Study On QSAR Models For Predicting Toxicity And Water Quality Criteria Of Heavy Metals

Environmental pollution with heavy metals has been an increasingly serious problem.There are few standard methods available,especially in marine environment,for metal toxicity testing.In addition,systematic and comprehensive testings are costly and time-consuming.Therefore,only rarely experimental data have been published,which severely restricts the capacity of people especially the government regulators to assess water quality and to make sound environmental management decisions.As one of the most important non-experimental methods in current toxicology,quantitative structure-activity relationship(QSAR)study can be used to quickly predict and estimate toxicities of various metals based on limited experimental data,and is regarded as an efficient tool for deriving water quality criteria and assessing their ecological risks.In the present study,toxicity data of metals for marine organisms(six phyla and eight families)were collected and screened to develop QSAR models with physicochemical properties by use of linear regression methods.Toxicities of 21 metals or metalloids for which experimental data were missing were then predicted,and used to build species sensitivity distribution(SSD)curves.And continuous criteria concentrations(CMCs)of these metals were derived.We examined relationships between physicochemical properties of metals,environmental conditions and toxicities of metals to marine organisms.The influence of temperature and salinity were investigated.And risks of Hg2+,Cd2+,Cu2+,Zn2+,Ni2+ and Cr3+ in various temperatures and salinities were then assessed.In addition,relationships between toxicities of 16 metal oxide nanoparticals(MeNOPs)to E.coli and 26 physicochemical characteristics were developed to establish a nano-QSAR model.Then,we predicted cytotoxicities of 51 MeNOPs in the periodic table,and analysed potential toxicological mechanisms.In this paper,QSAR approach was used to predict toxicities of MeNOPs as well as metals,and was demonstrated to be an efficient alternative to the experimental methods.It may be as a supplement to predict toxicities,criteria and standards of metallic pollutants.