Dissecting mitotic phosphatases network and the effect of molecular glues to combat cell proliferation

Background

Many cancers are associated with an alteration in protein function and regulation. While protein phosphatases represent potential therapeutic targets because of their essential role in cell signalling, survival and cell division, they usually do not have any pockets for inhibitors. New drugs to protein phosphatases targeting protein interfaces are in phase 1 and 2  clinical trials for cancer treatment (Vainonen et al, 2021). To understand the effect of these drugs and how resistance may arise and be targeted, we need to know what are the interactors of phosphatase targets and what are the mutational hotspots. The project will focus on the mitotic serine/phosphatase PP2A-a tumour suppressor, evaluated as a target in various cancers. PP2A activity is essential for cells to complete cell division and control cell survival.

The current strategy is to stabilize the inactive or active conformation of PP2A to interfere with PP2A activity, which display promising an-tumour activities. In particular, the small molecule LB100, acts by stabilizing inhibited PP2A and is in phase 2 clinical trials for astrocytoma and glioblastoma, and phase 1 for advanced solid tumours while other drugs stabilize PP2A regulatory complexes. To ensure the efficacy of this treatment strategy, the interactors and downstream pathways affected by PP2A inactivation need to be determined to understand how pathways may adapt to overcome the drug treatment and limit cytotoxic side effects. The ultimate goal of the project is to understand the PP2A interaction network and identify regions that can be targeted to improve current molecular glues and predict glue resistance.

Aims

1. Design a methodology to chemically couple PP2A to mitotic interactors with high spatial and temporal control. The phosphatase will be fused to a HALO tag and photocrosslinkers will be added so that the temporal and spatial localization are optimized before crosslinking.
2. Identify key PP2A interactors that allow cell cycle progression and cell survival, using a light-induced chemical crosslinking approach, using light-activated small molecule crosslinking HALO-ligands. The conjugated hit proteins will be purified using anti-HALO-tag modified agarose beads and identify by MS. The protein complexes and interactors will then be characterized, and the activity and affinity of known PP2A molecular glues such as LB100 will be determined for the PP2A complexes biophysically.
3. Define key interaction surfaces of PP2A complexes, and evaluate their therapeutic potential for PP2A molecular glues targeting. The student will use machine learning algorithms predicting protein interactions, molecular modelling and bioinformatics to define PP2A interacting network driving the completion of the cell cycle. Using genomic tools, the student will mine the gnomAD database to identify benign mutations in PP2A. Combined with available data on human PP2A pathogenic mutations, they will identify mutation hotspots and conserved regions that could be more specific drug targets for molecular glues (Vainonen et al, 2021) to stabilize PP2A holocomplexes. The student will use computational modelling to predict PP2A mutations that could result in drug resistance and cancer relapse. Selected clinically relevant mutations will be tested to check whether the mutations interfere with molecular glue/drug binding.

Training Outcomes

1. Biophysics and cell biology: The student will gain experience in cell biology using Hela cells and glioblastoma cells cultured in the Welburn lab, fluorescence microscopy establishing a methodology to crosslink PP2A interactors with high spatial and temporal precision. This will provide a strong foundation for identifying interactors using photocrosslinkers optimized by Hulme.  The complexes-molecular glue interactions will be characterized through biophysical studies, complemented with Alphafold prediction of interactions.
2. Chemistry: The student will chemistry methodology designing photoreactive small molecule crosslinkers and mass spectrometry to analyze proteomic data
3. Structural biology, biochemistry and modelling to analyze protein complex structures and effect of mutations on structure and protein interactions
4. Computational biology and machine learning: The student will acquire knowledge in computational modelling to predict protein complexes structure, and analyze genomic data from human population to analyze the effect of mutations using modelling and in supervised machine learning techniques.

References

Vainonen et al, 2021. Druggable cancer phosphatases. Science translational medicine.

Leonard et al, 2020. Selective PP2A enhancement through biased heterotrimer stabilization. Cell.