Computational Biology Studies Move Researchers Closer to Treating Rare Form of Alzheimer’s Disease

Published February 23, 2023

By Kimberly Mann Bruch

According to the Centers for Disease Control and Prevention, approximately 6.5 million Americans live with Alzheimer’s disease, which is the most common cause of dementia in older adults. University of Kansas (KU) researchers who study familial Alzheimer's disease (FAD) – a rare form of Alzheimer's disease caused by changes in inherited genes – turned to Expanse at the San Diego Supercomputer Center at UC San Diego to conduct computational biology studies on how mutations of a critical protein enzyme called γ-secretase could be treated to better control thought, language and memory.

KU graduate students Hung Do, Kushal Koirala and Apurba Bhattarai, led by Associate Professor of Molecular Biosciences Yinglong Miao, used Expanse to build the first dynamic models that show the activation and substrate processing of γ-secretase in the absence and presence of various FAD mutations. The team also worked with KU Professor of Medicinal Chemistry Michael Wolfe and Postdoctoral Researcher Sujan Devkota to compare his laboratory’s biochemical experimental data with Miao’s team’s computational biology data. The highly consistent findings have been published in ACS Central Science, Journal of American Chemical Society, and more recently in the Nature Communications Biology Journal.

“These studies provide a mechanistic basis to guide our rational drug discovery efforts of γ-secretase for treating Alzheimer’s disease,” Miao said. “Our laboratories were able to combine cutting-edge accelerated molecular simulations using Expanse GPUs with highly complementary mass spectrometry and western blot biochemical experiments to investigate the functional mechanisms of substrate cleavage by γ-secretase and effects of these FAD mutations.”

Wolfe explained that the γ-secretase complex, called “the proteasome of the membrane,” uses water within the otherwise water-repelling environment of cell membranes to detach a certain protein, known as APP, to form a smaller protein called beta-amyloid, which deposits as plaques in the brain in FAD. “Because scientists have already discovered that mutations in γ-secretase and APP lead to early-onset FAD, computational biologists like the Miao group work with medicinal chemists like my group on how to treat this devastating disease. Our recent studies get us one step closer,” Wolfe said.

The original study built on new cryo-EM structures of γ-secretase, cutting-edge simulation techniques such as Gaussian accelerated molecular dynamics (GaMD) developed in Miao’s lab to decipher functional mechanisms of γ-secretase, and effects of FAD mutations, and the unique expertise of Wolfe and his lab to examine effects of FAD mutations on all the cleavage events of APP by γ-secretase.

In the most recent study, the researchers combined advanced GaMD simulations and biochemical experiments to determine the effects of six representative FAD mutations (P117L, I143T, L166P, G384A, L435F, and L286V) in the γ-secretase catalytic subunit Presenilin-1 (PS1). Biochemical experiments showed that all six FAD mutations rendered g-secretase less active for a specific cleavage of APP.

“Our simulations and experiments combined have provided unprecedented mechanistic insights into how PS1 FAD mutations affect structural dynamics and enzyme-substrate interactions of g-secretase and APP,” said Miao, adding that the results from this current work help his team pursue three long-term research goals.

“Our primary objectives are to obtain comprehensive understanding of the structural dynamics and functional mechanisms of γ-secretase in the wildtype and disease-causing mutant forms; determine the mechanisms of recognition and processing of various substrates by γ-secretase; and design selective and potent drug molecules of γ-secretase to treat Alzheimer’s disease and other related human diseases,” Miao said. “Supercomputers like Expanse allow us to accomplish these goals – one simulation at a time.”

This work was funded by the National Science Foundation (award no. 2121063)  and the National Institutes of Health (award no. AG66986). Supercomputing time on Expanse was funded by the NSF’s Extreme Science and Engineering Discovery Environment (allocation no. MCB180049).