A 500 million-year-old neural induction pathway

A study by a team of biologists at the University of California, San Diego suggests that a signalling pathway which regulates patterning of the nervous system during development evolved over 500 million years ago from the common ancestor shared by invertebrates and vertebrates.

In a paper published today in PLoS Biology, Claudia Mizutani and her colleagues report that the mechanism by which the bone morphogenetic protein (BMP) pathway specifies the identity of neural cell types along the dorso-ventral (D-V) axis of the developing nervous system is highly evolutionarily conserved, occurring in the same way in fruit flies as it does in chicks.

BMPs are a family of proteins, including Decapentaplegic (Dpp), Dorsal and (in humans and other vertebrates) Sonic Hedgehog, which are involved in neural induction, the process by which neural cells differentiate from ectoderm during embryogenesis. BMPs are morphogens, secreted proteins whose local concentrations provide positional information that determines the fate of cells during development. The concentration gradients of morphogens are involved in the subdivision of the neuroectoderm into regions containing specified cell types, with each cell type being specified by a particular concentration of a given morphogen.

BMPs are transcription factors, or proteins that alter the pattern of gene expression in cells which are sensitive to them. The repertoire of genes expressed by a cell determines the differentiation pathway along which that cell will proceed, with each pathway giving rise to a specific type of neuron. In the vertebrate nervous system, for example, graded BMP signalling generates sensory neurons in the dorsal spinal cord and motor neurons in the ventral spinal cord.

mizutani.jpgThe image on the left, taken from the PLoS paper, shows the normal expression domains of vnd, ind, and msh in the developing nervous system of the fruit fly Drosophila melanogaster. These genes are expressed in three adjacent stripes in the developing nerve cord, with the stripe pattern being regulated by a concentration gradient of the Dorsal protein.

“Because of the dominant role of the gradient of Dorsal protein, it has not been possible to directly test the role of BMPs in patterning nerve tissue in fruit flies,” says Mizutani. “We genetically reconstruct[ed] embryos that had a uniform concentration of Dorsal throughout. Then we could examine how neural patterning was affected by a BMP gradient.”

In order to determine the effects of Dpp in neural induction, Mizutani and her colleagues created a strain of fruit flies with a uniform concentration of Dorsal in the D-V axis, and a narrow longitudinal stripe of Dpp expression. By inactivating the other genes involved in D-V axis patterning, and expressing Dpp in a well-defined region where it would not normally be present, the effects of Dpp could be examined. A technique called fluorescent multiplex in situ hybridization, developed by the authors, was used to show that the expression domains of vnd, ind, and msh were expanded. Thus, it was demonstrated that Dpp signalling alone is sufficient to pattern the D-V axis of the developing fruit fly nervous system.

The actions of Dpp were then investigated in explants of nervous tissue from chick. Sonic hedgehog, the vertebrate homolog of Dorsal, was added to the Petri dishes in which these tissues were grown, so that the concentration gradient of the protein was abolished.

Again, Dpp signalling alone was sufficient to induce the expression of neuronal cell markers in the chick tissues. For example, grafting BMP-expressing cells onto explanted chick nervous tissue was enough to organize a pattern of cell types, with increasingly dorsal cell types being induced in areas with progressively higher BMP concentration.

“We have provided the first evidence for a common role of BMPs in establishing the pattern of gene expression along the dorsal-ventral axis of the nervous system of vertebrates and invertebrates,” says Ethan Bier, senior author of the paper describing the work. The authors speculate that BMP signalling was sufficient to pattern the D-V axis of a simple ancestral organism which existed 500 million years ago, but insufficient to pattern the developing nervous system of more complex organisms which evolved later.

The role of BMPs in D-V patterning of the vertebrate neural tube has been known for over 10 years, so the findings of this study are not at all surprising. It is, however, interesting to ponder the evolutionary origins of neural developmental mechanisms, and it is, perhaps, novel to show that a Dpp can pattern the neural tube on its own.

Related posts:


4 thoughts on “A 500 million-year-old neural induction pathway

  1. Your nice summary captures key aspects of our work, however, I thought I might point out two things. First, BMPs are secreted signals in the TGF-beta superfamily which act via heteromeric transmembrane receptors to alter gene expression. In the case of the fly, this signaling has two major effects. First it activates genes appropriate for development of dorsal non-neural cell types and second, it represses expression of neural genes. This latter repressive function of BMP signaling is thought to be ancestral since it occurs in both flies and vertebrates.

    The second point I would like to raise is that while it is true that BMPs have been known to pattern the neural tube for more than a decade, the prevailing view as illustrated in most textbooks is that they do so by activating expression of genes in the dorsal region of the spinal cord such as the MsX1/2 genes. I have always felt that this idea was a bit counterintutive since earlier in development during the a phase ofter referred to as neural induction BMP signaling represses neural genes in both flies and vertebrates, as mentioned above. So, the oddity was that BMPs were then being proposed to switch the sign of their function during the next developmental stage when the nervous system was being subdivided into three primary domains. What our analysis in flies shows is that the same basic mechanism of BMPs repressing nerual genes is acting both during neural induction and latter during patterning of the nervous system. The only difference is that during the earlier stage the levels of BMP signaling are so high in the non-neural ectoderm that they inhibit expression of all neural genes whereas later during neural patterning the graded BMP levels in the developing nervous system are much lower and hence become limiting. Under these conditions of limiting levels of BMPs some neural genes are more effectively repressed than others. Those most senstive to the repression are expressed further from the BMP source while the less sensitive genes can tolerate higher levels of signaling that prevail closer to the source. By this dosage senstive effect BMPs can help create patternrf gene expression in the nervous system. Our finding that BMPs can organize gene expression in apolar chick neural plate explants suggests that a similar mechanism may acting in vertebrates. Indeed nearly all the published literature regarding the dosage senstive effects of BMPs in the vertebrate nervous system is eaqually consistent with our model in flies (i.e., BMP mediated dosage dependent repression). One of the interresting areas of future research will be to see if the chain of ventral dominant cross inhibition among the neural identity genes vnd, ind, and msh, which helps refine the expression patterns in the fly nervous system also has been conserved in vertebrates. If so, then I think our model will most likely turn out to be correct since the combined action of BMP mediated repression and ventral dominance would account best for the existing data.

  2. Pingback: GNIF Brain Blogger

  3. Pingback: The Synapse, Volume 1, Issue 7 - 52dot’s Blog

Comments are closed.