Functional profiling and computational models to better understand DDR processes and guide improvements in radiotherapy

Background: The propensity of deficient DNA response and repair components in cancer cells are exploited in radiotherapy treatments.

Irradiation leads to damage of cellular molecules, the most prominent being breaks in nuclear DNA which can culminate in cell death or senescence if levels are above threshold of the cell’s capacity to repair these.

The DNA Damage Response (DDR) is a dynamic reconfiguring of chromatin organisation and behaviour, driven by a complex cascade of proteins and post-translational modifications (PTMs), such as phosphorylation, that are activated upon DNA damage detection.

Systematic approaches to experimental map and computational model signalling networks offers a powerful capability to gain a deeper mechanistic understanding of the regulatory processes that orchestrate DDR. This information can be exploited to guide improvements in radiotherapy treatments.

Aims: We will systematically characterise the molecular changes that occur in DDR deficient cells and signalling events that are activated upon IR perturbation.

High content proteomics datasets will be generated and integrated on a mechanistic network model of DDR, which will be deployed to conduct virtual screens and predict clinical responses.

Through this work we will establish a computational platform that can integrate diverse experimental data and produce a valuable resource for the RadNet community to use.

Methods: We will draw on our combined expertise in functional genomics, proteomics and computational biology, to produce a network model of DDR.

Leading edge quantitative mass spectrometry will be applied to systematically characterise proteome and phosphoproteome changes across DDR deficient cell models, at baseline and upon IR perturbation.

We will apply statistical approaches to derive robust molecular signatures for DDR deficiency states and to associate the changes to pathways, genes or perturbation states. These data will be embedded in an executable computational model of DDR.

We will implement iterative cycles between experimental and computational components to train and validate a mechanistic model of DDR.

How research will be used: By extending our understanding of DDR processes and how deficient genetic states respond to IR, we will be able enhance the therapeutic efficacy radiotherapy.

Firstly, our data driven executable network model of DDR will be used to conduct virtual drug and CRISPR screens to identify new sensitising treatments.

Secondly, we will further use these predictive models to identify radio-resistance mechanisms and to select biomarkers to monitor IR therapy.

Thirdly, we incorporate information on patient backgrounds to accurately model clinical data and predict patient treatment outcomes.