The role of enhancers and transcription factors in the reorganisation of chromatin architecture

In this proposal we are asking fundamental questions about the mechanisms employed to activate gene expression within the normal context of chromosomes. The mammalian genome encompasses about 1 metre of DNA which is assembled as chromatin and is compacted ~10,000 fold to fit it all into the cell nucleus as condensed chromatin. The genome is packaged as repeating units of structures termed nucleosomes which are made up of bundles of histone proteins surrounded by 147 base pairs (bp) of DNA which makes 1.7 turns around the outside of the nucleosome. Nucleosomes are themselves separated from each other by a short linker of about 40 bp, and are organised as a highly regular array with a repeat length of ~180 - 200 bp. In order to activate gene expression it is necessary for genes to undergo many distinct levels of chromatin modification and decondensation. The whole process of gene activation is tightly controlled by control modules on the DNA, namely proximal promoter elements where mRNA synthesis starts and distal enhancer elements. These elements function by first interacting with specific proteins named transcription factors that bind to DNA, which then recruit complexes that modify and remodel chromatin, and at the same time engage with the transcription apparatus that reads the genetic information required for the synthesis of gene products. Gene activation can involve the displacement of nucleosomes from regulatory elements and significant modification of the whole nucleosome array encompassing genes and their enhancers. Our laboratory is actively engaged in defining the transcription factors and chromatin modifying complexes that interact with and activate transcription of the human GM-CSF gene. This is a gene that functions during haemopoiesis and within the immune system to control the growth and function of specific classes of white blood cells termed granulocytes and macrophages. We have shown that this gene is very tightly regulated and is induced ~10,000 fold upon stimulation of pro-inflammatory pathways in cells such as T cells and mast cells, which also represent part of the immune system. In this context, GM-CSF functions to receive and send signals between different parts of the immune system, and it is essential that its expression is carefully controlled. In our studies we have identified an inducible transcriptional enhancer that is located 3000 bp upstream of the GM-CSF gene. We have shown that this enhancer responds to activation of the T cell antigen receptor, and undergoes extensive remodelling at the level of chromatin structure. We have shown that the recruitment of inducible transcription factors such as NFAT and AP-1, or developmentally regulated factors such as GATA-2, leads to the eviction of nucleosomes from the enhancer and profound remodelling of the flanking nucleosomes. We find that at least 4000 bp of DNA (~ 20 nucleosomes) is restructured such that (i) the regular array is disrupted and randomised, and (ii) the average nucleosome repeat length is reduced from the normal value of 180-190 bp to a much shorter value of 150-160 bp. It is not known how such relatively small enhancer and promoter elements can alter the structure of the chromatin fibre over such long distances. We have indications that this may be done by a combination of (i) transcription initiated not from the normal promoter element but from the enhancer, and (ii) chromatin remodelling complexes that are recruited by the enhancer and then spread across the locus. In this proposal we wish to test whether these mechanisms do indeed operate within the GM-CSF gene, and we want to work out the precise details of the remodelling process. We also aim to determine whether additional genetic elements exist that block the spread of chromatin remodelling from active genes into neighbouring genes.

This proposal addresses mechanisms controlling gene activation by enhancers at the level of chromatin structure, and it focuses on inducible and developmentally regulated nucleosome reorganisation mediated by NFAT and AP-1, and/or GATA-2 across the human GM-CSF enhancer. (1) We find that transcription factors mediate nucleosome displacement in the enhancer, and that enhancer activation by distinct pathways results in different patterns of nucleosome positioning. (2) We find that enhancer activation leads to nucleosome mobilisation across 4 kb of DNA. Chromatin which previously existed as a regular array of positioned nucleosomes becomes randomised, and the average nucleosome repeat length is reduced from 180-190 bp to just 150-160 bp. This represents a profound restructuring of the locus and hints at a largely undefined but very important level of chromatin decondensation that accompanies the activation of genes by enhancers. (3) We find that the enhancer can recruit RNA polymerase II in an inducible fashion and initiate the synthesis of a non-coding RNA. We now aim to (1) Use GM-CSF transgenic mice as a model to identify the nature of the chromatin modifications, and the chromatin modifying complexes responsible for chromatin decondensation and nucleosome mobilisation. We will also define the kinetics of nucleosome mobilisation, and the rate and processivity of the spread of active chromatin across the GM-CSF locus upon enhancer activation. (2) Use episomal constructs of the GM-CSF locus in cells such as T cells, mast cells and erythroid cells, as tools to define the regulatory elements, transcription factors and remodelling complexes that direct nucleosome mobilisation and enhancer function. (3) Determine whether transcription initiating within the enhancer plays a role in nucleosome mobilisation, and determine whether either transcription terminators or other structural elements in the GM-CSF locus can block the spread of active chromatin.