Prenatal development of the composite occipito-atlanto-axial synovial joint cavity in the dog.код для вставкиСкачать
THE ANATOMICAL RECORD 216423-433 (1986) Prenatal Development of the Composite OccipitoAtlanto-Axial Synovial Joint Cavity in the Dog ALASTAIR G. WATSON, ALEXANDER DE LAHUNTA, AND HOWARD E. EVANS Department of Veterinary Anatomy, New York State College of Veterinary Medicine, Cornell University, Ithaca, NY 14853 (A.G.W, A.d L., H.E.E.); Orthopaedic Research Laboratory, Shriners Hospital for Crippled Children, Tampa, FL 33612 (A.G.W) ABSTRACT In the dog, the synovial cavities of the atlanto-occipital and atlantoaxial joints communicated to form a single large composite joint cavity. The prenatal development of this composite occipito-atlanto-axial joint cavity was studied by examining 26 serially sectioned dog embryos and fetuses that ranged in size from 19 to 68 mm crown-rump length, and were between 30 and 42 days of gestational age. In the composite occipito-atlanto-axial joint, trilaminar interzones developed a t 1922 mm (30-31 days), joint cavities opened a t 27-32 mm (33-34 days), and the atlantooccipital and atlanto-axial cavities first communicated a t 48 mm (37 days). The cranio-vertebral junction has evolved from a relatively simple intervertebral type of articulation, as in fishes, into a markedly specialized triosseous complex in mammals-the occipito-atlanto-axial complex (Watson, 1981). This unique bony complex has become greatly modified and highly specialized in its structure and function. The general features of the phylogenetic and ontogenetic development of the cranio-vertebral junction have been covered in standard texts of comparative anatomy (for example, Goodrich, 1930; Romer and Parsons, 1986), and more detailed accounts are also available (Gadow, 1933; de Beer, 1937; Evans, 1939; Sensenig, 1957; Jenkins, 1969, 1971). These studies, nevertheless, have tended to deal more with comparisons of adult morphology, or with only one part of this junctional complex. Recent work has emphasized that, in mammals at least, the occipital bones, atlas, and axis have direct phylogenetic, ontogenetic, structural, and functional interrelationships (Mayhew et al., 1978; Watson, 1981; Watson et al., 1985a,b, 1986b). Moreover, these studies suggest that the cranio-vertebral junction may be better represented by the occipito-atlanto-axial complex rather than by the atlanto-occipital or atlanto-axial concepts. The embryonic development of the cranio-vertebral region has been reviewed and documented by O’Rahilly and associates (O’Rahilly and Meyer, 1979; Muller and O’Rahilly, 1980; O’Rahilly et al., 1980, 1983; O’Rahilly and Miiller, 1984a). These detailed accounts are based on precise reconstruction of staged human embryos and are limited to specimens from the embryonic period proper. The greater part of synovial joint development, however, occurs during the fetal period. Specific studies on the sequential development of the synovial joints of the occipito-atlanto-axial complex in mammals have not been found. In adult dogs it has been shown that the cavities of the atlanto-occipital and atlanto-axial joints are interconnected to form a single composite occipito-atlantoaxial joint cavity (Watson et al., 1986~). The objective of 0 1986 ALAN R.LISS, INC. this present study was to determine when the synovial cavities of the occipito-atlanto-axial complex developed and when their various joint spaces coalesced to form the single joint cavity. This was achieved by studying serially sectioned dog embryos, some of which were Beagles of known age. MATERIALS AND METHODS Serial sections of 26 dog embryos and fetuses from the Cornell collection of serially sectioned embryos (Evans and Sack, 1973) were examined. Dog embryos have not been classified into a n adequate series of “developmental stages,” as has been done for human embyros (O’Rahilly, 1973, 1979), and embryonic length by itself is not necessarily a n exact indicator of the stage of development. External measurements have been used as a n indicator of gestational age for dog embryos of the Beagle breed: in addition, estimates of crown-rump length for each day of gestation along with external morphological features and ossification patterns were determined (Evans, 1974). In these Beagles, the embryonic period extended up to day 35 postinsemination, at which time embryos were 35 mm in crown-rump length; embryos of larger breeds may be a little longer. The fetal period extended from day 36 to birth (60-63 days), a t which time the pups averaged 160 mm crown-rump length (Evans, 1979). Nevertheless, in developmental studies of domestic mammalian embryos, crown-rump length is the measurement most widely used (Evans and Sack, 1973), and thus the dog embryos and fetuses examined in this study were arranged into a series based on crown-rump length. More recently, it has been recommended that for human embryos and fetuses the greatest length is a more satisfactory measurement to use than the traditional crownReceived March 31, 1986; accepted July 8, 1986. Address inquiries and reprint requests to Dr. A. de Lahunta, Department of Anatomy, NYS College of Veterinary Medicine, Cornell University, Ithaca, NY 14853. 424 A.G. WATSON, A. DE LAHUNTA, AND H.E. EVANS OCCIPITO-ATLANTO-AXIALJOINT DEVELOPMENT IN DOG rump length (ORahilly and Muller, 1984b). The implications of this proposal for studies of nonhuman embryos have yet to be determined. Six of the embryos in this present study were purebred Beagles of known age (Evans, 1974); the other embryos were of unknown breeding and their ages were estimated from growth curves (Evans and Sack, 1973; Evans, 1979).The 26 embryos and fetuses ranged in size from 19 to 68 mm crown-rump length, and were estimated to be between 30 and 42 days of gestational age. The atlanto-occipital joint is at a flexion point between the head and the neck; this flexion is particularly marked in small embryos. Because of this marked flexion, there is also a marked change in orientation of the longitudinal axis from the head to the neck. For the purpose of this study, the plane of section was taken with reference to the axis, except for description of the atlanto-occipital joint by itself, in which case the plane of section was taken with reference to the base of the skull or the occipital condyle. The fixative was usually formalin, and double embedding in celloidin and p a r f i i n or in paraffin alone was used. The sections were 10-25 pm in thickness, sectioned in the sagittal, transverse, or dorsal planes, and most were stained with hematoxylin and eosin or with carmine. Representative sections were photographed. OBSERVATIONS In the dog embryo, four synovial cavities were found in the occipito-atlanto-axial complex: 1) the composite occipito-atlanto-axial joint cavity herein described; 2 and 3) the right and left joint cavities between the right and left articular processes of the axis and the third cervical vertebra; and 4) a synovial bursa between the dens and the transverse atlantal ligament. The development of the latter three synovial cavities are described elsewhere (Watson et al., 1986a). Composite Occipito-Atlanto-AxialJoint Cavity The occipito-atlanto-axial joint cavity is a composite cavity formed by the coalescence of the atlanto-occipital joint cranially, with the atlanto-axial joint cavity cau- A. AB, B, BC, D, 0, S, X, Abbreviations Atlas Body of atlas Brain Basicranium, caudal end Dens Occipital condyle Cervical spinal cord Body of axis Fig. 1. Photomicrographs from serially sectioned dog embryos showing the development of the lateral atlanto-occipitaljoint between the occipital condyle and the cranial articular fovea of the atlas. In the sagittal sections (a,b,e) cranial is to the left and dorsal is to the top of the page. In the dorsal sections (c,d,f,g)cranial is to the top and lateral is to the right of the page. a) Sagittal section of a 19-mm embryo showing the trilaminar interzone. ~ 6 0 b) . Higher power view of a similar area from the same embryo in “a.” x 140. c) Dorsal section of a 21.5-mm embryo showing early joint cavitation. ~ 6 0d). Dorsal section of a 21-mm embryo showing intercellular spaces in the intermediate layer of the interzone. x90. e) Sagittal section of a 27-mm embryo showing an open joint cavity. X60. f7 Dorsal section of a 29-mm embryo showing joint cavitation. ~ 6 0 g) . Dorsal section of a 35-mm embryo showing a well established joint cavity. ~ 6 0 . 425 dally. Each of these two cavities develop from three separate synovial joint spaces; a bilateral pair of joint spaces and a median space, which initially forms one cavity (atlanto-occipital) cranial to the atlas, and another cavity (atlanto-axial) caudal to the atlas. Subsequently these two cavities coalesce ventral to the cranial tip of the dens to form the common occipito-atlanto-axial joint cavity (Table 1). Atlanto-Occipital Joint The atlanto-occipital joint cavity develops initially from three separate joint spaces: 1)bilaterally, a pair of lateral atlanto-occipital joint spaces, between the occipi. tal condyles and the cranial articular foveae of the atlas; and 2) a median atlanto-occipital joint space between the caudal end of the basicranium, the cranial surface of the body of the atlas, and the ventral surface of the cranial tip of the dens. Subsequently, these three joint spaces coalesce and form a single atlanto-occipital joint. The atlanto-occipital cavity communicates caudally with the joint space (atlanto-axial) between the dens and the body of the atlas. Various stages in the development of the atlanto-occipital joint are illustrated in Figures 1-3. At 19 mm the occipital condyles and the atlas are formed by cartilage, each surrounded by a clearly differentiated condensation of cells-the perichondrium. The area between each cartilage is composed of a band of cells that forms the joint interzone. The joint interzones of the lateral atlanto-occipital joints are trilaminar and clearly developed over the full width and depth of the condyles and cranial articular foveae (Fig. la). The central or intermediate layer of the interzone develops as a relatively loose aggregation of cells with small intercellular spaces; the intermediate layer thus appears “lighter” in these sections (Fig. lb). In contrast, the layer of the interzone adjacent to the cartilage develops as a relatively dense aggregation of cells, the chondrogenic layer, which is continuous with the perichondrium; the chondrogenic layer thus appears “darker” in these sections. In addition, the joint capsule is more developed dorsally than ventrally (Fig. lb), and in the intercondylar area the intermediate layer shows the earliest loosening stage (Fig. 2a). In the 21-22.5-mm embryos the intermediate layer is well developed in the pair of lateral joints. There are numerous elongate and stellate cells, and prominent intercellular spaces, which are more prominent than the lateral part (Figs. Id, 30. At 21.5 mm some of the intracellular spaces are coalesced to form larger spaces (Fig. lc). The medial part of the joint capsule is more prominent than the lateral part (Figs. lc,d). In the median joint, the mesenchyme between the basicranium and the dens is looser than at 19 mm, and a trilaminar interzone is present (Figs. 2b, 5a,b), which is continuous laterally with the medial part of the interzone of the lateral joints. At 21-22.5 mm there is thus a common trilaminar joint interzone connecting the pair of lateral parts with the median part of the atlanto-occipital joint, and also connects caudally with the interzone of the dentoatlantal joint. At 27-29 mm the intercellular spaces coalesce to form large spaces (Fig. 10 and in most there is a well defined lateral joint space, with flattened cells lining the cartilaginous articular surfaces (Fig. le). These joint spaces 426 A.G. WATSON, A. DE LAHUNTA, AND H.E. EVANS 427 OCCIPITO-ATLANTO-AXIALJOINT DEVELOPMENT IN DOG TABLE 1. Summary of the relationship between embryo length (mm) and four developmental stages of synovial cavities in the occipito-atlanto-axialjoint of the dog Cavity Atlanto-occipital Lateral Median’ Atlanto-axial Lateral Median’ Homogeneous interzone - DeveloDmental stages Initial Trilaminar appearance interzone of joint spaces Coalescence of joint spaces 19 19 22 22 29 27-29 38-39 19 19 21-22 21-22 29 29 32-39 32-38 ‘The lateral atlanto-occipital joint cavities communicated with the median atlanto-occipital cavity at 48 mm, and the lateral atlanto-axialjoint cavities communicated with the median atlanto-axial cavity at 58 mm. The median atlanto-occipital joint cavity communicated with the median atlanto-axial cavity at 48 mm, and all cavities communicated at 58 mm to form the composite occipito-atlanto-axialjoint cavity. are relatively free of cells, cover the full width of the occipital condyles, and extend medially to the level of the lateral border of the dens (Figs. 2d,f, 3h). At the ventral part of the atlanto-occipital joint, the median, or intercondylar area consists of loose mesenchyme with only a few small intercellular spaces (Figs. 2c,d, 3i). In the 32-35-mm embryos and larger fetuses the lateral joint spaces are wide open, and the joint capsule is prominent over the dorsal and medial aspects (Fig. lg, 3b,c,e).In contrast, the ventral and lateral aspects of the joint capsule are poorly defined (Figs. lg, 3c). Medially, however, there is some development of a ventral capsule between the basicranium and the body of the atlas, and the joint space is formed at 32 mm (Fig. 2e). This medial joint space is clearly defined in the 38-39-mm fetuses. The lateral atlanto-occipital joint spaces communicate medially with the median joint space at 48 mm (Fig. 3g). This communication is caudal to the basicranium through the intercondylar area. Furthermore, this lateral-to-lateral communication also communicates caudally with the now well developed joint space between the dens and the atlas body. This caudal communication is ventral to the cranial tip of the dens, at which point Fig. 2. Photomicrographs from serially sectioned dog embryos showing the development of the median atlanto-occipitaljoint between the caudal end of the basicraniurn, the cranial tip of the dens and the cranial surface of the body of the atlas. The development of the median atlanto-axial joint between the dens and the atlas body is also seen. All sections were cut sagittally; cranial is to the left, and dorsal (spinal cord and brain) is to the top of the page. In the median sections (ac,e,g) the notochord is seen passing through the center of the dens and more caudally, through the axis body. a) Median section of 19-mm embryo showing the interzone between skull and dens. The perichondria between the dens and atlas body are poorly differentiated. x60. b) Median section of a 21.5-mm embryo showing dense interzones in the atlanto-occipital and atlanto-axial joints. ~ 6 0c). Median section of a 29-mm embryo showing a small joint space between the dens and atlas body. x 60. d) Sagittal section at lateral edge of the dens, in same 29-mm embryo as “c.” At this level the joint interzone showed small spaces. ~ 6 0e). Median section of a 32-mm embryo showing prominent joint spaces between the dens and the atlas body and the caudal skull. ~ 6 0 0. Sagittal section in same 29-mm embryo as “c” and “d,” but section is lateral to dens and medial to the neural arches. The joint cavity caudoventral to the base of the skull is the medial extent of the lateral atlanto-occipitaljoint cavity. X60. g) Median section of 67-mm fetus showing well established joint cavities, which are connected ventral to the cranial tip of the dens. ~ 3 0h) . Sagittal section lateral to the dens in same 67-mm fetus in “g,” showing that the joint cavities are not continuous lateral to the dens. ~ 3 0 . the pair of lateral and the median atlanto-occipitaljoint spaces and the dento-atlantal joint spaces intercommunicate (Fig. 3g). In one 52-mm fetus a small band of tissue from each side of the dens to the atlas body and the medial part of the occipital condyle persisted and thus prevented the joint space intercommunication. In the larger fetuses the intercommunication is prominent, but only extends the width of the dens (Fig. 2g,h). Atlanto-Axial Joint The atlanto-axial joint cavity, like the atlanto-occipital joint cavity, develops initially from three joint spaces: 1) bilaterally, a pair of lateral atlanto-axial joint spaces between the caudal articular foveae of the atlas and the cranial articular surfaces of the axis; and 2) a medial atlanto-axial (dento-atlantal) joint space between the ventral articular surface of the dens and the fovea dentis on the dorsal surface of the body of the atlas. Subsequently, these three joint spaces coalesce and form a single atlanto-axial joint cavity. Various stages in the development of the atlanto-axial joint are illustrated in Figures 2-5. At 19 mm the joint interzones are homogeneous. The median joint shows a “lighter” interzone where the perichondria merge between the ventral surface of the dens and the dorsal surface of the body of the atlas (Fig. 2a); whereas the lateral joints show a “darker” interzone owing to a dense aggregation of cells (Fig. 3a). The interzone is trilaminar in the caudal part of the median joint and in the dorsolateral parts of the lateral joints in 21-22.5-mm embryos (Fig. 4a). The other areas of the joints are represented by dense aggregations of cells in a homogeneous interzone (Figs. 2b, 3a,f). At 27 mm the interzone is trilaminar and well developed throughout the three parts of the joint. At 29 mm, in the median joint, there is a small but well developed joint space (Figs. 2c, 5c) that extends laterally almost for the full width of the dens (Fig. 2d). The lateral atlanto-axial joints each have a well developed joint space that extends the full width and depth of the adjacent articulating cartilages (Figs. 3h, 4b). The joint areas between the median and lateral joints are composed of well developed trilaminar interzones with an intermediate layer of loose mesenchyme (Figs. 2f, 4c,d). In the 32-38-mm embryos the median joint has expanded to the full width and length of the dens (Figs. 2e, 5e), and is not connected to the atlanto-occipital nor to the lateral atlanto-axial joints, although the interme- 428 A.G. WATSON, A. DE LAHUNTA, AND H.E. EVANS OCCIPITO-ATLANTO-AXIALJOINT DEVELOPMENT IN DOG diate layer of the interconnecting interzone has a few small intercellular spaces. The lateral atlanto-axial joints each have a prominent joint cavity and joint capsule (Fig. 3b,d). At 52 mm some of the spaces in the interconnecting intermediate layer are coalesced t o form larger spaces (Fig. 4e), but the median joint cavity (Fig. 50 is not continuous with the lateral joint cavities (Fig. 40. At 58 mm and in the larger fetuses, the median atlanto-axial joint cavity is in broad communication caudally with the pair of lateral cavities. In these fetuses, the median cavity extends as a C-shaped cavity to the middorsoventral depth of the dens (Fig. 5g), and caudally has a broad connection with each lateral atlantoaxial cavity (Fig. 4g). In the sagittal plane the median joint cavity is wide open and extends the full width and length of the dens, and is continuous cranially with the median atlanto-occipital joint cavity (Fig. 2g,h). DISCUSSION Synovial Cavity Development The prenatal development of the synovial joint cavities in the occipito-atlanto-axial complex of the dog followed the typical sequence of developmental stages that have been previously described for synovial joints in the limbs of man and other mammals (Haines, 1947; Barnett et al., 1961; Mitrovic, 1978; O’Rahilly and Gardner, 1978). Studies on the development of the synovial cavities of the occipito-atlanto-axial complex in mammals have not been found, although some aspects of the development of joints in this region in chicken embryos have been reported (Murray and Drachman, 1969). The histological interpretation of synovial joint development is often complicated by preparation artifacts, particularly the compression of the intermediate layer of the interzone, which is thought to be caused by the expansion of the adjacent cartilages (Haines, 1947).This Fig. 3. Photomicrographs from serially sectioned dog embryos showing the development of the lateral atlanto-occipital and lateral atlantoaxial joints. In the sagittal sections (a-e) cranial is to the left and dorsal is to the top of the page. In the dorsal sections (f,h,i) cranial is to the top of the page. a) Sagittal section of a 19-mm embryo cut at the level of the occipital condyles, showing homogeneous interzones in the lateral atlanto-occipital and atlanto-axial joints. X60. b) Sagittal section at the level of the occipital condyles from a 32-mm embryo showing the orientation of the occipito-atlanto-axialcartilages and the open joints. Higher power views in “c” and “d.” ~ 2 0c). Higher power view of the atlanto-occipital joint from the same 32-mm embryo in “b.” There is no condensation for the ventral joint capsule. x60. d) Higher power view of the atlanto-axial joint from the same 32-mm embryo in “b.” The joint is wide open and the capsules are prominent. ~ 6 0 e) . Sagittal section at the level of the occipital condyle from a 67-mm fetus showing the well developed joint cavities. ~ 3 0 . Dorsal 0 section at the level of the dens from a 22.5-mm embryo, showing the trilaminar interzone in the atlanto-occipital joints and the homogeneous stage in the atlanto-axial joints. ~ 2 0g) . Transverse section of a 48-mm fetus showing the continuity of the lateral atlanto-occipital joint cavities across the median, and also the continuity of the atlanto-occipital cavity with the median atlanto-axial cavity ventral to the cranial tip of the dens. The middle area of the atlanto-occipital joint cavity on the left is obliterated by a compression artifact. ~ 2 0h) . Dorsal section at the level of the dens from a 25-mm embryo showing open lateral joint cavities. This embryo was markedly flexed and was 4 mm shorter than its litter mates although its “stage” of development was similar to those at 29 mm. ~ 2 0 i). Dorsal section ventral to the dens from the same 25-mm embryo in “h.” The atlanto-occipitaljoint cavities are not continuous across the median. ~ 2 0 . 429 and other artifacts may suggest that a particular joint is at an earlier or later stage of development than it actually is; however, the interpretive problems associated with these preparation artifacts have been elucidated (Haines, 1947). Even though a number of preparation artifacts were encountered during this study of serially sectioned dog embryos, a sequence of developmental stages for the synovial cavities was established (Table 1).The number of dog embryos within each stage was limited, and only six of the 26 embryos and fetuses studied were of known gestational age. Consequently, the sizes assigned to any one developmental stage should be considered as provisional. More dog embryos and fetuses of appropriate sizes, and preferably of known age and breed, are required to firmly establish the ages and sizes for the developmental stages of the occipito-atlanto-axialjoint in the dog. In this prenatal study of the dog, the lateral atlantooccipital joint cavity developed in advance of the other joint cavities in the complex. Notwithstanding the limitations of preparation artifacts and the limited number of embryos examined, the relatively earlier development of this particular joint may be real. Most of the movements between the head and the neck in adult dogs occur at the atlanto-occipital joint (flexion and extension) (Evans and Christensen, 1979), and thus as spontaneous movements develop in the embryo, this joint may be expected to develop a little before the others in this region in response to the earlier functional demands upon it. Contraction of skeletal muscles are necessary for spontaneous movements and for normal joint development in embryos (Drachman and Sokoloff, 1966). And conversely, experimentally induced neuromuscular paralysis in chicken embryos has lead to improper joint development and extensive cartilaginous union of the atlanto-occipital joint and between adjacent vertebrae (Sullivan, 1966; Murray and Drachman, 1969). The role of flexion and extension of the head on the neck should be studied further. Correlative studies of joint and muscle development, and body movements in embryos and fetuses are thus needed. Alternatively, one could suggest that since the somites develop in a craniocaudal sequence (man-Sensenig, 1957; O’Rahilly and Meyer, 1979; dog-Holst and Phemister, 1970; Phemister and Holst, 1974),the somites that contribute to the atlanto-occipital joint are older than the somites that contribute to the atlanto-axial joint, and as a result of the apparently older anlagen the atlanto-occipital joint should develop prior to the atlanto-axial joint. Even though somites do develop in a craniocaudal sequence, the occipital bones, the vertebral centra, and the vertebral neural arches do not ossify in a craniocaudal sequence in either man (Bagnall et al., 1977; Ford et al., 1982)or dog (Evans, 1974). Intervertebral Muscles and Positions of Intervertebral Joints In the development of a series of typical trunk vertebrae, a myotome initially lies adjacent to each sclerotome, but when the sclerotomal cells migrate and together with cells from the perichordal tube form the vertebral anlagen, each myotome becomes intervertebra1 in position-spanning from one vertebra to the next (Dalgleish, 1985). The muscles of the vertebral column, which are derived from the myotomal (somitic)cells, can thus produce bending movement between one vertebra 430 A.G. WATSON, A. DE LAHUNTA, AND H.E. EVANS OCCIPITO-ATLANTO-AXIALJOINT DEVELOPMENT IN DOG and the next (Hamilton and Mossman, 1972). The position and basic form of synovial joints are determined by intrinsic factors independent of movement (O’Rahilly and Gardner, 1978). Experimental studies suggest that movement is necessary, however, for the initiation and maintenance of cavitation in joints of birds (Sullivan, 1966; Murray and Drachman, 1969), although the precise role of movement in the development of joints in mammals has not been clarified (O’Rahilly and Gardner, 1978). Nevertheless, even though a myotome does span between two vertebrae, it does not necessarily follow that a myotome should span equally across the intervertebral joint in the adult. It has been proposed (O’Rahilly and Meyer, 1979) that in man, at least, the intervertebral disc and the vertebral arch with its articular processes are derived from the caudal part of the sclerotome. This asymmetry may have arisen from differential growth of the two halves of the somite, or from an asymmetrical location of the intervertebral joint condensation. Nonetheless, most accounts of the early development of the vertebral column have depended on the resegmentation (Neugliederung) concept-a concept that has been challenged and rejected (Verbout, 1976, 1985; Dalgleish, 1985). Furthermore, it has been shown that during the development of the thoracic vertebrae in mice embryos, the vertebral bodies and their intervertebral discs developed from the cells of the perichordal tube and the neural arch and transverse processes were Fig. 4. Photomicrographs from serially sectioned dog embryos showing the development of the atlanto-axial joint in transverse sections (dorsal is to the top of the page). a) Overall view from a 21-mm embryo showing the relationship of the spinal cord to the atlas and axis. The medial area of the joint interzone is looser, whereas the lateral areas show denser aggregations of cells. ~ 2 0b). Overall view from a 29-mm embryo showing joint cavitation in the lateral atlanto-axial joints. ~ 2 0c). Overall view, at a more cranial level, from the same embryo in “b.” Joint cavitation has occurred laterally and ventrally, although compression artifacts (dense bands dorsolateral) obliterate some of the spaces. X30. d) Overall view of the atlantal-axial joints in 35-mm embryo showing a well developed median joint cavity. The joint area dorsolaterally is obliterated by compression and overlap artifacts. There is a trilaminar interzone between the median and lateral joints. X30. e) Transverse section of the area between the lateral and median parts of the atlanto-axial joint in a 52-mm fetus. This area lateral to the caudal part of the dens shows cavitation in the intermediate layer. ~ 6 0 .Same 0 52-mm fetus as in “e” but sectioned caudal to the base of the dens. The cavity is the craniomedial extent of the lateral atlanto. Lateral atlanto-axial joint cavity lined by axial joint cavity. ~ 6 0 g) flattened cells in a 68-mm fetus. ~ 6 0 . 431 derived from the caudal sclerotorne half (Dalgleish, 1985).In addition, it was also shown that myotomal cells bridged the gap between the neural arches and thus, at least initially, bridged the intervertebral joint of the thoracic vertebrae. Intervertebral movements are therefore possible at this early stage for the appropriate initiation and maintenance of cavitation in the synovial joints of the vertebral column. The interpretation of original intervertebral boundaries in the occipito-atlanto-axialcomplex is further compounded because the vertebrae in this complex are greatly modified, and the somitic or sclerotomal boundaries for the individual bony elements are uncertain. The intervertebral muscles in this region have also been greatly modified. Occipito-Atlanto-Axial Concept Earlier studies suggested that, in mammals at least, the cranio-vertebral junction was constituted by more than the atlanto-occipital area: the occipital bones, the atlas, and the axis were all involved (Mayhew et al., 1978; Watson, 1981). Evidence has been accumulating that the occipito-atlanto-axialcomplex is perhaps a more appropriate designation for the cranio-vertebral junction. First, in studies of congenital malformations of the bones of this junction in several domestic species, it was shown that in all cases the occiput, atlas, and axis all were malformed and both the atlanto-occipital and the atlanto-axial joints were affected (Mayhew et al., 1978; Watson, 1979;Watson and Mayhew, 1986;Watson et al., 1985a,b).To reflect the triosseous nature of these specific malformations, we proposed that they be classified as occipito-atlanto-axial malformations. Second, investigations of the normal development of the ossification centers in this region have shown that there are direct ontogenetic and phylogenetic interrelationships of the atlas, axis, and the proatlas elements (Watson, 1981; Watson et al., 1986b). Third, a gross anatomical study in the dog has shown that the atlanto-occipital and the atlantal-axial joint cavities are interconnected to form a single composite synovial cavity (Watson et al., 19864. Fourth, at the cranio-vertebral junction there are muscles and ligaments, which unite the occiput to the axis, as well as others which link occiput to atlas, and atlas to axis (Evans and Christensen, 1979). And now, this study on the prenatal development of the composite occipito-atlanto-axial synovial joint cavity has shown the interrelationships and the interconnections of these joint spaces a t the cranio-vertebral junction. There is thus a significant interdependence for the developmental, structural, and functional interrelationships of the occipito-atlanto-axialcomplex. ACKNOWLEDGMENTS The authors thank Dr. Robert F. Smith for taking the photomicrographs and Wendie M. Smith for printing them. This work was supported by funds from the New York State College of Veterinary Medicine, Cornell University, and from the Shriners Hospitals for Crippled Children. 432 A.G. WATSON, A. DE LAHUNTA, AND H.E. EVANS OCCIPITO-ATLANTO-AXIALJOINT DEVELOPMENT IN DOG LITERATURE CiTED Bagnall, K.M., P.F. Harris, and P.R.M. Jones (1977) A radiographic study of the human fetal spine. 2. The sequences of development of ossification centres in the vertebral column. J. Anat., 124:792-802. Barnett, C.H., D.V. Davies, and M.A. MacConaill(1961) The prenatal and postnatal development of synovial joints. In: Synovial Joints. Their Structure and Mechanics. Thomas, Springfield, IL, Chapter II.4, pp. 127-138. Dalaleish, A.E. (1985)A study of the develoDment of thoracic vertebrae in the mouse assisted by autoradiography. Acta Anat. (Basel), 122:91-98. de Beer, G.R. (1937)The Development of the Vertebrate Skull. Oxford University Press, New York. Drachman, D.B., and L. Sokoloff (1966) The role of movement in embryonic joint development. Dev. Biol., 14:401-420. Evans, F.G. (1939) The morphology and functional evolution of the atlas-axis complex from fish to mammals. Ann. N.Y. Acad. Sci., 39:29-104. Evans, H.E. (1974) Prenatal development of the dog. In: 24th Gaines Vet. Symp. New York State Veterinary College, Ithaca, NY, pp. 1828. Evans, H.E. (1979) Reproduction and prenatal development. In: Miller’s Anatomy of the Dog, 2nd ed. H.E. Evans, G.C. Christensen, eds. Saunders, Philadelphia, Chapter 2, pp. 13-77. Evans, H.E., and W.O. Sack (1973) Prenatal development of domestic and laboratory mammals: Growth curves, external features and selected references. Anat. Histol. Embryol., 211-45. Evans, H.E., and G.E. Christensen (1979) The axial skeleton; ligaments and joints of the vertebral column; muscles of the vertebrae. In: Miller’s Anatomy of the Dog, 2nd ed. Saunders, Philadelphia, p. 163, pp. 232-240,303-316. Ford, D.M., K.D. McFadden, and K.M. Bagnall (1982) Sequence of ossification in human vertebral neural arch centers. Anat. Rec., 203:175-178. Gadow, H.F. (1933)The Evolution of the Vertebral Column. A Contribution to the Study of Vertebrate Phylogeny. J.F. Gaskell, H.L.H.H. Green, eds. Cambridge University Press, London. Goodrich, E.S. (1930) Studies on the Structure and Development of Vertebrates. MacMillan, London. Haines, R.W. (1947) The development ofjoints. J. Anat., 81:33-55. Hamilton, W.J., and H.W. Mossman (1972) Skeletal System. In: Hamilton, Boyd and Mossman’s Human Embryology, 4th ed. Williams & Wilkins, Baltimore, Chapter 14, pp. 526-547. Holst, P.A., and R.D. Phemister (1970) Prenatal canine development: Description of normal developmental stages. Annual Report for 1969, Collaborative Radiological Health Laboratory, Colorado State University, Fort Collins, CO, pp. 102-106. Jenkins, F.A. (1969)The evolution and development of the dens of the Fig. 5. Photomicrographs from serially sectioned dog embryos showing the development of the median atlanto-axial joint cavity ventral to the dens in transverse section (dorsal is to the top of page). Stages in the development of the transverse atlantal ligament and its bursa are also seen (d-g). a) Section from a 21-mm embryo, at the cranial end of the dens showing its relationship to the interzone at the intercondylar occipital notch. x60. b) Higher power view of central interzone area from “a,” The mesenchyme is loosely aggregated. X140. c) Section through the middle of the dens in a 29-mm embryo showing joint cavitation between the ventral surface of the dens and the dorsocranial part of the atlas body. ~ 6 0dl . Section through dens in the same 29mm embryo as in “c” but at a more cranial level. Note the bursa of transverse ligament. x90. e) Transverse section of a 34-mm embryo showing the early joint cavity ventral to the dens, and the transverse atlantal ligament and its bursa. ~ 3 0f). Transverse section of a 52-mm fetus showing the open joint cavity ventral to the dens. ~ 6 0g). Transverse section of a 59-mm fetus showing a well-established median atlantal-axial joint cavity between the dens and the body of the atlas. Vascular synovial membrane and joint capsule are seen lateral to the dens. ~ 6 0 . 433 mammalian axis. Anat. Rec., 164t173-184. Jenkins, F.A. (1971) The postcranial skeleton of African cynodonts. problems in the early evolution of the mammalian postcranial skeleton, ~ ~ 1Peabody 1 , M ~Nat, ~ Hist., , 36;1-216, Mavhew. LG,, A.G. Watson. and J,A, ~~i~~~~(1978) Congenital occiD. ito-atlanto-axial malformations in the horse. Equine Vet. J., lOt103113. Mitiovic, D. (1978) Development of the diarthrodial joints in the rat embryo. Am. J. Anat., 151:475-486. Muller, F.,and R. O’Rahilly (1980)The human chondrocranium at the end of the embryonic period proper with particular reference to the nervous system. Am. J. Anat., 159:33-58. Murray, P.D.F., and D.B. Drachman (1969) The role of movement in the development of joints and related structures: The head and neck in the chick embryo. J. Embryol. Exp. Morphol., 22:349-371. O’Rahilly, R. (1973)Developmental Stages in Human Embryos, Including a Survey of the Carnegie Collection. Part A: Embryos of the First Three Weeks (Stages 1 to 9). Carnegie Institute, Washington, DC, Publ. No. 631,167 pp. O’Rahilly, R. (1979) Early human development and the chief sources of information on staged human embryos. Eur. J. Obstet. Gynecol. Reprod. Biol., 9:273-280. O’Rahilly, R. and E. Gardner (1978)The embryology of movable joints. In: The Joints and Synovial Fluid. L. Sokoloff, ed. Academic Press, New York, Vol. 1, Chapter 2, pp. 49-103. O’Rahilly, R., and D.B. Meyer (1979) The timing and sequence of events in the development of the human vertebral column during the embryonic period proper. Anat. Embryol. (Berl), 157:167-176. O’Rahilly, R., and F. Miiller (1984a) The early development of the hypoglossal nerve and occipital somites in staged human embryos. Am. J. Anat., 169:237-257. ORahilly, R., and F. Miiller (198413) Embryonic length and cerebral landmarks in staged huyman embryos. Anat. Rec., 209:265-271. O’Rahilly, R., F. Muller, and D.B. Meyer (1980) The human vertebral column at the end of the embryonic period proper. 1,The column as a whole. J. Anat., 131565-575. O’Rahilly, R., F. Muller, and D.B. Meyer (1983) The human vertebral column at the end of the embryonic period proper. 2. The occipitocervical region. J. Anat., 136:181-195. Phemister, R.D., and P.A. Holst (1974) Prenatal development of the dog. Annual Report for 1973, Collaborative Radiological Health Laboratory, Colorado State University, Fort Collins, CO, pp. 5659. Romer, A.S., and T.S. Parsons (1986) The Vertebrate Body, 6th ed. Saunders, Philadelphia. Sensenig, E.C. (1957) The development of the occipital and cervical segments and their associated structures in human embryos. Contrib. Embryol., 36(248): 141-152. Sullivan, G.E. (1966) Prolonged paralysis of the chick embryo, with special reference to effects on the vertebral column. Aust. J. Zool., 14:1-17. Verbout, A.J. (1976) A critical review of the ‘Neugliederung’ concept in relation to the develoument of the vertebral column. Acta Biotheor. (Leiden), 25:219-2b. Verbout, A.J. (1985) The development of the vertebral column. Adv. Anat. Embrvol. Cell Biol., 90:l-122. Watson, A.G. i1979) Congenital occipitoatlantoaxial malformation (OAAM) in a dog. Anat. Histol. Embryol., 8:187. Watson, A.G. (1981) The Phylogeny and Development of the OccipitoAtlas-Axis Complex in the Dog. W.D. thesis, Cornell University, Ithaca, NY. Watson, A.G., and I.G. Mayhew (1986) Familial congenital occipitoatlanto axial malformation 10AAM) in the Arabian horse. Spine, 11:334-339. Watson, A.G., M.A. Hall, and A. de Lahunta (1985a) Congenital occipito-atlanto-axial malformation (OAAM) in a cat. Compend. Cont. Educ. Prac. Vet., 7:245-254. Watson, A.G., J.H. Wilson, A.J. Cooley, G.A. Donovan, and C.P. Spencer (1985b) Occipito-atlanto-axial malformation with atlanto-axial subluxation in an ataxic calf. J . Am. Vet. Med. Assoc., 187340742. Watson, A.G., A. de Lahunta, and H.E. Evans (1986a) Prenatal development of the bursa of the transverse atlantal ligament in the dog. (in prep). Watson, A.G., H.E. Evans, and A. de Lahunta (1986b) Ossification of the atlas-axis complex in dog. Anat. Histol. Embryol., 15122-138. Watson, A.G., H.E. Evans, and A. de Lahunta (1986~)Gross morphology of the composite occipito-atlas-axis joint cavity in the dog. Anat. Histol. Embryol., 15~139-146.