|Home | About | Journals | Submit | Contact Us | Français|
Secondary cartilage occurs at articulations, sutures, and muscle attachments, and facilitates proper kinetic movement of the skeleton. The induction and maintenance of secondary cartilage requires mechanical stimulation and accordingly, its evolutionary presence or absence reflects species-specific variation in functional anatomy. Avians illustrate this point well. In conjunction with their distinct adult mode of feeding via levered straining, duck develop a pronounced secondary cartilage at the insertion (i.e., enthesis) of the mandibular adductor muscles on the lower jaw skeleton. An equivalent cartilage is absent in quail, which peck at their food. We hypothesized that species-specific pattern and a concomitant dissimilarity in the local mechanical environment promote secondary chondrogenesis in the mandibular adductor enthesis of duck versus quail. To test our hypothesis we employed two experimental approaches. First, we transplanted neural crest mesenchyme (NCM) from quail into duck, which produced chimeric “quck” with a jaw complex resembling that of quail, including an absence of enthesis secondary cartilage. Second, we modified the mechanical environment in embryonic duck by paralyzing skeletal muscles, and by blocking the ability of NCM to support mechanotransduction through stretch-activated ion channels. Paralysis inhibited secondary cartilage, as evidenced by changes in histology and expression of genes that affect chondrogenesis, including members of the FGF and BMP pathways. Ion channel inhibition did not alter enthesis secondary cartilage but caused bone to form in place of secondary cartilage at articulations. Thus, our study reveals that enthesis secondary cartilage forms through mechanisms that are distinct from those regulating other secondary cartilage. We conclude that by directing the musculoskeletal patterning and integration of the jaw complex, NCM modulates the mechanical forces and molecular signals necessary to control secondary cartilage formation during development and evolution.
Mechanical forces play an essential role in shaping bone and cartilage during development. The differentiation of one type of cartilage, termed secondary cartilage, is a special phenomenon mostly associated with the dermal bones of the cranial skeleton. Secondary cartilage arises after osteogenesis and formation of the primary cartilaginous skeleton at articulations, sutures, and muscle attachments (Beresford, 1981; de Beer, 1937; Hall, 2005; Murray, 1963; Murray and Smiles, 1965). Secondary cartilage relies on mechanical stimulation and its evolution within a given taxon is linked to species-specific differences in functional morphology, especially in relation to feeding (Beresford, 1993; Hall, 1978, 1979, 1986; Stutzmann and Petrovic, 1975; Zweers, 1974). In mammals, secondary cartilages include the condyle and coronoid process of the mandible, whereas in the highly kinetic skulls of birds they also comprise the cartilaginous linings of articulations like that found along the proximal medial surface of the quadratojugal bone, which forms a joint with the quadrate in the upper jaw (Hall, 1984; Hall and Hanken, 1985; Moore, 1981; Novacek, 1993). Secondary cartilage is initiated and maintained via mechanical forces like compressive loading (i.e., cyclic hydrostatic pressure), as revealed by experimental manipulation in animal models (Asano, 1986; Copray et al., 1985; Fang and Hall, 1997; Hall, 1967, 1968; Murray and Smiles, 1965; Stutzmann and Petrovic, 1975).
To understand the mechanistic contributions of musculoskeletal pattern and biomechanical forces to the induction of secondary cartilage, we conducted a series of experiments using quail and duck embryos, which exhibit considerably different craniofacial morphologies. In support of their distinct mode of feeding via levered straining, duck and other Anseriformes develop a prominent bony process along the lateral aspect of the lower jaw (Baumel, 1993; Buhler, 1981; Van den Heuvel, 1992; Zweers, 1974; Zweers et al., 1977b). This process is homologous to the coronoid process on the mandible of mammals, including humans (Coues, 1887). In Galliformes such as the chick and quail, which peck at their food, the coronoid process appears as a slight bony ridge along the dorsal margin of the lower jaw (Chamberlain, 1943; Fitzgerald, 1969; Jollie, 1957; Lucas and Stettenheim, 1972; McLeod, 1964; Shufeldt, 1909) (Fig. 1A, B). The coronoid process of birds functions as the attachment site for the aponeurosis of the mandibular adductor muscle fibers (Fig. 1C, D, G, H), which are the primary jaw-closing muscles (Baumel, 1993; Buhler, 1981; George and Berger, 1966). In duck, the coronoid process forms first as secondary cartilage (de Beer, 1937; de Beer and Barrington, 1934) lateral to the surangular bone and within the insertion of the mandibular adductor (arrows, Fig. 1F and and2A).2A). This secondary cartilage is unusual in developing at an enthesis rather than an articulation (Murray, 1963). An equivalent secondary cartilage is absent in the mandibular adductor enthesis of quail (asterisk, Fig. 1E; ;2C2C).
In the current study we tested the hypothesis that species-specific differences in jaw morphology and a corresponding dissimilarity in the local mechanical environment induce the formation of secondary cartilage in the mandibular adductor enthesis of duck versus quail. We employed two experimental approaches. First, we transformed the duck jaw complex to resemble that found in quail by transplanting neural crest mesenchyme (NCM) from quail to duck (Fig. 1I, J). NCM is the source of species-specific patterning information for the jaw skeleton (Eames and Schneider, 2008; Jheon and Schneider, 2009; Merrill et al., 2008; Schneider, 2005, 2007; Schneider and Helms, 2003; Tucker and Lumsden, 2004) as well as the accompanying musculature (Tokita and Schneider, 2009). NCM gives rise to bone and cartilage and also produces muscle connective tissues including ligaments, tendons, fascia, and epi- and endomysia (Couly et al., 1993; Köntges and Lumsden, 1996; Le Lièvre and Le Douarin, 1975; Noden and Schneider, 2006; Noden, 1978, 1983b). In contrast, jaw muscles are derived from cranial paraxial mesoderm (Couly et al., 1992; Evans and Noden, 2006; Noden, 1983a; Noden and Francis-West, 2006; Noden and Trainor, 2005). Second, we altered the mechanical environment in duck by either paralyzing the skeletal musculature or by blocking mechanotransduction through stretch-activated ion channels (SAC). Paralysis experiments in chick embryos (Fang and Hall, 1995; Hall, 1979; Murray, 1963; Murray and Smiles, 1965) or suturing the jaw shut in mice (Habib et al., 2005), lead to a loss of secondary cartilage in the jaw. Similarly, in chondrocyte culture experiments under conditions of mechanical loading, treatment with SAC inhibitors such as gadolinium (Gd3+) reduces chondrocyte proliferation (Wu and Chen, 2000). We adapted the use of gadolinium in ovo to assess the ability of cells to sense and transduce stresses via SAC at the mandibular adductor enthesis and quadratojugal joint in duck.
To determine the effects of our transplants and treatments on secondary chondrogenesis, we performed anatomical reconstructions and histological analyses, and assayed for the expression of genes known to be involved in skeletogenesis including Sox9, Col2a1, Runx2, Fgfr2, and Bmp4. Our experiments reveal that the induction of secondary cartilage within the mandibular adductor muscle enthesis relies upon NCM-dependent changes in musculoskeletal pattern that are species-specific and that likely shape the local mechanical force environment. Moreover, we find that enthesis secondary cartilage appears to arise via different molecular, histogenic, and biomechanical mechanisms than secondary cartilage that forms at bony articulations. Thus, NCM has played a critical role in directing the structural and functional integration of the jaw apparatus during the course of vertebrate evolution.
Fertilized eggs of Japanese quail (Coturnix coturnix japonica) and white Pekin duck (Anas platyrhynchos) were purchased from AA Labs (Westminister, CA) and incubated in a humidified chamber at 37°C until reaching Hamburger and Hamilton (HH) 9.5 (Hamburger and Hamilton, 1951). Embryos were handled following University and NIH guidelines. Eggs were windowed and embryos visualized with Neutral Red (Sigma). Unilateral populations of NCM extending from the caudal forebrain to the rostral hindbrain (Fig. 1I) were grafted orthotopically from quail to duck (Lwigale and Schneider, 2008). Unilateral transplants provided an internal control on the un-operated host side. Flame-sharpened tungsten needles and Spemann pipettes were used for surgical operations (Schneider, 1999). Donor graft tissue was positioned and inserted into a host with an equivalent region of tissue excised. For additional controls, orthotopic grafts or sham operations were made within each species. Controls were incubated alongside chimeras to ensure that stages of grafted cells were accurately assessed. A combination of morphological characters was used to stage embryos, with emphasis on post-cranial and other structures unaltered by surgery.
Embryos were fixed in Serra’s (100% ethanol:37% formaldehyde:glacial acetic acid, 6:3:1) overnight at 4°C, paraffin embedded, and cut into 10 µm sections. Representative sections were stained with Milligan’s Trichrome (Presnell and Schreibman, 1997) for visualization of cartilage, bone, tendon, and muscle. Three-dimensional images of first arch jaw muscles and portions of associated skeletal elements were generated via reconstruction of serial sections using the WinSurf software package (SURF driver, Hawaii) (Tokita and Schneider, 2009). To detect quail cells in chimeric embryos, sections were immunostained with the quail nuclei-specific Q¢PN antibody (Fig. 1J) (1:10, Developmental Studies Hybridoma Bank (DSHB)) (Schneider, 1999). Whole embryos were stained with Alcian blue and Alizarin red and cleared with glycerol (Wassersug, 1976).
In situ hybridization was performed as described previously (Albrecht et al., 1997; Schneider et al., 2001). Sections adjacent to those used for histological and immunohistochemical analyses were hybridized with 35S-labeled chicken riboprobes to genes expressed during chondrogenesis (Sox9, Col2a1, Fgfr2 and Bmp4) and osteogenesis (Runx2). Sections were counterstained with a fluorescent nuclear stain (Hoechst Stain; Sigma).
Duck embryos between HH29 and HH36 were paralyzed by administering a 0.5ml solution of 10mg/ml decamethonium bromide (Sigma) in Hank's balanced salt solution (HBSS). in ovo, as previously described (Hall, 1986). Decamethonium bromide is an agonist for the acetylcholine receptor on the post-synaptic membrane of the neuromuscular junction (Drachman, 1971). Exposure to decamethonium bromide produces a depolarizing cascade that blocks synaptic transmission in skeletal muscle but does not interfere with cardiac or smooth muscle contraction (Macharia et al., 2004). Solution was dispersed into the albumin over the developing embryo. A dose-response curve was generated and doses higher than 20 mg/ml were found to be lethal for duck embryos at these stages. Controls were given 0.5 ml Hank’s balanced saline solution (HBSS, Sigma). During embryo collection, paralysis was confirmed by observing beak and head movement, as well as by removing embryos from their shells and assaying for reflex muscular activity after hind limb extension.
Duck embryos between HH31 and HH36 were treated with a 0.5 ml solution of 1 mg/ml GdCl3 (Sigma-Aldrich 439770) in HBSS. Gadolinium (Gd3+) has been used in vitro to inhibit mechanotransduction via stretch-activated ion channels in populations of mechanically stimulated chondrocytes (Park et al., 2002; Wu et al., 2001; Wu and Chen, 2000). Solution was dispersed into the albumin over the developing embryo. Controls were treated with HBSS. A dose-response curve was generated and doses above 2.5 mg/ml were found to be lethal for duck embryos at these stages.
To understand the relationship between species-specific morphology and secondary cartilage formation, we performed unilateral transplants of NCM from quail to duck embryos stage-matched at HH9.5 (Fig. 1I). In resultant chimeric quck collected at HH38, secondary cartilage developed within the mandibular adductor enthesis along the surangular bone on the duck host side of the mandible, with an equivalent size and orientation as that found in control duck (n=7; Fig. 2A,B). However, secondary cartilage was absent in the enthesis on the quail donor side like that observed in control quail (Fig. 2B,C). To analyze the effects of NCM on the spatial orientation and morphology of the enthesis, we generated and compared three-dimensional reconstructions of the surangular bone and mandibular adductor muscle enthesis. We found that the mandibular adductor muscle inserted along the dorsal margin of the surangular bone in control quail (Fig. 2G), whereas in control duck, this muscle inserted laterally on the surangular (Fig. 2D). Moreover in duck, the enthesis was relatively broader and had a more extensive attachment along the surangular than in the quail, where the enthesis remained thin throughout its length and had a more restricted insertion. In addition, the mandibular adductor muscle inserted more distally along the surangular in duck, whereas in quail this insertion was more proximal to the jaw joint. In quck chimeras at HH36 (n=5; Fig. 2F), the enthesis on the quail donor-derived side was thin and inserted dorsally on the surangular, like that observed in quail (n=5; Fig. 2G). On the duck host side, the lateral position and robust morphology of the enthesis was equivalent to that seen in control duck (n=5; Fig. 2D,E).
Histological analyses confirmed these significant species-specific differences in the relative orientation, size, and shape of the mandibular adductor muscle enthesis between quail and duck. Correspondingly, in chimeric quck at HH36 (Fig. 2J), the enthesis was much narrower and less developed on the donor side, like in control quail (Fig. 2K). On the host side, the enthesis was much wider and triangular shaped, as observed in control duck (Fig. 2I). We stained adjacent sections with the anti-quail Q¢PN antibody and found no quail-derived cells on the duck host side (Fig. 2M) but abundant quail-derived cells throughout the bone, cartilage, and muscle connective tissues on the donor side (Fig. 2N). In particular, we observed that the fibrous aponeurosis and enthesis of the mandibular adductor muscle on the quck donor side formed from quail NCM, but the muscle itself was derived from the duck host.
To identify molecular changes that accompanied the species-specific transformation of the mandibular adductor enthesis in quck, we analyzed the expression of genes known to be required for cartilage development. In particular, we used section in situ hybridization to detect mRNA for Sox9 and Col2a1. Sox9 transcripts appeared within the enthesis on the host side of HH36 chimeric quck, in the same domain as that observed in control duck (Fig. 2P,Q). However, Sox9 was neither expressed in the enthesis on the quck donor side, nor in the corresponding enthesis of control quail (Fig. 2R,S). Col2a1 transcripts were detected throughout the enthesis on the quck host side as in control duck (Fig. 2T,U). However on the donor side of quck, Col2a1 expression was confined to a narrow band along the mandibular adductor muscle aponeurosis (Fig. 2V) like that observed in control quail (Fig. 2W).
To assess the extent to which the formation of enthesis secondary cartilage depends upon the mechanical environment, we performed a series of paralysis experiments using decamethonium bromide. In the mandibular adductor enthesis of duck, the first histological evidence of a cartilaginous condensation and the earliest expression of chondrogenic molecular markers can be detected at HH33 (data not shown). Paralysis of duck at HH31 completely inhibited the formation of secondary cartilage within the mandibular adductor enthesis as evidenced by cleared and stained specimens and 3D reconstructions (n=9; Fig. 3A,B). Enthesis secondary cartilage was present in control duck (Fig. 3C). However, duck paralyzed at HH34, which is immediately after secondary chondrogenesis can be detected histologically, developed greatly reduced enthesis cartilage (n=5; Fig. 3D). Again, duck treated at HH31 showed no histological evidence of enthesis secondary cartilage formation and significant muscle atrophy (n=12; Fig. 3E). Analysis of gene expression in HH38 control duck revealed that Sox9 and Col2a1 continued to be expressed within the enthesis secondary cartilage (Fig. 3F,I). These expression domains were substantially reduced in duck paralyzed at HH34 (Fig. 3G,J) and completely absent in duck paralyzed at HH31 (Fig. 3H,K).
To determine whether the effects of paralysis were specific to enthesis secondary cartilage, or to secondary cartilages in general, we analyzed the formation of secondary cartilage at the articulation between the quadratojugal bone and the quadrate (Fig. 1E,F; Fig. 3L). In control duck embryos at HH38, we observed a large domain of secondary cartilage bounded by periosteum at the proximal medial margin of the quadratojugal (Fig. 3L). This cartilage expressed Sox9 and Col2a1 (Fig. 3O and data not shown). Quadratojugal secondary cartilage was not detected in HH38 duck paralyzed at HH31 and HH34 but instead we observed well-ossified bone (Fig. 3M,N). The lack of quadratojugal secondary cartilage was confirmed by an absence of Col2a1 expression, despite the fact that Col2a1 expression was maintained in the adjacent quadrate cartilage (Fig. 3P,Q).
To ascertain the extent to which enthesis secondary cartilage depends upon mechanical force transduction via stretch-activated ion channels, we treated developing embryos with gadolinium. Treated duck embryos formed enthesis secondary cartilage along the surangular bone just like in control duck. Histological analyses of embryos at HH36 revealed no changes in the development or morphology of enthesis secondary cartilage, or in the expression of Sox9 and Col2a1 relative to controls (n=3; Fig. 4A,D,G and data not shown). In contrast, in the same gadolinium-treated embryos, we observed well-ossified bone in place of quadratojugal secondary cartilage (Fig. 4B,C). The absence of quadratojugal secondary cartilage in gadolinium-treated embryos was also coincident with a down-regulation of Sox9 and Col2a1 relative to controls (Fig. 4E,F,H,I).
To investigate whether enthesis and quadratojugal secondary cartilages differ in other ways on the molecular level during normal development and in response to gadolinium treatment, we analyzed the expression of members of signaling pathways known to mediate the formation of cartilage and bone. Bmp4 expression was unaffected in enthesis secondary cartilage but appeared down-regulated in quadratojugal secondary cartilage relative to controls (Fig. 4J,K,L). Fgfr2 was expressed in developing enthesis secondary cartilage following gadolinium treatment (Fig. 4M) as in controls (data not shown), but Fgfr2 was not expressed in the body of the quadratojugal secondary cartilage of either control or treated embryos (Fig. 4N,O). Instead, Fgfr2 expression was restricted to a narrow domain within the periosteum of the quadratojugal bone. Runx2 was not expressed in the developing enthesis secondary cartilage following gadolinium treatment (Fig. 4P) or in controls (data not shown). Runx2, which was expressed only within the periosteum of the quadratojugal bone in control embryos (Fig. 4Q), became up-regulated within the ossified quadratojugal secondary cartilage bone following gadolinium treatment (Fig. 4R).
In this study, we tested if species-specific differences in jaw morphology and a corresponding dissimilarity in the local mechanical environment promote the formation of secondary cartilage in the mandibular adductor enthesis of duck versus quail. Our results reveal that the patterning of the adductor muscle insertion by NCM induces the development of the coronoid process via secondary chondrogenesis within the enthesis. We conclude that this is primarily due to a shift in the anatomical insertion of the mandibular adductor muscle (Fig. 5A) but NCM-mediated changes to other aspects of muscle morphology such as size and shape could also produce species-specific differences in the local mechanical environment and contribute to the induction of secondary cartilage. Previous studies have shown that the fibers of the mandibular adductor muscle insert on the lateral side of the surangular bone in duck (Zweers, 1974; Zweers et al., 1977b), whereas the same group of muscle fibers insert on the dorsal surface of the surangular in quail (Baumel, 1993; Van den Heuvel, 1992). Our anatomical and histological analyses confirm such observations, and our transplant experiments demonstrate that these species-specific differences in the shape, orientation, and insertion point of the mandibular adductor muscle are patterned by NCM. In particular, chimeric quck have quail-like mandibular adductor enthesis morphology on the quail donor-derived side, and duck-like morphology on the duck host-derived side. These results support our previous work demonstrating that species-specific muscle morphology arises through the actions of NCM-derived skeletal and muscular connective tissues (Tokita and Schneider, 2009) and reinforce the notion that muscles and their connective tissues are mechanistically linked during development and evolution (Evans and Noden, 2006; Grenier et al., 2009; Kardon, 1998; Mathew et al., 2011; Noden, 1983b, 1986; Noden and Francis-West, 2006; Noden and Trainor, 2005; Rinon et al., 2007; Tokita, 2004; Tokita et al., 2007). But the current experiments extend these conclusions further and show that such NCM-dependent control over muscle pattern has secondary consequences for the NCM-derived skeleton itself, via the effects of muscle on skeletal pattern. The force-dependent formation of secondary cartilage within the duck mandibular adductor enthesis is a functionally relevant example of this phenomenon. The duck enthesis secondary cartilage is replaced by bone after hatching (data not shown) and becomes the coronoid process of the mandible.
The coronoid process plays an important role in the functional morphology of the duck jaw apparatus (Buhler, 1981; Zweers, 1974; Zweers et al., 1977b). The predominant mode of feeding in duck occurs via a suction pressure pump system that involves both the lingual apparatus and the jaw complex, and enables filtering of small food items by the levered straining of water (Buhler, 1981; Olson and Feduccia, 1980; Van den Heuvel, 1992; Zweers, 1974; Zweers et al., 1977a; Zweers et al., 1977b). Levered straining is performed through rapid opening and closing of the mandible, which requires sudden acceleration and significant force (Zweers et al., 1977b). The relatively large forces required in this mode of feeding contrast with the quail, which only needs to peck and swallow its food (Buhler, 1981; Van den Heuvel, 1992). Consequently, duck have evolved relatively large mandibular adductor and depressor muscles (Fig. 5B), which function in double-coupled kinesis, whereby each muscle contributes to lengthening and pre-loading of the opposing muscle around the fulcrum of the quadrate (Zweers et al., 1977b). Lengthening and pre-loading of both muscle groups enables the rapid contraction and acceleration of the jaw. A broad insertion area further enhances the functional leverage of this muscle in closing of the jaw. These features allow rapid straining in the duck, approaching 25 times a second (Zweers, 1974).
Qualitative and quantitative analyses of the avian jaw apparatus have described the complex patterns of movement as well as determined force vectors for the mandibular adductor muscles during feeding (Beecher, 1962; Bock, 1964; Bock, 1999; Bock and Kummer, 1968; Bout and Zweers, 2001; Buhler, 1981; Fisher, 1955; Hoese and Westneat, 1996; Lederer, 1975; Meekangvan et al., 2006; Smith, 1993; Zusi, 1967; Zusi, 1993; Zweers, 1982). In quail and duck, the adductor muscles exert a dorsal-caudal oriented force during closing of the jaw (Zweers, 1974; Zweers et al., 1977b). However, the difference in anatomical location of the mandibular adductor muscle insertion between these species likely generates uniquely different force environments. Presumably, a shift in the orientation of the mandibular adductor muscle insertion from a lateral position in duck, to dorsal one in quck changes the resulting force environment at the enthesis. The anatomical outcome of this shift in chimeric quck is an absence of secondary cartilage. Secondary cartilage within the duck enthesis extends lateral to the plane of the mandible. The fibers of the adductor muscle originate dorsal-caudally, inserting at a right angle to the direction of the applied force. In this configuration the enthesis would likely experience both tension and shear. However, in control quail and on the donor side of chimeric quck the enthesis inserts dorsally on the surangular, and so in this orientation, the enthesis would solely be under tension. Computational modeling may help explain how biomechanical factors contribute to secondary chondrogenesis in duck versus quail. We would predict that NCM plays an active role in establishing the musculoskeletal geometry, and in response to the nature of the resultant mechanical force (compression versus tension), NCM differentiates into fibroblasts, tenocytes, and/or chondrocytes. Quite remarkably of course, all of these musculoskeletal patterning events occur in the context of embryonic motility, prior to any functional use, and in anticipation of later species-specific feeding behavior.
The results from our paralysis experiments demonstrate that the induction of enthesis secondary cartilage requires active skeletal muscle contraction. In this respect, the developing enthesis responds to mechanical force in the same manner as secondary cartilages at articulations (Hall, 1979, 1986; Murray, 1963; Murray and Smiles, 1965). Furthermore, we observe that inhibited movement limits the growth of pre-existing secondary cartilage within the enthesis, exactly as observed at articulations containing secondary cartilage (Buxton et al., 2003; Hall, 1986; Murray and Smiles, 1965). During development, the mechanical stimulation required for induction of secondary cartilage arises from embryonic motility. In the chick, the start of chondrogenic condensation coincides with the time course of embryonic motility (Hall, 1986; Hamburger et al., 1965). Movement is first observed at three days of incubation (HH21) and is initially isolated to the neck. At day 4 (HH23), there are waves of movement along the body axis (Hamburger and Balaban, 1963). Motility in the legs and wings begins at 6.5 days (HH29) and by 7.5 days (HH31), embryos become responsive to external stimuli (Hamburger and Oppenheim, 1967). Complex movements such as “beak clapping” are seen in later stages (HH41 to hatching) and coordinated pre-hatching movements begin around day 17 (HH43) (Hamburger and Balaban, 1963; Hamburger et al., 1965; Hamburger and Narayanan, 1969). Strikingly similar patterns of embryonic movement and levels of activity have been described in duck, and at comparable stages of development (Oppenheim, 1970). Given that the progression and extent of embryonic motility appears to be conserved between these species, then the critical determinants of enthesis secondary cartilage induction most likely arise from the anatomical organization of the jaw apparatus and the ensuing consequences of differences in the local force environment. So in chimeric quck, while the time course and extent of embryonic motility (i.e., the source of mechanical force), would be equivalent on the donor and host sides of the embryo, the distribution of forces required to induce secondary cartilage would likely be dissimilar on one side of the jaw versus the other due to species-specific differences in the geometry of the jaw apparatus. Yet subtle variations in embryonic motility may exist between quail and duck, and also contribute to the formation of enthesis secondary cartilage. For example, increasing embryonic motility either by raising incubation temperature or by injecting embryos with 4-aminopyridine (4-AP), which results in hyperactivity, has profound morphological effects including augmented bone length and muscle mass, and articular cartilage phenotypes such as larger joint cavities (Hammond et al., 2007).
In response to forces produced by embryonic motility as well as from normal activities during post-embryonic life, musculoskeletal tissues develop a complex internal stress, which is set by tissue geometry and matrix material properties. Matrix material properties are biologically defined by signaling pathways that employ factors such as Runx2 (Chang et al., 2010). In turn, these physical cues can direct cell differentiation (Balooch et al., 2005; Engler et al., 2006). Stresses experienced by cells vary in type (e.g., tensile or compressive) as well as in frequency (e.g., cyclical or static). The ways in which cells sense and transduce stresses at the mandibular adductor enthesis in duck versus quail remain to be determined, but mechanical stress is known to maintain the proliferation and differentiation of chondrocytes (Archer et al., 2006). Moreover, the action of compressive force has been shown to induce secondary cartilage in various anatomical environments, and there are many reports of enhanced proliferation of chondrogenic precursor cells in response to compressive mechanical stress both in vivo and in vitro (Fang and Hall, 1995, 1999; Hall, 1979, 1986; Wu et al., 2001; Wu and Chen, 2000). Such data suggest that entheses are constantly tuning their mechanical responses during development (Benjamin and Ralphs, 1998; Li et al., 2006; Robbins et al., 1997). In tendons and ligaments, cartilage or fibrocartilage forms in zones of compression where ligaments wrap around bones (Benjamin et al., 2006; Carter and Beaupré, 2001; Li et al., 2006; Robbins et al., 1997). Cartilage does not ordinarily form at tendon insertions because tendons are generally subject to tension (i.e., deviatoric stress). However, anatomical situations exist in which the tendon insertion must flex considerably during active movements (Benjamin et al., 2006; Kardon, 1998). Enthesis secondary cartilages are found in the bovine Achilles tendon and the rat medial collateral ligament (Benjamin and Ralphs, 1998; Gao et al., 1996). Flexing exerts hydrostatic pressure on entheses and can induce cartilage (Benjamin and Ralphs, 1998; Benjamin et al., 2006; Li et al., 2006).
Our histological analyses indicate that the chondrogenic condensation of the adductor enthesis in duck develops within the fibrous aponeurosis of the mandibular adductor muscle. In contrast, secondary cartilage at bony articulations forms via a well-defined process in which periosteal progenitor cells become chondrocytes under the action of biomechanical force (Archer et al., 2006; Buxton et al., 2003; Hall, 1979, 1986). Our experiments suggest that enthesis secondary cartilage develops in a manner distinct from that of secondary cartilage at articulations but the specific mode of mechanotransduction by which mechanical force is converted into a molecular signal to induce enthesis secondary cartilage remains unclear. Through mechanotransduction, cells can distinguish among physical signals and translate them into intracellular responses. Chondrocyte mechanoreceptors include intracellular ion channels and integrins, which bridge physical stresses and intracellular responses (Millward-Sadler and Salter, 2004; Takahashi et al., 2003). Once mechanoreceptors receive a physical stress, an intracellular signal is transduced, which can result in tissue remodeling.
We blocked mechanotransduction by using gadolinium (Gd3+) in duck embryos. Gadolinium is a potent blocker of mechanogated ion channels, and in particular those that are stretch-activated (Hamill and McBride, 1996; Park et al., 2002). While we observed that gadolinium treatments disrupted secondary cartilage formation in the quadratojugal, we saw no effect on secondary cartilage in the mandibular adductor enthesis. This difference may reflect distinct modes of mechanotransduction and/or activation by discrete molecular programs for each type of secondary cartilage (Buxton et al., 2003; Fang and Hall, 1995; Hall, 1986; Shibata and Yokohama-Tamaki, 2008). For example, cartilage formation in tendons and aponeuroses subject to pressure is accompanied by expression of Sox9, but not Runx2 (Li et al., 2006). Sox9 is a chondrogenic transcription factor that is required for expression of Col2a1, which is an early cartilage matrix constituent (Bell et al., 1997; Eames et al., 2003; Eames et al., 2004; Healy et al., 1996; Zhao et al., 1997). The absence of Runx2 expression in enthesis secondary cartilage further indicates that the chondroctyes develop from within the fibrous aponeurosis, rather than the periosteum. This is in contrast to secondary cartilages at articulations, which arise from a pre-osteoblast Sox9 and Runx2 co-expressing population that lies within the periosteum at articulation sites (Buxton et al., 2003; Eames et al., 2004; Hall, 1986; Hall, 2005; Shibata and Yokohama-Tamaki, 2008).
When we compare enthesis secondary cartilage to secondary cartilage at the quadratojugal articulation, we find Bmp4 transcripts at early stages in both sites. In contrast, we observe Fgfr2 expression only within the enthesis. FGF and BMP signaling up-regulate Sox9 and promote chondrogenic differentiation (Govindarajan and Overbeek, 2006; Healy et al., 1999; Murakami et al., 2000; Shum et al., 2003; Uusitalo et al., 2001l). Fgfr2 is a receptor for Fgf4, which functions upstream of Sox9 (Bobick et al., 2007; Govindarajan and Overbeek, 2006; Murakami et al., 2000; Petiot et al., 2002). Bmp4 may help direct the differentiation of fibrocartilage from tendon under hydrostatic pressure (Robbins et al., 1997; Sato et al., 1999) and Bmp4 can rescue secondary cartilage formation in Runx2-deficient mice suggesting that Runx2 expression is not required for the induction of secondary cartilage (Fukuoka et al., 2007). We also find Fgfr2 expression only within the chondrogenic enthesis in control duck and on the host side in chimeric quck. Fgfr2 expression is absent in the aponeurosis and enthesis of quail and the donor side of quck. Secondary chondrogenesis at entheses may thus proceed via a Runx2– and SAC–independent pathway involving FGF and BMP signaling (Fig. 5C,D). By identifying molecular and cellular processes through which NCM responds to mechanical stimulation and undergoes secondary chondrogenesis, this study helps characterize the complex relationships that connect musculoskeletal structure to function, as well as reveals underlying developmental mechanisms that have shaped the vertebrate jaw complex during evolution.
We thank Kristin Butcher, Johanna Staudinger, and Maren Caruso for technical assistance; Ralph Marcucio, Andrew Jheon, Jane Yu, Erin Ealba, Stephanie Gline, Simon Tang, Carol Chen, Christian Mitgutsch, Tamara Alliston, and Jeff Lotz for helpful discussions; and Thomas Dam at AA Lab Eggs. The Q¢PN antibody was obtained from the Developmental Studies Hybridoma Bank, maintained by University of Iowa under the auspices of the NICHD. Supported in part by a Harvard Presidential Scholars Award to C.S.; NIH F32 DE016778 to B.F.E.; Grants-in-Aid of JSPS Fellowship to M.T. (18002260); and NIDCR R03 DE014795 and R01 DE016402, NIAMS R21 AR052513, and March of Dimes 5-FY04-26 to R.A.S.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.