Genes, forces, and forms: Mechanical aspects of prenatal craniofacial development

Abstract

Current knowledge of molecular signaling during craniofacial development is advancing rapidly. We know that cells can respond to mechanical stimuli by biochemical signaling. Thus, the link between mechanical stimuli and gene expression has become a new and important area of the morphological sciences. This field of research seems to be a revival of the old approach of developmental mechanics, which goes back to the embryologists His (1874), Carey (1920), and Blechschmidt (1948). These researchers argued that forces play a fundamental role in tissue differentiation and morphogenesis. They understood morphogenesis as a closed system with living cells as the active part and biological, chemical, and physical laws as the rules. This review reports on linking mechanical aspects of developmental biology with the contemporary knowledge of tissue differentiation. We focus on the formation of cartilage (in relation to pressure), bone (in relation to shearing forces), and muscles (in relation to dilation forces). The cascade of molecules may be triggered by forces, which arise during physical cell and tissue interaction. Detailed morphological knowledge is mandatory to elucidate the exact location and timing of the regions where forces are exerted. Because this finding also holds true for the exact timing and location of signals, more 3D images of the developmental processes are required. Further research is also required to create methods for measuring forces within a tissue. The molecules whose presence and indispensability we are investigating appear to be mediators rather than creators of form. Developmental Dynamics 235:1219–1229, 2006. © 2006 Wiley-Liss, Inc.

INTRODUCTION

Molecules, Tissues, and Forces

Our current knowledge of molecular signaling during craniofacial development is advancing rapidly (Thorogood,1993; Thorogood et al.,1998; Depew et al.,2002; Francis-West et al.,2003). The homeobox code has been shown to be crucial in determining the polarity of the head and the patterning of the arrangement of the anatomical features that make up the craniofacial complex (Sharpe,1995). There is, however, a deep gap between the rapid progress in elucidating the regulative interdependencies within molecular cascades and the availability of detailed knowledge on the morphological level (Radlanski,2003).

Research reports focusing on molecular signaling often exclude the morphological aspect. However, no tissue grows alone; thus, the neighborhood of tissues may be the decisive factor in creating organic form. It is obvious that these changes also involve forces and mechanical interactions of the neighboring tissues. This aspect has been virtually neglected in recent years.

MECHANICS AND EMBRYOLOGY

The discussion has now been reopened on an interdependency between mechanical stimuli, molecular signals, and tissue differentiation (Henderson and Carter,2002; Benjamin and Hillen,2003; Brouzes and Farge,2004; Curtis,2005; de la Fuente and Helms,2005). It was as early as in 1874 when Wilhelm His sen. (His,1874) explained our body form as a consequence of mechanical forces exerted by tissues growing in a confined space. Eben J. Carey was also convinced that forces act during morphogenesis when he explained the formation of the annular muscles of the colon and esophagus (Carey,1920a, b, 1935).

In the 1950s, Blechschmidt created a collection of three-dimensional (3D) reconstructions (Blechschmidt,1954) of human embryos in Göttingen (Germany), which is still on display in large showcases (Fig. 1). This enabled him to compare the different stages of development at the same scales. He was very specific in describing growth changes in terms of local increments of tissue volume, displacement of tissue, local recession of tissue mass, invagination, and evagination. From his observations, he drew conclusions on the differential growth of different tissues in different regions. He found that expansion in one region was accompanied by the formation of curvatures, folds, bulges, brims, and crimples in adjacent or even distant regions (Blechschmidt,1960). Some cells or tissues were in a favorable condition and continued to grow, whereas others were displaced or were confined and compressed within a limited space. He described these ontogenetic processes as developmental movements of the tissues. Blechschmidt examined local changes on the level of cellular differentiation for a correlation with the overall changes of body shape, and he established general embryological rules (Blechschmidt,1948). For example, he found that cartilage develops in regions of dense mesenchyme (densation fields), whereas bone arises in regions of detraction and muscles in regions of dilation (Blechschmidt,1978, 2004).

Figure 1. The Blechschmidt Collection of Human Embryos in the Anatomical Institute of Göttingen University (Germany).

The mechanical view of developmental anatomy suggests that forces play a fundamental role in tissue differentiation and morphogenesis. Blechschmidt (1948) could only assume molecular interactions, as they have only been demonstrated in recent years. We now know that cartilage is formed under the influence of SOX9 (Akiyama et al.,2002), which is found in regions of pressure, and that bone is formed under the influence of RUNX2 (Mundlos et al.,1997), which is found in regions of tensile strain (Carter et al.,1998; de la Fuente and Helms,2005). We think the return to this view of developmental mechanics, including forces, cell forms, and their location, will enable better understanding of the morphological aspects of craniofacial development.

OUTCOME OF GROWTH

When cells grow, and particularly when they increase in number, they cannot stay in place and, thus, inevitably will exert an influence on adjacent cells and on the extracellular matrix in their neighborhood. This extracellular matrix is viscoelastic and will be distorted by the movement of the cells it contains. Distortion of the extracellular matrix may trigger further cell movement. Cells can respond in terms of chemotaxis and galvanotaxis. They can also move down a density gradient (diffusion), move up an adhesive gradient (haptotaxis), be guided and inhibited by contact, or just float away (convection; Murray,2000). Monolayer in vitro cultures showed that chemotactic responsiveness was highly dependent on cell density (Gormar et al.,1990; Szabo et al.,2001).

Response to mechanical stimulation is a basic biological phenomenon. Nearly all cells process mechanical input and respond to it by inducing and modulating biochemical pathways. If the average mechanical load is increased in tissues, some tissues can increase their performance and often increase their bulk by cell division (Jones et al.,1995). There is more detailed evidence that cells can respond to mechanical forces (Chen and Grinnell,1997). Cells that are stretched proliferate in vitro (Kippenberger et al.,1999, 2000). Investigations on the molecular level revealed that mechanical forces at the interface between the cell and the extracellular matrix can activate stretch-induced changes in ion channels. Cell contractions at adherence junctions activate cell membrane-associated secondary messenger pathways and lead to growth -factor-like activities that influence cellular proliferation and protein synthesis (Silver et al.,2003). The diversity of responses reflect the genotype of the cell and the mechanical demands of the resident tissue (Banes et al.,1995).

The link between mechanical stimulus and gene expression represents a new and important area of morphological sciences. Stretch is an important mechanical signal for the production of more actin and myosin filaments and the addition of new sarcomeres in series and in parallel (Goldspink,1999). That mechanical aspect as one of the main factors responsible for creating organic form can be found in the older anatomic literature (His,1874; Carey,1920b; Blechschmidt,1948). Blechschmidt regarded the whole of human morphogenesis as a closed system with living cells as the players and biological, chemical, and physical laws as the rules. He was convinced he could explain how the form of the whole body arises and how tissues differentiate by mechanical forces.

Tissue Differentiation as a Consequence of Unequal Growth

Numerous varying conditions of tissue differentiation have been described (Blechschmidt,1960). Three examples will be given to illustrate the special features of the regions in which cartilage, bone, and muscles form, as viewed by Blechschmidt (1948, 2004).

Cartilage Formation

Cartilaginous formation of early skeletal components is a good example of outside–inside differentiation: The outer form influences what happens inside. Deeper-seated mesenchymal tissue undergoes a loss of water from the intercellular substance (Blechschmidt,1948). This loss is because, in that early stage, the outer cells have better access to nutrients, grow better, and put pressure on the central cells. The mesenchymal cells, thus, are condensed. Cells in these so-called densation fields can be identified as precartilage cells. These condensed cells, however, impede themselves more and more in terms of cellular uptake and release of molecular metabolites the larger the densation field becomes. As a consequence, these cells develop a high osmotic pressure, because their intracellular catabolites have a high molecular weight. As water now starts to flow into these cells from their surroundings, they differentiate into globular chondrocytes. Because these cells are lined up in the center of a mesenchymal bulge, they assume a longitudinal arrangement. This structure also extends in a longitudinal direction, termed “Stemmkörperfunktion” in German, which can be translated as “piston-like function,” as suggested by Freeman (Blechschmidt,2004). This finding is a general process that takes place during early skeletal development of the arms, legs, ribs, and spinal column. Meckel's cartilage in the first visceral arch also arises in this way. Further growth stretches not only the cartilaginous formation itself but also the adjacent tissue by exerting forces as it pushes on either end (pressure). This push gives rise to a shearing force.

Bone Formation (Desmal Ossification)

Desmal ossification is observed when an expanding cartilaginous core slides against its surrounding tissue, thus exerting a shearing force. This process occurs during formation of the mandibular bone next to Meckel's cartilage and the long bones of the arms and legs, and it also applies to the frontal and parietal bones, when expanding brain tissue slides against the overlying skin. Blechschmidt (2004) called these regions where expanding tissues slide against each other “detraction fields.”

Muscle Formation

Possibly the most striking example of embryonic development as developmental movements of tissues against each other is the orientation of muscle fibers. It is apparent that all the arm and leg muscles run in a longitudinal direction, reflecting the main direction of the skeletal components underneath. There is one muscle in the arm, however, the quadrate pronator muscle, which runs transversely, stretched out between the ulna and radius. It should be obvious that the orientation of the muscles reflects the direction of growth of the underlying skeletal elements, in this case, the moving apart of the ulna and radius. If we go through the whole body, we always find a correlation between the muscle fiber direction and the developmental movements of the underlying skeletal components. For example, the autochthonous muscles of the vertebrae reflect the vertical, transverse, and oblique expansion of each vertebra. The trapezius muscle shows the expansion of the shoulder area in a transverse direction as well as the vertical expansion of the upper vertebral column (Fig. 2). Also, if we superimpose the direction of the masticatory muscles, we may as well draw the growth vectors of the enlarging skull (Figs. 3, 4A,B).

Figure 2. Diagram of the trapezius muscle, right half. The orientation of the muscle fibers reflects vertical growth of the cervical, thoracic, and lumbar spine and transverse growth of the shoulder (arrows). Taken from Blechschmidt,2004, with permission of North Atlantic Books.

Figure 3. Diagram of the head and neck region. The orientation of the sternocleidomastoid muscle reflects growth of the cervical spine. The increasing angle of the mandible (arrow) during the descent of the viscera is concomitant with the formation of the masseter and the medial pterygoid muscles. Taken from Blechschmidt,1960, with permission of Karger, Basel.

Figure 4. A,B: The muscles of mastication (B) take on the same orientation as the growth vectors of the skull (A). The image in B (without arrows) is taken from (Sobotta,1993) with permission of Urban and Schwarzenberg, Munich.

Blechschmidt (2004) drew the conclusion that muscles arise in regions of dilation. Muscle fibers arise wherever mesenchymal tissue is available that can be dilated by adjacent expanding tissue (Carey,1920a, b, 1921, 1936).

Folding, Molding, and Cell Form as Consequences of Unequal Growth

Developmental movements of tissues lead to unequal growth. The consequent forces created lead to differentiation of tissues. Unequal growth was held responsible for the creation of form during embryogenesis (His,1874; Carey,1920b; Blechschmidt,2004). Different growth rates of adjacent tissues are thought to cause folding and bulging.

As a consequence, there are no really flat surfaces in the human body. True cuboidal or columnar epithelial cells are rarely found, except in textbook diagrams (Blechschmidt,2004).

When an epithelium expands in a confined space, it can bulge upward or downward and invaginate into the mesenchyme. In doing so, the cells at the margin of the curvature become wedge-shaped. This epithelial tissue is thus called a “wedged epithelium” (Blechschmidt,1948, 1955, 1960, 2004). It is obvious that the cells exert mutual pressure in a lateral direction, which leads to bulging. Typical wedge epithelia are dental lamina, the buds of the dental primordia (Fig. 5A–C), or the acini of glands. In other regions, e.g., the colon or esophagus, unequal growth leads to stretching of the outer cell layers (Carey,1920b).

Figure 5. A: Horizontal section through the region of the dental lamina of a human embryo of 21-mm crown–rump length (CRL; Carnegie stage 20). The numbers 1–3 indicate the epithelial condensations of the dental primordia at the bud stage. Hematoxylin–eosin (HE) stained. VL, vestibular lamina; OC, oral cavity; T, tongue. B: At the bulges of the tooth buds, a wedge epithelium (W) dominates, while the epithelium of the dental lamina in the regions between the tooth buds shows a cuboidal (C) form. HE stained. C: A three-dimensional reconstruction of the mandibular region of the same embryo of 21-mm CRL (Carnegie stage 20). The vestibular lamina (Lv) and the dental lamina (green) with the dental primordia in their bud stages (i₁, i₂, c, m₂) grow in a space confined anteriorly by the lip and posteriorly by Meckel's cartilage (CM). Mand, mandible. Taken from Radlanski,1993, with permission of Quintessence Publishing Group. Scale bars = 500 μm in A,C, 200 μm in B.

CRANIOFACIAL GROWTH

In craniofacial development, a typical example of active morphogenetic forces is flexure of the head during early embryonic development. Rapid growth of neuronal tissue and the expanding anlage of the brain in the cranial region cause the early head to bend forward. The tissue in the facial region must now grow in a confined space between the anlagen of the forebrain and the heart (Fig. 6A,B). Thus, the early human face is characterized by a series of bulges that run predominantly in an horizontal direction. We identify the maxillary bulge, the first visceral arch, which will give rise to the mandible, and further below, we see the subsequent visceral arches.

Figure 6. A: Human face of an embryo of 16-mm crown–rump length (CRL; Carnegie stage 19). The face is characterized by bulges and furrows running predominantly in a horizontal direction. B: Thus, they reflect pressure exerted by the expanding brain (b) and the expanding heart (h), with the face in between. Taken from Blechschmidt,1960, with permission of Karger, Basel.

According to the embryological rules of Blechschmidt (2004), we should find a mesenchymal, and later on, a cartilaginous condensation in the center of these arches. The nasal capsule arises in the maxillary arch, and cartilaginous centers in the first, second, and subsequent visceral arches give rise to Meckel's cartilage and the anlagen of the hyoid bone and larynx. These cartilaginous brackets, following their orientation in early stages, reflect the outer form and extension of the bulge of the visceral arch. This finding corroborates Blechschmidt's principle of outside–inside differentiation (Blechschmidt,2004).

When we compare earlier stages of craniofacial morphogenesis (Fig. 7A) with later stages (Fig. 7B), it becomes obvious that the facial region is initially much smaller and that the expansion of the brain tissue dominates. We may draw conclusions as to forces compressing the facial region while observing growth proportions at different stages. Densated mesenchymal tissue in the center of the bulges gives rise to precartilaginous cell condensations. Cartilage has the ability to expand rapidly, while maintaining the basic form. In this stage of development, the cartilaginous structures take control, and while expanding, they counteract the compression force that originally formed them.

Figure 7. Outlines of the head of a human embryo of 21-mm crown–rump length (CRL; Carnegie stage 20) and a fetus of 40-mm CRL. Blue, cartilaginous skeletal components of the craniofacial region. Brown, bone. Yellow, brain. Black, contour of the head. A: The development of the expanding brain (arrow) leads to a spatial impediment of the tissue within the visceral arches, in whose centers the cartilaginous skeleton forms. B: Later, facial height, length, and width are increased by rapid expansion of the cartilaginous structures, predominantly Meckel's cartilage and the nasal capsule (arrows). Thus, the cartilaginous parts of the facial skeleton reverse the action that led to their formation.

It is undisputed that the mesenchymal cells in the center of the visceral arches live in different environmental conditions than those close to the epithelial surface. Light microscopy shows increased tissue density in the center of the bulges, but a pressure increase in the center would be difficult to measure. However, the facial region straightening and enlarging relative to the brain region presumes pressure factors are operating.

Formation of the Mandible

When Meckel's cartilage develops into mature expanding cartilage, it moves along its longitudinal axis, increasing rapidly in anteroposterior length (Radlanski et al.,1994). A triple curvature of Meckel's cartilage occurs, which gives it a swinging outline (Kjaer et al.,1999; Radlanski et al.,2003). This wavy outline of Meckel's cartilage can be interpreted as a reaction to further elongation against a limiting outer skin. At its anterior end, the mesenchymal tissue clearly appears compressed and the cells are longitudinally arranged with the appearance of being floated away (Fig. 8). Taken together, these findings suggest a force originating from the longitudinal expansion. Although there are many morphological indications of forces molding cells and tissues, we would only obtain satisfactory evidence if they could be measured in vivo.

Figure 8. Vertical, parasagittal section through the mandibular region of a human embryo, 37-mm crown–rump length (CRL). The enlarged cells of the expanding Meckel's cartilage (MC) cause the surrounding mesenchymal cells to float around its anterior end. i₁, dental primordium of the deciduous central incisor; VL, vestibular lamina; M, early formation of mandibular bone. Scale bar = 100 μm.

The first bone formations of the mandible occur at the buccal sides of Meckel's cartilage (Fig. 9A). Our 3D reconstructions of the early development of the mandible (Radlanski et al.,2003) show that this region may well be characterized by a sliding action of the elongating Meckel's cartilage against the skin with the mesenchymal tissue in between being sheared (Fig. 9B–D). According to Blechschmidt (2004), mesenchymal tissues form bone in these regions of detraction. Anterior and posterior expansion of the mandible along Meckel's cartilage proceeds as long as shearing forces continue to be exerted there. Also, the encroachment of bony tissue around Meckel's cartilage with its partial embrasure is concomitant with detraction of the mesenchyme caused by expansion of the cartilaginous frame.

Figure 9. A: The computer-aided three-dimensional reconstruction shows a human embryo of 21-mm crown–rump length (CRL; Carnegie stage 20) with the components of the craniofacial region made visible through the transparent skin. B–D: Meckel's cartilage, mandibular bone, and the alveolar inferior nerve in a human embryo of 21 mm CRL (B), 25-mm CRL (Carnegie stage 22; C), and 66 mm CRL (D), seen from an anterior, lateral, 45 degree oblique view. The image in A is taken from Radlanski, (2003), with permission of Blackwell Munksgaard, Copenhagen, and those in B–D are taken from Radlanski et al., (2003), with permission of Springer Verlag, Berlin. Scale bar = 1,000 μm.

During these developmental stages, the mandible alters its shape. It elongates anteroposteriorly, while expanding transversally. The gonial angle and the ramus of the mandible form, when the whole embryo straightens. Muscle tissue is formed from mesenchymal tissue that is not detracted but dilated under the direct mechanical influence of the expanding mandible (Carey,1920b; Blechschmidt,1948, 2004). The arrangement of muscles of the floor of the mouth during early prenatal stages (Fig. 10A,B) exactly reflects these changes in size and form (Radlanski et al.,2001). Thus, fibers of the mylohyoid muscle extend in a transversal direction as the mandible becomes wider; the geniohyoid and digastric (anterior belly) muscles reflect the forward thrust of the mandible, and the genioglossus muscle bundles reflect the vertical stretch with the increase of the tongue. Blechschmidt (1948, 2004) was convinced that the epithelium is one of the driving forces that create form and that the underlying mesenchyme is under the influence of the developing movements (and forces) of the epithelium. The increased size and the formation of the tongue, rising from the floor of the oral cavity as a protrusion, thus, is dependent on the molding function of the bulging epithelium.

Figure 10. A,B: Mandibular region with the muscles of the floor of the mouth in a human embryo of 25-mm crown–rump length (CRL; A), cranial view, and a human fetus of 68 mm CRL (B), viewed from a posterior direction. The orientation of the muscle fibers matches the growth pattern of the mandible, which is in a sagittal, transverse, and vertical direction. Taken from Radlanski et al., (2001), with permission of Gustav-Fischer-Verlag, Jena. Scale bar = 500 μm.

Formation of the Oral Cavity

An embryo with a CRL of 14 mm (early sixth week, Carnegie stage 18) has an oronasal cavity with a gentle in- and out-contour of the epithelium lining the oronasal capsule (Fig. 11A). The epithelium grows faster than the oronasal cavity itself, and the spatial impediment causes elevations and invaginations to increase and become more accentuated (Fig. 11B). As soon as bulging alleviates the spatial impediment, the epithelial bulges expand into the oronasal cavity or the underlying mesenchymal tissue. The central elevation is the tongue, which expands in vertical, transverse, and anteroposterior directions. The mesenchyme is dilated, thus giving rise to the formation of the intralingual muscles in corresponding directions. On either side of the tongue, the epithelium grows inward, leading to formation of the sublingual and the submandibular glands. Further laterally, another invagination of the oral epithelium leads to formation of the dental and vestibular laminae. At the buccal corner, another epithelial duct turns inward for the parotid gland.

Figure 11. A–C: Histological frontal sections through the oral cavity of human embryos of 14-mm crown–rump length (CRL; Carnegie stage 18; A) and 26-mm CRL (Carnegie stage 22; B), and a fetus of 68-mm CRL (C). In B, the lines lead to the primordium of the nasal septum (a), to Meckel's cartilage (b), to the mandibular bone (c), to the masseter muscle (d), to the mylohyoid muscle (e), and to the fusion of the mylohyoid muscles (f). A and B are taken from Blechschmidt, (1960), with permission of Karger, Basel, and C is from the Radlanski collection. Scale bar = 1 mm.

In the maxillary region, another band of invagination forms the upper dental and vestibular lamina. Palatal processes arise through further expansion of the epithelium that turns toward the oronasal cavity. In addition, a perpendicular epithelial fold at the top of the oronasal cavity gives rise to the nasal septum. Again, the epithelium grows faster than the cavity, so it bulges again and forms the prominences of the nasal conchae. Also, the lowest prominences, the palatal processes, become larger. As soon as the embryo stretches, when vertical growth increases in the neck (mediated by expansion of the cervical vertebral column), the infrahyoidal muscles arise as well as the gonial angle, and the tongue is pulled away from the oronasal cavity. Further growth and the elasticity of the palatal processes (Ferguson,1981) allow them to swing up and fuse in the center of the palate (Fig. 11C).

Location of Maxillary Bone Formation

There are also indications that Blechschmidt's concepts of developmental mechanics may apply in the maxillary region, although he himself did not describe them. In the center of the (epithelial) nasal septum, the mesenchymal tissue is likely to be condensed in such a way that a “condensation field” gives rise to a cartilaginous septum that expands according to cartilage behavior in a vertical and anteroposterior direction. This process definitely contributes toward enlarging the midfacial region, and it also shears the mesenchymal tissue next to it and under the septal epithelium. This determines the bone-forming region. Further condensation centers developing in the region of the future nasal conchae contribute toward the formation of the nasal capsule. This in turn adds to expansion of the midface in corresponding directions. Furthermore, the region of initial maxillary bone formation can be characterized as detraction fields of the expanding nasal capsule, eye bulbs, and oral epithelium shearing the remaining mesenchyme. The early maxilla, as shown by 3D reconstructions (Radlanski et al.,2000), has the form of a small pyramid, wedged exactly into the remaining space just described (Fig. 12A,B). The adjacent mesenchyme shows signs of flowing motion (possible shearing forces) parallel to the expanding organs mentioned.

Figure 12. A: Maxillary bone formation (ochre) in a human fetus of 54-mm crown–rump length (CRL), seen from an oblique 45 degrees posterior, cranial, and lateral view. P, palatal bone formation. Computer-aided three-dimensional reconstruction. The outlines of the eyes (e) and the nasal capsule (n) are superimposed to show the spatial relationships. B: Diagram of the craniofacial components in a human fetus of 42-mm CRL, frontal view. Red arrows show the direction of growth and expansion of the tissues (eyes, gray; nasal capsule and Meckel's cartilage, blue; oral cavity, pink) next to bone. They thus can exert shearing forces (red dotted arrows), which may cause the mesenchyme to react by bone formation (brown). Scale bars = 500 μm in A, 1 mm in B.

DISCUSSION

Pure Mechanics?

At first glance, it is difficult to give credence to the very unusual mechanical interdependencies between cell growth, spatial impediment and subsequent folds and bulges (His,1874), and the concept of tissue differentiation in fields of specific mechanical stress (Carey,1920a, b, 1921; Blechschmidt,2004). It seems too mechanistic and too simplistic. Where are the intrinsic biological capacities of the cells, and how does this concept fit into our contemporary knowledge of molecular signaling?

The changes of form and differentiation of tissues can be seen as biological responses of the cells, not as pure mechanics. The key idea, however, is that the cells do not follow an intrinsic program of differentiation during morphogenesis but are able to react to the different situations into which they have been maneuvered by their own growth and that of their neighboring cells (Carey,1920a, b, 1921; Blechschmidt,2004). The genetic code provides the basis for condensed mesenchymal cells to react by forming cartilage. Other cells subjected to shearing forces react by forming bone, and still others respond to being stretched by forming muscle fibers.

Cartilage Formation

We know that many molecular signals or transcription factors are indispensable during formation of cartilage. Sox9 is a downstream mediator for chondrogenic commitment, as shown in zebrafish (Yan et al.,2002) and chickens (Healy et al.,1996), and it can be counteracted by Msx2 (Takahashi et al.,2001), as shown in mice. Dlx5 also seems to be necessary, because Dlx5 knockout mice have hypoplastic nasal capsules (Depew et al.,1999).

There may be many more signaling molecules expressed during differentiation of cartilage tissue, and further studies may elucidate their indispensability. These molecules should be seen as mediators of cartilage formation, whereas the creation of cartilage form under the influence or at least in the presence of these molecules has not been addressed by signaling studies. Is there anything that contradicts the concept of mechanics and development? It is maintained that mesenchymal cells in specific locations (i.e., where mesenchyme is being condensed by growth of adjacent tissue) will form cartilage, and the form of the cartilage is determined by the spatial arrangement of the neighboring tissue. In terms of differentiation, cartilage is formed in response to pressure changes within the tissue (Copray et al.,1986; Francis-West et al.,2003). The differentiation of cartilage in these locations can be identified by the molecular signals just discussed. How and where does it start? Is Sox9 expression typical for mesenchymal cells being compressed in the center of a visceral arch or in any other places where cartilage is being formed? It would be beneficial if we could measure the forces in play at precisely the time when SOX9 is expressed.

Bone Formation

A major regulator found during bone formation is Runx2 (Mundlos et al.,1997; Kundu et al.,2002; Komori,2003; Lian and Stein,2003). Bone formation is a very complex process with many more signals active during matrix formation (Sodek et al.,2002; Holliday et al.,2003) and mineralization (Linkhart et al.,1996; Yamashiro et al.,2004). Again, our knowledge of molecular cascades per se does not give us any information about the control of bone form. However, we may specify our questions with the aim of bridging the gap between molecular biology and morphogenesis: How do mesenchymal cells that start to form bone differ from those that do not form bone?

They differ in many points, one being location. Thus we can ask whether location may have an influence on these mesenchymal cells. The start signal to form bone in these specific regions may be conveyed as a water-soluble protein. On the other hand, the start signal may also be transmitted by the physical conditions; for bone, these conditions would be the shearing forces exerted on the mesenchyme by the developmental movements of adjacent tissues. There is evidence that mesenchymal cells about to be transformed into osteoblasts are arranged in such a way as to float in the mesenchyme under the influence of a shearing action (our own histological findings and Blechschmidt,1948, 1960). The mesenchymal cells in these regions may react to this specific condition by expressing Runx2, for example, to start the cascade that leads to ossification. In the absence of one or more molecular signaling factors in knockout mice, the cells subject to the forces that trigger bone formation cannot do their bone-forming job.

Other putative factors, e.g., the mandatory presence of nerves in the region of early bone formation (Kjaer,1998), can also be explained as a detraction between the more or less taut nerves leashing and shearing the mesenchymal tissue, which responds with the bone-forming cascade at that specific spot.

Of interest are not only the starting points of ossification of a bone but also the points where the form of the bone is initially created. When examining the formation of the mandible in morphological detail, we see that bone extends along Meckel's cartilage in an anterior and posterior direction; it spreads underneath Meckel's cartilage and flows over Meckel's cartilage in an oral direction, thus forming a gutter between an inner and an outer bony partition (Radlanski and Klarkowski,2001; Radlanski et al.,2003). This finding may be explained by ongoing shearing forces, which can be envisioned when the different developmental stages of the mandible are compared in 3D at the same scale. It is difficult, however, to explain the early formation of the coronoid and condylar processes. The condylar process as an Anlagerungsgelenk (There is no English term; it may be translated as “approaching joint”) will remain more enigmatic (Youdelis,1966; Burdi,1992; Naidoo,1993; Radlanski et al.,1999a), whereas the coronoid process may result from muscle insertion and traction, as in later developmental stages (Spyropoulos,1977; Uchida et al.,1994).

Muscle Formation

Whereas the molecular interactions leading to muscle formation in the limbs have been unraveled in greater detail, there are very few data on the craniofacial muscles (Francis-West et al.,2003). Among many others, Myf5 has been recognized as a major myogenic determination factor controlled by distinct promoters in the head and trunk (Noden et al.,1999; Hadchouel et al.,2000; Summerbell et al.,2000).

There also seems to be evidence that cranial neural crest cells migrating to sites where they will form the corresponding muscles are preprogrammed (Noden,1983, 1988, 1991; Vaglia and Hall,1999; Francis-West et al.,2003). There are transplantation experiments that corroborate this concept, while others lead to an unclear judgment (Wahl and Noden,2001), questioning whether cranial neural crest cells migrate actively at all.

The literature does not indicate to what extent the direction and arrangement of muscle bellies might be regulated by signals (Robson,1993). The cause of the developmental movement of the muscles is unknown, e.g., extension of the rectus lateralis muscle of the eye (Francis-West et al.,2003). The review suggests that the direction of that muscle might be a result of cranial flexing (Wahl et al.,1994). In human development, the direction and extension of the muscles responsible for eye motility have been seen as a consequence of the expansion of the growing eyeball (Blechschmidt,1960), and it has been reported more recently that stretch of mesenchymal cells leads to formation of muscle fibers (Goldspink,1999). It is not clear, however, why the muscle bellies concentrate as four bundles adjacent to the eye.

There is not much overlap between our knowledge of the molecular signals associated with differentiation of muscle tissue and the control of orientation of the muscle fibers and muscle bellies. The arrangement of the muscles of the floor of the mouth recently described for human prenatal development (Radlanski et al.,2001) correlated well with the concept of dilated mesenchymal tissue. Our measurements of the growing mandible closely matched the orientation of muscle fibers in any direction: anteroposterior, transverse, or vertical (Radlanski et al.,2003).

PERSPECTIVES

There has been significant progress in understanding the signaling network controlling the patterning and development of the face (Francis-West et al.,1998; Mina,2001). It is believed that specific signaling molecules dispersed in the mesenchymal regions command the cells to differentiate (Mina et al.,1991). Most recent results describing human gene expression changes that occur during early stages of development with particular emphasis on craniofacial development can be found at http://hg.wustl.edu/COGENE. It is now obvious that studies elucidating the molecular signaling cascades during tissue differentiation do not explain the morphological aspect involved in creating the form of the organs. It seems as though the molecules are mediators rather than creators of form.

Knockout experiments demonstrate the essential requirement of the signaling molecules (Young et al.,2000; Gimara et al.,2004), and rescue experiments show that it is possible to restore the signaling cascade (Johnston and Nüsslein-Volhard,1992; Pispa et al.,1999; Bei et al.,2000). This finding does not affect the credibility of the concept that cells react to irritation by mechanical forces (Banes et al.,1995; Carter et al.,1998; Goldspink,1999; Murray,2000; Szabo et al.,2001; de la Fuente and Helms,2005). The biochemical signal cascade, which is part of the differentiation process, can be initiated by mechanical forces.

We do not know how genes are switched on and off in nature. We assume that forces generated by growth play a role. However, not much detailed information is available on the changes in form during morphogenesis. Future studies should elucidate the interdependency between mechanical stimuli and molecular signals and how they lead to differentiation. They should identify mechanosensors that may be responsible for this signal transduction and develop novel methods for measuring such forces within a tissue (de la Fuente and Helms,2005).

For this purpose, it is also necessary to create 3D reconstructions showing the growth, the proportional changes, the developmental movements, the spatial arrangement, and the surroundings of the evolving craniofacial complex, preferably for human development (Radlanski et al.,1999b; Radlanski,2003). We think that an increased number of studies will be devoted to comparative embryology. An interspecies comparison can be made in studies on craniofacial development in terms of form, 3D arrangement, mechanical interactions, and signaling cascades. This knowledge will enable us to gain a greater insight into the way in which similar signaling cascades create different forms. This approach may help to bridge the gap and correlate the morphological changes with the molecular signaling.

When it becomes possible to independently extract signal cascades from the force systems to which they are linked in nature, molecular medicine may become practicable in the field of craniofacial development, particularly in preventing malformation syndromes of the head and neck at early developmental stages.

EXPERIMENTAL PROCEDURES

The human histological material was obtained from legal and spontaneous abortions according to German law. The material was dehydrated, embedded in paraffin, sectioned at 10 μm, stained with hematoxylin–eosin, and reconstructed in 3D using the software AnalySIS (SIS, Münster, Germany).

Acknowledgements

We thank Mrs. B. Danielowski and Mrs. I. Schwarz for their skillful technical assistance in processing the embryos and the 3D reconstructions. This work has been partially funded by the Deutsche Forschungsgemeinschaft (Ra 428/1-3), and the studies have been conducted within the frame of the COST actions B8 and B23.