Deep Learning‐Driven Exploration of Pyrroloquinoline Quinone Neuroprotective Activity in Alzheimer's Disease

Abstract

Abstract Alzheimer's disease (AD) is a pressing concern in neurodegenerative research. To address the challenges in AD drug development, especially those targeting Aβ, this study uses deep learning and a pharmacological approach to elucidate the potential of pyrroloquinoline quinone (PQQ) as a neuroprotective agent for AD. Using deep learning for a comprehensive molecular dataset, blood–brain barrier (BBB) permeability is predicted and the anti‐inflammatory and antioxidative properties of compounds are evaluated. PQQ, identified in the Mediterranean‐DASH intervention for a diet that delays neurodegeneration, shows notable BBB permeability and low toxicity. In vivo tests conducted on an Aβ₁₋₄₂‐induced AD mouse model verify the effectiveness of PQQ in reducing cognitive deficits. PQQ modulates genes vital for synapse and anti‐neuronal death, reduces reactive oxygen species production, and influences the SIRT1 and CREB pathways, suggesting key molecular mechanisms underlying its neuroprotective effects. This study can serve as a basis for future studies on integrating deep learning with pharmacological research and drug discovery.

Keywords

• Alzheimer's disease
• deep learning
• neuroprotective activities
• pyrroloquinoline quinones