The mammalian cortex is a highly organised structure responsible for receiving, integrating and sending complex signals required for higher order cognition, language, complex behaviour and precise motor control. Development of the cortex is a tightly regulated process and divergences from the normal trajectory of cortical development can have devastating consequences on individuals. Among the myriad molecular regulators of cortical development, the axon guidance receptor RYK acting via the WNT family of ligands has been shown to be necessary for normalcortical development.
RYK was initially shown to regulate axon guidance within the Drosophila ventral nerve cord (Bonkowsky and Thomas 1999; Yoshikawa et al. 2003). This lab has previously identified a callosal phenotype resulting from loss of Ryk in the mouse (Keeble et al. 2006). In Ryk-/- embryos, callosal axons successfully cross the midline but are then unable to project into the contralateral cortex and thus form contralateral axon bundles. We have also shown that WNT5a is the chemorepulsive ligand responsible for the guidance of Ryk-expressing callosal axons away from the midline.
The goal of this PhD project was to gain further understanding into how Ryk and Wnt5a regulate callosal projection neuron production and axon guidance. The transcription factor Satb2, is expressed in the majority of callosal projection neurons (Alcamo et al. 2008) and we have performed SATB2 immunohistochemistry to specifically identify the SATB2-positive (SATB2+ve) population of callosal projection neurons compared to the SATB2-negative (SATB2-ve) population of cortical neurons (which includes subcerebral projection neurons, corticothalamic projection neurons and cortical interneurons). We performed an analysis of neurite outgrowth features in SATB2+ve callosal projection neurons and SATB2-ve non-callosal neurons from E18 Ryk+/+ and Ryk-/- embryos. Loss of Ryk led to increased axon outgrowth in callosal projection neurons and non-callosal neurons suggesting that loss of Ryk did not impair the ability of cortical neurons to project axonal processes in vitro. WNT5a and RYK together exerted opposing effects in callosal and non-callosal neurons, specifically inhibiting axon outgrowth in callosal projection neurons while enhancing axon outgrowth in non-callosal neurons.
We also discovered that Ryk controlled the production of callosal projection neurons during cortical development. There were significantly more SATB2+ve callosal projection neurons in the E18 Ryk-/- cortex compared to wildtype embryos, showing that cortical neuron production was disrupted in the absence of Ryk. These data suggested that inappropriate specification of SATB2+ve neurons was responsible, at least in part, for the callosal phenotype we observed in Ryk-/- embryos. Satb2, expressed in callosal projection neurons represses Ctip2, which subsequently specifies a subcerebral projection neuron fate (Alcamo et al. 2008). We have also examined the number of CTIP2-positive (CTIP2+ve) subcerebral projection neurons in the E18 Ryk+/+ and Ryk-/- cortex. Although Satb2 inhibits Ctip2, we observed no decrease in the number of CTIP2+ve neurons in E18 Ryk-/- embryos compared to wildtype.
One possible mechanism for the increase in callosal projection neurons in Ryk-/- embryos was the deregulation of WNT signalling. We observed increased β-catenin in the E18 Ryk-/- cortex compared to wildtype, suggesting that Ryk was required to maintain lower levels of cytoplasmic β-catenin during the time when callosal projection neuron specification was occurring. Using an in vitro luciferase assay to measure β-catenin transcriptional activity, we showed that WNT3a, also a RYK ligand, but not WNT5a, activated canonical WNT/β-catenin signalling in cortical neurons from E18 Ryk+/+ and Ryk-/- embryos. These data suggested that WNT/β-catenin signalling was an important mechanism regulating cortical neuron development. We also showed that WNT5a modulated WNT3a-dependent WNT/β-catenin signalling in cortical neurons, suggesting there was a complex interplay between canonical and non-canonical WNT signalling during callosal projection neuron specification. Finally, we showed that WNT5a, via RYK, was able to inhibit nuclear translocation of β-catenin, specifically in SATB2+ve callosal projection neurons, implicating WNT5a-RYK-dependent inhibition of WNT/β-catenin signalling as a potential mechanism for regulation of callosal projection neuron production.
In conclusion, we have identified Ryk as a novel regulator of callosal neuron production. We have also shown that Ryk modulates canonical WNT/β-catenin signalling, implicating this pathway as a potential mechanism for Ryk regulation of callosal neuron specification. Given the importance of appropriate cortical neuron specification during nervous system development and the wide-ranging actions of canonical WNT/β-catenin signalling during cortical development we conclude that identification of Ryk as a key regulator of these processes is an important discovery that warrants further investigation.