Neural induction 1 Lecture 1

The fertilized egg, or zygote, of an animal is the single cell from which all tissues will develop. Here, we will focus primarily on the amphibian Xenopus laevis, which has fruitfully served as an experimental laboratory animal for many decades to investigate the embryonic development of an animal. The early stages of embryonic development are fairly similar across vastly divergent animal groups. Firstly, repeated cell divisions of the zygote form the blastula. As more and more cells are generated within the blastula, a cavity devoid of cells opens (the blastocoel) which sports a hole: the blastopore. At this stage, there is a single-cell layer that wraps around all, most, or some of the embryo (depending on the animal group, and generally depending on how large the yolk reserves are). Gastrulation is the developmental stage at which the three main embryonic layers form: ectoderm, mesoderm and endoderm. The animal pole side becomes theectoderm, and the vegetal pole the endoderm. Cells delaminating from the boundary between the ectoderm and endoderm migrate and enter the cavity of the embryo through the blastopore, forming the mesoderm. Experimental embryology was for much of the XX century the approach of choice to investigate the nature of embryonic development. Early experiments used slicing techniques to isolate chunks of the early embryo, and grafting techniques to transplant tissue from one embryo to another, often at a different place or time of development. An early experiment in Xenopus consisted in isolating the animal cap before or after gastrulation, and then observe the fate of cells in the isolated animal cap. Before gastrulation, the isolated animal cap develops into epidermis. After gastrulation, the isolated animal cap develops into neural tissue. During gastrulation, as the mesoderm organizes into a layer and folds under the ectoderm (in the involuting marginal zone or IMZ), the mesoderm releases neural inducers (noggin, chordin, follistatin) and the nearby ectoderm acquires a neural identity, forming the neurogenic region: a region of the ectoderm near the blastopore from which the entire nervous system will develop. Grafting experiments by Spemann and Mangold discovered early on (in 1924!) that a small region near the blastopore held unique powers of induction. By grafting a small patch of tissue around the blastopore from a pigmented donor embryo onto a region far from the blastopore of a non-pigmented host embryo, the induction of not one but two body axes, each with its own neural tube, was observed. Furthermore, the secondary body axis was not only formed of pigmented cells from the donor, but also non-pigmented ones from the host, indicating that the donor cells had the capability to organize the host cells into forming a new body axis. Hence, the ”Spemann and Mangold organizer” was identified. Now into the molecular era, similar embryo surgery experimental setups were used to test the neural inducer capabilities of various compounds. Two kinds of inductions on sliced animal caps were observed: (1) indirect, where a mesoderm was induced which then secondarily induced neural tissue; and (2) direct, where the animal cap acquired neural identity. 1 NST IB Neurobiology – Neural determination Lecture 1 – Neural induction Experiments by Harland et al. in whole Xenopus embryos proved to be an additional excellent test bench for testing inducer molecules. Exposing the pre-gastrulation embryo to UV light resulted in a ventralized embryo lacking neural structures. On the other hand, treatment with lithium (Li+) hyperdorsalized the embryo, presenting oversized neural structures at the expense of reduced ventral structures. The extraction of messenger RNA (mRNA) from a lithium-treated embryo and its application to an UVtreated embryo rescued the ventralization and the embryo developed somewhat normally. Follow up experiments isolated messenger RNA (mRNA, in the form of cDNA) from the organizer region around the blastopore and, by cloning the cDNA into a library, amplifying each individual type of cDNA, then each cDNA could be tested for neural induction properties. In this way, the experimental rescue of the UV-treated embryo could be reproduced with a single cDNA, that of the noggin gene, which was demonstrated sufficient. The expression of the noggin protein and its application to an isolated animal cap led to the direct induction of the ectoderm to neural fate. In addition to noggin, further genes were found to have neural inducer capabilities. Sasai et al. 1994 reported on the identification of chordin, a newly identified gene from a screen (i.e., the testing of many possible genes, in this case from a pool of cDNA extracted from animal caps) that encoded an extracellular protein with potent axis-forming activities, including the ability to recruit neighbouring cells into secondary axes. The third gene is follistatin: the Follistatin protein binds to Activin, preventing it from binding to the Activin Receptor– a transmembrane protein whose intracellular domain is responsible for signal transduction that ultimately leads to the activation of additional genes that effect the induction of epidermal cell fate. A truncated Activin Receptor, incapable of signal transduction, effectively prevents epidermal cell fate, inducing then the neural cell fate. This was demonstrated by using isolated blastula animal caps that, when left alone, develop into epidermis. But when injected with the mRNA for the truncated Activin Receptor, then the neural cell fate is induced. Many more genes participate of neural cell fate induction. Experiments with dissociated animal caps–that is, animal caps are sliced out and then their cells separated with each other by manipulating the level of Ca2+ in the culture medium–showed that, when dissociated, animal cap cells differentiate into neurons. But dissociated animal cap cells exposed to the BMP4 protein differentiate to epidermal cells instead, just like the intact, non-dissociated animal cap. Taken together, the current model of neural induction in the early embryo is the following. In the extracellular matrix, BMP proteins (or their homologues in invertebrates) bind to the BMP receptors and, via the phosphorilation of the Smad intracellular protein, inhibit the transcription of the Zic1 gene, promoting then epidermal cell fate. But the release by the mesoderm of noggin, chordin or follistatin proteins which diffuse through the extracellular matrix of the animal cap ectoderm then results in the blocking of BMP, which then cannot bind to the BMP receptor, and therefore Smad isn’t phosphorilated, and the Zic1 gene is transcribed. Together with the FGF protein in the extracellular matrix which, via the FGF receptor, activates the ERK protein which in turn activates the transcription of the Zic3 gene, then Zic1 and Zic3 together promote the transcription of neural genes and the induction of neural cell fate (see summary in Fig. 1.22 of the Sanes book). The genetic program for neural induction is evolutionarily ancient. Homologues for the genes described above for neural induction in vertebrates exist in the invertebrate genome. In particular, the dpp (”decapentaplegic”) gene identified in Drosophila is a homologue of the BMP genes, and the sog (”short gastrulation”) gene is a homologue of chordin. In the vinegar fly Drosophila, the blastula is similarly organized as an external epithelium with a region where dpp expression dominates (the future dorsal side of the animal) and another region where sog dominates (the ventral side) that will get folded in (invaginated) to give rise to the mesoderm. Regions adjacent to invaginated mesoderm sit between the dpp (the BMP homolog) and sog (the chordin homolog) domains and, like in vertebrates, become a neurogenic region that will give rise to the nervous system. Interestingly, in insects, and more generally in invertebrates, the ectodermal region that will give rise to the nervous system is ventral, whereas in vertebrates is dorsal. This dorso-ventral inversion was 2 NST IB Neurobiology – Neural determination Lecture 1 – Neural induction famously noted in the 18th century by Geoffroy St-Hillaire when eating a lobster. As there are many similarities between insects and mammals, so there are many differences. In Drosophila, once gastrulation occurs and the mesoderm takes its place inside the embryo, the nervous system is seeded by delaminating neuroblasts: that is, specific cells within the neuroectoderm (the region of the ectoderm that acquires a neural fate) detach from the ephithelium and start stereotyped sequences of asymmetric cell divisions that give rise to specific complements of neuronal progenitors, which ultimately differentiate into all the neuronal cell types, forming the entire nervous system. Whereas, in amphibians such as Xenopus, in mammals, and more generally in all vertebrates, rather than by delamination, the nervous system forms by the invagination (another in-folding) of the neuroectoderm which forms the neural tube. A few cells from the dorsal neural tube, though, do delaminate: these are the neural crest cells. And the notochord, a mesodermal structure that runs the ventral length of the neural tube, releases retinoic acid and Shh which induce the formation of the ventral plate in the ventral side of the neural tube (see Lecture 2). Above, most details of neural induction in the blastula and gastrula stages of embryonic development were explained on the basis of experiments and details of the amphibian Xenopus, a laboratory animal very suitable for experimental embryology given its dimensions and characteristics. Mammals, though, present notable differences. In mammals, blastula’s blastopore takes the form of a slit named the primitive streak (which ends in the primitive pit). Through the primitive streak, migrating epiblasts (the undifferentiated cells on the surface of the blastula) enter the blastocoel (the space inside the blastula) and are induced as the endoderm. A second wave of migration through the primitive streak form the mesoderm internally. (See the Saladin et al. book for details, at Chapter 4, Human Development, subsection on Stages of Prenatal Development.) Once the 3 embryonic layers (ectoderm, mesoderm and endoderm) have formed, events proceed in the mammalian embryonic development overall quite similarly to the amphibian embryo, with an invaginating neural tube over a mesodermal notochord and the delaminating neural crest cells at the interface between the fully invaginated neural tube and the dorsal epidermis. 3