5gene (Lacombe et al., 1994).

Activin bA, another member of the TGFb family, isexpressed in the mesenchyme of developing mouse andrat faces (Feijen et al., 1994; Roberts and Barth, 1994).Knockout of its function in mice results in cleft palate andloss of lower incisors (Matzuk et al., 1995a); the lack ofskeletal abnormalities suggests that the cleft palate is dueto a defect in growth of the facial primordia (Matzuk et al.,1995a). A null mutation offollistatin results in hard palateabnormalities similar to those seen inactivin bA knockoutmice, suggesting that follistatin acts as an activin agonist invivo, despite it acting as an activin antagonist in vitro (Mat-zuk et al., 1995b). Follistatin can also act as a BMP-2/BMP-4 antagonist in vitro and thefollistatin knockout has addi-tional defects to those seen in theactivin bA knockout insome regions of the body. This suggests that follistatin canmodulate the activity of several members of the TGFb

family. Knockout of the activin type 2 receptor,ActRcII,also causes craniofacial defects, including hypoplasia of themandible and cleft palate in a low percentage of animals(Matzuk et al., 1995c).

2.3. Hedgehog family

Three members of the Hedgehog family have been iden-tified in humans, mice and chicks: Sonic hedgehog (Shh),Indian hedgehog (Ihh) and Desert hedgehog (Dhh)(reviewed by Hammerschmidt et al., 1997). Shh signallinghas been implicated in the patterning of several vertebratestructures including the limb bud, neural tube and somites,whereas Ihh is thought to control chondrocyte differentia-tion in developing cartilage elements. In developing mouseand chick faces,Shh is expressed in the epithelia of someof the facial primordia (Fig. 3D; Martı´ et al., 1995; Walland Hogan, 1995; Barlow and Francis-West, 1997; Helmset al., 1997). Knockout ofShhfunction in mice and muta-tion of the SHH gene in humans result in holoprosence-phaly, characterized by abnormal forebrain and facialdevelopment (Belloni et al., 1996; Chiang et al., 1996;Roessler et al., 1996). In the knockout mice, the skeletalstructures of the face are almost completely absent,although the branchial arches do appear to be relativelynormal at E9.5. As any perturbation in the developmentof the brain will have a knock-on effect on the face, it isnot yet clear if Shh has a direct role in facial patterning.However, functional studies have shown that the facialepithelia which expressShhcan induce digit duplicationswhen grafted into the anterior region of a developing limb

bud, showing that this facial epithelium does have pattern-ing ability (Helms et al., 1997).

Mutations in members of the hh signal transduction path-way can also result in facial abnormalities. InDrosophila,hh signals through the transmembrane receptor patched(ptc) and the transcription factor, cubitus interruptus (ci)protein (Alexandre et al., 1996; Domıńguez et al., 1996).A similar signalling cascade exists in vertebrates (seereview by Ingham, 1998). In humans, haplo-insufficiencyof PTC results in Gorlin’s syndrome which is characterizedby cleft palate and odontogenic cysts in the mandible con-sistent with the expression of thePtchomologues in the firstbranchial arch in mice and chicks (Goodrich et al., 1996;Hahn et al., 1996a,b; Johnson et al., 1996; Unden et al.,1996; Platt et al., 1997).

In vertebrates, there are at least three homologues of ci,known as Gli, Gli-2 and Gli-3 (Ruppert et al., 1988; Hui etal., 1994). In the chick limb bud andXenopusneural tube,Gli expression is induced by Shh and Gli is thought tomediate the downstream effects of Shh signalling, as inDrosophila (Marigo et al., 1996; Lee et al., 1997). In con-trast, in the developing chick limb bud,Gli-3 is downregu-lated following Shh mis-expression and it is has beenproposed that Gli-3 is a repressor ofShh expression(Masuya et al., 1995; Marigo et al., 1996; see review byRuiz i Altaba, 1997). In the embryonic mouse face,Gli-2and-3 are expressed in migrating neural crest and all threeGlis are expressed in early facial mesenchyme, later becom-ing associated with skeletogenesis (Walterhouse et al.,1993; Hui et al., 1994). If and how these genes mediate/affect hedgehog signalling in the developing face isunknown. Knockout ofGli-2 and mutation ofGli-3 inmice results in facial abnormalities (Mo et al., 1997).These phenotypes are distinct, with knockout ofGli-2resulting in truncation of the distal part of the maxilla andmandible, the absence of the incisors and clefting due to theloss of the presphenoid bone, maxillary and palatine shelves(Mo et al., 1997). In contrast, in theGli-3 mutant, the max-illary region is enlarged, the external nasal processes aresmaller and some clefting occurs. Double knockout ofboth Gli-2 and Gli-3 in mice has a more severe effect onmandibular development, with several additional bonesbeing hypoplastic, suggesting that in some regions of thedeveloping face, there is functional redundancy betweenGli-2 and Gli-3 (Mo et al., 1997).

In humans, partial loss ofGLI3 function due to deletion ortranslocation of one copy ofGLI3 results in Grieg’s syn-drome, which is usually characterized by a wide foreheadand nasal bridge (Vortkamp et al., 1991). The clinicallydistinct Pallister-Hall syndrome is associated with amutation inGLI3 which truncates the protein (Kang et al.,1997).

2.4. Wnt family

The Wnt gene family consists of at least 15 members.


They have been shown to have a number of roles duringdevelopment, including patterning the central nervous sys-tem and the dorso-ventral axis of the limb bud (reviews byMcMahon, 1992; Johnson and Tabin, 1997) and modulatingchondrocyte differentiation in chick limbs and faces (Rud-nicki and Brown, 1997). At a cellular level, some membersof the family have been shown to modulate cell-adhesion,proliferation and gap-junctional communication (Tsuka-moto et al., 1988; Olson et al., 1991; Dickinson et al.,1994; Hinck et al., 1994; Bradley and Brown, 1995; seealso review by Moon et al., 1993).Wnt-5a, -10a, -10b and-11 are expressed in the developing facial primordia ofmice, whereas expression ofWnt-1 and -2 has not beendetected in similar studies (Gavin et al., 1990; Parr et al.,1993; Christiansen et al., 1995; Hollyday et al., 1995; Wangand Shackleford, 1996).Wnt-5aand-11have been shown tobe expressed in the mesenchyme of the facial primordia byin situ hybridization (Parr et al., 1993; Christiansen et al.,1995; Hollyday et al., 1995).Wnt -10aand-10bhave beendetected at 12.5 days of embryonic development by North-ern analysis (Wang and Shackleford, 1996) but localizationof their expression has not been determined. The roles of theWnt genes during facial development are not yet known.However, knockout ofLef-1, a gene involved in Wnt signaltransduction and which is expressed in the migrating neuralcrest and later in the facial primordia, results in the loss ofteeth and the mesencephalic nucleus of the trigeminal nerve,although there are no detectable abnormalities in the skele-tal structures of the face (Oosterwegel et al., 1993; vanGenderen et al., 1994).

2.5. Jagged

Jagged 1 and 2 (also known as serrate 1 and 2), membraneligands for the notch receptor family, are known to have arole in facial development. In mice,jagged 1is expressed indeveloping maxillary and mandibular primordia (Mitsiadiset al., 1997). In humans, haplo-insufficiency ofJAGGED1causes Alagilles syndrome characterized by a triangularshaped face resulting from defects in mandible, foreheadand nose development (Alagille et al., 1975, 1987; Li etal., 1997; Oda et al., 1997). Recently, truncated forms ofserrate have been shown to act in a dominant-negative fash-ion duringDrosophilawing development (Hukriede et al.,1997). Although, some cases of Alagilles syndrome are dueto truncation of the JAGGED1 protein which may act asdominant-negatives during facial development, it is mostlikely that Alagilles syndrome is due to haplo-insufficiencyas loss of theJAG1gene has the same phenotype as expres-sion of the truncated forms.

Jagged 2(serrate 2) is expressed in the epithelium of the1st branchial arch in day 9 mouse embryos, with expressionbeing detectable in other branchial arches by day 10 (Sidowet al., 1997; Valsecchi et al., 1997; Jiang et al., 1998). Inaddition, expression ofjagged 2has been detected aroundthe developing nasal pits. Mutation ofjagged 2, which

causes a single amino-acid change in the protein, is thoughtto be responsible for the mousesyndactylism(sm) mutant(Sidow et al., 1997). This mutant is characterized by limbdefects, but has no apparent facial abnormalities. However,complete knockout ofjagged 2function does result in facialdefects, predominantly characterized by a failure of thepalatal shelves to elevate which subsequently fuse withthe developing tongue (Jiang et al., 1998). In addition, inthe jagged 2knockout there are limb defects which aresimilar to but more severe than those in thesm mutant.This suggests that thesmmutation acts as a hypomorphicmutant allele and that the developing limb is more sensitiveto changes in jagged 2 expression.

2.6. Endothelins

Endothelin 1 (ET-1) is a small polypeptide of 21 aminoacids which is proteolytically activated by endothelin-con-verting enzyme (ECE-1). Both are expressed in theepithelium and paraxial mesoderm of developing firstand second branchial arches, whilst one of the receptorsfor ET-1, EtA, is expressed in a complementary pattern,i.e. in neural crest derived mesenchyme (Kurihara et al.,1994; Barni et al., 1995; Maemura et al., 1996; Clouthieret al., 1998). In addition, at earlier stages,EtA expressionhas been detected in migrating neural crest with expres-sion of Ece-1 in the surrounding mesenchyme (Clouthieret al., 1998; Yanagisawa et al., 1998). The expression ofthe ligands of EtA (Et-1 and Et-2) during these earlystages has not been reported. TheEt-1, Ece-1and EtAknockout mice have very similar phenotypes with theskeletal structures derived from the first and second archesbeing smaller and there is cleft palate (Kurihara et al.,1994; Clouthier et al., 1998; Yanagisawa et al., 1998).Analysis of theEt-1 mutant arches shows that they aresmaller at 10.5 days of embryonic development, suggest-ing that these abnormalities are, in part, due to an earlydefect in facial development.EtA is expressed in themigrating neural crest cells and ifEt-1 is also expressedat this stage, this defect may arise during neural crestmigration. Alternatively, this early abnormality could bedue to changes in epithelial-mesenchymal interactions.Since ET-1 is mitogenic for differentiated melanocytes(Swope et al., 1995) and since another member of theendothelin family, ET-3, has been shown to increase pro-liferation of trunk neural crest, the defects in facial devel-opment in theEt-1 knockout mice are likely due to adecrease in proliferation of the cranial neural crest(Lahav et al., 1996; Stone et al., 1997). This proposedmitogenic effect may be direct or may be by modulationof the expression/activity of other mitogenic factors. Forexample, in rat atrial cardiocytes, ET-1 increases thesecretion of the endopeptidase 24.11, a zinc-containingmetalloendopeptidase, which is expressed in craniofacialmesenchyme and modulates the activity of various signal-ling molecules either through proteolytic activation and/or


degradation (Fukuda et al., 1988; Spencer-Dene et al.,1994). The defects seen in theEt-1, Ece-1andEtA knock-out mice are similar to human syndromes collectivelyknown as CATCH 22, which are characterized by facedefects in addition to cardiac and other defects. It hasbeen proposed that mutation in a component of the ET-1 signalling pathway may be responsible for some of theCATCH 22 syndromes (Clouthier et al., 1998).

2.7. PDGFs

PDGFs also have a role in facial development.Pdfgareceptor (Pdgfra) is expressed in the migrating neuralcrest and facial mesenchyme, whilst its ligand,Pdgfa, isexpressed in the epithelium of the branchial arches (Morri-son-Graham et al., 1992; Orr-Urtreger and Lonai, 1992;Schatteman et al., 1992). Knockout ofPdgfra results inmidline defects and the under-development of the facewith the loss of some bones, whilst neuronal derivativesof the neural crest are unaffected (Soriano, 1997). At E10,there is increased apoptosis in the head and branchial archesand it has been proposed that PDGFa promotes survival of asubset of neural crest which give rise to non-neuronal deri-vatives (Soriano, 1997).

To understand the roles of these signalling moleculesduring facial development, functional studies need to beperformed to determine how members of these genefamilies control outgrowth and patterning and how thesedifferent signalling pathways interact. For example, dothese different signalling factors/pathways control differentaspects of facial patterning, as in the developing limbbud?

3. Target genes

Knockout studies have shown that homeobox-containinggenes control the development and patterning of manystructures and thus, homeobox-containing genes which areexpressed in the developing face are strong candidates tocontrol facial patterning. Genes of theHox cluster, whichhave been shown to mediate patterning in the axial andappendicular skeleton (reviewed by Krumlauf, 1994a), arenot expressed during development of the facial primordia.However,Hoxa-2is expressed in the neural crest derivativesthat give rise to the second arch (Hunt et al., 1991b,c; Princeand Lumsden, 1994; Couly et al., 1996; reviewed by Hunt etal., 1991a; Krumlauf, 1994b) and knockout ofHoxa-2results in the transformation of second arch structures toform proximal first arch derivatives, showing that thisHoxgene is involved in patterning the second branchial arch(Gendron-Maguire et al., 1993; Rijli et al., 1993). Recentstudies have shown that the expression of theHox clustergenes are the same during craniofacial development inhumans, mice and chick, confirming the validity of usingmouse and chicks as models for understanding the role of

HOX genes during human facial development (Vieille-Grosjean et al., 1997).

Although the classical Hox genes are not expressed in thedeveloping facial primordia, several other homeobox-con-taining genes are expressed and these are therefore candi-dates for patterning the facial primordia. These genesincludeMsx-1and2, MHox, Dlx-1–6, Barx-1 and -2, Otx-2, GH6; Bicoid-homologues such asGsc-1, Ptx-1andReig(also calledBrx-1 andOtlx-2); the paired-box homologues,Pax-3, -6, -7, Cart-1and the paired-related genes such asS8andUncx-4.1(Table 1 and see below for references. Alsosee Opstelten et al., 1991; Stadler and Solursh, 1994; Tis-sier-Seta et al., 1995; Jones et al., 1997; Mansouri et al.,1997; Wu et al., 1997). Many of these homeobox-containinggenes are expressed in the neural-crest derived mesenchymeof the developing facial primordia, marking specific regionsof the developing face. As the neural crest cells are classi-cally thought to pattern the developing face (but see Section5), much attention has focused on these genes as they maycontrol patterning of the developing facial primordia.Knockout of several of these genes results in distinct defectsin facial development, supporting this idea (see below;Table 2 and Fig. 4). As different skeletal structures areaffected by the various knock-outs, the combinatorial pat-tern of gene expression may pattern the developing face, ashas been suggested for Dlx genes during development of thefirst branchial arch (Fig. 4; Qiu et al., 1995, 1997). Of par-ticular interest areDlx-2, MhoxandOtx-2, whose knockoutresults in atavistic skeletal changes in the shape of bonesand/or the formation of additional bones, such that theyresemble those of more primitive ancestors (see referencesbelow). In addition to providing functional evidence thatthese genes play a role in patterning the facial primordia,this also suggests that either the evolution of these genes ortheir differential expression in different species has resultedin changes in the patterning of craniofacial structures(reviewed by Kuratani et al., 1997).

In addition, to homeobox-containing genes expressed inthe ectomesenchyme, others such asTlx-1 andEn-2expres-sion are thought to mark presumptive myoblasts in the firstarch (Gardner et al., 1988; Davis et al., 1991; Logan et al.,1993; Raju et al., 1993; Roberts et al., 1994; Dear et al.,1995). As the neural crest-derived mesenchyme patterns thedeveloping musculature, these genes are unlikely to have adirect role in facial patterning. This idea is supported byknockout of eitherTlx-1 or En-2 in mice, which do nothave any detectable consequences on face, including facialmuscle, development (Millen et al., 1994; Roberts et al.,1994).

Expression of these homeobox-containing genes,together with those transcription factors that are known tobe involved in facial development, is presented in Table 1.As with the signalling factors, it is not always clear whichstages of development are affected by mutations. However,genes known to be involved in specific aspects of develop-ment are shown in Fig. 2.


3.1. Homeobox-containing genes

3.1.1. Msx genesMsx-1andMsx-2are related to the muscle-segmentation

genemsh in Drosophila. They are expressed in migratoryand post-migratory neural crest and appear to mediateepithelial-mesenchymal signalling interactions (Hill et al.,1989; Robert et al., 1989; Mackenzie et al., 1991a,b; Suzukiet al., 1991; Mina et al., 1995). Knockout ofMsx-1functionin mice causes abnormalities in the development of all thefacial structures, ranging from the loss of structures such asthe shelves of the palatine and maxillary bones, to a slightshortening of other bones such as the maxilla and mandible(Fig. 4; Table 2; Satokata and Maas, 1994). In addition, theteeth do not develop beyond the bud stage (Satokata andMaas, 1994).Msx-1/Msx-2double mutant mice exhibit amore severe face phenotype, suggesting a functional redun-dancy between these two genes (R. Maas, pers. commun.).

There are human syndromes due to mutation in thesegenes. Mutation in one copy of theMSX1 homeoboxdomain in humans results in the loss of premolar andmolar teeth, but there are no other detectable abnormalitiesin facial development (Vastardis et al., 1996), whilst a muta-tion in one copy of theMSX2 gene, which increases thestability of binding of the protein to the MSX2 bindingsite, TAATTG, causes Boston-type craniosynostosis (Jabs

et al., 1993; Liu et al., 1996; Ma et al., 1996; reviewed byWilkie, 1997). A similar phenotype is found in transgenicmice overexpressing theMsx-2gene, again suggesting thatincreased activity of Msx-2 can cause craniosynostosis (Liuet al., 1995).

3.1.2. Dlx gene familyThe Dlx genes comprise a multi-gene family of at least

six members in mice (Nakamura et al., 1996; Stock et al.,1996). Based on their overlapping expression patterns andfunctional analyses, it has been proposed that they have arole in patterning the branchial arches (Bulfone et al., 1993;Akimenko et al., 1994; Robinson and Mahon, 1994;Simeone et al., 1994;). They are homologous to theDis-tal-lessgene inDrosophila, which is required for the for-mation of the distal region of the legs, mouth parts andantennae (Cohen et al., 1989).Dlx-1 and -2 are expressedin migratory neural crest. In mice, onlyDlx-1 andDlx-2 areexpressed in the developing maxillary arch, whilstDlx-1–6are expressed in overlapping domains in the developingmandibular primordia (Dolle´ et al., 1992; Bulfone et al.,1993; Robinson and Mahon, 1994; Qiu et al., 1995, 1997;Zhang et al., 1997). Knockout ofDlx-1 affects the develop-ment of the palatine and pterygoid bones, whereas knockoutof Dlx-2 function in mice affects the development of theseas well as the maxillary, squamosal and jugal bones (Fig. 4;

Table 2

Skeletal structures of the mouse face and skull which can be affected by knockouts of various homeobox genes

Otx-2 Msx-1 Gsc Dlx-1 Dlx-2 Mhox Pax-6 Pax-7

Exoccipital (5)Supraoccipital (4) +BasioccipitalInterparietal (3)Otic capsuleMalleus (21) + + +Incus (19) + + +Tympanic (6) + +Parietal (2) +Alisphenoid (17) + + + + +Squamosal (18) + +Pterygoid (9) + + + +Basisphenoid (8) + +Frontal (1) + + +Orbitsphenoid +Presphenoid + +Vomer + +Palatine (12) + + + + +Jugal (11) + +Maxilla (13) + + + + + +Lachyrmal +Nasal capsule (15) + + + + +Nasal (16) + +Premaxilla (14) + +Mandible (10) + + + +Gonial (7) + +

The numbers next to the name of the element refer to the position of the element in the mouse skull shown in Fig. 4. The data from thePax-6mutant is fromanalysis of the ratSey/Seymutant. It should also be noted that the defects listed for theOtx-2mutant are for the heterozygous, not the homozygous, mutant.(+) Denotes skeletal structures of the mouse face and skull which can be affected by knockouts of various homeobox genes (see text for references).


Fig. 4. Diagram showing the skeletal structure of the mouse skull. Those elements affected by knockouts or mutation of the various homeobox genes areshaded. Frontal (1); parietal (2); interparietal (3); supra-occipital (4); exoccipital (5); tympanic (6); gonial (7); basisphenoid (8); pterygoid (9); mandible (10);jugal (11); position of the palatine (12); maxilla (13); premaxilla (14); nasal capsule (15); nasal (16); alisphenoid (17); squamosal (18); incus (19); stapes(20); malleus (21). The palatine bone is obscured by the maxilla; its position is marked only as a guide. For the sake of clarity, the basi-occipital andthepetrosal bones have been omitted. As this a lateral view, more medial bones such as the presphenoid, orbitosphenoid and vomer are obscured. For details ofthe effects on these elements see Table 2.


Table 2; Qiu et al., 1995, 1997). In theDlx-2 knockout, thereis an apparent replacement of alisphenoid with an ectopiccartilaginous structure which resembles the palatoquadrateof more primitive vertebrates. The squamosal and jugalbones also appear to be replaced by four bones. No addi-tional skeletal structures are affected in the double knockoutof Dlx-1 and Dlx-2. However, in these mutants, the max-illary molars do not form, showing a functional redundancybetweenDlx-1 andDlx-2 during the development of theseteeth. The Dlx-1/Dlx-2 knockout skeletal defects arerestricted to derivatives of the maxillary primordia suggest-ing that there is functional redundancy of Dlx-1 and-2 withother Dlxs in the developing mandibular primordia. It hasbeen suggested that some of the changes seen in theDlx-2mutant are atavistic and that the evolution ofDlx-2 resultedin changes in the shape/pattern of the facial bones (Qiu etal., 1995, 1997).

3.1.3. Otx-2Otx-2 is expressed in the migrating neural crest and the

mesenchyme of the first arch and frontonasal mass in mice(Ang et al., 1994; Osumi-Yamashita et al., 1994). Knockoutof Otx-2 function usually results in death of the embryosbefore E10 with the homozygous mutants completely lack-ing the presumptive rostral head, indicating that Otx-2 isessential for head development (Acampora et al., 1995;Matsuo et al., 1995; Ang et al., 1996). Haplo-insufficiencyin the heterozygote mice results in a wide spectrum ofdefects, ranging from the lack of a head to malformed andsmaller jaws (Fig. 4; Table 2; Matsuo et al., 1995). In someheterozygous mutants, there are changes in the alisphenoidand incus such that they resemble the more primitive pala-toquadrate cartilage. Furthermore, there are other possibleatavistic changes, such as the apparent representation of thenasal septum by two rod-like structures (Matsuo et al.,1995).

3.1.4. Bicoid homologuesTwo bicoid homologues,Goosecoid(Gsc-1) andMhox,

have been knocked out in mice which result in facialdefects. As these genes are expressed in the facial primor-dia, but not at earlier stages of facial development, thisshows that they are important in later stages of facial devel-opment such as controlling epithelial-mesenchymal interac-tions (Fig. 2).

In mice,Gsc -1is expressed in post-migratory crest in thedeveloping facial primordia and knockout of its function inmice affects most of the facial region (Fig. 4; Table 2; Gauntet al., 1993; Rivera-Perez et al., 1995; Yamada et al., 1995).A related homeobox-containing gene, pituitary homeobox-1,ptx-1, is expressed in a complementary pattern toGsc-1inthe developing mandible and nasal cavities in E11.5 andE13.5 mice (Lanctoˆt et al., 1997), although the significanceof this is unknown. In humans, Rieger’s syndrome, whichhas mild craniofacial deformaties including odontogenicabnormalities, is due to mutation in another bicoid-related

homeobox gene,RIEG (Semina et al., 1996). Analysis ofthe expression patterns of the mouse and chick homologues(known asRieg, Brx-1 or Otlx-2), shows that it is expressedin the oral epithelium of the maxillary and mandibular pri-mordia between E8.5–11 and at stage 18 of embryonicdevelopment, respectively (Semina et al., 1996; Kitamuraet al., 1997; Mucchielli et al., 1997). Therefore, the odonto-genic defects and the facial abnormalities seen in Rieger’ssyndrome are likely due to defective epithelial-mesenchy-mal interactions. However, asRieg is also expressed in thedeveloping head mesoderm and later in the developingbrain, it is also possible that these defects are due to defec-tive mesoderm development and/or neural crest formationand migration.

Mhox(also calledPrx-1) is expressed in the mesenchymeof the developing facial primordia in mice and chicks (Cser-jesi et al., 1992; Nohno et al., 1993; Kuratani et al., 1994).Knockout ofMhoxfunction in mice affects the developmentof skeletal elements derived from the maxillary primordiaand the proximal part of the mandibular primordia (Fig. 4;Table 2; Martin et al., 1995).

3.1.5. Paired-related genesThe paired box family consists of nine members which

are transcription factors containing a conserved proteinmotif, the paired box and, with the exception of Pax-1and-9, a homeobox or partial homeobox domain (reviewedby Dahl et al., 1997). During development they are thoughtto have several roles, including controlling epithelial-mesenchymal transitions, differentiation and proliferation.At least five members of the paired box gene family,Pax-1,-3, -6, - 7 and -9, are expressed in the developing facialprimordia (Goulding et al., 1991; Jostes et al., 1991;Walther and Gruss, 1991; Dietrich and Gruss, 1995; Grind-ley et al., 1995; Mansouri et al., 1996; Neubu¨ser et al., 1995,1997). To date, fourPax mutants or knockouts have beencharacterized. Thesmall eye, Pax-6, mutant is the most wellstudied in terms of craniofacial development, and at leastsome of the facial defects are associated with abnormalneural crest migration (see references below). Studies ofthe Pax-3mutant,splotch, have shown that Pax-3 controlstrunk neural crest migration.Pax-3and-7 are expressed inregions of the neural tube/developing brain which will giverise to neural crest and in the migrating neural crest cellsthemselves, raising the possibility the facial defects in thesemutants/knockouts may be due to failure of epithelial-mesenchymal transformations or neural crest migration,but this has yet to be determined.

The Pax-1mutant,undulated, is not associated with theloss of facial skeletal structures, but they are shorter thannormal and the angle between the upper and lower jaw isabnormal (Dietrich and Gruss, 1995). AsPax-1is expressedin the mesenchyme of the facial primordia, thePax-1 facedefects may be due to defective epithelial-mesenchymalinteractions during later facial development.

The homozygous mousePax-3mutant,splotch, does not


have facial defects, possibly due to compensation by othermembers of the gene family that are expressed in similarregions (Morell et al., 1992). However, on a differentgenetic background, there are craniofacial abnormalities inthesplotchmutant and it has been suggested that at least twogene products interact to modulate the function of Pax-3(Asher et al., 1996a). Studies of its role in trunk neuralcrest have shown that Pax-3 can modulate the expressionof cell-adhesion molecules. The failure of trunk neural crestcells to migrate is associated with abnormally high levels ofN-CAM proteins, some of which are incorrectly processed(Moase and Trasler, 1991) and the ectopic expression ofversican, a chondroitin sulphate proteoglycan, which inhi-bits neural crest migration (Henderson et al., 1997). In anormal embryo,Pax-3andversicanare expressed in com-plementary domains along pathways of trunk neural crestmigration and it is proposed that Pax-3 negatively regulatesversicanexpression, therefore allowing neural crest migra-tion (Henderson et al., 1997). It is possible that Pax-3 mayaffect facial development by modulating cranial neural crestmigration.

The small eye(Sey) mice and rat mutants are due to amutation inPax-6 (Hill et al., 1991; Matsuo et al., 1993).These animals have abnormalities of the eyes and nasal pitsand an ectopic bar of cartilage forms which is thought tosubstitute for the lateral nasal capsule (Fig. 4; Table 2; Kauf-man et al., 1995; Osumi-Yamashita et al., 1997). Studies inSeyrat embryos have shown that the anterior midbrain crestmigrates abnormally (Matsuo et al., 1993; Osumi-Yama-shita et al., 1997) and studies ofSeymice have suggestedthat some of this crest population dies during migration(Grindley et al., 1995).Pax-6is not expressed in the anteriormidbrain crest cells, but it is expressed in the rostral headectoderm and forebrain, suggesting that environmental cuesfor migration are abnormal in theSeymutants. This has beenconfirmed by heterotopic crest transplants in which labelledmutant midbrain crest have been grafted into wildtype hostembryos and vice versa. In wildtype hosts, mutant cellsmigrate along the normal pathway whereas the wildtypecells migrate abnormally in the mutant hosts (Osumi-Yama-shita et al., 1997). Therefore, secreted molecules which con-trol neural crest migration appear to be altered inSeymutants. Furthermore, other experiments have shown thatwildtype cells and mutant cells do not mix in chimeric mice,again suggesting that there are cell surface changes inSeymutants (Quinn et al., 1996). Candidate genes that may beaffected include those encoding cell-adhesion molecules,which are known to be important in neural crest migration,and some of which have promoters that can bind membersof the Pax gene family (reviewed by Bronner-Fraser, 1993;Kozmik et al., 1992; Chalepakis et al., 1994; Holst et al.,1997). In thePax-7 knockout, as with thePax-6 mutant,development of the nasal septum/capsule is affected andin addition, the maxillary bone is reduced in size (Fig. 4;Table 2; Mansouri et al., 1996).

As with the naturally occurringPax-3andPax-6mouse

mutations, there are human syndromes due to mutation inthese genes. In humans, haplo-insufficiency ofPAX6resultsin Aniridia, a complete or partial absence of the iris (Nelsonet al., 1984; Ton et al., 1991), or Peter’s anomaly (Hanson etal., 1994). Haplo-insufficiency ofPAX3 in humans, due tomissense mutations or deletions in thePAX3gene, results inWaardenburg’s syndrome type 1, characterized by deafnessand some craniofacial anomolies (Tassabehji et al., 1992). Adistinct syndrome, craniofacial-deafness-hand syndrome,which includes deafness and hypoplasia of the nasalbones, is due to an asparagine to lysine mutation in thepaired box domain ofPAX3(Asher et al., 1996b). Interest-ingly, mutation of the same amino-acid, but this time tohistidine, results in Waardenburg syndrome type 3, whichis not characterized by the loss or hypoplasia of the nasalbone (Hoth et al., 1993).

Another paired-class homeobox-containing gene isCart-1. In rat embryos,Cart-1 is expressed in the early develop-ing head mesenchyme, later the branchial arches and finallyin regions of chondrogenesis (Zhao et al., 1994). In mice,knockout ofCart-1 results in a hypoplastic facial skeletonand the loss of the cranial vault, the latter associated with afailure of the neural tube to close at the midbrain level (Zhaoet al., 1996). The early migrating neural crest appears to benormal, but there is increased apoptosis of the forebrainmesenchymal cells, which may account for the developmentof some of the hypoplastic facial structures, suggesting thatCart-1 is needed for survival of some cells.

3.2. Other transcription factors

Other transcription factors that are not homeobox-con-taining genes but which have been shown to have a rolein facial development include Glis (discussed above),twist, ski and AP-2. Twist is a transcription factor contain-ing a basic helix-loop-helix motif which is expressed in themesenchyme of the mouse developing face, with high levelsof expression in the first arch (Wolf et al., 1991; Fu¨chtbauer,1995; Stoetzel et al., 1995; Gitelman, 1997). Knockout oftwist function in mice is lethal before E10.5. The develop-mental defects include failure of the neural tube to close,abnormal head mesenchyme and branchial arch formation(Chen and Behringer, 1995). In humans, Saethre–Chotzensyndrome, characterized by cleft palate, abnormal nasal sep-tum formation, maxillary hypoplasia and craniosynostosis,is due to mutation of one copy of theTWIST gene (ElGhouzzi et al., 1997; Howard et al., 1997).

AP-2 (activating protein-2) is expressed in the cranialneural crest, with knockout of its function in mice resultingin severely hypoplastic facial skeletal structures, which areseparated at the midline and exencephaly (Schorle et al.,1996; Zhang et al., 1996; Shen et al., 1997). Migration ofthe neural crest cell populations appears to be normal inthese mice. However, there is increased cell death in thebranchial arches, suggesting that later phases of facialdevelopment are affected, possibly epithelial-mesenchymal


interactions. The midline defects appear to be due to failureof the neural plate to fold correctly, which may be due toincreased apoptosis which occurs in theAp-2 mutants(Schorle et al., 1996). AP-2 is thought to have a numberof roles during development, including regulating matrixsynthesis, apoptosis andMsx-1 gene expression, althoughin the Ap-2 knockout, Msx-1 expression appears to beunchanged in the facial primordia at E10.5 (Schorle et al.,1996; reviewed by Morriss-Kay, 1996). The abnormalitiesin facial development in theAp-2 knockout may be due tochanges in a number, if not all, of these processes (reviewedby Morriss-Kay, 1996). N-CAM expression is increased inthe neural tube ofAp-2 mutants. Interestingly,Pax-3mutants also have increased levels of N-CAM expressionin the trunk neural tube. Furthermore, the eye defect inAp-2mutants has some similarities to that in thePax-6 mutant(reviewed by Morriss-Kay, 1996). This may suggest thatPax and Ap-2 expression are part of a common pathwayin some regions of embryonic development.

Knockout ofski, a proto-oncogene, has a similar pheno-type to Ap-2, i.e. exencephaly and facial clefting, but incontrast only a few of the skull bones are severely affected(Berk et al., 1997). However, as with theAp-2 knockout,derivatives of the mandibular skeleton are less affected thanmore anterior regions. Analysis of E9.5 mutant embryosshows that there is increased apoptosis of the cranial neuraltube and craniofacial mesenchyme, with predominant cell-death in the frontonasal mass region. This is in contrast totheAp-2mutant, where predominant apoptosis is detected inthe midbrain/hindbrain region, which is associated with theforming branchial arches (Schorle et al., 1996). As with theAp-2mutant, neural crest migration appears normal.

Another gene that is involved in facial development istreacle which is predicted to encode a phosphorylatednuclear protein (Wise et al., 1997). Mutation of this geneis thought to be responsible for Treacher–Collins syndrome,characterized by hypoplasia of some skeletal structuresderived from the first arch, cleft palate and ear abnormalities(Dixon et al., 1996). The expression of this gene duringfacial development is not known, so it is unclear whichstages of facial development are affected by mutation ofthis gene.

4. Signalling interactions

To understand facial development, we need to knowwhich signalling pathways co-ordinate facial outgrowthand patterning, and how these signalling pathways are inte-grated. It is clear that many factors expressed in the face alsohave other roles during vertebrate development and thereare likely to be similar signalling cascades. However, muchof the vertebrate head arose through the evolution of thecranial crest (for a review see Gans and Northcutt, 1983)and although some functions of genes and their homologueshave been conserved during evolution (for example, eyeless

and Pax-6 during the development of the eye inDrosophilaand mice, respectively) it is as yet unclear if all the signal-ling cascades have been conserved and whether there areany novel factors which are involved in face development.

Outgrowth of the facial primordia is dependent on epithe-lial-mesenchymal signalling interactions (Wedden, 1987;Richman and Tickle, 1992), but the role of these interac-tions in controlling patterning is unknown and there is stillvery little data about which signalling factors can regulatethe expression of the homeobox-containing genes. Forexample, although we know that FGFs can partially main-tain outgrowth of the facial primordia, we do not yet knowwhich genes are regulated by FGF expression. However, wedo know that BMP-2 and BMP-4 can activate/maintain theexpression ofMsx-1andMsx-2 in the mesenchyme in thedeveloping facial primordia (Fig. 3C; Barlow and Francis-West, 1997). Furthermore, in the maxillary primordium,ectopic application of BMP-2 and BMP-4 can extend theexpression ofMsx-1 in the mesenchyme, which is asso-ciated with an extension ofShhin the overlying epithelium(Fig. 3C,D; Barlow and Francis-West, 1997). As Shh prob-ably activates the expression ofBmp-2 in the maxillarymesenchyme, there may be a signalling loop betweenShhin the epithelium andBmp-2and Msx-1 in the underlyingmesenchyme. In the chick mandibular primordia,Shhexpression is not detectable until much later in developmentand therefore, Shh is unlikely to have a role in patterningand growth of the chick mandibular primordia. However, inthe mandibular primordia, ectopic application of BMP-2 orBMP-4 can extend the expression ofMsx-1in the mesench-yme andFgf-4 in the overlying epithelium, suggesting thatthere is a signalling loop between mesenchymal expressionof Bmp-2and epithelial expression ofFgf-4 (Barlow andFrancis-West, 1997).

Recombination experiments between facial mesenchymeand limb ectoderm have shown that, overall, the epitheliumis interchangeable and reasonable development/differentia-tion can occur in all the facial primordia with ectopic epithe-lia (Richman and Tickle, 1989). Furthermore, facialmesenchyme, unlike flank mesenchyme, is able to maintainthe structure of the AER for at least 48 h (Richman andTickle, 1992). The AER is a region of thickened ectodermnormally responsible for maintaining outgrowth of thedeveloping limb bud. This ability of facial mesenchyme tomaintain the AER in particular, suggests that there are atleast some common epithelial-mesenchymal signallingpathways between developing limb buds and facial primor-dia. Molecules which are involved in signalling interactionsthat are expressed in both the AER and facial epitheliuminclude Fgf-4, Fgf-8, jagged 2, Bmp-2and Bmp-4, whilstMsx-1, Msx-2, Bmp-2andWnt-5agenes are expressed in themesenchyme of both the limbs and facial primordia. Theability of BMP-2 to extend the epithelial expression ofFgf-4in the mandibular primordia is also seen in the developinglimb bud, again indicating that there are parallels in thesignalling cascades (Duprez et al., 1996; Barlow and Fran-


cis-West, 1997). However, experiments in which maxillarymesenchyme has been grafted into the developing limb andvice versa have suggested that the signalling interactions arenot totally identical (Brown et al., 1993). This idea is sup-ported by comparing the expression of some of the putativekey players such asShh, Bmp-2, FgfsandMsx in relation toeach other in the developing face and limb. Furthermore, atleast one important signalling molecule in limb develop-ment, Wnt-7a, is not expressed in the chick developingfacial primordia (Francis-West, unpublished data). How-ever, in the latter case, it is possible that other members ofthe Wnt gene family which are expressed in the developingfacial primordia have equivalent roles to Wnt-7a in thedeveloping limb bud.

As facial structures and teeth both arise by signallinginteractions between the epithelium and ectomesenchyme,studies performed on the developing tooth also may provideinsight into signalling networks that co-ordinate facialdevelopment. In the developing tooth bud, both BMP-2/-4and members of the FGF family can activate/maintainmesenchymal expression ofMsx-1 and Msx-2 (Vainio etal., 1993; Chen et al., 1996). In addition, BMP-4 can acti-vateBmp-4and the transcription factors,Lef-1 andEgr-1.Whilst both FGFs and BMPs can activate/maintainMsx

gene expression and requireMsx gene function to mediatesome of their effects, they can also act differentially in theirability to activate other genes (Chen et al., 1996). For exam-ple, BMPs can induce/maintainnotch-1, -2 and -3, but notjagged 1, gene expression, whilst FGFs have the converseeffect (Mitsiadis et al., 1997). BMPs and FGFs also actdifferentially in their ability to activatePax-9 expression.FGF-8 inducesPax-9 expression but BMP-4 prevents thisinduction, suggesting that the two signalling pathways areantagonistic (Neubu¨ser et al., 1997).Pax-9appears to definethe presumptive tooth mesenchyme, suggesting that BMP-4/FGF-8 signalling may determine this region (Neubu¨ser etal., 1997). Interestingly, there is a similar complementarypattern ofBmp-4andFgf-8 expression in the chick devel-oping maxillary and mandibular primordia (Fig. 5C,D;Bogardi, Barlow and Francis-West, unpublished data).How this differential expression ofFgf-8andBmp-4affectsgene expression in the underlying mesenchyme and subse-quent development of these primordia, remains to be deter-mined but it suggests that antagonistic interactions betweenBMP and FGF signalling may have a more general role indevelopment.

Very little is known about the Wnt signalling pathwayduring facial development; the expression patterns of the

Fig. 5. Whole mount in situ hybridisation showing the expression ofBmp-4(A,C) andFgf-8 (B,D) at stage 12 (A,B) and stage 16 (C,D).Bmp-4andFgf-8areinitially expressed in overlapping domains across the tip of the mandibular primordia but then become expressed in complementary domains at the medialtips and the lateral edge, respectively. At stage 16,Bmp-4 and Fgf-8 are also expressed in complementary domains in the maxillary primordia. Thisdifferential expression may be related to the different combinations of crest that populate each primordia. md, mandibular primordia; mx, maxillaryprimordia.


Wnt gene family members still needs to be fully determined.Functional studies also need to be performed to determine ifthere is a signalling loop betweenWntandShhgene expres-sion, as there is in the developing limb bud, which maycontrol outgrowth and patterning of the developing facialprimordia. Alternatively, Shh and members of the Wnt genefamily may act synergistically, as they do during the induc-tion of myogenesis during somite development (Mu¨nster-berg et al., 1995).

Most functional studies of signalling interactions thatcontrol homeobox gene expression during facial develop-ment have focused on theMsxgene family, but the signal-ling factors that regulateDlx and most other homeobox geneexpression in the developing facial primordia are notknown. It will be important to analyse these signalling inter-actions, not only becauseDlx genes are important in pattern-ing, but because recent studies have shown that members ofthe Msx gene family can form heterodimers with membersof theDlx family, and with other transcription factors suchas Miz1 (Wu et al., 1997; Zhang et al., 1997).Dlx/Msxheterodimers have different transcriptional activities invitro than homodimers. Since members of theMsx andDlx gene families are expressed in overlapping domainsduring facial development (Zhang et al., 1997), fates ofdifferent regions of the developing facial primordia maybe determined by the specific gene combinations that theyexpress. To understand facial patterning, we will need toidentify the signalling factors that maintain or induce theexpression of these homeobox-containing genes and thendetermine how the expression of these genes influencesfacial development.

5. Patterning of the facial primordia-neural crest versusenvironmental signals?

Patterning of the developing face is probably controlledby co-ordinated signalling between the mesenchyme andepithelium, as in many other vertebrate structures. How-ever, the individual roles of these tissues in patterning thefacial primordia are far from understood. Furthermore, dif-ferent mechanisms may pattern different axial levels. Theresults from a number of studies suggest that at least somepatterning information is inherent to even premigratoryneural crest cells. When an anuran (toad) cephalic neuralfold was transplanted to a urodele (newt) host, the visceralarch skeleton which developed on the operated side was ofcharacteristically donor morphology, despite having devel-oped in a host mesoderm and epithelial branchial arch envir-onment (Wagner, 1949). Furthermore, Horstadius andSellman (1946) reciprocally transplanted mandibular andgill arch neural crest by rotation of the neural plate inanother urodele and found that, while each could followthe others’ migratory routes, some morphological featuresof the cartilaginous elements which developed were moreappropriate to the original source of the neural crest than to

their new location, particularly with respect to fusion withother cartilaginous elements. Even more striking are theresults of Noden (1983) regarding heterotopic transplanta-tions of quail or chick premigratory first arch neural crest tochick second or third arch levels. Graft-derived ectopic car-tilages and bones developed in virtually all cases; the shapesand arrangements of these elements were highly character-istic of proximal mandibular arch elements.

This neural crest pre-patterning hypothesis is compli-cated, however, by the finding that rostral mesencephalicneural crest, which normally forms maxillary, periocularand frontonasal skeletal structures, produce the same super-numerary mandibular arch elements as rostral metencepha-lic neural crest (the normal source of the mandibularelements) when transplanted to second or third arch levels(Noden, 1983). Furthermore, transplantation of mandibularneural crest to more rostral levels (Noden, 1978) results in anormal head skeleton, although the same graft would pro-duce ectopic mandibular structures if transplanted caudally.This shows that in the chick, in addition to the first archcrest, the crest which gives rise to structures anterior of themandible is ‘pre-patterned’ to form mandibular structures.However, normally this ‘mandibular pre-patterning’ is mod-ified by environmental signals during development to formthe anterior structures of the head (Noden, 1983).

The pre-patterning of the first arch crest is not absolute.For example, in the majority of cases with supernumerarymandibular elements, normal second and third arch struc-tures were also produced, with most being of chimericdonor/host origin (Noden, 1983). This again suggests thatenvironmental signals can modify the fate of the crest.There may also be a population effect, whereby respecifica-tion can occur, but only if the population of ectopic neuralcrest cells is not too large. Another technique to divertneural crest cells to ectopic arches is the extirpation of pre-migratory neural crest, which subsequently regenerates toyield a normal skeleton. Labelling studies have shown thatunder certain conditions, the regenerated neural crest cellsare derived from neural crest immediately adjacent to theexcised region, both anterior and posterior (Couly et al.,1996; Saldivar et al., 1997). In one such study (Couly etal., 1996), first arch crest cells repopulated the second arch,either in its entirety, or in combination with neural crestcells derived from caudal hindbrain levels. Second archelements developed normally in all cases and supernumer-ary mandibular elements were not observed, implicating thearch environment for control of patterning. However, regu-lated neural crest cells may have different properties tothose from which they were derived; neural crest ablationhas been shown to alter gene expression of neural crest cellseven at a distance from the ablation site, prior to their emi-gration (Saldivar et al., 1997). The first arch-derived cellsthat repopulated the second arch after ablation may there-fore have lost original pre-patterning information and /orbecome more susceptible to environmental cues.

Two more lines of evidence suggest that the environment


can alter the fate of the crest. In neural crest rotation studiesin which second and third arch neural crest were recipro-cally redirected, normal second arch structures wereformed, but there was an almost complete absence of thirdarch structures (Saldivar et al., 1997). In both cases, the crestpopulations do not appear to be pre-patterned and the nor-mal development of the second arch from third arch crestsuggests that environmental signals are important in pattern-ing. However in this case, the third arch environment is notsufficient to promote third arch skeletogenesis from secondarch crest. The different results in the two arches followingintroduction of ectopic neural crest may be due to actualspecific differences in the neural crest populations or archenvironments, or may be a function of the direction of trans-plantation, i.e. neural crest cells may be more difficult torespecify when transplanted caudally than rostrally. Inanother study, early and late migrating crest have beenshown to have equivalent potential, forming structuresappropriate to their new environment when transplantedinto older or younger hosts (Baker et al., 1997). All thisdata suggests that the neural crest is plastic and can respondto environmental signals changing fate under the appropri-ate circumstances. These signals could be from the adjacentneural tube or neural crest populations. Alternatively, thecranial neural crest may receive signals from the mesodermor ectoderm.

Ectoderm could also directly pattern skeletal structuresby the localized production of factors which could eitherentrap determined chondrogenic cells, or directly promotechondrogenesis in underlying neural crest cells. Type IIcollagen expression is found in a pattern which prefiguresthe future elements of the neurocranium, at stages whenchondrogenesis-promoting interactions are known to beoccurring (Thorogood et al., 1986; Wood et al., 1991). Ithas been proposed as a candidate molecule with a major rolein this model, although other molecules would also have tobe involved to explain its presence in non-chondrogenicregions of the embryo. However, this model is specific forpatterning neurocranial elements and would not apply to thevisceral skeleton (reviewed by Hanken and Thorogood,1993; Thorogood, 1993).

At later stages of development, the epithelium probablyhas an, as yet, unidentified role in controlling patterning ofthe visceral structures. Whether or not the neural crest cellsare pre-patterned, they could still induce/maintain theexpression of factors which have a role in patterning. Sup-port for this possibility comes from work by Hunt et al.(1991c), who showed thatHoxb-2 is initially expressed inthe migrating neural crest cells of the second arch, but issubsequently also expressed in the ectoderm overlying theHoxb-2expressing mesenchyme. Alternatively, the paraxialmesoderm may induce/maintain the expression of epithelialfactors that control patterning. In either case, this would beanalogous to the situation seen in the limb bud where dif-ferent tissues control one aspect of patterning at differentstages. During limb development, signals from the meso-

derm induce the expression ofEn-1andWnt-7ain the pre-sumptive ventral and dorsal ectoderm, respectively. Thesemolecules subsequently control patterning of the d-v axis ofthe developing limb bud (Michaud et al., 1997). Therefore,first the mesoderm, then the ectoderm have instructive rolesin controlling patterning of the d-v axis of the limb.

Finally, whilst the role of the paraxial mesoderm in earlysignalling interactions is unknown, myogenic cells whichare derivatives of the paraxial mesoderm appear to be pat-terned by the neural crest populations. This has been shownin experiments in which the first arch crest has been trans-planted caudally. Analysis of the muscles formed from theparaxial mesoderm showed that they are appropriate to theectopic mandibular structures (Noden, 1983). Another studyin which somitic or pre-somitic mesoderm, which normallygive rise to trunk and limb muscles, was transplanted intothe midbrain region, some donor myogenic cells were pat-terned according to the local environment again indicatingthat the myogenic cells have no inherent patterning ability(Noden, 1986). Similar experiments in the developing limbbud have shown that the muscles are patterned accordingto their local environment. The mechanisms that controlmuscle patterning in both systems are far from understood.However, ET-1 may be involved in ectomesenchyme-paraxial mesoderm signalling interactions in the branchialarches:Et-1 is expressed in the paraxial mesoderm whilstits receptor,EtA, has been detected in the adjacent ecto-mesenchyme (Maemura et al., 1996; Clouthier et al.,1998).

6. Unravelling facial development

To fully understand facial patterning, we need to firstunderstand the mechanisms which control outgrowth andpatterning of the facial primordia at different stages of facialdevelopment. Second, we need to identify signalling path-ways involved in these processes. Ultimately, we will needto know how these signalling pathways interact to co-ordi-nate facial development. However, there is still a lack ofknowledge about very basic aspects of facial development.For example, when are the skeletal elements of the facialprimordia specified/patterned? This problem is particularlyevident during the development of the maxillary primor-dium, which gives rise to a series of bones. If we assumethat some aspects of patterning must operate during out-growth of the facial primordia, we still do not know if theskeletal elements formed from the developing maxillaryprimordia are patterned/specified sequentially, as occurswith the skeletal elements in the developing limb budwhere the more proximal elements are specified and formedprior to the development of the distal structures. Alterna-tively, although less easy to envisage, the skeletal elementsfrom the maxillary primordia may all be specified simulta-neously.

We will also need to address signalling interactions dur-


ing neural crest migration. For example, the environmentappears to be important in patterning many of the neuralcrest populations. The tissues and signalling factorsinvolved in these interactions are not known.

Another question that needs to be resolved is whether theepithelium has any role in patterning the facial primordia.Whilst ectoderm-mesenchyme recombination experimentshave shown that the epithelia are all interchangeable, allow-ing outgrowth and appropriate differentiation of cartilage ormembrane bone, whether the epithelium can control facialpatterning has not been fully explored. Furthermore, if theepithelium has a role in facial patterning, what are the sig-nalling factors?Shhis expressed in the epithelium of someof the developing facial primordia and is therefore, a strongcandidate but its patterning role has yet to be demonstrated.

Other fundamental questions that need to be addressedinclude determining how the neural crest populations influ-ence gene expression in the epithelium once they havereached the facial primordia. For example, do differentpopulations of crest maintain/induce the differential expres-sion of signalling molecules in the epithelium? In the chick,during early facial development,Bmp-4 and Fgf-8 areexpressed in complementary domains in both the develop-ing mandibular and maxillary primordia (Fig. 5). In thedeveloping chick maxillary primordia, their expressionappears to be associated with the arrival of neural crest inthe forming arch suggesting the incoming crest may inducetheir expression (Fig. 5). A similar association between theinduction of Fgf-8 expression following the arrival of themigrating neural crest has also been reported in the mouse(Heikinheimo et al., 1994). How are these differentialexpression patterns set up and are they related to the differ-ent populations of neural crest that form the primordia? Inthe mouse, antagonistic interactions between BMP-4 andFGF-8 signalling are thought to define the presumptive den-tal region in the mandibular primordia. Furthermore, morerecently antagonistic BMP/FGF signalling interactions havebeen shown to be involved in limb and pituitary develop-ment (Buckland et al., 1998: Ericson et al., 1998). Thus, anunderstanding of how these reciprocal patterns ofBmp-4and Fgf-8 expression are determined should give us afurther understanding of facial patterning and may alsogive us insight into signalling interactions between differentpopulations of crest.

The maxillary and mandibular primordia are made up ofthe neural crests from several distinct populations which, onthe whole, remain segregated giving rise to distinct skeletalstructures. However, in some regions of the developingface, distinct crest cell populations intermingle. How is pat-terning between these different crest populations co-ordi-nated, is there one crest population that is dominant, i.e.instructive over another or is it via epithelial signals?Furthermore, is there a region within the developing facialprimordia analogous to the polarising region in the limb thatsignals to specify the development of neighbouring cells?Although, it would seem likely that there must be some such

region, it has not yet been identified within any of the facialprimordia and may not exist.

Finally, other unresolved questions do not relate to pat-terning but concern differentiation. The cranial crest, unlikethe trunk crest, has chondrogenic potential and acquires itprior to migration. Furthermore, the ability to form dermalbone is exclusive to the cranial crest in higher vertebrates.We still do not understand the signalling interactionsinvolved in these processes and yet these are fundamentalto the development of the head.

Knockout approaches, overexpression and ectopicexpression studies will continue to give insight into howthe facial primordia are patterned, and studies in thechick, mouse and zebrafish will all contribute significantlyto our understanding of facial development. The identifica-tion of promoters to specifically mis-express genes orknockout expression, via acre/loxP system, in distinctareas of the face would be a powerful tool to study signal-ling interactions in facial development in mice. In addition,the identification and analysis of zebrafish mutants withcraniofacial defects will also provide significant insightsinto signalling interactions that control facial development(Piotrowski et al., 1996; Schilling et al., 1996a). Alreadysuch mutants have been used to determine when a gene iscrucial during facial development (Schilling et al., 1996b).The identification of the mutant genes will provide furtherimportant stepping stones to answer questions about facedevelopment.

Acknowledgements

We would like to thank YiPing Chen, Darrell Evans,Brian Hall, Paul Sharpe, Moya Smith, Peter Thorogoodand Jenny Whiting for their many helpful comments andinsights on the manuscript. Finally, we thank Martin Cha-perlin for invaluable help with this manuscript. Work in theauthors’ laboratory was supported by the MRC and TheWellcome Trust.

References

Acampora, D., Mazan, S., Lallemand, Y., Avantaggiato, V., Maury, M.,Simeone, A., Bruˆlet, P., 1995. Forebrain and midbrain regions aredeleted inOtx2−/− mutants due to a defective anterior neuroectodermspecification during gastrulation. Development 121, 3279–3290.

Akimenko, M.-A., Ekker, M., Wegner, J., Lin, W., Westerfield, M., 1994.Combinatorial expression of three zebrafish genes related toDistal-Less: part of a homeobox gene code for the head. J. Neuroscience 14,3475–3486.

Alagille, D., Odievre, M., Gautier, M., Dommergues, J.P., 1975. Hepaticductular hypoplasia associated with characteristic faces, vertebral mal-formations, retarded physical, mental, and sexual development, andcardiac murmur. J. Pediatr. 86, 63–71.

Alagille, D., Estrada, A., Hadchouel, M., Gautier, M., Odievre, M.,Dommergues, J.P., 1987. Syndromic paucity of interlobular bile ducts(Alagille syndrome or arteriohepatic dysplasia): review of 80 cases. J.Pediatr. 110, 195–200.


Alexandre, C., Jacinto, A., Ingham, P.W., 1996. Transcriptional activationof hedgehogtarget genes inDrosophila is mediated directly by theCubitus interruptus protein, a member of the GLI family of zinc fingerDNA-binding proteins. Genes Dev. 10, 2003–2013.

Ang, S.L., Conlon, R.A., Jin, O., Rossant, J., 1994. Positive and negativesignals from mesoderm regulate the expression of mouse, Otx2 in ecto-derm explants. Development 120, 2979–2989.

Ang, S.-L., Jin, O., Rhinn, M., Daigle, N., Stevenson, L., Rossant, J., 1996.A targeted mouseOtx2 mutation leads to severe defects in gastrulationand formation of axial mesoderm and to deletion of rostral brain.Development 122, 243–252.

Asher, J.H., Harrison, R.W., Morell, R., Carey, M.L., Friedman, T.B.,1996a. Effects ofPax3 modifier genes on craniofacial morphology,pigmentation, and viability: a murine model of Waardenburg syndromevariation. Genomics 34, 285–298.

Asher, J.H., Sommer, A., Morell, R., Friedman, T.B., 1996b. Missensemutation in the paired domain of PAX3 causes craniofacial-deafness-hand syndrome. Hum. Mutat. 7, 30–35.

Baker, C.V.H., Bronner-Fraser, M., Le Douarin, N.M., Teillet, M.-A.,1997. Early- and late-migrating cranial neural crest cell populationshave equivalent developmental potential in vivo. Development 124,3077–3087.

Barlow, A.J., Francis-West, P.H., 1997. Ectopic application of BMP-2 andBMP-4 can change patterning of developing chick facial primordia.Development 124, 391–398.

Barni, T., Maggi, M., Fantoni, G., Serio, M., Tollaro, I., Gloria, L.,Vannelli, G.B., 1995. Identification and localization of endothelin-1and its receptors in human fetal jaws. Dev. Biol. 169, 373–377.

Basilico, C., Moscatelli, D., 1992. The FGF family of growth factors andoncogenes. Adv. Cancer Res. 59, 115–165.

Belloni, E., Muenke, M., Roessler, E., Traverso, G., Siegel-Bartelt, J.,Frumkin, A., Mitchell, H.F., Donis-Keller, H., Helms, C., Hing, A.V.,Heng, H.H.Q., Koop, B., Martindale, D., Rommens, J.M., Tsui, L.-C.,Scherer, S.W., 1996. Identification ofSonic hedgehogas a candidategene responsible for holoprosencephaly. Nat. Genet. 14, 353–356.

Bennett, J.H., Hunt, P., Thorogood, P., 1995. Bone morphogenetic protein-2 and -4 expression during murine orofacial development. Arch. OralBiol. 40, 847–854.

Berk, M., Desai, S.Y., Heyman, H.C., Colmenares, C., 1997. Mice lackingtheski proto-oncogene have defects in neurulation, craniofacial pattern-ing, and skeletal muscle development. Genes Dev. 11, 2029–2039.

Bradley, R.S., Brown, A.M.C., 1995. A soluble form of Wnt-1 protein withmitogenic activity on mammary epithelial cells. Mol. Cell. Biol. 15,4616–4622.

Brickell, P., Thorogood, P., 1997. Retinoic acid and retinoic acid receptorsin craniofacial development. Cell. Dev. Biol. 8, 437–443.

Bronner-Fraser, M., 1993. Mechanisms of neural crest cell migration.BioEssays 15, 221–230.

Brown, J.M., Wedden, S.E., Millburn, G.H., Robson, L.G., Hill, R.E.,Davidson, D.R., Tickle, C., 1993. Experimental analysis of the controlof expression of the homeobox-geneMsx-1 in the developing limb andface. Development 119, 41–48.

Brown, J.M., Robertson, K.E., Wedden, S.E., Tickle, C., 1997. Alterationsin Msx-1andMsx-2expression correlate with inhibition of outgrowth ofchick facial primordia induced by retinoic acid. Anat. Embryol. 195,203–207.

Buckland, R.A., Collinson, J.M., Graham, E., Davidson, D.R., Hill, R.E.,1998. Antagonistic effects of FGF4 on BMP induction of apoptosis andchondrogenesis in the chick limb bud. Mech. Dev. 71, 143–150.

Bulfone, A., Kim, H.-J., Puelles, L., Porteus, M.H., Grippo, J.F.,Rubenstein, J.L.R., 1993. The mouseDlx-2 (Tes-1) gene is expressedin spatially restricted domains of the forebrain, face and limbs in mid-gestation mouse embryos. Mech. Dev. 40, 129–140.

Celli, G., LaRochelle, W.J., Mackem, S., Sharp, R., Merlino, G., 1998.Soluble dominant-negative receptor uncovers essential roles for fibro-blast growth factors in multi-organ induction and patterning. EMBO J.17, 1642–1655.

Chalepakis, G., Wijnholds, J., Giese, P., Schachner, M., Gruss, P., 1994.Characterization of Pax-6 and Hox-1 binding to the promoter regionof the neural cell-adhesion molecule. L1. DNA Cell Biol. 13, 891–900.

Chen, Y., Bei, M., Woo, I., Satokata, I., Maas, R., 1996.Msx1 controlsinductive signaling in mammalian tooth morphogenesis. Development122, 3035–3044.

Chen, Z.F., Behringer, R.R., 1995. Twist is required in head mesenchymefor cranial neural tube morphogenesis. Genes Dev. 9, 686–699.

Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal,H., Beachy, P.A., 1996. Cyclopia and defective axial patterning in micelacking Sonic hedgehoggene function. Nature 383, 407–413.

Christiansen, J.H., Dennis, C.L., Wicking, C.A., Monkley, S.J., Wilkinson,D.G., Wainwright, B.J., 1995. MurineWnt-11andWnt-12have tempo-rally and spatially restricted expression patterns during embryonicdevelopment. Mech. Dev. 51, 341–350.

Clouthier, D.E., Hosoda, K., Richardson, J.A., Williams, S.C.,Yanagisawa, H., Kuwaki, T., Kumada, M., Hammer, R.E.,Yanagisawa, M., 1998. Cranial and cardiac neural crest defects inendothelin-A receptor-deficient mice. Development 125, 813–824.

Coffin-Collins, P.A., Hall, B.K., 1989. Chondrogenesis of mandibularmesenchyme from the embryonic chick is inhibited by mandibularepithelium and by epidermal growth factor. Int. J. Dev. Biol. 33,297–311.

Cohen, M.M., Kreiborg, S., 1996. A clinical study of the craniofacialfeatures in Apert syndrome. Int. J. Oral. Maxillofac. Surg. 25, 45–53.

Cohen, S.M., Bro¨nner, G., Kuttner, F., Ju¨rgens, G., Ja¨ckle, H., 1989.Dis-tal-lessencodes a homeodomain protein required for limb developmentin Drosophila. Nature 338, 432–434.

Couly, G.F., Le Douarin, N.M., 1985. Mapping of the early neural primor-dium in quail-chick chimeras. I. Developmental relationships betweenplacodes, facial ectoderm, and prosencephalon. Dev. Biol. 110, 422–439.

Couly, G.F., Le Douarin, N.M., 1987. Mapping of the early neural primor-dium in quail-chick chimeras. II. The prosencephalic neural plate andneural folds: implications for the genesis of cephalic human congenitalabnormalities. Dev. Biol. 120, 198–214.

Couly, G., Le Douarin, N.M., 1990. Head morphogenesis in embryonicavian chimeras: evidence for a segmental pattern in the ectoderm cor-responding to neuromeres. Development 108, 543–558.

Couly, G., Coltey, P., Le Douarin, N.M., 1993a. The developmental fate ofthe cephalic mesoderm in quail-chick chimeras. Development 114, 1–15.

Couly, G.F., Coltey, P.M., Le Douarin, N.M., 1993b. The triple origin ofskull in higher vertebrates: a study in quail-chick chimeras.Development 117, 409–429.

Couly, G., Coltey, P., Eichmann, A., Le Douarin, N.M., 1995. The angio-genic potentials of the cephalic mesoderm and the origin of brain andhead blood vessels. Mech. Dev. 53, 97–112.

Couly, G., Grapin-Botton, A., Coltey, P., Le Douarin, N.M., 1996. Theregeneration of the cephalic neural crest, a problem revisited: the regen-erating cells originate from the contralateral or the anterior and posteriorneural fold. Development 122, 3393–3407.

Crossley, P.H., Martin, G.R., 1995. The mouseFgf8gene encodes a familyof polypeptides and is expressed in regions that direct outgrowth andpatterning in the developing embryo. Development 121, 439–451.

Cserjesi, P., Lilly, B., Bryson, L., Wang, Y., Sassoon, D.A., Olson, E.N.,1992. MHox: a mesodermally restricted homeodomain protein thatbinds an essential site in the muscle creatine kinase enhancer.Development 115, 1087–1101.

Dahl, E., Koseki, H., Balling, R., 1997.Pax genes and organogenesis.BioEssays 19, 755–765.

Davis, C.A., Holmyard, D.P., Millen, K.J., Joyner, A.L., 1991. Examiningpattern formation in mouse, chicken and frog embryos with an En-specific antiserum. Development 111, 287–298.

Dear, T.N., Colledge, W.H., Carlton, M.B.L., Lavenir, I., Larson, T.,Smith, A.J.H., Warren, A.J., Evans, M.J., Sofroniew, M.V., Rabbits,


T.H., 1995. TheHox11gene is essential for cell survival during spleendevelopment. Development 121, 2909–2915.

Dewulf, N., Verschueren, K., Lonnoy, O., Moreń, A., Grimsby, S., Vande-Spiegle, K., Miyazono, K., Huylebroeck, D., ten Dijke, P., 1995. Dis-tinct spatial and temporal expression patterns of two type 1 receptors forbone morphogenetic proteins during mouse embryogenesis.Endocrinology 136, 2652–2663.

Dickinson, M.E., Krumlauf, R., McMahon, A.P., 1994. Evidence for amitogenic effect of Wnt-1 in the developing mammalian central nervoussystem. Development 120, 1453–1471.

Dickson, M.C., Martin, J.S., Cousins, F.M., Kulkarni, A.B., Karlsson, S.,Akhurst, R.J., 1995. Defective haematopoiesis and vasculogenesis intransforming growth factorb1 knockout mice. Development 121,1845–1854.

Dietrich, S., Gruss, P., 1995.undulatedphenotypes suggest a role ofPax-1for the development of vertebral and extravertebral structures. Dev.Biol. 167, 529–548.

Dixon, J., Edwards, S.J., Gladwin, A.J., Loftus, S.K., Bonner, C.A.,Koprivnikar, K., Wasmuth, J.J., Dixon, M.J., 1996. Positional cloningof a gene involved in the pathogenesis of Treacher-Collins syndrome.Nat. Genet. 12, 130–136.

Dolle, P., Price, M., Duboule, D., 1992. Expression of the murineDlx-1homeobox gene during facial, ocular and limb development.Differentiation 49, 93–99.

Domınguez, M., Brunner, M., Hafen, E., Basler, K., 1996. Sending andreceiving the hedgehog signal: control by theDrosophila Gli proteinCubitus interruptus. Science 272, 1621–1625.

Drucker, B.J., Goldfarb, M., 1993. Murine FGF-4 gene expression is spa-tially restricted within embryonic skeletal muscle and other tissues.Mech. Dev. 40, 155–163.

Dudley, A.T., Lyons, K.M., Robertson, E.J., 1995. A requirement for bonemorphogenetic protein-7 during development of the mammalian kidneyand eye. Genes Dev. 9, 2795–2807.

Dunn, N.R., Winnier, G.E., Hargett, L.K., Schrick, J.J., Fogo, A.B.,Hogan, B.L.M., 1997. Haploinsufficient phenotypes inBmp4heterozy-gous null mice and modification by mutations inGli3 and Alx4. Dev.Biol. 188, 235–247.

Duprez, D., Kostakopoulou, K., Francis-West, P.H., Tickle, C., Brickell,P.M., 1996. Activation of Fgf-4 and HoxD gene expression by Bmp-2expressing cells in the developing chick limb. Development 122, 1821–1828.

El Ghouzzi, V., Le Merrer, M., Perrin-Schmitt, F., Lajeunie, E., Benit, P.,Renier, D., Bourgeois, P., Bolcato-Bellemin, A.-L., Munnich, A.,Bonaventure, J., 1997. Mutations of theTWISTgene in the Saethre-Chotzen syndrome. Nat. Genet. 15, 42–46.

Ericson, J., Norlin, S., Jessel, T.M., Edlund, T., 1998. Integrated FGF andBMP signaling controls the progression of progenitor cell differentiationand the emergence of pattern in the embryonic anterior pituitary.Development 125, 1005–1015.

Feijen, A., Goumans, M.J., van den Eijinden-van Raaij, A.J.M., 1994.Expression of activin subunits, activin receptors and follistatin in post-implantation mouse embryos suggests specific developmental functionsfor different activins. Development 120, 3621–3637.

Francis-West, P.H., Tatla, T., Brickell, P.M., 1994. Expression patterns ofthe bone morphogenetic proteinsBmp-4and Bmp-2 in the developingchick face suggest a role in outgrowth of the primordia. Dev. Dyn. 201,168–178.

Fuchtbauer, E.M., 1995. Expression of M-twist during post-implantationdevelopment of the mouse. Dev. Dyn. 204, 316–322.

Fukuda, Y., Hirata, Y., Yoshimi, H., Kojima, T., Kobayashi, Y.,Yanagisawa, M., Masaki, T., 1988. Endothelin is a potent secretagoguefor atrial natriuretic peptide in cultured rat atrial myocytes. Biochem.Biophys. Res. Commun. 155, 167–172.

Gans, C., Northcutt, R.G., 1983. Neural crest and the origin of vertebrates:a new head. Science 220, 268–274.

Gardner, C.A., Darnell, D.K., Poole, S.J., Ordahl, C.P., Barald, K.F.,1988. Expression of an engrailed-like gene during development of the

early embryonic chick nervous system. J. Neurosci. Res. 21, 426–437.

Gaunt, S.J., Blum, M., De Robertis, E.M., 1993. Expression of the mousegoosecoidgene during mid-embryogenesis may mark mesenchymal celllineages in the developing head, limbs and body wall. Development117, 769–778.

Gavin, B.J., McMahon, J.A., McMahon, A.P., 1990. Expression of multi-ple novel Wnt-1/int-1-related genes during fetal and adult mousedevelopment. Genes Dev. 4, 2319–2332.

Gendron-Maguire, M., Mallo, M., Zhang, M., Gridley, T., 1993.Hoxa-2mutant mice exhibit homeotic transformation of skeletal elementsderived from cranial neural crest. Cell 75, 1317–1331.

Gitelman, I., 1997. Twist protein in mouse embryogenesis. Dev. Biol. 189,205–214.

Goodrich, L.V., Johnson, R.L., Milenkovic, L., McMahon, J.A., Scott,M.P., 1996. Conservation of thehedgehog/patchedsignaling pathwayfrom flies to mice: induction of a mousepatchedgene by hedgehog.Genes Dev. 10, 301–312.

Goulding, M.D., Chalepakis, G., Deutsch, U., Erselius, J.R., Gruss, P.,1991. Pax-3, a novel murine DNA binding protein expressed duringearly neurogenesis. EMBO J. 10, 1135–1147.

Green, M.C., 1951. Further morphological effects of the short ear gene inthe house mouse. J. Morphol. 88, 1–21.

Grindley, J.C., Davidson, D.R., Hill, R.E., 1995. The role of Pax-6 in eyeand nasal development. Development 121, 1433–1442.

Hahn, H., Christiansen, J., Wicking, C., Zaphiropoulos, P.G.,Chidambaram, A., Gerrard, B., Vorechovsky, I., Bale, A.E., Toftgard,R., Dean, M., Wainwright, B., 1996a. A mammalianpatchedhomolog isexpressed in target tissues ofsonic hedgehogand maps to a regionassociated with developmental abnormalities. J. Biol. Chem. 271,12125–12128.

Hahn, H., Wicking, C., Zaphiropoulous, P.G., Gailani, M.R., Shanley, S.,Chidambaram, A., Vorechovsky, I., Holmberg, E., Unden, A.B., Gillies,S., Negus, K., Symth, I., Pressman, C., Leffell, D.J., Gerrard, B.,Goldstein, A.M., Dean, M., Toftga˚rd, R., Chenevix-Tren, G.,Wainwright, B., Bale, A.E., 1996b. Mutations of the human homologof Drosophila patched in the nevoid basal cell carcinoma syndrome.Cell 85, 841–845.

Hammerschmidt, M., Brook, A., McMahon, A.P., 1997. The world accord-ing to hedgehog. Trends Genet. 13, 14–21.

Hanken, J., Thorogood, P., 1993. Evolution and development of the verte-brate skull: the role of pattern formation. Trends Ecol. Evol. 8, 9–14.

Hanson, I.M., Fletcher, J.M., Jordon, T., Brown, A., Taylor, D., Adams,R.J., Punnett, H.H., van Heyningen, V., 1994. Mutations at thePAX6locus are found in heterogeneous anterior segment malformationsincluding Peters’ anomaly. Nat. Genet. 6, 168–173.

Hartung, H., Feldman, B., Lovec, H., Coulier, F., Birnbaum, D., Goldfarb,M., 1997. Murine FGF-12 and FGF-13: expression in embryonic ner-vous system, connective tissue and heart. Mech. Dev. 64, 31–39.

Haub, O., Goldfarb, M., 1991. Expression of the fibroblast growth factor-5gene in the mouse embryo. Development 112, 397–406.

Hebert, J.M., Basilico, C., Goldfarb, M., Haub, O., Martin, G.R., 1990.Isolation of cDNAs encoding four mouse FGF family members andcharacterisation of their expression patterns during embryogenesis.Dev. Biol. 138, 454–463.

Heikinheimo, M., Lawshe´, A., Shackleford, G.M., Wilson, D.B.,MacArthur, C.A., 1994.Fgf-8 expression in the post-gastrulationmouse suggests roles in the development of the face, limbs and centralnervous system. Mech. Dev. 48, 129–138.

Helms, J.A., Kim, C.H., Hu, D., Minkoff, R., Thaller, C., Eichele, G.,1997. Sonic hedgehog participates in craniofacial morphogenesis andis down-regulated by teratogenic doses of retinoic acid. Dev. Biol. 187,25–35.

Henderson, D.J., Ybot-Gonzalez, P., Copp, A.J., 1997. Over-expression ofthe chondroitin sulphate proteoglycanversicanis associated with defec-tive neural crest migration in thePax3mutant mouse (splotch). Mech.Dev. 69, 39–51.


Hill, R.E., Jones, P.F., Rees, A.R., Sime, C.M., Justice, M.J., Copeland,N.G., Jenkins, N.A., Graham, E., Davidson, D.R., 1989. A new familyof mouse homoebox-containing genes: molecular structure, chromosomallocation, and developmental expression of Hox-7.1. Genes Dev. 3, 26–37.

Hill, R.E., Favor, J., Hogan, B.L.M., Ton, C.C.T., Saunders, G.F., Hanson,I.M., Prosser, J., Jordan, T., Hastie, N.D., van Heyningen, V., 1991.Mouse Small eyeresults from mutations in a paired-like homeobox-containing gene. Nature 354, 522–525.

Hinck, L., Nelson, W.J., Papkoff, J., 1994. Wnt-1 modulates cell-celladhesion in mammalian cells by stabilizing B-catenin binding to thecell adhesion protein cadherin. J. Cell. Biol. 124, 729–741.

Hogan, B.L.M., 1996. Bone morphogenetic proteins: multifunctional reg-ulators of vertebrate development. Genes Dev. 10, 1580–1594.

Hollyday, M., McMahon, J.A., McMahon, A.P., 1995.Wnt expressionpatterns in chick embryo nervous system. Mech. Dev. 52, 9–25.

Holst, B.D., Wang, Y., Jones, F.S., Edelman, G.M., 1997. A binding sitefor Pax proteins regulates expression of the gene for the neural celladhesion molecule in the embryonic spinal cord. Proc. Natl. Acad.Sci. USA 94, 1465–1470.

Horstadius, S., Sellman, S., 1946. Experimentalle Untersuchungen uber diedetermination des Knorpeligen Kopfskelettes bei Urodelen. Nova ActaRegiae Soc. Sci. Upsaliensis (ser. 1V) 13, 1–170.

Hoth, C.F., Milunsky, A., Lipsky, N., Sheffer, R., Clarren, S.K., Baldwin,C.T., 1993. Mutations in the paired domain of the human PAX3 genecauses Klein-Waardenburg syndrome (WS-III) as well as Waardenburgsyndrome type I (WS-I). Am. J. Hum. Genet. 52, 455–462.

Howard, T.D., Paznekas, W.A., Green, E.D., Chiang, L.C., Ma, N., DeLuna, R.I.O., Delgado, C.G., Gonzalez-Ramos, M., Kline, A.D., Jabs,E.W., 1997. Mutations inTWIST, a basic helix-loop-helix transcriptionfactor, in Saethre-Chotzen syndrome. Nat. Genet. 15, 36–41.

Hui, C.-C., Slusarski, D., Platt, K.A., Holmgren, R., Joyner, A.L., 1994.Expression of three mouse homologs of theDrosophilasegment polaritygenecubitus interruptus, Gli, Gli-2, andGli-3, in ectoderm and meso-derm-derived tissues suggests multiple roles during postimplantationdevelopment. Dev. Biol. 162, 402–413.

Hukriede, N.A., Gu, Y., Fleming, R.J., 1997. A dominant-negative form ofSerrateacts as a general antagonist ofNotch activation. Development124, 3427–3437.

Hunt, P., Whiting, J., Muchamore, I., Marshall, H., Krumlauf, R., 1991a.Homeobox genes and models for patterning the hindbrain and branchialarches. Development Suppl. 1, 187–196.

Hunt, P., Gulisano, M., Cook, M., Sham, M.-H., Faiella, A., Wilkinson, D.,Boncinelli, E., Krumlauf, R., 1991b. A distinctHox code for the bran-chial region of the vertebrate head. Nature 353, 861–864.

Hunt, P., Wilkinson, D., Krumlauf, R., 1991c. Patterning the vertebratehead: murine Hox 2 genes mark distinct subpopulations of premigratoryand migrating cranial neural crest. Development 112, 43–50

Ingham, P.W., 1998. Thepatchedgene in development and cancer. Curr.Opin. Genet. Dev. 8, 88–94.

Iulianella, A., Lohnes, D., 1997. Contribution of retinoic acid receptorgamma to retinoid-induced craniofacial and axial defects. Dev. Dyn.209, 92–104.

Jabs, E.W., Muller, U., Li, X., Ma, L., Luo, W., Haworth, I.S., Klisak, I.,Sparkes, R., Warman, M.L., Mulliken, J.B., Snead, M.L., Maxson, R.,1993. A mutation in the homeodomain of the human MSX2 gene in afamily affected with autosomal dominant craniosynostosis. Cell 75,443–450.

Jabs, E.W., Li, X., Scott, A.F., Meyers, G., Chen, W., Eccles, M., Mao, J.,Charnas, L.R., Jackson, C.E., Jaye, M., 1994. Jackson-Weiss and Crou-zon syndromes are allelic with mutations in fibroblast growth factorreceptor 2. Nat. Genet. 8, 275–279.

Jiang, R., Lan, Y., Chapman, H.D., Shawber, C., Norton, C.R., Serreze,D.V., Weinmaster, G., Gridley, T., 1998. Defects in limb, craniofacial,and thymic development in Jagged2 mutant mice. Genes Dev. 12,1046–1057.

Johnson, R.L., Rothman, A.L., Xie, J., Goodrich, L.V., Bare, J.W.,Bonifas, J.M., Quinn, A.G., Myers, R.M., Cox, D.R., Epstein, E.H.,

Scott, M.P., 1996. Human homolog of patched, a candidate gene forthe basal nevus syndrome. Science 272, 1668–1671.

Johnson, R.L., Tabin, C.J., 1997. Molecular models for vertebrate limbdevelopment. Cell 90, 979–990.

Johnston, M.C., 1966. A radioautographic study of the migration and fateof cranial neural crest cells in the chick embryo. Anat. Rec. 156, 143–155.

Johnston, M.C., Bronsky, P.T., 1995. Prenatal craniofacial development:new insights on normal and abnormal mechanisms. Crit. Rev. Oral Biol.Med. 6, 368–422.

Jones, F.S., Kioussi, C., Copertino, D.W., Kallunki, P., Holst, B.D.,Edelman, G.M., 1997.Barx2, a new homeobox gene of theBar classis expressed in neural and craniofacial structures during development.Proc. Natl. Acad. Sci. USA 94, 2632–2637.

Jostes, B., Walther, C., Gruss, P., 1991. The murine paired box gene,Pax7,is expressed specifically during the development of the nervous andmuscular system. Mech. Dev. 33, 27–38.

Kaartinen, V., Cui, X.-M., Heisterkamp, N., Groffen, J., Shuler, C.F.,1997. Transforming growth factor-b3 regulates transdifferentiation ofmedial edge epithelium during palatal fusion and associated degradationof the basement. Dev. Dyn. 209, 255–260.

Kaartinen, V., Voncken, J.W., Shuler, C., Warburton, D., Bu, D.,Heisterkamp, N., Groffen, J., 1995. Abnormal lung development andcleft palate in mice lacking TGF-b3 indicates defects of epithelial-mesenchymal interaction. Nat. Genet. 11, 415–421.

Kang, S., Graham, J.M., Olney, A.H., Biesecker, L.G., 1997. GLI3 frame-shift mutations cause autosomal dominant Pallister-Hall syndrome. Nat.Genet. 15, 266–268.

Kaufman, M.H., Chang, H.-H., Shaw, J.P., 1995. Craniofacial abnormal-ities in homozygousSmall eye(Sey/Sey) embryos and new-born mice. J.Anat. 186, 607–617.

King, J.A., Marker, P.C., Seung, K.J., Kingsley, D.M., 1994. BMP5 andthe molecular, skeletal, and soft-tissue alterations in short ear mice. Dev.Biol. 166, 112–122.

Kingsley, D.M., 1994. The TGF-b superfamily: new members, new recep-tors, and new genetic tests of function in different organisms. GenesDev. 8, 133–146.

Kingsley, D.M., Bland, A.E., Grubber, J.M., Marker, P.C., Russell, L.B.,Copeland, N.G., Jenkins, N.A., 1992. The mouse short ear skeletalmorphogenesis locus is associated with defects in a bone morphogeneticmember of the TGFb superfamily. Cell 71, 399–410.

Kitamura, K., Miura, H., Yanazawa, M., Miyashita, T., Kato, K., 1997.Expression patterns ofBrx1 (Reiggene),Sonic hedgehog, Nkx2.2, Dlx1and Arx during zona limitans intrathalamica and embryonic ventrallateral geniculate nuclear formation. Mech. Dev. 67, 83–96.

Kontges, G., Lumsden, A., 1996. Rhombencephalic neural crest segmenta-tion is preserved throughout craniofacial ontogeny. Development 122,3229–3242.

Kozmik, Z., Wang, S., Dorfler, P., Adams, B., Busslinger, M., 1992. Thepromoter of the CD19 gene is a target for the B-cell-specific transcrip-tion factor BSAP. Mol. Cell. Biol. 12, 2662–2672.

Kronmiller, J.E., Upholt, W.B., Kollar, E.J., 1991. Expression of epidermalgrowth factor mRNA in the developing mouse mandibular process.Arch. Oral Biol. 36, 405–410.

Krumlauf, R., 1994a.Hox genes in vertebrate development. Cell 78, 191–201.

Krumlauf, R., 1994b. Hox genes and pattern formation in the branchialregion of the vertebrate head. Trends Genet. 9, 106–112.

Kuratani, S., Martin, J.F., Wawersik, S., Lilly, B., Eichele, G., Olson, E.N.,1994. The expression pattern of the chick homeobox genegMHoxsug-gests a role in patterning of the limbs and face and in compartmentali-zation of somites. Dev. Biol. 161, 357–369.

Kuratani, S., Matsuo, I., Aizawa, S., 1997. Developmental Patterning andEvolution of the mammalian viscerocranium: genetic insights into com-parative morphology. Dev. Dyn. 209, 139–155.

Kurihara, Y., Kurihara, H., Suzuki, H., Kodama, T., Maemura, K., Nagai,R., Oda, H., Kuwaki, T., Cao, W.-H., Kamada, N., Jishage, K., Ouchi,


Y., Azumi, S., Toyoda, Y., Ishikawa, T., Kumada, M., Yazaki, Y., 1994.Elevated blood pressure and craniofacial abnormalities in mice deficientin endothelin-1. Nature 368, 703–710.

Lacombe, D., Toutain, A., Gorlin, R.J., Oley, C.A., Battin, J., 1994. Clin-ical identification of a human equivalent to the short ear (se) murinephenotype. Ann. Genet. 37, 181–191.

Lahav, R., Ziller, C., Dupin, E., Le Douarin, N.M., 1996. Endothelin-3promotes neural crest cell proliferation and mediates a vast increase inmelanocyte number in culture. Proc. Natl. Acad. Sci. USA 93, 3892–3897.

Lanctot, C., Lamolet, B., Drouin, J., 1997. Thebicoid-related homeopro-tein Ptx1 defines the most anterior domain of the embryo and differ-entiates posterior from anterior lateral mesoderm. Development 124,2807–2817.

Le Douarin, N.M., Ziller, C., Couly, G.F., 1993. Patterning of neural crestderivatives in the avian embryo: in vivo and in vitro studies. Dev. Biol.159, 24–49.

Lee, D.C., Rochford, R., Todaro, G.J., Villarreal, L.P., 1985. Develop-mental expression of rat transforming growth factor-a mRNA. Mol.Cell. Biol. 5, 3644–3646.

Lee, Y.M., Osumi-Yamashita, N., Ninomiya, Y., Moon, C.K., Eriksson,U., Eto, K., 1995. Retinoic acid stage-dependently alters the migrationpattern and identity of hindbrain neural crest cells. Development 121,825–837.

Lee, J., Platt, K.A., Censullo, P., Ruiz i Altaba, A., 1997. Gli1 is a target ofSonic hedgehog that induces ventral neural tube development.Development 124, 2537–2552.

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

Le Lievre, C.S., Le Douarin, N.M., 1975. Mesenchymal derivatives of theneural crest: analysis of chimeric quail and chick embryos. J. Embryol.Exp. Morphol. 34, 125–154.

Li, L., Krantz, I.D., Deng, Y., Genin, A., Banta, A.B., Collins, C.C., Qi,M., Trask, B.J., Kuo, W.L., Cochran, J., Costa, T., Pierpont, M.E., Rand,E.B., Piccoli, D.A., Hood, L., Spinner, N.B., 1997. Alagille syndrome iscaused by mutations in humanJagged1, which encodes a ligand forNotch1. Nat. Genet. 16, 243–251.

Lillie, F.R., 1952. Development of the chick. H.L. Hamilton (Ed.), Holt,New York.

Liu, Y.H., Kundu, R., Wu, L., Luo, W., Ignelzi, M.A., Snead, M.L.,Maxson, R.E., 1995. Premature suture closure and ectopic cranialbone in mice expressingMsx2 transgenes in the developing skull.Proc. Natl. Acad. Sci. USA 92, 6137–6141.

Liu, Y.H., Ma, L., Kundu, R., Ignelzi, M., Sangiorgi, F., Wu, L., Luo, W.,Snead, M.L., Maxson, R., 1996. Function of theMsx2 gene in themorphogenesis of the skull. Ann. N Y Acad. Sci. 785, 48–58.

Logan, C., Khoo, W.K., Cado, D., Joyner, A.L., 1993. Two enhancerregions in the mouseEn-2 locus direct expression to the mid/hindbrainregion and mandibular myoblasts. Development 117, 905–916.

Lohnes, D., Mark, M., Mendelsohn, C., Dolle´, P., Dierich, A., Gorry, P.,Gansmuller, A., Chambon, P., 1994. Function of the retinoic acid recep-tors (RARs) during development (I) Craniofacial and skeletal abnorm-alities in RAR double mutants. Development 120, 2723–2748.

Lumsden, A., Sprawson, N., Graham, A., 1991. Segmental origin andmigration of neural crest cells in the hindbrain region of the chickembryo. Development 113, 1281–1291.

Luo, G., Hofmann, C., Bronckers, A.L.J.J., Sohocki, M., Bradley, A.,Karsenty, G., 1995. BMP-7 is an inducer of nephrogenesis and is alsorequired for eye development and skeletal patterning. Genes Dev. 9,2808–2820.

Lyons, K.M., Hogan, B.L.M., Robertson, E.J., 1995. Colocalization ofBMP 7 and BMP 2 RNAs suggests that these factors cooperativelymediate tissue interactions during murine development. Mechs. Dev.50, 71–83.

Ma, L., Golden, S., Wu, L., Maxson, R., 1996. The molecular basis ofBoston-type craniosynostosis: the Pro148-His mutation in the N-term-inal arm of theMSX2homeodomain stabilizes DNA binding without

altering nucleotide sequence preferences. Hum. Mol. Genet. 5, 1915–1920.

Mackenzie, A., Ferguson, M.J.W., Sharpe, P.T., 1991a.Hox-7 expressionduring murine craniofacial development. Development 113, 601–611.

Mackenzie, A., Leeming, G.L., Jowett, A.K., Ferguson, M.J.W., Sharpe,P.T., 1991b. The homeobox gene Hox 7.1 has specific regional andtemporal expression patterns during early murine craniofacial embryo-genesis, especially tooth development in vivo and in vitro. Development111, 269–285.

Maemura, K., Kurihara, H., Kurihara, Y., Oda, H., Ishikawa, T., Copeland,N.G., Gilbert, D.J., Jenkins, N.A., Yazaki, Y., 1996. Sequence analysis,chromosomal location, and developmental expression of the mouse pre-proendothelin-1 gene. Genomics 31, 177–184.

Mansouri, A., Stoykova, A., Torres, M., Gruss, P., 1996. Dysgenesis ofcephalic neural crest derivatives inPax7− / − mutant mice. Development122, 831–838.

Mansouri, A., Yokota, Y., Wehr, R., Copeland, N.G., Jenkins, N.A., Gruss,P., 1997. Paired-related murine homeobox gene expressed in the devel-oping sclerotome, kidney and nervous system. Dev. Dyn. 210, 53–65.

Marigo, V., Johnson, R.L., Vortkamp, A., Tabin, C.J., 1996. Sonic hedge-hog differentially regulates expression ofGLI and GLI3 during limbdevelopment. Dev. Biol. 180, 273–283.

Martı, E., Takada, R., Bumcrot, D.A., Sasaki, H., McMahon, A.P., 1995.Distribution of Sonic hedgehog peptides in the developing chick andmouse embryo. Development 121, 2537–2547.

Martin, J.F., Bradley, A., Olson, E.N., 1995. Thepaired-like homeoboxgeneMHox is required for early events of skeletogenesis in multiplelineages. Genes Dev. 9, 1237–1249.

Masuya, H., Sagai, T., Wakana, S., Moriwaki, K., Shiroisha, T., 1995. Aduplicated zone of polarizing activity in polydactylous mouse mutants.Genes Dev. 9, 1645–1653.

Matovinovic, E., Richman, J.M., 1997. Epithelium is required for main-taining FGFR-2 expression levels in facial mesenchyme of the devel-oping chick embryo. Dev. Dyn. 210, 407–416.

Matsui, Y., Zsebo, K.M., Hogan, B.L.M., 1990. Embryonic expression of ahaematopoietic growth factor encoded by theSl locus and the ligand forc-kit. Nature 347, 667–669.

Matsuo, I., Kuratani, S., Kimura, C., Takeda, N., Aizawa, S., 1995. MouseOtx2 functions in the formation and patterning of rostral head. GenesDev. 9, 2646–2658.

Matsuo, T., Osumi-Yamashita, N., Noji, S., Ohuchi, H., Koyama, E.,Myokai, F., Matsuo, N., Taniguchi, S., Doi, H., Iseki, S., Ninomiya,Y., Fujiwara, M., Watanabe, T., Eto, K., 1993. A mutation in the Pax-6gene in ratsmalleye is associated with impaired migration of midbraincrest cells. Nat. Genet. 3, 299–304.

Matzuk, M.M., Kumar, T.R., Vassalli, A., Bickenbach, J.R., Roop, D.R.,Jaenisch, R., Bradley, A., 1995a. Functional analysis of activins duringmammalian development. Nature 374, 354–356.

Matzuk, M.M., Lu, N., Vogel, H., Sellheyer, K., Roop, D.R., Bradley, A.,1995b. Multiple defects and perinatal death in mice deficient in follis-tatin. Nature 374, 360–363.

Matzuk, M.M., Kumar, T.R., Bradley, A., 1995c. Different phenotypes formice deficient in either activins or activin receptor type II. Nature 374,356–360.

McMahon, A.P., 1992. TheWnt family of developmental regulators.Trends Genet. 8, 236–242.

Meyers, G.A., Orlow, S.J., Munro, I.R., Przylepa, K.A., Jabs, E.W., 1995.Fibroblast growth factor receptor 3 (FGFR3) transmembrane mutationin Crouzon syndrome with acanthosis nigricans. Nat. Genet. 11, 462–464.

Michaud, J.L., Lapointe, F., Le Douarin, N.M., 1997. The dorsoventralpolarity of the presumptive limb is determined by signals produced bythe somites and by the lateral somatopleure. Development 124, 1453–1463.

Millan, F.A., Denhez, F., Kondaiah, P., Akhurst, R.J., 1991. Embryonicgene expression patterns of TGF-b1, b2, and b3, suggest differentdevelopmental functions in vivo. Development 111, 131–143.


Millen, K.J., Wurst, W., Herrup, K., Joyner, A.L., 1994. Abnormalembryonic cerebellar development and patterning of postnatal foliationin two mouseEngrailed-2mutants. Development 120, 695–706.

Mina, M., Gluhak, J., Upholt, W.B., Kollar, E.J., Rogers, B., 1995. Experi-mental analysis ofMsx-1andMsx-2gene expression during chick man-dibular morphogenesis. Dev. Dyn. 202, 195–214.

Mitsiadis, T.A., Henrique, D., Thesleff, I., Lendahl, U., 1997. MouseSerrate-1 (Jagged -1): expression in the developing tooth is regulatedby epithelial-mesenchymal interactions and fibroblast growth factor-4.Development 124, 1473–1483.

Mo, R., Freer, A.M., Zinyk, D.L., Crackower, M.A., Michaud, J., Heng,H.H.-Q., Chik, K.W., Shi, X.-M., Tsui, L.-C., Cheng, S.H., Joyner, A.L.,Hui, C.-C., 1997. Specific and redundant functions ofGli2 andGli3 zincfinger genes in skeletal patterning and development. Development 124,113–123.

Moase, C.E., Trasler, D.G., 1991. N-CAM alterations in splotch neuraltube defect mouse embryos. Development 113, 1049–1058.

Moon, R.T., DeMarais, A., Olson, D.J., 1993. Responses to Wnt signals invertebrate embryos may involve changes in cell adhesion and cellmovement. J. Cell. Sci. Suppl. 17, 183–188.

Morell, R., Friedman, T.B., Moeljopawiro, S., Hartono, Soewito andAsher, J.H., 1992. A frameshift mutation in the HuP2 paired domainof the probable human homolog of murinePax-3 is responsible forWaardenberg syndrome type 1 in an Indonesian family. Hum. Mol.Genet. 1, 243–247.

Morrison-Graham, K., Schatteman, G.C., Bork, T., Bowen-Pope, D.F.,Weston, J.A., 1992. A PDGF receptor mutation in the mouse (Patch)perturbs the development of a non-neuronal subset of neural crest-derived cells. Development 115, 133–142.

Morriss-Kay, G., 1993. Retinoic acid and craniofacial development: mole-cules and morphogenesis. BioEssays 15, 9–15.

Morriss-Kay, G.M., Sokolova, N., 1996. Embryonic development and pat-tern formation. FASEB J. 10, 961–968.

Morriss-Kay, G.M., 1996. Craniofacial defects in AP-2 null mutant mice.BioEssays 18, 785–788.

Moses, H.L., Serra, R., 1996. Regulation of differentiation by TGF-b.Curr. Opin. Genet. Dev. 6, 581–586.

Mucchielli, M.L., Mitsiadis, T.A., Raffo, S., Brunet, J.-F., Proust, J.-P.,Goridis, C., 1997. Mouse Otlx2/RIEG expression in the odontogenicepithelium precedes tooth initiation and requires mesenchyme-derivedsignals for its maintenance. Dev. Biol. 189, 274–284.

Muenke, M., Schell, U., 1995. Fibroblast-growth-factor receptor mutationsin human skeletal disorders. Trends Genet. 11, 308–313.

Munsterberg, A.E., Kitajewski, J., Bumcrot, D.A., McMahon, A.P., Lassar,A.B., 1995. Combinatorial signaling by Sonic hedgehog and Wnt familymembers induces myogenic bHLH gene expression in the somite. GenesDev. 9, 2911–2922.

Nakamura, S., Stock, D.W., Wydner, K.L., Bollekens, J.A., Takeshita, K.,Nagai, B.M., Chiba, S., Kitamura, T., Freeland, T.M., Minowada, J.,Lawrence, J.B., Weiss, K.M., Ruddle, F.H., 1996. Genomic analysis of anew mammalian distal-less gene: Dlx7. Genomics 38, 314–324.

Nelson, L.B., Spaeth, G.L., Nowinski, T.S., Margo, C.E., Jackson, L.,1984. Aniridia. A review. Surv. Ophthalmol. 28, 621–642.

Neubuser, A., Koseki, H., Balling, R., 1995. Characterization and devel-opmental expression ofPax9, a paired-box-containing gene related toPax1. Dev. Biol. 170, 701–716.

Neubuser, A., Peters, H., Balling, R., Martin, G.R., 1997. Antagonisticinteractions between FGF and BMP signalling pathways: a mechanismfor positioning the sites of tooth formation. Cell 90, 247–255.

Noden, D.M., 1975. An analysis of migratory behaviour of avian cephalicneural crest cells. Dev. Biol. 42, 106–130.

Noden, D.M., 1978. The control of avian cephalic neural crest cytodiffer-entiation. I. Skeletal and connective tissues. Dev. Biol. 67, 296–312.

Noden, D.M., 1983. The role of the neural crest in patterning of aviancranial skeletal, connective and muscle tissues. Dev. Biol. 96, 144–165.

Noden, D.M., 1986. Patterning of avian craniofacial muscles. Dev. Biol.116, 347–356.

Noden, D.M., 1988. Interactions and fates of avian craniofacial mesench-yme. Development 103 Suppl.,121–140.

Nohno, T., Koyama, E., Myokai, F., Taniguchi, S., Ohuchi, H., Saito, T.,Noji, S., 1993. A chicken homeobox gene related toDrosophila pairedis predominantly expressed in the developing limb. Dev. Biol. 158,254–264.

Oda, T., Elkahloun, A.G., Pike, B.L., Okajima, K., Krantz, I.D., Genin, A.,Piccoli, D.A., Meltzer, P.S., Spinner, N.B., Collins, F.S.,Chandrasekharappa, S.C., 1997. Mutations in the humanJagged1gene are responsible for Alagille syndrome. Nat. Genet. 16, 235–242.

Olson, D.J., Christian, J.L., Moon, R.T., 1991. Effect of wnt-1 and relatedproteins on gap junctional communication inXenopusembryos. Science252, 1173–1176.

Oosterwegel, M., van de Wetering, M., Timmerman, J., Kruisbeek, A.,Destree, O., Meijlink, F., Clevers, H., 1993. Differential expression ofthe HMG box factorsTCF-1 andLEF-1 during murine embryogenesis.Development 118, 439–448.

Opitz, J.M., Gilbert, E.F., 1982. CNS anomalies and the midline as a‘developmental field’. Am. J. Med. Genet. 12, 443–455.

Opstelten, D.-J.E., Vogels, R., Robert, B., Kalkhoven, E., Zwartkruis, F.,de Laaf, L., Destree, O.H., Deschamps, J., Lawson, K.A., Meijlink, F.,1991. The mouse homeobox gene,S8, is expressed during embryogen-esis predominantly in mesenchyme. Mech. Dev. 34, 29–42.

Orr-Urtreger, A., Avivi, A., Zimmer, Y., Givol, D., Yarden, Y., Lonai, P.,1990. Developmental expression ofc-kit, a proto-oncogene encoded bythe W locus. Development 109, 911–923.

Orr-Urtreger, A., Lonai, P., 1992. Platelet-derived growth factor-A and itsreceptor are expressed in separate, but adjacent cell layers of the mouseembryo. Development 115, 1045–1058.

Osumi-Yamashita, N., Ninomiya, Y., Doi, H., Eto, K., 1994. The contri-bution of both forebrain and midbrain crest cells to the mesenchyme inthe frontonasal mass of mouse embryos. Dev. Biol. 164, 409–419.

Osumi-Yamashita, N., Kuratani, S., Ninomiya, Y., Aoki, K., Iseki, S.,Chareonvit, S., Doi, H., Fujiwara, M., Watanabe, T., Eto, K., 1997.Cranial anomaly of homozygous rSey rat is associated with a defectin the migration pathway of midbrain crest cells. Dev. Growth Diff. 39,53–67.

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

Patstone, G., Pasquale, E.B., Maher, P.A., 1993. Different members of thefibroblast growth factor receptor family are specific to distinct cell typesin the developing chicken embryo. Dev. Biol. 155, 107–123.

Pelton, R.W., Nomura, S., Moses, H.L., Hogan, B.L.M., 1989. Expressionof transforming growth factorb2 RNA during murine embryogenesis.Development 106, 759–767.

Pelton, R.W., Saxena, B., Jones, M., Moses, H.L., Gold, L.I., 1991. Immu-nohistochemical localization of TGF-b1,TGF-b2, and TGF-b3, in themouse embryo: expression patterns suggest multiple roles duringembryonic development. J. Cell. Biol. 115, 1091–1105.

Peters, K.G., Werner, S., Chen, G., Williams, L.T., 1992. Two FGF recep-tor genes are differentially expressed in epithelial and mesenchymaltissues during limb formation and organogenesis in the mouse.Development 114, 233–243.

Piotrowski, 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., Nu¨sslein-Volhard, C.,1996. Jaw and branchial arch mutants in zebrafish II: anterior arches andcartilage differentiation. Development 123, 345–356.

Platt, K.A., Michaud, J., Joyner, A.L., 1997. Expression of the mouseGliand Ptc genes is adjacent to embryonic sources of hedgehog signalssuggesting a conservation of pathways between flies and mice. Mech.Dev. 62, 121–135.

Pratt, R.M., 1987. Role of epidermal growth factor in embryonicdevelopment. Curr. Top. Dev. Biol. 22, 175–193.


Prince, V., Lumsden, A., 1994. Hoxa-2 expression in normal and trans-posed rhombomeres: independent regulation in the neural tube andneural crest. Development 120, 911–923.

Proetzel, G., Pawlowski, S.A., Wiles, M.V., Yin, M., Boivin, G.P.,Howles, P.N., Ding, J., Ferguson, M.W.J., Doetschman, T., 1995. Trans-forming growth factor-b3 is required for secondary palate fusion. Nat.Genet. 11, 409–414.

Qiu, M., Bulfone, A., Martinez, S., Meneses, J.J., Shimamura, K.,Pedersen, R.A., Rubenstein, J.L.R., 1995. Null mutation ofDlx-2 resultsin abnormal morphogenesis of proximal first and second branchial archderivatives and abnormal differentiation in the forebrain. Genes Dev. 9,2523–2538.

Qiu, M., Bulfone, A., Ghattas, I., Meneses, J.J., Christensen, L., Sharpe,P.T., Presley, R., Pedersen, R.A., Rubenstein, J.L.R., 1997. Role of theDlx homeobox genes in proximodistal patterning of the branchialarches: mutations of Dlx-1, Dlx-2, Dlx-1 and Dlx-2 alter morphogenesisof proximal skeletal and soft tissue structures derived from the first andsecond arches. Dev. Biol. 185, 165–184.

Quinn, J.C., West, J.D., Hill, R.E., 1996. Multiple functions forPax6 inmouse eye and nasal development. Genes Dev. 10, 435–446.

Raju, K., Tang, S., Dube, I.D., Kamel-Reid, S., Bryce, D.M., Breitman,M.L., 1993. Characterization and developmental expression ofTlx-1,the murine homolog ofHOX11. Mech. Dev. 44, 51–64.

Reardon, W., Winter, R.M., Rutland, P., Pulleyn, L.J., Jones, B.M.,Malcolm, S., 1994. Mutations in the fibroblast growth factor receptor2 gene cause Crouzon syndrome. Nat. Genet. 8, 98–103.

Richman, J.M., Tickle, C., 1989. Epithelia are interchangeable betweenfacial primordia of chick embryos and morphogenesis is controlled bythe mesenchyme. Dev. Biol. 136, 201–210.

Richman, J.M., Crosby, Z., 1990. Differential growth of facial primordiain chick embryos: responses of facial mesenchyme to basic fibroblastgrowth factor (bFGF) and serum in micromass culture. Development109, 341–348.

Richman, J.M., Tickle, C., 1992. Epithelial-mesenchymal interactions inthe outgrowth of limb buds and facial primordia in chick embryos. Dev.Biol. 154, 299–308.

Richman, J.M., Herbert, M., Matovinovic, E., Walin, J., 1997. Effect offibroblast growth factors on outgrowth of facial mesenchyme. Dev. Biol.189, 135–147.

Rijli, F.M., Mark, M., Lakkaraju, S., Dierich, A., Dolle, P., Chambon, P.,1993. A homeotic transformation is generated in the rostral branchialregion of the head by disruption ofHoxa-2, which acts as a selectorgene. Cell 75, 1333–1349.

Rivera-Perez, J.A., Mallo, M., Gendron-Maguire, M., Gridley, T.,Behringer, R.R., 1995.goosecoidis not an essential component of themouse gastrula organizer but is required for craniofacial and ribdevelopment. Development 121, 3005–3012.

Robert, B., Sassoon, D., Jacq, B., Gehring, W., Buckingham, M., 1989.Hox-7, a mouse homeobox gene with a novel pattern of expressionduring embryogenesis. EMBO J. 8, 91–100.

Roberts, C.W.M., Shutter, J.R., Korsmeyer, S.J., 1994.Hox11controls thegenesis of the spleen. Nature 368, 747–749.

Roberts, V.J., Barth, S.L., 1994. Expression of messenger ribonucleicacids encoding the inhibin/activin system during mid- and late-gestationrat embryogenesis. Endocrinology 134, 914–923.

Robinson, G.W., Mahon, K.A., 1994. Differential and overlapping expres-sion domains ofDlx-2 and Dlx-3 suggest distinct roles forDistal-lesshomeobox genes in craniofacial development. Mech. Dev. 48, 199–215.

Roelen, B.A.J., van Rooijen, M.A., Mummery, C.L., 1997. Expression ofALK-1, a type 1 serine/threonine kinase receptor, coincides with sites ofvasculogenesis and angiogenesis in early mouse development. Dev.Dyn. 209, 418–430.

Roessler, E., Belloni, E., Gaudenz, K., Jay, P., Berta, P., Scherer, S.W.,Tsui, L.-C., Muenke, M., 1996. Mutations in the humanSonic Hedge-hog gene cause holoprosencephaly. Nat. Genet. 14, 357–360.

Rudnicki, J.A., Brown, A.M.C., 1997. Inhibition of chondrogenesis byWntgene expression in vivo and in vitro. Dev. Biol. 185, 104–118.

Ruiz i Altaba, A., 1997. Catching a glimpse of hedgehog. Cell 90, 193–196.

Ruppert, J.M., Kinzler, K.W., Wong, A.J., Bigner, S.H., Kao, F.T., Law,M.L., Seuanez, H.N., O’Brien, S.J., Vogelstein, B., 1988. The GLI-Kruppel family of human genes. Mol. Cell. Biol. 8, 3104–3113.

Saldivar, J.R., Sechrist, J.W., Krull, C.E., Ruffins, S., Bronner-Fraser, M.,1997. Dorsal hindbrain ablation results in rerouting of neural crestmigration and changes in gene expression, but normal hyoiddevelopment. Development 124, 2729–2739.

Sanford, L.P., Ormsby, I., Gittenberger-de Goot, A.C., Sariola, H.,Friedman, R., Boivin, G.P., Cardell, E.L., Doetschman, T., 1997.TGFb2 knockout mice have multiple developmental defects that arenon-overlapping with other TGFb knockout phenotypes. Development124, 2659–2670.

Satokata, I., Maas, R., 1994.Msx1deficient mice exhibit cleft palate andabnormalities of craniofacial and tooth development. Nat. Genet. 6,348–355.

Schatteman, G.C., Morrison-Graham, K., van Koppen, A., Weston, J.A.,Bowen-Pope, D.F., 1992. Regulation and role of PDGF receptora-subunit expression during embryogenesis. Development 115, 123–131.

Scherson, T., Serbedzija, G., Fraser, S., Bronner-Fraser, M., 1993. Regu-lative capacity of the cranial neural tube to form neural crest.Development 118, 1049–1061.

Schilling, T.F., Walker, C., Kimmel, C.B., 1996a. Thechinlessmutationand neural crest interactions in zebrafish jaw development. Develop-ment 122, 1417–1426.

Schilling, 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., Nu¨sslein-Vol-hard, C., 1996b. Jaw and branchial arch mutants in zebrafish 1: bran-chial arches. Development 123, 329–344.

Schorle, H., Meier, P., Buchert, M., Jaenisch, R., Mitchell, P.J., 1996.Transcription factor AP-2 essential for cranial closure and craniofacialdevelopment. Nature 381, 235–238.

Semina, E.V., Reiter, R., Leysens, N.J., Alward, L.M., Small, K.W.,Datson, N.A., Siegel-Bartelt, J., Bierke-Nelson, D., Bitoun, P., Zabel,B.U., Carey, J.C., Murray, J.C., 1996. Cloning and characterization of anovel bicoid-related homeobox transcription factor gene,RIEG,involved in Rieger syndrome. Nat. Genet. 14, 392–399.

Shen, H., Wilkie, T., Ashique, A.M., Narvey, M., Zerucha, T., Savino, E.,Williams, T., Richman, J.M., 1997. Chicken transcription factor AP-2:cloning, expression and its role in outgrowth of facial prominences andlimb buds. Dev. Biol. 188, 248–266.

Shum, L., Sakakura, Y., Bringas, P., Luo, W., Snead, M.L., Mayo, M.,Crohin, C., Millar, S., Werb, Z., Buckley, S., Hall, F.L., Warburton, D.,Slavkin, H.C., 1993. EGF abrogation-inducedfusilli-form dysmorpho-genesis of Meckel’s cartilage during embryonic mouse mandibular mor-phogenesis in vitro. Development 118, 903–917.

Sidow, A., Bulotsky, M.S., Kerrebrock, A.W., Bronson, R.T., Daly, M.J.,Reeve, M.P., Hawkins, T.L., Birren, B.W., Jaenisch, R., Lander, E.S.,1997. Serrate2 is disrupted in the mouse limb-development mutantsyndactylism. Nature 389, 722–725.

Simeone, A., Acampora, D., Pannese, M., D’Esposito, M., Stornaiuolo, A.,Gulisano, M., Mallamaci, A., Kastury, K., Druck, T., Huebner, K.,Boncinelli, E., 1994. Cloning and characterization of two members ofthe vertebrateDlx gene family. Proc. Natl. Acad. Sci. USA 91, 2250–2254.

Soriano, P., 1997. The PDGFa receptor is required for neural crest devel-opment and for normal patterning of the somites. Development 124,2691–2700.

Spencer-Dene, B., Thorogood, P., Nair, S., Kenny, A.J., Harris, M.,Henderson, B., 1994. Distribution of, and a putative role for, the cell-surface neutral metallo-endopeptidases during mammalian craniofacialdevelopment. Development 120, 3213–3226.

Stadler, H.S., Solursh, M., 1994. Characterization of the homeobox-con-


taining geneGH6 identifies novel regions of homeobox gene expressionin the developing chick embryo. Dev. Biol. 161, 251–262.

Stock, D.W., Ellies, D.L., Zhao, Z., Ekker, M., Ruddle, F.H., Weiss, K.M.,1996. The evolution of the vertebrate Dlx gene family. Proc. Natl. Acad.Sci. USA 93, 10858–10863.

Stoetzel, C., Weber, B., Bourgeois, P., Bolcato-Bellemin, A.L., Perrin-Schmitt, F., 1995. Dorso-ventral and rostro-caudal sequential expressionof M-twist in the postimplantation murine embryo. Mech. Dev. 51, 251–263.

Stone, J.G., Spirling, L.I., Richardson, M.K., 1997. The neural crest popu-lation responding to endothelin-3 in vitro includes multipotent cells. J.Cell. Sci. 110, 1673–1682.

Sun, D., Vanderburg, C.R., Odierna, G.S., Hay, E.D., 1998. TGFb3 pro-motes transformation of chicken palate medial edge epithelium tomesenchyme in vitro. Development 125, 95–105.

Suzuki, H.R., Padanilam, B.J., Vitale, E., Ramirez, F., Solursh, M., 1991.Repeating developmental expression of G-Hox 7, a novel homeobox-containing gene in the chicken. Dev. Biol. 148, 375–388.

Swope, V.B., Medrano, E.E., Smalara, D., Abdel-Malek, Z.A., 1995. Longterm proliferation of human melanocytes is supported by the physiolo-gical mitogens a-melanotropin, endothelin-1, and basic fibroblastgrowth factor. Exp. Cell. Res. 217, 453–459.

Tassabehji, M., Read, A.P., Newton, V.E., Harris, R., Balling, R., Gruss,P., Strachan, T., 1992. Waardenburg’s syndrome patients have muta-tions in the human homologue of thePax-3 paired box gene. Nature355, 635–636.

Thomas, J.T., Kilpatrick, M.W., Lin, K., Erlacher, L., Lembessis, P.,Costa, T., Tsipouras, P., Luyten, F.P., 1997. Disruption of humanlimb morphogenesis by a dominant negative mutation inCDMP1.Nat. Genet. 17, 58–64.

Thorogood, P., Bee, J., von der Mark, K., 1986. Transient expression ofcollagen type II at epitheliomesenchymal interfaces during morphogen-esis of the cartilaginous neurocranium. Dev. Biol. 116, 497–509.

Thorogood, P., 1993. Differentiation and morphogenesis of cranial skeletaltissues. In: Hanken, J., Hall, B., editors. The Skull, Vol. 1. UniversityChicago Press, pp. 112–152.

Thorogood, P., 1998. The head and the face. In: Thorogood, P., editor.Embryos, Genes and Birth Defects. Wiley, pp. 197–229.

Tickle, C., Eichele, G., 1994. Vertebrate limb development. Ann. Rev.Cell Biol. 10, 121–152.

Tissier-Seta, J.P., Mucchielli, M.L., Mark, M., Mattei, M.G., Goridis, C.,Brunet, J.F., 1995. Barx1, a new mouse homeodomain transcriptionfactor expressed in cranio-facial ectomesenchyme and the stomach.Mech. Dev. 51, 3–15.

Ton, C.C.T., Hirvonen, H., Miwa, H., Weil, M.M., Monaghan, P., Jordan,T., van Heyningen, V., Hastie, N.D., Meijers-Heijboer, H., Drechsler,M., Royer-Pokora, B., Collins, F., Swaroop, A., Strong, L.C., Saunders,G.F., 1991. Positional cloning and characterization of a paired box- andhomeobox-containing gene from the aniridia region. Cell 67, 1059–1074.

Tsukamoto, A.S., Grosschedl, R., Guzman, R.C., Parslow, T., Varmus,H.E., 1988. Expression of the int-1 gene in transgenic mice is associatedwith mammary gland hyperplasia and adenocarcinomas in male andfemale mice. Cell 55, 619–625.

Unden, A.B., Holmberg, E., Lundh-Rozell, B., Sta˚hle-Backdahl, M.,Zaphiropoulos, P.G., Toftga˚rd, R., Vorechovsky, I., 1996. Mutationsin the Human Homologue ofDrosophila patched (PTCH)in basalcell carcinomas and the Gorlin syndrome: Different in vivo mechanismsof PTCH inactivation. Cancer Res. 56, 4562–4565.

Vainio, S., Karavanova, I., Jowett, A., Thesleff, I., 1993. Identification ofBMP-4 as a signal mediating secondary induction between epithelial-mesenchymal interactions during early tooth development. Cell 75, 45–58.

Valsecchi, V., Ghezzi, C., Ballabio, A., Rugarli, E.I., 1997. JAGGED2: aputative Notch ligand expressed in the apical ectodermal ridge and insites of epithelial-mesenchymal interactions. Mech. Dev. 69, 203–207.

van Genderen, C., Okamura, R.M., Farinãs, I., Quo, R.-G., Parslow, T.G.,

Bruhn, L., Grosschedl, R., 1994. Development of several organs thatrequire inductive epithelial-mesenchymal interactions is impaired inLEF-1-deficient mice. Genes Dev. 8, 2691–2703.

Vastardis, H., Karimbux, N., Guthua, S.W., Seidman, J.G., Seidman, C.E.,1996. A humanMSX1homeodomain missense mutation causes selectivetooth agenesis. Nat. Genet. 13, 417–421.

Vieille-Grosjean, I., Hunt, P., Gulisano, M., Boncinelli, E., Thorogood, P.,1997. Branchial HOX gene expression and human craniofacialdevelopment. Dev. Biol. 183, 49–60.

Vogel, A., Rodriguez, C., Izpisuá-Belmonte, J.-C., 1996. Involvement ofFGF-8 in initiation, outgrowth and patterning of the vertebrate limb.Development 122, 1737–1750.

Vortkamp, A., Gessler, M., Grzeschik, K.H., 1991. GLI3 zinc-finger geneinterrupted by translocations in Greig syndrome families. Nature 352,539–540.

Wagner, G., 1949. Die Bedeutung der neuralleiste fu¨r die kopfgestaltungder amphibienlarven. Rev. Suisse Zool. 56, 519–620.

Wall, N.A., Hogan, B.L.M., 1995. Expression of bone morphogeneticprotein-4 (BMP-4), bone morphogenetic protein-7 (BMP-7), fibroblastgrowth factor (FGF-8) and sonic hedgehog (SHH) during branchial archdevelopment in the chick. Mech. Dev. 53, 383–392.

Walterhouse, D., Ahmed, M., Slusarski, D., Kalamaras, J., Boucher, D.,Holmgren, R., Iannaccone, P., 1993.gli, A zinc finger transcriptionfactor and oncogene, is expressed during normal mouse development.Dev. Dyn. 196, 91–102.

Walther, C., Gruss, P., 1991.Pax-6, a murine paired box gene, is expressedin the developing CNS. Development 113, 1435–1449.

Wanaka, A., Milbrandt, J., Johnson, E.M., 1991. Expression of FGF recep-tor gene in rat development. Development 111, 455–468.

Wang, J., Shackleford, G.M., 1996. MurineWnt10aandWnt10b: cloningand expression in developing limbs, face and skin of embryos and inadults. Oncogene 13, 1537–1544.

Wedden, S.E., 1987. Epithelial-mesenchymal interactions in the develop-ment of chick facial primordia and the target of retinoid action.Development 99, 341–351.

Wilcox, J.N., Derynck, R., 1988. Developmental expression of transform-ing growth factors alpha and beta in mouse fetus. Mol. Cell. Biol. 8,3415–3422.

Wilke, T.A., Gubbels, S., Schwartz, J., Richman, J.M., 1997. Expression offibroblast growth factor receptors (FGFR1, FGFR2, FGFR3) in thedeveloping head and face. Dev. Dyn. 210, 41–52.

Wilkie, A.O.M., Morriss-Kay, G.M., Jones, E.Y., Heath, J.K., 1995. Func-tions of fibroblast growth factors and their receptors. Curr. Biol. 5, 500–507.

Wilkie, A.O.M., 1997. Craniosynostosis: genes and mechanisms. Hum.Mol. Genet. 6, 1647–1656.

Winnier, G., Blessing, M., Labosky, P.A., Hogan, B.L.M., 1995. Bonemorphogenetic protein-4 is required for mesoderm formation and pat-terning in the mouse. Genes Dev. 9, 2105–2116.

Winter, R.M., 1996. What’s in a face?. Nat. Genet. 12, 124–129.Wise, C.A., Chiang, L.C., Paznekas, W.A., Sharma, M., Musy, M.M.,

Ashley, J.A., Lovett, M., Jabs, E.W., 1997.TCOF1 gene encodes aputative nucleolar phosphoprotein that exhibits mutations in Treacher-Collins syndrome throughout its coding region. Proc. Natl. Acad. Sci.USA 94, 3110–3115.

Wolf, C., Thisse, C., Stoetzel, C., Thisse, B., Gerlinger, P., Perrin-Schmitt,F., 1991. TheM-twist gene ofMus is expressed in subsets of mesoder-mal cells and is closely related to theXenopus X-twiand theDrosophilatwist genes. Dev. Biol. 143, 363–373.

Wood, A., Ashhurst, D.E., Corbett, A., Thorogood, P., 1991. The transientexpression of type II collagen at tissue interfaces during mammaliancraniofacial development. Development 111, 955–968.

Wu, L., Wu, H., Sangiorgi, F., Wu, N., Bell, J.R., Lyons, G.E., Maxson, R.,1997. Miz1, a novel zinc finger transcription factor that interacts withMsx2 and enhances its affinity for DNA. Mech. Dev. 65, 3–17.

Yanagisawa, H., Yanagisawa, M., Kapur, R.P., Richardson, J.A.,Williams, S.C., Clouthier, D.E., de Wit, D., Emoto, N., Hammer,


R.E., 1998. Dual genetic pathways of endothelin-mediated intercellularsignaling revealed by targeted disruption of endothelin convertingenzyme-1 gene. Development 125, 825–836.

Yamada, G., Mansouri, A., Torres, M., Stuart, E.T., Blum, M., Schultz,M., De Robertis, E.M., Gruss, P., 1995. Targeted mutation of the murinegoosecoidgene results in craniofacial defects and neonatal death.Development 121, 2917–2922.

Yamaguchi, T.P., Conlon, R.A., Rossant, J., 1992. Expression of the fibro-blast growth factor receptor FGFR-1/flg during gastrulation and seg-mentation in the mouse embryo. Dev. Biol. 152, 75–88.

Yamaguchi, T.P., Rossant, J., 1995. Fibroblast growth factors in mamma-lian development. Curr. Opin. Genet. Dev. 5, 485–491.

Zhang, H., Bradley, A., 1996. Mice deficient for BMP2 are nonviable andhave defects in amnion/chorion and cardiac development. Development122, 2977–2986.

Zhang, H., Hu, G., Wang, H., Sciavolino, P., Iler, N., Shen, M.M., Abate-Shen, C., 1997. Heterodimerization of Msx and Dlx Homeoproteinsresults in functional anatagonism. Mol. Cell. Biol. 17, 2920–2932.

Zhang, J., Hagopian-Donaldson, S., Serbedzija, G., Elsemore, J., Plehn-Dujowich, D., McMahon, A.P., Flavell, R.A., Williams, T., 1996.Neural tube, skeletal and body wall defects in mice lacking transcriptionfactor AP-2. Nature 381, 238–241.

Zhao, G.-Q., Eberspaecher, H., Seldin, M.F., de Crombrugghe, B., 1994.The gene for the homeodomain-containing proteinCart-1 is expressedin cells that have chondrogenic potential during embryonicdevelopment. Mech. Dev. 48, 245–254.

Zhao, Q., Behringer, R.R., de Crombrugghe, B., 1996. Prenatal folic acidtreatment suppresses acrania and meroanencephaly in mice mutant forthe Cart1 homeobox gene. Nat. Genet. 13, 275–283.


Documents

Signalling interactions during facial development