A team led by neurobiologist Prof. Dr. Hilmar Bading at Heidelberg University has identified a key molecular process that drives the progression of Alzheimer’s disease. Working with researchers from Shandong University (China), the scientists used a mouse model of Alzheimer’s to show that a harmful protein interaction causes brain cells to die, leading to cognitive decline. The findings point to new possibilities for developing more effective treatments.
The harmful protein interaction involves two components previously studied: the NMDA receptor and the TRPM4 ion channel. NMDA receptors play an essential role in communication between nerve cells and are located on the cell surface, both at synapses and in areas outside these junctions. They are activated by glutamate, a key neurotransmitter.
When NMDA receptors function within synapses, they support neuron survival and help maintain cognitive function. However, when TRPM4 interacts with NMDA receptors outside synapses, it alters their behavior in a harmful way. Together, they form what researchers describe as a “death complex” that can damage and kill nerve cells, explains Hilmar Bading, director of the Institute of Neurobiology at Heidelberg University’s Interdisciplinary Center for Neurosciences (IZN).
Experimental Drug Breaks the Toxic Protein Link
The study found that this neurotoxic NMDAR/TRPM4 complex appears at much higher levels in Alzheimer’s mice compared to healthy ones. To target this mechanism, the researchers used a compound called FP802, a “TwinF Interface Inhibitor” previously developed by Prof. Bading’s team.
In mouse experiments, FP802 successfully disrupted the interaction between TRPM4 and NMDA receptors. The molecule binds to the “TwinF” interface where the two proteins connect, preventing them from interacting and effectively breaking apart the toxic complex.
Slowed Disease Progression and Preserved Memory
“In Alzheimer’s mice treated with the molecule, disease progression was markedly slowed,” states Dr. Jing Yan, formerly part of Prof. Bading’s team and now with FundaMental Pharma. The treated animals showed far less of the typical cellular damage associated with Alzheimer’s. This included reduced loss of synapses and less structural and functional damage to mitochondria, the powerhouses of the cell.
Importantly, learning and memory abilities remained largely intact. The researchers also observed a significant drop in beta-amyloid buildup in the brain, a hallmark of Alzheimer’s disease.
A New Treatment Strategy Beyond Amyloid
Prof. Bading emphasizes that this approach differs from traditional Alzheimer’s strategies. “Instead of targeting the formation or removal of amyloid from the brain, we are blocking a downstream cellular mechanism, the NMDAR/TRPM4 complex, that can cause the death of nerve cells and — in a disease-promoting feedback loop — promotes the formation of amyloid deposits,” he explains.
Earlier research by the team showed that FP802 also provides neuroprotective effects in models of amyotrophic lateral sclerosis (ALS), another neurodegenerative disease involving the same protein interaction.
Future Potential and Next Steps
The researchers believe this inhibitor could represent a broadly applicable strategy for slowing or stopping neurodegenerative diseases such as Alzheimer’s and ALS. However, Prof. Bading cautions that clinical use is still far in the future. “The previous results are quite promising in the preclinical context, but comprehensive pharmacological development, toxicological experiments, and clinical studies are needed to realize a possible application in humans,” he says.
Efforts are now underway, in collaboration with FundaMental Pharma, to further refine FP802 for potential therapeutic use.
Funding and Publication
The research received support from the German Research Foundation, the European Research Council, the former Federal Ministry of Education and Research, the National Natural Science Foundation of China, and the east Chinese province of Shandong. The findings were published in the journal Molecular Psychiatry.
