Researchers from the University of Arizona College of Medicine – Tucson and Harvard University have utilized the power of artificial intelligence (AI) to delve deep into the human brain, mapping the molecular changes that occur in healthy neurons as Alzheimer’s disease progresses. The findings, published in Nature Communications Biology, shed light on potential drug targets and causes of the neurodegenerative disorder.

Alzheimer’s disease poses a significant challenge in the field of medicine, with its complex nature leading to irreversible symptoms such as dementia, memory loss, and personality changes. While drugs can alleviate certain symptoms, discovering a cure has remained elusive, mainly due to the unclear cause of the disease.

Through an AI algorithm developed by Dr. Rui Chang, an associate professor of neurology, the research team analyzed tissue samples from over 2,000 Alzheimer’s brains sourced from a national database. The algorithm drew upon a vast repository of genetic and molecular information, resulting in a comprehensive computational network model of the human brain. This enabled the researchers to map the interactions of whole-genome genes and track the sequential changes in their relationships during the development of Alzheimer’s, offering insights into the disease’s origins and the molecular pathway from health to illness.

Dr. Chang likens the progression from a healthy state to Alzheimer’s to a watercourse, where the accumulation of amyloid plaques and tau tangles—the abnormal structures found in the Alzheimer’s brain—occurs downstream as a response to issues upstream.

Targeting the downstream effects directly, such as the abnormal structures, has proven ineffective in clinical trials. This suggests that they are more likely consequences rather than the cause of Alzheimer’s. Dr. Chang emphasizes the importance of targeting the disease upstream and concurrently addressing multiple targets. To achieve this, understanding the entire landscape is crucial, and AI serves as a novel method to unravel complex data into a network model, providing a clear overview of upstream events and identifying genes that trigger downstream effects like amyloid plaques and tau tangles. These upstream genes hold promise as better targets for potential therapies.

Using AI, Dr. Chang identified 19 neuron-specific genetic points along the Alzheimer’s pathway that seem to push neurons closer to a diseased state. Collaborators at Harvard validated these genes’ role in Alzheimer’s development by deactivating them in stem cell-derived neurons. Their experiments revealed that 10 of these genes influenced the production of plaques and tangles, positioning them as potential drug targets for Alzheimer’s treatment.

Once gene targets are identified, the next step involves finding drugs that can effectively interact with those targets. Dr. Chang employed 3D computer models to assess if existing molecules and drugs align with potential drug targets, akin to a key fitting into a lock. This comprehensive approach, examining thousands of targets simultaneously, accelerates drug development and discovery, narrowing down the field to around 3,000 drug candidates of interest.

Dr. Chang expresses his enthusiasm for the journey from mathematical algorithms and data analysis to clinical studies, with the ultimate goal of benefiting patients. This research represents a significant step forward in understanding Alzheimer’s disease and advancing potential treatments, offering hope for the millions affected by this challenging condition.

Through virtual screenings, millions of compounds, including FDA-approved drugs, natural products, and small molecules, were evaluated against the identified targets. The team is now investigating several small molecules further, and with the support of a National Institutes of Health grant, they are preparing for clinical trials on three of the compounds. Human trials are expected to commence in the near future.

By Impact Lab