Q J Med 2002; 95: 193-194
© 2002 Association of Physicians
Biologic |
Get ahead!
The development of the head, jaws and neck are complex processes requiring the integration of mesodermal and ectodermal growth, the control of morphogenetic movement and the formation of the neural crest. In evolutionary terms, one might suppose that these activities depended on progressively acquired functions, with an increasingly complex scheme of genetic activity. But like so much of development, making a head requires inhibitions; by default, we form a metameric (segmented) trunk like the middle of an earthworm, rather than a head. This process has to be stopped, and the way in which this is brought about illustrates some fundamental points about differentiation.
The head cannot begin to develop until antero-posterior and dorsoventral axes are formed, and this requires the suppression of the function of a number of genes. The acquisition of a dorsoventral axis requires further inhibitory actions, mainly by bone morphogenetic protein (BMP) inhibitors. BMPs are highly conserved and involved in dorsoventral patterning in all vertebrates and many invertebrates; they have been found in every animal so far sequenced, from worms to humans. They are ventralizing—only in their absence, or when they are inhibited, do dorsal structures form.
So how does the regional development of a vertebrate body begin? The first cells that migrate through the primitive streak (having lost their E-cadherin and become detached from their neighbours) will become the pharyngeal endoderm. The next group does not move so far anteriorly; they form a mass that will give rise to the prechordal plate and the head mesenchyme, raised up in the mid-line. The next group will form the notochord as they move forward to meet the prechordal plate. In brief (and not entirely accurately) the forebrain and midbrain come from the cells forward of the node and the hindbrain and trunk from cells behind it.
At the onset of these events, the organizing region expresses a number of genes that produce proteins that prevent the expression of signalling factors. These genes include Noggin, Chordin, Crescent and Dickkopf, which are all BMP inhibitors. Mice lacking Chordin and Noggin, for example, develop without forebrain, nose and facial structures. Cerberus, a specific BMP inhibitor, is part of the mechanism that prevents trunk-like development, allowing the development of a head. These events are early; the establishment of the major axes is determined by maternally active genes in pre-blastocyst stages of mammalian development.1 Gardner2 has suggested that the elliptical form of the mouse four-cell-stage conceptus is the first indication of the establishment of an animal-vegetal axis.
Following this phase, some large blocks of anterior or posterior and dorsal or ventral tissues arise, and more repressor functions act after the establishment of this primary metameric pattern. Activation of the so-called ‘gap’ genes (which produce ‘gaps’ in the regular pattern of segmentation) allow the specialized development of particular segments.3 The gap genes have to be continuously active throughout development, and the selector genes that set up the programme are generally maintained in operation by other genes that also work by repressing inhibitors. These changes to the segmental pattern are further refined by interactions between the gap genes that modify the simple posterior/anterior divisions into zones of specific gene expression. ‘Pair-rule’ genes, genes that control two segment units (in flies the pair-rule genes act, in some mutants, to produce an embryo with every other segment missing) then direct differentiation within the segments (see below). In this way, the regional identity of the brain is established along its anterior posterior axis after interactions between mesoderm and ectoderm. The anterior part of the developing neural tissue is established as a result of inhibition of Wnt and BMPs by specific antagonists. This diminishes posteriorly, and Wnt expression gradually imposes a more posterior fate. Ectopic activation of Wnt can inhibit formation of the head,4 and overexpression of antagonists can induce secondary heads along the axis.5
Since the development of the head depends on the suppression of the normal segmental pattern of development, it is interesting to look at how things are managed in the early chordates, where organized development of the anterior end of the organism is more variable. Some of this group develop in a way that suggest they have decided to do without a head—the precursor forms have a notochord but do not make much use of it, phylogenetically speaking. Homeobox genes act on the ‘determined’ head process to form the primary divisions of the forebrain, midbrain and hindbrain and pattern the rhombomeres (the segmented, non-somitic units of mid and hindbrain development). Comparison of homeobox (Hox) genes between vertebrates and their closely related invertebrates such as Amphioxus shows two major changes: duplication of the complexes and amplification of their roles and expression. The homeobox genes have shown little structural change over 500 million years of evolution, and retinoic acid signalling is a central mechanism controlling their action. Amphioxus has a notochord, paired somites and a dorsal neural tube, but not much of a head. Its homeobox genes are homologues of the first ten groups of vertebrate Hox genes, but further duplications were necessary before the neural crest and hence (and this is a generous and interpretive ‘hence’) jawed vertebrates appeared.
Extensive study has shown that the Hox gene domains shift in parallel with vertebrate anatomy. Although segmental identities may vary, the plan is uniform: the forelimb develops at the cervical to thoracic transition, and the hindlimb at the lumbosacral transition, regardless of variation in the number of segments in these regions. The extent of this variation is evident in the fishes. Cartilaginous fishes may have several hundred vertebrae; in teliosts (bony fish), a typical number is 48. Long-bodied teliosts (eels) may have 200, snakes have hundreds, and birds are in the range 37–53. Dolphins have a greater variability in number (up to 100) than terrestrial vertebrates, who generally have a stable and consistent number of presacral vertebrae, although the number in the tail may vary from three to 47. The anterior boundaries of expression of members of the HoxD family mark the transitions between a number of boundaries of vertebral development (lumbar/sacral HoxD 9–10, sacral/caudal HoxD 11–12) and HoxC 6 defines the cervical/thoracic transition. This is true even if these transitions are in different somitic positions in different vertebrates. Carroll has pointed out that (in the mouse) the thoracic/lumbar transition which is associated with expression of HoxA 9, HoxB 9 and HoxC 9 is not accompanied by HoxD 9 expression. He makes the point that the thoracic/lumbar transition is not general among tetrapods, and that this change in anatomy of the trunk may underlie the evolution of the tetrapods from the fish—again, the modification of a standard pattern.
In Amphioxus, a homologue of the engrailed gene (a pair rule gene) is expressed in the posterior half of each of the first eight segments (remember, the animal has a notochord but no vertebrae). It is thus reasonable to assume that the Urbilateria (the ancestors of the protostomes and deuterostomes) were segmented. Engrailed homologues in the zebrafish, chick and mouse, are expressed in subsets of cells in the somites after their formation but the Amphioxus homologue forms bands of expression before any morphological signs of segmentation. This may be the first establishment of a mechanism for specialization at the ‘head end’. The first eight segments in Ampioxus form differently from the remaining 40 or so; they arise from the pinching out of pockets from the gut, while the others form in a manner closely resembling their origin in vertebrates (by pinching off from a block of mesenchyme). Genes from Amphioxus will operate in vertebrate embryos to pattern the rhombomeres—the segmental units of the developing hindbrain. Experiments by Manzanares et al.6 have shown that genes from this animal can pattern the neural tube, neural crest cells and placode derivatives in higher vertebrates, a compelling statement of conservation of function. As Amphioxus does not have migratory neural crest cells or placodes, it must be that either a precursor population of the neural crest exists in these animals, or that the duplication of the homeobox genes that is a feature of the vertebrate genome allowed the development of this population.
More about the face and jaws another time.
References
1. Berry CL. Making a pushmi-pullyu: the left or right handed model. Paed Devel Path1999; 2:491–6.
2. Gardner RL. Specification of embryonic axes begins before cleavage in normal mouse development. Development2001; 128:839–47.[Abstract]
3. Muller J, Gaunt S, Lawrence PA. Function of the polycombe protein is conserved in mice and flies. Development1999; 121:2847–52.[Abstract]
4. Kim CH, Oda T, Itoh M, et al. Repressor activity of Headless/Tcf3 is essential for vertebrate head formation. Nature2000; 407:913–16.[Medline]
5. Sirotkin HI, Dougan ST, Schier AF, Talbot WS. Bozozok and squint act in parallel to specify dorsal mesoderm and anterior neuroectoderm in zebrafish. Development2000; 127:2583–92.[Abstract]
6. Manzanares M, Wada H, Itasaki N, Trainor PA, Krumlauf R, Holland PWH. Conservation and elaboration of Hox gene regulation during evolution of the vertebrate head. Nature2000; 408:854–7.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Berry Variety QJM, April 1, 2002; 95(4): 259 - 260. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||