Mechanistic and AI-assisted evaluation of paclitaxel-loaded HSPC liposomes for targeted breast cancer therapy
摘要
Paclitaxel (PCL) faces translational barriers including poor aqueous solubility (log P = 3.0) and dose-limiting toxicity. Hydrogenated soy phosphatidylcholine (HSPC) liposomes were engineered exploiting the lipid’s high phase transition temperature (Tm ~ 52 °C) for membrane rigidity and sustained drug retention. Synthesized via thin-film hydration (HSPC:cholesterol:PCL = 7:2:0.1), nano-formulations achieved a mean diameter of 142.3 ± 8.7 nm and PDI of 0.178 ± 0.01 by dynamic light scattering, a zeta potential of − 28.4 ± 1.5 mV determined by electrophoretic light scattering, 87.6 ± 2.1% encapsulation efficiency, and controlled drug release (78.4 ± 5.2% over 72 h). Against MCF-7 breast adenocarcinoma cells, liposomal PCL demonstrated time-dependent cytotoxic amplification with IC₅₀ improving from 0.525 µg/mL (24 h) to 0.09 µg/mL (48 h)—54% superior to free PCL (IC₅₀ = 0.139 µg/mL). Mechanistic studies were conducted, revealing (i) 2.96-fold G2/M phase accumulation (63.23% vs. 21.38% control, p < 0.001), (ii) mitochondrial apoptosis with 2.8-fold caspase-3 and 3.1-fold caspase-9 activation, (iii) 4.2-fold enhanced cellular uptake, and (iv) anti-metastatic activity with 62% migration and 68% invasion inhibition. Complementary computational analyses included the following: principal component analysis (PCA) revealing multivariate separation of treatment groups (PC1, 78.3% variance; PC2, 16.4%); Grad-CAM visualization localizing morphological features associated with apoptosis; and exploratory Random Forest modeling demonstrating time and formulation type as primary determinants of cytotoxic potency. Three-month stability assessment showed minimal drift in physicochemical parameters (particle size, + 10.2%; encapsulation efficiency, − 6.3%), supporting formulation robustness for preclinical studies.