The role of transcription factor Pax6 in glutamatergic versus GABAergic cell fate determination in developing cerebral cortex

Lead Research Organisation: University of Edinburgh
Department Name: Biomedical Sciences


The brain contains two major classes of neuron: (1) excitatory neurons, which transmit signals between cells by activating the neurons they connect to; (2) inhibitory neurons, which help keep the excitatory neurons in check by dampening down their activity. Both types are needed in the correct numbers for the brain to work normally. As well as doing different things, excitatory and inhibitory neurons look different and have different chemistries. These differences are created because, early in their existence, each type activates its own distinct set of genes that specify the way in which it develops. We have evidence implicating a protein called Pax6 in the high-level control of this specification process. Pax6 is a transcription factor, meaning that it binds to DNA at specific sites and controls many other genes. We have new data indicating that one of Pax6's most important functions during early embryonic development of the cerebral cortex is to prevent the activation of genes that would turn cells into inhibitory neurons and promote the activation of genes that would turn cells into excitatory neurons.

In mouse, 70-80% of neurons in mature cerebral cortex are excitatory. They are generated within the cortex from cells that contain high levels of Pax6. The other 20-30% are inhibitory. In normal development, they are generated outside the cortex by cells that do not express Pax6 and they then migrate into the cortex. That is, the mouse cortex does not normally make its own inhibitory neurons. We have discovered that if Pax6 is removed specifically from cortical cells in the embryo, when they are just starting to make cortical neurons, these cells undergo a highly abnormal, rapid and powerful activation of genes that would be expected to promote the development of inhibitory neurons. In other words, they respond to Pax6 removal by altering their program of gene activation from a pro-excitatory to a pro-inhibitory one.

Our finding raises two important questions. (1) What happens to cortical cells that undergo this early reprogramming? Do they go on to form mature inhibitory neurons, or do they later revert to an excitatory type, or do they make cells with a mixture of properties? (2) How does Pax6 normally prevent a pro-inhibitory and promote a pro-excitatory program of gene activation in the embryonic cortex? We aim to provide answers to these questions.

To address the first question, we shall remove Pax6 from specifically the cortex in the embryo, using genetic methods, and examine the consequences at progressively older ages through to adulthood. At each age, we shall look at what genes are activated, whether the mutated cells contain proteins associated with inhibitory or excitatory neurons, what the cells look like and what their electrical properties are. To address the second question, we shall test a hypothesis supported by previous evidence suggesting that removal of Pax6 might promoted inhibitory neuron development because it allows abnormal cortical activation of another transcription factor, called Gsx2, that is not normally present in cortex and which can itself activate a pro-inhibitory program. In other words, we suggest that Gsx2 might mediate many of Pax6's actions. To test this hypothesis, we shall simultaneously remove Pax6 and prevent Gsx2 activation and then examine the consequences for inhibitory vs excitatory neuron development in the cortex. We shall assess the effects on neuronal type using methods mentioned above.

In summary, we aim to provide new knowledge about the ways in which the cerebral cortex develops a correct balance of inhibitory vs excitatory neurons and also about the capacity of cortical neurons to alter how they develop when perturbed genetically.

Technical Summary

The development of a correct balance between the numbers of excitatory (glutamatergic) and inhibitory (GABAergic) neurons in the cerebral cortex is essential for its normal function. Previous work indicates that Pax6 suppresses a pro-GABAergic and promotes a pro-glutamatergic program of gene expression in developing cortex. We have discovered that, if Pax6 is deleted acutely and specifically from cortical progenitors at the onset of corticogenesis, these progenitors switch from a pro-glutamatergic to a pro-GABAergic program of gene expression.

We shall address the following questions. (1) What happens to cortical cells that undergo this early Pax6-loss-induced reprogramming? (2) What are the mechanisms by which Pax6 normally represses a pro-GABAergic and promotes a pro-glutamatergic program of gene expression in the embryonic cortex?

To address question 1, we shall follow, across a range of prenatal and postnatal ages up to adult, the effects of inducing Pax6 loss (using tamoxifen-induced conditional mutation) at two embryonic stages, one early in corticogenesis and one later, to test the long-term effects of early reprogramming and how they vary with the age of deletion. We shall study the effects of deletion on gene expression, neurochemistry, neuronal morphology and electrophysiological properties.

To address question 2, our priority will be to test the hypothesis, which is supported by previous evidence, that many of Pax6's actions on the pro-GABAergic and pro-glutamatergic programs might be mediated through its repression of the transcription factor Gsx2. We shall use a genetic approach to block changes in the expression of Gsx2 (and possibly other transcription factors) that result from acute Pax6 loss and might mediate its effects. Cytological, biochemical and next-generation sequencing methods will be used to assess the outcomes of attempts to reverse the Pax6-loss-induced reprogramming of cortical progenitors.

Planned Impact

The main beneficiaries fall into several categories:

(i) The project will benefit staff working directly on the project or staff and students indirectly associated with it (e.g. other postdoctoral fellows, undergraduate, masters and PhD students in our group), enhancing either directly or indirectly their research and allowing them to continue, or promoting their ability to start, a successful career in science. In addition, staff working directly on the project will benefit from the additional skills learned through collaboration on the electrophysiological properties of reprogrammed cells. These skills are new to the postdoctoral scientist employed here (Dr Manuel) and their acquisition will significantly broaden her employment potential.

(ii) Academic researchers working in the field of developmental neuroscience in the UK and abroad, including those in our immediate professional circle carrying out similar research on cortical development, will benefit from the details of the new knowledge generated, the new high-throughput data deposited in publically-available databases and, possibly, a new line of transgenic mice.

(iii) Our work is also likely to benefit academic researchers working on the programming, reprogramming and plasticity of progenitors and the cells derived from them, be they neuronal or non-neuronal, for generating cells of specific types for repair of damaged tissues un humans. We shall interact regularly with Edinburgh's large group of researchers working on stem cell biology and tissue repair at the Scottish Centre for Regenerative Medicine (SCRM).

(iv) Our work on the actions of high level regulators of development such as Pax6 and Gsx2 is likely to be of great interest to groups of mathematicians, engineers, computer scientists and biologists engaged in blue skies research aimed at using lessons learned from biology to inspire new way of generating computers. We have in the past engaged, and shall continue to engage, with these groups through workshop attendance (CapoCaccia Cognitive Neuromorphic Engineering Workshop during Year 3 of the grant period).

(v) Our work will have health-related benefits, since there are significant numbers of humans with mutations of the gene at the centre of the study, Pax6. In the past, we have contributed presentations to annual meetings that include patient-groups suffering the effects of PAX6 mutation ( and We have observed that these patients have welcomed, and gained comfort from, a deeper understanding of their condition. We shall present our new findings to these groups and to clinician scientists who are trying to identify ever more precisely the neurological abnormalities in these patients (see Pathways to Impact for further details). Results from our studies are likely to provide information guiding their search.

(vi) Public beneficiaries are likely to be numerous. We exploit an extensive range of options for communicating widely with the public, including: presentations at local schools and non-scientific special-interest groups; displays in the publically-accessible Anatomy Museum at the University (funded by the Federation of European Neuroscience Societies [FENS]); authorship of educational websites (e.g. Teaching & Learning Guide with, Wiley and iTunesU on Mapping and Manipulating the Brain [] and a Website publication funded by FENS on Phrenology (; consultancy work with the BBC. We expect to publicize our research advances via the University websites (Years 2-3), actively promote our advances through University press releases (Years 2-3), and put on a display of our major findings and their implications in the Anatomy Museum, enhancing the impact of the display through engagement with a University-funded Art-Neuroscience group known as FUSION (of which the Price is the convener).
Description Pax6 is expressed throughout cortical development. Without it, significant numbers (~30%) of cortical-born cells activate genes encoding pro-GABAergic transcription factors. Our evidence indicates that Pax6 acts cell autonomously to suppress the competence of cortical progenitors to activate these transcription factors. This suppression is in place when cortical progenitors start to make neurons, as shown by our unpublished RNA-seq data on the effects of triggering cortex-specific Pax6 deletion at the start of cortical development. The details of this activation vary spatially and temporally: for example, Dlx1/2 and Gsx2 appear initially only laterally and later spread to other regions whereas Prdm13 and Ptf1a are activated only medially.
I hypothesize that Pax6 acts on the genome of cortical progenitors to protect them against the effects of extracellular signals that might otherwise reprogram them to a GABAergic fate. This is attractive because it would explain why only ~30% of Pax6-null cortical-born cells reprogram to become GABAergic and why the programs they adopt vary. The hypothesis predicts that the levels and the types of signals that an unprotected progenitor receives, which are likely to vary spatially and temporally, will determine whether or not a progenitor reprograms and, if it does, which pro-GABAergic transcription factors it activates. For example, as interneurons enter lateral cortex and later spread more widely, the signals they produce (such as Shh) might trigger cells in their vicinity to activate Dlx1/2 and Gsx2, explaining the spatiotemporal pattern of ectopic expression of these transcription factors described above. The fact that Prdm13 and Ptf1a are activated medially could be explained by their exposure to high levels of signals enriched medially, such as Bmps and Wnts.
Exploitation Route The results could be taken forward by other researchers interested in the mechanisms of cortical development. The full impact of the research will be clearer once we have completed the work.
Sectors Healthcare