AI-Driven Optimization of Curcumin-Loaded Polymeric Nanoparticles for Targeted Therapy in Alzheimer's Disease

Alzheimer's disease (AD) remains one of the most devastating and therapeutically recalcitrant neurodegenerative disorders worldwide, characterized by progressive amyloid-β plaque deposition, neurofibrillary tau tangles, neuroinflammation, and oxidative stress culminating in irreversible cognitive decline. Curcumin, the principal bioactive constituent of Curcuma longa, possesses extraordinary neuroprotective, anti-amyloidogenic, anti-inflammatory, and antioxidant properties; however, its clinical translation has been profoundly constrained by poor aqueous solubility, rapid systemic metabolism, and particularly inadequate blood-brain barrier (BBB) penetration. Encapsulation within polymeric nanoparticles (PNPs) constructed from biocompatible and biodegradable polymers such as poly(lactic-co-glycolic acid) (PLGA) and polycaprolactone (PCL) offers a compelling strategy to circumvent these pharmacokinetic deficiencies. Nevertheless, the intrinsically high-dimensional and nonlinear nature of nanoformulation design—governed by complex interdependencies among polymer concentration, surfactant type, solvent ratios, drug-to-polymer ratios, and processing parameters—renders empirical optimization approaches inadequate. This study presents the systematic development, artificial intelligence (AI)-assisted optimization, and in vitro evaluation of curcumin-encapsulated PLGA-PCL polymeric nanoparticles engineered for targeted BBB traversal in Alzheimer's disease therapy. Artificial neural network (ANN) and response surface methodology (RSM) frameworks, integrated within a Quality by Design (QbD) paradigm, were employed to identify formulation optima predictively and with mechanistic resolution. Optimized nanoparticles exhibited a mean hydrodynamic diameter of 142.3 ± 6.8 nm, zeta potential of −28.7 ± 1.4 mV, encapsulation efficiency of 87.4 ± 2.1%, and sustained drug release exceeding 72 hours conforming to Korsmeyer-Peppas kinetics. In vitro BBB permeability assays on hCMEC/D3 monolayers demonstrated an apparent permeability coefficient (Papp) of 18.6 × 10−6 cm/s, with SH-SY5Y neuroprotection assays confirming substantial cytoprotective activity against amyloid-β 1-42-induced neurotoxicity. ANN models attained a predictive R2 of 0.9891, outperforming classical RSM approaches. These findings validate AI-driven computational optimization as a transformative paradigm in nanoformulation science and position curcumin-loaded PNPs as a scientifically compelling candidate for ADtargeted nanomedicine.

Hali tarjima qilinmagan