Abstract
This study utilizes ultrasound-assisted enzymatic extraction technology to optimize the extraction of antitumor bioactive compounds from Sargassum polycystum and evaluates their antitumor effects and mechanisms in non-small cell lung cancer. Multi-objective optimization was performed by combining response surface methodology and non-dominated sorting genetic algorithm, precisely controlling key extraction parameters such as the liquid-to-solid ratio, ethanol concentration, enzyme concentration, and pH value, which significantly enhanced the extraction yields of phenolic acids and lactone compounds. Compared to the RSM optimization, the NSGA-II optimization further refined the combination of parameters including the liquid-to-solid ratio, ethanol concentration, enzyme concentration, and pH value. As a result, the total phenolic acid yield increased by 3.34% (20.4928 mg/g vs. 19.83 mg/g), and the lactone yield increased by 0.60% (24.6062 mg/g vs. 24.46 mg/g). The optimized extract significantly inhibited tumor growth in the NCI-H1975 cell line, primarily by targeting the cathepsin B (CTSB) pathway to regulate cell proliferation and apoptosis. Biological evaluations analysis further confirmed the high affinity of the active components for CTSB and identified a multi-target mechanism that regulates the cell cycle, apoptosis, and metastasis, enhancing the extract's antitumor efficacy. The study demonstrates that ultrasound-assisted enzymatic extraction combined with multi-objective optimization not only effectively enhances the antitumor activity of S. polycystum but also provides important scientific evidence for its development as a potential antitumor drug, showcasing the broad application potential of ultrasound-assisted extraction technology in natural product development.