Final Report: Genotoxicity and Occurrence Assessment of Disinfection By-Product Mixtures in Drinking Water

EPA Grant Number: R825956
Title: Genotoxicity and Occurrence Assessment of Disinfection By-Product Mixtures in Drinking Water
Investigators: Minear, Roger A. , Plewa, Michael J.
Institution: University of Illinois at Urbana-Champaign
EPA Project Officer: Hahn, Intaek
Project Period: September 1, 1997 through August 31, 2000 (Extended to November 16, 2001)
Project Amount: $378,088
RFA: Drinking Water (1997) RFA Text |  Recipients Lists
Research Category: Water , Drinking Water


The main objective of this research project was to assess whether the single-cell gel electrophoresis (SCGE-COMET) assay using transgenic mammalian cells could predict human risks from drinking water disinfection byproducts (DBPs), rather than bacterial genotoxicity tests. We investigated if it is sufficiently responsive to evaluate DBPs produced from different disinfection processes and under different conditions. The objective was founded on the premise that mammalian cells may be more representative of direct cytotoxic and genotoxic affects from DBPs on humans than the "traditional" bacterial assays. The specific objectives of the research were to: (1) calibrate Salmonella typhimurium and transgenic mammalian cells genotoxicity assays using regulated DBPs; (2) compare the relative genotoxicities of chlorinated versus brominated DBPs; (3) compare the relative genotoxicities of chlorination byproducts (CBPs) versus brominated ozonation byproducts (OBPs); (4) compare the relative genotoxicities of DBPs derived from singular versus combination (sequential) use of ozonation and chlorination; and (5) provide a DBP occurrence database for extrapolating genotoxicity results to current disinfection practice.

Summary/Accomplishments (Outputs/Outcomes):

Sample Characterization and Methods Development. As a part of the project, a new analytical method was developed to enable differentiation between the chlorine and bromine components of total organic chlorine (TOX). In addition, a post-column reaction method was refined to aid in determining low concentrations of bromate in ozonated samples. These and accompanying refinements in standard U.S. Environmental Protection Agency (EPA) methods for halogenated compounds and a derivitization method for aldehydes were used to characterize the DBPs produced from selected waters and Suwannee River Fulvic Acid samples in terms of the known compounds TOX, TOCl, and TOBr. The percent of TOX attributable to known compounds was evaluated for chlorination, chloramination, bromination, and chlorine dioxide treatment. In all cases, the majority of TOX was attributed to unknown compounds. Chlorination yielded the greatest fraction of identifiable compounds, but also yielded the highest TOX when compared to the other three disinfection treatments.

S. typhimurium Cytotoxicity Assay. In addressing the research objectives for this research project, we found that the commonly used biological assays would require large amounts of DBPs. In addition, the sensitivity of the assays was not sufficient to allow a quantitative comparison of the cytotoxicity and genotoxicity of DBPs in bacterial and mammalian cell systems. We altered our approach and developed two novel, microplate methods to assess the cytotoxicity of DBPs in S. typhimurium and Chinese hamster ovary (CHO) cells.

Although the S. typhimurium plate incorporation test (the standard Ames test) is an excellent rapid qualitative mutagenicity assay, it cannot quantitatively determine the cytotoxicity of test agents. Therefore, we developed a rapid, microplate-based, semi-automatic cytotoxicity assay to quantitatively compare selected DBPs. For the DBPs analyzed in this research project, we have cytotoxicity data that cover more than 8 log-orders of concentration. We calculated the concentration that induced a reduction of S. typhimurium growth by 50 percent from a regression analysis of the concentration-response curves. This concentration was referred to as the percent C½ value and may be thought of as the concentration that causes 50 percent of the killing of the bacteria. The percent C½ value is useful in comparing the relative S. typhimurium cytotoxicity of DBPs. The rank order (from high to low) of the cytotoxicity of the DBPs and other chemicals was: 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX) >> bromoacetic acid (BA) > bromoform (BF) > dibromoacetic acid (DBA) > chloroacetic acid (CA) > tribromoacetic acid (TBA) > chloroform (CF) > ethylmethanesulfonate (EMS) > trichloroacetic acid (TCA) > ethanol > dichloroacetic acid (DCA) > potassium bromate (KBrO3 ) > dimethylsulfoxide (DMSO). The alkylating agent and potent mutagen EMS was used as a positive control, and DMSO was used as a solvent.

S. typhimurium Mutagenicity Assay. The data generated by the S. typhimurium cytotoxicity assay were used to establish a broad concentration range for each DBP. To compensate for the different treatment regimens used in the cytotoxicity assay and the preincubation mutagenicity assay, we calculated the concentration-hours (M-h) of exposure. The preincubation assay was used with S. typhimurium strains TA98, TA100, and RSJ100 with and without hepatic microsomal activation (S9). Cells were treated with each DBP or complex mixture, plated on selective medium (VB), and incubated for 72 hours at 37°C. Only reverse mutants (revertants) can grow into colonies that are counted. The concentration-response curves were determined and the mutagenic potency was determined (revertants/ mole) for each DBP. The rank order for the mutagenic potency for each DBP and Salmonella strain was as follows:

For strain TA98 -S9: MX > BA > DBA > CA > DCA, with TBA, BF, TCA, and CF not mutagenic.

For strain TA98 +S9: MX > BF > BA > DBA with TBA, CA, DCA, TCA, and CF not mutagenic.

For strain TA100 -S9: MX > BA > EMS > DBA > DCA > CA, with TBA, BF, TCA and CF not mutagenic.

For strain TA100 +S9, BA > BF > CA > DCA, with TBA, MX, TCA, and CF not mutagenic.

For strain RSJ100 -S9 was: BF > DCA, with BA, DBA, TBA, MX, CA, TCA, and CF not mutagenic.

For strain RSJ100 +S9 none of the DBP standards expressed a mutagenic response.

It appears that the rank order of the DBPs was the same for the cytotoxicity and mutagenicity; however, a quantitative analysis of the DBPs that expressed a mutagenic potency value showed that they were not significantly correlated when the percent C½ and mutagenic potency values were compared (r = 0.35; p <  0.49). Although a quantitative analysis of cytotoxicity is essential for a precise comparison of the mutagenicity of agents in S. typhimurium, cytotoxicity alone is not a suitable predictor for the mutagenic strength of a DBP.

Structure function analysis of the brominated versus the chlorinated analogs of the haloacetic acids showed that brominated acetic acids expressed a higher mutagenic potency. Based on the adjusted mutagenic potency values, BA was 150 times more mutagenic than CA. With TA100 S9, the mutagenic potency of the haloacetic acids was inversely related to the number of halogen substituent groups. BA was 36 times more mutagenic than DBA, as calculated with the adjusted mutagenic potency values. In summary, we demonstrated a rapid, quantitative, microplate method to determine the cytotoxicity of DBPs. The cytotoxicity data were integrated with the mutagenicity data derived from a preincubation test such that adjusted mutagenic potency values were calculated. The differential cytotoxicity expressed by the DBPs indicates that a cytotoxicity analysis enhanced the sensitivity of the mutagenicity data, which resulted in increased precision for comparing the relative mutagenic strengths of the DBPs.

Mammalian Cell Cytotoxicity Assay. We completed the development and implementation of a rapid, semi-automated cytotoxicity assay employing mammalian cells. The microplate mammalian cytotoxicity assay allowed for the measurement of a large number of replicate samples and provided for a high degree of reproducibility. The following DBPs were evaluated for their chronic cytotoxicity: BA, DBA, TBA, CA, DCA, TCA, bromonitromethane (BNM), dibromonitromethane (DBNM), tribromonitromethane (TBNM), MX, and KBrO3. We established a rank order of the cytotoxicity of DBPs in CHO cells and compared it to that of S. typhimurium. The rank order of DBPs in CHO cells, based on percent C½ values, was: BNM > DBNM > TBNM > BA >> MX > DBA > CA > KBrO3 > TBA > DCA > TCA. The rank order of DBPs in S. typhimurium was: MX >> BA >> DBA > CA > TBA > TCA > DCA >> KBrO3. The CHO and Salmonella data were not significantly correlated (r = 0.20; P < 0.48).

The reversal of the order ranking of BA and MX by an appreciable amount in the mammalian system verses the bacterial assay should be noted.

Mammalian Cell Genotoxicity Assay. We modified the SCGE assay to determine the genotoxic activity of specific DBPs. SCGE is an effective predictor of mutagenic and carcinogenic activity in animals, including humans. The SCGE assay detects genomic DNA damage in the form of DNA strand breakage, alkali-labile sites, and incomplete excision repair sites in single cells. The SCGE concentration-response curves of the DBPs were generated with differential genotoxicity expressed over four orders of magnitude. With the exception of DCA and TCA, the DBPs analyzed were more genotoxic than the positive control mutagen EMS. We established a rank order of the genotoxicity of the DBPs in CHO cells by determining the concentration within the positive trend in a concentration response curve that represented the 50 percent of the genetic damage as measured by tail moment values. The rank-ordered DBP concentrations as a function of DNA damage detected were: BA 18 µM, DBNM 26 µM, TBNM 60 µM, TCNM 100 µM, BNM 140 µM, MX 238 µM, CA 318 µM, DCNM 488 µM, DBA 1.76 mM, TBA 4.08 mM, EMS 5.8 mM, DCA 19.02 mM; TCA was not genotoxic in this assay. Cytotoxicity is a useful and valid biological endpoint. This study represents the largest comparative cytotoxicity analysis of drinking water DBPs in mammalian cells. There is a statistically significant correlation between DBP cytotoxicity and genotoxic potency in mammalian cells (r = 0.99; P < 0.0001). The appreciable difference between the genotoxicity of BA and MX, which are reversed in position when compared to bacterial assay also is significant.

Cytotoxicity of Organic Fractions Isolated from Disinfected Water Samples. Six organic fractions isolated from the disinfected water samples were analyzed for cytotoxicity using both S. typhimurium and CHO cells. The rank order of these samples for cytotoxicity in S. typhimurium was 1018 > 1017 >1014 > 1015 > 1016 > 1100, while the rank order for CHO cells was 1018 > 1017 > 1015 > 1016 > 1014 > 1100. In a qualitative sense, the cytotoxicity of the organic fraction samples was similar. A Pearson's Correlation analysis was conducted on the cytotoxic potency (based on the percent C½ values) in S. typhimurium and CHO cells; and they were significantly correlated (r = 0.97; P < 0.002). The mammalian cells were more sensitive to the toxicity of each sample as compared to their corresponding bacterial cytotoxicity. These data indicate that for complex mixtures of organic materials isolated from disinfected water samples, the bacteria and CHO cells were effective monitors for cytotoxic responses.

S. typhimurium Mutagenicity of Organic Fractions Isolated From Disinfected Water Samples. The rank order of the mutagenic potency of the isolated organic samples was 1018 > 1019 1020 > 1102 > 1100, and the rank order based on the cytotoxicity of the samples was 1018 > 1019 > 1020 > 1100 > 1102. By brief observation, the rank order is similar. However, a Pearson Correlation analysis indicated that a significant quantitative relationship between the rank order of the mutagenic potency of these samples and the percent C½ values did not exist (r = 0.65; P < 0.24). These data indicate that using the quantitative biological methods that we employed with the standard DBP, we were able to enhance the sensitivity of the S. typhimurium mutagenicity assay.


From the research results, a number of conclusions can be drawn:

1. Quantitative rapid mutagenicity assays incorporating cytotoxicity measurements of DBPs enhanced the sensitivity of routine "Ames" test methods.

2. Mutagenic potency values can be modified with appropriate cytotoxicity factors to better compare the mutagenicity of DBPs in S. typhimurium.

3. A quantitative mammalian cell cytotoxicity assay can determine the relative ranking of DBPs.

4. A quantitative, microplate SCGE method can analyze DBPs for their relative genotoxic potency in mammalian cells.

5. From observation and by rank order, it appears that a qualitative relationship exists between Salmonella cytotoxicity and mutagenicity of the DBPs. However, comparing the quantitative data (percent C½ values and mutagenic potency values using a Pearson's Product Moment Correlation Test) there was no significant correlation between Salmonella cytotoxicity and mutagenicity of DBPs (r = 0.35; P 0.49).

6. From observation and by rank order, a qualitative relationship between CHO cell cytotoxicity and genotoxicity of the DBPs can be found. Comparing the quantitative data (percent C½ values and SCGE potency values), there was a highly significant correlation between CHO cell cytotoxicity and genotoxicity of DBPs (r = 0.99; P < 0.001).

7. DBP mutagenicity in Salmonella and genotoxicity in CHO cells were qualitatively related; however, a quantitative comparison of Salmonella mutagenic potency and CHO SCGE potency values of DBPs demonstrated that they were not significantly related (r = -0.28; P < 0.58).

8. Isolated organic fractions of the waters with defined natural organic matter (NOM) and treated with different disinfectants were evaluated for their cytotoxicity in S. typhimurium and CHO cells and for their mutagenicity in S. typhimurium. The cytotoxic response in S. typhimurium indicated a range of toxicity that was comparable for MX > BA < CA.

9. The mutagenicity of the organic fractions was comparable to concentration-response curves between BA and CA.

10. The percent C½ CHO cell cytotoxicity values of the organic fractions ranged from 5.4 to 177 ug C/mL.

11. The CHO cell genotoxic potency of a limited number of extracts was moderately correlated with that of CHO cell cytotoxicity (r = 0.76).

12. No relationship was found between TOX of the organic extracts and the CHO cytotoxicity, or between TOX and CHO mutagenic potency.

13. MX was the most cytotoxic and mutagenic DBP tested in S. typhimurium; however, it was not the most potent genotoxic agent in mammalian cells. Monobromoacetic acid was more than an order of magnitude more genotoxic than MX.

Recommendations. The following recommendations have been formulated:

1. In this research, the importance of integrating analytical chemistry and analytical biology in the study of the relative toxicity and risks of drinking water DBPs has been demonstrated. This suggests that all future studies should contain extensive chemical data, in conjunction with toxicity evaluations.

2. Mammalian cell assays for cytotoxicity and genotoxicity that were developed or modified under the auspices of this research project are well suited for the analysis of DBPs, and should be employed to determine their relative biological risks.

3. Quantitative measurement of cytotoxicity is an important biological endpoint. The relative cytotoxicity of DBPs can be determined by the use of rapid, microplate-based assays that were developed in this project for S. typhimurium and mammalian cells. Cytotoxicity also is important to quantitatively determine, because it can influence the outcome of genotoxicity assays. Without accounting for the cytotoxicity of the DBP or TOX samples, the sensitivity of a specific genotoxicity assay could be compromised.

4. The fact that the data from a highly used S. typhimurium mutagenicity test and the CHO SCGE genomic DNA damage test did not correspond, indicates that short-term bacterial assays may not be the most suitable biological indicators for human risk of DBPs. It is recommended that S. typhimurium cytotoxicity and mutagenicity tests be used as one of a series of biological assays to evaluate DBPs or the complex organic products of the water disinfection processes. Mammalian cell tests that include a quantitative measurement of long-term cytotoxicity and the induction of genomic DNA damage are highly recommended for the analysis of DBPs. It also is recommended that the toxicity of organic agents isolated from disinfected water is monitored.

5. Data presented in this research project indicate that between 50 percent and 98 percent of the DBPs generated by common disinfection processes are not chemically defined. Thus, the majority of DBP organic materials found in TOX are unknown. These agents must be evaluated for their cytotoxicity and genotoxic activity before and after metabolism in mammalian cell systems. The information presented here clearly indicates the class of DBPs to which humans are exposed are those that are categorized as "unknown for TOX." These agents and their metabolic products should have a high priority for future research on their toxic characteristics.

6. The EPA has prepared a list of high-priority DBPs, based on their structure, occurrence, and inferred toxicity, that may pose a public health threat. Most of these DBPs have not been properly evaluated for their biological impacts. The combination of quantitative cytotoxicity and genotoxicity assays developed in this project would be helpful in evaluating these priority DBPs for their relative risks.

7. Finally, the genotoxicity research on DBPs is focused primarily on the ability of these agents to induce DNA damage. This only provides one facet of genotoxic risk. Future research that would quantitatively determine the impact of DBPs on DNA repair mechanisms is needed before we have a more complete understanding of the relative human health risks posed by these ubiquitous agents in our drinking water.

Journal Articles on this Report : 5 Displayed | Download in RIS Format

Other project views: All 25 publications 9 publications in selected types All 5 journal articles
Type Citation Project Document Sources
Journal Article Kargalioglu Y, McMillan BJ, Minear RA, Plewa MJ. Analysis of the cytotoxicity and mutagenicity of drinking water disinfection by-products in Salmonella typhimurium. Teratogenesis, Carcinogenesis, and Mutagenesis 2002;22(2):113-128.
abstract available  
R825956 (Final)
  • Abstract from PubMed
  • Journal Article Plewa MJ, Kargalioglu Y, Vankerk D, Minear RA, Wagner ED. Development of quantitative comparative cytotoxicity and genotoxicity assays for environmental hazardous chemicals. Water Science & Technology 2000;42(7-8):109-116.
    abstract available   full text available
    R825956 (Final)
  • Abstract: Water Science & Technology
  • Other: Water Science & Technology PDF
  • Journal Article Plewa MJ, Kargalioglu Y, Vankerk D, Minear RA, Wagner ED. Mammalian cell cytotoxicity and genotoxicity analysis of drinking water disinfection by-products. Environmental and Molecular Mutagenesis 2002;40(2):134-142.
    abstract available  
    R825956 (Final)
  • Abstract from PubMed
  • Journal Article Zhang X, Minear RA. Decomposition of trihaloacetic acids and formation of the corresponding trihalomethanes in drinking water. Water Research 2002;36(14):3665-3673.
    abstract available   full text available
    R825956 (Final)
    R826834 (Final)
  • Abstract from PubMed
  • Full-text: Science Direct Full Text
  • Other: Science Direct PDF
  • Journal Article Zhang X, Echigo S, Lei H, Smith ME, Minear RA, Talley JW. Effects of temperature and chemical addition on the formation of bromoorganic DBPs during ozonation. Water Research 2005;39(2-3):423-435.
    abstract available   full text available
    R825956 (Final)
  • Abstract from PubMed
  • Full-text: Science Direct HTML
  • Other: Science Direct PDF
  • Supplemental Keywords:

    Chinese hamster ovary cells, CHO, mammalian cells, single-cell gel electrophoresis, single-cell gel electrophoresis, SCGE, water disinfection byproducts., RFA, Health, Scientific Discipline, Waste, Water, Environmental Chemistry, Genetics, Health Risk Assessment, Environmental Microbiology, Risk Assessments, chemical mixtures, Drinking Water, monitoring, occurrence monitoring, Safe Drinking Water, single cell electrophoesis assay, microbial risk assessment, human health effects, chlorinated by products, exposure and effects, disinfection by-product mixtures, disinfection byproducts (DPBs), exposure, community water system, natural organic matter, genotoxicity, human exposure, treatment, bacterial genotoxicity, dietary ingestion exposures, drinking water contaminants, water treatment, drinking water treatment, cytotoxic effects, drinking water system, ozonation, DNA microarrays

    Progress and Final Reports:

    Original Abstract
  • 1998
  • 1999 Progress Report
  • 2000 Progress Report
  • 2001