Search

The mammalian cortex is composed of two main types of neuron. Principal cells (PCs) are glutamatergic and use glutamate as the main neurotransmitter. Interneurons are mostly GABAergic and use GABA as their main neurotransmitter. PCs are the output units of the cortex and comprise ~80-85% of the mouse cortex. Interneurons modulate the activity of PCs and comprise ~20-15% of the mouse cortex. These two broad cell types originate from separate regions of the developing nervous system. Early embryo development is dependent on the balance between cell proliferation and differentiation. The process of cortical progenitor pool expansion, division and production of postmitotic cells starts at around embryonic day (E) 10.5. PCs arise from the dorsal telencephalon while interneurons arise from the ventral telencephalon. The process of telencephalon regionalization is regulated by differential expression of transcription factors (TFs). Pax6 is one of such TFs. Importantly, it is expressed strongly in the PC dorsal Emx1 lineage but much less so or not at all in interneuron ventral lineages. Here I investigated the postnatal characteristics of a Cre-loxP recombinant mouse that had Pax6 removed from the Emx1-lineage at E11.5. I targeted L5 and L2/3 of the somatosensory cortex and recorded their intrinsic electrophysiological properties. The results suggested that PCs can be specified without the need for Pax6 as the intrinsic properties of neurons in layers 5 and layers 2 and 3 were not significantly different than those of control cells. I explored an ectopic population of cells from the Emx1-lineage that expressed GABAergic markers in the embryo using electrophysiology and immunostaining. The results indicated that ectopic cells do not develop much electrical activity and stay in a state of arrested development. I investigated the consequences of deleting Pax6 in Emx1-lineage progenitor cell cultures. The results revealed the possibility that Emx1-lineage progenitors can produce interneurons in the absence of Pax6 if Sonic Hedgehog is activated. Taken together, the results support the hypothesis that Pax6 protects cortical PC specification by inhibiting interneuron-inducing signals.

The development of the diencephalon can be summarised as a process in which cells that initially appear similar give rise to a complex structure that contains a variety of cell groups called nuclei. It involves two stages: the early patterning of the diencephalic prosomeres and the later formation of the individual nuclei. It has been shown that transcription factors and morphogens regulate the first stage but their further effects on the second stage remain unclear. The ventral lateral geniculate nucleus (vLGN) is involved in the visual system and is shown to have complex origins from the thalamus, the zona limitans intrathalamica (ZLI) and the prethalamus. The transcription factor Pax6 is involved in the development of brain structures including the cortex, the diencephalon and the major axonal tracts in the forebrain by playing a multifaceted role in patterning, proliferation, differentiation, migration and axon guidance. It is known that Pax6 is essential in setting up the prosomeric boundaries in the developing diencephalon but its role in the formation of individual nuclei has not yet been explored. By using a conditional Pax6 knock-out mouse driven by Zic4Cre with a green fluorescent protein (GFP) reporter showing the Cre activity, the formation of the thalamic nuclei vLGN, dorsal lateral geniculate nucleus (dLGN) and VP (ventral posterior nuclei) was examined in postnatal day 0 (P0) Pax6+/+, Pax6fl/+ and Pax6fl/fl pups. Using this mouse model, I found an increase in nuclear volume at the rostral level and a global decrease in cell density in the P0 Pax6fl/fl vLGN, whereas in the dLGN an increase of GFP+ve cell proportion was observed. In Pax6fl/+, I found an increase in GFP+ve cell proportion in the caudal part of the vLGN and across the dLGN. No significant change was observed in the VP in either the Pax6fl/+ or the Pax6fl/fl. The defects in the vLGN and dLGN could be caused by: 1. disruption of the expression of patterning factors such as Shh and Nkx2.2; 2. cell proliferation defcts and abnormal apoptosis; 3. ocular developmental defects; 4. failure in cell sorting/migration; 5. cell fate change. During my PhD, I tested the first three theories and explored the fourth but was not able to pursue the last due to the time limit of the project. To test the hypothesized mechanisms underlying those defects seen in the vLGN and dLGN, I performed BrdU labelling to study the time origin of cells that contribute to these two nuclei and discovered that E11.5 and E12.5 are the main ages when these cells and the GFP+ve subpopulation are born. Then I carried out experiments to examine the cell proliferation and cell apoptosis in the thalamus (pTH-R, rostral part of the progenitor zone of the thalamus, and pTH-C, caudal part of the progenitor zone of the thalamus) and the prethalamus (Pth) from E11.5 to E13.5 and found: 1. the proliferation rate decreased in the pTH-R in Pax6fl/+ at E11.5; 2. the growth fraction decreased in both pTH-C and pTH-R in E12.5 Pax6fl/fl; 3. there is no change in cell proliferation in the GFP+ve subpopulation; 4. no abnormal apoptosis is observed in either the whole cell population or the GFP+ve subpopulation. Judging by the amplitude of the change in proliferation in the pTH-R and pTH-C at E11.5 and E12.5, it is unlikely that these changes alone are responsible for the phenotypes seen in P0 vLGN and dLGN. Then I examined the expression patterns of Shh and Nkx2.2 and the expansion of both was observed in Pax6fl/fl at both E12.5 and E13.5, which could explain the volume change of the vLGN but not the change in the proportion of GFP+ve subpopulation in both the vLGN and dLGN. Then I continued to examine if the ocular input from the retinal ganglionic cells are severely affected by the deletion of Pax6 and found no gross change in the conditional mutants, which rejected the ocular developmental defects theory. At the end of my PhD, I performed a BrdU short-term survival experiment and a brain slice culture combined with live imaging experiment to explore the possibility of abnormal cell migration causing the vLGN and dLGN phenotypes and found that the cells moving along the border of the thalamus and prethalamus move faster in the Pax6fl/fl than in the Pax6fl/+, but rather than moving directly toward the lateral surface of the diencephalon, they take a detour. These findings indicate that the deletion of Pax6 causes minor changes in the proliferation of E11.5 to E13.5 diencephalon and expansion of regional marker expression such as Shh and Nkx2.2, which could potenially affect the volume and change the proportion of GFP+ve cells in P0 vLGN and dLGN. Migration defects caused by Pax6 could also contribute to the phenotype observed in those two nuclei. Potential cell fate change caused by Pax6 deletion could be another factor that contributes to the defects in the conditional mutants. More work needs to be done to test the migration defect and cell fate change hypotheses in future.

The regulation of gene expression resulting in the formation of the mammalian cerebral cortex is tightly regulated by a group of transcription factors. The deletion of any one of these transcription factors results in numerous defects whose nature and severity depends on the role of the transcription factor in the regulation of complex gene regulatory networks involved in development. There is currently relatively little knowledge about the gene networks that these transcription factors control and how they exert their regulatory effects. The paired-box transcription factor Pax6 has been identified as a master regulator of gene networks involved in cortical development and its deletion results in numerous cortical defects such as an abnormally thin cortical plate and a vastly expanded proliferative zone. Previous work in our lab identified a list of candidate genes that are likely to be regulated by Pax6 in the developing cortex. Members of the Notch signalling pathway were potential Pax6 targets of particular interest since Notch signalling plays a crucial role in the maintenance of neural progenitor cells during development and consequently plays a critical role during corticogenesis. Our work aims to identify the regulatory relationship between Pax6 and Notch ligands Dll1 and Jag1 during cortical development. Analysis by flow cytometry and double labelling analysis of both gene and protein expression has provided insight into the relationship between Pax6 and Dll1 in progenitor cell subpopulations during cortical development. In situ hybridisation and qPCR results confirmed that loss of Pax6 causes loss of Dll1 expressing cells and downregulation of Jag1, indicating that both ligands are regulated by Pax6. Bioinformatic screening and analysis by luciferase assay suggests that Jag1 is a likely candidate to be a direct target of Pax6.

During development the nuclei of the thalamus form reciprocal connections with specific regions within the cortex. These connections give rise to the thalamocortical tract. The processes by which axons of the thalamocortical tract are guided to their target regions are poorly understood. It has been shown that diffusible or membrane bound factors can have a chemoattractive or chemorepulsive effect on the tip or growth cone of the axon. Thalamocortical axons may also be guided along ‘pioneer’ axon populations that form a scaffold along which axons may grow. The transcription factor Pax6 has been shown to have a role in a variety of developmental processes such as neuronal patterning, proliferation, migration and axon guidance. It is known that Pax6 is involved in the development of the thalamocortical tract but its exact role is unknown. To explore the role that Pax6 plays in the development of the thalamocortical tract I have used two different mouse models, the small eye (Pax6Sey/Sey) mouse which lacks functional Pax6, and a conditional Pax6 knock-out (Pax6cKO) mouse made using a Gsh2 Cre line that specifically reduces Pax6 expression in the ventral telencephalon and prethalamus. Using the Pax6Sey/Sey mouse I show that thalamocortical axons do not enter the ventral telencephalon in the absence of Pax6 and that a small number of axons incorrectly enter the hypothalamus. In addition axons found within the ventral telencephalon of the mutant do not originate from the thalamus but instead originate from cells within the ventral telencephalon itself. I have found that the expression of guidance molecule Robo2 is reduced in the Pax6Sey/Sey mouse, which may explain why thalamocortical axons enter the hypothalamus. When Pax6 expression is reduced at the prethalamus and ventral telencephalon using the Pax6cKO mouse I show that the majority of thalamocortical axons reach the cortex normally but some axons become disorganized within the thalamus. Pioneer axons which emanate from the prethalamus normally guide thalamocortical axons through the diencephalon but in the Pax6cKO I report that these axons are reduced which may explain the disorganization of thalamocortical axons within the thalamus. Taken together the data from these two models demonstrate that for the thalamocortical tract to form normally Pax6 expression is required in both the cells of the thalamus and in cells that lie along the route of the tract. In addition I provide evidence that Pax6 may influence axon guidance by controlling the expression of guidance molecules and the development of pioneer axon tracts.