ARTICLE IN PRESS ZOOLOGY Zoology 110 (2007) 212–230 www.elsevier.de/zool Embryonic development of Python sebae – I: Staging criteria and macroscopic skeletal morphogenesis of the head and limbs Julia C. Boughnera,1,2, Marcela Buchtováa,2, Katherine Fua, Virginia Diewertb, Benedikt Hallgrı́mssonc, Joy M. Richmana, a Department of Oral Health Sciences, Life Sciences Institute, University of British Columbia, Life Sciences Centre, 2350 Health Sciences Mall, Vancouver, BC, Canada V6T 1Z3 b Department of Oral Health Sciences, Faculty of Dentistry, University of British Columbia, 2199 Wesbrook Mall, Vancouver, BC, Canada V6T 1Z3 c Department of Cell Biology and Anatomy, Faculty of Medicine, Heritage Medical Research Centre, 3330 Hospital Drive NW, Calgary, Alta., Canada T2N 4N1 Received 24 August 2006; received in revised form 21 January 2007; accepted 23 January 2007 Abstract This study explores the post-ovipositional craniofacial development of the African Rock Python (Python sebae). We first describe a staging system based on external characteristics and next use whole-mount skeletal staining supplemented with Computed tomography (CT) scanning to examine skeletal development. Our results show that python embryos are in early stages of organogenesis at the time of laying, with separate facial prominences and pharyngeal clefts still visible. Limb buds are also visible. By 11 days (stage 3), the chondrocranium is nearly fully formed; however, few intramembranous bones can be detected. One week later (stage 4), many of the intramembranous upper and lower jaw bones are visible but the calvaria are not present. Skeletal elements in the limbs also begin to form. Between stages 4 (day 18) and 7 (day 44), the complete set of intramembranous bones in the jaws and calvaria develops. Hindlimb development does not progress beyond stage 6 (33 days) and remains rudimentary throughout adult life. In contrast to other reptiles, there are two rows of teeth in the upper jaw. The outer tooth row is attached to the maxillary and premaxillary bones, whereas the inner row is attached to the pterygoid and palatine bones. Erupted teeth can be seen in whole-mount stage 10 specimens and are present in an unerupted, mineralized state at stage 7. Micro-CT analysis reveals that all the young membranous bones can be recognized even out of the context of the skull. These data demonstrate intrinsic patterning of the intramembranous bones, even though they form without a cartilaginous template. In addition, intramembranous bone morphology is established prior to muscle function, which can influence bone shape through differential force application. After careful staging, we conclude that python skeletal development occurs slowly enough to observe in good detail the early stages of craniofacial skeletogenesis. Thus, reptilian animal models will offer unique opportunities for understanding the early influences that contribute to perinatal bone shape. r 2007 Elsevier GmbH. All rights reserved. Keywords: African Rock Python; Cartilage; Skeletogenesis; Embryo staging; Craniofacial development Corresponding author. E-mail address: richman@interchange.ubc.ca (J.M. Richman). Current address: Department of Cell Biology and Anatomy, Faculty of Medicine, Heritage Medical Research Centre, 3330 Hospital Drive NW, Calgary, Alta., Canada T2N 4N1. 2 These authors contributed equally to this work. 1 0944-2006/$ – see front matter r 2007 Elsevier GmbH. All rights reserved. doi:10.1016/j.zool.2007.01.005 ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 Introduction Reptile biology has been investigated for over a century by many different workers, yet there remains a paucity of studies on embryos (Parker, 1879; Kamal and Hammouda, 1965a–c; De Beer, 1937). The exceptions are that reptilian embryos have been used to study tooth and palate development (Lemus et al., 1980, 1986; Ferguson, 1981a, b; Westergaard and Ferguson, 1987) and temperature-dependent sex determination (Wibbels et al., 1998; Torres-Maldonado et al., 2001, 2002). More recently, nonavian reptiles, including snakes (Cohn and Tickle, 1999) and turtles (Nagashima et al., 2005; Ohya et al., 2005), are becoming popular animal models among evolutionary developmental biologists. The more recent studies are beginning to clone genes and look at expression patterns in reptiles (Kuraku et al., 2005; Ohya et al., 2005). Compared to mammalian and chicken models, much less is known about reptilian genomes (Matsuda et al., 2005). In order to identify the most interesting molecular questions, it is necessary to see where the structural differences lie between reptilian and other amniote embryos. It is from this anatomical variation and our previous knowledge of development in other vertebrates that we can formulate hypotheses about genetic and morphological change during evolution. Here, we characterize a non-venomous, egg-laying snake, Python sebae, as a developmental model with special emphasis on craniofacial development. We selected an oviparous member of the Boidae snake family for several reasons. First, pythons lay large clutches of eggs (40–100). Second, embryos are accessible during the time when craniofacial skeletal differentiation and odontogenesis are ongoing. Third, nonvenomous snakes such as the python lack the specialized fangs and supporting jaw modifications of venomous snakes. Although constrictor snakes have their own suite of synapomorphic jaw characters related to swallowing large prey whole, their jaw architecture may be more typical of the class Serpentes. The morphology of the snake skull incorporates key reptilian features of the jaw and palatal complex, such as a doubly articulated jaw joint and a natural cleft palate. These features have been retained from the most primitive amniotes, and should inform us about both snake and amniote evolution. Other than the late embryonic and prehatching development of P. sebae (Branch and Patterson, 1975), there are little developmental data available for this snake. The most comprehensive, detailed studies of cranial and gnathic morphology, musculature and function are limited to adult pythons (P. sebae, P. molurus; Frazzetta, 1959, 1966). These careful analyses of adult python morphology (Fig. 1) are a solid foundation for our investigation of the prehatching developmental morphology of P. sebae. 213 To facilitate developmental work on snakes, we need to develop a staging system that is generally applicable to other members of this class. Excellent classic and widely used staging tables have been published for mouse (Theiler, 1972), chicken (Hamburger and Hamilton, 1951) and turtle (Yntema, 1968). There are also three staging tables for embryonic development in snakes: the viviparous garter snake, Thamnophis sirtalis (Zehr, 1962); the viviparous asp viper, Vipera aspis (Hubert and Dufauré, 1968); and, most recently, the oviparous monocled cobra, Naja kaouthia (Jackson, 2002). Further, incomplete staging data were published for the viviparous brown water snake Natrix taxispilota (Franklin, 1945). However, to date no data are available for oviparous, non-venomous snakes. Therefore, one of our aims is to provide as complete information as possible on the external and internal morphology of P. sebae to facilitate the staging of other oviparous snakes. In addition to staging, the other main focus of this paper is craniofacial skeletal development. The adaptation of skeletal patterns during evolution is particularly interesting in the skull where gain and loss of fenestrations and joint morphology are important means of classifying an amniote (De Beer, 1937; Carroll, 1988). However, relatively little comparative work has been carried out on the developing craniofacial complex in embryos. Here we use whole-mount skeletal staining and computed tomography (CT) scanning to investigate the ossification of each of the skull bones and cartilages in P. sebae. We find that due to the relatively slow development of ectothermic reptiles, the timing of ossification can be resolved down to the level of individual bones. This is a distinct advantage as compared to endothermic, more rapidly developing amniotes. This work is our first of two papers on P. sebae, the second focusing on microscopic anatomy and cellular dynamics. We have separated the macroscopic from the microscopic data in order to make it more convenient for other investigators who want to use our data to study oviparous, non-venomous snakes. Materials and methods Python egg acquisition and incubation We obtained P. sebae eggs from the Rainforest Reptile Refuge (Surrey, British Columbia, Canada). Eggs were laid in July 2004 and June 2006 after an undisturbed period of in utero incubation of approximately 8 weeks (P. Springate, pers. comm.). Of the clutch of 40 eggs laid in 2004, we discarded 13 dead or infertile eggs and incubated the surviving clutch in a sand-filled bucket in warm, humid conditions ARTICLE IN PRESS 214 J.C. Boughner et al. / Zoology 110 (2007) 212–230 Fig. 1. Adult skull of Python sebae. (A) Cranial and upper jaw skeleton, lateral view, (B) lateral and (C) medial views of the lower jaw skeleton, (D) palatal view of the cranium. Arrows show axes that are used in the current paper. (A) reproduced from Frazzetta (1966) with permission of John Wiley & Sons, Inc.; (B–D) adapted from Frazzetta (1966). Abbr.: a – angular; bo – basioccipital; bs – basisphenoid; co – coronoid; d – dentary bone; ec – ectopterygoid; eo – exoccipital; f – frontal bone; mx – maxilla; n – nasal bone; p – parietal bone; pa – prearticular process of the compound bone; pf – prefrontal bone; pl – palatine bone; pm – premaxilla; po – post-orbital bone; pr – prootic; ps – parasphenoid; pt – pterygoid; q – quadrate; s – stapes; sm – septomaxilla; so – supraorbital bone; soc – supraoccipital bone; sp – splenial bone; st – supratemporal; su – surangular process of the compound bone; tc – trabeculae cranii; v – vomer. (day temperature about 30 1C, night temperature about 25 1C) for 1 month at the Rainforest Reptile Refuge. We then transported all remaining eggs to the Richman Lab, where they were incubated at 30 1C in a modified chicken egg incubator packed with moist vermiculite, a highly water-absorbing material that properly cushioned and hydrated the eggs (minimum 80% humidity). In 2006, we obtained 21 eggs from the same python mating pair and were able to collect five stage 1 specimens. Unfortunately, the other embryos were dead upon collection. The normal total incubation period for P. sebae is between 80 and 90 days. All animal work was reviewed and approved by the UBC Animal Ethics Committee, certificate no. A04-0271. described (Plant et al., 2000). No whole-mount skeletal data were available for stage 9 due to damage to the collected specimen. CT scanning Two fixed embryos (stage 6; stage 10) were CTscanned to visualize skeletal structure in three dimensions (3D) at the 3D Morphometrics Centre, University of Calgary. The stage 6 embryo was scanned using a Skyscan 1072 100 kv micro-CT at 5 mm resolution. The stage 10 embryo was scanned using a Scanco Viva-CT 40 scanner at 25 mm resolution, which is more suitable for larger animals. Fixation, processing and skeletal staining Embryos were removed from the egg and immersed in ice-cold phosphate-buffered-saline immediately. Embryo weight and length were measured, and pictures were taken of external head, body and tail morphology. For whole-mount skeletal staining, embryos were fixed in 100% ethanol for 4 days followed by 100% acetone for 4 days. Embryos were stained for bone and cartilage with Alizarin Red/Alcian Blue solution and were cleared in glycerol and potassium hydroxide as previously Results Here, we report the embryonic development of the oviparous, non-venomous African Rock Python (P. sebae). We first describe staging criteria based on external characteristics visible upon recovery of the embryo (Table 1). In addition, we describe the internal development of the craniofacial complex at each of the stages using whole-skeletal staining and micro-CT ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 Table 1. embryos Staging characteristics for post-oviposition snake Stage External morphological characteristics 1 Mandibular process does not extend rostral to eye, endolymphatic ducts present, no scales on body or head, 5–6 body coils Mandibular process extends rostral to eye Mandibular process extends midway between eye and upper jaw, scales visible on body only, 3–4 body coils Mandibular process lines up with upper jaw, 3 body coils Cervical flexure 4901, heart protrudes from body cavity Scales visible on the head, heart fully enclosed inside body cavity, 2–3 body coils Fusion of eyelids in the center of eye, 2 body coils Cervical flexure 1201, body wall closed, pigmentation starting on body and head, 2 body coils Cervical flexure 1801, brain not visible through skin, 1–2 body coils, pigmentation pattern is clear but color is pale Endolymphatic ducts not visible, embryo resembles neonate python, with the head pigmentation being darker than the body, 1–2 body coils. Egg tooth visible intraorally 2 3 4 5 6 7 8 9 10 scanning. The timing and sequence of appearance of skull bones is very dynamic and takes place over the first month and a half of post-ovipositional (po) development. The sequence of ossification is therefore diagnostic of a particular stage in development. We also report growth curves so, together with our other analyses, it will be possible to predict the extent of tissue development for a particular stage (Table 2, Figs. 2A and B). 215 External morphology of P. sebae In general, the main characters used for staging include the relative length of the mandible in relation to the maxilla (a measure that, together with skeletal data, provides some indication of how far an embryo has progressed); the visibility of the developing brain, heart and the closure of body wall musculature; the angle of the cervical flexure; and the presence or appearance of the eyelid, scales and endolymphatic ducts (Table 1). Weight and length were also used to help stage embryos. Weight gain in P. sebae is exponential with increasing embryonic stage (Fig. 2A), whereas the relationship between embryo length (head to tail) and stage is linear (Fig. 2B). Observation of certain characteristics depends on translucency of the embryo and thus it is best if fresh specimens are examined. These features include the visibility of the endolymphatic duct, the brain and the heart. Stage 1 (1–3 days po; Figs. 3A and B; Table 2). Scales are totally absent from the body. Retinal pigmentation is light. The paired endolymphatic ducts are visible as two white opacities, on the dorsal surface of the head (Fig. 3B). The facial prominences are visible, including the nasal slit delineating the frontonasal mass and lateral nasal prominences. In addition, the paired maxillary and mandibular prominences could be seen. From the lateral view, the tip of the mandibular prominences lined up with the rostral edge of the eye. The second and third pharyngeal arches and clefts are visible (Fig. 3A and data not shown). Bilateral limb buds are found in the pelvic area (Fig. 4A, B). Stage 3 (11–12 days po; Figs. 3C and D; Table 2). In fresh embryos, the brain is visible through the transparent neurocranial tissues (Fig. 3C). The length of the mandibular process extends rostrally and usually is midway between the eye and the tip of the upper jaw (Fig. 3D). The upper facial prominences are completely Table 2. External features of our sample of Python sebae at different stages of development Stage Days postoviposition No. of specimens collected Body length (cm) Body weight (g) Mean head length (mm) No. of complete body coils Endolymphatic ducts visible Body wall closed Angle of cervical flexure (degrees) 1 3 4 6 7 8 9 10 1–3 11, 12 18 33 44 54 61 75 5 14 4 3 3 1 1 1 No data 10.0–13.5 (10) 14.0–15.0 (2) 16.0–19.5 22.0 (1) 26.0 37.0 48.0 0.3–0.5 0.82–1.6 (10) 1.5–2 (2) 4.2–6.1 9.4 (1) 8.7 26.0 50.5 6.0 8.0 (10) 10.5 (2) 12.8 15.2 (1) 17.9 NA NA 5–6 3–4 3 2.5 2 2 1–2 1–2 Yes Yes Yes Yes Yes Yes Yes No No No No No No Yes Yes Yes 901 901 901 90–1201 90–1201 120–1801 1801 1801 Summary of Python sebae embryo mean body length (tip of upper jaw to tip of tail) and weight, and rostro-caudal head length (measured from mesencephalon to tip of upper jaw in stages 1–6 embryos and measured crown of head to tip of upper jaw in stages 7–10 embryos). Embryonic stages were defined by several external morphological characteristics, the principal of which are listed here. In stages 3, 4 and 7, a subset of embryos was measured (number in parentheses). ARTICLE IN PRESS 216 J.C. Boughner et al. / Zoology 110 (2007) 212–230 Stage 7 (44 days po; Figs. 3I and J; Table 2). The body wall musculature is still open along the ventral midline and the heart remains visible through the body wall. The brain is still visible through the skull. Distinct scales are present on both the body and head (Fig. 3J). The transparent upper and lower eyelid edges have fused together. The limbs and hemipenes are not as prominent as at stage 6 (Fig. 4I). Stage 8 (54 days po; Figs. 3K and L; Table 2). The body wall musculature is now completely fused along the ventral midline. However, the heart remains visible through the body wall (Fig. 3K). The pigmentation on the head and body is distinct. Stage 9 (61 days po; Figs. 3M and N; Table 2). The heart remains visible through the fused body wall. The pigmentation of the scales on the head and body now assumes a species-specific pattern but is still lightly colored. The brain is no longer visible through the skin. Stage 10 (75 days po; Figs. 3O and P; Table 2). The heart is no longer visible through the fused body wall (Fig. 3O). Furthermore, the endolymphatic ducts are no longer visible through the skin of the head (Fig. 3O). The embryo closely resembles a neonate python, with complex head and body pigmentation (Figs. 3O and P). Upon opening the mouth, the egg tooth is erupted and visible on the inside tip of the upper jaw (not shown). Fig. 2. Embryonic growth charts for Python sebae. (A) Embryo weight (g) versus post-oviposition developmental age (days). Embryo weight gain was exponential over time. (B) Embryo length (cm) versus age (days). Increase in embryo length was linear over time. (A and B) Embryonic stages (S) are listed for corresponding data points. fused. The retina is well pigmented; however, pigment cells are absent from the skin. Scales are only visible along the dorsal body and not on the head. As hindlimbs are rudimentary in modern snakes, external limb size and morphology do not change from stage 3 and cannot be used for staging embryos (Figs. 4C, E, G, I and K). The hemipenes are visible medial to the limb bud (Fig. 4C). Stage 4 (18 days po; Figs. 3E and F; Table 2). The body wall musculature remains open in the ventral midline, through which the heart is visible (not shown). Body scales are present both dorsally and ventrally (not shown). The eyelids remain open but cover about half of the eye (Fig. 3F). Stage 6 (33 days po; Figs. 3G and H; Table 2). Head scales are visible for the first time in the lateral lower jaw and lateral cranium (not shown). While the skin of the head is translucent and unpigmented, both eyes have dark pigmentation. Eyelids form a band of approximately 1/3 the width of eye radius (Fig. 3H). Limb development At stage 1, the python limb bud shape corresponds to that of a stage-21 chicken. However, the chicken limb bud has a much more prominent apical ridge at a similar stage (Hamburger and Hamilton, 1951; Raynaud, 1985; Cohn and Tickle, 1999; Figs. 4A and B). By stage 3, there is no evidence of an ectodermal ridge (Fig. 4C). There are no skeletal elements yet present in the limb buds (Fig. 4D). At stage 4, on each side of the pelvic area, two simple and rod-like hindlimb cartilages have begun to form between the body and tail, immediately caudal to the last rib (Fig. 4F). Both elements are aligned with each other along the same axis, parallel to the body. In the limb at stage 6, two cartilages are present caudal to the last ribs: the stylopod of the hindlimb (vestigial femur); and a pelvic cartilage with three processes representing the ischium, ilium and pubis elements (Fig. 4H). Thus, the shape of the pelvic element is tripartite and more complex at this stage compared to stage 4. Bilaterally at stage 7, the hindlimb bud cartilages have not progressed any further as compared to stage 6 appendages (Figs. 4I and J). Thus, final limb skeletal pattern is established at stage 6 in the python. ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 217 Fig. 3. External body (left column) and head (right column) morphology, illustrating external morphological characters used to stage Python sebae embryos. Specimens in a–c, e, g–p were photographed fresh, specimens in d and f were photographed in 70% ethanol following fixation. The second pharyngeal arch and facial prominences were visible as a separate structure only in stage 1 embryos (A and B). The endolymphatic duct (eld) was present from stage 1 (B and F, arrow) and is best seen in either fresh (A and B) or stained and cleared specimens (E ii). Body coil number (cl) decreases through development beginning with a high of up to 6 coils in stage 1 embryos (A) and decreasing to 1–2 coils (O) by stage 10. The body wall (bw) gradually encloses the internal organs and is complete by stage 8. By stage 6, the eyelid covers a band approximately 1/3 of the eye radius (bar in H). Pigment pattern (pg) became obvious from stage 8 onwards (K–P). Scale bar for whole bodies ¼ 1 cm, bar in b ¼ 2.5 mm and for (D, F, H, J, L, N, and P) ¼ 5 mm. Abbr.: bw – body wall; sel – eyelid; eld – endolymphatic duct; ht – heart; mdp – mandibular prominence; mxp–maxillary prominence; ns – nasal slit; p2 – second pharyngeal arch; sc – scale. Macroscopic craniofacial skeletal development of P. sebae Stage 1–3 (1–11 days po). Considerable development had taken place inside the mother prior to oviposition. Although stage 1 embryos were well into organogenesis, no skeletal tissues were detected with whole-mount staining (not shown; Buchtová et al., 2007). Skull pattern is initially established with the cartilaginous chondrocranium; therefore, we expected that the chondrocranium would have almost completely formed by stage 3 (Table 3). However, the nasal capsule had not formed in three of the six stage 3 specimens that we analyzed. In these more developed specimens, the nasal capsule, trabeculae cranii, occipital cartilage, quadrate, otic capsule and Meckel’s cartilage had formed (Figs. 5A–C). Long rods of trabecular cartilage extended rostrally from the level of the vomeronasal organ, behind the eyes. Rostrally, the paired trabecular cartilages were fused in the midline (Fig. 5C). In the lower jaw, paired Meckel’s cartilages were well separated and unfused at the symphysis (Figs. 5A and B). The quadrate was developing dorsomedially to Meckel’s cartilage and ventrolaterally to the otic capsule (Fig. 5A). The parachordal plate, including the occipital cartilage, was well developed (Fig. 5C). Many of the chondrocranial elements subsequently undergo endochondral ossification at later stages (Table 3). In addition to endochondral bones, numerous neural crest-derived intramembranous bones form directly from condensing mesenchyme, which is typical for ARTICLE IN PRESS 218 J.C. Boughner et al. / Zoology 110 (2007) 212–230 Fig. 4. Morphology of the residual limb buds (lb) of Python sebae. External features are visible in (A–C, E, G, I and K) and skeletal morphology in (D, F, H and J). An opaque ectodermal ridge is visible in fresh stage 1 specimens (A, B, arrows) but not in older stage embryos (C, E, G). Box in A is illustrated at higher power in (B). Hindlimb skeletal development started at stage 4 (F) just caudal to the ribs. A pelvic rudiment and femur have formed. (G–J) Lateral view of tail and limb buds. The pelvic and femoral rudiments are fully formed at stage 6 and slightly increased in size at stage 7. (K) This specimen may be female since the limb buds and hemipenes are more recessed into the body. Abbr.: fm – femur; hp – hemipenes; lb – limb bud; pv – pelvic rudiment; r – last rib. gnathostomes. The earliest bone to ossify is the pterygoid (Fig. 5A; Table 4). The palatine and other jawbones could not reliably be seen in stage 3 wholemount stained specimens (Fig. 5A; Table 4). Stage 4 (18 days po). The same chondrocranial elements present at earlier stages were represented in stage 4 specimens and none had begun to ossify (Table 3). Furthermore, the cartilages of the trabeculae cranii and nasal capsule were fused bilaterally at this stage (Figs. 5D and G). Many of the intramembranous bones were present at stage 4. Ossification of the rod-like maxillary bone had J.C. Boughner et al. / Zoology 110 (2007) 212–230 Table 3. 219 Endochondral bones and persistent cartilages in the craniofacial skeleton Persistent cartilages Meckel’s cartilage Nasal capsule Vomeronasal cartilage Trabeculae cranii Orbital cartilage Endochondral bones Quadrate Hyoid Parachordal plate (later basisphenoid, basioccipital, exooccipital, supraoccipital) Otic capsule (later prootic bone) Stage 1 Stage 3 Stage 4 Stage 6 Stage 7 Stage 8 Stage 10 Mcs   Mcs  + +  + + + +  + + + + + + + + + + + + + + + + + + + + + + Mcs Mcs + + + + + + + + + Os Os + Os (basisphenoid) Os Os Os Os Os Os + + + + Os Os Os Summary of the embryonic stages at which the cartilages of the craniofacial skeleton are first observed in Python sebae. Persistent cartilages remained intact until the end of the study (stage 10) with the exception of the posterior end of the trabeculae cranii, which undergoes perichondrial ossification after stage 8. Note that stage 8 specimens were only analyzed in tissue sections but were included in the table for completeness. Key: (+) cartilage present, () cartilage not yet developing, mcs – mesenchymal condensation in sections, Os – ossification. begun at the lateral edges of the upper jaw (Figs. 5D, F and G). Only one small ossification center was recognizable in this stage. The triangularly shaped palatine bone and pterygoid bones could easily be detected in whole-mount stained stage 4 embryos (Figs. 5D–G). Ossification of the nasal bones had begun (Fig. 5D, Table 4). Of the cranial vault bones, the prefrontal and frontal bones had just begun to ossify bilaterally (Fig. 5D). The snake has a doubly jointed jaw articulation that increases the gape in order to swallow large prey. Instead of the quadrate bone or condyle nesting into a fossa (i.e. the glenoid fossa), the snake quadrate articulates with the lateral side of the supratemporal bone of the skull (Frazzetta, 1966). At stage 6, the supratemporal bone was beginning to develop from a single ossification center, lateral to the otic capsule and dorsocaudal to the cartilaginous quadrate (Fig. 5E). It was now possible to see some of the mandibular bones. Moving from proximal to distal (caudal to rostral), the surangular and prearticular processes of the compound bone as well as the splenial bone had formed (Figs. 5D–F). Distally, the dentary bone was ossifying directly lateral to Meckel’s cartilage (Figs. 5D and F). All of these bones were separate but closely related to Meckel’s cartilage as seen in other amniotes. Stage 6 (33 days po). The key difference between stage 6 and earlier embryos is that endochondral ossification is starting in many of the cartilages (Table 3). Several new intramembranous bones had formed between stages 4 and 6 including the premaxilla, vomer, ectopterygoid and parietal bones (Fig. 6A, Table 4). In addition to using whole-mount staining to assess skeletal development (Fig. 6), we used high-resolution micro-CT prior to staining (Figs. 7A–H). The advantage of this method is that it was possible to digitally dissect individual skull bones away from surrounding tissues; however, the less radio-dense cartilages were not captured using micro-CT. Rostrally, the triangular premaxilla appeared as a single bone (Figs. 6A, C and 7A–C, E). Caudally, the premaxilla was tripartite, composed of separate processes arising on either side of the main body of the bone (Fig. 7E). There were two lateral palatine processes and a single midline nasal process. In between these processes was the distal end of the trabecular cartilage, which at this stage was joining with the nasal capsule (Figs. 6A and C). The maxillary bone extended from the premaxilla to the pterygoid (Figs. 6C and 7B–D). The palatine bone had developed a pterygoid process, although the palatine and pterygoid bones were not yet in contact with each other (Figs. 6D, G and 7D). For the first time, the ectopterygoid was visible dorsal to the maxillary bone and lateral to the pterygoid bones (Figs. 6C and 7C, D, F). The paired septomaxillae were beginning to ossify ventrolaterally to the vomers (Figs. 6C and 7B, C, H). The nasal bones were now ossifying in the midline, dorsal and exterior to the nasal capsular cartilage (Figs. 6A and 7A–C). Calvarial bones that were ossifying included the prefrontal, frontal and parietal bones (Figs. 6A and 7A–D). The prefrontal bone was ossifying in three distinct but loosely connected parts (Figs. 7B–D). One process extended medially towards the midline of the frontal bone. A second process extended laterally to form part of the roof of the orbit. A third process extended ventrolaterally towards the palato-maxillary arch, which later articulates with this process. The frontal bone was ossifying dorsal to the prefrontal bone (Figs. 6A and 7A–C). ARTICLE IN PRESS 220 J.C. Boughner et al. / Zoology 110 (2007) 212–230 Fig. 5. Stages 3 and 4 – Skeletal anatomy of craniofacial structures. (A, D, E) lateral views, (B, F) ventral views, mandible intact, (C, G) palatal views, mandible deflected. (A–C) Chondrocranium is nearly fully formed in this specimen including the nasal capsule, trabeculae cranii, quadrate, otic capsular cartilage and Meckel’s cartilage. The vomeronasal cartilage is not visible yet. One membranous bone is visible (pterygoid) although not well-stained. (D–G) Many intramembranous bones are easily visible at this stage, including maxilla, prefrontal, nasal, palatine, pterygoid, surangular, splenary, prearticular and dentary. (E) shows a higher power view of the jaw joint. Scale bars for (A–D, F and G) ¼ 2 mm and for (E) ¼ 1 mm. Abbr.: d – dentary bone; f – frontal bone; h – hyoid; mc – Meckel’s cartilage; mx – maxilla; n – nasal bone; ncs – nasal capsule; ocs – otic capsule; pa – prearticular process of the compound bone; pc – parachordal plate; pf – prefrontal bone; pl – palatine bone; pt – pterygoid; q – quadrate; sp – splenial bone; st – supratemporal; su – surangular process of the compound bone; tc – trabeculae cranii. The lower jaw bones arose around Meckel’s cartilage (Figs. 6A, B and 7A, B). The dentary bone was bi- and tripartite in sections across more caudal levels in the transverse plane. The surangular and prearticular processes of the compound bone continued to ossify lateral and medial to Meckel’s cartilage, respectively. The angular and coronoid bones were visible for the first time. The angular bone was caudal to the splenial bone. The coronoid bone was located dorsal to Meckels’ cartilage and caudal to the splenial bone. The intramandibular joint between the dentary bone and the more proximal lower jaw bones was forming at stage 6. Stage 7 (44 days po). Analysis of one recently dead specimen in whole-mount stained preparation indicated that the parasphenoid and supraorbital were the only ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 Table 4. 221 Sequence of ossification of intramembranous craniofacial bones Stage 1 Stage 3 Stage 4 Stage 6 Stage 7 Stage 8 Stage 10 Mandibular complex Dentary Compound bone (prearticular process) Compound bone (surangular process) Splenial Angular Coronoid Supratemporal — — — — — — — 2 2 — — — — 2 1 1 1 1 — — 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Maxillary complex Palatine Pterygoid Maxillary Premaxilla Vomer Ectopterygoid Septomaxilla — — — — — — — 2 1 — — — — — 1 1 1 2 2 — — 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Cranial vault Nasal Parietal Prefrontal Frontal — — — — — — — — 1 2 1 — 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Basicranial Parasphenoid Postorbital Supraorbital — — — — — — — — — — 1 — 1 1 2 1 1 1 1 1 1 Schedule of appearance of craniofacial intramembranous bones of Python sebae as detected with either Alizarin Red staining in whole-mount or Picosirius Red staining in sections. Once formed, bones remained intact until the end of the study (stage 10). Stage 8 specimens were only examined in section. However, we are confident that the same bones would be visible in whole mount at stage 8 that were already visible at stage 7. Key: (—) bone not observed, (1) bone present in whole-mount stain, (2) bone present in histological sections only. new membranous bones to form at this stage, although only the parasphenoid could be detected in wholemount stained specimens (Figs. 6D and F; Table 4). Ossification of the quadrate, basisphenoid and prootic cartilages also had started (Table 3). The vomer bone had three processes: the horizontal body, and the medial and lateral vertical lamina (not shown). Rostrally, the shape of the maxillary bone was more complex, owing to the formation of processes for muscle attachments dorsally and ventrally (Figs. 6D and E). Further caudally, the maxillary bone was tubular in shape. The centrally located parasphenoid was first observed in the midline between the pterygoid and palatine bones (Fig. 6F). The frontal bone was ossifying dorsally and extended ventrally immediately lateral to the brain to form part of the braincase and the medial wall of the orbit (Fig. 6D). Several unerupted teeth in the upper jaw were sufficiently mineralized to be visible (Fig. 6D). At stage 7, the spatial arrangement of the lower jaw bones was similar to the arrangement found in the adult python (Figs. 1B and C), where the coronoid (dorsal), prearticular (medial), angular (ventral) and surangular (ventrolateral) were arranged around Meckel’s cartilage (Figs. 6D and E). Stage 10 (75 days po). The formation of craniofacial and lower jaw cartilages and bones was complete at stage 10, 10–20 days before hatching (Figs. 8A–G; Tables 3 and 4). The CT images clearly showed the near-complete ossification of the craniofacial bones (Figs. 8A–D). Following CT scanning, the specimen was hemisected and one half was embedded in paraffin, while the other half was stained in whole mount. The maxillary bone was structurally complex (Figs. 8A–G), with many fenestrations on its lateral aspect and numerous processes for muscle attachment. At stage 10, the lateral process of the palatine bone articulated with the prefrontal bone and contacted the maxillary bone (Figs. 8B and D). The rostral part of Meckel’s cartilage was intact (Figs. 8E–G), consistent with the common persistence of this cartilage into post-natal stages in reptiles. An intramandibular joint was present on each side of the mandible between the dentary bone and the compound bone (Figs. 8A and D). The caudal dentary bone formed a bifurcation, into which the surangular process of the compound bone was recessed ARTICLE IN PRESS 222 J.C. Boughner et al. / Zoology 110 (2007) 212–230 Fig. 6. Stages 6 and 7 – Skeletal anatomy of craniofacial structures. (A–C) The same stage 6 specimen as in Fig. 7, stained in whole mount after CT scanning. (A) lateral view, (B) ventral view, mandible intact, (C) palatal view, mandible retracted. All membranous bones are present by this stage. (D–F) Stage 7 skull in whole mount. Specimen has been distorted in medial–lateral plane due to fixation after death of embryo in ovo. Nonetheless, more advanced development can be seen as compared to the stage 6 specimen. (D) The single, midline, parasphenoid bone can be seen for the first time at stage 7. This bone is also clearly visible in the palatal view (F). Unerupted upper teeth are visible (arrowheads in D). All mandibular bones are well developed. Scale bar for (A–F) ¼ 2.0 mm. Abbr.: a – angular; bs – basisphenoid; co – coronoid; d – dentary bone; ec – ectopterygoid; eld – endolymphatic duct; f – frontal bone; imj – intramandibular joint; mc – Meckel’s cartilage; mx – maxilla; n – nasal bone; p – parietal bone; pa – prearticular process of the compound bone; pf – prefrontal bone; pl – palatine bone; pm – premaxilla; po – post-orbital bone; pr – prootic; ps – parasphenoid; pt – pterygoid; q – quadrate; st – supratemporal; su – surangular process of the compound bone; tc – trabeculae cranii; v – vomer. (Figs. 8A and D). At stage 10, the first tooth generation was erupted and five bones were tooth-bearing: the premaxilla, maxillary, palatine, pterygoid and dentary (Figs. 8A–D). Thus, there were two rows of teeth in the upper jaw and one in the lower jaw. Discussion Our study of P. sebae further demonstrates that the general process of skeletal differentiation is conserved, while the pattern of bones – the ossification sequence, number, morphology, positions – is quite divergent among reptiles, mammals and birds. Importantly, because the python develops more slowly than a mammalian model animal, it is simpler to identify the order of ossification for each of the membranous bones in this snake. For example, in mouse, many dermal bones begin to ossify within about 24 h of each other, beginning at embryonic day 15 (Johnson, 1933) (Table 6). ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 223 Fig. 7. Micro-CT scans of the skull of Python sebae at stage 6. Cartilage does not have sufficient radiodensity to be detected. (A) Lateral view with soft tissues illustrated around bones. (B) Tilted, lateral view showing both sides of the skull. The characteristic morphology of most bones is visible at this stage. (C) Hemicrania show internal structures in medial view. This view reveals the position of the vomer and septomaxilla, bones that are obscured from view in whole-mount preparations. (D) Palatal view of skull with the mandible digitally removed, showing more clearly the position of the ectopterygoid. (E–H) Digitally dissected bones. (E) Premaxilla – dorsal (i) and lateral (ii) views. (F) Left pterygoid and ectopterygoid – ventral (i) and lateral (ii) views. (G) Pterygoid and ectopterygoid – bilateral, dorsal view. (H) Vomer and septomaxillae – anterior (i) and vomer, lateral (ii) views. These two bones are at the limit of resolution of the CT machine. Scale bar in (B) applies to (A–D) ¼ 1.0 mm, bar in (E) applies to (E–H) ¼ 1.0 mm. Abbr.: a – angular; co – coronoid; d – dentary bone; ec – ectopterygoid; f – frontal bone; mx – maxilla; n – nasal bone; p – parietal bone; pa – prearticular process of the compound bone; pf – prefrontal bone; pl – palatine bone; pm – premaxilla; pt – pterygoid; sm – septomaxilla; sp – splenial bone; st – supratemporal; su – surangular process of the compound bone; v – vomer. ARTICLE IN PRESS 224 J.C. Boughner et al. / Zoology 110 (2007) 212–230 Fig. 8. Stage 10 – skeletal and microscopic anatomy of craniofacial structures. (A–D) 3D reconstructions using micro-CT scans of the skull – (A) lateral view, (B) dorsal view, (C) medial view with mandible digitally removed, (D) ventral view with mandible digitally removed. The adult bone morphology is present at this stage (compare Fig. 1). All sutures and joints have formed including the intramandibular and maxillary–premaxillary joints. The cranial vault sutures are fusing. After CT analysis, the same embryo was cut in half, the left half processed for whole-mount staining (E–G). Even though similar information can be obtained from CT or whole-mount stained specimens, the CT scans have a distinct advantage in that the intact specimen can be digitally dissected and the specimen is preserved intact for future microscopic examination. (E) Lateral view, (F) medial view, (G) ventral view, mandible intact. Scale bar for (A–G) ¼ 5 mm. Abbr.: a – angular; bo – basioccipital; bs – basisphenoid; c – compound bone; co – coronoid; d – dentary bone; ec – ectopterygoid; f – frontal bone; imj – intramandibular joint; mc – Meckel’s cartilage; mx – maxilla; n – nasal bone; ncs – nasal capsule; p – parietal bone; pa – prearticular process of the compound bone; pc – parachordal plate; pf – prefrontal bone; pl – palatine bone; pm – premaxilla; po – post-orbital bone; pr – prootic; ps – parasphenoid; pt – pterygoid; q – quadrate; s – stapes; sm – septomaxilla; so – supraorbital bone; soc – supraoccipital bone; sp – splenial bone; st – supratemporal; su – surangular process of the compound bone; th – tooth; v – vomer. A staging system for oviparous snakes based on external morphology We found relatively close congruence of our python data with studies of egg-laying venomous snakes (Haluska and Albrecht, 1983; Jackson, 2002). Since oviparous snakes lay their eggs during organogenesis stages, it is important to determine where in the entire spectrum of snake development our sample fits. One of the most comprehensive studies of early embryonic snake development was carried out on the viviparous T. sirtalis sirtalis (the garter snake) where 1225 specimens were collected (Zehr, 1962). The morphology of the stage 26 garter snake corresponds very closely to that of ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 the stage 1 python and includes the following features: the external naris had not formed, the grooves between the pharyngeal arches had not completely filled in, and there were 5.5–6.25 coils (Zehr, 1962). Our assessment also agrees with that of others (Holtzman and Halpern, 1991) who in the process of analyzing in detail the olfactory organs of the garter snake also presented detailed staging criteria for the face. Others (Haluska and Albrecht, 1983) compared a 14-day Elaphe obsoleta embryo (akin to stage 3 in the present study) to a stage 28 garter snake. Taken together, our stage 1 embryos are between stages 23 and 28 in the Zehr staging system. There were a total of 37 stages to reach the appearance of a hatchling garter snake and we have arrived at a similar number in our study (assuming stage 26 at oviposition followed by an additional 10 distinct stages). Other viviparous snakes have also been staged carefully from inception but a larger number of stages were described (Hubert and Dufauré, 1968; Hubert, 1985). The main differences between the study by Zehr (1962) and those of Hubert (Hubert and Dufauré, 1968; Hubert, 1985) are about 6 or so extra stages during gastrulation and neurulation. Similar headfold stage embryos are listed as stage 11 in Zehr (1962) but as stage 16 in V. aspis (Hubert and Dufauré, 1968). Based on our evaluation of the extra stages, it seems sufficient to recognize those of the garter snake as encompassing the major morphological changes during early development. We agree that the absolute number of stages is not the critical point but rather the features represented at each stage. The utility of a systematic and standardized staging guide is the ability to scientifically evaluate how far along a specimen has progressed in the continuum of its development. On a practical note, reptilian development is very susceptible to external temperature fluctuations, and staging is a more consistent way to describe development versus actual days incubation (Holtzman and Halpern, 1991). Our comparison of different snake studies shows that different investigators recognize similar numbers of time points at which new morphological features could be identified (a total of 37, the last 10 of which are represented in oviparous species). Taken together with data from other studies, there is sound external validation for our staging table. Missing data points in the present collection of P. sebae Even though we were not able to collect specimens between 3 and 11 days, we can make several predictions about snake development at stage 2 from our data. First, we can reasonably assume that stage 2 begins around 5 days po. Judging by the very limited ossification of the dentary, palatine and pterygoid bones in our stage 3 embryos, we would not expect to see these 225 bones with whole-mount skeletal staining in stage 2 specimens. Second, since our stage 4 specimens were collected at 18 days and our stage 6 embryos at 33 days, stage 5 would be about 25 days po. It is the day 25 time point when we predict that the ossification of certain bones will first become apparent. In a study on E. obsoleta (Haluska and Albrecht, 1983), two intermediate stages of 24 and 30 days incubation were collected, corresponding to late stage 4 and approximately mid stage 5 in our study. At 24 days the ectopterygoid had begun to ossify, whereas at 30 days the septomaxilla, exooccipital, compound and angular bones were ossifying. Thus it is very likely that in P. sebae the ectopterygoid and angular bones would be present at approximately 25 days po (stage 5). However, based on our data, the frontal bone and coronoid would not have been present any earlier than stage 6. We emphasized collection at earlier stages of development in this study and as a result we are lacking some details in older specimens, in particular stage 9 specimens. One possibility is that stages 9 and 10 specimens may be similar in terms of skeletal development and could be collapsed into a single stage. However, we reject this idea since there were enough key differences externally to distinguish a stage 9 from a stage 10 embryo. In summary, the data in Table 4 capture the initiation of all the bones and little will change once additional older specimens are collected. The presence or absence of intramembranous bones can be used as a tool to assess how far development has progressed. Rudimentary hindlimb development in P. sebae In our study, hindlimbs were present as small buds at oviposition, along which an opaque ridge could be seen in the youngest specimens (approximately 18 h after laying). The apical ectodermal ridge (AER) was slightly thickened in relation to the rest of the limb ectoderm (Raynaud, 1985) in sections of new-laid, reticulated python embryos. However, in another study, sections of python limbs 24 h po had no AER, nor were genes characteristic of the AER expressed in python limb buds (Cohn and Tickle, 1999). These differences could be due to slight variations in egg-laying times. At the very least, if a ridge does form along the limb bud ectoderm, then it is very short-lived in the python. A raised AER is required for normal limb outgrowth in birds. Experimental extirpation of the AER produced truncated wings. Depending on the stage of AER removal, either a stylopod (humerus or femur) alone, stylopod plus zeugopod (lower limb bones) or stylopod, zeugopod and truncated digits develop (Saunders, 1948; Summerbell, 1974). It is important to note that even if the AER is removed just after it has formed (H–H stage ARTICLE IN PRESS 226 Table 5. J.C. Boughner et al. / Zoology 110 (2007) 212–230 Craniofacial skeletal development in various snakes Gygax (1971) Franklin (1945) Haluska and Albrecht (1983) Jackson (2002) Richman lab Stage (dpo) Natrix tessellata Stage (dg) Natrix spp, Heterodon contortix, Diadophis punctatis p., Opheodrys vernalis Stage (dpo) Elaphe obsolete Stage (dpo) quadrivittata Naja kaouthia Stage (dpo) Python sebae 1 (1–2) No ossification pl, pt (1–23) No data 1 (2–6) No data 1 (1–3) No ossification (24) ec, pl, pm, pt 2 (6–9) No data 2 (4–10)* No data 3 (21) pl, pt, pm, 1–11 (1–26) ec, d mx, p, f, v, 12 (27) sm n, pf, pof, ps 13 (28) su 3 (9–15) ar, d, ec,pl, pq, pm, pt 3 (11,12) 4–8 (26–37) No new bones 14 (29) a, d, eo, pm, (30) v a, ar, com, eo, mx, sm 4 (15–20) No data 4 (18) 15 (33) n, pf, pr (36) 5 (20–23) No data 5 (20–30)* 16 (34) mx (41) bo, f, n, p, pf, sp, st, v bs, d, q com (prearticular process), d, pl, pt, st com (surangular process), mx, n, p, pf, pm, sp, v No data 6 (22–25) No data 6 (33) 17 (35) 18 (36) pof, q ar, co (48) (49) po pr 7 (24–28) 8 (28–38) 7 (44) 8 (54) 19 (41) bo, bs, ps, soc NA (52) ps 9 (38–51) mx,st bo, bs, com, eo, f, pl, p, pof, pf, pr, q, sp, soc, st No data a, co, ec, f, sm, po, bo, eo, soc, pr bs, ps, q so, h 9 (60–70)* No data (59) s, soc 10 (51–hatching) o, ps 10 (75) No new bones 2 (13) NA (25–29) No new ossification Comparison of the embryonic stages and/or days at or by which the ossification of the endochondral and dermal craniofacial and lower jaw bones is first observed in various snake species for which data are published. Not all bones can be represented for each snake taxon due to lack of data. In some cases, we were able to examine the figures and find that some bones were misidentified. A closer examination of Fig. 15 in Jackson (2002) reveals that the maxillary bone is actually the palatine and this is indicated in the table here. Abbr.: a – angular; ar – articular process; bo – basioccipital; bs – basisphenoid; co – coronoid process; com – compound; d – dentary; dg – days gestation for viviparous species; dpo – days postoviposition for oviparous species; ec – ectopterygoid; eo – exoccipital; f – frontal; h – hyoid; mx – maxillary; n – nasal; o – occipital; p – parietal; pl – palatine; pf – prefrontal; pm – premaxilla; po – postorbital; pof – postfrontal; pq – palatoquadrate; pr – prootic; ps – parasphenoid; pt – pterygoid; q – quadrate; s – stapes; sm – septomaxilla; so – supraorbital; soc – supraoccipital; sp – splenial; st – supratemporal; su – surangular process; v – vomer; * – predicted days. 18; Jurand, 1965), the stylopod still forms. This AER stripping data together with more recent fate maps of the scapula (equivalent to the pelvic girdle) in chicken (Huang et al., 2000; Wang et al., 2005) indicate that in the python most of the skeletal elements form independently of signals from the AER. In the past, snakes and their predecessors did have limbs (Coates and Ruta, 2000; Tchernov et al., 2000; Apesteguia and Zaher, 2006). Some of the more recently discovered snake fossils even show that hindlimbs with digits formed (Apesteguia and Zaher, 2006). These fossil data predict that primitive snakes had a robust AER, similar to modern amniotes with digits. Species-specific temporal sequence of craniofacial bone development in snakes We compared some of the best-available data on initiation times of neurocranial and facial bones among many different snakes (Table 5) and found that the timing of onset of ossification varied between species. Some of these differences may be due to the sensitivity of various detection methods; whole-mount stains were used in some papers (Franklin, 1945; Jackson, 2002), whereas in the most detailed studies a combination of whole-mount and histological analysis was employed (Haluska and Albrecht, 1983). We also ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 Table 6. 227 Comparison of craniofacial development amongst amniotes Python (Python sebae) Chicken (Gallus gallus) Mouse (Mus musculus var. alba) Stage (days postoviposition) Source: Richman lab H–H stage (days) Source: Erdmann (1940) Gestational day+hours 1 (1–3) No ossification 13–29 No ossification 15 2 No data 30–31 (7) Nasal, pterygoid, quadratojugal, surangular 15 to 15+11 3 (11–12) Compound (prearticular process), dentary, palatine, pterygoid, supratemporal 34 (8) 16 4 (18) Compound (surangular process), maxillary, nasal, parietal, prefrontal, premaxilla, splenial, vomer No data Angular, coronoid, ectopterygoid, frontal, septomaxilla, postorbital, basioccipital, exooccipital, supraoccipital, prootic Basisphenoid, parasphenoid, quadrate Supraorbital, hyoid 35 (9) Angular, basisphenoid, dentary, frontal, jugal, maxillary, palatine, premaxilla, quadrate, splenial Exoccipital Sources: 1 – Johnson (1933) 2 – Wirtschafter (1960) 3 – Evans and Sack (1973) Basioccipital, dentary, exoccipital, frontal, parietal, vomer1 Maxillary, palatine, premaxilla, pterygoid processes of basisphenoid1,2 Nasal, zygoma2 17+1 Nasal1 36 (10) 37 (11) Parietal Basioccipital, supraoccipital, vomer 18 Neonate (19 days) No new bones3 40 (14) Articular Hatchling (20–21) No new bones 5 6 (33) 7 (44) 8 (54) 9 (61) 10 (75) Hatchling (80–90) No data No new bones No data The embryonic stages at which ossification of the dermal skull bones is first observed in python, chicken and mouse. Not all dermal bones are represented for each taxon. present histological data in our accompanying paper (Buchtová et al., 2007) and identify in Table 4 which bones can only be seen with this method. We have taken into account this additional information from sections when preparing Table 5. Therefore, the cobra has a similar sequence of ossification to the python, with the maxillary, palatine and pterygoid bones forming before other membranous bones (Kamal et al., 1970). The exception in the sequence of upper jaw bone ossification is the ectopterygoid, which forms much later in python than in cobra. Some other species with shorter prehatching periods such as Natrix tessellata (the venomous and oviparous dice snake, 2 month period), have a more compressed time frame of bone ossification. Shortly after laying (1–2 days), the premaxilla, palatine, pterygoid, ectopterygoid and dentary bones can already be detected in tissue sections (Gygax, 1971). However, even when we account for this more advanced development, the dice snake has a couple of differences in ossification sequence, with the premaxilla and ectopterygoid forming earlier than in our python sample. After accounting for missing time points and allowing for lack of sensitivity in certain methods (i.e. histology is more sensitive than whole-mount staining, no cartilage is visible in CT scans), we conclude that in snakes the most conserved features are that the palatine and pterygoid are always among the first intramembranous bones of the face to ossify. Other bones such as the premaxilla and ectopterygoid are more variable among taxa and may be more specific to the family of snake being studied. ARTICLE IN PRESS 228 J.C. Boughner et al. / Zoology 110 (2007) 212–230 CT scanning is of a sufficiently high resolution to reveal the complex architecture of very immature, small bones. Thus, we were able to compare the initial morphology of the early ossification center to the final bone form. From observations on the septomaxillary, maxillary, premaxillary and vomer bones, a template of adult bone morphology is visibly established at the outset of prenatal development. Thus, there are intrinsic patterning cues that operate early in organogenesis. No doubt post-natal muscle activity builds bone thickness in certain areas (Figs. 1A–D), but the overall bone shape is preserved from its early organogenesis in Python. Comparison of python craniofacial skeletal development to that of other amniotes The sequence of ossification and morphology of the python craniofacial complex is more similar to birds than it is to mammals. In avian reptiles, the pterygoid is among the first bones to ossify along with the quadratojugal and surangular (Erdmann, 1940). Furthermore, the pterygoid of the bird is quite similar to the snake pterygoid in that it is a separate bone, abutting the palatine bone rostrally and the quadrate laterally (Richman et al., 2006). In contrast, in mammals the pterygoid plates are thought to be processes of the sphenoid bone (Presley and Steel, 1978). This greater similarity between birds and snakes is to be expected, as birds are a recent radiation of reptiles (Pough et al., 2004; Zhou, 2004). What is interesting is that python and chicken facial bone morphologies remain so similar despite the evolution of the specialized bird beak. Since the pterygoid is the first skull bone to form in birds (Erdmann, 1940) and snakes, we wondered how widespread this characteristic is amongst reptilian taxa. The osteology of two turtle species has been studied in detail and, in both cases, the pterygoid begins to ossify after the maxilla and dentary bones (Sheil, 2003, 2005). Used as an outgroup, these turtle data suggest that the early ossification of the pterygoid is a conserved feature limited to snakes, birds and possibly other squamates and not necessarily to other reptilian classes. We found large differences in the sequence and timing of onset of ossification in mammals as compared to reptiles (Table 6). In mouse, the fetal period is relatively short and ossification occurs between days 15 and 17 of gestation (Johnson, 1933; Wirtschafter, 1960; Evans and Sack, 1973). Thus it is hard to determine the sequence of ossification to as fine a level as in slower-developing reptiles. Perhaps the use of other tools such as molecular markers for early commitment to an osteogenic lineage is a better way to study the sequence of ossification in mouse (Ducy et al., 1997; Otto et al., 1997; Stricker et al., 2002). Preparing the groundwork for future studies on snakes Snakes are relatively rarely studied at embryonic stages and there is no standard accepted species for developmental studies as there is for birds (chicken, quail) or mammals (mouse). Comparisons between species are difficult if there is no standardized set of observations taken at different times during development. For example, by using chicken staging (Hamburger and Hamilton, 1951) it was possible to compare finch or duck to chicken even though incubation periods differ (Schneider and Helms, 2003; Abzhanov et al., 2004; Wu et al., 2004). The staging table and skeletal data presented here will facilitate comparisons between embryos of different snake species in future studies. The craniofacial region is one of the most important areas for determining phylogenetic relationships between animals. Thus we should include non-avian reptiles, a branch of the evolutionary tree that has largely been overlooked. While we must be cautious when extrapolating studies of adult, highly derived, living snakes to reptilian and vertebrate evolution, the embryos are where true homologies can be identified. Thus, this paper on python embryo development will form an important foundation for further studies on reptilian, amniote and vertebrate evolution. Acknowledgments We are grateful to Paul Springate at The Rainforest Reptile Refuge, Surrey, British Columbia, whose expert reptile husbandry made this research possible. We thank Jeff Dunn and the Medical Imaging technical staff at the University of Calgary, Faculty of Medicine, for their 3D imaging expertise and assistance. In particular, Wei Liu (University of Calgary) kindly provided technical assistance with computed micro-tomography. This work was first funded by CIHR and later by NSERC Grants to J.M.R. J.M.R. is a Michael Smith Distinguished Scholar. References Abzhanov, A., Protas, M., Grant, B.R., Grant, P.R., Tabin, C.J., 2004. Bmp4 and morphological variation of beaks in Darwin’s finches. Science 305, 1462–1465. Apesteguia, S., Zaher, H., 2006. A Cretaceous terrestrial snake with robust hindlimbs and a sacrum. Nature 440, 1037–1040. Branch, W.R., Patterson, R.W., 1975. Notes on the development of embryos of the African Rock Python, Python sebae (Serpentes: Boidea). J. Herpetol. 9, 243–248. ARTICLE IN PRESS J.C. Boughner et al. / Zoology 110 (2007) 212–230 Buchtová, M., Boughner, J.C., Fu, K., Diewert, V.M., Richman, J.M., 2007. Embryonic development of Python sebae. II. Craniofacial microscopic anatomy, cell proliferation and apoptosis. Zoology 110, doi:10.1016/ j.zool.2007.01.006. Carroll, R.L., 1988. Vertebrate Paleontology and Evolution. W.H. Freeman and Company, New York. Coates, M., Ruta, M., 2000. Nice snake, shame about the legs. Trends Ecol. Evol. 15, 503–507. Cohn, M.J., Tickle, C., 1999. Developmental basis of limblessness and axial patterning in snakes. Nature 399, 474–479. De Beer, G.R., 1937. The Development of the Vertebrate Skull. Oxford University Press, Oxford (Reprinted in 1985 by The University of Chicago Press, Chicago). Ducy, P., Zhang, R., Geoffroy, V., Ridall, A.L., Karsenty, G., 1997. Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 89, 747–754. Erdmann, K., 1940. Zur Entwicklungsgeschichte der Knochen im Schädel des Huhnes bis zum Zeitpunkt des Ausschlüpfens aus dem Ei. Z. Morphol. Ökol. Tiere 36, 315–400. Evans, H.E., Sack, W.O., 1973. Prenatal development of domestic and laboratory mammals: growth curves, external features and selected references. Anat., Histol., Embryol. 2, 11–45. Ferguson, M.W., 1981a. Review: the value of the American alligator (Alligator mississippiensis) as a model for research in craniofacial development. J. Craniofac. Genet. Dev. Biol. 1, 123–144. Ferguson, M.W., 1981b. The structure and development of the palate in Alligator mississippiensis. Arch. Oral Biol. 26, 427–443. Franklin, M.A., 1945. The embryonic appearance of centres of ossification in the bones of snakes. Copeia 2, 68–72. Frazzetta, T.H., 1959. Studies on the morphology of the skull in the Boidae (Serpentes). I. Cranial differences between Python sebae and Epicrates cenchris. Bull. Mus. Comp. Zool. 119, 451–472. Frazzetta, T.H., 1966. Studies on the morphology and function of the skull in the Boidae (Serpentes). II. Morphology and function of the jaw apparatus in Python sebae and Python molurus. J. Morphol. 118, 217–295. Gygax, P., 1971. Development, morphology and function of the venom gland (Duvernoy’s gland) of Natrix tessellata. Acta Trop. 28, 226–274. Haluska, F., Albrecht, P., 1983. The cranial development of Elaphe obsoleta (Ophidia, Colubridae). J. Morphol. 178, 37–55. Hamburger, V., Hamilton, H., 1951. A series of normal stages in the development of the chick embryo. J. Morphol. 88, 49–92. Holtzman, D.A., Halpern, M., 1991. Incorporation of 3Hthymidine in the embryonic vomeronasal and olfactory epithelial of garter snakes. J. Comp. Neurol. 304, 435–449. Huang, R., Zhi, Q., Patel, K., Wilting, J., Christ, B., 2000. Dual origin and segmental organisation of the avian scapula. Development 127, 3789–3794. Hubert, J., 1985. Embryology of the Squamata. Wiley, Toronto. 229 Hubert, J., Dufauré, J.P., 1968. Table de development de la vipere aspic, Vipera aspis. Bull. Soc. Zool. France 93, 135–148. Jackson, K., 2002. Post-ovipositional development of the monocled cobra, Naja kaouthia (Serpentes: Elapidae). Zoology 105, 203–214. Johnson, M.L., 1933. The time and order of appearance of ossification centres in the albino mouse. Am. J. Anat. 52, 241–271. Jurand, A., 1965. Ultrastructural aspects of early development of fore-limb buds in chick and mouse. Proc. R. Soc. Lond. Ser. B—Biol. Sci. 162, 387–405. Kamal, A.M., Hammouda, H.G., 1965a. The development of the skull of Psammophis sibilans. I. The development of the chondrochranium. J. Morphol. 116, 197–245. Kamal, A.M., Hammouda, H.G., 1965b. The development of the skull of Psammophis sibilans. II. The fully formed chondrocranium. J. Morphol. 116, 247–296. Kamal, A.M., Hammouda, H.G., 1965c. The development of the skull of Psammophis sibilans. III. The osteocranium of a late embryo. J. Morphol. 116, 297–310. Kamal, A.M., Hammouda, H.G., Mokhtar, F.M., 1970. The development of the osteocraniium of the Egyptian Cobra: I. The embryonic osteocranium. Acta Zool. 1970, 1–17. Kuraku, S., Usuda, R., Kuratani, S., 2005. Comprehensive survey of carapacial ridge-specific genes in turtle implies cooption of some regulatory genes in carapace evolution. Evol. Dev. 7, 3–17. Lemus, D., Paz de la Vega, Y., Fuenzalida, M., Illanes, J., 1980. In vitro differentiation of tooth buds from embryos and adult lizards (L. gravenhorsti): an ultrastructural comparison. J. Morphol. 165, 225–236. Lemus, D., Coloma, L., Fuenzalida, M., Illanes, J., Paz de la Vega, Y., Ondarza, A., Blanquez, M.J., 1986. Odontogenesis and amelogenesis in interacting lizard–quail tissue combinations. J. Morphol. 189, 121–129. Matsuda, Y., Nishida-Umehara, C., Tarui, H., Kuroiwa, A., Yamada, K., Isobe, T., Ando, J., Fujiwara, A., Hirao, Y., Nishimura, O., Ishijima, J., Hayashi, A., Saito, T., Murakami, T., Murakami, Y., Kuratani, S., Agata, K., 2005. Highly conserved linkage homology between birds and turtles: bird and turtle chromosomes are precise counterparts of each other. Chromosome Res. 13, 601–615. Nagashima, H., Uchida, K., Yamamoto, K., Kuraku, S., Usuda, R., Kuratani, S., 2005. Turtle–chicken chimera: an experimental approach to understanding evolutionary innovation in the turtle. Dev. Dyn. 232, 149–161. Ohya, Y.K., Kuraku, S., Kuratani, S., 2005. Hox code in embryos of Chinese soft-shelled turtle Pelodiscus sinensis correlates with the evolutionary innovation in the turtle. J. Exp. Zool. Part B—Mol. Dev. Evol. 304B, 107–118. Otto, F., Thornell, A.P., Crompton, T., Denzel, A., Gilmour, K.C., Rosewell, I.R., Stamp, G.W., Beddington, R.S., Mundlos, S., Olsen, B.R., Selby, P.B., Owen, M.J., 1997. Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89, 765–771. Parker, W.K., 1879. On the structure and development of the skull in the common snake, Tropidonotus natrix. Philos. Trans. R. Soc. Lond. (Biol.) 169, 385–417. ARTICLE IN PRESS 230 J.C. Boughner et al. / Zoology 110 (2007) 212–230 Plant, M.R., MacDonald, M.E., Grad, L.I., Ritchie, S.J., Richman, J.M., 2000. Locally released retinoic acid repatterns the first branchial arch cartilages in vivo. Dev. Biol. 222, 12–26. Pough, F.H., Andrews, R.M., Cadle, R.M., Crump, M.L., Savitzky, A.H., Wells, K.D., 2004. Herpetology. Pearson Prentice Hall, Upper Saddle River. Presley, R., Steel, F.L., 1978. The pterygoid and ectopterygoid in mammals. Anat. Embryol. 154, 95–110. Raynaud, A., 1985. Development of limbs and embryonic limb reduction. In: Gans, C. (Ed.), Biology of the Reptilia, Development B. Wiley, Toronto, pp. 117–148. Richman, J.M., Buchtová, M., Boughner, J.C., 2006. Comparative ontogeny and phylogeny of the upper jaw skeleton in amniotes. Dev. Dyn. 235, 1230–1243. Saunders, J.W., 1948. The proximo-distal sequence of origin of the parts of the chicken wing and the role of the ectoderm. J. Exp. Zool. 108, 363–403. Schneider, R.A., Helms, J.A., 2003. The cellular and molecular origins of beak morphology. Science 299, 565–568. Sheil, C.A., 2003. Osteology and skeletal development of Apalone spinifera (Reptilia: Testudines: Trionychidae). J. Morphol. 256, 42–78. Sheil, C.A., 2005. Skeletal development of Macrochelys temminckii (Reptilia: Testudines: Chelydridae). J. Morphol. 263, 71–106. Stricker, S., Fundele, R., Vortkamp, A., Mundlos, S., 2002. Role of Runx genes in chondrocyte differentiation. Dev. Biol. 245, 95–108. Summerbell, D., 1974. A quantitative analysis of the effect of excision of the AER from the chick limb-bud. J. Embryol. Exp. Morphol. 32, 651–660. Tchernov, E., Rieppel, O., Zaher, H., Polcyn, M.J., Jacobs, L.L., 2000. A fossil snake with limbs. Science 287, 2010–2012. Theiler, K., 1972. The House Mouse: Development and Normal Stages from Fertilization to 4 Weeks of Age. Springer, New York. Torres-Maldonado, L., Moreno-Mendoza, N., Landa, A., Merchant-Larios, H., 2001. Timing of SOX9 downregulation and female sex determination in gonads of the sea turtle Lepidochelys olivacea. J. Exp. Zool. 290, 498–503. Torres Maldonado, L.C., Landa Piedra, A., Moreno Mendoza, N., Marmolejo Valencia, A., Meza Martinez, A., Merchant Larios, H., 2002. Expression profiles of Dax1, Dmrt1, and Sox9 during temperature sex determination in gonads of the sea turtle Lepidochelys olivacea. Gen. Comp. Endocrinol. 129, 20–26. Wang, B., He, L., Ehehalt, F., Geetha-Loganathan, P., Nimmagadda, S., Christ, B., Scaal, M., Huang, R., 2005. The formation of the avian scapula blade takes place in the hypaxial domain of the somites and requires somatopleurederived BMP signals. Dev. Biol. 287, 11–18. Westergaard, B., Ferguson, M.W., 1987. Development of dentition in Alligator mississippiensis. Later development in the lower jaws of embryos, hatchlings and young juveniles. J. Zool. 212, 191–222. Wibbels, T., Cowan, J., LeBoeuf, R., 1998. Temperaturedependent sex determination in the red-eared slider turtle, Trachemys scripta. J. Exp. Zool. 281, 409–416. Wirtschafter, Z.T., 1960. The Genesis of the Mouse Skeleton: A Laboratory Atlas. Charles C Thomas, Springfield, IL. Wu, P., Jiang, T.X., Suksaweang, S., Widelitz, R.B., Chuong, C.M., 2004. Molecular shaping of the beak. Science 305, 1465–1466. Yntema, C.L., 1968. A series of stages in the embryonic development of Chelydra serpentina. J. Morphol. 125, 219–251. Zehr, D.R., 1962. Stages in the normal development of the common garter snake, Thamnophis sirtalis sirtalis. Copeia 1962 (2), 322–329. Zhou, Z., 2004. The origin and early evolution of birds: discoveries, disputes, and perspectives from fossil evidence. Naturwissenschaften 91, 455–471. DEVELOPMENTAL DYNAMICS 233:721–738, 2005 RESEARCH ARTICLE Embryonic Staging System for the Short-Tailed Fruit Bat, Carollia perspicillata, a Model Organism for the Mammalian Order Chiroptera, Based Upon Timed Pregnancies in Captive-Bred Animals Chris J. Cretekos,1 Scott D. Weatherbee,2 Chih-Hsin Chen,1 Nilima K. Badwaik,3 Lee Niswander,2† Richard R. Behringer,1* and John J. Rasweiler, IV3,4 There are approximately 4,800 extant species of mammals that exhibit tremendous morphological, physiological, and developmental diversity. Yet embryonic development has been studied in only a few mammalian species. Among mammals, bats are second only to rodents with regard to species number and habitat range and are the most abundant mammals in undisturbed tropical regions. Bat development, though, remains relatively unstudied. Here, we describe and illustrate a staging series of embryonic development for the short-tailed fruit bat, Carollia perspicillata, based on embryos collected at timed intervals after captive matings. As Carollia can be readily maintained and propagated in captivity and is extremely abundant in the wild, it offers an attractive choice as a chiropteran model organism. This staging system provides a framework for studying Carollia embryogenesis and should prove useful as a guide for embryological studies of other bat species and for comparisons with other orders of mammals. Developmental Dynamics 233:721–738, 2005. © 2005 Wiley-Liss, Inc. Key words: mammal; Eutheria; Chiroptera; Phyllostomidae; embryogenesis; limb development; embryo staging; model organism Received 3 November 2004; Revised 4 January 2005; Accepted 4 January 2005 INTRODUCTION There are approximately 4,800 species of mammals currently living on Earth (Nowak, 1999). These mammals are divided into three subclasses: Prototheria (or monotremes), Metatheria (or marsupials), and Eutheria. These three subclasses are further divided into 26 or more orders. Unlike all other animals, mammals nourish their young with milk, possess body hair, and have three middle ear bones. As a class, mammals display enor- mous diversity in form and function. From the tiny Kitti’s hog-nosed bat (Craseonycteris thonglongyai) that weighs between 1.5 and 2 grams to giant 150 ton Blue whales (Balaenoptera musculus), mammals walk, run, jump, swim, glide, or fly through The Supplementary Material referred to in this article can be found at http://www.interscience.wiley.com/jpages/1058-8388/suppmat 1 Department of Molecular Genetics, University of Texas M.D. Anderson Cancer Center, Houston, Texas 2 Howard Hughes Medical Institute, Developmental Biology Program, Memorial Sloan-Kettering Institute, New York, New York 3 Department of Obstetrics and Gynecology, Weill Medical College of Cornell University, New York, New York 4 Department of Obstetrics and Gynecology, State University of New York Downstate Medical Center, Brooklyn, New York Grant sponsor: National Institutes of Health; Grant numbers: CA09299; HD07325; HD08720; HD28592; HD28592; Grant sponsor: National Science Foundation; Grant number: IBN 0220458; Grant sponsor: Barnts Family; Grant sponsor: Department of Obstetrics and Gynecology, Weill Medical College of Cornell University; Grant sponsor: MSKCC Cancer Center. † Dr. Niswander’s present address is Howard Hughes Medical Institute, University of Colorado Health Sciences Center, Department of Pediatrics, Section of Developmental Biology, Aurora, CO 80045. *Correspondence to: Richard R. Behringer, Department of Molecular Genetics, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030. E-mail: rrb@mdanderson.org DOI 10.1002/dvdy.20400 Published online 28 April 2005 in Wiley InterScience (www.interscience.wiley.com). © 2005 Wiley-Liss, Inc. 722 CRETEKOS ET AL. nearly every terrestrial, aquatic, or aerial habitat on the planet (Nowak, 1999). However, most of our knowledge of development comes from studying a single group of mammals: the rodents (order Rodentia). Short generation time, large numbers of progeny, modest husbandry requirements, and a long history of genetics have all contributed to several members of Rodentia (predominantly mice and rats) becoming the major models for laboratory study of mammalian biology. Consequently, rodent embryology has been studied in great detail, whereas comparatively little is known about the other 25⫹ orders of mammals. Although knowledge gained from the study of mouse and rat development is of immense value, and much of this knowledge is likely to be applicable to mammals in general, it should be noted that the rate and fecundity of rodent propagation is accomplished through a set of highly specialized and species-specific reproductive and developmental adaptations. Thus, the lessons learned using rodents to model mammalian reproduction and development may be misleading due to the very characteristics that have facilitated and promoted rodent models for laboratory research. This suggests that a complete picture of mammalian development cannot be obtained by studying one or two species from within a single order (Eakin and Behringer, 2004). Morphological, physiological, and molecular comparisons between diverse species beyond the standard rodent models will be necessary to truly understand mammalian embryonic development. As the second largest order of mammals in terms of the number of recognized species (more than 1,000; Simmons, 2001), bats are also one of the most successful with respect to geographic range and biological diversity. Bats are extremely important from agricultural and ecological perspectives for control of pest insects and as pollinators and seed dispersers. Bats are of interest to epidemiologists as known and suspected carriers of a variety of pathogens. Finally, the bat’s unique abilities of powered flight and echolocation are fascinating from biomechanical, auditory, and neuroscience perspectives. Bats are all members of the order Chiroptera. According to Walker’s Mammals of the World (Nowak, 1999), the order Chiroptera is divided into 2 superfamilies (Mega- and Microchiroptera), 18 families, and 186 genera. The characteristic feature of bats is the presence of wings, making them the only mammals capable of powered flight. Several other groups of mammals such as flying squirrels and flying lemurs do not in fact fly but rather glide. Bats’ wings are membranous and supported by the skeletal elements of the limbs and tail. The first digit of the forelimb is short, usually possesses a claw, and is not enclosed within the wing, whereas the other four digits are significantly elongated relative to the thumb, are usually clawless, and support a large portion of the wing membrane. The third (and longest) digit of the forelimb is generally approximately equal in length to the height of the animal from head to foot. Bats display numerous additional adaptations for flight, including light and slender long bones, a robust pectoral girdle and ribcage to support the wings, and a prominent keel extending from the ventral midline of the sternum for the attachment of enlarged pectoral flight muscles. Most bats are nocturnal, and aerial navigation in darkness is guided by means of a phenomenon known as echolocation, another unique feature of bats (Neuweiler, 2000). These bats emit vocal sounds through the mouth and/or nose as they fly. The sounds are generally at frequencies above the limit of human hearing and are reflected back to the bat in flight. Neural processing of these reflected echoes enable the bat to avoid obstacles and to locate food in darkness. Echolocation has been found in all bat families investigated thus far, but not all species of bats echolocate (Nowak, 1999). The short-tailed fruit bat, Carollia perspicillata (commonly known as Seba’s short-tailed bat or simply Carollia), would appear to be the bat species of choice for molecular developmental and embryological studies. This bat is a predominantly fruit-eating microchiropteran, which also takes some insects and floral parts, belonging to the American leaf-nosed bats (family Phyllostomidae). This is one of the largest and most successful groups of bats. Carollia occurs from northern Argentina to southern Mexico and is probably the most abundant mammal inhabiting the humid lowland tropics of the New World (McLellan, 1984; Fleming, 1988; Eisenberg, 1989; Rasweiler, unpublished observations). First described in 1758 by Linnaeus in his classic Systema naturae, Carollia’s behavioral ecology and reproductive physiology have been well described since (Kleiman and Davis, 1979; Fleming, 1988; Cosson and Pascal, 1994; Badwaik and Rasweiler, 2000; de Mello and Fernandez, 2000; Rasweiler and Badwaik, 1999). Carollia adults weigh approximately 19 grams and have a wingspan of approximately 21–25 cm. Carollia in the wild usually breeds twice per year and gives birth to a single offspring that can weigh up to 32% of its mother’s postpartum mass (Fleming, 1988; Cloutier and Thomas, 1992; Rasweiler and Badwaik, 1999). Gestation is normally 113 to 120 days (Rasweiler and Badwaik, 1997; Rasweiler, unpublished observations) and weaning occurs between 6 and 8 weeks postpartum (Fleming, 1988). Juveniles reach adult mass at 10 –13 weeks, and both males and females achieve sexual maturity in the wild between 1 and 2 years of age (Fleming, 1988). Simple, efficient, and economical methods are now available for maintaining, breeding, and propagating Carollia in the laboratory (Rasweiler and Badwaik, 1996). Furthermore, it is routine to safely capture and handle animals on a daily basis, for example to check for breeding activity. In captivity, the females are polyestrous and can be successfully bred at any time of the year. Finally, large numbers of young can be reared and become fertile. A staging series is a fundamental tool for developmental studies in any species (McCrady, 1938; Streeter, 1942; Nieuwkoop and Faber, 1967; Hamburger and Hamilton, 1951; Hendrickx, 1971; Eyal-Giladi and Kuchar, 1976; Theiler, 1989; Mate et al., 1994; Kimmel et al., 1995; Selwood and Hickford, 1999; Iwamatsu, 2004). This is in part because, even among genetically homogeneous populations of animals such as inbred strains of mice or clonal strains of fish, there is some variability in the rate of development between individual progeny (Streisinger et al., 1981; Theiler, 1989; BAT EMBRYO STAGING SYSTEM 723 Downs and Davies, 1993; Kimmel et al., 1995). Such a tool is particularly important for species displaying characteristics such as significant variability in normal gestation length, developmental delays, or where specimens are often harvested from the wild and the day of fertilization is unknown. Staging by morphological criteria relative to a standard series minimizes the effects of developmental variability and delay, obviates the requirement for knowing when conception occurred, and facilitates comparison between independent studies. Moreover, a morphology-based staging system allows comparison to different bat species and with other mammals. A series of timed laboratory matings were arranged to generate normally developing Carollia embryos of known gestational age. These specimens were then used to create a standardized staging system based on the classic Carnegie system for human development (Streeter, 1942; O’Rahilly and Müller, 1987). Here, we illustrate this staging system and describe a set of criteria for staging embryos on the basis of the morphology of anatomical features that are easily distinguished either in freshly dissected or fixed specimens under the dissecting microscope. This staging system provides the foundation for further embryological studies in Carollia and other bat species. RESULTS Variation in Gestation Length and Rate of Development Specimens carried by females born, reared, and mated in captivity anchor the staging series presented here. Females in this group have gestation periods of 113–120 days, and this represents the normal (nondelayed) gestation period for this species (Rasweiler and Badwaik, 1997; Rasweiler, unpublished observations). To stage normal development beyond the primitive streak stage, embryos are examined that fall close to the lines of best fit for changes in uterine diameter and embryo size as gestation progresses in captive-reared and -bred females (Supplementary Figure S1, which can be viewed at http://www.interscience. wiley.com/jpages/1058-8388/suppmat; Badwaik and Rasweiler, 2001). It is important to note that some variability in embryo size and/or developmental state was observed among the specimens collected at all time points. There exists up to 24 hr of variation in the time of ovulation, and up to 2 days of variation in oviduct transit time, relative to the onset of breeding, between individual pregnancies (de Bonilla and Rasweiler, 1974; Olivera et al., 2000; Rasweiler and Badwaik, unpublished observations). This finding presumably accounts for some of the observed variability in normal gestation length (7 days) and in developmental state on any particular day post coitum (dpc). Normal differences in the rate of development between individuals may also contribute to this variability. Additional variability can be introduced by the occurrence of developmental delay at the primitive streak stage in captive-bred animals. Such delays are more common and sometimes of substantial length in wildcaught, captive-bred females. They are much less frequent and tend to be of shorter duration in captive-reared and -bred females. In a very few cases, the delays in captive animals have exceeded 100 days. All of the available evidence indicates that the delays occurring in captive-bred animals are induced by stress (Rasweiler and Badwaik, 1997, unpublished observations). Delays also occur under natural conditions in the wild. On the West Indian island of Trinidad, most adult female Carollia are reproductively synchronized and carry two pregnancies in succession. The first, initiated late in the rainy season, includes a period of developmental delay estimated to be at least 44 –50 days. The timing of this delay is apparently controlled by some unknown seasonal factor. The second, conceived in most parous females at a postpartum estrus during the dry season, usually progresses without significant delay (Rasweiler and Badwaik, 1997; Badwaik and Rasweiler, 2001; Rasweiler et al., unpublished observations). Staging The majority of specimens shown below were collected from timed mating of captive bred and reared females for which the day of initial insemination was determined. Between two and six specimens (combined n ⫽ 29), which fell near the lines of best fit for changes in uterine diameter and embryo size with time (Badwaik and Rasweiler, 2001) were examined for stages 12, 14, 15, 16, 17, 18, 20, 22, and 24. Table 1 summarizes the measurements and key features for all 29 specimens; the Supplementary Figure S1 illustrates the rate of growth in uterine diameter and embryo mass. The individual pictured (the reference standard) is the most developmentally advanced specimen examined for each time point. In two cases, stages 16 and 18, a reference standard specimen, having become fragile after fixation and prolonged storage, was damaged in handling. In these cases, we replaced the damaged specimen with a morphologically stage-matched specimen collected from the wild. The individuals pictured for stage 10, 11, and 13 were collected from the wild and ordered in this series according to the number of somite pairs present. The individual pictured for Stage 19 was collected from the wild and staged by comparison with the preceding and subsequent stages. The actual time of gestation for these specimens is unknown. Stages 1–9: Fertilization to Open Neural Plate (Not Shown) The Carollia equivalents of Carnegie Stage 1 (fertilization), Stage 2 (twocell to morula), Stage 3 (free blastocyst), Stage 4 (attaching blastocyst), and Stage 5 (implantation) previously have been described thoroughly (de Bonilla and Rasweiler, 1974; Badwaik et al., 1997; Oliveira et al., 2000; Rasweiler et al., 2002). Useful descriptions of the Carollia equivalents of Carnegie Stage 6 (primitive streak), Stage 7 (notochordal process), Stage 8 (primitive pit), and Stage 9 (open neural plate) will require histological and/or ultrastructural analyses beyond the scope of this study. During 724 CRETEKOS ET AL. TABLE 1. Summary of Bat Embryo Reference Specimens* Stage 12 14 15 16 17 18 20 22 24 Age (dpc) Key features Forelimb buds form; tail bud forms; caudal neuropore closes; 3 pharyngeal arches; 21–29 somite pairs. Retinal pigment; nasal pits; end of somitogenesis; propatagium and plagiopatagium primordia; hindlimb AER; 36–40 somite pairs. Hand plate and footplate form; lens vesicle; auditory hillocks; premaxillary centers. Nose-leaf primordium; pinna and tragus form; forelimb digital condensations, uropatagium primordium. Tongue protruding; cervical flexure straightens; hindlimb interdigit tissue receding; eyes begin to close. Free thumb; head and body smoother, rounder; eyes half-closed; postaxial flexure at wrist; calcar. Distal forelimbs overlap over face; head larger; eyelids cover pigmented retina; claw primordia form. Prominent, triangular nose-leaf; eyelids reopening; wing membranes corrugated; claws pigmented, hooked. Fetal period commences; eyes completely open; face and noseleaf pigmenting. 40 44 46 50 54 60 70 80 90 Uterus diameter (mm) Crown–rump length (mm) Mass (mg) 5.75 (⫹/⫺ 0.64) 3.4 (⫹/⫺ 0.42) 4.3 (⫹/⫺ 1.7) 6.2 3.7 5.5 5.3–6.2 3.1–3.7 3.1–5.5 6.95 (⫹/⫺ 0.44) 5.35 (⫹/⫺ 0.24) 24.6 (⫹/⫺ 3.6) 6.9 5.5 27.1 6.9–7.0 5.2–5.5 22.0–27.1 8.65 (⫹/⫺ 1.20) 9.5 7.8–9.5 12.06 (⫹/⫺ 1.45) 7.45 (⫹/⫺ 0.92) 8.1 6.8–8.1 8.66 (⫹/⫺ 1.05) 56 (⫹/⫺ 13) 65 47–65 110 (⫹/⫺ 30) 12.5 9.2 124 10.3–14.0 7.3–10.1 60–130 13.45 (⫹/⫺ 1.34) 9.15 (⫹/⫺ 1.34) 114 (⫹/⫺ 45) 14.4 10.1 146 2.5–14.4 8.2–10.1 82–146 16.32 (⫹/⫺ 0.98) 12.35 (⫹/⫺ 1.16) 278 (⫹/⫺ 83) 17.2 13 345 15.0–17.2 10.9–13.9 185–358 20.0 (⫹/⫺ 3.54) 16.35 (⫹/⫺ 1.06) 617 (⫹/⫺ 156) 22.5 17.1 727 17.5–22.5 15.6–17.1 507–727 23.03 (⫹/⫺ 2.68) 20.02 (⫹/⫺ 0.26) 1527 (⫹/⫺ 208) 24 20.3 1457 20.0–25.1 19.9–20.4 1363–1760 23.53 (⫹/⫺ 0.64) 24 22.8–24.0 21.13 (⫹/⫺ 0.06) 21.2 21.1–21.2 2097 (⫹/⫺ 199) 2327 1980–2327 *Uterus diameter, crown–rump length, and mass for the most developmentally advanced embryo (pictured in Fig. 2 and 3) collected at each gestational age is shown in bold type; the average for all normally developing specimens examined for each age is shown above (⫹/⫺ standard deviation); the range is shown below. dpc, days post coitum; AER, apical ectodermal ridge. these early stages, the morphology of Carollia embryos is planar in geometry, like that of human and most other mammals, but unlike the rodent egg cylinder (Streeter, 1942; Butler and Juurlink, 1987; O’Rahilly and Müller, 1987; Theiler, 1989; Downs and Davies, 1993; Behringer et al., 2000). Stage 10: Neural Fold Fusion Stage 10 embryos appear as an hourglass-shaped multilayered region (embryo proper) surrounded by the optically clearer extraembryonic membranes (Fig. 1A–C). The long axis of the opaque region defines the anterior–posterior (A-P) axis of the embryo. The most prominent features at this stage are the forming neural tube and somites (Fig. 1A). Somite formation begins during Stage 9. The first pair of somites form at the junction of the head and trunk, and as development proceeds, additional somite pairs form progressively from anterior to posterior. Stage 10 is defined by the presence of 4 –12 pairs of somites. The open neural plate folds medially and fuses dorsally to form the neural tube. Early in Stage 10 the neural tube is fused at the level of the trunk. By the end of this stage, fusion extends anterior to the hindbrain region and posterior to around the level of the newest-formed somite pair. As the lateral edges of the anterior neural plate elevate relative to the midline, and more somite pairs form in the posterior trunk, the previously planar embryo bends toward ventral in the midbrain region (cranial flexure) and in the region posterior to the newest somite pair. This bending of the embryo toward the ventral side reaches maximum curvature during late Stage 11 or early Stage 12 (Fig. 1D–G), after which the embryo progressively, but never completely, straightens. By the middle of this stage, a pair of slight depressions, called the optic sulcii, are seen in the forebrain region of the neural plate. Fig. 1. Stage 10 –13. A–C show two specimens at Stage 10 (4 –12 somites). A: Dorsal view of a four-somite embryo collected from a wild-caught female. The extraembryonic membranes have been removed to expose the dorsal surface of the embryo. B,C: Lateral view with dorsal to the left (B) and dorsal view (C) of an eight-somite embryo collected from a wild-caught female. D,E: Stage 11 (13–20 somites); lateral view with dorsal to the left (D) and dorsal view (E) of an 18-somite embryo collected from a wild-caught female. F,G: Stage 12 (21–29 somites); lateral view with dorsal to the left (F) and dorsal view (G) of a 26-somite embryo collected from a wild-caught female. H,I: Stage 13 (30 –35 somites); lateral view with dorsal to the left (H) and dorsal view (I) of a 31-somite embryo collected from a wild-caught female. aer, apical ectodermal ridge; al, allantois; cn, caudal neuropore; crf, cranial flexure; ela; endolymphatic appendage; ga, glossopharyngeal arch; fl, forelimb bud; h, heart; ha, hyoid arch; hl, hindlimb bud; lp, lens placode; ma, mandibular arch; nt, neural tube; opc, optic cup; ope, optic evagination; otv, otic vesicle; rn, rostral neuropore; so, 1st somite; tb, tail bud; v, ventricle. All specimens are oriented with anterior at the top. Scale bars ⫽ 1 mm in A, in C (applies to B,C), in E (applies to D,E), in G (applies to F,G), in I (applies to H,I). 726 CRETEKOS ET AL. The optic sulcii will form the rudiments of the eyes. The otic placodes, or rudiments of the ears, are visible in the ectoderm overlying the hindbrain. The heart tube and allantois are apparent at the anterior and posterior ventral midline, respectively, by the end this stage (Fig. 1B). Stage 11: Rostral Neuropore Closure The embryo is increasingly curved toward ventral. The posterior end of the embryo spirals past the head, usually but not invariably to the right side (Fig. 1D,E). Stage 11 is defined by the presence of 13–20 pairs of somites. This stage is marked by the closure of the anterior end of the neural tube, or rostral neuropore (Fig. 1E), when there are 15–16 pairs of somites formed. Cranial flexure reaches approximately 90 degrees by the middle of this stage (Figs. 1D, 6A) and the ventricles of the brain (Fig. 1E) are readily apparent. The first two pharyngeal arches (mandibular and hyoid) become distinct to either side of the hindbrain region and begin to extend distally (Figs. 1D, 6A). The heart tube is larger, segmented into two distinct chambers, and looped rightward (Fig. 1E). The optic sulcii, seen as evaginations from the lateral forebrain after neural tube closure (Figs. 1D,E, 6A), become more prominent, and by the end of this stage, have begun to form optic vesicles. The otic placodes invaginate and take on a saclike morphology (Figs. 1D, 6A), but remain open until the early part of the next stage. The allantois expands, becoming spherical in morphology and highly vascularized (Fig. 1D,E). By the end of the stage, the allantois extends ventrally from the posterior trunk region and makes contact with the chorion (Rasweiler and Badwaik, 1997). Stage 12: Forelimb Bud The corkscrew-like curl of the main body axis reaches maximum extent by early Stage 12 (Figs. 1F,G, 2A–D). This stage is defined by the presence of 21–29 pairs of somites and marked by the first appearance of forelimb buds adjacent to the 7th through 12th somite when 21–22 pairs of somites are present (Figs. 1F,G, 2B,C). The allantois is fused with the chorion early in this stage, initiating the chorioallantoic placenta (Rasweiler and Badwaik, 1997; Rasweiler et al., 2000; Badwaik and Rasweiler, 2001). The optic evaginations enlarge and approach the overlying ectoderm (Figs. 1G, 6B), but the retina and lens are not yet seen. The otic vesicles close early in Stage 12 and take on a spherical morphology (Figs. 1F, 2A, 6B). The posterior end of the neural tube, or caudal neuropore, closes by the time 25 somite pairs are formed, and a tail bud is present (Figs. 1F, 2C). The first and second pharyngeal arches are larger and extend distally (Figs. 1F, 2A, 6B). A third pair of pharyngeal arches (glossopharyngeal) appears immediately posterior to them in the first half of this stage (Figs. 2A, 6B). By the end of this stage, the distal end of this outgrowth is partly concealed under the second (hyoid) arch (Fig. 1F). A regular heartbeat, with blood circulating through the pharyngeal region and around the neural tube, is seen in freshly dissected specimens by the time there are 26 somite pairs formed. Stage 13: Hindlimb Bud Overtly similar in overall appearance to those of the previous stage, Stage 13 embryos are defined by the presence of 30 –35 pairs of somites (Fig. 1H,I). This stage is marked by the first appearance of hindlimb buds adjacent to the 22nd through 28th somite pairs when there are 30 –31 pairs of somites (Fig. 1H,I). The first two pharyngeal arch pairs are larger and show distinct proximal and distal parts: A cleft, called the oral groove, is visible on the anterior margin of the first arch, dividing the distal mandible (or lower jaw) component from the proximal maxilla (or upper jaw and palate) component by the time there are 30 pairs of somites formed (Figs. 1H, 6C). The second, or hyoid, arch is dumbbell shaped when viewed laterally (Figs. 1H, 6C). The third arch extends under the hyoid and by the end of this stage only the most proximal portion is visible in lateral view (Figs. 1H, 6C). From the second half of the stage, an apical ectodermal ridge (AER) is seen at the distal edge of the forelimb bud (Fig. 1H). The optic evaginations form optic cups, and lens placodes are seen in the overlying ectoderm by the end of this stage (Figs. 1H, 6C). The otic vesicle loses its spherical shape as the endolymphatic appendage initiates dorsally (Figs. 1H, 6C). Stage 14: Pigmented Retina Defined by the presence of 36 – 40 pairs of somites, the first appearance of faint scattered pigment in the retina of the eye marks the start of this stage, and during its course, retinal pigmentation is progressively more uniform (Figs. 2E, 6D). The curve of the main body axis is interrupted by a sharp angle at the head/trunk junction called the cervical flexure (Fig. 2E–H). The axis is also straighter, and by the end of this stage, the head and tail are no longer adjacent to each other. The first pharyngeal arch is further divided into maxillary and mandibular components by the deepening oral groove (Figs. 2E,H, 6D). A pair of nasal pits are clearly evident just anterior to the distal end of the maxilla by the time there are 36 somite pairs (Figs. 2E, 6D). The hindlimb bud is approximately as wide as it is long, whereas the forelimb bud is distinctly longer than it is wide (Fig. 4A,B). When there are 37 somite pairs, a prominent bulge appears proximally on the anterior edge of the forelimb bud (Figs. 2E, 4A). It is from this region that the primordium of the propatagium, or the portion of the wing membrane that will eventually stretch between the shoulder and wrist, emerges. A second bulge appears at the junction of the posterior edge of the forelimb bud and flank (Fig. 2H). Again based on position and subsequent morphogenesis, this structure is the primordium of the plagiopatagium, or the portion of the wing membrane that will eventually stretch from the end of the fifth digit of the forelimb to the ankle (Fig. 4K,L). When there are 38 somite pairs, the heart appears to be enclosed by the body wall but is still protruding ventrally (Fig. 2E). The rudiment of the genital tubercle first becomes evident as a midline thickening of the ventral body wall at the level of the hindlimb buds, between the umbilicus and tail. AERs are present at the dis- Fig. 2. Stage 12–17. A–T: The first column (A,E,I,M,Q) shows lateral views with dorsal to the left, the second column (B,F,J,N,R) shows ventral views, the third column (C,G,K,O,S) shows dorsal views of the head and trunk, and the fourth column (D,H,L,P,T) shows dorsal views of the trunk and tail. A–D: A Stage 12 specimen collected from a timed mating at 40 days post coitum (dpc) with 26 somite pairs formed. E–H: A Stage 14 (36 – 40 somites) specimen collected from a timed mating at 44 dpc, with 37 somite pairs formed. I–L: A Stage 15 specimen collected from a timed mating at 46 dpc. M–P: A Stage 16 specimen collected from a wild-caught female and morphologically stage-…
Purchase answer to see full attachment