The mandibular symphysis is the conspicuous syndesmosis that connects the two hemimandibles at the chin among jawed animals. Developed by membranous ossifications around the rostral ends of Meckel's cartilages, the joint is a key component of the head and has provided important developmental, functional, and evolutionary insight into the feeding apparatus in sharks (Gerry et al., 2008), metamorphosing amphibians (Svensson and Haas, 2005), birds (Hogg, 1983) and mammals (Scapino, 1981; Hogue and Ravosa, 2001; Herring et al., 2008; Williams et al., 2008), among other vertebrates. The development of a fused bony symphysis is acknowledged to be a critical adaptation related to the complicated masticatory behaviors evolved in primates and humans (Hylander et al., 1998, Lieberman and Crompton, 2000, Ravosa et al., 2008), in which the joint transfers balancing side muscle forces to that of the working side, or bite point (Crompton and Hylander, 1986) by maintaining a relatively narrow, rigid fibrocartilaginous pad between the dentaries. Finally, the joint offers important insight into connective tissue plasticity and short-term adaptations of the craniofacial skeleton (Ravosa et al., 2007). The role of Meckel's cartilage and other joint tissues in the architecture of the mandibular symphysis is poorly understood in vertebrates, and particularly among lepidosaurs, which rely more on soft rather than bony tissues to build the joint (Bellairs, 1984). Despite the use of symphyseal bony features (e.g., Meckelian groove and intramandibular septum) in phylogenetic analyses (e.g., Estes et al., 1988; Lee and Scanlon, 2001; Evans, 2003), only Bellairs (1984) touched briefly on the topic of postnatal, soft symphyseal tissue morphology in his closing remarks at his Festschrift symposium. However, his remarks, and later those of Kley (2006) focused primarily on the relatively flexible joint tissues of snakes and only anecdotally about the joint's broader morphological diversity within Lepidosauria.
Like the symphyses of mammals and other vertebrates, it is reasonable to hypothesize that the mandibular symphysis' form and function in lepidosaurs (Sphenodon, lizards, amphisbaenians, snakes,) also bears ecomorphological and evolutionary significance. Lepidosaurs evolved a large diversity of feeding behaviors including herbivory, durophagy, macropredation, among other more opportunistic behaviors (Schwenk, 2000; Reilly and McBrayer, 2007). Among these behaviors, lepidosaur clades evolved two relatively distinct, integrated feeding repertoires: lingual prehension-palatal crushing-akinesis (iguanids and skinks) and jaw prehension-puncture crushing-cranial kinesis (e.g., anguimorphs, lacertoids, and geckos) (Metzger, 2002; Reilly and McBrayer, 2007). The intrinsic flexibility of the lizard skull, or cranial kinesis, has gained wide recognition as a model system for analyzing feeding and craniofacial functional morphology in extant, as well as extinct reptiles (see Rieppel, 1993; Herrel et al., 2000; Metzger, 2002; Holliday and Witmer, 2008). Cranial kinesis is hypothesized to facilitate prey processing and tooth occlusion, while mechanically insulating portions of the skull during unilateral biting, which induces torsion and bending upon the skull and mandibles (Herrel et al., 2000; Metzger, 2002; Moazen et al., 2009). Because the lizard symphysis is probably a moderately mobile joint like some intracranial joints, it is expected that it also facilitates occlusion while transmitting joint reaction forces. Lingual feeders, as opposed to jaw-apprehending taxa, rely heavily on intrinsic (e.g., muscles and papillae) but also extrinsic structures (e.g., hyoid musculoskeletal system) of the tongue to apprehend and transport prey intraorally (Schwenk, 2000). M. genioglossus, the muscle responsible for tongue protraction and m. intermandibularis, which raises the floor of the mouth, attach to the caudal surface of the symphysis and rostromedial portions of Meckel's cartilage, respectively; thus it is expected that the joint and its tissues should exhibit specializations to lingual feeding among iguanids and geckos in particular. Finally, the phylogenetic relationships among squamate clades are contentious depending on the use of morphological (e.g., Conrad, 2008; Fig. 1) or molecular data (e.g., Townsend et al., 2004; Vidal and Hedges, 2005), thus new insight into cranial soft and hard tissues may help further clarify evolutionary patterns.
Bellairs and Kamal (1981) and Bellairs (1984) only briefly touched on symphyseal (i.e., the intraramal hinge) anatomy of a small sample of snakes and lizards, noting there was much potential in further exploring the nature of the joint. Aside from the above work, there are only anecdotal references to the anatomy of the joint (e.g., Ramaswami, 1946; Torres-Carvajal, 2003). Here we follow their work and report on new microanatomical and imaging data that illustrate the soft tissues in the symphysis among a broad assemblage of adult lizards. We analyzed the organization of connective tissue fibers and Meckel's cartilage within a phylogenetic context. Symphyseal composition varies discretely among particular lizard clades, exhibits ecorelevant diversity, and illustrates morphological adaptations that lend insight into the phylogenetic relationships and evolution of groups within Squamata.
MATERIALS AND METHODS
Eleven mature individuals representing nine clades of squamate (Fig. 1; Table 1) were housed in an environmental chamber at 27°C, 85% humidity and fed mealworms, crickets or mice ad libitum for over a year during the course of locomotor experiments (NSF IOB 0520100, Reilly and Biknevicius, Ohio University). Approximately 5 weeks prior to euthanasia, individuals were injected intravenously via the caudal vein with three series of fluorochrome dyes (Calcein, Oxytetracycline HCL, Oxytetracycline HCL; 20 mg/kg dissolved and filtered in 0.9% saline) every 11 days (Williams and Holliday, OU IACUC U06-09). Four days after the final fluorochrome injection, the animals were euthanized. Because formalin dilutes labeling dyes, after euthanasia, specimens were immediately stored in 100% and then 70% ethanol for histological processing and later analyses of bone deposition rates (Holliday, unpublished data). Heads were micro-CT-scanned at 45-μm slice thickness (GE eXplore Locus in vivo Small Animal MicroCT Scanner, Ohio University). Three dimensional models of symphyses were rendered with Amira v4.1 (Visage Imaging) and Geomagic (v8.0) (Fig. 2A–D). Additional, salvaged individuals (Table 1) were also obtained, variably micro or medical CT-scanned, and similarly histologically sampled to complement the lab-raised dataset. Dietary and feeding habits of represented lizard taxa were collected from the literature.
Table 1. Body masses, feeding ecology, qualitative symphyseal characteristics, and an accompanying figure key of lepidosaur taxa used in analysis
Data collected from McBrayer and Corbin (2007) and Reilly and McBrayer (2007).
Mandibular symphyses were harvested using a rotary Dremel cutting tool or clippers, fixed, and dehydrated with graded ethanol solutions. Specimens were cleared with xylene and methyl methacrylate (MMA) and then infiltrated with three solutions of MMA and dibutyl phthalate (DBP) with vacuum over the course of 3 weeks. Specimens were then embedded in a fresh solution of MMA, DBP, and Perkadox-16 (P16) and allowed to polymerize using a combination of room temperature, refrigeration, waterbath, and an oven over the course of 2 weeks. Serial thin-section microtomy (∼5 microns) was performed using a motorized rotary microtome (Leica Microsystems) and a tungsten-carbide knife (Dorn and Hart Microedge). Alternating groups of three sections were then deplastified and stained using Safranin O/Fast green, Goldner's trichrome (e.g., Fig. 2E), von Kossa/MacNeal's tetrachrome (e.g., Fig. 2F) staining techniques. The fourth slide of three sections in each series was left unstained for fluorescent imaging or later staining. Slides were viewed using standard and polarized light microscopy. Unstained specimens of select taxa (Pogona, Eumeces, V. niloticus) were also viewed using linear polarized light (Leica DM750 Polarizing Light Microscope) and grayscale images were rendered in Amira where an 8-bit color map was applied to better visualize fiber orientation (Fig. 3).
The lizard symphysis is organized into several domains of tissues oriented along the rostrocaudal and dorsoventral axes of the joint. Meckel's cartilage and a band of loose woven tissues occupy the ventral half of the joint whereas fibrocartilage, denser, fibrous connective tissues, and/or looser parallel-fibered tissues occupy the dorsal half (Figs. 2–4). Meckel's cartilage is more prominent in the caudal half of the symphysis whereas fibrous connective tissues unite the rostral portion of the joint. The muscles mm. genioglossus and intermandibularis attach to the rami of Meckel's cartilage, the former attaching to the rostralmost portions of the cartilage either directly or via a broad tendon (Figs. 2H, 4H). The ventral bony portions of the symphysis in geckos taper laterally away from the midline giving the joint an inverted “V” shape (Fig. 4K), whereas the bony joints in other taxa are more vertical and parallel with their opposite fellow until gradually tapering laterally along their ventral margins.
The dorsal half of the symphysis is spanned by fibrous connective tissue in the center of the joint in all taxa. The fibers are generally oriented parallel to one another within the horizontal or axial planes of the joint (Fig. 3) and are shallowly embedded into the bone via extrinsic fibers (e.g., Fig. 3C) surrounded by Sharpey-fiber bone rich with glycosaminoglycans as seen with Safranin O/Fast green staining (Fig. 5C). In Varanus exanthematicus, a nodular fibrocartilaginous pad occupies the rostrodorsal portion of the joint and excavates circular fossa on each hemimandible (Fig. 2). The pad is surrounded dorsally and ventrally by parallel fibers. In Gerrhosaurus, a large, well-defined bundle of mineralized tissue is embedded within the fibers spanning the relatively narrow joint (Fig. 4F) whereas in Eumeces, portions of the fibrocartilage show more diffuse signs of mineralization of the tissues via von Kossa staining (Fig. 4A). The ventral portion of these dorsal fibers then often attach directly into the rostral portions of Meckel's cartilage. The ventral fibrous portion of the joint is composed largely of woven fibers that are deeply embedded in a thick field of Sharpey-fiber bone and radiate about the ventromedial angle of the dentary while passing orthogonal to the contralateral fibers in the sagittal and axial planes (Fig. 3). The rostral portion of the symphysis in Nerodia appears to possess a fibrocartilaginous nodule, perhaps similar to that described by Bellairs (1984); however due to tissue decomposition, these histological data were not as clear as was hoped.
Meckel's cartilage approaches the symphysis from the ventromedial surface of the dentary where it enters into the joint via the medially open Meckelian groove (Figs. 4, 5). Meckel's cartilage tapers rostrally, terminating as an endochondral ossification front along the rostromedial surface of the dentary (Fig. 5). The ossification fronts are evidenced by columnar arrangement of chondrones near the bony margin of the dentary, the deposits of calcified tissues that often complement the encroaching cartilages, and the appearance of osteocytes near the chondroosseous junction (Fig. 5D–J). The ventral, parallel-fibered portions of the fibrous joint often attach directly to the perichondrium of the cartilages. The paired Meckel's cartilages remain separate in many taxa including Eumeces, Varanus, Tupinambis, Lepidophyma (Fig. 4). Bellairs (1984) noted the symphyses of Chamaeleo, as well as Lacerta and Anguis, possessed fibrocartilage, but did possess fused Meckel's cartilages. However, in the iguanians studied here (Oplurus, Callisaurus, Pogona), and geckoes (Oedura, Hemitheconyx, Gecko), the cartilages are fused across the midline (Fig. 5). However, geckoes and iguanids differ in how they fuse Meckel's cartilages. The cartilages are continuous with one another with no clear separation between adjacent elements in geckoes (Fig. 5G). In iguanids, there is a clear septum between the two Meckel's cartilages where the two rami meet at the midline (Fig. 5A–C), but then diverge as separate crura rostrally to rejoin the dentary as separate ossification fronts (Fig. 5C, H–I). In the taxa with fused Meckel's cartilages, m. genioglossus attaches to the midline of the fused pairs, whereas in taxa with unfused cartilages the muscles appear to either attach along the medial surface of the cartilage, or to a midline raphe of connective tissue. Finally, in Oplurus, paired accessory ossifications (os mentale) are present along the ventral borders of the cartilages (Fig. 4C).
The distribution of fibrous and cartilaginous elements in the lizard symphysis suggests an association with the function and behavior of the joint (Kley, 2006). The fibrocartilaginous pads and tightly packed, parallel-fibered dorsal portion of the symphysis suggests this portion of the joint largely experiences compression while restricting separation of the joint. The ventral, woven-fibered, Sharpey-fiber bone-bounded parts of the joint likely experience shear forces and may accommodate independent movements of the individual mandibles with the dorsal portion serving as a hinge. The variably mineralized, narrower fibrocartilaginous pads in the harder-biting taxa, Eumeces, Gerrhosaurus, and iguanids (Herrel et al., 2007) suggest that these joints are indeed stiffer than those of the loose-fibered, kinetic varanids and geckos. Indeed, the fibrous portion of the joint bears the morphology of typical symphyses menti found among most mammalian vertebrates. However, in addition to possessing a fibrous symphysis, geckos and iguanids have fused Meckel's cartilages, which may constrain movements of the caudal portion of the joint as well as forming a scaffold for m. genioglossus during tongue movement in these two lingual-feeding clades (Fig. 6). Interestingly, Kley (2006) observed rotational and hingelike movements in the symphysis of the basal snake Leptotyphlops, despite it having a midline cartilaginous nodule connected to Meckel's cartilages. Thus, the fused cartilages in these lizards may not offer significant stiffening of the joint, though there may be a mobile joint between the nodule and Meckel's cartilages. On the other hand, both iguanids and geckos evolved apomorphic lingual behaviors that may rely on increased joint stiffening for hyolingual muscles that a fused Meckel's cartilage may provide. Iguanids evolved translational tongue protrusion (Reilly and McBrayer, 2007) in which the tongue is projected out of the mouth toward prey items (Meyers and Nishikawa, 2000) whereas geckos engage in drinking behaviors in which the tongue is extended into aqueous substrates (Jamniczky et al., 2009). Both behaviors require extensible, hydrostatic hyolingual systems that differ significantly from the chemosensation-based tongue extension behavior evolved in anguimorphs and snakes. Although these hypotheses require additional testing, the microanatomy indicates that the lizard symphysis is a dynamically loaded joint with a complex arrangement of different connective tissues.
In addition to their functional significance, the fusion of Meckel's cartilages in iguanids and geckoes suggests convergent events in lizard evolutionary morphology. Whereas phylogenetic relationships determined from morphological characters (Conrad, 2008; Fig. 1) find iguanids and geckoes to fall near the base of the clade, molecular analyses find iguanids to be allied with the anguimorphs (varanids, snakes, etc.) (Vidal and Hedges, 2005), which is a position that is quite distant from the basal geckoes. Interestingly, the histological data suggest that indeed these two clades fuse the cartilages via different mechanisms. The lack of a septum in geckos indicates that the cartilages likely fused early in ontogeny, whereas iguanids may fuse them later, and less completely. Bellairs and Kamal (1981) and Bellairs (1984) noted that Meckel's cartilages fuse during embryonic development in some taxa (e.g., Lacerta), but then become unfused in adult lizards. This suggests that the role of these cartilages in symphyseal construction may vary during ontogeny and additional developmental and comparative data are necessary to better understand the picture. Moreover, the septated cartilages in iguanians may be evolutionarily significant if indeed they did evolve from a stock of autarchoglossan lizards (Vidal and Hedges, 2005), many of which possess unfused, mobile symphyses (Fig. 6). Bellairs (1984) and Kley (2006) identified cartilaginous nodules (i.e., intraramal nodules) in the symphyses of some snakes that are similar to the linkages described here. However, Kley (2006) showed that the nodal cartilages lacked a perichondrium and characteristic mucopolysaccharides of Meckel's and other hyaline cartilages and thus were not homologous to Meckel's cartilages. This finding suggests that the midline fibrocartilaginous nodule present in the V. exanthematicus specimen may be of similar origin to the nodules in snakes (e.g., Nerodia, Fig. 4L), though not as discretely formed whereas geckos and iguanids posses hyaline-derived cartilaginous fusions.
Our understanding of the evolution of symphyseal functional morphology in lepidosaurs is relatively poor without complementary analyses of additional extant and fossil taxa. Moreover squamates have a rich fossil record that shows marked skull diversity within the clade (e.g., Evans, 2003; Mo et al., 2010), thus this article is likely just scratching the surface regarding symphyseal morphology. The rhynchocephalian, and commonly (mis)used primitive model for squamates, Sphenodon, is akinetic, displays jaw and lingual prehension behaviors (Reilly and McBrayer, 2007) and has a fibrous symphysis that allows some movement of the joint (Jones et al., 2009). The Meckelian fossa, which houses the distal portion of the cartilage is small, however, the morphology of Meckel's cartilage remains unstudied in this taxon. The fossil rhynchocephalian Gephyrosaurus (Evans, 1980) has an extraordinarily large Meckelian fossa at the symphyseal surface suggesting the cartilages may not only be huge but also fuse across the midline. Therefore, these new data suggest that extant lizards, as well as fossil taxa, possess a wide variety of soft-tissue morphologies that likely reflect different local functional and developmental environments. The combination of hyaline and fibrocartilages, as well as elastic fiber and bony contributions, suggest the symphysis is a complicated joint that may shed light on the adaptive plasticity and evolution of skeletal tissues in lepidosaurs.
The authors thank Steve Reilly, Susan Williams, and Lawrence Witmer at Ohio University, for access to specimens, equipment, and OUμCT scanning facilities. They thank Rudy Wang and Catherine Rushton at Marshall University for access to microscopy equipment. The development of this project benefited from discussions with Eric McElroy, Tobin Hieronymus, Robert Hikida, Dawn Holliday, Marc Jones, Lance McBrayer, Matt Vickaryous, and Anthony Herrel. Thanks to Jessie Maisano and NSF ATOL Deep Scaly (EF-0334961) for sharing Pogona micro-CT data and Phil Allman (Florida Gulf Coast University) for providing salvage specimens. Many thanks to the students of the 2009 Skeletal Tissues Biology class, for weathering the instructor's new found interest in symphyseal structure and function with patience and good humor. They also thank two anonymous reviewers for their input.