Exploring Metal Oxide Nanomaterials QSAR Modeling And Identification Of Key Features Contributing To The Cytotoxicity

Metal oxide nanomaterials are widely used in many areas, their toxic effects to humans can not be ignored. However, it is not feasible to test each new emerging nanomaterial with a complete set of toxicity assessment procedure, which will cost a lot of resources. So the application of high-throughput screening method in nanomaterials toxicity assessment will enable assess nanomaterials toxicity rapidly and efficiently, which is meaningful to provide some information for screening lower toxicity nanomaterials and optimization designation of new nanomaterials.This study first characterized the physicochemical properties of 16 metal oxide nanomaterials, including the primary morphology, specific surface area, hydration particle size and zeta potential of 16 metal oxide nanomaterials, as well as quantum chemistry parameters of these metal oxide nanomaterials obtained from published article, which were used as structural descriptors. Then the study measured the media inhibitory concentration (IC50) of 16 metal oxide nanomaterials towards normal liver cells (L02 cells) and human liver cancer cells (Hep G2 cells),which used as toxicology endpoints. On the basis of these structural descriptors and IC50 of 16 metal oxide nanomaterials towards L02 cells and Hep G2 cells, it built up two prediction models successfully. In the end, the study selected two metal oxide nanomaterials, which has significant toxicity difference, to explore the key factors in influencing the toxicity of metal oxide nanomaterials, to provid researchers a deeper understanding of quantitative nanostructure activity relationship(QNAR) prediction model and some information on further optimizing designation of new nanomaterials.Through this study, we built a metal oxide nanomaterials toxicity prediction model which can be used to predict the oxide nanomaterials toxicity towards L02 cell (n= 12, F= 23.0, R2= 0.8, p< 0.05); and another metal oxide nanomaterials prediction model which can be used to predict the oxide nanomaterials toxicity towards Hep G2 cells (n=12, F=10.5, R2=0.7, p< 0.05). These two models used three Quantum-mechanical descriptors (hard, Ec and shift) of metal oxide to predict the toxicity of nanomaterials towards L02 cells and Hep G2 cells respectively, and R2 of the two models are higher than 0.6, which meets the requirments of model building. Besides our toxicity experiments results also proved that the key features contributing to the cytotoxicity is related to the chemical property of metal oxide nanomaterials, such as the releasement of nanomaterials.Combining the model with the experimental results, the study showed that the main factors influencing metal oxide nanomaterials toxicity may be are some chemical properties of the main compounds composition of nanomaterial, such as the release of metal ions in the cell, which reminded researchers to control the stability of the nanomaterial such as reducing metal ion releasement when designing nanomaterial can reduce the toxicity of metal nanometer materials to a certain extent.