sign in sign up
<<
Home , Modularity (biology)

Developmental 332 (2009) 61–69

Contents lists available at ScienceDirect

Developmental Biology

journal homepage: www.elsevier.com/developmentalbiology

Review Modularity, comparative embryology and evo-devo: Developmental dissection of evolving body plans

Shigeru Kuratani ⁎

Evolutionary Morphology Research Group, RIKEN Center for , 2-2-3 Minatojima Minamimachi, Chuo-ku, Kobe 650-0047, Japan article info abstract

Article history: Modules can be defined as quasi-autonomous units that are connected loosely with each other within a Received for publication 18 March 2009 system. A need for the concept of modularity has emerged as we deal with evolving in Revised 18 May 2009 evolutionary developmental research, especially because it is unknown how are associated with Accepted 19 May 2009 anatomical patterns. One of the strategies to link genotypes with could be to relate Available online 23 May 2009 developmental modules with morphological ones. To do this, it is fundamental to grasp the context in which certain anatomical units and developmental processes are associated with each other specifically. By Keywords: identifying morphological modularities as units recognized by some categories of general as Modularity established by comparative anatomy, it becomes possible to identify developmental modules whose genetic Evolutionary novelty components exhibit coextensive expressions. This permits us to distinguish the evolutionary modification in Morphological homology which the identical morphological module simply alters its shape for , without being decoupled from the functioning network (‘coupled modularities’), from the evolution of novelty that involves a Developmental constraints heterotopic shift between the anatomical and developmental modules. Using this formulation, it becomes possible, within the realm of Geoffroy's homologous networks, to reduce morphological homologies to developmental mechanistic terms by dissociating certain classes of modules that are often associated with actual shapes and functions. © 2009 Elsevier Inc. All rights reserved.

The framework of bones being the same in the hand of a man, wing of as a possible link between the two. Modules are defined to represent a bat, fin of the porpoise and leg of the horse, - the same number of semiautonomous units or elements that are connected loosely with vertebrae forming the neck of the giraffe and of the elephant, - and others in a system (Raff, 1996; Carroll et al., 2001; Schlosser and innumerable other such facts, at once explain themselves on the Wagner, 2004; Klingenberg, 2008). There can be several different theory of descent with slow and slight successive modifications. modules defined in different contexts, such as those defined as The Origin of Species by Means of (Darwin, 1859) genetic, morphological, developmental and functional modules (for examples see Wagner et al. (2007), Klingenberg (2008)). Typically, …characters controlled by identical genes are not necessarily the animal body is made up of anatomical units that can often evolve homologous… The converse is no less instructive… homologous independently. For example, bats have obtained a wing based on the structures need to be controlled by identical genese and homology of forelimb module and gnathostomes have acquired jaws differentiated phenotypes does not imply similarity of genotypes. from the mandibular arch. Embryonic units such as these are also the de Beer (1971) sites at which regulatory genes are specifically expressed. Importantly, a discrete and specific regulation of genes is often Introduction associated with an anatomical unit. In the present review, the concept of modularities is considered in the context of development and Understanding the evolutionary changes involved in development evolution, as a possible conceptual tool to connect morphological and is not easy or simple. Changes can be seen at every hierarchical level of developmental units. The central issue is: how can we grasp the the developmental programs, anatomical patterns, or in the genes that interrelationships between different types of modules and how they function in morphogenetic development. To fill the gap between have changed through evolutionary processes? Because the hierarch- phenotypes and genes, the concept of modules has recently attracted ical organization of genes and any given anatomical structure do not the attention of researchers in evolutionary development (‘evo-devo’) correlate with each other, a conceptual framework has to be established to deal with the two variables simultaneously. To grasp ⁎ Fax: +81 78 306 3370. the modular structures in evo-devo biology from a top-down E-mail address: [email protected]. perspective, the discussion below focuses mainly on the relevance of

0012-1606/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2009.05.564 62 S. Kuratani / Developmental Biology 332 (2009) 61–69 comparative morphology to the field, including a consideration of how the body plan of animal phyla (von Baer, 1828; Raff, 1996; see below). developmental and morphological modules behave in the acquisition Thus, Geoffroy's rule can be applied primarily to animals that share of evolutionary novelties. thesamebasicbodyplanbasedonthesamemorphological modularity. What this implies is that the embryonic development Morphological modules: ideas from comparative embryology of various animals belonging to a is under a certain developmental constraint to show limited changes in the connectiv- The concept of morphological homology is profoundly associated ities of organs (Fig. 1A; Maynard-Smith et al., 1985; for further with morphological modularity. Typically, animal bodies are con- elaboration see Müller and Wagner (1991), Hall (1994), Wagner and structed based on a body plan, which is made of comparable Müller (2002), Schwenk and Wagner (2003), Galis and Sinervo elements that we perceive as homologous modules to build up the (2003), Kuratani (2003); also see below). Modules and constraints plan. These include the various germ layers and their derivatives, are thus tightly linked with each other at various levels of segmental units, tissue and cell types. These units form even more and evolution. Morphological modularity is clear in complicated functional modules or systems as seen in highly organized animal groups such as arthropods and chordates. vertebrate limbs or nervous systems. Thus, the morphological modu- Comparative anatomists such as Richard Owen in 1848 and Carl larity itself can constitute a hierarchical system (Woodger, 1945; Gegenbaur in 1898 elaborated the concepts of homologies (morpho- reviewed by Hall (1998)). logical homologies) into systematized categories, which still help us Examples of homologous morphological modules relevant to evo- recognize the nature of morphological modularity. devo are best exemplified by the works of comparative morphology. First, morphological homology can be categorized into special and By the early nineteenth century, comparative anatomy had revealed general homologies. ‘Special homology’ refers to a relationship that the animal body is constructed from a set of morphologically between equivalent modules (or sets of modules) belonging to homologous units commonly found in a taxon, and that these units different animals, as we see between our arm and the bat wing are found in common connectivities. Geoffroy St. Hilaire (1818) called (Fig. 1B). This type of homology is divided further into complete and this the ‘principe des connexions’, which is still regarded as the incomplete types (Fig. 2). Of these, ‘complete homology’ means a simplest definition of morphological homology (Hall, 1998). This is perfect match of homologous modules, as seen in elements in the also a context in which archetypes have been postulated to represent forelimbs among amniote species, whereas ‘incomplete homology’ is a

Fig. 1. Constraint and homology. (A) Various faces of different mammals share the same basic arrangement of organs, based on a common developmental program. Thus, morphological homology is conserved through the ancestral developmental constraint for the mammalian facial patterning program, which is hard to override through evolution. (B) Various hindlimbs seen in various : 1 and 2, amphibians; 3–7, reptiles; 8, a bird; 9–16, mammals. Complete homologies can be established for many of the hindlimbs shown here. A is from Haeckel (1874), B is from Haeckel (1902). S. Kuratani / Developmental Biology 332 (2009) 61–69 63

derm, and induction of epibranchial placodes (Begbie et al., 1999; Holzschuh et al., 2005; see also Kuratani and Kirby (1991, 1992) for pharyngeal arch formation). In other words, somitomeric and branchiomeric modules in the vertebrate body are established under the influence of generative constraints (Wagner, 1994) derived from primary segmental modules in embryos, the somites and pharyngeal pouches, respectively (see Kuratani (2008) for generative constraint and serial homology).

Links between morphological and developmental modules

As mentioned above, there can be various types of modularities in developmental phenomena. Among these, the subdivisions of general homology (Homodynamie, Homotypie, Homonymie and Homonomie) are associated with specific functions of control genes that are often called ‘genetic toolkits’ (reviewed by Carroll et al. (2001) and by De Robertis (2008)). Importantly, these toolkit genes are not only homologous within a given phylum or lower taxa of animals, but also widespread across various phyla (reviewed by Carroll et al. (2001); and see Garcia-Fernàndez (2005) for Hox codes). These genes are more or less specifically associated with particular developmental fi Fig. 2. Incomplete homology. Pectoral nofScymnus (A) and a generalized pattern for events or embryonic morphological modules in evolutionarily con- the forelimb (B). These two skeletal modules are homologous as the pectoral fins of jawed vertebrates; however, not all of the morphological elements or polarities served manners. found in the tetrapod pattern can be identified in the Scymnus fin. From Gegenbaur Typically, the modular nature of gene expression coding associated (1898). with some organogenetic events is often discussed in various contexts, such as the well-known Hox codes along the anteroposterior (AP) axis relationship such as that found between pectoral fins in sharks and the and the orderly expression patterns of metazoan homeotic selector tetrapod forelimb. The latter modules are identical as ‘pectoral fins’, genes. These genes form a cluster in the genome as the result of but the submodules in the tetrapod forelimb, such as the humerus, tandem duplication. The Hox cluster itself has undergone duplications ulna, radius and carpals, are not always found in the shark. in vertebrates (reviewed by Garcia-Fernàndez (2005) and Carroll General homology, on the other hand, refers to a classification of (2008)). The coordinated expression patterns of Hox genes become iterating body parts to describe a body plan systematically. Under this most conspicuous at the organogenetic stages of embryogenesis, also category, we can see various modular properties of the anatomical known as the phylotypes of a given taxonomic unit. This expression plan. Best known is the serial homology (‘Homodynamie’) that refers pattern thus represents a tool to define the body plan of animals to the relationships among iterating series of equivalent develop- mental units and their derivatives. Somitomerism (metamerism of somitic derivatives) and branchiomerism (metamerism of pharyngeal arch derivatives) are known in vertebrates (neuromeres are also recognized in the vertebrate central nervous system, but they are not always equivalent along the entire neuraxis). Serially homologous units can be called homodynamic modules after the terminology of Gegenbaur (1898). As the second category of general homology, Homotypie refers to the relationships between counterparts that are distributed with bilateral symmetry (Fig. 3). Thus, many organs are homotypic in bilaterians except for secondarily asymmetrical digestive tracts or morphological parts of the heart (and also the staggered dorsal spines of the Stegosaurus and myotomes of amphioxus). Similarly, Homony- mie refers to a relation among segmental elements found in a single serial homologue, such as teeth growing in the jaw or fingers in a tetrapod limb (Fig. 3). Finally, like Homonymie, Homonomie refers to the serially homologous elements found in equivalent positions of the fore- and hindlimbs (for example, the humerus and femur in tetrapods; Fig. 3). In these categories of morphological modules, we can already identify the relevant embryonic phenomena rather easily. For example, we now know that segmental patterns of vertebrae, spinal nerves and myotomes stem from the segmental patterns of somites. Not only do the structures differentiate directly from the somites but also the neural crest-derived cells and spinal motor axons are patterned through the presence of the somite to assume segmental distribution patterns: they are both inhibited by the caudal halves of somites (Keynes and Stern, 1984; Rickmann et al., 1985). Similar morphogenetic or priority is known for the pharyngeal Fig. 3. Categories of general homology. The morphological relationships called Homo- pouches responsible for pharyngeal arch development, division of typie, Homonymie and Homonomie are illustrated, based on the anatomy of salamander neural crest-derived ectomesenchyme and pharyngeal arch meso- limbs described by Gegenbaur (1898). For details, see the text. 64 S. Kuratani / Developmental Biology 332 (2009) 61–69

(Duboule, 1994). The Hox code can be regarded as a developmental (Fig. 4). Comparative observation of expression patterns module, not simply as the colinear expression patterns of the genes. suggests that the morphological homologies of the vertebrae are Instead, it acts as a developmental system to function in a specific associated not with the segmental numbers but with the Hox genes morphological context as the homeotic selector tool for specification expressed there: indeed the names of expressed Hox genes on of elements along the embryonic axes of various organ systems. provertebrae can almost be used as anatomical terms. Thus, we A similar category covers gene expression coding in the dorsoven- know that Hoxc6 is commonly responsible for the cervical–thoracic tral (mediolateral) axis of the central nervous system in eubilateria transition in many different vertebrates (Burke et al., 1995; Ohya et al., (Denes et al., 2007) and the coordinated expressions of a number of 2005; Fig. 4) and the elongated axial skeleton of the python can be genes in the prosencephalic neuromeres (reviewed by understood as an enumeration of the vertebrae specified as those in Echevarría et al. (2003) and Lowe et al. (2003)). As shown in the the thoracic region by the change of Hox code (Cohn and Tickle, 1999). hypothetical ‘ureubilateria’, there are many more modules wide- In other words, segmental organization (somitogenesis) and AP spread and highly conserved across animal phyla, making all the specification (Hox gene regulation) are primarily decoupled from each eubilaterians look similar and possess one common body plan, as other (Burke et al., 1995; also see Goodrich (1930) and de Beer (1937); Geoffroy once imagined (reviewed by Appel (1987)). Once the Fig. 5) and this fact liberates the classical morphologists from the module is stabilized at the end of the organogenetic period of agony of trying to reconcile homology with segmental numbers. development, the gene expression profiles define well-established As another possible case, the alignment of visceral arch elements anatomical domains or even cell types (‘homologous cell types’ and myotomic elements used to be an issue of debate in comparative according to Arendt (2008)). Such modules are evolutionarily stable morphology regarding the primary organization of the vertebrate and expected to serve as an ideal starting point to compare head. For typical ‘segmentalists’, the mesoderm in the head is also developmental and morphological modules. segmented into blocks, like somites in the trunk, each of which was An important aspect of developmental modularity is that it can assumed to correspond to each pharyngeal arch. Thus, branchiomer- potentially translate comparative morphological concepts—especially ism (metamerism of the arches) and somitomerism (metamerism of the categories of general homologies—into terms of developmental the paraxial mesodermal segments) were thought to be identical. dynamism. This is because many of the developmental modules are Comparative embryology, however, has shown that the branchiomer- associated and coextensive with specific morphological modules ism and somitomerism are decoupled both developmentally and (character identity network; see Wagner (2007)). First, among the evolutionarily—they can vary in number and position and can shift categories of general homologies, Homotypie is linked to the Nodal- independently (de Beer, 1922; Goodrich, 1930). Thus, these two Pitx2 regulatory cascade to determine left–right asymmetry, homo- metamerisms represent two different morphological modules. dynamic modules (=serial homologues) to elements specified by the Modularity is reflected in comparative morphology, namely the Hox code modules, homonymic modules to the visceral arches specified composition of the body plan at semantic and cognitive levels. by the Dlx code (Depew et al., 2002) and homonomic modules to fore- However, in the evo-devo context, it is the key to detecting and hindlimb buds by Tbx4/5 and Hox genes (Dollé et al., 1989; associations between morphological units and relevant developmen- Yokouchi et al., 1991; Takeuchi et al., 1999; Rodriguez-Esteban et al., tal phenomena or mechanisms. In this, there arises the possibility of 1999; also see Naiche and Papaioannou (2007)). In every case, the connecting genes and phenotypes without losing morphological expression domains and functions of such toolkit genes are concep- semantics. More importantly, as we have seen in a few cases, tually coextensive with the morphological modules found in the modularity provides the cue to ‘stiffness’ and ‘softness’ in evolutionary vertebrate body plan. This indicates that comparative morphology changes through its autonomy. Thus, modules are relevant not only to parallels the developmental formulation of the animal body plan. the concepts of morphological homology or body plan but also to the Typically, the Hox code is parallel to the position-dependent developmental constraints defined as bias in evolutionary changes. morphological specification (morphological homologies) of vertebrae and the visceral arches (Burke et al.,1995). Comparative morphological Novelties and constraints: the lamprey and jaw evolution concepts relevant to these morphological modules are the serial homologies (Owen, 1848; Gegenbaur, 1898). Association between A tight association or coupling between developmental and morphology and development extends even to the peculiar morpho- morphological modules often stems from the phenomenon termed logical patterns obtained by genetic experiments, such as transforma- an ‘epigenetic trap’ (Wagner, 1989a,b): a type of structural network tion of segments by shifting the Hox code (Rijli et al., 1993; Gendron- created by gene regulatory mechanisms and embryonic tissue Maguire et al.,1993). In this, disruption of the Hoxa2 gene that specifies interactions. Because a developmental network like this consists of the hyoid arch in vertebrate embryos leads to the transformation of interdependent elements specified by the very interactions them- that arch into the identity of the first element that is defined by the Hox selves, it becomes tremendously hard to override by selection. This code-default state (Rijli et al., 1993; Couly et al., 1998). Thus, when the results in conservative morphological patterns associated with the developmental module is shifted, the anatomical pattern is also shifted expression of conservative genes, as typically seen in the vertebrate in a parallel manner. This association alludes strongly to the eye or tetrapod limbs. As this conservation serves as a developmental mechanical link between these two modules to assure the develop- constraint to maintain morphological modules, we can predict mental background of homology, as suggested above. whether morphological homologies will be preserved perfectly Discovery of the Hox code was tremendously meaningful for through evolution. For example, in the wing of a bat we can observe understanding the morphology of vertebrates, as it revealed tight every anatomical element (morphological module) found in the linkages between particular developmental and morphological mod- human arm, because the bat could not escape from mammalian ules (Kessel et al., 1990; reviewed by Kessel (1992)). Actually, from developmental constraint in its limb patterning, even if the wing has comparative anatomy, morphological homologies of the vertebrae attained a tremendously distinct function. Naturally, if evolutionary develop at different axial levels in various animals, and different novelty is to be obtained by overriding the ancestral developmental numbers of cervical, thoracolumbar, sacral and caudal vertebrae arise constraints, we may be able to see a disruption of morphological along the AP axes of different animals. Thus, morphologically homologies in the rise of novelties (see Wagner and Altenberg (1996), equivalent vertebrae develop from different numbers of somites in Wagner and Müller (2002) and Wagner et al. (2007) for a similar each animal, a paradox that was a typical dilemma for those pure consideration). In other words, morphological and developmental morphologists such as Gegenbaur (1887) and Fürbringer (1897) who modules have to be able to decouple from time to time, which seems tried to reduce this type of homology to serial segmental numbering to be the case. S. Kuratani / Developmental Biology 332 (2009) 61–69 65

Fig. 4. Homeotic shift of vertebral identities and Hox coding. (A) In a classical comparative anatomical concept, it was believed by Gegenbaur (1887) and Fürbringer (1897) that the segments and spinal nerves given the same serial numbers should represent morphologically identical vertebrae and nerves. Thus, the protometameric cranium was assumed to contain certain vertebral segments in all the vertebrate species. Even in cyclostomes, latent protometameric cranial segments were believed to be present already among the trunk vertebrae. Secondarily assimilated vertebrae were called part of the auximetameric cranium and were regarded as homologues of primary trunk vertebrae. (B) Explanation by de Beer (1937). Like Goodrich (1930), de Beer thought that different numbers of segments could be ‘transformed’ to become occipital bones in each animal lineage. (C) Current understanding of segmental numbers and homologies is closer to the idea of ‘transformation’. Here, the morphological homologies of vertebrae are coupled not with the segmental numbers but with the homologies of Hox genes expressed in provertebrae, whose number can shift anteroposteriorly along the axis in evolution. Modified from Burke et al. (1995). 66 S. Kuratani / Developmental Biology 332 (2009) 61–69

Fig. 5. Coupling and decoupling of modules in heterotopic evolution of the vertebrate jaw. Top. Distribution pattern of cephalic ectomesenchymal modules in lamprey and gnathostome embryos. Identical sets of crest cell populations (soc, supraoptic crest cells; ioc, infraoptic crest cells; mc, mandibular arch crest cells) can be identified, based on topographical relationships of the mesenchyme with commonly found embryonic structures such as the eye, the premandibular mesoderm and the first pharyngeal pouch (p). Based on Horigome et al. (1999) and Kuratani et al. (1999, 2001). Bottom. Developmental and morphological modules are illustrated diagrammatically with the shapes of the oral apparatus in both of the animals at the bottom. These modules do not necessarily represent a ‘true’ chronological sequence of developmental events. Of the various developmental modules, the Hox code module has been coupled with the morphological ectomesenchymal modules in a conserved manner. Thus, Hox2 expression is always associated with the hyoid crest cells (hc) and the mandibular and premandibular ectomesenchyme are defined by the Hox code-default state. In this way, the Hox module has been conserved morphologically in vertebrate embryogenesis. The oral patterning module is based on a conserved epithelial–mesenchymal tissue interaction involving a set of growth factors, FGF8/17 and BMP2/4 and their target homeobox genes, Dlx and Msx. This module is recognized commonly in the lamprey and gnathostome embryos, but it functions in different sets of morphological modules (broken lines to show spatial expansions of the module). This module is rather tightly coupled with the functional shapes of the developed oral apparatus, in which BMP2/ 4-Msx domains (purple) always specify the dorsoventral tips of the vertebrate oral margin. This heterotopic evolution of the vertebrate jaw appears to have been based on decoupling between the ectomesenchymal–Hox module and the oral shaping module. It is thus only in gnathostomes that the oral apparatus can be called a ‘mandibular arch derivative’.Note that the concept of morphological homology can be seen only in the ectomesenchymal–Hox module, not in the oral shaping module. The similarity between the lamprey lips and gnathostome jaws can be called homoplasy, or ‘analogy’ in the sense of Owen (1848), in that these organs resemble each other only for similar functions but are derived from morphologically nonequivalent embryonic primordia. Based on Shigetani et al. (2005). Abbreviations: llp, lower lip; md, lower jaw; mx, upper jaw; ulp, upper lip.

One of the best examples to see the coupling and decoupling of embryos have shown that the lamprey oral apparatus consists of the morphological modules is in the heterotopic evolution of the jaw in upper lip derived from the premandibular module and the lower lip gnathostomes (Shigetani et al., 2002). Gnathostomes, or jawed derived from the mandibular arch module. By contrast, as mentioned vertebrates, possess dorsoventrally articulated jaws (upper and above, the gnathostome jaw consists of upper and lower jaws both lower) mostly derived from a morphological module, the mandibular derived from the mandibular arch module (Kuratani et al., 2001, 2004; arch, which contains Hox-negative, neural crest-derived ectome- Fig. 5). senchyme. Note that only a minor involvement of the premandibular The mesenchymal composition of the early embryonic head per se ectomesenchyme, another module rostral to the mandibular arch, has is quite similar between the lamprey and gnathostomes. With respect been implied for the formation of the rostralmost part of the upper to the common embryonic anlagen—such as eyes, pharyngeal pouches jaw in some gnathostome embryos (Wada et al., 2005; also see Lee et and the otic pit—mesodermal and crest-derived mesenchymal tissues al. (2004) and Cerny et al. (2004)). In the lamprey, an extant jawless are distributed in a similar pattern, showing a shared morphological vertebrate, the upper and lower lips arise in larval stages, but these modularity in the embryonic head mesenchyme; at this stage, lips do not correspond to the upper and lower jaws, respectively (if morphological homology is perfectly satisfied for each morphological they did, lampreys could no longer be classified as ‘agnathans’). module (Horigome et al., 1999; Kuratani et al., 1999; Fig. 5). Comparative observations between lamprey and gnathostome Homologous regulatory genes are also expressed in these animals in S. Kuratani / Developmental Biology 332 (2009) 61–69 67 consistent manners, as seen in the En- and Hox-homologues. Others Coupling and decoupling: homology and modularity are expressed in the central nervous system (Holland et al., 1993; Neidert et al., 2001; Murakami et al., 2002; Takio et al., 2004, 2007; If we admit the morphological homologies of cephalic mesench- Matsuura et al., 2008). These observations imply that many develop- ymal domains in the early embryos between the lamprey and mental modules are tightly linked to morphological modules by gnathostomes, we have to abandon any idea of homology between morphologically homologous expressions. However, differences arise the lamprey oral apparatus (ammocoete upper and lower lips) and in the assignment of mesenchymal modules to the patterning of the gnathostome jaws. If they were homologous, the homologue of the oral region, which takes place later in development. gnathostome trabecula would not be present in the lamprey, because The patterning of the oral region is best monitored by the the trabecula is supposedly derived from the premandibular module expression of Fgf8 homologues in the ventral head ectoderm, (Kuratani et al., 2004). Similarly, if we define homologies of cephalic corresponding to the future stomodaeum. The ectodermally derived mesenchymal components solely by their topographical relationships FGF8 induces its downstream target genes such as Dlx1 in the with embryonic structures, expression domains of homologous adjacent ectomesenchyme, which in gnathostomes will specify the regulatory genes are no longer associated with the morphologically hinge region of the jaw. Another growth factor, BMP4, is released from homologous mesenchyme among vertebrates. In other words, the ectoderm covering the dorsal and ventral tips of the jaws to morphological modules and developmental modules are not coupled activate its target gene Msx1 in the ectomesenchyme at the tip of the during evolution, so that morphologically homologous modules are upper and lower jaws. Thus, in the gnathostome, the mandibular arch not always coupled with the same developmental modules. The acquires a gradient of homeobox gene expression along the AP (hinge- above-noted refers to this uncoupling. cap) axis and this pattern is prefigured by the distribution of growth In the above example—the development of the oral region—there factors in the ectoderm (Neubüser et al., 1997; Ferguson et al., 2000; are no homologies that satisfy every aspect of developmental Shigetani et al., 2000; Depew et al., 2002). These interactions, or a identities, so that homologous structures do not always express or developmental module, further leads to the proximodistal specifica- are not always patterned by homologous genes. In early development, tion of the gnathostome jaw and to the patterning of dentition; embryonic anatomical units are all homologous between the lamprey inactivation of distal BMP4 activity leads to the development of molar and gnathostomes. The Hox code is also consistent with the latter as teeth in place of incisors (Tucker et al., 1998). far as the genes for PG2 and PG3 are concerned. In addition, the Importantly, the above developmental module not only functions mandibular arch is defined as such by the Hox code-default state in the axis specification in the established oral apparatus but also is (Takio et al., 2004) and the Hox code developmental module is responsible for the specification of the extent of the oral region in consistent with the morphological modules. As we have seen above, early stages (Shigetani et al., 2000). Thus, the same developmental our trouble is that the term ‘oral’ is coupled with the developmental module is associated with different morphogenetic contexts at module, whereas the morphological module clarifies the identities different developmental stages, which explains the absence of a jaw (homologies) of embryonic primordia (Fig. 5). Thus, the functional homologue in the lamprey. Molecules homologous to those in shape of a given structure is irrelevant to morphological homologies, gnathostomes also function in oral patterning in the lamprey, based as has long been emphasized by comparative morphologists (Owen, on an evolutionarily shared molecular cascade involving Fgf8/17, 1848). This is consistent with the function of the regulatory genes Bmp2/4, Dlx and Msx homologues (Shigetani et al., 2002). This involved in this developmental module: thus, Fgf8/Dlx expression is developmental module is therefore homologous between the lamprey roughly associated with the functional oral domain and Bmp4/Msx and gnathostomes. Actually, mammalian recombinant FGF8 and BMP4 expression with the oral fringes (distal tips). This suggests that the applied to the lamprey embryo are capable of inducing shape of the ‘mouth’ had already been established as a developmental ectopic expressions of lamprey Dlx and Msx genes (endogenous target module before it was applied to the morphological module of the genes), respectively. However, the expression domains of these genes mandibular arch of the gnathostomes (Shigetani et al., 2005). are dissimilar between lamprey and gnathostome embryos. Thus, The upstream developmental factor that induced the above LjFgf8/17 is upregulated in a morphologically more extensive area of ‘decoupling’ is not well understood. Possibly, a spatial shift of the the embryonic ectoderm of the lamprey than in gnathostomes, and oral endoderm occurred before the heterotopy. The oral endoderm- LjDlx upregulation is seen not only in the mandibular arch ectome- derived factor SHH might define the localization of the FGF8–BMP4 senchyme but also in the crest cells in the so-called premandibular module in the stomodaeum (Brito et al., 2006, 2008). It is also domain (Kuratani, 2004). This shows that morphological modules are possible that the oral shaping module is coupled developmentally not coupled evolutionarily with the same developmental modules in with the patterning of the nasal and adenohypophysial placodes: in the differentiation of the vertebrate oral apparatus (Fig. 5). the cyclostomes, olfactory and adenohypophysial placodes arise as a Because of the uncoupling of morphological and developmental single medial placode called the nasohypophysial plate (reviewed by modules, we no longer see a perfect comparison between the lamprey Kuratani et al. (2001)). Thus, there is only one nasal placode in the and gnathostome oral regions. The gnathostome mouth is formed of lamprey, confluent with the caudally located adenohypophysis that the mandibular arch module, but the lamprey mouth involves the is not specified as a part of the oral ectoderm but is rostral to it. In premandibular module as its large dorsal component. Therefore, we this context, Uchida et al. (2003) found that, of the genes known to have suggested a heterotopic shift or restriction of the same be involved in the induction of the Rathke's pouch in amniotes, only epithelial–mesenchymal interaction (FGF–BMP developmental mod- those encoding for signaling molecules have shifted their expression ule) to the mandibular arch module in the gnathostome ancestor, spatially (), while the expressions of transcription- allowing the formation of the gnathostome-type oral apparatus factor-encoding genes are always associated with equivalent cell (Shigetani et al., 2002; Kuratani, 2005; also see Gilbert (2006) and types between the lamprey and gnathostomes. In other words, the Mallatt (2008) for other interpretations). Here, the developmental evolution of the jaw has been coupled evolutionarily with the double module is associated with the shape of the ‘functional oral apparatus’, nostril state (diplorhiny) of the vertebrate head and is thus to be not with morphologically homologous modules. ‘Heterotopy’ is a term regarded as a part of placodal patterning in early vertebrate coined by Haeckel (1874, 1875) to refer to the spatial shift of evolution in which heterotopic regulatory evolution appears to development in evolution (just as heterochrony means the temporal have played a central role (Kuratani and Ota, 2008; also see Janvier shift of the developmental timetable through evolution). Heterotopy (1996)). In this evolutionary event, the developmental module of the developmental module is seen in the agnathan-to-gnathostome might have utilized factors such as FGF8/17, BMP2/4 and SHH; their transition. spatial application does not appear to have been coupled with the 68 S. Kuratani / Developmental Biology 332 (2009) 61–69 morphological module of crest-derived ectomesenchyme, leading to Arendt, D., 2008. The evolution of cell types in animals: emerging principles from – ‘ ’ molecular studies. Nat. Rev. Genet. 9, 868 882. disruption of the morphological homology of the mouth in de Beer, G.R., 1922. The of the head in Squalus acanthias. Quart. J. Microsc. vertebrates. Sci. 66, 457–474. Homology has always been the key concept for comparing de Beer, G.R., 1937. The Development of the Vertebrate Skull. Oxford Univ Press, London. de Beer, G.R., 1958. Embryos and Ancestors. Oxford Univ. Press, Oxford. developmental phenomena in evo-devo research. It is true that de Beer, G.R., 1971. In: Head, J.J., Lowenstein, O.E. (Eds.), Homology, An Unsolved there are ample instances in which morphological identities do not Problem, Oxford Biology Readers. Oxford Univ. Press, Oxford. coincide with the genes expressed, with cell lineages from which the Begbie, J., Brunet, J.F., Rubenstein, J.L., Graham, A., 1999. Induction of the epibranchial – organs originate, or even with the developmental processes by which placodes. Development 126, 895 902. Bellairs, A.D'A., Gans, C., 1983. A reinterpretation of the amphisbaenian orbitosphenoid. certain structures are formed (reviewed by de Beer (1958, 1971), Nature 302, 243–244. Bellairs and Gans (1983), Hall (1994, 1998), Larsson and Wagner Brito, J.M., Teillet, M.A., Le Douarin, N.M., 2006. An early role for sonic hedgehog from (2002), Scholtz (2005)). What seems to be missing intellectually is foregut endoderm in jaw development: ensuring neural crest cell survival. Proc. Natl. Acad. Sci. U. S. A. 103, 11607–11612. the reintegration of comparative morphological concepts and those of Brito, J.M., Teillet, M.A., Le Douarin, N.M., 2008. Induction of mirror-image super- developmental biology into evo-devo, by establishing relationships numerary jaws in chicken mandibular mesenchyme by Sonic Hedgehog-producing between developmental and morphological modules, or possibly by cells. Development 135, 2311–2319. Burke, A.C., Nelson, C.E., Morgan, B.A., Tabin, C., 1995. Hox genes and the evolution of integration of the concepts of developmental and morphological vertebrate axial morphology. Development 121, 333–346. homologies. Carroll, S.B., 2008. Evo-devo and an expanding evolutionary synthesis: a genetic theory Tissue interactions during embryogenesis doubtless serve to of morphological evolution. Cell 134, 25–36. Carroll, S.B., Greiner, J.K., Weatherbee, S.D., 2001. From DNA to Diversity. Blackwell Sci., anchor developmental modules with related and conserved Victoria. morphological homologies. Similarly, linkages have been main- Cerny, R., Lwigale, P., Ericsson, R., Meuleman, D., Epperlein, H.H., Bronner-Fraser, M., tained by the autoregulation of some transcription-factor-encoding 2004. Developmental origins and evolution of jaws: new interpretation of “maxillary” and “mandibular”. Dev. Biol. 276, 225–236. genes in developmental compartments (segments) after interme- Cohn, M.J., Tickle, C., 1999. Developmental basis of limblessness and axial patterning in diate developmental stages. These linkages assure, for example, the snakes. Nature 399, 474–479. anatomical patterns of embryos at the phylotypic stages and Couly, G., Grapin-Botton, A., Coltey, P., Ruhin, B., Le Douarin, N.M., 1998. Determination homeobox gene expression patterns in the mesenchyme to of the identity of the derivatives of the cephalic neural crest: incompatibility between Hox gene expression and lower jaw development. Development 125, establish position-dependent identities of each segment thereafter, 3445–3459. many of which are laid down through global interactions in the Darwin, C., 1859. The Origin of Species by Means of Natural Selection. John Murray, phylotype. Thus, it is conceivable that the conserved gene regula- London. Denes, A.S., Jékely, G., Steinmetz, P.R., Raible, F., Snyman, H., Prud homme, B., Ferrier, tion of signaling modules would result in the integrity of conserved D.E., Balavoine, G., Arendt, D., 2007. Molecular architecture of annelid nerve cord embryonic anatomical patterns, as the possible developmental supports common origin of nervous system centralization in bilateria. Cell 129, bases for Geoffroy's principle of connectivity. As far as this type 277–288. Depew, M.J., Lufkin, T., Rubenstein, J.L., 2002. Specification of jaw subdivisions by Dlx of regulatory network is trapped epigenetically, gene expression genes. Science 298, 371–373. patterns will serve as cues for morphological homologization. Dollé, P., Izpisúa-Belmonte, J.C., Falkenstein, H., Renucci, A., Duboule, D., 1989. However, as we have seen, this type of network can also result in a Coordinate expression of the murine Hox-5 complex homoeobox-containing genes during limb . Nature 342, 767–772. large-scale heterotopy of morphological modules with minor shifts Duboule, D., 1994. Temporal colinearity and the phylotypic progression: a basis for the of developmental modules (as in the case of jaw evolution in stability of a vertebrate Bauplan and the evolution of morphologies through gnathostomes), as suggested by their effects in the scheme of the heterochrony. Dev. Suppl. 1994, 135–142. ‘ ’ Echevarría, D., Vieira, C., Gimeno, L., Martínez, S., 2003. Neuroepithelial secondary epigenetic landscape proposed by Waddington (1975). organizers and cell fate specification in the developing brain. Brain Res. Brain Res. By defining morphological and developmental modules sepa- Rev. 43, 179–191. rately and by realizing several different factors to maintain these Ferguson, C.A., Tucker, A.S., Sharpe, P.T., 2000. Temporospatial cell interactions modules through evolution, we will be able to avoid confusion in the regulating mandibular and maxillary arch patterning. Development 127, 403–421. concepts of homologies. Morphological homologies are statements Fürbringer, M., 1897. Über die spino-occipitalen Nerven der Selachier und Holocephalen about relationships among developmental–morphological module ind ihre vergleichende Morphologie. Festschr. für Carl Gegenbaur 3, 349–788. complexes, which form a phylotype-like morphological network as Galis, F., Sinervo, B., 2003. Conserved early embryonic stages. In: Hall, B.K., Olson, W.M. fi (Eds.), Keywords and Concepts in Evolutionary Developmental Biology. Harvard exempli ed by Geoffroy. However, they do not always represent Univ. Press, Cambridge, pp. 43–52. structural entities with identical functions. Homologization of these Garcia-Fernàndez, J., 2005. The genesis and evolution of homeobox gene clusters. Nat. modules should be done primarily between the equivalent module Rev. Genet. 6, 881–892. Gegenbaur, C., 1887. Die Metamerie des Kopfes und die Wirbeltheorie des Kopfskelets. complexes belonging to the latter networks, according to compara- Morphol. Jb. 13, 1–114. tive morphological strategies. Molecular or developmental homo- Gegenbaur, C., 1898. Vergleichende Anatomie der Wirbeltihiere mit Berücksichtung der logies are for developmental modules that are conserved at various Wirbellosen. Verlag von Wilhelm Engelmann, Leipzig. Gendron-Maguire, M., Mallo, M., Zhang, M., Gridley, T., 1993. Hoxa-2 mutant mice levels for various reasons. Developmental and morphological exhibit homeotic transformation of skeletal elements derived from cranial neural modules together constitute our concept of the body plan, which crest. Cell 75, 1317–1331. also has evolved in a hierarchical manner with time. Therefore, Geoffroy Saint-Hilaire, E., 1818. Philosophie Anatomique (tome premiere). J.B. Baillière, fi fi Paris. identi cation of modules in their correct context should be the rst Gilbert, SF., 2006. Developmental Biology, 7th Ed. Sinauer Associates, Inc, Sunderland, task for evo-devo studies. Massachusetts. Goodrich, E.S., 1930. Studies on the Structure and Development of Vertebrates. Acknowledgments McMillan, London. Haeckel, E., 1874. Anthropogenie oder Entwickelungsgeschichte des Menschen. Keimes- und Stammesgeschichte. Wilhelm Engelmann, Leipzig. I thank Takao Suzuki for critical reading of the manuscript and Haeckel, E., 1875. Die Gastrea und die Eifurchung der Thiere. Jena Z. Naturwiss. 9, – Yuko Hirofuji for preparation of the figures. 402 508. Haeckel, E., 1902. Natürliche Schöpfungs-Gechichte-Gemeinverständliche wissenschaf- tliche Verträge über die Entwickelungs-Lehre. Druck und Verlag von Georg Reimer, Berlin. References Hall, B.K., 1994. Homology: The Hierarchical Basis of Comparative Biology. Academic Press, San Diego. Appel, T.A., 1987. The Cuvier-Geoffroy Debate: French Biology in the Decades before Hall, B.K., 1998. Evolutionary Developmental Biology, 2nd Ed. Chapman & Hall, London. Darwin (Monographs on the History and Philosophy of Biology). Oxford Univ. Press, Holland, N.D., Holland, L.Z., Honma, Y., Fujii, T., 1993. expression during New York. development of a lamprey, Lampetra japonica: a possible clue to homologies S. Kuratani / Developmental Biology 332 (2009) 61–69 69

between agnathan and gnathostome muscles of the mandibular arch. Dev. Growth Ohya, Y.K., Kuraku, S., Kuratani, S., 2005. Hox code in embryos of Chinese soft-shelled Differ. 35, 153–160. turtle Pelodiscus sinensis correlates with the evolutionary innovation in the turtle. Holzschuh, J., Wada, N., Wada, C., Schaffer, A., Javidan, Y., Tallafuss, A., Bally-Cuif, L., J. Exp. Zool. (Mol. Dev. Evol.) 304B, 107–118. Schilling, T.F., 2005. Requirements for endoderm and BMP signaling in sensory Owen, R., 1848. On the Archetype and Homologies of the Vertebrate Skeleton. London. neurogenesis in zebrafish. Development 132, 3731–3742. Raff, R.A., 1996. The Shape of . The University of Chicago Press, Chicago. Horigome, N., Myojin, M., Hirano, S., Ueki, T., Aizawa, S., Kuratani, S., 1999. Development Rickmann, M., Fawcett, J., Keynes, R.J., 1985. The migration of neural crest cells and the of cephalic neural crest cells in embryos of Lampetra japonica, with special growth of motor axons through the rostral half of the chick somite. J. Embryol. Exp. reference to the evolution of the jaw. Dev. Biol. 207, 287–308. Morphol. 90, 437–455. Janvier, P., 1996. Early Vertebrates. Oxford Scientific Publications, New York. Rijli, F.M., Mark, M., Lakkaraju, S., Dierich, A., Dollé, P., Chambon, P., 1993. Homeotic Kessel, M., 1992. Respecification of vertebral identities by retinoic acid. Development transformation is generated in the rostral branchial region of the head by 115, 487–501. disruption of Hoxa-2, which acts as a selector gene. Cell 75, 1333–1349. Kessel, M., Galling, R., Gruss, P., 1990. Variations of cervical vertebrate after expression of De Robertis, E., 2008. Evo-devo: variations on ancestral themes. Cell 132, 185–195. a Hox-1.1 transgene in mice. Cell 61, 301–308. Rodriguez-Esteban, C., Tsukui, T., Yonei, S., Magallon, J., Tamura, K., Izpisua-Belmonte, Keynes, R.J., Stern, C.D., 1984. Segmentation in the vertebrate nervous system. Nature J.C., 1999. The T-box genes Tbx4 and Tbx5 regulate limb outgrowth and identity. 310, 786–789. Nature 398, 814–818. Klingenberg, C.P., 2008. Morphological integration and developmental modularity. Ann. Schlosser, G., Wagner, G.P., 2004. Modularity—in development and evolution. Univ. Ecol. Evol. Syst. 39, 115–132. Chicago Press, Chicago. Kuratani, S., 2003. Evolutionary developmental biology and vertebrate head segmenta- Scholtz, G., 2005. Homology and : pattern and process in comparative tion: a perspective from developmental constraint. Theory Biosci. 122, 230–251. developmental biology. Theory Biosci. 124, 121–143. Kuratani, S., 2004. Evolution of the vertebrate jaw: comparative embryology reveals the Schwenk, K., Wagner, G.P., 2003. Constraint. In: Hall, B.K., Olson, W.M. (Eds.), Keywords developmental factors behind the evolutionary novelty. J. Anat. 205, 335–347. and Concepts in Evolutionary Developmental Biology. Harvard Univ. Press, Kuratani, S., 2005. Cephalic crest cells and evolution of the craniofacial structures in Cambridge, pp. 52–61. vertebrates: morphological and embryological significance of the premandibular- Shigetani, Y., Nobusada, Y., Kuratani, S., 2000. Ectodermally-derived FGF8 defines the mandibular boundary. Zoology 108, 13–26. maxillomandibular region in the early chick embryo: epithelial–mesenchymal Kuratani, S., 2008. Is the vertebrate head segmented?—evolutionary and developmental interactions in the specification of the craniofacial ectomesenchyme. Dev. Biol. 228, considerations. Integ. Comp. Biol. 48, 647–657. 73–85. Kuratani, S.C., Kirby, M.L., 1991. Initial migration and distribution of the cardiac neural Shigetani, Y., Sugahara, F., Kawakami, Y., Murakami, Y., Hirano, S., Kuratani, S., 2002. crest in the avian embryo: an introduction to the concept of the circumpharyngeal Heterotopic shift of epithelial–mesenchymal interactions for vertebrate jaw crest. Am. J. Anat. 191, 215–227. evolution. Science 296, 1316–1319. Kuratani, S.C., Kirby, M.L., 1992. Migration and distribution of the circumpharyngeal Shigetani, Y., Sugahara, F., Kuratani, S., 2005. Evolutionary scenario of the vertebrate crest cells in the avian embryo: formation of the circumpharyngeal ridge and E/C8+ jaw: the heterotopy theory from the perspectives of comparative and molecular crest cells in the vertebrate head region. Anat. Rec. 234, 263–280. embryology. BioEssays 27, 331–338. Kuratani, S., Ota, G.K., 2008. The primitive versus derived traits in the developmental Takeuchi, J.K., Koshiba-Takeuchi, K., Matsumoto, K., Vogel-Hopker, A., Naitoh-Matsuo, program of te vertebrate head: views from cyclostome developmental studies. M., Ogura, K., Takahashi, N., Yasuda, K., Ogura, T., 1999. Tbx5 and Tbx4 genes J. Exp. Zool. (Mol. Dev. Evol.) 310B, 294–314. determine the wing/leg identity of limb buds. Nature 398, 810–814. Kuratani, S., Horigome, N., Hirano, S., 1999. Developmental morphology of the cephalic Takio, Y., Pasqualetti, M., Kuraku, S., Hirano, S., M., Rijli, F., Kuratani, S., 2004. Lamprey mesoderm and re-evaluation of segmental theories of the vertebrate head: Hox genes and the evolution of jaws. Nature OnLine 429, 1 p following 262. http:// evidence from embryos of an agnathan vertebrate, Lampetra japonica. Dev. Biol. www.nature.com/cgi-taf/DynaPage.taf?file=/nature/journal/v429/n6989/full/ 210, 381–400. nature02616_fs.htm006C. Kuratani, S., Nobusada, Y., Horigome, N., Shigetani, Y., 2001. Embryology of the lamprey Takio, Y., Kuraku, S., Kusakabe, R., Murakami, Y., Pasqualetti, M., Rijli, F.M., Narita, Y., and evolution of the vertebrate jaw: insights from molecular and developmental Kuratani, S., Kusakabe, R., 2007. Hox gene expression patterns in Lethenteron perspectives. Phil. Trans. Roy. Soc. 356, 15–32. japonicum embryos insights into the evolution of the vertebrate Hox code. Dev. Biol. Kuratani, S., Murakami, Y., Nobusada, Y., Kusakabe, R., Hirano, S., 2004. Developmental 308, 606–620. fate of the mandibular mesoderm in the lamprey, Lethenteron japonicum: Tucker, A.S., Matthews, K.L., Sharpe, P.T., 1998. Transformation of tooth type induced by comparative morphology and development of the gnathostome jaw with special inhibition of BMP signaling. Science 282, 1136–1138. reference to the nature of trabecula cranii. J. Exp. Zool. (Mol. Dev. Evol.) 302B, Uchida, K., Murakami, Y., Kuraku, S., Hirano, S., Kuratani, S., 2003. Development of the 458–468. adenohypophysis in the lamprey: evolution of the epigenetic patterning programs Larsson, H.C.E., Wagner, G.P., 2002. Pentadactyl ground state of the avian wing. J. Exp. in organogenesis. J. Exp. Zool. (Mol. Dev. Evol.) 300B, 32–47. Zool. (Mol. Devel. Evol.) 294, 146–151. von Baer, K.E., 1828. Entwicklungsgeschichte der Thiere: Beobachtung und Reflexion. Lee, S.H., Bédard, O., Buchtová, M., Fu, K., Richman, J.M., 2004. A new origin for the Born Träger, Königsberg. maxillary jaw. Dev. Biol. 276, 207–224. Wada, N., Javidan, Y., Nelson, S., Carney, T.J., Kelsh, R.N., Schilling, T.F., 2005. Lowe, C.J., Wu, M., Salic, A., Evans, L., Lander, E., Stange-Thomann, N., Gruber, C.E., Hedgehog signaling is required for cranial neural crest morphogenesis and Gerhart, J., Kirschner, M., 2003. Anteroposterior patterning in hemichordates and chondrogenesis at the midline in the zebrafish skull. Development 132, the origins of the chordate nervous system. Cell 113, 853–865. 3977–3988. Mallatt, J., 2008. The origin of the vertebrate jaw: neoclassical ideas versus newer, Waddington, C.H., 1975. The Evolution of an Evolutionist. Cornell Univ. Press, Ithaca, development-based ideas. Zool. Sci. 25, 990–998. New York. Matsuura, M., Nishihara, H., Kokubo, N., Kuraku, S., Kusakabe, R., Okada, N., Kuratani, S., Wagner, G.P., 1989a. The origin of morphological characters and the biological basis of Tanaka, M., 2008. Identification of four Engrailed genes in the Japanese river homology. Evolution 43, 1157–1171. lamprey, Lethenteron japonicum. Dev. Dyn. 237, 1581–1589. Wagner, G.P., 1989b. The biological homology concept. Annu. Rev. Ecol. Syst. 20, 51–60. Maynard-Smith, J., Burian, R., Kauffman, S., Alberch, P., Campbell, J., Goodwin, B., Lande, Wagner, G.P., 1994. Homology and the mechanisms of development. In: Hall, B.K. (Ed.), R., Raup, D., Wolpert, L., 1985. Developmental constraints and evolution. Quart. Rev. Homology: The Hierarchical Basis of Comparative Biology. Academic Press, San Biol. 60, 265–287. Diego, pp. 273–299. Müller, G.B., Wagner, G.P., 1991. Novelty in evolution: restructuring the concept. Annu. Wagner, G.P., 2007. The developmental genetics of homology. Nat. Rev. 8, 473–479. Rev. Ecol. Syst. 22, 229–256. Wagner, G.P., Altenberg, L., 1996. Perspective: complex and the evolution of Murakami, Y., Ogasawara, M., Satoh, N., Sugahara, F., Myojin, M., Hirano, S., Kuratani, S., . Evolution 50, 967–976. 2002. Compartments in the lamprey embryonic brain as revealed by regulatory Wagner, G.P., Müller, G.B., 2002. Evolutionary innovations overcome ancestral gene expression and the distribution of reticulospinal neurons. Brain Res. Bull. 57, constraints: a re-examination of character evolution in male sepsid flies (Diptera: 271–275. Sepsidae). Evol. Dev. 4, 1–6. Naiche, L.A., Papaioannou, V.E., 2007. Tbx4 is not required for hindlimb identity or post- Wagner, G.P., Pavlicev, M., Cheverud, J.M., 2007. The road to modularity. Nat. Rev. 8, bud hindlimb outgrowth. Development 134, 93–103. 921–931. Neidert, A.H., Virupannavar, V., Hooker, G.W., Langeland, J.A., 2001. Lamprey Dlx genes Woodger, J.H., 1945. On biological transformations. In: Le Gros Clark, W.E., and early vertebrate evolution. Proc. Natl. Acad. Sci. U. S. A. 98, 1665–1670. Medawar, P.B. (Eds.), Essays on Growth and Form Presented to D'Arcy Neubüser, A., Peters, H., Balling, R., Martin, G.R., 1997. Antagonistic interactions between Thompson. Cambridge Univ. Press, Cambridge, pp. 95–120. FGF and BMP signalling pathways: a mechanism for positioning the site of tooth Yokouchi, Y., Sasaki, H., Kuroiwa, A., 1991. Homeobox gene expression correlated with formation. Cell 90, 247–255. the bifurcation process of limb cartilage development. Nature 353, 443–445.