Control of differentiation onset in the early embryo


1.5 day old embryo

During embryonic development the spinal cord is laid down progressively in a head to tail (rostral to caudal) sequence and is derived entirely from a unique population of epiblast cells that remain at the tail/caudal end. There is evidence that this cell group contains a resident cell population with stem cell-like properties and so it is known as the stem zone. Once cells leave the stem zone they begin to differentiate and as development proceeds the temporal events of differentiation become spatially separated along the extending body axis. So, while stem zone cells all proliferate, some cells in the neural tube now stop dividing and form the first neurons.

We have exploited this sequential nature of spinal cord generation to identify distinct steps in neural differentiation and to investigate the molecular mechanisms that control them. Some of our work focuses on identifying the signalling pathways that direct differentiation. Fibroblast growth factors (Fgfs) including Fgf8 and Fgf4 expressed in the stem zone/tail end keep cells undifferentiated and maintain them in the cell cycle via Notch signalling (Diez del Corral et al. 2002; Akai et al. 2005). FGF signalling also helps to specify the stem zone cell population as it forms in the early embryo and maintains expression of a unique combination of stem zone genes (Delfino-Machin et al. 2005). These include genes characteristic of both early neural and mesodermal tissue. Cells that form mesoderm ingress through a unique tail-end structure called the primitive streak, while cells that will form neural tissue remain in the epiblast layer.

As the body elongates some epiblast cells furthest away from the FGF source begin to differentiate; they cease to express some stem zone genes and turn on new ones that promote neural differentiation and patterning. We have found that signals from the paraxial mesoderm (tissue that lies next to the newly formed spinal cord) provide a signal, Retinoic acid, which both attenuates FGF signalling and drives expression of neuronal and patterning genes (Diez del Corral et al. 2003). Conversely, we have found that FGF signalling prevents the expression of a key enzyme (Raldh2) that mediates Retinoid production in the paraxial mesoderm. The mutual opposition of FGF and Retinoid signalling thus controls and coordinates the onset of differentiation in mesodermal and neural tissue as the body elongates during development.

FGF signalling represses not only neuronal differentiation genes but also the expression of ventral patterning genes (Diez del Corral et al. 2003). These have been shown to act in a combinatorial fashion to specify different types of neurons. A key regulator of ventral patterning genes is the morphogen sonic hedgehog (Shh) and we have shown that onset of this gene in the neuroepithelium is also controlled by FGF signalling. So, opposition of FGF and Retinoid signalling is a general differentiation switch, which regulates onset of neural differentiation and specification of neuronal cell type as well as neuron birth.

Embryonic development is governed by a relatively small number of signalling pathways and the specificity of signalling outcome must therefore rely on unique integration between these pathways in different contexts. We have therefore begun a series of studies to address how further signalling pathways interact with the FGF/Retinoid signalling switch. We have found that canonical Wnt signalling mediates the transition from FGF to Retinoid signalling (Olivera-Martinez and Storey, 2007) and does so in a surprising way. While the Wnt8c is induced by FGF signals in the stem zone and persists in the newly formed neural tissue it acts, as FGF declines to promote Retinoid synthesis in the neighbouring paraxial mesoderm. This then leads to inhibition of FGF signalling and loss of Wnt8c. This action of Wnt8c at the transition from FGF to Retinoid signals depends on the loss of FGF signalling from the paraxial mesoderm prior to its loss in the neuroepithelium.

This FGF/Retinoid signalling switch is conserved in other tissues in the developing embryo, such as the extending limb bud. Elevated FGF and impaired Retinoid signalling are also characteristic of many cancers and this signalling switch may therefore represent a fundamental step that regulates the differentiation status of cells. Current work seeks to elucidate further the molecular mechanism of this "differentiation switch" by linking signalling pathways with the regulation of higher order chromatin organisation (Patel et al 2013).