Abstract
The gross anatomy of the human embryo, far from being an mere catalog of obscure and transitory structures, is of direct relevance for all clinicians and researches involved in the study and treatment of craniofacial anomalies. What we have long needed is an understandable gross anatomy of the embryo. What has been missing are terms of reference that allow us to visualize the physical steps involved in head and neck construction. When we understand the processes of embryogenesis in terms of the mechanisms that produce this gross anatomy, we can see in congenital aberrations of the system clues as to the pathogenesis of these anomalies.
The human embryo is constructed from a series of developmental fields which are identifiable by their neurovascular relations. Ultimately, the definition of each field can be traced backward to the earliest stages of its development. Aberrant size or absence of a field leads to predictable structural changes in that zone in later stages and a final morphologic result at birth. Understanding the sequence of field deficiency has bidirectional consequences: it leads us further inward to work out the genetic sequences which produce the field, in the first place; and it leads us outward towards a more developmentally based and innovative therapeutics.
This chapter presents the rationale for a gross anatomy of the embryo based on a map of developmental fields in terms of their relationships to the central nervous system. Since the mid 1980s, breakthroughs in genetic mapping have led to a new model of neuroembryology, in which structure units of development, neuromeres, with identifiable anatomic boundaries based on a unique code of homeotic genes stretch forward and backward from the anterior terminus of the hindbrain at Hensen’s node to create the neuraxis. The genetic sequence in the midline responsible for this definition also acts to define at the time of gastrulation a series of exits sites for migrating cells wherein they acquire a homeotic identify in register with the CNS, a registry which is faithfully reproduced in their innervation and function. Subsequently, cells are grouped into intermediate structures, pharyngeal arches, each with its only genetic x–y–z coordinates defining distinct zones. Cells enter the arches much like marbles falling into a pocket. As they do so, they acquire an identity based on those coordinates. Developmental fields result, each with a distinct neurovascular supply. Each is repositioned at a unique time and place to form tissues in a coherent anatomy design. The learning objective of this chapter is to gain a familiarity with the neuromeric model and to appreciate how it can be used to construct the true gross anatomy of the embryo.
Access provided by Autonomous University of Puebla. Download chapter PDF
Similar content being viewed by others
Keywords
- Gastrulation
- Neurulation
- Somitomere
- Ectomere
- Prosencephalon
- Mesencephalon
- Rhombencephalon
- Somite
- Pharyngeal arch
- Homeotic genes
- Developmental field
- Neural crest
- Piriform fossa
- Cleft lip
Introduction
Developmental anatomy is the bedrock upon which all treatment methods for cleft and craniofacial anomalies must be based. Traditionally, facial development has been considered independent of brain development, but recent advances in molecular genetics demonstrate a more intimate neuroembryological relation than was previously appreciated. The well-known dictum of DeMeyer in 1964, “the face predicts the brain,” can be inverted to “the brain predicts the face,” as mechanisms of induction are now better understood.
Many important insights into neuroembryology can be deduced from craniofacial anomalies and the results obtained by surgical intervention for their correction. Why should a closer collaboration between craniofacial surgeons and neurologists be of mutual benefit? Most facial anomalies represent defects in specific developmental fields. The success or failure of surgical manipulations permits a more accurate understanding of just exactly where these fields exist and how they behave.
When a deficient developmental field is released from normal surrounding fields, subsequent facial growth can be anticipated to be more normal. By the same token, persistent patterns of relapse after surgery point an accusatory finger at the site of the pathology. Nowhere is this more apparent than in the treatment of routine clefts of the lip and palate. Despite advances in presurgical orthodontics and operative techniques, we continue to be faced with results that deteriorate over time; most of our patients requiring multiple secondary interventions. Such a model cannot be correct. As an anatomist, I have always struck with by the inability of descriptive embryology to answer very fundamental questions about clefts, unaddressed even in the most contemporary of texts. First, if the mechanisms of unilateral and bilateral cleft formation are the same, why are the surgical approaches so different? Second, what explains the spectrum of severity of clefts? Third, what relation exists between cleft palate and cleft lip? Fourth, why are isolated palatal clefts more likely to be associated with additional birth defects? Finally, if isolated genes are to blame for clefts, why don’t we find evidence of such mis-expression all over the body; how is it possible for genetic mis-expression to be unilateral?
The purpose of this chapter is to introduce a new and clinically relevant model of craniofacial pathogenesis based on concepts of developmental biology and neuroembryology that are as yet little known in medicine. Much of the relevant literature is less than 20 years old. Many key discoveries have come from technologies as yet in evolution. All clinicians dealing with congenital anomalies are witnesses to nature’s variations. These pathologies may well constitute the great “Rosetta stone” of developmental biology; on the faces of our patients are written the hieroglyphics of embryology. Properly translated these can create, for the first time, a new “gross anatomy of the embryo.” Craniofacial surgeons and neurologists have an indispensable role to play in the discovery of this knowledge…for those who confront the experimentations of nature are those closest to her secrets.
Our central tenet is this. All craniofacial tisues, be they mesenchyme (neural crest and paraxial mesoderm), ectoderm (skin), or endoderm (mucosa), originate at specific sites of the embryo during gastrulation. These sites correspond to the embryonic segmentation units of the central nervous system from which their innervation proceeds. These developmental units are called neuromeres, the anatomic boundaries of which are defined by a series of position-specifying homeotic genes. Thus, if we know the innervation of a structure, we can deduce where its component tissues were originally produced. Many important inferences arise from this system. Applications of neuromeric anatomy provide a potential embryonic “map” of all craniofacial structures with important implications for diagnosis and surgery. Let us consider two examples to see how this works.
Exclusive of the cranial base (basisphenoid and posterior) and parietal bone, the craniofacial skeleton is made exclusively from neural crest. Thus, the cell populations producing the ethmoid and presphenoid all originate from the neural folds in genetic register with the first rhombomere (abbreviated r1). While they are still in residence within the neural folds, the identity of these neural crest cells (NCC) is generalized at first. When the NCCs migrate from the fold, they are simply destined for the basal and frontal part of the forebrain. But upon receiving additional signals from the local environment, they become further specified as tissues of anterior cranial base and frontonasal complex.
In similar fashion, the neural crest cells of the neural folds in register with rhombomeres 2 and 3 populate the rostral and caudal aspects, respectively, of the first arch. Here they are further organized by a set of distal-less, Dlx genes into regional identifiers. We shall discuss this process further along. Neurovascular supply develops within the pharyngeal arches; this network defines the future developmental fields. When the process of embryonic folding positions the first arch with respect to the future face, each population supplied by a branch of the V2 stapedial artery (an addition onto the internal maxillary artery) will come into contact with an epithelial program and form a soft tissue-bone complex. This is reflected in the spatial order of the neurovascular pedicles, each accompanied by a sensory branch of V2 that fans out from the pterygopalatine fossa to their respective destination. Most distal along of these is the medial nasopalatine axis giving rise to the premaxilla and vomer. Just behind it, the lateral nasopalatine and descending palatine axes supply the inferior turbinate and palatine bones, respectively. Maxilla and zygoma are next being fed from the anterior superior alveolar and zygomatic axes. The squamous temporal, mandible, malleus, and incus are r3 neural crest bones and have their own neurovascular supply. All of these fields develop in accordance with the extracranial stapedial system as it forms a hybrid with internal maxillary of the external carotid. This vascular anatomy will be discussed in detail in Chaps. 6 and 7.
Disturbances at a particular neuromeric level can affect individual or multiple fields, causing to be deficient or absent. Thus, isolated cleft of the secondary hard palate (unassociated with cleft lip) represents a deficiency state of the vomer field. This occurs as a spectrum. As the vomer is progressively smaller, it lifts away from the plane of the palatal shelves and the cleft extends forward toward the incisive foramen. This is intrinsic genetic specification of the vascular system which is clinically relevant. In Treacher-Collins syndrome, multiple r2 and r3 developmental fields of the midface are affected: the maxilla, zygoma, and the mandible are all small. Each of these has some degree of compromise in its neurovascular axis. Vomer and premaxilla are unaffected as their medial nasopalatine axis remains intact (Figs. 1.1 and 1.2).
Developmental fields form in a specific spatiotemporal sequence. Each one builds upon its predecessors. Making a face is much akin to assembling a house with magical pieces of Lego®, each one of which will grow over time. Imagine a Lego house made from 20 pieces (4 on the floor and 5 stories high). All pieces are growing independently. If a cornerstone piece is removed, the 19 remaining pieces undergo a deformation and the house tilts into the deficiency site. The missing Lego® piece in cleft lip is the premaxilla. Surgical reconstruction of this field is the key to repair of the common cleft.
We shall explore these concepts further during the course of this chapter.
Definitions
-
Anterior visceral endoderm (AVE): Described in mammals; the AVE constitutes the inducer of the forebrain.
-
Chondral neurocranium: Skull dedicated to protection of the brain and formed in cartilage. (1) Axial cranial base = ethmoid, presphenoid, basisphenoid, basioccipital, exoccipital, and supraoccipital. The first two bones are formed by neural crest. The basisphenoid comes from somitomere 1. The remaining bones are formed from the sclerotomes of somites 1–5 (the occipital somites). (2) Lateral cranial base = temporal bone (petrous and mastoid fields). These bones are non-sclerotomal and originate from somitomeres 6 and 7.
-
Chondral viscerocranium: Skull dedicated to supporting the pharyngeal arches. Larynx may represent a transformation of previous visceral cartilages.
-
Frontonasal prominence (obsolete): Neural crest of the caudal prosencephalic neural folds (prosomeric levels p1, p2, p3, and p4) migrates forward to populate the rostral prosencephalic neural folds (prosomeric levels p5 and p6). Note that these rostral folds are “sterile,” i.e., they contain neither cells of the CNS nor neural crest. For this reason, they are referred to as nonneural ectoderm (NNE). NNE gives rise to the optic, olfactory, and adenohypophyseal placodes, as well as to the epidermis of the forehead, nose, and vestibular lining. The arrival of prosencephalic neural crest beneath the NNE results in creation of a dermis for the forehead (p4 zone), the external nose (p5 zone), and the internal nose (p6 zone). As the frontal lobes grow, the skin envelope is pushed forward and medially on both sides of the midline. The resulting configuration is a p4 forehead, a p5 orbital roof, and a p6 ethmoid-nasal complex. These are bilateral structures connected in the midline. There are two p5 optic placodes and two p6 nasal placodes. Initially, both fronto-orbital-nasal masses are widely separated. The initial orbital angle is about 180°. A combination of brain growth, embryonic folding, and the differential growth of the PNC mesenchyme (p6 < p5 < p4) results in a dramatic apoptosis of the midline MNC sphenethmoid complex and causes a narrowing of the orbital angle to its final configuration of 90°.
-
Intermediate mesoderm (IM): This mesenchyme lies medial to LPM and lateral to PAM; unclear if a separate population or induced from LPM by medial genes; forms urogenital system; sequential formation of urinary organs in craniocaudal order via nephrotomes: pronephros, mesonephros, and metanephros. Origin of pronephros is at fourth cervical neuromere.
-
Lateral plate mesoderm (LPM): First wave of mesoderm at gastrulation migrates to most distal position of the trilaminar disc; has two potential layers, an inner visceral layer (LPMv) and an outer somatic layer (LPMs). These are split by intraembryonic coelom into two layers. LPMv forms cardiac mesoderm and much of peripheral vasculature and all gut musculature below neck. Unsplit LPM in neck forms cervical esophagus. Neuromeric coding of trunk LPMv may follow the sympathetic nervous system and ventral unpaired branches of the aorta. LPMs forms the fascial envelope of the trunk and all bones of the extremities including the clavicle, the scapula, and the pelvic girdle. LPMs and PAM share common hox genes and segmental innervation.
-
Membranous neurocranium: Skull dedicated to protection of the brain and formed in membrane. All bones except parietal are exclusively neural crest. Parietal receives mesenchyme from somitomere 4; includes the bones of the nasal and orbital cavities.
-
Membranous viscerocranium: Skull dedicated to supporting the pharyngeal arches and formed in membrane—includes lisphenoid, inferior turbinate, maxilla, palatine, zygoma, mandible, and ear bones.
-
Neuromere: Individual developmental zone of the embryonic neural tube. Boundaries defined by a “barcode” of overlapping zones of gene expression. Each neuromere provides spatial segmentation to mesoderm existing outside the neural tube at that level. All neural crest, paraxial mesoderm, and somatic lateral plate mesoderm originating from a given neuromeric level will bear the same Hox code.
-
Paraxial mesoderm (PAM): Final wave of mesoderm to exit the primitive streak; remains closest to neuraxis and is induced by genes from the notochord and neural tube. Because it contains angioblasts, it immediately forms the primitive perineural plexus. Most anterior PAM forms the dorsal aorta, the muscles of the pharyngeal arches, the cranial base caudal to the pituitary, the pharyngeal arch muscles, and the. Derivatives of PAM are discussed in Chap. 2.
-
Placode: Specialized areas of epithelium seen on the surface of the embryo in close association with the brain. (1) Required by special somatic afferent systems of smell, sight, hearing/balance. Contain neurons with eventual connection to the CNS. (2) Surrounding all placodes are zones of tissue with functional importance for the eventual development of a sensory unit. (3) All placodes remodel and “sink” into the embryonic head; in this manner they internalize functional tissue. (4) Once internalized placodes undergo structural alteration (p6 adenohypophyseal = anterior pituitary, p6 nasal = olfactory bulbs, p5 optic = lens, r2–r3 trigeminal = gasserion ganglion, r5–r6 facio-acoustic = cochlea and vestibular apparatus, r7–r11 statoacoustic placodes (in fishes) produce ventral nuclei of taste ganglia). (5) In all cases, specialized tissue of the sense organ (placode) must make contact with cranial nerve structures in order for perception to occur.
-
Prechordal plate mesoendoderm (PCM): First cells to exist from Hensen’s node form first an endoderm and then a mesoderm. This zone lies directly beneath the future midbrain and forebrain; may form the extraocular muscles innervated by the rostral nucleus of cranial nerve III (inferior rectus, medial rectus, and inferior oblique). In avian model, PCM is a potent inducer of forebrain.
-
Somite: Organizational unit of paraxial mesoderm, each is derived from a precursor somitomere. The first somite arises from Sm8. The units are completely separated by epithelium and roughly cubic in structure. Each somite has three laminar derivatives: (1) an internal lamina, the sclerotome, forms the vertebrae and ribs; (2) an external lamina, the dermatome, forms those dermal structures associated with each neuromeres; (3) an intermediate lamina, the myotome, is divided into a dorsomedial epaxial zone and a ventrolateral hypaxial zone.
-
Somitomere: Initial and transient form of segmentation of the paraxial mesoderm corresponding to a given neuromeric level. Somitomeres are incompletely segmented. Each contains level-specific paraxial mesoderm that has exited from the primitive streak during gastrulation. Somitomeres undergo secondary transformation to derivative structures. The walls of the dorsal aortae are formed from the entire length of somitomeres. They provide the mesenchyme for striated muscles and contribute to the cranial base of the middle cranial fossa and petrous complex. In the occipital regions, Sm8–Sm11 contributes for the muscles of the larynx and pharynx prior to formal transition to somites. Beginning with the Sm8 and continuing through Sm11, each somitomere forms first a pharyngeal arch and the remaining mesoderm is transformed into an occipital somite. All somitomeres from 12 onward are transformed exclusively into somites.
Zoologic Abbreviations
-
Alisphenoid (greater wing of sphenoid): AS
-
Basisphenoid: BS
-
Basioccipital: BO
-
Ear bones: Malleus (M), Incus (I), Stapes (Sp)
-
Epiotic: EpO
-
Exoccipital: EO
-
Frontal: PrF (prefrontal medial), PtF (postfrontal)
-
Lateral pterygoid plate: LPt (in continuity with AS)
-
Maxilla: Mx1 (incisor/canine), Mx2 lateral (premolars), Mx3 (molars)
-
Mandible: Mn1 (incisor/canine), Mn2 (premolars), Mn3 (molars), MnR (ramus)
-
Medial Pterygoid plate: Mpt (in continuity with the orbitosphenoid)
-
Palatine: Pl
-
Parietal: P
-
Premaxilla: PMx medial (central incisor), PMx lateral (lateral incisor), PMxF (frontal process)
-
Presphenoid: PS
-
Prootic: PrO
-
Pterygoid: Pt
-
Opisthotic: Op
-
Orbitosphenoid: OS
-
Supraoccipital: SOc (chondral), SOm (membranous)
-
Temporal: Tm (mastoid), Tp (petrous), Ts (squamous), Tt (tympanic)
-
Zygoma: PO (postorbital), J (jugal)
Developmental Fields: Lessons from the Common Labiomaxillary Cleft
The 4-D Theory of Cleft Formation
The pathologic anatomy of the labiomaxillary cleft is a four-dimensional problem. The principles of its surgical management must be conceptualized in the same manner. How the cleft site appears in the newborn is very different from its anatomy in utero. Indeed, the initiation of the cleft problem may occur as early in embryogenesis as gastrulation (the process by which the germ layers of the embryo are established) at 15–18 days gestation [1, 2]. At the time of initiation of the cleft site, four pathologic processes are unleashed; these processes exert their effects in a strict sequential order [3]. A spectrum of presentations is thereby produced ranging from the “cleft-lip nose” with an apparently normal lip, to a full-blown [4] (Figs. 1.3, 1.4, and 1.5).
The nature of this pathologic sequence has been identified [3,4,5,6,7,8,9,10]. First, a deficiency state exists in the functional matrix responsible for synthesizing the frontal process of the premaxilla that forms the piriform margin. Within this abnormal developmental field, insufficient bone volume results. This causes a stereotypical displacement pattern of the soft tissue envelopment on both sides of the cleft. If the deficiency state is significant enough, it will affect the ability of adjacent developmental fields to perform soft tissue closure of the nostril floor and lip. The resulting division further aggravates tissue displacement. Over time, the effects of deficiency, displacement, and division create a distortion of the soft tissue envelope. This results in an abnormal anatomy of the septum. Ongoing growth of the osteocartilagenous nasal vault, uncoupling of normal relationships between the skeletal elements, and aberrant force vectors exerted by the perioral musculature result in the characteristic “opening-up” of the cleft site so elegantly described by Delaire [11,12,13,14,15,16,17,18].
The above concepts of cleft formation are known as “4 dimensional theory.” These can be summarized by the pneumonic of “the 4 D’s,” each one of which corresponds to a dimension. Deficiency is axial. Displacement is coronal. Division is sagittal. Deficiency is temporal. Interestingly enough, the order of these processes follows the order of axis specification in the embryo: anteroposterior, then mediolateral, and finally right-left.
The Biologic Significance of Relapse
All surgeons involved with cleft care know full well the frustration of seeing well-executed repairs in infancy degenerate into a predictable sequence of secondary deformities requiring further correction. Even in the best of hands, reoperation rates may reach as high as 85% [19]. What exists here is not failure of technique, but an inadequate biologic model of the problem in the first place. If the pathologic anatomy of the cleft site hinges on a deficiency state in a specific developmental field, and if the surgical correction of the cleft does not include reconstitution of that defective field such that it will grow normally over time, and such that it will cease to perturb the growth of its neighboring fields, then all forms of cleft surgery are condemned over time to varying degrees of relapse!
Are there any grounds for optimism? If secondary deformities constitute a type of plastic surgery “lemon,” is it not possible to make of them some form of “lemonade?” The answer to this question is overwhelmingly positive because, in pediatric craniofacial surgery, all patterns of relapse point an unequivocal accusatory finger directly at the original pathology. Relapse is nothing more than the manifestation over time of a deficient developmental field. Knowledge of the embryology of the face conceived in terms of specific fields (zones of soft tissue and the bone that they produce) will enable surgeons to conceive new surgical approaches based, not on geometry, but upon developmental anatomy. Cut-and-paste tissue manipulation will be supplanted by biologic principles of field reconstitution and reassignment.
Beyond Descriptive Embryology: Developmental Fields and the Functional Matrix
It may come as a surprise to many readers that the drawings of cleft lip pathogenesis depicted in the most recent of texts reflect an understanding of facial embryogenesis that is nearly century and a half out-of-date. Descriptive embryology as a science began from pioneering observations by Wilhelm His in the 1870s using light microscopy and histologic staining [20, 21]. The approach was morphologic rather than cellular. (To the end of his career, His vigorously opposed the idea that genetic information could be contained in the nucleus.) Terms introduced by His such as “lateral nasal process” can be found in all textbooks. Yet who among us can define just what a “process” is? What are its constituent parts and from where in the embryo do they come? All surgeons are well aware that clefts, for example, occur in a comprehensible spectrum of presentations. Unfortunately, concepts such as “failure of fusion” or “failure of mesoderm penetration” are incapable of providing a rational explanation for the varying degrees of pathology. For most of us, embryology seems a mere jumble of terms with no clinical relevance. But the iron fact of the matter is this: without a detailed understanding of the developmental anatomy of the face based on modern developmental biology, genetics, comparative anatomy, and neuroembryology, pediatric plastic surgery is a collection of techniques in search of a science (Fig. 1.6 Wilhelm His).
Based upon clinical observations of secondary cleft patterns, this author arrived at the following hypotheses: (1) unidentified developmental fields might constitute the “building blocks” of the face; (2) a mesenchymal deficiency state in such a field would be characterized by an inadequate osteosynthetic capability; (3) an osseous deficiency state exists in the inferolateral piriform fossa/lateral nasal wall of cleft patients; (4) such a functional matrix deficiency state might account for the relapse pattern observed in cleft patients after primary repair.
Neurovascular Mapping of Developmental Fields: Nasomaxillary Model
To test these ideas, it seemed logical as a first approximation to study the relative contributions of the internal carotid artery (ICA) and external carotid artery (ECA) circulations to the skin and epithelium of the nasal fossa. Contrast injections into isolated internal carotid arteries in a series of aborted fetuses were performed. These results were first reported at the Plastic Surgery Educational Foundation awards program at the 2000 annual meeting in Los Angeles of the American Society of Plastic Surgeons [22] (Figs. 1.7 and 1.8).
The author found, to his surprise, that the upper border of the inferior turbinate combined with the skin/mucosa junction of the inferolateral piriform fossa just anterior to the inferior turbinate constituted a potential field interface zone. At this site, three distinct biologic systems (vascularization, innervation, and genetic programming) functioned in precisely the same manner. (1) The internal carotid supplied the mucosa (but not skin) of the lateral nasal wall, but only as far as the upper border of the inferior turbinate. The mucosa beneath the turbinate and the skin margin along the infracartilagenous nostril were un-perfused. (2) The sensory innervation followed the exact same distribution pattern! The epithelium supplied by the ICA corresponded to sensory supply from the first branch of the trigeminal, while that supplied by the ECA was innervated by V2. (3) The inferolateral piriform fossa also represents an interface zone between three entirely different developmental zones of the embryonic neural crest!
The clinical relevance of this model can be seen in developmental anatomy of the piriform fossa (Figs. 1.7 and 1.8). The lateral wall of the nasal cavity has an upper zone populated by prosencephalic neural crest (PNC), innervated by V1, and irrigated by ICA. The lower zone of the lateral nasal wall is populated by rhombencephalic neural crest (RNC), innervated by V2, and irrigated by ECA. The “breakpoint” between these two zones is the inferior turbinate. This outer lamina of the piriform margin belongs to the ascending process of the maxilla. The medial wall of the nasal cavity has an upper zone populated by PNC, innervated by V1, and irrigated by the ICA. This zone consists of the perpendicular plate of the ethmoid and the septum. The lower zone of the medial nasal wall is populated by mesencephalic neural crest (MNC), innervated by V2 and irrigated by the ECA. This zone contains the vomer and the premaxilla.
Details of the experimental data and their many implications appeared in print as a supplement to the Journal of Craniofacial Surgery in February 2002 [9]. The purpose of this publication was to describe a preliminary, but clinically relevant, “map” of developmental fields as they applied to the facial midline. Details of the medial nasal wall fit a model quite analogous to that of the lateral nasal wall. In this model, the septum and perpendicular plate of the ethmoid form a sharp developmental boundary with the vomer and premaxilla characterized by blood supply (ICA vs. ECA), innervation (V1 vs. V2), and neural crest origin (PNC vs. MNC/RNC). As we shall see, this model will permit us to assemble an accurate picture of how the premaxilla and maxilla develop and interface with one another. Isolated deficits in components of this system explain the pathologic anatomy of all forms of clefts.
The premaxillary developmental field has three subdivisions: central incisor, lateral incisor, and a vertical ascending process. Formation of an intact alveolar arch involves fusion between the premaxilla and the maxilla. This brings the two ascending processes (premaxilla and maxilla) into apposition and fusion. The piriform margin is therefore bicortical; it therefore serves as a buttress for reconstruction. Loss of the medial (premaxillary) ascending process leads to a scooped-out piriform fossa and is the cause of the cleft nasal deformity (Tessier #2 cleft). Loss of the lateral (maxillary) ascending process creates a cleft extending to the medial inferior orbital rim, but leaving the piriform fossa intact (Tessier #4 cleft). More extensive neural crest deficiency in the premaxillary field leads to loss of alveolar bone in a labiolingual gradient. This creates a cleft of the primary palate and, frequently, loss of the lateral incisor (Figs. 1.9 and 1.10).
Although this model description based on classifying tissues as ICA and ECA derivatives was immediately useful for mapping out the relative roles of premaxilla, vomer, and maxilla, in cleft formation, the vascular embryology behind it proved to be too simplistic. The arterial supply for these fields, in point of fact, represents a new system that arose with the transition between agnathic (jawless) fishes and gnathostomes (fishes with jaws). The process of converting the first and second gill arches to maxilla, mandible, and supporting hyoid bones required reassignment of neural crest from the gills to a new location and shape. This in turn required a disruption in the arterial supply to first and second arches and the creation of a new system.
The stapedial innovation arises from a remnant of the second aortic arch artery, the hyoid artery. This original stapedial stem passes through the stapes within the developing tympanic cavity and into intracranial and extracranial divisions. These are guided into position initially by the branches of cranial nerve VII and later by the branches of the trigeminal nerve. The intracranial division ultimately reaches the orbit where it connects with ophthalmic to form all arteries supplying the non-ocular tissues of the orbit and face. The extracranial division joins the distal-most branch of external carotid to form another hybrid system, the maxillo-mandibular complex. Arteries follow sensory branches of V2 and V3 to supply the neural crest structures of the jaws. We shall discuss this system at length at various junctures through the remainder of this book (Figs. 1.11 and 1.12).
Neural crest cells refer to those cells that arise during embryogenesis from a border zone between the more lateral zone ectoderm responsible for forming skin and the more axial zone of ectoderm responsible for forming the brain and spinal cord. Neural crest cells migrate widely and form many structures such as dermis, bone, and cartilage usually associated with mesoderm. These cells also form components of the nervous system such as Schwann cells and autonomic ganglia. For this reason, neural crest is often referred to as an ectomesenchyme. The great extent of its derivatives has often led authors to refer to it as “the fourth germ layer.” It should be born in mind that neural crest cells do not make their appearance until well after gastrulation is complete. The three traditional germ layers all are recognized at the time of gastrulation.
The behavior of neural crest cells in terms of their migration pathways and derivatives stems largely from what part of the neural folds they originate from. The names of these neural crest zones correspond to the original three parts of the developing central nervous system (learned by most of us and promptly forgotten). These are: the prosencephalon (forebrain), mesencephalon (midbrain), and rhombencephalon (hindbrain) (Fig. 1.13 Embryonic brain).
Neuromeres and Neuromeric Coding
Neuromeres: The Clinical Significance of the Neuromeric Map
All clinicians are familiar with the concept that certain cells born in the neural plates at the exact boundary between neural ectoderm and nonneural ectoderms are termed neural crest and that these cells have a very important role to play in development of the head and neck. Our model here is one in which neural crest can be “mapped” to very precise zones of origin. Each such zone corresponds to a developmental unit of the neural plate. These segmented units, called neuromeres, are distributed in a transverse fashion along the entire neural axis of the embryo [23,24,25,26,27]. The nervous system is thus divided into transverse developmental units just like the body of an earthworm. Each such segment has a genetic definition. Certain genes or combinations of genes express products only in a particular zone. The anatomic extent of these protein products constitutes the domain of the neuromere. Each neuromere has a certain neuroanatomic content characterized by nuclei and ascending or descending tracts. The neural crest sitting just outside the neural tube in the domain of a given neuromere will express the same defining set of proteins as those cells within the neural tube. Furthermore, neural crest cells from a given neuromeric level will supply certain zones of ectoderm and mesoderm. This model allows us to see the nervous system as the master integrative agent of development. In our discussion, it will be necessary to introduce a number of terms and concepts from neuroembryology (Fig. 1.14).
Before plunging into further details, let’s summarize the highlights of the neuromeric model, so we can stay oriented. The CNS of all vertebrates is divided into three classes of neuromeres [28,29,30,31,32]. The forebrain is formed from six prosomeres (formerly considered to be four in number). From caudal to cranial, these are numbered p1 to p6. Prosomeres p1–p3 belong to the diencephalon. They are subdivided into two tiers, dorsal (alar) and ventral (basal). The telencephalon is formed from prosomeres p4 to p6. The basal tier of p6 has much to do with the olfactory system, while basal p5 is associated with the visual apparatus.
Puelles and Rubenstein propose the midbrain to be constructed from two mesomeres m1 and m2. These contain, respectively, the superior and inferior colliculi. (An anatomical boundary between the two has not been demonstrated as it is in the borders between the rhombomeres.) In this model, the hindbrain is described as being made up from 12 well-defined rhombomeres r0 to r11. An alternative viewpoint held by Sarnat considers r0 to be the principle neuromere of the midbrain and r1 the neuromere of the isthmic region (metencephalon from which develops the pons and cerebellar cortex [33].
Neural crest from four neuromeres (m1–m2 and r0–r1) behave clinically as a midbrain neural crest. MNC is involved with the construction of the orbit. The midbrain per se has two mesomeres, m1 being rostral and much larger, while m2 (also called the preisthmus) abuts the isthums, a transition zone between midbrain and hindbrain.
Initially, the most rostral rhombomere of the hindbrain is r1, but a small zone emerges at its most anterior aspect that interfaces with m2. This zone corresponds to the isthmus itself and is termed r0. The remaining hindbrain neuromeres r2–r5 are dedicated to the metencephalon (pons) and r7–r11 for the medulla.
Neural crest originating from neural folds associated with rhombomeres r2 to r11 supplies the pharyngeal arch system. When the neural crest cells migrate, they swarm over the surface of the mesoderm lying just outside the neural tube. This mesoderm is called paraxial mesoderm (PAM) and is segmented in direct register with the neuromeric system. Each segment of PAM is called a somitomere (Sm) and is shaped like a ball. The first seven somitomeres (corresponding to r1–r7) are incompletely separated. All somitomeres from Sm8 caudally undergo anatomic rearrangement into somites. Thus, Sm8–Sm11 form the four occipital somites. Sm12 becomes the first cervical somite. Developmental biologists refer to mesoderm from Sm1 to Sm7 as cephalic mesoderm.
The mesenchyme of pharyngeal arch consists of neural crest from two rhombomeres plus PAM from a somitomere. The original number of pharyngeal arches in primitive aquatic vertebrates was seven. With the tetrapod transition to a land-based existence, this number was reduced to five, the first two being dedicated to the jaws and oral cavity and the last three to formation of the pharynx. We shall examine how neural crest and mesoderm are mapped out according to the neuromeric system.
Craniofacial Neural Crest
Neural crest migration is a physical process in which the cells move along pathways of least resistance (anatomic cleavage planes) or molecular “guide wires” such as vimentin. Neural crest cells swarm over the surface of PAM somitomeres like taffy poured over an apple. Neural crest cells migrate following three pathways. (1) They can move laterally outward in the plane between the nonneural ectoderm/endoderm and the somitomeres. (2) They remain interposed between the neural tube and the somitomeres. (3) They migrate ventral to the neural tube and then travel caudally stopping along their way to form the sympathetic chain (Fig. 1.15).
Craniofacial neural crest is organized according to the original three-part embryonic brain, each population of which has distinctive migration patterns. Rhombencephalic (hindbrain) neural crest is divided into 12 rhombomeres (r0–r11), the last 10 of which migrate segmentally into the pharyngeal arches and over the face and head. Mesencephalic (midbrain) neural crest populations arise from two mesomeres (m1 and m2) and from the most rostral hindbrain (r0–r1). It migrates as streams into the anterior cranial base, orbit and frontonasal zones. Prosencephalic neural crest is produced by the caudal three (of six) prosomeres (p1–p3). It migrates as a glacier-like sheet forward to populate the frontonasal skin (Fig. 1.16).
Fate mapping experiments show that the r1 neuromere is subdivided into a rostral zone that forms the cerebellar cortex and a caudal zone giving rise to the deep cerebellar nuclei. Neural folds corresponding to the caudal region of r1 neural plate give rise to neural crest with a unique fate, one with direct relevance to the formation of cleft lip and cleft palate. Rapid growth of the head causes the embryo to fold. This cephalic flexure forces the future eye to lie ventral to somitomeres 1–5. The eye has already been coated with midbrain neural crest which creates biologic zones of scleral. Thus, the extraocular myoblasts have a direct pathway to access the globe where they are organized by the r1 fascia and attach according to a strict spatiotemporal sequence.
The cephalic fate of midbrain neural crest includes formation of extraocular muscle fascia, all dura innervated by V1, and the presphenoid bone and the frontosphenethmoid complex. All these structures are supplied by the internal carotid artery via the stapedial system, the individual branches of which are programmed by V1. In the first pharyngeal arch r2 neural crest derivatives. These are supplied by the external carotid via branches of V2 stapedial.
Discovery of the rhombomeres came first, dating to the late nineteenth century, but their properties were not understood until the mid 1980s. Furthermore, the exact number of rhombomeres was uncertain. Their anatomic role at the most caudal medulla was unclear. Puelles in 2001 published work in the avian model demonstrating the existence of “pseudorhombomeres.” Previously, it was thought that the rhombomeric series terminated at r8 that included the entire spinal cord. Puelles’ work showed that the final number of rhombomeres was 12 [34, 35]. The existence of neuromeres in the more rostral CNS required advancements in gene mapping. These were not reported until 1993. Over the ensuing decade, further investigation of the neuromeric system has proceeded at a frenetic pace at neuroscience laboratories around the world. In this paper, we shall be using the Puelles and Rubenstein model as in its latest iteration (2015).
At this juncture, we must take note of a caveat that applies to concepts about prosomeres held widely within the scientific community. Many descriptions of brain anatomy extent in the literature assign four neuromeres to the forebrain, two for the telencephalon and two for the diencephalon. These are known respectively as T2/T1 and D2/D1. This nomenclature is elegantly presented in the brain development section of texts such as that of O’Rahilly and Muller [36,37,38]. Based on sophisticated gene mapping techniques previously unavailable, the Rubenstein-Puelles model has by no means been universally incorporated into the thinking of neurologists, neuropathologists, and researchers [39, 40]. However, from the standpoint of craniofacial surgery, the R-P model constitutes an extremely sensitive instrument with which to analyze patterns of deformity. Hence, we shall make use of this terminology in our discussion.
The clinical significance of the neuromeric model is that it enables us to map out the anatomic site of origin for all zones of ectoderm and mesoderm supplied by a given zone of the nervous system. The role of neural crest populations in those zones, specifically what structures they make, can be understood as well on the basis of their neuromere of origin. The premaxilla, for example, could develop from a precursor cell population of RNC along the neural fold corresponding to the second rhombomere. When these cells migrate into the first arch, those that populate the anterior and medial sector of the arch will be positioned such that they will come into contact with an spatially specific “program” in the epithelium which will direct them to form the bone fields. This sector will be specifically supplied by the medial nasopalatine branch of the arterial axis. A deficiency state in the population (be it of inadequate cell number, defective migration, abnormal post-migratory rates of mitosis, or cell death) will lead to a small or absent premaxilla. Furthermore, if the premaxillary MNC has several subsets, aligned in craniocaudal order along the neural fold (ascending process, lateral incisor, and central incisor), then the spectrum of deficiency states seen in the premaxilla of cleft patients can be understood as progressively greater degrees of disturbance in the premaxillary MNC precursor population. But let’s not get ahead of ourselves…to understand this system, we must first understand the scientific origins of the neuromeric map.
Craniofacial Mesoderm: Extraembryonic Versus Intraembryonic
The Little Appreciated Hypoblast
Formation of the hypoblast and preparation for gastrulation takes place during stage 4 (days 5–6). The process begins when cells of the blastoderm absorb water and then secrete it, causing a separation of blastomerm from the yolk, the subgerminal space. In this process, the deep cells of the blastoderm die creating a monolayer epiblast (light gray). This is known as zona pellucida. At the posterior pole of chick blastoderm, the deep cells persist, creating a thickening known as the zona opaca. This posterior marginal zone (PMZ) is thus bilaminar, having both primitive ectoderm (dark gray) and primitive endoderm (green). The anterior border of PMZ appears like a C-shaped sickle and is termed Koller’s sickle. Some of the deep cells (green) persist and are physically interposed between the PMZ and the yolk. These cells will give rise to the secondary hypoblast. The C-shaped configuration of the PMZ between the superficial cells and the deep cells is known as Koller’s sickle. It is from this point, during stage 6, that the primitive will originate (Fig. 1.17).
Later in stage 4, two events take place simultaneously: (1) cells of epiblast become transiently less adherent to each other. This “leaky” state permits epiblast cells of the zone pellucida to delaminate and fall into the subgerminal space, where they will link up again to form the primary hypoblast (light gray). (2) Cells below the zona opaca, the posterior marginal zone, proliferate forward beneath the epiblast, forming the secondary hypoblast (green) (Fig. 1.18).
By the end of stage 4, primary hypoblast (light gray) and secondary hypoblast (green) unite to form a single layer, thus creating the bilaminar embryo. In stage 5, hypoblast (primitive endoderm) will migrate outward to line the entire blastocoele and create the extraembryonic mesoderm that will be forming the placental circulation required to support the metabolic needs of the rapidly growing embryo during gastrulation. Note that at stage 6 primitive streak is initiated at Koller’s sickle and develops forward as far as the future anterior terminus of the notochord, i.e., at future neuromeres r0–r1. In so doing, the length of primitive streak is likely determined by a nested homeotic code within the epiblast that controls how far forward the primitive streak is permitted to extend (Fig. 1.19).
Accepted dogma about axis formation holds that primitive streak is the responsible event for determination of anterior-posterior, dorsoventral, and mediolateral relationships. In point of fact, axis determination may occur much earlier, at stage 4. Why should cells of the original blastoderm become asymmetrical? Specifically, why should a deeper layer of cells persist at the future posterior pole of the conceptus? What relationship might this have with the development of a homeotic gradient leading to the ultimate induction of the neuromeric segmentation system?
Let’s assume for the moment that the entire blastoderm consists of homogeneous cells which are pluripotent, capable of expressing the entire library of homeotic genes. Then let’s divide this population into two groups, cells that express a molecular marker, the HNK1 epitope, exhibiting a temporal and spatial distribution that can be related closely to the morphogenesis of tissues involved in the establishment of the craniocaudal axis. Blastoderm cells “choose up sides” with some being HNK1+ and the others being HNK1−. We can represent the positive cells as squares and the negative cells as triangles. A unidentified factor favors the accumulation of HNK1+ at the posterior pole, making the system instantly nonrandom. It could be that cell death of HNK1+ takes place centrally, but that the accumulation simply reflects their enhanced survival posteriorly. Or perhaps, migration of HNK1+ takes place. In any case, a deep layer of cells persists and the overlying single-layer epiblast is now a nonrandom structure, with electrochemical gradients both the AP and mediolateral dimensions. Further research will clarify these issues.
As we move to stage 5, the epiblast cells are pseudostratified columnar with apical microvilli facing the amniotic cavity and a definite basal lamina. Laterally, the epiblast cells of the embryo give way to extraembryonic epiblast cells that define the amnion. These cells are columnar. The reader is referred to Fig. 1.20 in which all tissues are color-coded (Fig. 1.20).
The stage 5 hypoblast spreads outward following the inner surface of the trophoblast and thus comes to line the entire cavity of the blastocyst. We can therefore distinguish two types of hypoblast, each of which has specific roles in development. The visceral hypoblast lies subjacent to the embryo proper. The cells are cuboidal. The surface they present to the blastocyst cavity is uniform with apical microvilli but no basal lamina. It is responsible for the formation of the midline primitive streak in stage 6, the signal event of gastrulation. Removal of hypoblast can result in multiple embryonic axes. In addition, visceral hypoblast is required for induction of the head region and for specification of germline cells. The parietal hypoblast forms the inner wall of the blastocyst and is histologically distinct, being composed of squamous cells. It quickly secretes a new layer of cells, the extraembryonic mesoblast, between it and the wall of the trophoblast. This results in a two-layer structure, the primary yolk sac.
Note that the term mesoblast is used until such time as embryonic tissues are in their correct final position. This layer is also called extraembryonic mesoderm.
Extraembryonic mesoblast proliferates wildly tracking upward along the confines of the trophoblast wall and therefore passing over the roof of the amniotic cavity. It is thus interposed beneath the cytotrophoblast of the future placenta. Recall that blood vessels are produced from mesoblast, wherever it is located. All tissues colored pink in Fig. 1.20 and angiogenic. At stage 5, small “fingers” of cytotrophoblast are inserted into the syncytiotrophoblast, but the primary villi do not have a mesodermal core.
Gastrulation begins at stage 6 (13–18 days). Extrembryonic mesoderm can be seen tracking around extraembryonic endoderm lining the yolk sac and extraembryonic ectoderm lining the amniotic cavity. Primitive streak develops at Koller’s sickle and extends forward about 2/3 of the length of the embryo. Cell migration occurs in two waves. At day 14, epiblast cells most proximal to primitive streak form definite intraembryonic endoderm, displacing hypoblast completely out from beneath the embryo. At day 16, ingression of intraembryonic mesoderm (red) takes place; it spreads outward and forward. By the end of stage 5, extraembryonic mesoderm has penetrated the overlying cytotrophoblast “fingers” to create a vascular core. This marks the beginning of maternal-embryonic circulation. Having made its contribution of cells to create the remainder of the embryo, the epiblast can rest on its laurels and is now called ectoderm proper.
Gastrulation takes place during embryonic stages 6 and 7 as the bilaminar embryo develops three germ layers. Gastrulation proceeds in cranial-caudal order. As soon as mesoderm is produced, it becomes organized in accordance with the neuromere with which it is in genetic register. This anatomy will be the subject of the subsequent chapter. For now, we shall follow the events that create developmental segmentation of paraxial mesoderm (Fig. 1.21 Gastrulation, Fig. 1.22 Erich Blechschmidt, and Fig. 1.23 Blechschmidt hypothesis).
Segmentation as a concept is highly intuitive. Repetitive functional units are observed in the centipede, in the multiple ribs of the snake, and in the dermatomes of the human thorax. All vertebrates possess the fundamental elements of a vertebral body, muscle units attached thereto, sensorimotor nerves arising at that same level to innervate those muscles, and well-defined geographic zones of skin. All the future spinal cord and all vertebrate embryos have primitive organizational blocks of mesoderm called somites [41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58]. Each somite is preceded by an earlier, incompletely segmented structure, called a somitomere (vide infra). These flank the neural tube from the cranial base to the tail. Each vertebra at a given neuromeric level is the combination of the caudal half-somite from the neuromeric level above immediately rostral to it and the cranial half-somite at that same neuromeric level. Human embryos possess 42–44 pairs of somites: 4 occipital somites contribute to the posterior brain case. These are followed by 8 cervical, 12 thoracic, 5 lumbar, 5 sacral, and 8–10 coccygeal somites. The rostral first cervical somite contributes to the foramen magnum and its caudal half to the atlas. For this reason, there are eight cervical nerves but only seven visible vertebrae.
Given the highly regular organization of the spinal cord, it seemed logical to anatomists of the nineteenth century that a similar pattern might exist in the CNS as well. Darwin’s concepts of evolution were bolstered by tremendous progress in paleontology and comparative anatomy. As the anatomic logic of vertebrate structure became defined, it was natural for scientists to look for commonalities in development as well. Neuroanatomists had noted the presence of small bulges on the surface of the embryo at the hindbrain region; these seemed to correlate in a regular way with cranial nerves. They also appeared to relate in some manner to the occipital somites and, rostrally, to the pharyngeal arches. These developmental zones of the rhombencephalon were termed rhombomeres, but their biologic rationale was uncertain. Furthermore, the physical presence of neuromeres appeared to die out at higher levels of the neuraxis.
Elaborate attempts were made by comparative anatomists to understand the organization of the head based on somites. In the 1930s, this culminated in the magnum opus of Goodrich based on fishes emphasizing dorsoventral relationships between mesoderm and cranial nerves [59]. This comparative approach was applied to the bones of the skull by Sir Gavin de Beer and later, in 1980, by Jarvik [60, 61]. Given the limitations of the data, progress in explaining the anatomy of the head and neck remained in a state of gridlock. As so often is the case in science, new technologies would be required before dramatic advances in knowledge could be made, and a new model, based on neuroembryology and genetics, would emerge.
The advent of scanning electron microscopy opened a new window on the morphology of development. Exhaustive work by Hinrichsen described the external development of the face [62]. Within the neural tube, SEM proved the existence of rhombomeres as segmental diverticula of the lateral wall. The boundaries of these neuromeres and their specific anatomic content were confirmed using contrast injections and immunofluorescence (Figs. 1.24 and 1.25) At the same time, Meier and Jacobson observed a somite-like organizational pattern of paraxial mesoderm (that mesoderm closest to the neural tube) in a wide variety of vertebrates [63,64,65,66,67,68,69,70]. In mammalian embryos, seven of these incompletely separated masses termed somitomeres were noted (Figs. 1.26 and 1.27). At the level of the eighth somitomere, a completely separate somite surrounded by an epithelial coat was noted. This is termed “the first occipital somite;” four such somites were documented rostral to the first cervical somite [71,72,73].
The anatomy of a somitomere consists of a whorl of PAM cells surrounding a central lumen. Though lacking the defined structures of a somite (dermatome, myotome, and sclerotome), the mesoderm of a somitomere has an intrinsic spatial orientation. Distal somitomeric PAM probably represents cells ingressing early at the neuromeric level during gastrulation, while PAM nearer the midline represents later-arriving cells. This spatial organization means that different PAM derivatives may come from distinct “compartments” of the somitomere. Each compartment may be characterized by the expression of different genes or by varying degrees of expression of the same gene. Alternatively, these differences may be translated into a specific migration sequence of PAM into its pharyngeal arch “target.”
Somitomeres provided a revolutionary new construct for the embryogenesis of the head. For the first time, an organization of mesoderm was observed that correlated with the previously described model of pharyngeal arches.
Somitomeres form in register with the rhombomeres of the hindbrain. As we shall see, the advent of genetic mapping to the neural plate provided a means to understand the anatomic relationships existing among the neuromeres of the medulla, the neural crest cells corresponding to each neuromeric level and the zones of outlying mesoderm innervate by those neuromeres and populated by those neural crest populations. The mesoderm of the human head was conceived as having 11 somitomeres divided into two distinct zones. The first 7 belong to what was termed head mesoderm. The next four somitomeres constitute the segmental plate. Segmental plate somitomeres go on to condense into somites while head somitomeres do not undergo a condensation. The transitional nature of the eighth somitomere is quite apparent.
The organization of paraxial mesoderm is tightly regulated pointing to the existence of a so-called “cellular clock”, with approximately four somitomeres being produced per day until all PAM is organized. Just at the time at which the 19th somitomere makes its appearance, the first somite transformation takes place at the eighth somitomere. A separation distance of 11 somitomeres between the last somite and the newest somitomere is maintained (Figs. 1.28 and 1.29).
A warning: readers who repair their textbooks of descriptive embryology and read about the transformation sequence of somitomeres to somites may be confused by the numbering system used in this paper. Many illustrations are based upon the avian 5-occipital somite model. Mammals have 4 occipital somites. Second, the number of rhombomeres described in most papers is only eight, with the eighth one being depicted as quite large. It remained for Puelles (vide infra) to describe the eighth rhombomere as broken up into four individual segments matching the cranial nerve anatomy and foramina of the skull; these he called “pseudorhombomeres.” With the accumulation of further evidence, the “pseudo” prefix has been dropped. Rhombomeres 8–11 are in direct genetic register with the occipital somites. Thus, readers studying over diagrams of somitomere-somite transformation will note the appearance of the first somite at Sm8 matched that of the 12th somitomere because a 4-occipital somite model is being used (Figs. 1.28 and 1.29 Carlson).
What happens in somitogenesis? A somitomere becomes a somite when it displays complete separation from its neighbors, when its outer layer of cells becomes an epithelium, and when it displays the three primary subunits of sclerotome, myotome, and dermatome. The first such transformation (that of the eighth somitomere to the first occipital somite) is externally difficult to recognize. The rostral half of the eighth somitomere is incompletely separated from the seventh somitomere. It behaves just like a somitomere. In contrast, the caudal half of the eighth somitomere undergoes a complete somitic transformation. The back end of the resulting first somite is thus completely separate from the rostral aspect of the second somite (ninth somitomere). Remaining somites are easily distinguished from one another (Fig. 1.30 Somite generation KW Tosney).
What are the unique characteristics of the 4 occipital somites? A brief digression is required here because recent cell labeling studies have conclusively demonstrated that somitogenesis begins at the eighth somitomere but in a subtle manner. The reader will recall that a typical somite is transformed from a cuboidal mass of mesoderm into bone, muscle, and dermis. In development, antecedents of these tissues within the somite occur as transient structures: the sclerotome, myotome, and dermatome. The sclerotome produces the chondral bone of the cranial base (basioccipital, exoccipital, and supraoccipital bones) and the vertebrae. The dorsal dermatome and dorsal myotome produce, respectively, the dermis and paraspinous muscles supplied by the posterior (dorsal) ramus of the spinal nerve. These are known as epaxial structures. The dermis of the remainder of the body wall and the extremities and their corresponding muscles are all hypaxial structures supplied by the anterior (ventral) ramus of the spinal nerve. The muscles arise from the ventral myotome. The dermis arises from lateral plate mesoderm.
In addition, the anatomic organization of occipital somites differs from that of all other somites. Although they have sclerotomes and ventral myotomes, they possess neither dorsal myotomes nor dermatomes. A dorsal myotome appears for the first time in the first cervical somite. This produces the muscles of suboccipital triangle. A true dermatome does not appear until the second cervical somite. For this reason, there is no skin innervated by C1. Although C1 is well-described in amphibians, in mammals it appears only transiently, providing neither sensory nor motor innervation in the mature state [74]. C1 is well-defined in amphibians. The dorsal branch of nerve C2 supplies the skin over the retroauricular skin, while the ventral branch of C2 supplies the mastoid and the rostral anterior triangle of the neck as part of the cervical plexus.
For these reasons, documentation of occipital somite anatomy requires a molecular, rather than a strictly morphologic, approach. When these structures are considered in terms of their associated rhombomeres, the anatomic pattern becomes quite clear. It is not surprising that the migration patterns of craniofacial muscles derived from somitomeres and somites are very different (Fig. 1.30).
Somitomeres and occipital somites follow similar trends in muscle formation [75,76,77,78,79,80,81]. Pharyngeal arches are hypaxial structures, the muscles of which are organized into two layers, deep and superficial. In this model, internally placed myoblasts of Sm7 produce the musculature of the soft palate, while externally placed myoblasts form stylopharyngeus and superior constrictor. Occipital somites make the muscles of the tongue from the internal plane of their myotomes. Sm8 (the first occipital somite) would be expected to produce a tongue muscle associated with the Sm7 palate; this is palatoglossus. Labeling studies show that the sternocleidomastoid and trapezius muscles originate from the external plane of the occipital somites. These two muscles logically represent the superficial plane of all four occipital myotomes. Both of these muscles originate from the osseous product of the Sm7, the mastoid process. Neuromeric theory allows for an accurate assignment of muscle origins according to their somitomere/somite of origin (based on the location of the motor nerve within the neuraxis). This approach to origin and insertion is rational and consistent, though in many cases it proves to be the reverse of that described in traditional textbooks of anatomy. We shall cover this subject in Chap. 9.
Comparative anatomy has much to contribute to our understanding of the significance of the r7/r8 transition zone. Fish brains are small. The positions of the roots of cranial nerve XI and the hypoglossal lie outside the piscine skull. These nerves can reach their targets without traversing the skull. The expansion of the vertebrate skull associated with tetrapods in the posterior fossa incorporated these nerves into the jugular foramen and four occipital foramina (subsequently reorganized in mammals into a single occipital foramen). This points to the transformation of somites 1–4 from a truncal to a cranial fate.
How Segmentation of the Mesoderm Matches that of the Neural Tube
Scanning EM studies in chicks by Meier in 1980 demonstrated that the segmentation of the somitomeres was also in register with segmentation of the intermediate mesoderm and lateral plate mesoderm (the somatic lamina) [65, 82]. Before continuing on, readers should note the following definitions having to do with varying regions of mesoderm. All forms of mesoderm come into being during gastrulation. The latter term refers to the process by which populations from a single layer of cells, the epiblast, rearrange themselves to form a three-layer “sandwich” consisting of ectoderm, mesoderm, and endoderm. A pear-shaped embryonic disc results; in the center of the disc an axially oriented neural plate develops. The ectoderm then becomes divided into two zones, neural ectoderm (the future nervous system) and surface ectoderm, the future epidermis of the body. The edges of the flattened neural plate roll up into cigar-like neural folds. These will ultimately contact each other in the dorsal midline seal up, forming the neural tube. This process is called neurulation [83, 84] (Fig. 1.31).
At the time of formation of the trilaminar embryonic disc (and before neurulation), mesoderm spreads out laterally to form various zones [85]. Those mesodermal cells closest to the neural tube and notochord are called paraxial mesoderm (PAM). PAM starts at the rostral tip of the notochord and is organized into somitomeres (and somites) all the way down to the tail. Mesoderm migrating further laterally stays flat; it is known as the lateral plate mesoderm (LPM). LPM has a natural separation plane (the embryo will eventually take advantage of this to form the intraembryonic coelom). The outer (dorsal) lamina of the LPM is the called somatic lateral plate mesoderm (LPMs for short). LPMs is responsible for forming all the non-axial bones of the body. From the clavicle on down, every bone (save the vertebrae) develops from LPMs. LPMs is overtly segmented in register with PAM. The inner (ventral) lamina of the LPM is called the visceral lateral plate mesoderm (LPMv for short). LPMv forms the mesoderm of the respiratory system and gut. It also forms the heart and blood vessels. After formation of PAM and LPM, an intercalated zone of intermediate mesoderm (IM) is formed; this runs down the length of the embryo as a segmented cord. IM is responsible for the production of the genitourinary system (Fig. 1.32).
Neural crest cells constitute a fourth germ layer of the embryo. These cells are found at the interface between the neural ectoderm and the surface ectoderm, i.e., atop the neural fold. In mammals, these neural cells “migrate” well before neural tube closure occurs, but after completion of gastrulation. Neural crest cells are essential to the formation of the aortic arch arteries. Since AA1 is fully formed by stage 9, the initiation of neural crest migration can be assigned to late stage 7 and is in full progress by stage 8. Note: the reader is forewarned that the concept of cellular migration may actually be a misnomer. An alternative viewpoint advocated by Anderson is that neural crest “movements” may represent local forms of cell proliferation that are physically directed by the microanatomy of their environment. In this model, the distinct manner in which neural crest populations reach their targets is a passive consequence of environmental passageways and roadblocks (vide infra).
Readers are warned here of another important revolution in biologic thinking. The germ layer system with which we are all so familiar is probably incorrect. It is certainly useful in terms of describing the organization of the vertebrate body via gastrulation and hence will probably persist in textbooks. But it is now known that a germ layer can produce cells associated with a different germ layer. For example, the hypoblast (the temporary second layer derived from the epiblast) lines the yolk sac which in turn produces the extraembryonic mesoderm (EEM). The intraembryonic mesoderm (IEM) produced later in time by gastrulation is in physical continuity with the EEM, but does not produce it! The endodermal lung bud emanating from the esophagus produces its own mesoderm. Thus, the mesenchyme of the lung does not result from the endoderm interacting with another (unknown) source of mesoderm, but actually is produced by the bud itself. Neural crest, of strictly ectodermal lineage, has the ability to produce structures such as fascia and bone normally considered outside the head and neck as strictly mesodermal derivatives (Fig. 1.33).
All physicians are aware that embryonic tissues seem to be organized into morphologic units called pharyngeal arches. With these definitions in mind, we can now explore how the somitomeric system fits this model. We shall then discuss the discovery of homeobox (Hox) genes provides the genetic basis for hindbrain segmentation. Thereafter, we shall complete our neuromeric map of the central nervous system by discussing how non-Hox genes define a more complex, but still logical, system of neuromeres rostral to the hindbrain. A population of neural crest is associated with each rhombomere, located at its respective sector along the neural fold. In like manner, each rhombomere is matched up with a corresponding somitomere. The exception to this is rhombomere 0. Neural crest from this neuromere interacts with prechordal plate mesoderm (PCM). In this way, the r0 neural crest provides the inferomedial sclera and the fascia for the inferior oblique, inferior rectus, and medial rectus, while the PCM provides the myoblasts (Fig. 1.30).
In humans, the paraxial mesoderm from the first 11 somitomeres and neural crest from r0 to r11 are organized into six head segments [86, 87]. The first segment is much larger than the rest. It receives mesoderm from somitomeres 1–3 and 5. It is aligned with mesomeres m1–m2 and r0–r1. The first segment does not interact with the pharyngeal arch system. It contributes to the orbit by forming all seven extraocular muscles. A new model (vide infra) posits this segment as constituting a new premandibular arch (herein abbreviated PA0). Each additional head segment is represented by a pharyngeal arch and consists of paired rhombomeres and a single somitomere (Table 1.1).
The model as referenced above was based strictly on somitomeres and is inaccurate. However, when corrected for the two rhombomeres per arch, it is plausible. The value of the model is that it is not strictly based on the arches and it takes into account mesenchyme that is dedicated to fronto-orbital structures (Fig. 1.34).
The reader will note that the arch numbering system for the pharyngeal arches differs from that of most texts. Each arch derives from two neuromeres. The ten rhombomeres of tetrapods (r2–r11) produce five pharyngeal arches. This paper proposes the following principle: neuromeres are never “lost” in evolution; their anatomic content is merely altered. Every line of evidence regarding the homeobox genes encoded in the hindbrain suggests that this system is strongly conserved throughout evolution. Thus, a terminology based on a mysterious “loss” of the fifth arch while the sixth is preserved is not logical. The neuromeric approach simplifies discussion of the comparative anatomy of pharyngeal arches [88].
Note that the first two pharyngeal arches are larger and exist as a dimer (pair). The mesenchyme (neural crest + paraxial mesoderm) of the second arch becomes confluent with, and is engulfed by, that of the first arch, like two slices of bread surround a slice of cheese. In this process, the ectoderm of PA2 is reduced to a small sector of skin covering the anterior external auditory canal. The endoderm of PA2 within the oral cavity is found transiently over a central swatch of the dorsum of the tongue, but is quickly absorbed. The last three pharyngeal arches are considerably smaller. The reason that they (PA3–5) are reduced in size is that some of their somitomeric mass must be shared to form somites, whereas the paraxial mesoderm of somitomeres 2–6 is completely dedicated to the formation of arch structures.
Recall that most bone from mesenchyme originating from r0 through r7 is not from PAM but from neural crest. Neural crest cells from each rhombomeric zone migrate downward from the neural fold and spill over the outer surface of their corresponding somitomere much like caramel poured over an apple. The tissue flows into pharyngeal arches. In each arch the neural crest from more cranial member of the rhombomere pair occupies the proximal/dorsal half of the arch, while the more caudal member is located in the distal/ventral half of the arch. An aortic arch artery supplies each pharyngeal arch transiently. The arterial axis enters each arch ventrally and exits dorsally to join with the ipsilateral dorsal aorta. Thus, the anterior and ventral aspect of each pharyngeal arch is metabolically more “mature.” For this reason, products of the distal arch such as bones and muscles appear before their proximal counterparts. The mandible (r3) is formed prior to the maxilla (r2). The facial muscles of the lower face (r5) appear earlier in time that those supplying the upper face (r4).
Although most craniofacial bones are formed from neural crest, PAM from the somitomeres has important derivatives as well. Sm1–3 do not participate in pharyngeal arch formation. PAM from Sm4 forms the parietal bone and likely basisphenoid. Sm5–Sm6 PAM produce the petrous temporal bone, while the PAM of the non-somitic part of Sm7 produces the mastoid process. Despite their lack of formal myotomes, the first seven somitomeres provide the myoblasts for many important craniofacial muscles.
Beginning with Sm7, all somitomeres change their configuration into fully separate somites having characteristic dermatomes, myotomes, and sclerotomes. The myotome of the first occipital somite produces two large external and ventral muscles, the sternocleidomastoid and the trapezius. Tongue musculature comes from the exclusively internal and ventral myotomes of the first through fourth occipital somites. The fascial envelopes of all muscles of the head and neck are derived from neural crest corresponding the neuromeric level of the myoblasts.
Molecular Basis of Segmentation
Introduction to Homeotic Genes
Vertebrates have a segmental body plan. Molecular biology has, since 1990, literally revolutionized our comprehension of the mechanisms by which segment formation and segment specialization are accomplished. The genes that control this development have an incredible degree of phylogenetic conservatism. Nucleotide sequencing studies show that genes found in primitive organisms such as Drosophila and worms exist in mammals as well and that these genes share common functions. A surprisingly limited number of genes are required. A given gene may play multiple roles depending upon the period of development in which it is expressed and the organ in which it is expressed. The molecular products of these genes can be broken down into two main types: (1) intracellular transcription factors, and (2) extracellular signaling molecules.
Transcription factors stay within the cell. They can bind to genes to initiate patterns of expression at key steps in development [89]. One class of transcription factors, the basic helix-loop-helix protein, has a short sequence of amino acids in which two alpha helices are separated by a “loop” of amino acids. Immediately next door to this sequence is a basic region that binds to DNA. The helix-loop-helix causes dimerization. This is typical in muscle development. Another class having an unusual geometry is the zinc finger protein. Symmetrically spaced units of cystine and histidine along the polypeptide chain bind to zinc ions via four ligands. When this happens, the four residues are drawn together and the polypeptide chain puckers up like a finger. This finger can subsequently insinuate itself into specific binding sites of DNA; activation of DNA sequences results.
In understanding embryonic segmentation and, ultimately, formation of the head, the most important class of transcription factors is the homeodomain proteins [24, 90,91,92,93,94,95,96,97,98,99,100]. These all have a helix-loop-helix configuration consisting of the same 61 amino acids. The DNA coding for this region, the homeodomain, is a unique sequence of 183 nucleotides known as the homeobox; every single gene producing this type of protein has the same sequence. Because of this molecular anatomy, these genes are called homeobox genes. Please note that in scientific terminology all genes are written in italics, whereas the protein product is not. Thus, the Sonic hedgehog protein is produced by the gene Sonic hedgehog (Shh for short). Note that when the gene is human, the entire abbreviation is capitalized, i.e., SHH.
As originally described in Drosophila, eight homeobox-containing genes exist on a single chromosome. These are divided into two sections, the anterior antennapedia complex and the posterior bithorax complex. Mammals possess 38 Hox genes analogous to those of the fruit fly. These are located on four different chromosomes and are arranged into 13 paralogous groups. What is amazing is that the genes are distributed anatomically along the chromosome. Hox genes for anterior mammalian segments are located at the 3′ end, while more caudal segments are at the 5′ end. Because the genes are activated and expressed in the 3′ to 5′ direction segment, formation in all organisms is a craniocaudal developmental sequence! (Figs. 1.35, 1.36, and 1.37).
Clinical Significance
This order has major experimental and evolutionary significance. Hox genes create characteristic morphology in each segment. Hox specification of neuromeric levels c7 and t1 creates the 12th and 13th somites; the 7th cervical vertebra results. At levels c8 and t1, a different Hox “barcode” results in a morphologically different structure, the first thoracic vertebra. When mutations in Hox genes occur, morphological variations are seen in the segmental structures that would normally have been expressed. A posterior-to-anterior transformation occurs when the contents of a give segment resemble those of the next most anterior structure. This is called a loss of function mutation. An anterior-to-posterior transformation occurs when the contents of a given segment resemble those of the next most posterior structure. This is called a gain of function mutation [101,102,103] (Fig. 1.37).
Retinoic acid (vitamin A) is a potent posteriorizing agent [104]. Exposure of mouse embryos to retinoic acid at a specific time in development causes an extra cervical vertebra to appear beneath the skull base! This structure, known as the proatlas, articulates with the skull with a single peg [105, 106]. The two condyles disappear. The cervical system, now consisting of 8 vertebrae, is also different. The morphology of the former atlas and axis reverts to that of standard murine cervical vertebrae. By the rules of vertebral formation, as the first of eight cervical vertebrae, the proatlas must be produced from the posterior half of the fourth occipital somite and the anterior half of the first cervical somite. The respective neuromeric levels are r11 and c1. Thus, the evolution of the foramen magnum and the double condylar occipito-atlantal joint was quite possible due to a “frameshift” mutation of the Hox code causing a shared “loss of function” between the fifth and sixth cervical somite. The paraxial mesoderm of these somites was “expropriated” from the proatlas to (1) expand the posterior braincase, (2) enable a double condyle system formerly between the proatlas and the second cervical vertebra to be “reassigned” to the skull base, and (3) reconfigure geometry of the foramen magnum (Figs. 1.38 and 1.39).
Potential evolutionary consequences of these changes are: (1) expansion of the posterior skull permitted better accommodation for an expanded visual cortex; (2) an increase in rotation, flexion, and extension of the skull was immediately useful for better predation and an augmented repertoire of feeding behaviors, and (3) better biomechanics for the occipitocervical junction adapted to tetrapod behavior on land…in primates this may even have facilitated the transition to an erect posture.
Hox genes are so-named because the homeobox-bearing genes are analogous to the antennapedia/bithorax complex. However, recent work has disclosed other gene families with no relation to Drosophila, but with the homeobox and additional conserved sequences. These include groups such as the Engrailed or Lim genes. The nine paired (Pax) genes all contain a paired domain of 128 amino acids; this is a DNA binding site [107]. POU genes (Pit-1, Oct-1, and Unc-86) all have a common 75 amino acid loop (for DNA binding) as well as a homeobox. As we shall see, many of these non-Hox homeobox genes have been used in the mapping of the vertebrate forebrain.
Some molecules produced by cells serve as extracellular “signal carriers.” Most inductions in embryology (interactions between tissues such as epithelium and mesenchyme) occur using peptide growth factors. The first molecule, nerve growth factor, was reported with great fanfare in the 1950s. Two large families of these molecules exist and their members are making their way into many scientific papers in our plastic surgery literature. The transforming growth factor-beta family is made up of more than 30 genes that are very active throughout embryogenesis and beyond. Induction of mesoderm, proliferation of myoblasts, and the invasive properties of angiogenic endothelium are all attributable to TGF-B products [108]. Nine genes make up the fibroblast growth factor family. FGF products perform a plethora of tasks from inducing growth of the limb bud to ensuring neuronal survival (and in their absence, neuronal death) [109]. These molecules are now widely used in and many applications have been made to cranial suture closure. A cautionary note however: the widespread and protean properties of the TGF-B’s, coupled with their ability to activate other cascades, make interpretation of laboratory results difficult at best. This is because, once again, the action of the very same gene can vary widely, for it depends heavily on the period of development and the location in which the gene is expressed.
In craniofacial development, one of the most potent signaling families, that of the hedgehog proteins, has been only recently isolated. In mammals, three forms of this protein (called Sonic hedgehog, Indian hedgehog, and Desert hedgehog) come from three different genes with the same name. Sites of action of Shh genes include the primitive node and notochord, the neural floorplate, the ectoderm of facial “processes” (there’s that pesky antiquated term again). Shared Shh activity in the apical ectoderm of the second pharyngeal arch and in the epithelial buds of the lungs may explain seemingly unrelated pathologic states [110,111,112,113,114,115,116].
Craniocaudal Pattern Formation
Gastrulation involves the creation of a three-layer “sandwich” of ectoderm on top of the future organism that comes from a single sheet of cells, the epiblast, sitting atop a nutrient source, the yolk sac (in the chick) or the gel-filled blastocoele cavity (in mammals). A two-layer sandwich comes first. In the avian model, epiblast cells are observed loosen their intercellular bonds. Cells “drop out,” from the epiblast in random fashion, falling down into the blastocoele cavity, where they reconnect to form a second, subjacent sheet [117, 118]. In mammals, the proliferation of epiblast leads to direct formation of an underlying second layer. In either model, the deeper cells are known as the hypoblast, or so-called primitive endoderm. A cellular cleavage plane now exists between the layers; this pathway will now permit gastrulation to happen. A groove up the midline axis of the epiblast forms, called the primitive streak. At the rostral end of the streak sits a small, elevated pit called Hensen’s node. Just like water draining from the bathtub down the drain, epiblast cells migrate toward the streak and enter the space below the epiblast. They are guided in their migrations by the basal lamina of the epiblast above. In so doing, these migrating cells are like arctic divers plunging through a hole in the ice and then navigating beneath the ice by following its undersurface.
At this juncture, we must raise a critical question, one with great philosophical implications: does cell “migration” exist and, if it does, how do the cells “know” where they should go? If we are conceived of cells following a route to a destination, how was the route set up in the first place? We wind up with the dilemma of attributing to cells a form of “knowledge” they cannot possess. Migration theories border close on the concept of Deus ex machina.
Fortunately, we are rescued from biochemical theism by the pioneering work of German embryologist Erich Blechschmidt (1880–1955) [119, 120]. Blechschmidt conceived embryologic events in very simple terms. Populations of cells in specific anatomic locations would expand their numbers in proportion to their relative access to nutrition. As the populations increase, identifiable anatomic structure within the embryo (ligaments and flexures) would channel the growth and ultimately affect the final three-dimensional form of these populations. Thus, the final shape of a structure is determined by physical factors, not a genetic program.
In the case of gastrulation, the Blechschmidt model postulates that the single layer epiblast is attached to an unyielding superstructure, the amnion. Around the periphery of the epiblast/amnion junction run two vascular structures, the vitelline veins. These run from the connecting stalk (or the yolk sac) forward to the primitive heart. This implies that the highest availability of nutrients to the cells of the epiblast occurs at the periphery. Multiplication of this peripheral population creates a physical force directed away from the constraining amnion and toward the midline. This force acts on the cells of the midline, causing them to ingress into the primitive streak. The midline epiblast cells are squeezed like toothpaste and forced out between the epiblast and hypoblast until they reach the periphery again. The situation is not unlike students lined up outside the door of a small theater. Once the doors open, the pressure from the back of the crowd will propel those students directly in front of the doors right into the theater and down to the very first row. Thus, the theater fills up from distal to proximal. We shall see that this same pattern recurs as the embryo “fills up” with mesoderm. The most peripheral mesoderm (extraembryonic or cardiogenic) comes from epiblast cells nearest the primitive streak. These are followed by lateral plate mesoderm cells and finally by the future paraxial mesoderm cells.
Cells ingressing from the epiblasts do so at four general spatial points: (1) Hensen’s node, (2) the remainder of the primitive pit, (3) the rostral streak, and finally (4) the caudal streak. The fate of a migrating cell (what layer it will belong to and how far out it shall migrate) is determined by its anatomic site of ingression and the timing of ingression. Another analogy might be that of paratroopers jumping out of a plane from four different doors. Standing in line before their respective doors, each receives instructions on his mission just prior to exiting. So it is that, from the apex of Hensen’s node, the very first cells to make their exit will become endoderm and, in the exact midline, the notochord. These are followed by the medial somitic mesoderm, the lateral somitic mesoderm, and finally, the lateral plate mesoderm. When gastrulation is complete at a given level of the embryonic axis, a caudal regression of Hensen’s node and of the streak takes place [83, 121,122,123,124,125,126,127].
The very first cells to ingress produce the foregut endoderm. Next follows mesoderm tip of the notochord and most anterior mesoderm called the prechordal plate mesoderm (PCM). Gastrulation occurs over time as a craniocaudal process of cellular ingression and reassignment. Sonic hedgehog produced at the PCM and at the notochord does two very important things. First, Shh from the PCM begins to specify the hindbrain, while Shh from the notochord begins to organize the somites, i.e., a decision is made between “brain” and “non-brain.” Furthermore, the ingressing cells at a given neuromeric level receive a Hox code “identity” that is identical with those epiblast cells that did not ingress. A unique pattern of Hox gene expression is thus a “barcode” that persists in all the cells that originated at the region, regardless of where they eventually migrate. This is readily seen in the form of the axial skeleton. Formation of a normal craniocaudal pattern of segment is reflected in the unique combination of Hox genes that specifies all vertebrae [102].
Noden applied these concepts higher in the neuraxis to the rhombencephalon by correlating patterns of gene product expression with cranial nerves and neural crest and rhombomeres [123]. A Hox “barcode” could be found for rhombomeric levels r3 and caudal, similar gene definitions using Krox-20 (the human form of this murine gene is EGR-2). Follistatin, Engrailed, and Wnt-1 permitted “mapping out” rhombomeres r0, r1, and r2 [128].
At rhombomere 0 (some texts describe this as the anterior border of r1), FGF-8 and Wnt-1 stimulate expression of engrailed genes En-1 and En-2. The farther one moves away from the “signaling center,” the lower the concentrations of En-1/En-2 proteins become. Indeed, r0 as the principle rhombomere of the mesencephalon expresses many important genes that determine the spatial organization of the brain. These molecules stimulate the rostral region to become mesencephalon (midbrain) and the caudal region at r1 to become metaencephalon (cerebellum). The midbrain subsequently forms two developmental units in which are located the superior and inferior colliculi for visual and auditory connections. Puelles and Rubenstein refer to these as m1 and m2, but for our purposes, since m2 (the preisthmic mesomere) is so small we will just refer to them collectively as m1.
In the 2013 iteration of the prosomeric model, early forebrain is divided into six prosomeres. Just like the rest of the spinal cord, hindbrain, and midbrain the prosencephalon possess a ventral (basal) zone and a dorsal zone (alar) zone. It should be noted that use of the terms basal and alar is a descriptive anlogy based upon the hindbrain. No sulcus limitans exist in the forebrain; therefore true “alar” or “basal” zones do not exist. We make use of this crude analogy simply to better understand the eventual differentiation of the prosencephalon into two major subdivisions. The caudal diencephalon is made from the alar and basal units of p1–p3 and contains the epithalamus, thalamus, and hypothalamus. The rostral diencephalon is formed from the basal units of p4–p6. This contains visual apparatus and olfactory lobe. The telencephalon is made from the three dorsal zones of p4, p5, and p6. It contains the cerebral cortex. Much progress in mapping the derivatives of the prosomeric zones is due to elegant work by Rubenstein and Puelles [32]. Terminology of the prosomeric model is described in Fig. 1.38.
The reader should note that recent changes in neuromeric mapping have been incorporated into the neuromeric model Consider r0 the boundary of midbrain and forebrain, “the organizer,” to be located at r0 [129]. The diencephalon is considered to have three prosomeres, whereas the secondary prosencephalon is described in terms of anatomic units. Figures 1.40 and 1.41 demonstrate these differences. For our purposes, since the anterior prosencephalic neural folds do not produce neural crest, and since it comes from above the diencephalon, we shall refer to PNC originating from prosomeres p1 to p3, but we will also describe the zones of the anterior neural folds populated by the PNC in term of p4, p5, and p6 (Fig. 1.42).
Neuromeric Basis of Neural Crest “Migration” and Fate
Now that we have a description of the neuromeric system firmly in mind, we shall need a schematic way to visualize it. Using a mouse embryo in saggital section, Rubenstein et al. have outlined the boundaries of the neuromeric zones. Key neuroanatomic landmarks are presented in a linear, schematic manner. Basal and alar regions of the entire neuraxis, including the prosencephalon, can be readily appreciated. These relationships are then projected onto an anatomic representation of the embryo. This author has modified these drawing by applying a color code in which each neuromeric zone along the neuraxis has its own specific color. When tracing out all the derivatives of rhombomere 3, all muscles and bones of the mandibular portion of the first pharyngeal arch will be depicted in green. This permits structures such as dermis and dura to be “mapped” backward to their origins at a specific neuromeric level. Let’s look at the neural folds in terms of five zones (Figs. 1.42, 1.43, 1.44, and 1.45).
Anterior Prosencephalic Neural Fold: Nonneural Ectoderm
Pioneering work by Couly and LeDourain established a detailed fate map of the avian neural plate [130,131,132,133,134,135,136,137,138,139]. In particular, they showed that neural crest cells do not occupy the entire length of the neural fold. The neural folds of the anterior prosencephalon (from p6 back to p4–p3) do not have neural crest cells.. The neural folds of the posterior prosencephalon (from p3 to p2 on backward) contain neural crest. All ectoderm lateral to the neural folds is destined to form epidermis. This ectoderm, plus that of the anterior prosencephalic neural folds, is termed nonneural ectoderm. We shall describe the anatomic content of the nonneural ectoderm in cranial-to-caudal order, i.e., from p6 to p4. The reader is advised that folding of the embryonic head coupled with growth of the forebrain causes radical changes in the spatial relationships of these three zones to occur, especially between p6 and p5. These changes will be described below.
The most anterior zones of nonneural ectoderm are termed p6. The first important structures of p6 are the adenohypophyseal placodes (AP). Described as singular, these fused structures occupy either side of the midline. Pathology in either AP explains the occurrence of epithelial tumor localized to either the right or left pituitary. Posterior to each AH placode, the nonneural ectoderm becomes the nasal epithelium (NE). This forms the epithelial lining of the nasopharynx all the way forward from Rathke’s pouch and Waldeyer’s ring to the true skin of the nasal vestibule. The NE makes a boundary with true skin of the nose and upper lip (in the chick known as upper beak epithelium UBE). Within the nasal epithelium lies the nasal placode (NP). This specialized epithelium gives rise to three classes of neurons, all of which migrate into the brain. The lateral NP contains olfactory neurons for conventional odor detection. The medial NP contains accessory olfactory neurons, those involved with the detection of complex chemical signals, pheromones being the prime example. Finally, the medial NP contains neurons associated with gonadotropin hormone releasing hormone (GnRH). Animal behaviors related to detection of sexual cues from urine and sniffing of genitalia relate to this system. GnRH is related to development of secondary sex characteristics. Kallman’s syndrome (anosmia and/or hypogonadotropic hypogonadism) results from anatomic defects in the nasal placode [140,141,142,143,144]. Note that the Spanish pathologist, Maestre de San Juan, first described this syndrome [145] (Fig. 1.44 Placodes).
The p5 zone of nonneural ectoderm comprises the skin of the nose and the philtrum of the upper lip. This upper beak epithelium (UBE), the term is derived from the avian model, constitutes the epidermis of the nose and philtrum of the upper lip, but not the forehead. The lateral boundaries of the UBE are marked by sensory innervation of V1 and lie medial to the arcade formed by the facial artery-angular artery. Just caudal (posterior) to the UBE lies the optic placode (OP). This forms the lens and is crucial for development of the globe (and ultimately for that of the orbit).
The p4 zone of nonneural ectoderm forms the calvarial ectoderm. The epidermis of the forehead and the frontal bone sweep back over the cerebrum within this zone. In birds, the frontal bone is huge, while the parietal bone is diminutive. Formation of the avian calvarium is exclusively from neural crest and is extremely rapid. Sutures are not observed; pathologic craniosynostosis does not occur. In mammals, the parietal bone forms from PAM of r2 and r3 derivation. The varying forms of synostosis all involve boundary between bones of dissimilar developmental derivation. These observations may have important implications for the relative incidence of craniosynostosis observed in humans.
Without a dermis for vascularization and support, epidermis becomes rather worthless. Indeed, all epithelia require a subjacent supporting layer. The production of all dermis and bone in the face is exclusively the responsibility of neural crest. So where does the neural crest come from and how does it “know” what to do? The answer to the first question lies in a clear conception of the three functional types of neural crest and of the manner in which they migrate [146, 147]. The answer to the second question relates to the “programming” function exerted upon the neural crest cells by the epithelial environments they inhabit [148,149,150]. We shall deal with the first question in the section below and discuss the second at the conclusion of this essay.
Posterior Prosencephalic Neural Fold: Neural Crest (PNC)
The first zone of neural crest extends over the posterior prosencephalon from p3 back to p1. This PNC moves forward in the subectodermal plane as a large vertical sheet of cells in the midline. The PNC migrates from the dorsal part of the lamina terminalis, from which also is formed the plate of corpus callosum. For this reason, p6 deficiency associated with holoprosencephaly can result in hypoplasia or absence of the intercanthal ligament; hypertelorism results. This is contrasted to the r1 deficiency state seen in anencephaly. The r1 component of the frontal bone is absent, the absence of r1 dura and neural crest pericytes profoundly affects forebrain development, and the sphenoid bone (an r1 structure) is small and misshapen. (For further details, see Chap. 5 on the forebrain commissure.)
Migration of p3–p1 neural crest beneath the nonneural folds of upper beak ectoderm produces frontonasal skin. We shall describe this skin as p6–p4. Subsequently, subcutaneous tissue are provided by MNC. These cells will form nasal cartilages in accordance with the underlying program of p6–p5 skin.
The most anterior cells of the sheet will reach the p6 skin where they will be “instructed” in situ to make upper lateral cartilages. Into this envelope, MNC neural crest from r1 migrates to form nasal septum, the alar and lateral cartilages, the perpendicular plate of the ethmoid, the ethmoid labyrinth and crista galli and, finally, the upper and middle turbinates. Sixth prosomere neural crest also “activates” the adenophypophyseal and nasal placodes. Without interaction with neural crest, these placodes are nonfunctional.
When MNC arrives in the p5 zone, it forms the nasal bones and the lower lateral cartilages. The nasal process of the frontal bone descends deep to the nasal bones to articulate with the facial process of the premaxilla. Here, to avoid future confusion, we must make a brief digression. The piriform rim is really a bilaminar structure. Its internal aspect is made from the nasal process of the frontal bone (p5) that extends downward just beneath the nasal bone. It abuts the frontal process of the premaxilla, PMxF, a product of r2 NCC supplied by medial/lateral nasopalatine axes. Overlying PMxF is the frontal process of the maxilla, MxF, an r2 derivative supplied by medial branch of the anterior superior alveolar axis. This abuts against the nasal bone itself. The bicortical structure of the piriform rim makes it stronger; it is capable of holding screws. Surgeons refer to the piriform as a “buttress” and use it for placement of fixation plates. Additional bones programmed by the p6 vestibular lining are the orbital lamina of the ethmoid bone and the lacrimal bone.
The boundary between the neural crest derivatives of zone p5 and those of zone p4 is uncertain at this time. The supraorbital margin of mammals and the roof of the orbit are p5, while the remainder of the forehead would logically be p4. Indeed, the ossification centers of the frontal bone appear in caudal to cranial order. In primitive tetrapods, orbital rim has two distinct fields, prefrontal (PrF) and post frontal (PtF), arranged on either side of the supraorbital neurovascular axis. These bones arise from MNC that arrives prior to that covering the frontal lobe. For this reason, they persist even when frontal bone per se is absent, as in cases of anencephaly.
Mesencephalic Neural Crest (MNC)
The MNC lies over the mesencephalon. Unlike PNC that moves as a large sheet, MNC proliferates as distinct streams, described as r0–r1, followed by m1–m2. Recall that the development of the mesencephalon is stimulated by FGF-8/Wnt-1 produced at the isthmus (levels r0 and r1). Thus, migration from r0 to r1 antedates that from m1 to m2. Neural crest associated with each of these neuromeres a remarkable expansion atop the rapidly growing mesencephalon. Although the midbrain is quite small in the mature state, the embryo is enormous. This pre-migratory MNC contributes to large surface areas of dura associated with the sensory distribution of V1.
Mesencephalic neural crest migrates in three successive streams in craniocaudal order. That from r0 and r1 is dedicated to development of extraocular connective tissues and the posterior orbital wall. Neural crest from r0 migrates first and its final destination is the most medial and cranial of all MNC. It makes no bone. It lies internal and caudal to the optic vesicle. It therefore produces the inferomedial sclera and the fascia of the corresponding extraocular muscles innervated from the cranial oculomotor nucleus (inferior oblique, inferior rectus, and medial rectus). A limited amount of dura over the base of the forebrain, the rhinencephalon, would logically come from r0 neural crest. The innervation pattern of the terminal nerve (cranial nerve 0) may well map out the contribution of r0 neural crest.
Next to migrate is MNC from r1. The presence of the optic vesicle forces this stream to assume a superolateral position within the future orbit. It also fills in the space behind the globe and lateral to the future optic nerve. The presphenoid bone is formed from r1 MNC via a chondral intermediate. So too is the lesser wing of the sphenoid, being derived from preexisting orbitosphenoid cartilage. As the superolateral half of the optic vesicle is enveloped by r1 MNC, its scleral coat is acquired as well as the fasciae for the remaining extraocular muscles innervated by the caudal oculomotor nucleus (superior rectus and levator palpebrae superioris). A great deal of r1 neural crest is involved with dura synthesis. It covers the entire cerebral cortex innervated by V1.
In the initial iteration of this model, consideration was given as to whether the neural crest cells had a predetermined identity according to their position along the neural fold. This is unlikely. The neuromeric population migrates into its destination where it becomes organized according to the genetic “street map” of the pharyngeal arch. Each sector of the arch is targetted by a vascular pedicle. In the case of the first arch, these are all branches of the internal maxillomandibular axis which were “added on” as the result of its union with the extracranial stapedial. And the spatial order of these branches may be determined by their position within the arch. Thus, considering the various StV2 branches fanning out from the pterygopalatine fossa, medial nasopalatine may represent the most distal branch; therefore, it peels off away from the other arteries to maxilla and zygoma, to follow a distinct route into the midline of the oral cavity.
Rhombencephalic Neural Crest (RNC)
Rhombencephalic neural crest does not travel in streams; at each neuromeric level, it proliferates or “migrates” as a segment into and around its corresponding somitomere and thence into its designated pharyngeal arch.. Furthermore, cells from one neuromere do not mingle with those of a neighboring neuromere. Under experimental conditions, mixing between even or odd numbered neuromeres is permissible but not in early life. Neural crest from r2 does not mix with that of r3. Because it obeys the “rules,” this neural crest from the most rostral rhombomere is named r2 RNC.
RNC can be subdivided into two zones. The rostral zone (RNCr) is described neural crest interacting with those somitomeres that do not go on to become full-fledged somites. These include the head mesoderm of Sm1 and the first three pharyngeal arches. As r2–r7 reorganize themselves into arches, the somitomeres seem to “pair up.” These dimers reorganize themselves into a single structure. Perhaps because the cellular volume of the first three pharyngeal arches is greater, neural crest migration into them takes longer to complete than that of more caudal zones. RNCr migration is not completed until somite-stage 14.
The caudal zone of rhombencephalic neural crest (RNCc) involves neuromeres responsible for populating the second group of pharyngeal arches. PA4–5 are smaller in size. It is therefore reasonable to expect that the neural crest population would take less time to migrate into these arches. Indeed, RNCc migration is complete by somite-stage 11. Later in this series, we shall speculate as to the manner in which the anatomy of a pharyngeal arch comes about and the manner in which this affects the order in which different fields within a pharyngeal arch are expressed (as reflected in the ossification sequence of craniofacial bones).
Spatial Reassignment of Nonneural Ectoderm
Head folding coupled with forebrain growth causes changes in the spatial relationships of nonneural ectoderm p6–p4. Prior to closure of the anterior neural folds, the p6 zone starts from the midline just rostral to the stomodeum, just in front of Rathke’s pouch, and projects backward (caudally) along the folds. The apical surface of this future nasal vestibular epithelium (NVE) faces upward (dorsally). With closure of the neural folds, both p6 zones come together in the midline. They acquire a population of PNC from the diencephalon at this time. Head folding forces this zone to roll forward and ventrally much like the tracks of an armored car. The NE that once faced anteriorly is now drawn into the future nasal cavity. At the leading edge of the NVE lies the AH placode and just behind it, the nasal placode (NP). The apical surface of the nasal vestibular epithelium now faces ventrally.
As the p6 NE and NP are pulled into the future nasal cavity, similar changes occur with the p5 zone of nonneural ectoderm. This zone of the upper beak epithelium is slated to become the future epidermis of the nose and upper lip. Being in continuity with the p6 NE, the p5 UBE is pulled forward. The topology of the nasal epithelium resembled the letter U turned 90 degrees onto its side. The upper limb of the U is the keratinized p5 epithelium of the nasal dorsum. Neural crest that has migrated immediately below the p5 epithelium will be programmed by it to form dermis and the nasal bones. The lower limb of the U is the nonkeratinized p6 epithelium of the nasal vestibule. The vomeronasal organ, located in the membranous septum, probably represents the remnant of the nasal placode. Neural crest that has migrated immediately above the p6 epithelium will be programmed by it to form the alar and triangular cartilages. The future nose consists of two such p6/p5 systems. The medial walls contain p6 neural crest. As these approximate, the perpendicular plate of the ethmoid and the nasal septum result.
Another way to visualize the topology of the nasal chamber is to imagine a condom placed on a flat surface. The tip of the reservoir represents the p6 nasal placode, while the remaining latex of the reservoir is the p6 nasal epithelium. All the rest of the condom is p5 nasal skin. If the tip of the reservoir is glued to the table and the condom inflated, an invagination will occur. This will place the nasal skin on the outside and the nasal epithelium on the inside.
Timing of Neural Crest Migration
Age in embryos is calculated by the presence of anatomic landmarks. Because somites appear at absolutely regular intervals and are readily counted, somite-stage is a reliable means to measure time in a developing embryo. Neural crest migrates from different anatomic sites at different times. To study this pattern, Osumi-Yamashita labeled neural crest populations from four zones: rostral rhomoboencephalon RNCr, caudal rhombencephalon RNCc, mesencephalon MNC, and caudal prosencephalon PNC. Neural crest migration from each zone was studied and the timing of its completion (as measured by somite stage) was determined.
At 11-somites, RNCr (neural crest from r2 to r7) migration was complete. This zone populates the first three pharyngeal arches in craniocaudal order. Thus, first arch bones such as the mandible appear before third arch derivatives such as the greater cornu of the hyoid. At the 14-somite stage, two distinct zones complete their migrations. RNCc (neural crest from r7 to r11) completes its migration in cranio-caudal order. MN, on the other hand (from m1 to m2 and r0 to r1), proceeds caudal to cranial. That is r1 migrates then m1. At the 16-somite stage, PNC migration is complete. PNC proliferation appears to proceed in caudal-to-cranial order, like toothpaste being squeezed from a tube. Thus, the p5 zone is populated by PNC prior to p6. Placodes are activated by the presence of underlying neural crest. Experimentally, it is known that the p5 optic placode certainly appears before the p6 olfactory placode.
The take-home message of this work is that cranial neural crest cells depart from the neural folds in strict order according to the neurologic maturation of the corresponding neuromeres. Thus, RNC from r2 to r11 migrates in cranio-caudal order. MNC gets started a bit later, because hindbrain induces midbrain. First, rhombomere, being biologically associated with midbrain, start first and is followed by the mesomere. Whether any significance can be placed upon cells coming from the diminutive neuromeres r0 and m2 is unknown, so we just concentrate on r1 and m1. Finally, since prosencephalon is induced after mesencephalon and develops in caudal to cranial fashion, it makes sense that PNC migration takes place last and follows the order p1 > p2 > p3. Once again, what significance can be attached to the localization of p1 vs. p2 in terms of the derivatives produced is unknown.
Pathologies tend to be more severe the earlier in development they strike. A very late “hit” on neural crest migration to the midline could cause a mild holoprosencephaly due to absence of the ethmoid complex without affecting the orbit. The more severe is this type of neurocrisopathy, the greater the degree of hypotelorism will result.
Fate of the Neural Crest: The Role of Epithelial “Programming”
In the preceding discussion, we have seen how neural crest derivatives receive instructions from the nonneural ectoderm as to what structures they should form. Many important surgical implications flow from these considerations. NVE (nasal vestibular epithelium) from p6 determines the size and shape of the upper lateral and septal cartilages of the nose, whereas the UBE (nasal skin) from p5 will affect the formation of lower lateral nasal cartilages and the nasal bones. Might a similar role be exerted by the foregut endoderm?
Recent experimental work from the laboratories of Couly and LeDouarin is of enormous clinical significance in this regard for it supports the hypothesis that foregut endoderm (FGE) plays the decisive role in the formation of neural crest bones and cartilages of the pharyngeal arches. First, let it be said that neural crest cells if left to themselves in the Petri dish will form cartilage. Next, it is necessary that we recognize that certain membranous bones of neural crest form via cartilaginous intermediates, while others form directly within membrane. By applying techniques of surgical extirpation and transplantation in the previously described quail chick chimera model, these workers mapped out distinct territories of FGE and found that specific zones were responsible for the production of specific bones and cartilages. This yielded a “map” of FGE in which endoderm destined to form Sessel’s pouch was found to underlie the frontonasal bud, while the remainder of FGE resulted from the summation of individual outgrowths of endoderm corresponding to the pharyngeal pouches.
Following our previous discussion of gastrulation, it is possible to conceive foregut endoderm as having a neuromeric code as well. The r1 zone is unclear because the vestibular lining is all a PNC/MNC complex supplied by V1. With closure of the palate, this mucosa becomes hidden from view, but it is histologically distinct from that of the p6 nasal epithelium. This is readily apparent on lateral dissection of the septum in which the mucoperiosteum stops at the border of the vomer, the innervation changes from V1 (the medial nasal nerves) to V2 (the sphenopalatine nerve), and the blood supply switches from internal carotid along the upper septum to the external carotid sphenopalatine coursing along the septo-vomerine border. All the remaining endoderm of the pharynx and larynx begins posterior to the buccopharyngeal membrane.
Couly and Le Douarin made an important distinction between neural crest cells emanating from differing parts of the neural fold in terms of the expression of Hox genes. They found that rostral neural crest corresponding to PNC and MNC constituted a domain in which Hox genes were not expressed (Hox-negative) These neural crest cells gave rise to the membranous bones of the neurocranium, the nasal capsule, and the first pharyngeal arch maxilla and mandible. Thus, the Hox-negative domain extended to r3. In contrast, the domain corresponding the pharyngeal arches 2–5 was considered Hox-positive and yielded the hyoid bone and cartilages of the visceral skeleton. Specific mapping of the hyoid-forming region of FGE revealed that extirpation of specific zones caused failure of formation of corresponding components of the hyoid. When the entire hyoid region was reverse 180 degrees, the orientation of the hyoid bone was reversed as well.
The take-home message of this work and previous studies by the same authors is that pharyngeal endoderm is required from neural crest cells to become cartilage or membranous bones based on cartilage, while ectoderm is required for neural crest to become membranous bone via the classic mechanism (no cartilaginous intermediate). Hox-negative neural crest is exclusively responsible for the generation of the facial skeleton, but does not possess the information required to pattern the skeleton. For this to occur, FGE is required. Defined areas of FGE induce the formation of specific bones and cartilages from cephalic neural crest. Information required to determine the axes of facial bones depend upon the spatial orientation of specific zones of FGE. The ability to respond to patterning cues from FGE is exclusive to non-Hox-expressing cephalic neural crest. Cells form different zones of the Hox-negative domain which behave in an equivalent manner. Due to enormous regenerative capability, up to 3/4 of the neural fold responsible for cephalic morphogenesis can be removed with no consequences. This implies that the pathways taken by PNC and MNC are not inherent to the neural cells per se, but are determined by the microenvironment through which the neural crest cells migrate/proliferate. Although the nature of the signaling employed by FGE to instruct neural crest is uncertain, it has been shown that the spatial orientation of branchial pouch endoderm is established prior to the migration of neural crest into the pharyngeal arches. Protein markers such as BMP7, FGF8, and Pax1 are found in different regions of the FGE irrespective of whether neural crest cells are present or not. Indeed, these polarities may well be established prior to formation of the pharyngeal arches, i.e., they may be determined by the spatiotemporal sequence of cell movements at gastrulation itself.
These considerations allow us to think of facial bones as the products of specific biosynthetic units or fields corresponding to the concept of functional matrix first elaborated by Moss years ago. Individual deficits of the facial bones as manifested in craniofacial clefts would thus result from very early insults or aberrations of either epithelial programming units, in the neural crest cells that come to populate them, or in the biochemistry of epithelial-mesenchyme interaction.
Ectomeres and Endomeres: A Final Note
Le Dourain et al. have mapped out developmental fate of craniofacial ectoderm. Recall that gastrulation does not take place anterior to r0–r1. The buccopharyngeal membrane maps to r1 ectoderm and r1 endoderm with no intervening mesoderm. With head flexion and the forward positioning of the pharyngeal arches, BPM is pulled inside the oral cavity. Its anterior surface is ectodermal. This means that all mucosa of the mouth anterior to BMP is ectoderm from r2 and r3, not endoderm. Furthermore, the future oral cavity is separated from the nasal cavity by a wall consisting of intraoral r2 ectoderm, a middle layer of MNC, and intranasal p6 vestibular mucosa. When the bottom of the nasal cavity disintegrates, prosencephalic ectoderm directly meets rhombencephalic ectoderm at the border between septum and perpendicular ethmoid (p6) versus vomer (r2). Upper eyelid skin maps to r1. The distribution pattern of V1 scalp demonstrates persistence of PNC skin in the midline all the way back to posterior fontanelle (Fig. 1.46).
Endoderm begins behind the BMP. After it undergoes apoptosis, the oropharynx is mapped with concentric rings first arch, with a small zone of r2 mucosa being dosal. The corresponding to r3 mucosa is much larger and lies rostral to the tonsil. Second arch mucosa is eliminated in the oropharynx such that third arch mucosa abuts first arch behind the tonsillar fossa.
Summary
We have now completed a brief synopsis of neuromeric organization. Neuromeres are developmental units of the nervous system with specific neurologic content. Outlying each neuromere are tissues of ectoderm, mesoderm, and endoderm that bear an anatomic relationship to the neuromere in three basic ways. This relationship is physical in that motor and sensory connections exist between a given neuromeric level and its target tissues. The relationship is also developmental because the target cells exit during gastrulation precisely at that same level. Finally, the relationship is chemical because the genetic definition of a neuromere is shared with those tissues with which it interacts. The model developed by Puelles and Rubenstein is used to describe the neuroanatomy of the neuromeres. Although important details of the model are currently being refined, it has immediate clinical relevance for practicing clinicians.
The physical size and shape of each neuromere are defined by the protein products of genes expressed within the confines of the neuromere. Many of the crucial genes in this system contain a unique sequence of DNA bases, leading to a stereotypical amino acid sequence known as the homeobox. Homeobox genes are master regulators of other genes because the homeobox unit unlocks other DNA sequences. Homeobox genes are divided into two classes: Hox genes are homeobox genes analogous to those originally described in Drosophila. Non-Hox homeobox genes possess the homeobox sequence of bases, but bear no relationship to the Drosophila system.
Neural crest developing in the neural fold above a given neuromere bears a similar relationship with that of neuromere. Neural crest cell populations are organized into three main groups depending upon their location along the neuraxis. The physical behavior of neural crest migration is determined by the microanatomy of each of these three environments. Neural crest from the caudal prosencephalon moves forward as a cohesive sheet and populates the nonneural ectoderm of the rostral prosencephalon. PNC is responsible for the mesenchyme producing the bones, cartilages, and connective tissues of the fronto-orbital-nasal mass. The nature, shape, and size of these derivatives is not inherent in the cells of the PNC, but results from the instructions given to PNC by specific “target” zones of ectoderm (termed p6, p5, and p4) and foregut endoderm (r0, r1, r2′) with which the neural crest cells interact.
Mesencephalic neural crest is associated with neuromeres r0, r1, and r2′ and travels as individual cells in anatomically distinct streams. This MNC is responsible for the bulk of the orbit, the sclera, and the sphenoid. Its most caudal portion (r2′) is of great clinical importance because it interacts with r2′ endoderm to produce the vomer and premaxilla. Deficits in this epithelial-mesenchymal system are responsible for the most common forms of clefting involving the primary palate, the lip, and the secondary palate (in association with cleft lip).
Rhombencephalic neural crest begins at neuromeric level r2 and continues to r11. Neural crest migration is segmentally segregated into pairs of outlying somitomeres and somites forming the five pharyngeal arches. This process occurs in cranio-caudal order and is completed over two time periods. Population of the first three arches (the rostral RNC or RNCr) coincides with that of MNC. A second wave (the caudal RNC or RNCc) populates pharyngeal arches 4 and 5.
This paper permits us to think about the head and neck in terms of neuromeric terminology. Relationships between the processes of neurulation and gastrulation have been presented to demonstrate the manner in which neuromeric anatomy is established in the embryo. We are now in a position to describe in detail the static anatomic structures that result from this system. The neuromeric “map” of craniofacial bones, dermis, dura, muscles, and fascia will be the subject of the next part of this series.
References
Carlson BM. Human embryology and developmental biology. St. Louis: Mosby; 1999.
Gilbert S. Developmental biology. 7th ed. Sunderland, MA: Sinauer Associates; 2003.
Carstens MH. The sliding sulcus procedure: simultaneous repair of unilateral clefts of the lip and primary palate—a new technique. J Craniofac Surg. 1999;10:415–34.
Carstens MH. The spectrum of minimal clefting: process-oriented cleft management in the presence of an intact alveolus. J Craniofac Surg. 2000;11:270–94.
Carstens MH. Correction of the unilateral cleft lip nasal deformity using the sliding sulcus procedure. J Craniofac Surg. 1999;10:346–64.
Carstens MH. Correction of the bilateral cleft using the sliding sulcus technique. J Craniofac Surg. 2000;11:137–67.
Carstens MH. Sequential cleft management with the sliding sulcus technique and alveolar extension palatoplasty. J Craniofac Surg. 1999;10:503–18.
Carstens MH. Functional matrix cleft repair: a common strategy for unilateral and bilateral clefts. J Craniofac Surg. 2000;11:437–69.
Carstens MH. Neural tube programming and craniofacial cleft formation. I. The neuromeric organization of the head and neck. Eur J Paediatr Neurol. 2004;8:181–210.
Carstens M. Functional matrix cleft repair: principles and techniques. Clin Plast Surg. 2004;31:159–89.
Delaire J. The potential role of facial muscles in monitoring maxillary growth and morphogenesis. In: Carlson DS, McNamara Jr JA, editors. Muscle adaptation and craniofacial growth. Craniofacial growth monograph no. 8, Center for Human Growth and Development. Ann Arbor, MI: University of Michigan; 1978. p. 157–80.
Delaire J, Precious D, Gordeef A. The advantage of wide subperiosteal exposure in primary surgical correction of labial maxillary clefts. Scan J Plast Reconstr Surg. 1989;22:147–51.
La DJ. cheilo-rhinoplastie primaire pour fente labiomaxllarie congenitale unilateral. Revue Stomatolo. 1975;76:193.
Delaire J. Theoretical principles and technique of functional closure of the lip and nasal aperture. J Maxillofac Surg. 1975;6:109.
Delaire J. The potential role of facial muscles in monitoring maxillary growth and porphogenesis. In: McNamara JA, editor. Monograph 38: “Cranial growth series”. Center for Human Growth and Development. Ann Arbor, MI: University of Michigan Press; 1978.
Delaire J, Precious S, Gordeef A. The advantage of wide subperiosteal exposure in primary surgical correction of labial maxillary clefts. Scand J Plast Reconstr Surg. 1993;22:710.
Markus AF, Delaire J, Smith WP. Facial balance in cleft lip and palate I. Normal development and cleft palate. Br J Oral Surg. 1992;30:287–95.
Precious DA, Delaire J. Surgical considerations in patients with cleft deformities. In: Bell WH, editor. Modern practice in orthognathic and reconstructive surgery. Philadelphia: WB Saunders; 1992. p. 390–425.
Mulliken JB. Repair of bilateral cleft lip: review, revisions, reflections. J Craniofac Surg. 2003;14:68–76.
His W. Beobachtungen zur Geshichte und Gamenbildung beim menschlichen embryo. Kgl Akad Wis. 1901;27.
His W. ABh sachs Ges (Akad) Wiss. 1902;27:347.
Carstens MH. Developmental anatomy of the facial midline. Plastic Surgery Educational Foundation awards presentation. American Society of Plastic Surgeons annual meeting, Los Angeles, CA, Nov 2002.
Berquist H. Studies on the cerebral tube in vertebrates: the neuromeres. Acta Zool. 1952;33:117–23.
Graham A, Papalopulu N, Krumlauf R. The murine and Drosophila homeobox gene complexes have common features of organization and expression. Cell. 1989;57:367–78.
Landmesser LT, editor. The assembly of the nervous system. New York: Liss; 1989.
Vaage S. The segmentation of the primitive neural tube in chick embryos (Gallsu domesticus). Adv Anat Embryol Cell Biol. 1969;41:1–88.
Butler AB, Hood M. Comparative vertebrate neuroanatomy. New York: Wiley-Liss; 1997.
Puelles L, Rubenstein JLR. Expression patterns of homeobox and other putative regulatory genes in the embryonic mouse forebrain suggest a neuromeric organization. Topics Neurosci. 1993;16:472–80.
Rubenstein JLR, Martinez S, Himamura K, Puellas L. The embryonic vertebrate forebrain: the prosomere model. Science. 1994;266:578–80.
Rubenstein JLR, Puelles L. Expression patterns of homeobox and other putative regulatory genes in the embryonic mouse forebrain suggest a neuromeric organization. Topics Neurosci. 1993;16:472–80.
Rubenstein JLR, Puelles L. Homeobox gene expression during development of the vertebrate brain. Curr Topics Dev Biol. 1994;29:1–63.
Puelles L, Rubenstein JLR. Forebrain gene expression domains and the evolving prosomeric model. Trends Neurosci. 2003;26:469–76.
Sarnat HB. Personal communication, Nov 2003.
Cambronero F, Puelles L. Rostrocaudal nuclear relationships in the avian medulla oblongata: a fate map with quail-chick chimeras. J Comp Neurol. 2000;427:522–45.
Wingate RJT, Lumsden A. Persistence of rhombomeric organization in the postsegmental hindbrain. Development. 1996;112:2143–52.
Muller F, O’Rahilly R. The timing and sequence of appearance of neuromeres and their derivatives in staged human embryos. Acta Anat. 1997;158:83–99.
O’Rahilly R, Gardner E. The timing and sequence of events in the development of the human nervous system during the embryonic period proper. Z Anat Entweck-Gesch. 1974;134:1–12.
O’Rahilly R, Muller F. Human embryology and teratology. 4th ed. New York: Springer-Verlag; 2001.
Figador MC, Stern CD. Segmental organization of embryonic diencephalons. Nature. 1993;363:630–4.
Larsen CW. Boundary formation and compartition in the avian diencephalon. J Neurosci. 2001;21:4699–711.
Bonner-Fraser M. Rostrocaudal differences within the somites confer segmental pattern to trunk neural crest migration. In: Ordahl CP, editor. Somitogenesis, part 1. San Diego: Academic Press; 2000. p. 279–96.
Brand-Saberi B, Whiting J, Ebesperger C, Christ B. The formation of somite compartments in the avian embryo. Int J Dev Biol. 1995;40:411–20.
Burke AC. Hox genes and the global patterning of the somatic mesoderm. In: Ordahl CP, editor. Somitogenesis, part 1. San Diego: Academic Press; 2000. p. 155–81.
Christ B, Ordahl P. Early stages of chick somite development. Anat Embryol. 1995;191:381–96.
Dietrich S, Schubert FR, Healy C, Sharpe PT, Lumsden A. Specification of hypaxial musculature. Development. 1998;125:2235–49.
Fan CM, Tessier-Levigne M. Patterning of mammalian somites by surface ectoderm and notochord: evidence for sclerotome induction by a hedgehog homolog. Cell. 1994;79:1175–86.
Hall BK. The embryonic development of bone. Am Sci. 1988;76:174–81.
Hall BK, Miyake T. Divide, accumulate, differentiate: cell differentiations in skeletal muscle revisited. Int J Dev Biol. 1995;39:881–93.
Kato N, Aoyama H. Dermamyotomal origin of the ribs as revealed by extirpation and transplantation experiments in chick and quail embryos. Development. 1998;125:3437–43.
Nowicki JL, Burke AC. Testing Hox genes by surgical manipulation. Dev Biol. 1999;210:238.
Ordahl CP. Myogenic lineages within the developing somites. In: Bernfeld M, editor. Molecular basis of morphogenesis. New York: Wiley-Liss; 1993. p. 165–70.
Pourquie O, et al. Lateral and axial signals involved in somite patterning: a role for BMP-4. Cell. 1996;84:461.
Pourquie O. Segmentation of the paraxial mesoderm and vertebrate somitogenesis. In: Ordahl CP, editor. Somitogenesis, part 1. San Diego: Academic Press; 2000. p. 82–106.
Tajbakhsh S, Spurle R. Somite development: constructing the vertebrate body. Cell. 1998;92:127–38.
Tam PPL, Behringer RR. Mouse gastrulation: the formation of a mammalian body plan. Mech Dev. 1997;68:3–25.
Tam PPL, Trainor PA. Specification and segmentation of the paraxial mesoderm. Anat Embryol. 1994;189:379–90.
Tam PPL, Goldman D, Camus A, Schoenwolf GC. Early events of somitogenesis in higher vertebrates: allocation of precursor cells during gastrulation and the organization of a meristic pattern in the paraxial mesoderm. In: Ordahl CP, editor. Somitogenesis, part 1. San Diego: Academic Press; 2000. p. 1–32.
Venters SI, Thornsteindottir S, Duxton MI. Early development of the myotome in the mouse. Dev Dyn. 1999;216:219–32.
Goodrich ES. Studies on the structure and development of vertebrates. London: Macmillan; 1930.
DeBeer GR. The development of the vertebrate skull. Oxford, 1937. Paperback reprint, University of Chicago; 1988.
Jarvik E. Basic structure and evolution of vertebrates. New York: Academic Press; 1980.
Hinrichsen K. The early development of morphology and patterns of development of the face in the human embryo, Advances in anatomy, embryology and cell biology, vol. 98. New York: Springer Verlag; 1985.
Meier SP. Development of the chick mesoblast: pronephros, lateral plate, and early vasculature. J Embrol Exp Morphol. 1980;55:291–306.
Meier SP. Morphogenesis of the chick embryo mesoblast: morphogenesis of the prechordal plate and early vasculature. J Embryol Exp Morphol. 1981;83:49–61.
Meier SP. The development of segmentation in the cranial region of vertebrate embryos. Scan Electron Microsc. 1982;3:1269–82.
Meier SP. The distribution of cranial neural crest cells during ocular morphogenesis. In: Daentl DA, editor. Clinical, structural and biochemical advances in hereditary eye research. New York: Alan R. Liss; 1982. p. 1–15.
Meier SP, Tam PPL. Metameric pattern in the embryonic axes of the mouse. I. Differentiation of the cranial region. Differentiation. 1982;21:95–108.
Meier SP. Somite formation and its relationship to metameric patterning of the mesoderm. Cell Differ. 1984;14:235–43.
Jacobson AG. Somitomeres: the primordial body segments. In: Bellairs R, Ede DA, Lash JW, editors. Somites in developing embryos. New York: Plenum Publ.; 1986.
Jacobson AG. Somitomeres: mesodermal segments of the head and neck. In: Hanken J, Hall BK, editors. The skull, Development, vol. I. Chicago: University of Chicago Press; 1993.
Huang R, Zhi Q, Ordahl CO, Christ B. The fate of the first avian somite. Anat Embryol. 1997;195:435–49.
Huang R, Zhi Q, Patel K, Wilting J, Christ b. Contribution of single somites to the skeleton and muscles of the occipital and cervical regions in avian embryos. Anat Embryol. 2000;202:375–83.
Hunter RM. The development of the anterior occipital somites in the rabbit. J Morphol. 1935;57:501–31.
Sarnat HB, Netsky MG. Evolution of the nervous system. 2nd ed. London: Oxford University Press; 1981. ISBN: 978-0195027761.
Noden DM. Origins of avian ocular and periocular tissues. Exp Eye Res. 1979;29:27–43.
Noden DM. The embryonic origins of avian cephalic and cervical muscles and associated connective tissues. Am J Anat. 1983;168:257–76.
Noden DM. The role of neural crest in patterning of avian cranial, skeletal, connective, and muscle tissues. Dev Biol. 1983;96:347–56.
Noden DM. Patterning of avian craniofacial muscles. Dev Biol. 1986;116:347–56.
Noden DM. Interactions and fates of avian craniofacial mesenchyme. Development (Suppl). 1988;103:121–40.
Noden DM. Origins and patterns of craniofacial mesenchymal tissues. J Craniofac Genet Dev Biol. 1991;11:192–213.
Bock WJ. The avian skeletomuscular system. Avian Biol. 1974;4:1119–257.
Meier SP. Development of the chick mesoblast: morphogenesis of the prechordal plate and cranial segments. Dev Biol. 1981;83:49–61.
Schoenwolf G. Cell movements in the epiblast during gastrulation and neurulation in avian embryos. In: Keller R, Clark Jr WH, Griffin F, editors. Gastrulation: movements, patterns, and molecules. New York: Plenum; 1991. p. 1–28.
Schoenwolf G. Neurulation: coming to closure. Trends Neurosci. 1997;20:510–7.
Larsen WJ. Human embryology. 2nd ed. New York: Churchill Livingstone; 1997. p. 49–71.
Christ B, et al. Segmentation of the vertebrate body. Anat Embryol. 1998;197:1–8.
DeRobertis EM, Oliver G, Wright CVE. Homeobox genes and the vertebrate body plan. Sci Am. 1990;263:46–52.
Kardong KV. Vertebrates: comparative anatomy, function, evolution. 3rd ed. New York: McGraw Hill; 2002.
Lobe CG. Transcription factors and mammalian development. Curr Topics Dev Biol. 1992;27:57–63.
Duboule D, editor. Guide to the homeobox genes. Oxford: Oxford University Press; 1994.
Hunt P, Krumlauf R. Deciphering the Hox code: clues to patterning branchial regions of the head. Cell. 1991;66:1075–8.
Hunt P, Krumlauf R. Hox codes and positional specification in vertebrate embryonic axes. Annu Rev Cell Biol. 1992;8:227–56.
Keynes R, Lumsden A. Segmentation and the origin of regional diversity in the vertebrate nervous system. Neuron. 1990;2:1–9.
Krumlauf P. Hox genes and pattern formation in the branchial region of the vertebrate head. Trends Genet. 1993;9:106–12.
Lumsden A, Krumlauf R. Patterning the vertebrate neuraxis. Science. 1996;274:1109–15.
Mavilio F. Regulation of vertebrate homeobox-containing genes by morphogens. Eur J Biochem. 1993;212:273–88.
Mcginnis W, Krumlauf R. Homeobox genes and axial patterning. Cell. 1992;68:283–302.
Puelles L, Rubenstein JLR. In: Rossant J, Tam PPL, editors. Mouse development: patterning, morphogenesis, and organogenesis. San Diego: Academic Press; 2002. p. 37–54.
Scott MP. Vertebrate homeobox nomenclature. Cell. 1992;71:551–3.
Stein S, Fritsch R, Lenmire L, Kessel M. Checklist: vertebrate homeobox genes. Mech Dev. 1996;55:91–108.
Kessel M. Respecification of vertebral identities by retinoic acid. Development. 1992;118:487–501.
Kessel M, Balling R, Gruss P. Variations of cervical vertebrae after expression of a Hox 1.1 transgene in mice. Cell. 1990;61:301–8.
Kessel M, Gruss P. Homeotic transformations of muringe vertebrae and concomitant alteration of Hox codes induced by retinoic acid. Cell. 1991;67:89–104.
Morris-Kay G. Retinoic acid and development. Pathobiology. 1992;60:264–70.
Jenkins FA Jr. The evolution and development of the dens of the mammalian axis. Anat Rec. 1969;164:174–84.
Kemp TS. The atlas-axis complex of the mammal-like reptiles. J Zool (London). 1969;159:223–48.
Wehr R, Gruss P. Pax and vertebrate development. Int J Dev Biol. 1996;40:369–77.
Kingsley DM. The TGF-B superfamily: new members, new receptors, and new genetic tests of function in different organisms. Genes Dev. 1994;8:133–46.
Wilkie AOM, et al. Function of FGFs and their receptors. Curr Biol. 1995;5:500–7.
Ahlgren SC, Bonner-Fraser M. Inhibition of Sonic hedgehog signaling in vivo results in craniofacial neural crest cell death. Curr Biol. 1999;9:1304–14.
Bellon F. Identification of Sonic hedgehog as a candidate gene responsible for holoprosencephaly. Nat Genet. 1996;14:353–6.
Helms JA, Kim CH, Hu D, Minkoff R, Thaller C, Eichele G. Sonic hedgehog participates in craniofacial morphogenesis and is down-regulated by teratogenic doses of retinoic acid. Dev Biol. 1997;187:25–35.
Hu D, Helms JA. The role of Sonic hedgehog in normal and abnormal craniofacial morphogenesis. Development. 1999;126:4873–84.
Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signal regulate multiple aspects of gastrointestinal development. Development. 2000;127:2763–72.
Roberts DJ, et al. Sonic hedgehog is an andodermal signal inducing BMP-4 and Hox genes during induction and regionalization of the chick hindgut. Development. 1995;121:3163–74.
Sukegawa A, Narita T, Kameda T, Saitoh K, Nohna T, Iba H, Yasugi S, Fukuda K. The concentric structure of the developing gut is regulated by Sonic hedgehog derived from endodermal epithelium. Development. 2000;127:1971–80.
Azar Y, Eyal-Giladi H. Marginal zone cells: the primitive streak-inducing component of the primary hyoblast in the chick. J Embryol Exp Morphol. 1979;52:79–88.
Eyal-Giladi H. The establishment of the axis in chordates: facts and speculations. Development. 1997;124:2285–890.
Blechschmidt E. The human embryo. Stuttgart: Schattauer; 1964.
Blechschmidt E. The beginnings of life. New York: Springer-Verlag; 1977.
Lemire L, Kessel M. Gastrulation and homeobox genes in chick embryos. Mech Dev. 1997;67:3–16.
Schoenwolf G, Garcia-Martinez V, Dias MS. Mesoderm movement and fate during avian gastrulation and neurulation. Dev Dyn. 1992;193:235–48.
Tam PPL, Beddington RSP. The formation of mesodermal tissues in the mouse embryo during gastrulation and early organogenesis. Development. 1987;99:109–26.
Tam PPL, Beddington RSP. Establishment and organization of germ layers in the gastrulating mouse embryo. In: Postimplantation development in the mouse. Ciba Found Symp. 1992;165:27–49.
Tam PPL, Williams EA, Chan WY. Gastrulation in the mouse embryo: ultrastructural and molecular aspects of germ layer morphogenesis. Microsc Res Tech. 1993;26:301–28.
Tam PPL, Zhou SX. The allocation of epiblast cells to ectodermal and germ-line lineage is influenced by the position of the cells in the gastrulating mouse embryo. Dev Biol. 1996;178:124–32.
Tam PPL, Parameswaran M, Kinder SJ, Weinberger RP. The allocation of epiblast cells to the embryonic heart and other mesodermal lineages: the role of ingression and tissue movement during gastrulation. Development. 1997;124:1631–42.
Noden DM. Vertebrate craniofacial development: the relation between ontogenetic process and morphological outcome. Brain Behav Evol. 1991;38:190–225.
Rhinn M, Brand M. The midbrain-hindbrain boundary organizer. Curr Opin Neurobiol. 2001;11:34–42.
Hall BK. The neural crest in development and evolution. New York: Academic Press; 1997.
Le Douarin NM, Kalcheim C. The neural crest. 2nd ed. Cambridge: Cambridge University Press; 1999.
Cobos I, et al. Fate map of the avian anterior forebrain at the 4 somite stage, based on the analysis of quail-chick chimeras. Dev Biol. 2001;239:46–67.
Couly GF, LeDourain NM. Mapping of the early neural primordium in quail-chick chimeras I. Developmental relationships between placodes, facial ectoderm, and prosencephalon. Dev Biol. 1985;110:422–39.
Couly GF, LeDourain NM. Mapping of the early neurala primordium in quail-chick chimeras II. The prosencephalic neural plate and neural folds: implications for the genesis of cephalic human congenital abnormalities. Dev Biol. 1987;120:198–214.
Couly GF, Le Douarin NM. Head morphogenesis in embryonic avian chimeras: evidence for a segmental pattern in the ectoderm corresponding to the neuromeres. Development. 1990;108:543–58.
Couly GF, Coulty PM, Le Douarin NM. The developmental fate of the cephalic mesoderm in quail-chick chimeras. Development. 1992;114:1–15.
Couly GF, Grapin-Botton PM, Coultey P, Le Douarin NM. The regeneration of the cephalic neural crest, a problem revisited: the regenerating cells originate from the contralateral or from the anterior and posterior neural fold. Development. 1996;122:3393–407.
Couly G, Grapin-Botton A, Coulty PM, Le Douarin NM. Determination of the identity of the derivatives of the cephalic neural crest: incompatibility between Hox gene expression and lower jaw development. Development. 1998;125:3445–59.
Fernandez-Garre P, et al. Fate map of the chicken neural plate at stage HH4. Development. 2002;129:2807–22.
Le Douarin NM, Catala M, Batini C. Embryonic neural chimeras in the study of vertebrate brain and head development. Int Rev Cytol. 1997;175:241–309.
La GG. dysplasie olfacto-genitale: agenesie des lobes olfactifs avec absence de development gonadique a la puberte. Acta Neuroveg. 1960;21:345–94.
Rugarli EI, Ballabio A. Kallman syndrome: from genetics to neurobiology. JAMA. 1993;270:2713–6.
Lieblich JM, Rogol AD, White BJ, Rosen SW. Syndrome of anosmia with hypogonadic hypogonadism (Kallman Syndrome): clinical and laboratory studies in 23 cases. Am J Med. 1982;73:506–19.
Molsted K, Kjaer I, Giwercman A, Vesterhauge S, Shakkbarek NE. Craniofacial morphology in patients with Kallman’s syndrome with and without cleft lip. Cleft Palate Craniofac J. 1197;34:417–34.
de San Juan AM. Teratologia: falta total de los nervfos olfatorios con anosmia en un individuo en quien existia una atrofia congenita de los testiculos y el miembro viril. El Siglo Med. 1856;3:211.
Serbedzija GN, Fraser SE, Bronner-Fraser M. Pathways of neural crest migration in the mouse embryo as revealed by vital dye labeling. Development. 1990;108:605–12.
Serbedzija GN, Bonner-Fraser M, Fraser SE. Vital dye analysis of cranial neural crest migration in the mouse embryo. Development. 1992;116:297–307.
Tan SS, Morriss-Kay GM. The development and distribution of the cranial neural crest in the rat embryo. Cell Tissue Res. 1985;240:403–16.
Tan SS, Morriss-Kay GM. Analysis of cranial neural crest cell migration and early fates in post-implantation rat chimeras. J Embryol Exp Morphol. 1986;98:21–58.
Gui T, Osama-Yamashita N, Eto K. Proliferation of nasal epithelia and mesenchymal cells during primary palate formation. J Craniofac Genet Dev Biol. 1993;13:250–8.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Carstens, M.H. (2023). Neuromeric Organization of the Head and Neck. In: Carstens, M.H. (eds) The Embryologic Basis of Craniofacial Structure. Springer, Cham. https://doi.org/10.1007/978-3-031-15636-6_1
Download citation
DOI: https://doi.org/10.1007/978-3-031-15636-6_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-15635-9
Online ISBN: 978-3-031-15636-6
eBook Packages: MedicineMedicine (R0)