* U. S. Environmental Protection Agency, Office of Research and Development, National Health and Environmental Effects Research Laboratory, Environmental Carcinogenesis Division, Research Triangle Park, North Carolina 27711; and
Department of Biology, North Carolina Central University, Durham, North Carolina 27707
Received September 27, 2000; accepted December 19, 2000
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: disinfection by-products; mutagenicity; Salmonella microsuspension assay; biotransformation; intestinal flora; enzymes..
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The enzyme activities of the normal mammalian intestinal microbial population can transform many promutagens and procarcinogens to their mutagenic and carcinogenic forms (Chadwick et al., 1992). For example, glycosides are compounds consisting of a non-sugar moiety (aglycone) bound to a sugar by an
- or ß-glycosidic linkage. Glycosides can enter the gut from two major sources, the diet or the liver (compounds detoxified by glucuronide formation in the liver are secreted into the intestine via the bile). The intestinal flora then can hydrolyze the ß-glucuronide bond, releasing the aglycones, some of which are toxic or carcinogenic (Goldin, 1986
). Repeated exposure to a compound via the enterohepatic circulation can amplify its biological activity. The enzymes of the intestinal flora (specifically, ß-glucosidase) hydrolyze cycasin, a nongenotoxic plant glycoside, to a genotoxic aglycone, methylazomethanol (MAM; Brown and Dietrich, 1979). Cycasin causes tumors in rats with normal intestinal flora, but not in germ-free rats (Laquer et al., 1967
). The principal glycosidases produced by the intestinal flora are ß-glucosidase, ß-galactosidase, and ß-glucuronidase. E. coli and Clostridium are associated with ß-glucuronidase activity. However, E. coli has low ß-glucosidase activity, with higher activities of this enzyme found with Bacteroides and Enterococcus faecalis (Hawksworth et al., 1971
).
Many azo dyes are used in the food and textile industries. The bacterial flora can reductively hydrolyze the azo bond by the action of azoreductase, which results in the formation of substituted aromatic amines, a number of which are well established carcinogens. The reduction of nitro groups by microbial nitroreductase in the intestine can be another source of aromatic amines (Goldin, 1986). Both nitroreductase and ß-glucuronidase activity are necessary for 2,6-dinitrotoluene genotoxicity (George et al., 1994
). Microbial dechlorinase activity may be involved in the metabolism of trichloroacetic and dichloroacetic acids in the gut (Moghaddam et al., 1996
). The DBPs dichloro-, dibromo-, or bromochloroacetic acid, administered in the drinking water of Fischer 344 rats, impacted the intestinal metabolism and microflora populations, indicating the potential for altered bioactivation of co-administered promutagens and procarcinogens (George et al., 2000
).
The objective of the current study is to determine whether in vitro cultures of the rat intestinal microbiota can metabolize monochloro-, dichloro-, trichloro-, monobromo-, dibromo-, tribromo-, and bromochloroacetic acids to intermediates that are more or less mutagenic than the parent compound. In addition, the effects of these DBPs on the intestinal microbial populations and their metabolism are studied, with emphasis on the enzymes often involved in the bioactivation of promutagens and procarcinogens.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Growth curves.
Forty-eight-h growth curves were performed in vitro with rat cecal homogenate and with each of 4 strains isolated from the rat cecum. Adult male CDF rats from Charles River Laboratory (Raleigh, NC) were provided food (Purina 5001, Purina Mills, St. Louis, MO) and water ad libitum. Rats were asphyxiated with CO2 and taken into an anaerobic chamber (Coy Laboratory Products, Inc., Grass Lake, MI) where the cecum was removed and placed into 20 ml of prereduced VPI buffer (2.0 g gelatin, 0.5 g cysteine, 500 ml deionized water, and 500 ml aqueous solution containing 0.1 g anhydrous CaCl2, 0.1 g anhydrous MgSO4, 0.5 g K2HPO4, 5.0 g NaHCO3, and 1.0 g NaCl) (Holdeman et al., 1977), and minced with scissors. Forty µl of this cecal homogenate was added to 10 ml of PYG media along with 0.1 ml of a 100 mg/ml aqueous solution of DBP (filter-sterilized) for a final concentration of 1 mg/ml. Tubes (2 per treatment) were vortexed and incubated anaerobically at 37°C. At 0, 3, 6, 12, 24, and 48 h of incubation, a 0.1 ml aliquot of culture was removed, diluted in prereduced phosphate-buffered saline, and plated onto selective media for enumeration of flora populations. Enumerated populations were: lactose-fermenting enteric bacilli on MacConkey agar (MAC), enterococci on KF streptococcus agar (KF), total anaerobes and facultatives on Brucella laked-blood agar (BA), anaerobic gram-negative bacilli on Brucella laked blood agar + vancomycin and kanamycin (VK), and lactobacilli on Rogosa agar (ROG). MAC and KF plates were incubated aerobically for 24 and 48 h, respectively. BA, VK, and ROG plates were incubated anaerobically for 72 h.
The 4 cecal isolates, chosen for growth curves because of their abundance, were identified as Bacteroides distasonis, Bacteroides uniformis, Clostridium bifermentens, and Lactobacillus johnsonii, using the methods and software of the Microbial Identification System (MIDI, Inc., Newark, DE). Overnight cultures (25 ml) of each strain were grown in PYG broth. Cultures were centrifuged, washed in 5 ml of reduced salts solution, then resuspended in 1 ml of reduced salts solution. To each 10 ml PYG tube was added 0.1 ml of a 100 mg/ml aqueous solution of DBP and 0.1 ml of bacterial suspension. Tubes (2 per treatment) were incubated and samples removed for dilution and plating (on BA only) as described previously. Growth curves for the isolates also were performed in salts solution (5 mg CaCl2 -3881µ2H2O, 4 mg anhydrous MgSO4, 20 mg K2HPO4, 20 mg KH2PO4, 200 mg NaHCO3, and 1.0 g NaCl in 1 liter deionized, distilled H2O, [Microbial Identification System operating manual, p. C-4, MIDI, Inc., Newark, DE]) instead of PYG, to determine if the isolates could maintain growth by utilizing the DBP as a carbon source.
Enzyme assays.
PYG tubes were inoculated in triplicate with a DBP (1 mg/ml) and with cecal homogenate or a cecal isolate, as described above for the growth curves. Because of its toxicity, MBA also was tested at a second concentration of 0.1 mg/ml. After 15 h of incubation, cultures were centrifuged, the pellet washed in 5 ml prereduced VPI buffer, and resuspended in 10 ml VPI buffer. A 1 ml aliquot of each sample was transferred into 9 ml of 0.85% saline for dilution and plating (enumeration on BA only), and for APIZYM (bioMerieux Vitek, Inc., Hazelwood, MO) enzyme analysis following the manufacturer's instructions. The experiment was repeated twice for each DBP.
For determination of GLR, GAL, and GLU activities, 2 ml aliquots of each sample (pellet suspended in VPI buffer, see above paragraph) were transferred to 3 separate tubes, each containing 8 ml of VPI buffer. A rubber septum cap was placed on each tube, and tubes were removed from the chamber and placed on ice. A 37.5 µl aliquot of DMSO (culture controls and reagent controls) or the substrate mix containing 170 mg/ml of the nitroreductase competitor 3,4-dichloronitrobenzene and 130 mg/ml of the specific substrate GLR, GAL, or GLU was injected into each tube and incubated at 37°C with shaking for 1 h (Chadwick et al., 1995). Placing the tubes on ice terminated the reactions. Particulate matter was sedimented by centrifugation at 4°C. Tubes were again placed on ice and the OD was recorded at 405 nm using a Spectronic 20+ (Spectronic Instruments, Inc., Rochester, NY). The concentration of released p-nitrophenol was calculated from a p-nitrophenol standard curve.
For determination of nitroreductase, azoreductase, dechlorinase, and dehydrochlorinase activities, septum caps were placed on the tubes containing the remaining 3 ml of sample. The tubes were removed from the chamber and placed on ice. Twenty-five µl of DMSO only (culture controls and reagent controls) or 25 µl of substrate mix containing 25 mg/ml of 3,4-dichloronitrobenzene, 12.5 mg/ml of p,p'-DDT, and 23 mg/ml of methyl orange was injected through the septum cap of each tube. Tubes were incubated for 1 h at 37°C, then placed on ice to terminate the reaction. Samples were extracted, derivatized, and analyzed by gas chromatography according to the method of Chadwick et al. (1993).
Salmonella microsuspension bioassay.
A microsuspension modification of the Ames Salmonella reversion assay (DeMarini et al., 1989) was used to compare the mutagenicity of the DBP compounds/metabolites after incubation for various lengths of time, with or without the cecal microbiota. The microsuspension assay was chosen because of its greater sensitivity with a smaller mass of sample than that required by the standard plate incorporation assay. Growth curves in PYG were performed as previously described, with unopened sample tubes (2 per treatment) frozen (20°C) at 0, 3, 6, 12, and 24 h of incubation. The concentration of haloacetic acid in the PYG cultures was 1 mg/ml, with the exception of MBA, which was at 0.1 mg/ml. Samples were extracted with methyl tert-butyl ether (MTBE), blown to dryness, and dissolved in 100 µl DMSO, then frozen at 20°C until bioassay.
Based on the results of previous mutagenicity testing of these compounds in Salmonella (data not shown), tester strain TA100 was used for bioassay of the brominated compounds and strain TA104 was used for the chlorinated compounds. BCA was tested with strain TA100. The tester strains were grown overnight in Oxoid Nutrient Broth No. 2 and harvested by centrifugation. The pellet was resuspended in 0.015 M phosphate buffer (pH 7.4) at either a 5x (TA100) or a 1x (TA104, spontaneous counts were too high with a 5x culture) concentration of the overnight culture. To each bioassay tube (13 x 100 mm, glass disposable), we added 5 µl of the sample or sample dilution (1:2.5 dilution), 100 µl of 0.015 M phosphate buffer (without S9) or 100 µl S9 mix (3% v/v S9 concentration), and 100 µl of cells. Each sample was run in duplicate. Tube racks were covered with foil and incubated at 37°C without shaking for 90 min, then placed on ice. Molten top agar (2.5 ml) was added to each tube, which was vortexed and poured onto a VBME plate. Plates were incubated for 72 h at 37°C and counted on an Artex 880 colony counter. DMSO was used as the negative control. Sodium azide (1 µg/plate) was the positive control (S9) for strain TA100 and methyl glyoxal (25 µg/plate) for strain TA104. 2-Aminoanthracene (0.25 µg/plate) was the positive control (+S9) for TA100; 2-aminoanthracene (2.5 µg/plate) was the positive control for TA104. Assuming 100% recovery of the DBP after extraction, the dose of DBP would be 500 µg/plate, or 200 µg/plate for diluted samples (50 and 20 µg/plate for MBA). Actual recoveries (averages) are reported in Table 1.
|
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Quantitative Enzyme Assays
In cultures of rat cecal flora, the highest activity of the 7 enzymes measured (Fig. 1) was for ß-galactosidase, with the exception of the BCA treatment group where ß-glucosidase levels were very high (significantly higher than in the control group). In addition, dehydrochlorinase activity was significantly higher in the BCA treatment group than in the control group (Fig. 1
, Table 3
).
|
|
Semiquantitative Enzyme Assays
Semiquantitative analysis of 19 enzymes was performed using APIZYM assays (Fig. 2). Table 3
reports only those API results where statistical significance is shown. The following 11 enzymes were detected in control cultures of rat cecal flora: alkaline phosphatase, esterase (C4), esterase lipase (C8), leucine arylamidase, acid phosphatase, naphthol-AS-BI-phosphohydrolase,
-galactosidase, ß-galactosidase,
-glucosidase, ß-glucosidase, and N-acetyl-ß-glucosaminidase. Interestingly, ß-glucuronidase was not detected in control cultures but was detected in MCA-, MBA-, and DBA-treated cultures. These cultures also had the highest levels of ß-glucuronidase in the quantitative assays, but significance was not demonstrated. In most cases the API results correlated well with the quantitative enzyme results. However, because of the semiquantitative nature of the assay and the small sample sizes, statistical significance is difficult to demonstrate. Quantitative assays showed an elevated level of ß-glucosidase for BCA-treated cultures, and this effect was also seen using APIZYM, although it was not statistically significant. ß-Galactosidase activity was significantly reduced in cultures treated with TBA or MBA. Quantitative assays also showed a reduction in ß-galactosidase activity for these 2 treatment groups, but statistical significance was not shown. For the enzyme leucine arylamidase, activity was significantly lower in MCA-, MBA-, and DBA-treated cultures than in the control. Acid phosphatase activity was significantly lower in MCA- and MBA-treated cultures. Alkaline phosphatase activity was significantly elevated in BCA-treated cultures.
|
Salmonella Microsuspension Bioassays
No mutagenicity was detected for MCA, DCA, or TCA with S. typhimurium strain TA104, therefore these data are not shown. Mutagenicity ( 2-fold increase over spontaneous) was observed in TA100 for samples containing DBA or BCA with and without S9. TBA was mutagenic only in the presence of S9. Positive control values for TA100 (average of 4 experiments) were sodium azide (S9, 449 revertants/plate) and 2-aminoanthracene (+S9, 909 revertants/plate).
Undiluted samples containing cecal homogenate often were toxic to Salmonella when S9 was not present, especially at the later time points. Because toxicity was present without S9 for undiluted samples containing DBA and cecal homogenate, only the results for the diluted samples were plotted (Fig. 3a). Samples containing DBA only were mutagenic over the entire time course. Mutagenicity of the samples containing cecal homogenate only and samples containing DBA+ cecal homogenate (cec) dropped sharply by 6 h of incubation. With S9 added, the undiluted samples were more mutagenic than the diluted samples (Fig. 3b
). Samples containing DBA only were significantly more mutagenic than those containing cecal homogenate only, and mutagenicity for samples containing both the chemical and cecal homogenate were intermediate between the two, but not significantly different from either. This was the only sample set for which samples containing cecal homogenate alone were mutagenic.
|
The undiluted samples were more mutagenic than their dilutions for samples containing BCA (Figs. 3d and 3e). This was the only sample set with strain TA100 for which undiluted samples containing cecal homogenate were not toxic at 12 and 24 h. Without S9, undiluted samples containing BCA only or BCA + cec were mutagenic and significantly different from samples containing cecal homogenate only, but not from each other. The diluted samples were not mutagenic. With S9, both diluted and undiluted samples containing BCA only or BCA + cec were mutagenic and significantly different from the samples containing cecal homogenate only. BCA and BCA + cec samples differed significantly from each other, only for undiluted samples, at the 3 h time point, when the revertants/plate were significantly higher in samples containing both BCA and cecal homogenate. This difference was not present in the diluted samples.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Interestingly, the only chemical that was not toxic to the rat cecal microbiota (BCA) was also the only chemical to significantly alter the enzyme activity of the microbiota in vitro, resulting in increased ß-glucosidase and dehydrochlorinase activities. Induction of dehydrochlorinase activity could be due to the availability of the haloacetic acid for metabolism. Enterococcus faecalis and Bacteroides spp. have been associated with ß-glucosidase activity (Hawksworth et al., 1971). While the enterococci population in vitro was reduced by treatment with most of the haloacetic acids, enterococci numbers were not affected by BCA treatment, and ß-glucosidase activity increased. Evidence for the contribution of Bacteroides to the increase in ß-glucosidase activity with BCA treatment was seen in pure cultures of B. uniformis. In vivo, ß-glucosidase activity was increased in the large intestine, but not in the cecum, by DCA, DBA, and BCA (George et al., 2000
). Elevated enzyme activities indicate increased potential for bioactivation of promutagens and procarcinogens.
Azoreductase, nitroreductase, and dechlorinase activities were significantly reduced by one or more treatments in vivo, but not in vitro. Table 3 summarizes in vitro and previous in vivo (George et al., 2000
) week 5 results, indicating statistically significant changes in enzyme activity. DBP effects on enzyme activity in vivo often varied over the course of 5 weeks, making in vitro/in vivo comparisons difficult.
In cultures of B. distasonis and B. uniformis, but not the rat cecal cultures, ß-galactosidase activity was reduced for all treatment groups. In vivo, DCA, DBA, and BCA resulted in reduced cecal ß-galactosidase activity at weeks 1 and 3. However, by week 5, ß-galactosidase activity had returned to control levels in the DCA and DBA treatment groups, and had increased above control levels in the BCA treatment group.
APIZYM results correlated well with the quantitative results for those enzymes analyzed by both methods. Trends were similar, even though statistical significance was not always demonstrated. The haloacetic acids usually lowered the activities of additional enzymes assayed by the APIZYM methods. These enzymes are not known to be involved in the biotransformation of xenobiotics. In general, reduced enzyme activities may indicate a depressed cell metabolism, which in this case could be attributed to chemical toxicity. The lone exception again occurred with BCA-treated cultures of rat cecal flora where alkaline phosphatase activity was significantly elevated. BCA was not toxic to the rat cecal microbiota.
DBA, TBA, and BCA were mutagenic in TA100. After 3 h of incubation with BCA, when the culture of rat cecal flora was in exponential growth and should be most enzymatically active, a metabolite may have formed which was more mutagenic than BCA itself. However, overall, the co-incubation of the cecal microbiota with haloacetic acid either did not affect the mutagenicity of the chemical or reduced it.
In one sample set (with cecal microbiota derived from one animal) the cecal homogenate itself was mutagenic. Fecal mutagenicity has been well documented and is reviewed by Goldin (1986). The majority of fecal mutagens are fecapentaenes. A precursor, which is either a product of other bacteria in the colon, a result of diet, or a metabolite derived from the host, is necessary for fecapentaene production, in combination with Bacteroides. Fecal mutagenicity appears to be dependent on diet; humans on a typical "western" diet are at higher risk for colon cancer and have higher fecal mutagenicity than those on a vegetarian diet. However, all the rats in this study were on the same diet and housed identically. Therefore, individual genetic variability, or a differing flora composition due to a previous exposure are the most likely explanations for the occurrence of mutagenic cecal contents in only one animal.
Undiluted samples containing cecal homogenate often were toxic to Salmonella in the microsuspension assay. This was especially notable without S9 for the 12 and 24 h time points. By 12 h the cecal cultures were in stationary phase, during which growth is limited by the depletion of nutrients and the accumulation of toxic waste products in the medium. The addition of mammalian liver enzymes in the S9 mix resulted in detoxification of these samples.
In summary, high levels of haloacetic acid in the culture medium, with the exception of BCA, are toxic to the rat cecal microbiota, especially the enterococci. This effect also was seen in vivo in the intestine, and could increase the animals' susceptibility to disease. The mutagenicity of the haloacetic acids is not a result of biotransformation by the intestinal flora, but factors could be involved in the animal that cannot be mimicked in vitro (enterohepatic circulation, for example). BCA did elevate the activity of several enzymes involved in the biotransformation of xenobiotics, an effect that could potentially increase the biological activity of co-administered compounds. Therefore, future work should investigate this potential with compounds routinely found in drinking water.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
1 To whom correspondence should be addressed at U.S. EPA, 86 TW Alexander Drive, MD 68, Research Triangle Park, NC 27711. Fax: (919) 541-3966. E-mail: nelson.gail{at}epa.gov.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Bull, R. J., Sanchez, I. M., Nelson, M. A., Larson, J. L., and Lansing, A. J. (1990). Liver tumor induction in B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology 63, 341359.[ISI][Medline]
Chadwick, R. W., Allison, J. C., Talley, D. L., and George, S. E. (1995). Possible errors in assay for ß-glycosidase activity. Appl. Environ. Microbiol. 61, 820822.[Abstract]
Chadwick, R. W., George, S. E., and Claxton, L. D. (1992). Role of the gastrointestinal mucosa and microflora in the bioactivation of dietary and environmental mutagens or carcinogens. Drug Metab. Rev. 24, 425492.[ISI][Medline]
Chadwick, R. W., George, S. E., Kohan, M. J., Williams, R. W., Allison, J. C., Hayes, Y. O., and Chang, J. (1993). Potentiation of 2,6-dinitrotoluene genotoxicity in Fischer 344 rats by pretreatment with Aroclor 1254. Toxicology 80, 153171.[ISI][Medline]
DeAngelo, A. B., Daniel, F. B., Most, B. M., and Olson, G. R. (1996). The carcinogenicity of dichloroacetic acid in the male Fischer 344 rat. Toxicology 114, 207221.[ISI][Medline]
DeAngelo, A. B., Daniel, F. B., Stober, J. A., and Olson, G. R. (1991). The carcinogenicity of dichloroacetic acid in the male B6C3F1 mouse. Fundam. Appl. Toxicol. 16, 337347.[ISI][Medline]
DeMarini, D., Dallas, M. M., and Lewtas, J. (1989). Cytotoxicity and effect on mutagenicity of buffers in a microsuspension assay. Teratog. Carcinog. Mutagen. 9, 287295.[ISI][Medline]
DeMarini, D. M., Perry, E., and Shelton, M. L. (1994). Dichloroacetic acid and related compounds: Induction of prophage in E.coli and mutagenicity and mutation spectra in Salmonella TA100. Mutagenesis 9, 429437.[Abstract]
Fuscoe, J. C., Afshari, A. J., George, M. H., DeAngelo, A. B., Tice, R. R., Salman, T., and Allen, J. W. (1996). In vivo genotoxicity of dichloroacetic acid: Evaluation with the mouse peripheral blood micronucleus assay and the single cell gel assay. Environ. Mol. Mutagen. 7, 19.
George, S. E., Chadwick, R. W., Kohan, M. J., Allison, J. C., Williams, R. W., and Chang, J. (1994). Role of the intestinal microbiota in the activation of the promutagen 2,6-dinitrotoluene to mutagenic urine metabolites and comparison of GI enzyme activities in male Fischer 344 rats. Cancer Lett. 79, 181187.[ISI][Medline]
George, S. E., Nelson, G. M., Swank, A. E., Brooks, L. R., Bailey, K., George, M., and DeAngelo, A. (2000). The disinfection by-products dichloro-, dibromo-, and bromochloroacetic acid impact intestinal microflora and metabolism in Fischer 344 rats upon exposure in drinking water. Toxicol. Sci. 56, 282289.
Giller, S., Le Curieux, F., Erb, F., and Marzin, D. (1997). Comparative genotoxicity of halogenated acetic acids found in drinking water. Mutagenesis 12, 321328.[Abstract]
Goldin, B. R. (1986). In situ bacterial metabolism and colon mutagens. Annu. Rev. Microbiol. 40, 367393.[ISI][Medline]
Harrington-Brock, K., Smith, T. W., Parker, L. M., and Moore, M. M. (1992). Genotoxic effects of by-products of the chlorination of drinking water. Environ. Mol. Mutagen. 19(Suppl. 20), 24.
Hawksworth, G., Drasar, B. S., and Hill, M. J. (1971). Intestinal bacteria and the hydrolysis of glycosidic bonds. J. Med. Microbiol. 4, 451459.[ISI][Medline]
Holdeman, L. V., Cato, E. P., and Moore, W. E. (1977). Anaerobic Laboratory Manual, 4th ed. VPI Anaerobe Laboratory, Blacksburg, VA.
Koivusalo, M., Jaakkola, J. J., Vartiainen, T., Hakulinen, T., Karjalainen, S., Pukkala, E., and Tuomisto, J. (1994). Drinking water mutagenicity and gastrointestinal and urinary tract cancers: An ecological study in Finland. Am. J. Public Health 84, 12231228.[Abstract]
Laquer, G. L., McDanie, E. G., and Matsumoto H. (1967). Tumor induction in germ-free rats with methylazoxymethanol (MAM) and synthetic MAM acetate. J. Natl. Cancer Inst. 39, 355371.[ISI]
Moghaddam, A. P., Abbas, R., Fisher, J. W., Stavrou, S., and Lipscomb, J. C. (1996). Formation of dichloroacetic acid by rat and mouse gut microflora, an in vitro study. Biochem. Biophys. Res. Commun. 228, 639645.[ISI][Medline]
Morris, R. D., Audet, A.-M., Angelillo, I. F., Chalmers, T. C., and Mosteller, F. (1992). Chlorination, chlorination by-products, and cancer: A meta-analysis. Am. J. Public Health 82, 955963.[Abstract]
Parrish, J. M., Austin, E. W., Stevens, D. K., Kinder, D. H., and Bull, R. J.(1996). Haloacetate-induced oxidative damage to DNA in the liver of male B6C3F1 mice. Toxicology 110, 103111.[ISI][Medline]
U.S. EPA (1998). Drinking water priority rulemaking: Microbial and disinfection byproduct rules. EPA 815-F-980014 Environmental Protection Agency, Office of Water.
Van der Waaij, D. (1989). The ecology of the human intestine and its consequences for overgrowth by pathogens such as Clostridium difficile. Annu. Rev. Microbiol. 43, 6987.[ISI][Medline]
Van der Waaij, D., Berghuis-de Vries, J. M., and Lekkerkerk-van der Wees, J. E. (1971). Colonization resistance of the digestive tract in conventional and antibiotic-treated mice. J. Hyg. (Camb.) 69, 405411.