Integrating Capture-HiC with omic time course data to uncover the regulatory interactions modulated by genetic variation in disease

Rheumatoid arthritis (RA) is one of the most common chronic inflammatory diseases. A number of genetic differences related to RA have been discovered through Genome Wide Association Studies (GWAS) where large genetic datasets are analysed to statistically associate genetic changes in the human population (usually differences at single nucleotides, called SNPs) with disease risk. However, to make progress it is now important to improve our understanding of how these disease-associated SNPs affect the functions of cells and tissues, so that we can better understand the disease mechanism and ultimately develop more effective medicines. A major obstacle to understanding the effect of these disease-associated SNPs is that many of them lie far from the main control regions of protein-coding genes (known as promoters) in terms of the linear DNA sequence of the genome. There is strong evidence that some of these SNPs lie in enhancer regions which are distal control regions of DNA that come into contact with promoters through looping of the DNA sequence, i.e. these enhancer regions appear distant in terms of DNA sequence but may be close in physical space. One of the applicants, Peter Fraser, is a leading expert in investigating how DNA folds and has developed novel experimental methods to allow this to be tested. This allows us to investigate how the DNA regions associated with RA interact and control genes in human cells. In this project we will combine this technique with measurements of gene activity and enhancer activity in human cells over time. We will look specifically at stimulated T-cells, cells which are involved in the immune system and are known to be important determinants of RA disease progression. We will use these data to build mathematical models describing how T-cells regulate gene expression through enhancer activity and enhancer-promoter interaction. These models will allow us to better understand how regulatory proteins, called transcription factors, bind to the DNA at enhancers and promoters to turn genes on or off. Finally, we will carry out experiments on T-cells derived from healthy human volunteers where genetic data (SNP calls) are already available to test whether the natural genetic variation we observe at SNPs identified through our analysis do have a strong effect on enhancer activity or enhancer-promoter interactions. This would then provide strong evidence for how these SNPs regulate specific genes and we will investigate the downstream cellular pathways to which these genes belong.

The project brings a leading RA genetics group together with a leading molecular biology group and a leading mathematical modeller to work closely together and utilise the most up to date knowledge to gain insight into the function of genes that cause RA.

It has become clear that T-cells are of fundamental importance in Rheumatoid arthritis (RA). In this project we will use time course data from stimulated primary human T-cells to uncover the regulatory function of RA-associated enhancers controlling target gene expression. Capture-HiC data enriched for promoters and disease-associated enhancers will be used to identify the physical interactions between enhancers and promoters and how these change through time. Nuclear RNA-Seq will be used to quantify nascent transcription. ATAC-Seq will be used to profile open chromatin and identify likely transcription factor binding (TF) events. We will collect data from these three assays from the same primary T-cell population over a 24 hour time period after stimulation. We will then use these data to develop a regulatory network model, using non-parametric Gaussian process methods to infer TF activities and mathematical models to describe the effect of TF-binding on promoter activity and nascent RNA production rates. We will score alternative regulatory network models through Bayesian model comparison in order to identify the most well-supported models of regulation for each target gene. Computational methods will be scaled up from previous smaller scale applications to deal with more complex enhancer-driven regulation involving large numbers of regulatory TFs. Methods will be published as open source computational tools for the community. Finally, we will validate our predictions through follow-up experiments on human data by comparing subjects with differences in SNPs implicated through our model-based analysis. We will use targeted ChIP-PCR to look at the effect of these SNPs on epigenetic marks and TF binding and targeted 3C to confirm enhancer-promoter interactions and look at whether these are altered according to genetic differences.

Our vision is to leverage the power of a systems biology approach to better understand the pathogenesis of RA and to translate this to clinical benefit. Having established expertise in genetics, functional genomics, bioinformatics, statistical modelling and chromatin biology, for the first time we will incorporate chromatin interaction, chromatin accessibility and gene expression data generated from time-course experiments in stimulated CD4+ T-cells, a cell type known to be crucial in disease development, to determine how disease genes interact to cause RA. This will provide unprecedented insight into how biological pathways are regulated as a result of combinations of disease associated genetic regions and allow subgrouping of patients by the primary pathway disrupted. In addition, the study will provide a molecular profile, based on the time-course transcriptomic data, to define how a CD4+ T-cell responds to stimulus. The impact on this expression profile will then be analysed on different RA risk genetic backgrounds. In turn, this will inform the stratified medicine agenda by allowing better targeting of available therapies according to the primary pathway mediating disease We will focus on RA, initially in CD4+ T-cells, but this approach has the potential to inform the study of all diseases with a genetic component to susceptibility and our strategy and tools can be adapted for a diverse set of cell types and stimulation regimes.

Potential non-academic beneficiaries of the research include:

Pharmaceutical companies seeking drug targets to prioritise for RA and other complex or immune disorders and patients who may ultimately benefit from the resulting medicines (time-frame: 10-15 years from lead to product)

Pharmaceutical companies or health service providers interested in targeting existing treatments as personalised therapies who will benefit from better stratification of RA and other complex diseases, and patients who may ultimately benefit (time-frame: 1-3 years)

Health service providers carrying out genetic risk profiling who would benefit from a better risk assessment and prevention strategies from genetic variation data, and patients and their families who may ultimately benefit (time-frame: 1-3 years)

Research staff employed on the project will also be beneficiaries as they will be provided with comprehensive research training. As detailed in 'Pathways to Impact', the researchers will benefit from professional and career development opportunities at the University of Manchester. Past experience from the applicants' laboratories has indicated that the skills acquired by the researchers have been, and will continue to be, transferable to careers in research, industry, the NHS, medical writing, education and funding agencies (beneficiaries: researchers employed on the programme; time-frame: ongoing; prospective employers; time-frame: 3-5 years).