Background:
Polymeric nanoparticles enable targeted drug delivery in cancer therapy, yet concerns regarding nanotoxicity and immune dysregulation limit their clinical translation. Artificial intelligence (AI) offers a promising approach to predict and optimize nanoparticle safety profiles prior to experimental validation.
Objective:
To develop an AI-assisted predictive framework for evaluating the nanotoxicological and immunopharmacological properties of folate-conjugated polymeric nanoparticles (FA-PNPs), followed by experimental validation in cancer and immune cell models.
Methods:
An AI-based random forest model was first developed to predict nanoparticle toxicity using physicochemical parameters. FA-PNPs (mean size: 112 ± 8 nm; zeta potential: −18.6 ± 2.1 mV) were then synthesized and characterized. Predicted outcomes were validated experimentally using MTT cytotoxicity assays (MCF-7 and Jurkat cells, 24–72 h), ROS generation (DCFH-DA assay), cytokine profiling (IL-6, TNF-α, IFN-γ via ELISA), and fluorescence-based cellular uptake studies. Non-functionalized polymeric nanoparticles (PNPs) were included as controls for comparative analysis.
Results:
The AI model demonstrated strong predictive performance (accuracy: 91.3%) and showed high correlation with experimental toxicity outcomes (R² = 0.87). FA-PNPs exhibited selective cytotoxicity toward MCF-7 cells (IC₅₀ = 42.7 µg/mL) compared to Jurkat cells (IC₅₀ = 91.5 µg/mL; p < 0.01). ROS generation increased by 2.8-fold in cancer cells versus 1.3-fold in immune cells. Pro-inflammatory cytokines IL-6 and TNF-α decreased by 34% and 29%, respectively, while IFN-γ increased by 21%. Compared to non-functionalized PNPs, FA-PNPs showed a 2.5-fold enhancement in cellular uptake.
Conclusion:
This study demonstrates that AI-guided prediction combined with experimental validation provides a robust framework for assessing nanoparticle safety and efficacy. FA-PNPs exhibit enhanced targeting, reduced inflammatory toxicity, and favorable immunomodulatory effects, supporting their potential as safer precision nanotherapeutics for cancer treatment.
