Professor Neil Carragher, Dr Joanna Birch, Professor Margaret Frame
Glioblastoma (GBM) is the most common and lethal brain tumour in adults, and current treatment options are not very effective. There is an urgent need to find new targets for therapeutic intervention. Based on our recent work, and that of others, ILK (Integrin Linked Kinase), which is generally found at sites where cells adhere to their environment, is a protein that GBM cells absolutely depend on for their malignant behaviour. We will therefore take new approaches to inhibit key functions of this vital protein and determine how best to combine its deficiency with other drugs, so that brain cancer cells are irreversibly impaired.
We will use the most up to date form of gene editing (so-called CRISPR/Cas9), to remove the ILK gene from human GBM cells that were recently derived from patients in Edinburgh. We will look for drugs that mimic ILK-deficiency by using complex, data-rich, microscopy-based ‘morphometric fingerprints’ combined with the latest Artificial Intelligence/Machine Learning approaches– a new way of using very detailed cellular feature analyses and data interrogation methods to classify cell phenotypes and drug responses. Crucially, we will try to find out how best to combine loss of ILK with drugs that enhance the biological effects of its loss; this is necessary because cancer cells rarely respond robustly to loss of function of a single protein target.
Therefore, we will use state-of-the-art drug discovery methods that “paint” cells and study hundreds of parameters relating to their shape, texture and state of growth by using high-throughput microscopy. This will be done initially in cells grown in 2D culture, with a view to expand into 3D culture to increase biological relevance. A unique series of well-understood chemical (drug) libraries that we have built for such purposes in our centre will be used. Agents that mimic, or work together with, ILK loss to enhance advantageous biological effects will be selected and the strongest hits prioritised to understand how they are working. To ensure enhancement of ILK loss effects is maintained with standard of care radiotherapy treatment, these selected hits will also be tested alongside clinically relevant doses of radiation using infrastructure available at the University of Glasgow under the guidance of Dr Joanna Birch.
This work is important because we need to discover which drugs to combine ILK-deficiency with to provide new opportunities for GBM patients. Our chemistry colleague (Prof Asier Unciti-Broceta) is generating small chemical inhibitors of ILK’s protein complexes which we will also test in combination with drugs identified by the chemical-genetic phenotypic screening strategy described above. We have a track-record in using cancer cell phenotyping coupled to chemical synthesis to develop an excellent drug candidate, recently licensed to Pharma (Nuvectis Inc, New Jersey; see Edinburgh-Nuvectis-deal).
The successful PhD candidate will receive training in the use of cutting-edge drug discovery technologies and data analysis approaches and will be well placed for pursuing a future career in translational research in academia or industry.
PhD main objectives and timeline
Year 1: Test hypotheses (on human CRISPR-ILK-deleted GBM cells derived from patients), of the potential synergy between ILK-deficiency and drug candidates such as MEK inhibitors previously identified in transformed NS cell model. Responses amongst a cell series representing distinct clinical subtypes will be compared with mutational data. In parallel, fully characterise ILK-deleted human GBM cell populations, and begin to optimize the protocols for Cell painting and high-content phenotypic imaging in these cells.
Year 2: Perform the bulk of the phenotypic combination chemical-genetic screens across target-annotated, FDA-approved and diverse chemical libraries and also a set of novel in-house compounds designed against the ILK-Pinch-Parvin complex. Phenotypic responses will be clustered by phenotypic signature and linked bioinformatically, and via Machine Learning, to known mutational drivers.
Year 3: Prioritise several (around 5) of the most promising hit agents, selected also for pharmalogical and drug-like properties. Perform radiation combination experiments, detailed mechanism-of-action studies, addressing target identity, and beginning in vivo experiments, to test effectiveness against GBM grown in mice.