Reinforcement Learning-Aided Design of ROS-Responsive Polymeric Nanocarriers for Hepatocellular Carcinoma Treatment

Hepatocellular carcinoma (HCC) is one of the most common malignant tumors worldwide. The main treatment methods include surgery and chemotherapy, but these often lead to significant drug resistance. Therefore, there is an urgent need to find new, low-toxicity, and efficient anti-HCC drugs. In this study, we developed a reactive oxygen species (ROS)-responsive nanoparticle drug delivery system to enhance the anti-tumor efficacy of radiotherapy through local ROS-activated chemotherapy. We synthesized a ROS-responsive polymer, PLA-TDE-PLA, for drug delivery, and encapsulated the newly synthesized compound 1 within it, constructing PLA-TDE-PLA@1. The non-polar sulfide residues in the drug delivery system are oxidized to polar sulfoxide units upon ROS activation, causing structural changes in the PLA-TDE-PLA carrying compound 1, including swelling and partial dissociation, leading to significant drug release. Cytological experiments showed that PLA-TDE-PLA@1 significantly inhibited the proliferation of HCC cells and exerted its effects by reducing the expression of Bcl-2. Through the molecular docking simulation, the proposed compound has been demonstrated to have good biological effect, further, based on the proposed compound, a generative study has been performed and some outstanding optimized molecules are analyzed in detail.