The number of vertebrae, and hence the number of segments or 'somites' in the body, is highly variable among different vertebrate species. For instance, frogs have 10 vertebrae, while many snakes have over 300. But what controls vertebra number in a given species and why does it vary so much between species? Gomez et al. propose that the number depends on a balance struck early in embryogenesis between the division of the body into somites and the overall rate of development. They establish this by showing snakes have a much greater segmentation clock speed, relative to embryo development as a whole, than lizards and other vertebrates with fewer somites. The vertebrate body axis is divided into a number of segments or 'somites', such as the number of vertebrae. But what controls vertebra number, and its variation between species? This paper postulates that the number depends on a balance between the division of the body into somites (the segmentation clock rate) and the overall rate of development, as established by showing that the large number of vertebrae in snakes comes from a much greater segmentation clock speed in snakes, relative to development as a whole, than in other vertebrates. The vertebrate body axis is subdivided into repeated segments, best exemplified by the vertebrae that derive from embryonic somites. The number of somites is precisely defined for any given species but varies widely from one species to another. To determine the mechanism controlling somite number, we have compared somitogenesis in zebrafish, chicken, mouse and corn snake embryos. Here we present evidence that in all of these species a similar ‘clock-and-wavefront’1,2,3 mechanism operates to control somitogenesis; in all of them, somitogenesis is brought to an end through a process in which the presomitic mesoderm, having first increased in size, gradually shrinks until it is exhausted, terminating somite formation. In snake embryos, however, the segmentation clock rate is much faster relative to developmental rate than in other amniotes, leading to a greatly increased number of smaller-sized somites.

Control of segment number in vertebrate embryos

The progressive generation of chick and mouse axial tissues - the spinal cord, skeleton and musculature of the body - has long been proposed to depend on the activity of multipotent stem cells. Here, we evaluate evidence for the existence and multipotency of axial stem cells. We show that although the data strongly support their existence, there is little definitive information about their multipotency or extent of contribution to the axis. We also review the location and molecular characteristics of these putative stem cells, along with their evolutionary conservation in vertebrates and the signalling mechanisms that regulate and arrest axis extension.

Stem cells, signals and vertebrate body axis extension.

Although cell death has long been recognized to be a significant element in the process of embryonic morphogenesis, its relationships to differentiation and its mechanisms are only now becoming apparent. This new appreciation has come about not only through advances in the understanding of cell death in parallel immunological and pathological situations, but also through progress in developmental genetics which has revealed the roles played by death in the cell lineages of invertebrate embryos. In this review, we discuss programmed cell death as it is understood in developmental situations, and its relationship to apoptosis. We describe the morphological and biochemical features of apoptosis, and some methods for its detection in tissues. The occurrence of programmed cell death during invertebrate development is reviewed, as well as selected examples in vertebrate development. In particular, we discuss cell death in the early vertebrate embryo, in limb development, and in the nervous system.

Programmed Cell Death in Development

The outgrowth of the vertebrate tail is thought to involve the proliferation of regionalised stem/progenitor cell populations formed during gastrulation. To follow these populations over extended periods, we used cells from GFP-positive transgenic chick embryos as a source for donor tissue in grafting experiments. We determined that resident progenitor cell populations are localised in the chicken tail bud. One population, which is located in the chordoneural hinge (CNH), contributes descendants to the paraxial mesoderm, notochord and neural tube, and is serially transplantable between embryos. A second population of mesodermal progenitor cells is located in a separate dorsoposterior region of the tail bud, and a corresponding population is present in the mouse tail bud. Using heterotopic transplantations, we show that the fate of CNH cells depends on their environment within the tail bud. Furthermore, we show that the anteroposterior identity of tail bud progenitor cells can be reset by heterochronic transplantation to the node region of gastrula-stage chicken embryos.

/pdf/localised-axial-progenitor-cell-populations-in-the-avian-11dqqy1cz6.pdf

Localised axial progenitor cell populations in the avian tail bud are not committed to a posterior Hox identity

The endogenous mechanism that determines vertebrate body length is unknown but must involve loss of chordo-neural-hinge (CNH)/axial stem cells and mesoderm progenitors in the tailbud. In early embryos, Fibroblast growth factor (FGF) maintains a cell pool that progressively generates the body and differentiation onset is driven by retinoid repression of FGF signalling. This raises the possibility that FGF maintains key tailbud cell populations and that rising retinoid activity underlies cessation of body axis elongation. Here we show that sudden loss of the mesodermal gene (Brachyury) from CNH and the mesoderm progenitor domain correlates with FGF signalling decline in the late chick tailbud. This is accompanied by expansion of neural gene expression and a similar change in cell fate markers is apparent in the human tailbud. Fate mapping of chick tailbud further revealed that spread of neural gene expression results from continued ingression of CNH-derived cells into the position of the mesoderm progenitor domain. Using gain and loss of function approaches in vitro and in vivo, we then show that attenuation of FGF/Erk signalling mediates this loss of Brachyury upstream of Wnt signalling, while high-level FGF maintains Brachyury and can induce ectopic CNH-like cell foci. We further demonstrate a rise in endogenous retinoid signalling in the tailbud and show that here FGF no longer opposes retinoid synthesis and activity. Furthermore, reduction of retinoid signalling at late stages elevated FGF activity and ectopically maintained mesodermal gene expression, implicating endogenous retinoid signalling in loss of mesoderm identity. Finally, axis termination is concluded by local cell death, which is reduced by blocking retinoid signalling, but involves an FGFR-independent mechanism. We propose that cessation of body elongation involves loss of FGF-dependent mesoderm identity in late stage tailbud and provide evidence that rising endogenous retinoid activity mediates this step and ultimately promotes cell death in chick tailbud.

/pdf/loss-of-fgf-dependent-mesoderm-identity-and-rise-of-36zp19s9vu.pdf

Loss of FGF-Dependent Mesoderm Identity and Rise of Endogenous Retinoid Signalling Determine Cessation of Body Axis Elongation

In the chick embryo, the tail bud reaches its maximum length at about stage 22 of Hamburger and Hamilton, after which it starts to regress By this stage the neural tube and notochord extend right to the tip of the tail, but the somites do not do so, the terminal tail bud mesoderm never becoming segmented The investigation is concerned with analysing why this mesoderm fails to segment When tail buds were explanted to the chorio-allantoic membrane, they continued to form somites only until the "correct" number had segmented, ie, the tail bud formed no more somites when isolated from the embryo than it would have formed if undisturbed Morphological studies suggest that in the normal embryo massive cell death overtakes the tail bud mesoderm before it can segment It is suggested therefore that cell death may be a contributory factor in preventing segmentation

An experimental and morphological analysis of the tail bud mesenchyme of the chick embryo

In the chick embryo the final number of somites is achieved at about stage 22 of Hamburger and Hamilton. By this time the neural tube and notochord have reached the tip of the tail bud but some paraxial mesoderm remains unsegmented. In this study using scanning electron microscopy we show that somitomeres are present in this mesoderm. This indicates that the terminal paraxial mesoderm of the tail bud may have the potential to form supplementary somites, though we do not as yet know what prohibits the completion of segmentation to the tip.

E. J. Sanders

Papers

An experimental and morphological analysis of the tail bud mesenchyme of the chick embryo

Somitomeres in the chick tail bud: an SEM study.