Study On Evaluation And Prediction For The Combined Toxicity Of Mixtures Using Luminescent Bacteria Test

A lot of scientific evidence proved that the biotoxicity of mixture might always be higher than the sum of each component and therefore the study of prediction and evaluation of mixed biotoxicity became hot topics. At present, the main evaluation models for the action of combined toxicity are the combined effect index methods,the dose addition(DA) model and independent action(IA) model, but each model has its own characteristics and applicable conditions, it is too difficult to bring the three models into a unified assessment system. The combined effect index methods were used for evaluating the action of combined effects of binary mixtures, but the different models often draw the inconsistent conclusions. DA and IA models are used to evaluating the action of components instead of predicting the mixture biotoxicity, as the mechanism of many chemicals on the organisms is not clear, DA or IA was selected mainly based on experience. Therefore, it is necessary to analyze the characteristics of the existing models to explore the ways which can improve mixture toxicity evaluation and prediction by theoretical and experimental studies.Among the biological toxicity tests, the luminescent bacteria toxicity assay was was conducted in this study which has been widely applied in the field of environmental toxicology due to the advantages of short culture time, easy operation and the test results comparable to higher organisms, such as fish, protozoa, etc. However, it is necessary to optimize the sensitivity and reproducibility of the luminescent bacteria toxicity testing which is often affected by the test environment, especially the state of bacterial suspension.This study is part of the project of National Natural Science Foundation of China. Firstly, the optimization of luminescent bacteria biotoxicity test was carried out using Vibrio qinghaiensis sp. Q67 as testing bacterium. From the aspect of the influence of culture, growth stage and bacterial density, the sensitivity and reproducibility of Q67 toxicity test was evaluated. Finally, the method of toxicity test using Q67 was established with high sensitivity and reproducibility. Secondly, the limitations of toxic unit method(TU) was analyzed based on a series of binary mixtures of typical chemicals, and then the model of the nonlinear TU(TU’ model) was established according to the definition of the combined effects. Then a calculation method which can predict the action of combined effects of environmental samples was established with TU’ model. Thirdly, the characteristic parameters which can determine the applicability of DA were proposed on the basis of the dose-response characteristics of single chemical, binary mixtures and multi-component mixtures. These achievements are the foundations for the unified mixture toxicity evaluation and prediction system.The main work and achievements are as follows:(1) In order to improve the reproducibility and sensitivity of luminescent bacteria toxicity test, Q67 as the testing bacteria was selected and the influence of the culture and storage time of bacteria was studied. The parameters of Q67 suspension, such as pH, relative light unit(RLU) and bacterial density(D600), were detected to discuss the reasons of the great variations of biotoxicity. It was found that the bacterial density was the key factor for the reproducibility. According to the analysis of the growth process of Q67,it was found that as long as the same bacteria density of the bacterial suspension, the Q67 with different storage time are in the same stage of the logarithmic growth phase. All the five detected chemicals including heavy metals and organic compounds and three industrial wastewaters have the highest toxicity when the D600 of Q67 bacterium suspension is 2.0 and the reproducibility is good. When the D600 is greater or less than 2.0, Q67 toxicity tests are either insensitive or appearing strong stimulated with luminescence. Thereby the D600 = 2.0 was confirmed as a control indicator for bacteria culture using in the subsequent research.(2) The improved model was applied to evaluate the action of combined biotoxicity of mixture accurately. The joint action of binary mixtures of five chemicals with different dose-response curves(DRCs) was analyzed on the basis of the combined effect index methods- Toxic Unit model(TU). It was found that TU model was only appropriate to accurately evaluate the joint action of substances with the linear DRCs but not adequate for those with the non-linear DRCs. The non-linear TU’ model was developed according to the definition of joint action and the framework of TU model. The TU’ model was more appropriate for evaluating the joint action of multi-component mixture and natural mixture. The calculation method to predict joint action of mixture was constructed based on the individual, binary toxicity of the components and the combined effects of a binary mixture evaluated by TU’ model. The results showed that: the DRCs reflect the mode of chemicals on the tested organism to some extent. It‘s prone to appearing antagonism or synergy when the mixture was composed by substances with different sharp of DRCs. This method has been testified to be quite qualified for predicting the joint action using a series of experiments.(3) The index which could determine the applicability scope of DA was proposed by analyzing the acute toxicity of five chemicals and their binary and five-component mixtures to Q67. The results showed that DA model seriously overestimated the biotoxicity of all the mixtures, but IA model could effectively predict the biotoxicity of mixtures. The mathematical analysis was introduced for the DRC of chemicals. It was found that the biotoxicity of chemicals with the same toxic unit is closely related with k·ECx(ECx is the concentration of the component that provokes x% effect when it was applied individually, k is the value of derivative function of dose-response equation at the concentration ECx). Only when the chemicals have similar k·ECx values, they could have similar effect on organism and could be replaced by each other. The suitability of k·ECx as the discrimination indicator of was confirmed by experimental and literature data. If the mixture contains substances whose k·ECx value is much larger than the rest of the components, DA model may overestimate the biotoxicity of the mixture. Otherwise, DA model may underestimate the biotoxicity of the mixture. Only when the k·ECx value of each component in the mixture are almost equal, DA model can accurately predict the mixture toxicity.These results have important theoretical significance and application value for improving the accuracy of mixture toxicity evaluation and prediction.