The cis-regulatory logic of the ground state for neural specification

All cells in the body have the same genetic information in their DNA yet different cells become specialised for different functions. They achieve this by differential expression (reading) of the DNA information according to their context. To become a particular cell type during development, cells integrate signals from their neighbours and their own history and progress through a series of sequential steps that lead them to specialise. After embryonic development most cells stop dividing and they can no longer change fate. Understanding this processes by which cells make these decisions are critical because future therapeutic intervention in humans and animals will most likely involve harnessing the potential to re-specify cell identity so that we can help the body to regenerate or heal itself.

One problem is that the mechanisms that regulate the differential reading of the DNA are very complex: many of them take place at the same time and affect each other. Understanding how this works is therefore a daunting task. Until recently it was only possible to tackle this problem by studying one gene at a time, making it very difficult to understand the true complexity of gene interactions. Therefore it typically took many years even to build a basic understanding of any one of these decisions. Here we combine a number of recently available "next generation" techniques allowing many genes to be assessed at the same time, the active "control panels" (enhancers) of each one to be identified in specific cell populations, and to study the effect of changing one gene on hundreds of others at the same time. This information will be integrated to build a model for a critical developmental decision. We have chosen the decision by which cells go from being pluripotent to being specified as central nervous system (brain or spinal cord) or sensory precursors, partly because it seems that the initial processes of this decision may be very fundamental and perhaps even common to many other such cell fate changes. Using this strategy we expect to be able to uncover the basic gene regulatory interactions that govern these steps with only 2 post-doctoral researchers in 3 years, establishing an efficient methodology for tackling other problems in future.

In addition the project will generate a number of important resources that will be made publicly available, including data about the state of activity of all genes (and their enhancers) and the places and times during development at which they are activated. The project will also compare cells during normal development with stem cells, to determine the extent to which the decisions made by these cells are similar. In future this information will be invaluable to be able to direct stem cells (as well as normal body cells) to particular types in order to repair injury or disease, or to study the responses of particular types of cells to different potential treatments in culture.

This project aims to characterise a state of cells specified to become neural, pre-placodal and/or early epiblast, which also appears to accompany the transition from pluripotency to a "primed" state of ES cells that predisposes them to a neural or neural crest-like fate. We will use chick embryos as a model (which offers a number of important advantages) and a combination of new techologies in parallel, which should enable us very efficiently to construct a Gene Regulatory Network of gene interactions that define this state. Specifically we will:

1. use RNA-seq to compare the transcriptome of embryonic cells exposed to signals from the organiser, of pre-placodal ectoderm and of epiblast from pre-streak embryos;

2. use ChIP-seq to identify active enhancers in the same samples. This combination of approaches will allow us to prioritise genes, selecting those represented in both datasets and common to the 3 conditions;

3. starting with the transcription factors (TFs) represented in the above selected set, we will validate their expression at appropriate stages of development by in situ hybridisation;

4. make reporter constructs for enhancers for the selected, in-situ validated TFs and test them in vivo using a rapid cloning and screening strategy;

5. use bioinformatics to analyse the enhancers to identify key transcription factors; these will be related back to 1 and 2 above;

6. use NanoString analysis for testing interactions between key sets of transcription factors to generate a preliminary interactome (GRN).

7. analyse the transition from pluripotency to the "primed" state of chick ES cells as in 1 and 2 above, and the genes and enhancers compared to the draft GRN.

8. use the information from 5-7 above iteratively to refine the network.

We expect to be able to generate a fairly complete GRN within 3 years with only 2 postdocs.

This project is not only multidisciplinary but also cuts across 7 key BBSRC strategic priority areas:
ageing, animal health, the 3Rs, synthetic biology, technology development, data driven biology, systems approaches.

The benefits to academics in many disciplines are summarised above. Briefly:

The methodology should be applicable to many problems: a biological question is transformed into inter-related genome-wide screens and data then used to reduce the number of important genes for further study (no "candidate genes"). The GRN is a model with predictive power and will be made public on www.Biotapestry.org, for users to explore the effects of changing the state of different genes in the network, etc. It can therefore serve as an important teaching/training aid for students and professionals in many disciplines. The "omics" and perturbation (NanoString) data will be made publicly available and make up resources useful for reference for many problems and help to annotate the genome and Gene Ontologies. Using different tools on the same biological problem will make the resources amenable for cross-reference, increasing their value for Systems Biology. All of these benefits should occur during the project or shortly after its end. Beyond this, the project could generate information to understand cell fate transitions more generally, contributing to the generation of tools facilitating the manipulation of cell fate in vitro or in vivo by pointing at crucial genes and interactions.

The project should also have benefits outside academia although it is likely to take a little longer for these to bear fruit. Specifically we can envisage the most likely benefits to include:

* interdisciplinary training and provision of highly skilled individuals:
* training of PDRAs employed on the grant will not only equip them with scientific skills, but also with transferable skills applicable to other areas including organisational, cross-disciplinary interaction, problem solving, modelling complex scenarios. This will contribute to the UK economy by providing highly skilled personnel for the private sector
* improve international reputation of UK science and collaboration

The network generated can be used for teaching and training medical, veterinary and other practitioners and general public.
* it will be an interactive teaching tool - useful for dissemination, teaching about cell fate/stem cells, etc.
* particularly useful for modelling situations
* moving towards a comprehensive model of human and animal physiology - the network is adaptable to other situations involving these genes
* young scientists at 6th form level - these will be attracted to our labs through the Nuffield and similar schemes

Concerning the 3Rs, a predictive model can help to streamline the design of biological experiments and thus reduce the number of animals required to test hypotheses. Our publicly available data will also help because it will be a unique example of datasets matched for the same biological situation which will have much greater value.

Medical, veterinary and other animal and human health applications, especially regeneration and repair:
* the genes could potentially help to identify endogenous stem cells that could be harnessed for repair or other therapies
* the genes and their regulatory mechanisms could be manipulated to control cell fate in vivo or in vitro
* the information generated has potential to help generate patient-specific cells specified as pre-neural for therapy or to test treatments

In pathology or other laboratories (including in the commercial/industrial sector, pharmaceutical industry etc):
* grow cells and manipulate their fates, or "dial up" the pre-neural state - use these to test effects of drugs or other treatments on specific cell types
* use the cells to develop new drugs/treatments