Computational Prediction of Hotspots in Protein Misfolding for Rational Immunotherapy

Abstract

Physics-based algorithms can predict the misfolding mechanisms of proteins involved in aggregation-related diseases, including SOD1 whose misfolding template-directed conversion is involved in Amyotrophic lateral sclerosis and PrPc, wherein propagation of the misfolded protein is central to the the prion diseases. We have recently developed an algorithm capable of predicting thermodynamically likely regions for misfolding, by employing modeling which involves both atomistic interactions and surface-area based coarse-graining, along with a heterogeneous dielectric function inside the protein. Predictions based upon the algorithm are consistent with recent immunological assays that have uncovered disease-specific epitopes in SOD1 and prion protein, and point to diagnostic and therapeutic applications. This research was performed in collaboration with Dr. Neil Cashman at the Brain Research Centre, University of British Columbia, and involved joint supervision of M.D./Ph.D. student Will Guest.