Organic mutagens and drinking water in The Netherlands : a study on mutagenicity of organic constituents in drinking water in The Netherlands and their possible carcinogenic effects

English. Several mutagenic and carcinogenic organic compounds have been detected in Dutch surface waters and in drinking water prepared from these surface waters. Although the levels of these compounds in drinking- and surface water are relatively low, in general below μg per litre, it appeared that organic concentrates tested in the Ames/microsome assay, showed mutagenic activity in 50 ml surface- and 500 ml drinking water.<p/>Such a result however was not expected based on the concentration of organic mutagens identified in these waters. Therefore the conclusion had been drawn that a number of unknown organic mutagens in combination with the identified mutagens were responsable for the level of mutagenic activity. With this in mind, an extensive investigation was carried out in an attempt to answer the following questions :<br/>- do drinking waters prepared from surface water, groundwater or a mixture of both show mutagenic activity;<br/>- do different water treatment processes during the preparation of drinking water influence the mutagenic activity;<br/>- what kind of physical-chemical properties have the organic compounds in drinking water concentrates showing mutagenic activity in the Ames test;<br/>- do mutagenic organic concentrates prepared from drinking water show carcinogenic properties?<p/>Against this background an inventory study was made on the presence of mutagenic activity in drinking water of 18 cities in The Netherlands.<p/>Besides this study, the influence of different water treatment processes on the mutagenic activity was examined in a number of water works. Furthermore, an attempt was made to characterize the organic compounds which are responsable for the mutagenic activity. Finally, a carcinogenicity study was carried out to see whether mutagenic drinking water concentrates induce carcinogenic effects in rats.<p/>Chapter 1, describes the aim of this investigation and the factors which have led to the present concern regarding the possible toxic effects of organic (micro)pollutants in drinking water in The Netherlands.<p/>Furthermore a summary is given of the literature on the identification of individual compounds and some data are given of the mutagenic and (suspect) carcinogenic organic compounds which are detected in surface- and drinking water in The Netherlands.<p/>In Chapter 2, materials and methods are described which were used for the inventory study on the presence of mutagenic activity in drinking water of 18 cities (Chapter 3), the characterization of organic mutagens (Chapter 4) and for a carcinogenicity study with mutagenic drinking water concentrates (Chapter 5).<p/>This chapter describes the XAD resins which have been used for concentrating organic mutagens from water, the concentration procedures with the aid of XAD resins and freeze drying, a mutagenicity assay viz. the Ames Salmonella/microsome assay (Ames test). The methods for chemical analysis which are used in the 18 city survey, fractionating techniques like thinlayer chromatography (TLC), high performance liquid chromatography (HPLC) and gelfiltration are described. The treatment, dose levels and experimental design and conduct which are used in a carcinogenicity study are explained. Finally, statistical procedures which have been used for analyzing tumour incidence in the carcinogenicity experiment and relating chemical parameters in drinking water to mutagenic activity in organic concentrates prepared from these drinking waters are described.<p/>In Chapter 3, the need for concentrating organic mutagens is explained. Mutagenicity results (Ames test) of five surface waters in The Netherlands are presented, in which the organic mutagens are concentrated with the XAD procedures. Depending on the concentration factor applied (103-4.103) it appeared that in all five surface waters mutagenic activity could be detected. The river Rhine showed the highest mutagenic activity viz. a doubling of revertants in 50 ml water.<p/>To be certain that the XAD procedure is a selective concentration method for organic mutagens, this procedure was compared with a freeze drying technique. The results of this comparative investigation showed that the XAD procedure is a reliable method for concentrating organic mutagens from surface water. on the basis of the surface water results, the XAD procedure was also applied for concentrating organic mutagens from drinking water. It was found that drinking water of four out of six cities showed mutagenic activity in volumes varying from 0.5 to 3 litre.<p/>To see whether the mutagenic results obtained in the six cities are representive for drinking water in The Netherlands, an extensive inventory study on the presence of mutagenic activity in several types of drinking water was carried out. In this study eighteen cities (twenty drinking waters) were investigated three times for the presence of mutagenic activity and the following chemical parameters : AOCL, EOCl, THM, VOCl, TOC, Tot. N, over a period of 2 years. The chemical parameters were measured to see whether a relationship could be found between one or more of these parameters and mutagenic activity.<p/>The selection of the drinking water was based on the water source (ground water, surface water) the storage facility (dune filtration, bankfiltration, storage reservoir) and the application of a chlorine treatment during the preparation of drinking water. The results of this study showed that in fourteen of the twenty drinking waters, mutagenic activity could be detected in volumes varying from 0.5 to 3 litre. When the cities were classified according to their water source, storage facility and type of treatment, it appeared that only three of the fifteen cities which prepare their drinking water from surface water or a mixture of surface- and groundwater, did not show mutagenic activity. Two of the five cities which use groundwater as a drinking water source showed mutagenic activity although in one city the activity was marginal.<p/>Correlating the chemical parameters with the mutagenic activity, it was demonstrated that AOCL showed the highest correlation with the direct mutagenic activity in strain TA 98 and TA 100. It was also shown that a chlorine treatment applied during drinking water preparation correlated well with the direct mutagenic activity in strain TA 98 and TA 100. In addition, chemical parameters which showed a significant difference (p< 0.01) in concentration between mutagenic and non mutagenic samples (Mann-Whitney) and a significant correlation (p< 0.01, Kendall tau) with mutagenic activity were used for straight curve fitting. From these fitted regressions, concentration levels were calculated above which a doubling of revertants may be expected.<p/>In the second part of Chapter 3 (3.4) the influence of different treatment processes on the mutagenic activity and some chemical parameters were investigated in three waterworks. Application of a chlorine treatment, generally increased the direct and promutagenic activity, but the amount of increase proved to be dependent on the type of water chlorinated. The use of ozone in the preparation of drinking water decreased the mutagenic activity in the water. The amount of reduction was dependent on the type of water ozonated. Dune filtration greatly reduced the mutagenic activity.<p/>Slow sand filtration could not be evaluated, because of the toxicity of the organic concentrates for the bacterial strains. Filtration over active carbon filters which operated for about 1 year, reduced the mutagenic activity below the detection level. Carbon filters which operated more than 1.5 years in a pilot plant, showed a break-through of mutagenic activity. This result suggests that carbon filters are able to remove organic mutagens only for a short time.<p/>From the results of the chemical parameters before and after the different treatment processes it appeared that the level of AOCL, behaved very similar with the mutagenic activity in the neutral fraction. These results confirm those obtained in the eighteen city survey and support the idea that A0C1 might be a useful indicator for mutagenic activity in drinking water.<p/>The physical-chemical characterization of organic mutagens in drinking water concentrates is described in Chapter 4. The first approach was to examine the influence of the pH on the adsorption behaviour of organic mutagens on the XAD resins. Lowering the original pH of drinking water with HCl to pH = 2- 3, revealed the presence of another class of organic mutagens the so-called acid fraction which hardly adsorb on the XAD resins at pH = 7.5.<p/>Further it was investigated whether different organic solvents are able to eluate the organic mutagens from the XAD column. it was found that diethylether only eluted a minor part of the mutagenic activity from the XAD resins and subsequent elution with acetone eluted the major part of the activity.<p/>This result showed that it is not likely that the organic mutagens in the acetone fraction are identical with the already identified organics in the ether fraction in this drinking water. Furthermore it appeared that the organic mutagens arenot only less volatile but also resistent to boiling. Another approach to characterize the organic mutagens was to apply fractionation techniques. Using TLC and HPLC, it was found with HPLC analysis that the mutagenic activity was present predominantly in two fractions. Finally, gelfiltration on Sephadex LH20, showed that organic mutagens which demonstrated mutagenic activity with strain TA 98, had a molecular weight in the order of 200.<p/>Chapter 5, presents the results of a carcinogenicity study, in which Wistar SSP TOX rats were exposed to mutagenic drinking water concentrates of one city in The Netherlands. Drinking water concentrates were prepared every week using the XAD concentration procedure and the organic concentrates (DMSO) were mixed with the non mutagenic drinking water of the Wistar SSP TOX rats.<p/>Dosage levels were based on multiples of expected human exposure levels. In the calculation the average human exposure was assumed to be approximately 29 ml/kg bw./day. Furthermore it was assumed that a rat of 250 gram would consume about 30 ml water per day. In the experiment rats were divided into four groups (50 males and 50 females per group), a control group and groups which received respectively 10, 30 and 90 times the human exposure level in their drinking water.<p/>During the experiment (106 weeks) the water consumption of the rats was measured weekly. Body weights were recorded weekly in the first two months of the experiment and once a month thereafter. The mutagenic activity of concentrates which were mixed with non mutagenic drinking water were measured after a week. It was found that the mutagenic activity with strain TA 98 could be recovered for 60-90%, while the TA 100 activity hardly could be recovered. The latter was explained by the observed partial decrease of TA 100 activity after mixing the mutagenic drinking water concentrate with drinking water.<p/>During the experiment it appeared that the assumed water consumption of 30 ml per day per rat, on which the dose level was based, not was reached.<p/>Moreover, it appeared that Wistar SSP TOX rats became much heavier than was expected and therefore the actual dose levels in ml/kg bw. of the three exposed groups were not 10, 30 and 90 times the expected human exposure levels, but 4.5, 14 and 40 for male rats respectively and 7, 22 and 68 for female rats. Exposing Wistar SSP TOX rats to these dose levels for 106 weeks, did not result in a significant increase (p <0.05) in tumour induction.<p/>Furthermore it was shown, that the development and types of tumours were similar in the treated and control groups. The number of animals with tumours and the animals which died of tumours in the exposed groups was not significantly different (p < 0.05) from the control group. The negative results in this carcinogenicity study indicate that mutagenic drinking water concentrates did not contain very potent carcinogens in effective concentrations like for instance N-Nitrosodiethylamine because when these kind of compounds are present, the carcinogenicity should be positive at the dose levels tested.<p/>On the other hand one cannot exclude the presence of weak carcinogens in the mutagenic concentrates, because one cannot detect a carcinogenic effect at the dosis tested. Based on the carcinogenicity results, an estimation of a risk factor was made for the people who consumed this mutagenic drinking water.<p/>For the estimation the following three assumptions had been made:<br/>- when organic carcinogens had been present in drinking water, they were present solely in the mutagenic drinking water concentrates<br/>- results obtained in the present carcinogenicity study with rats may be extrapolated to man<br/>- linear extrapolation of high dosis to low dosis give correct results.<p/>When all these assumption are correct it appeared after statistical analysis that less than 356 of 110.000 people might be at risk. On this basis the contribution of drinking water, if there is a contribution at all, seems relatively small viz. less than 1. 1 % in comparison to the expected tumour incidence of 33.000. The latter value is based on an average tumour incidence of 30 per 100, due to background processes.<p/>To improve, however the reliability of an estimation of the risk factor for people who consume polluted drinking water, organic concentratres of drinking water which have not been included in the present carcinogenicity study should be examined for complete carcinogenicity in combination with higher doses of concentrates of organic mutagens as used in this study.<p/>Besides this investigation it is recommended that both organic concentrates will be tested for tumour promotion and initiation activity as well as identification of the organic responsable for these effects should receive priority. Finally Chapter 6 describes a number of conclusions emerging from this investigation. The data presented show that mutagenic activity was detectable in drinking water of many cities in The Netherlands. The number of cities which showed mutagenic activity in their drinking water was the least where groundwater was used as a drinking water source. Therefore it is recommended that groundwater is preferable over other sources of drinking water and the quality of the groundwater should be protected very carefully so that oxidation and disinfection with a chlorine treatment will not be necessary. This recommendation is not only based on the eighteen city survey but also from data obtained in waterworks which prepare their drinking water mainly from surface water, because mutagenic activity is significantly increased by a chlorine treatment which may often result in mutagenic activity in the end product. It is recommended further that the organic mutagens should be identified in order to evalute them with respect to their possible toxic properties. Also higher dosis of mutagenic drinking water concentrates in combination with concentrates prepared from the rest of the organics in drinking water, should be tested for carcinogenicity and chronic toxicity for a more reliable risk estimation. Besides, this investigation it is recommended that both organic concentrates should be investigated for tumour promotion and initiation activity.

Dutch; Flemish. Verslag van een inventarisatie van de mutatieverwekkende en kankerverwekkende werking van het leidingwater van 18 Nederlandse steden (waarbij ratten als testdieren werden gebruikt) na verschillende zuiveringsstappen