A new mathematical framework for Aβ peptide assembly in Alzheimer’s disease

Biography

I completed my integrated masters in Theoretical Physics (MPhys) at the University of Sheffield in Summer 2018. My background is rooted in fluid mechanics, quantum physics and biophysics. During my master’s year I undertook a project with Dr Buddhapriya Chakrabarti (Department of Physics & Astronomy) and Dr Rosemary Staniforth (Department of Molecular Biology and Biotechnology) wherein I was simulating proteins in the context of Alzheimer’s disease. I am continuing with this objective in my PhD as a member of the same research groups from October 2018 – September 2022 as a member of the White Rose DTP program having recently attended the Ulm Meeting 2019 on “The biophysics of amyloid formation” and the International Soft Matter Conference (ISMC 2019).

My Project 

Misfolded proteins that do not fold to their native state often lead to disease pathology [1]. Examples of such a phenomenon in humans include neuropathologies such as Alzheimer’s, Parkinson’s and several other prevalent diseases. Understanding the cause of such protein misfolding diseases, and thereby developing strategies for a cure, is thus crucial from a fundamental, technological and societal point of view. Fibrillar “amyloid” species accumulate as extracellular plaques or intracellular tangles and act as sinks for neurotoxic soluble species. Our data show that current models of protein polymerisation are inapplicable to amyloids such as AB. This is because historical models were developed for proteins with a narrow range of folds, while amyloids have complex folding landscapes and a vast array of conformations.

While the mechanism of protein misfolding and the cause of pathologies are still unclear, computer simulations based on a relatively small number of experimental observations have produced ground breaking insights, as recognised by a Nobel prize to Levitt, Karplus and Warshel in 2013.[2] Our proposal is to use techniques that will remove bias in our search for mechanisms, producing emergent behaviours from simulation rather than simply confirming or negating existing theories. This represents a step-change in our approach to amyloid formation. To date, theoreticians have not taken full advantage of structural and kinetic data. The novelty of this proposal is its “multi-scale” nature and the combination of top-down and bottom up approaches.

All-atom simulations are typically viewed as the standard in simulation work. Whilst atomistic simulations are valuable, they are incredibly computationally expensive. Coarse graining is the process of representing multiple atoms as one interaction site thereby reducing the number of degrees of freedom a system has. This can be carried out on a variety of length scales dependent upon the system in question, varying from united atom models to generating elastic networks for quaternary protein structures. Our primary aims in this project are to create coarse grained models of amyloid proteins with the goal of understanding protein misfolding processes in general and to make quantitative predictions of the Aβ self assembly into fibrillar aggregates and also the formation of oligomeric structures.

[1] Knowles, T. P. J., et al. (2014). Nat. Rev., Molecular Cell Biology 15, 384-396.

[2] Levitt, M. (2014). Angew. Chem. Int. Ed. 53, 10006-10018