Mutagenicity and Genotoxicity of Organic Extracts from Finished Water with Different Treatment Process

Purification of surface water is widely practiced with conventional water treatment processes like coagulation-flocculation, sedimentation, filtration, and disinfection. Some reports have specified that conventional wastewater purification processes do not effectively remove many chemical contaminants, and that treatment may increase the mutagenicity or genotoxicity of wastewater^[1]. Moreover, these treatment processes are increasingly experiencing operational difficulties owing to the widespread pollution of water resources^[2].

According to recent water quality monitoring results, the source water in our city contains many organic pollutants, such as 6-chlorobenzene, dichloroacetic acid, trichloroacetic acid, atrazine and benzo (a) pyrene etc. However, it is not clear whether these organic pollutants can be effectively removed by traditional water treatment processes. There has been no in-depth and comprehensive research on water quality testing and assessment in the city. Some pollutants can interact with naturally occurring chemicals in the water and their toxicity may be increased^{[3, 4]}. Therefore, it is impossible to predict toxic properties from routine physical/chemical measurements with a sufficient level of safety. It is important to include biological toxicity assays in drinking water monitoring programs. In this study, water samples from six water plants in the city were collected and concentrated and organic extracts were tested for their mutagenicity and genotoxicity by using the Ames test, a microtiter fluctuation test, and a micronucleus test.

Treated water from six water plants (samples marked A, B, C, D, E, and F) were collected between April and July 2009. The water source and characteristics of the treatment process in the six water plants are shown in Supplementary Table S1 (available in www.besjournal.com).

From each water plant, a 150-L water sample was collected and hydrochloric acid was added to adjust the pH to 2.0. Sampling was performed according to the recommended standard method^[5]. The water samples were concentrated using ion exchange adsorption with an XAD-7 macroporous resin according to the manufacturer's guidelines. After resin activation, the water was passed through XAD-7 to adsorb organic pollutants. The velocity of flow was controlled at 30 mL/min. Organic matter was eluted with 300 mL dichloromethane/n-hexane (v:v, 85:15) and 200 mL acetone (at a rate of 3-5 mL/min). The organic solvents were evaporated to a small volume at 30 ℃ under reduced pressure with a rotary evaporator and then dried by blowing with a nitrogen stream. The dry residue was re-dissolved in dimethylsulfoxide (DMSO). Further dilutions were performed as necessary. Samples were stored in a freezer at -20 ℃ until use.

The Ames test was carried out with Salmonella typhimurium strains TA98 and TA100 with and without in vitro metabolic activation. S. typhimurium strains TA98 and TA100 were supplied by the Ames Laboratories California, USA. Liver S9 fractions obtained from polychlorinated biphenyls (PCBs)-induced Sprague-Dawley rats were used as an exogenous metabolic activation system. Overnight grown cell suspensions (approximately 10⁹ cfu/mL) of TA98 and TA100 were prepared. All tests were performed in triplicate plates for two independent experiments. Positive controls were included in all experiments in order to confirm strain sensitivity and S9 activity. Each sample was tested with three concentrations containing 0.25, 0.5, or 1 L equivalent of concentrated organic matter per plate respectively. S9, used as the metabolizing system, was added at 500 μL per plate. Once prepared, each culture was incubated for 20 min at 37 ℃. After the first incubation period, agar was mixed into the cultures, and each tube was poured onto the surface of a Petri dish prepared with the standard medium. The plates were then incubated for two days at 37 ℃. After the 2^nd incubation, histidine-reverted colonies on each plate were counted with an Artek model 880 automatic colony counter. Statistical analysis of the Ames test was conducted as suggested by the EPA Gene-Tox work group report for the mutation assay^[6]. The plates were prepared in triplicate for every test sample, and the result presented is the mean of triplicate counts (± standard deviation). The mutagenic effect was evaluated from the number of revertant colonies per plate. The sample was considered positive only if the mutation rate (MR) was ≥ 2.00 (where MR = mutant colonies on test plate/spontaneous mutant colonies on negative control plate).

Water samples from water plant C were measured by a microtiter fluctuation test, which was carried out as described by Gatehouse^[7] with Salmonella typhimurium strains TA98 and TA100 without in vitro metabolic activation. After incubation at 37 ℃ for 24 h, 20 μL of bromocresol purple was added to each well. The culture was grown for a further 72 h and the number of reactive wells was observed. A reaction was considered positive when the culture in the well turned from purple to yellow. Triplicate plating was done at each dose level. In addition, the relative sensitivity of the Ames test and microtiter fluctuation test was compared.

In the mouse micronucleus assay, NIH male Swiss mice (provided by the Laboratory Animal Center of Guangdong Province, 5 mice per treated group) weighing 18-20 g were used. The post chlorine-added water sample from water plant D was measured with the micronucleus test. The organic extract was diluted with DMSO to prepare 6 concentrations containing 0.2, 0.8, 3.1, 12.5, 50, and 200 L equivalent/kg BW, respectively.

Each mouse was given a dose of 10 mg/kg BW sample through intragastrical administration (IG) according to body weight. The mice were administrated twice with an interval of 24 h and sacrificed 6 h after the second administration. Cyclophosphamide (CP) at 40 mg/kg body weight was used as the positive control and the negative control group received the vehicle only. Bone marrow from the mouse femurs was smeared on a glass slide. Two slides were prepared per mouse. Slides were left to air dry, fixed with absolute methanol, and stained with Giemsa and May- Grunwald strain solutions. The number of micronuclei in polychromatophilic erythrocytes (PCE) was counted. The same test was conducted at a concentration of 12.5 L/kg BW for the organic extracts from all six water plants as mentioned above.

A total of 1, 000 PCE/slide were scored with a light microscope for the evaluation of micronucleus (MN) frequencies. The sample was considered positive for micronucleation when the mean MN frequency was at least more than twice the spontaneous frequency.

Statistical analyses of the data were performed by a Chi-square test (χ² test) and regression analysis using SPSS 18.0 software (IBM, chicago). The Chi-square test (χ² test) was conducted to analyze the microtiter fluctuation assay results and micronucleus assay results. Regression analysis was performed to analyze the specific mutagenic activity of organic matter.

In the experimental concentration range, samples from the six water plants were all positive in the Ames test and the positive results were as follows: A: TA98(-S9), TA100(-S9); B: TA98(-S9); C: TA98(+S9), TA98(-S9), TA100(-S9); D: TA98(+S9), TA98(-S9); E: TA98(+S9), TA98(-S9), TA100(+S9), TA100(-S9); F: TA98(+S), TA98(-S9), TA100(-S9). The specific activity parameter method^[8] was applied to compare the mutagenicity of water samples from different sources. The specific activity is defined as the number of revertant colonies per unit volume (number of revertant colonies/L-equivalent, rev/L-equivalent). The number of revertant colonies in the Ames test (MN value) for each water sample was used to plot the dose-response curve and the slope of the linear part indicates the specific activity of the water sample (Table 1).

Positive results were obtained from the microtiter fluctuation test for all water samples measured for TA98 and TA100 at different concentrations (Table 2 and Supplementary Figure S1 available in www.besjournal.com). The results of chlorinated water samples from water plant C showed positive results for TA98 and TA100 at 0.25 L/plate (P < 0.05). The figure shows that R² of the number of positive wells for TA98 and TA100 was 0.9529 and 0.9295, respectively, and there was a clear dose-response relationship. For the water plant C measured by both Ames test and microtiter fluctuation test, positive results were obtained from the latter in lower concentrations.

Figure Supplementary Figure S1.  Dose-response relationship in microtitre fluctuation test for water sample from water plant C. The number of positive cells R² of TA98 and TA100 is 0.9529 and 0.9295 respectively, and there is a clear dose-response relationship.

There was a significant difference in micronucleus frequency between each dose group and the negative control group except the 0.2 and 0.8 L/kg BW dose group (Supplementary Table S2 available in www.besjournal.com). The 12.5 L/kg BW dose group was the most sensitive and chosen to measure water samples from the six water plants. Micronucleus frequency of all the water samples from the different plants at a dose of 12.5 L/kg BW were significantly higher than that of the negative control (P < 0.05) (Table 3).

The organic extracts from the treated water showed mutagenic activity in both the S. typhimurium strains, TA98 and TA100, but they had greater specific activity without S9 than that with S9, which suggested that the water plants in the city contained higher direct mutagens than indirect mutagens. The mutagenicity of the water samples from different water sources, with different disinfectants, or different treatment process were pairwise compared (Supplementary Table S3 available in www.besjournal.com).

A comparison of plants A and E, which had different water sources, indicated that river water was better to be used as source water than a mixture of reservoir and river water in the city. The imperfect ecological structure of the semi-artificial reservoir, with cumulative pollutant deposition, and the production of disinfection by-products (DBPs) during the chlorination process may have cause the increased mutagenicity of this water.

A comparison of plants D and E, which had different disinfectants, indicated that chlorine dioxide was better than liquid chlorine as a disinfectant. It has been found that liquid chlorine disinfection can produce carcinogenic substances such as trihalomethanes (THMs), while chlorine dioxide disinfection can effectively destroy the organic pollutants in water and can decrease THMs and other substances^[9].

A comparison of plants A and F, which have different treatment processes, indicated that there were pros and cons between the conventional chlorination and the booster chlorination. Pre-chlorination can improve the effect of coagulation, it also tends to increase the formation of DBPs. One previous study showed that booster chlorination increased direct frameshift mutagens in tap water^[10]. A similar result was obtained in this study. A comparison of plants A and B indicated that activated carbon filtration could reduce the mutagenic activity in the water treatment process. It can remove a variety of pollutants (such as hydrocarbons, PAHs, etc.) in water effectively^{[11, 12]}.

Microtiter fluctuation test was more sensitive than the Ames test in detecting mutagenicity, consistent with a previous report^[13]. Compared to the Ames test, a microtiter fluctuation test may be more suitable for the mutagenicity test of water samples.

In this study, the results of the micronucleus test and Ames test were fundamentally uniform. The results from these in vitro tests were further supported by the in vivo test results. The two experiments can reflect the genotoxic effects of organic matter in water comprehensively from the genetic endpoints of gene mutation and chromosomal aberration, respectively.

In summary, the organic compounds from six water plants in the city were obviously mutagenic. The following measures can be taken to reduce the formation of mutagenic organic substances: using river water as source water; eliminating pre-chlorination; replacing liquid chlorine with chlorine dioxide as a disinfectant; and applying activated carbon filtration. However, this study was conducted in one particular city, so it is difficult to make extensive conclusions. In addition, there are many other factors that influence the content of organic compounds in water, such as chlorine, reaction time, and type of pollutants. Next, we need to conduct a comprehensively detailed exploration and research to evaluate the combined effects of various factors and to look for optimal disinfection procedures.

We wish to express our gratitude to the 'Research Base for Environment and Health in Shenzhen Center for Disease Control and Prevention, Chinese Center for Disease Control and Prevention' and the participants for their cooperation in this study.