Research Projects

Despite decades of research, Alzheimer’s disease (AD) remains without a definitive cure, with current treatments only slowing disease progression. This failure can be traced to our limited understanding of the amyloid aggregation cascade, which describes the self-assembly of proteins into oligomers of different sizes, and culminating in fibrils. Recent studies highlight amyloid-beta (Aβ) oligomers as the primary toxic species responsible for pathogenesis, shifting the research focus from mature fibrils to these smaller, transient structures. However, some critical questions remain largely unanswered: What are the secondary structures of oligomers, and how are they impacted by the binding of divalent metal ions? Does fibrillar order emerge early or late in the aggregation cascade? A generally accepted hypothesis is that oligomers exhibit their toxic effect by disrupting neuronal membranes. While multiple mechanisms of toxicity have been proposed, their molecular details remain elusive. To address these intriguing questions, we propose to create a high-fidelty computational framework by integrating rare-event sampling with multi-resolution models. The first strand of our research program will focus on predicting the structures of oligomers at the all-atom resolution. We will use the basin-hopping global optimization technique to predict the structures of oligomers of different sizes. Since basin-hopping is less sensitive to the presence of kinetic traps, it is likely to be more efficient in locating global minima (lowest-energy structures) as compared to previous schemes, which deploy different flavors of molecular dynamics. We will also ascertain how divalent metal ions affect oligomer structures, and specifically whether they abrogate or reinforce interactions critical for self-assembly. The next strand of our research program will elicit the microscopic details underlying the assembly of oligomers, and their maturation into protofibrils. As these transitions span a hierarchy of length and time-scales, atomistic simulations would be intractable. To address these bottlenecks, we will develop an intermediate resolution coarse-grained model that would provide a fine balance between accuracy and speed-up. By constructing kinetic transition networks, we will provide mechanistic and kinetic insights into the assembly process. Finally, we will develop a coarse-grained membrane model, and use it in conjunction with our protein model to probe the various modes of oligomer-membrane interactions. Coarse-graining will enable us to unravel binding modes that were not previously observed in atomistic simulations due to insufficient sampling. We anticipate that our research will lead to a better molecular understanding of the aggregation cascade, and inspire more rational drug design. Additionally, our research promises to unravel generic rules underlying protein self-assembly, which could be exploited in the context of material design and nanofabrication.