USC Neuroscientists Secure NIH Grant to Map Early Alzheimer’s Using Advanced Multiscale Hippocampal Models

USC scientists receive NIH grant to create multiscale brain models, targeting specific neuron loss in the hippocampus to find early Alzheimer's treatments.

By: AXL Media

Published: Apr 10, 2026, 8:08 AM EDT

Source: Information for this report was sourced from EurekAlert!

Decoding the Architecture of Memory Loss

A specialized team at the USC Mark and Mary Stevens Neuroimaging and Informatics Institute is launching a five year study to resolve a fundamental mystery regarding Alzheimer’s disease progression. The research, supported by an NIH R01 award, focuses on the hippocampus, the brain region primarily responsible for memory formation and one of the first areas to suffer damage. Led by Michael S. Bienkowski, an assistant professor of physiology and neuroscience, the project seeks to understand why certain neurons within the hippocampus deteriorate rapidly while others stay resilient. This investigation aims to move beyond general observations of toxic protein buildup to identify the specific cellular vulnerabilities that trigger dementia.

The Power of Multiscale Computational Modeling

Central to this initiative is the development of a multiscale model, a sophisticated digital framework that connects biological data across various levels of complexity. This system allows researchers to observe how minute changes in gene activity and cellular circuits ripple upward to affect large scale brain networks. According to Bienkowski, the project acts as a virtual testbed, enabling the team to simulate disease mechanisms at scales that are currently impossible to observe in living human patients. By bridging the gap between microscopic cellular stress and macroscopic network failure, the model provides a comprehensive view of how Alzheimer's disrupts the brain's internal communication systems.

From Gene Atlases to 3D Circuitry

The study leverages Bienkowski’s prior development of the Hippocampus Gene Expression Atlas, a detailed map that classifies neuron populations by their genetic activity and wiring patterns. To enhance this framework, the team will utilize 3D circuit reconstruction and advanced molecular imaging to analyze both donated human brain tissue and mouse models. This collaborative effort includes computational experts Gianluca Lazzi and Jean-Marie Bouteiller, who bring specialized knowledge in modeling the retina and hippocampus. Their combined work will help pinpoint the exact cell types that exhibit signs of structural damage or altered gene expression long before actual neuron death occurs.