Researchers, spearheaded by Mina Zhang and Dabaowen Zhang from the UC Irvine Joan and Don Bren School of Population Health, have constructed the most comprehensive maps yet detailing how genes causally govern each other within distinct brain cell types impacted by Alzheimer’s disease.
Employing SIGNET, a novel machine learning framework they developed that pinpoints causal links rather than mere genetic associations, they uncovered pivotal biological pathways potentially responsible for memory loss and cerebral degeneration.
This study, featured in Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association, also brought to light previously unidentified genes that could serve as future therapeutic targets.
Alzheimer’s disease, the leading cause of dementia, is projected to affect nearly 14 million Americans by 2060. Although numerous genes related to the condition, such as APOE and APP, have already been identified, the precise mechanisms by which these genes disrupt normal brain function remain incompletely understood.
“Different brain cell types contribute uniquely to Alzheimer’s, but their interactions at the molecular level have been opaque,” stated co-author Professor Ming Zhang. “Our work furnishes cell-type-specific regulatory maps for Alzheimer’s in the brain, shifting the research focus from observing correlations to establishing the causal mechanisms actively driving disease progression.”
To generate these maps, the team analyzed single-cell molecular data extracted from the brain samples of 272 participants in longitudinal studies on memory and aging. They engineered SIGNET as a scalable, high-throughput computational method that integrates single-cell RNA sequencing and whole-genome sequencing data to reveal causal relationships among all genes.
The researchers established causal gene regulatory networks for six major brain cell categories. This capability allowed them to single out which genes are likely commanding others—a feat traditional correlation-based tools cannot reliably achieve.
“Most gene mapping tools can show which genes are linked, but they can’t discern which genes are actually the driving force behind the changes,” explained co-author Professor Dabaowen Zhang. “Some existing methods also impose unrealistic assumptions, such as overlooking feedback loops between genes. Our approach leverages information encoded in the DNA itself to uncover the true causal relationships between genes in the brain environment.”
The scientists determined that the most significant genetic perturbations in Alzheimer’s afflict excitatory neurons—the nerve cells that transmit activating signals—and their analysis of nearly 6,000 causal interactions indicated a widespread remodeling within these cells as the disease advances.
Furthermore, they pinpointed hundreds of “hub genes” functioning as central controllers, influencing numerous other genes and likely playing a critical role in initiating detrimental changes. These genes present promising new targets for early detection and therapeutic intervention. Additionally, the team uncovered novel regulatory roles for well-known genes, such as APP, which exerts significant control over other genes within inhibitory neurons.
Crucially, the investigators validated these findings utilizing an independent cohort of human brain samples, enhancing confidence that these gene interdependencies reflect genuine biological mechanisms involved in Alzheimer’s pathogenesis. SIGNET holds potential utility for investigating a wide array of other complex conditions, including cancer, autoimmune disorders, and psychiatric illnesses.
About the Author
