SNPS AND TF BINDING · Globally connecting single nucleotide variations to transcriptional regulation in vivo

Globally connecting single nucleotide variations to transcriptional regulation in vivo

Final report for project 'Globally connecting single nucleotide variations to transcriptional regulation in vivo' (SNPs and TF binding) Period covered: 1 October 2009 - 30 September 2012 Summary of proposed objectives: The overarching goal of this project was to understand how subtle changes in regulatory genetic sequence could generate phenotypic differences between closely related species via influencing binding of regulatory proteins - transcription factors (TFs) - to the deoxyribonucleic acid (DNA). Changes in TF-DNA interaction have, in turn, impact on the recruitment of complexes that transcribe the relevant genetic information. My aim was to interrogate the drivers responsible for moderating the impact of the sequence changes on TF binding. Understanding how sequence divergence in non-coding parts of the genome relate to the phenotype differences in closely related species can help us understand the origin and driving forces behind phenotypic diversity. Aim one: Identify the in vivo binding sites of regulatory proteins - TFs - within the genomes of closely related inbred mouse strains. Aim two: Connect the changes in TF binding location to the changes in the underlying and surrounding sequence. Propose the possible causative link between the sequence and TF binding changes for each of the locations genome-wide. Aim three: Test our hypothesis by eliminating one of the drivers of TF binding change. Summary of achieved objectives: Aim one: I have experimentally determined more than 50 000 binding sites genome-wide for three TFs (CEBPa, HNF4a and FoxA1) in mouse species separated by less than six million years of evolution, all in duplicate, to acquire the highest quality data, allowing me to achieve an unprecedented resolution for interrogating TF binding among closely related mammals. By high-throughput sequencing of the DNA fragments each TF interacts with, I was able to extract two crucial parameters: both the precise and reproducible in vivo binding locations for each TF, as well as the intensity of the binding, representing the average probability that the TF binds to that genomic location within the population of liver cells for each species. Aim two: First, in order to assess the initial steps in the evolution of transcriptional regulation, we compared how well both the location and binding intensity for each in vivo TF binding site is preserved over six million years of evolution, representing less than 3% of sequence divergence. While a conserved location indicates that the same physical position between divergent mice species is continuously occupied, similar intensity represents the probability that the TF binds to this location in a similar number of liver cells. We observed an extensive change in both the physical location and the binding intensity among the five closely related mouse species for all three TFs. While changes in the short sequence that is directly bound by each TF can explain a significant fraction, it cannot explain more than a third of all the changes in TF binding we observe between the mouse species. In addition, our experimental design allowed us to interrogate binding of all three TFs simultaneously. We found that TFs that bind to the same location exhibit coordinated changes in binding intensity. This implies that the driver behind changes in TF binding acts locally and often affects the whole cluster of TFs. Aim three: Since the intensity of TFs bound within the same cluster undergoes coordinate changes during evolution, this suggests that the binding of one TF has the potential to influence binding of the other TFs within the same cluster. I have tested the hypothesis by using a mouse that has one of the TFs (CEBPa) excised from the genome (a knock-out mouse). While CEBPa is an important TF for liver development and function, we do not see a profound effect on the stability of clusters of TF binding containing all three TFs. These clusters are bound in most of the cells and are evolutionary stable, containing possibly many more TFs that may be better candidates for the pioneering function of chromatin opening and recruitment of TFs other than CEBPa. Possibly, other TFs, similar to CEBPa (CEBPß and CEBP?) can compensate for CEBPa loss. I am currently testing if excising HNF4a will have similar or drastically different effect.

Data: CORDIS, © European Union