- Email: [email protected]
Application of semipermeable membrane device for assessing toxicity in drinking water
Chemosphere 61 (2005) 1691–1699 www.elsevier.com/locate/chemosphere
Application of semipermeable membrane device for assessing toxicity in drinking water G. Gilli
b
a,*
, T. Schiliro` a, C. Pignata a, D. Traversi a, E. Carraro b, C. Baiocchi c, R. Aigotti c, D. Giacosa d, E. Fea a
a Department of Public Health and Microbiology, University of Turin, Via Santena, 5bis, 10126 Turin, Italy Department of Science and Advanced Technology, University of Eastern Piedmont A. Avogadro, Alessandria, Italy c Department of Analytical Chemistry, University of Turin, Italy d SMAT, Societa` Metropolitana Acque Torino, Italy
Received 18 October 2004; received in revised form 14 March 2005; accepted 23 March 2005 Available online 12 May 2005
Abstract Semipermeable membrane devices (SPMDs) mimic passive diffusive transport of bioavailable hydrophobic organic compounds through biological membranes and their partitioning between lipids and environmental levels. Our study was developed on a surface water treatment plant based in Turin, Northern Italy. The investigated plant treats Po River surface water and it supplies about 20% of the drinking water required by Turin city (about one million inhabitants). Surface water (input) and drinking water (output) were monitored with SPMDs from October 2001 to January 2004, over a period of 30 days. The contaminant residues, monthly extracted from SPMDs by dialysis in organic solvent, were tested with the MicrotoxTM acute toxic test and with the Ames mutagenicity test. Same extracts were also analyzed with gaschromatography—mass spectrometry technique in order to characterise the organic pollutants sampled, especially Polycyclic Aromatic Hydrocarbons (PAHs). Although the PAHs mean concentration is about one hundred times lower in the output samples, the mean toxic units are similar in drinking and surface water. Our data indicate that the SPMD is a suitable tool to assess the possible toxicity in drinking water. 2005 Elsevier Ltd. All rights reserved. Keywords: Semipermeable membrane devices (SPMDs); Toxicity; Genotoxicity; Polycyclic aromatic hydrocarbon (PAH)
1. Introduction The need of drinking water, due to the growth of urbanization and industrialization of the urban areas, has determinated a rise of surface resources for drinking
* Corresponding author. Tel.: +39 011 6705810; fax: +39 011 6705872. E-mail address: [email protected] (G. Gilli).
water. This evidence has caused an increased risk related to microbiological and chemical contaminations of the aquatic environment (WHO, 2004). Many studies have reported the presence of a variety of hazardous compounds in treated drinking water (Kraybill, 1981; Peters et al., 1990; Rehena et al., 1996; Sadiq and Rodriguez, 2004). These chemicals result sometimes from raw water pollution and eutrophication, but they can also be produced during water treatment: chemical oxidants, reacting with natural organic matter in raw water (humic and
0045-6535/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2005.03.085
1692
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
fulvia acids), can lead to the formation of disinfection by-products (De Marini et al., 1995; Sadiq and Rodriguez, 2004). The level of exposure to such toxicants or mutagens through drinking water is lower than other pathway, but this is maintained throughout life and can represent a risk factor for human population (Carraro et al., 2000). Chemical and biological evaluations through reliable sampling and analysing methods are compelled. The resources for the protection of the environment are limited, so the implementation of lowtech and cost-effective monitoring methods is often feasible. Such methods should allow to monitor chemicals fate and concentration in the environment, to evaluate their effects and to assess the potential hazard for both the ecosystem and human health (Sabaliunas et al., 1999). Semipermeable membrane devices (SPMDs) can be used as indicator of bioavailability of chemical pollutants, they are commonly used as a time integrated measure of aqueous concentration of persistent hydrophobic chemicals (Huckins et al., 1990), including PAHs, pesticides dioxins and PCBs. SPMDs accomplish two tasks simultaneously: provide a highly reproducible sampling matrix, largely unaffected by most environmental stressors that affect biomonitoring organisms; mimic the bioconcentration of organic contaminants in organismsÕ fatty tissues, thus reflect the true extent of exposure of living organism to chemical contaminants in the aquatic environment. The uptake will reflect the relative composition and concentration of pollutants in the matrix. In these devices, the uptake of chemicals is based on the process of passive partitioning of compounds between water and the synthetic lipid triolein that is enclosed in thin semipermeable polymeric membrane. The molecular size–exclusion limit of the polyethylene membrane is similar to that of biological membranes, while triolein constitutes a significant fraction of the lipidic pool of most aquatic organisms (Opperhuizen et al., 1985). Furthermore, they can be exposed to harsh environmental condition for long time and still remain operative, in contrast to most living organisms (Sodergren, 1990). In many cases, they proved to be better than aquatic organisms in terms of available chemicals for analysis (Sabaliunas et al., 1999). Some authors used SPMDs for in situ passive monitoring of aquatic contaminants (Sabaliunas et al., 1999; Wang et al., 2003; Richardson et al., 2003; Verweij et al., 2004). Most of these studies have focused on the identification and quantification of chemicals accumulated by SPMDs, but these analysis would be integrated with standard bioassay to measure toxic and mutagenic effects of accumulated pollutants in order to identify the possible hazard for environment and public health. Despite these promising results and numerous attractive qualities of SPMDs (i.e., their long-term stability, low cost, and ease of utilization), there are only limited published data pertaining to these passive sampling tools used in drinking water monitoring (IPSW, 2004).
In this study, we explored the potential of SPMDs as passive water samplers in surface and drinking water for assessing their toxic and mutagenic properties and the presence of PAHs. The results obtained by applying this method are presented, the aim was to assess both biological tests (acute toxicity test and genotoxicity test) and chemical analyses (GC–MS) on the same water sample.
2. Materials and methods 2.1. Characteristics of the surface treatment plant This study was developed with the co-operation of the Turin metropolitan water company (SMAT s.p.a.—Societa` Metropolitana Acque Torino) and undertaken in a Turin water treatment plant. The investigated plant treats up to 130 000 m3 per day of Po river surface water and it supplies about 20% of the drinking water required by Turin (about one million inhabitants). Briefly, after sedimentation and ozonation the water treatment process is divided into a ‘‘biological section’’, where the ammonia removal is performed in activated carbon filters, and a ‘‘chemical section’’, where the ammonia removal is performed using chlorine. In both sections, the treatment process is based on a clarification in Cyclofloc basin followed by two-beds biological activated carbon (BAC) or granular activated carbon (GAC) filtration. Water effluents from the two sections are collected in a tank and chlorinated with chlorine dioxide (0.2–0.4 mg/l) before entering a second tank, from which it is pumped in the distribution network. 2.2. Water sampling Standard SPMDs (Environmental Sampling Technologies Laboratories, US Patents 5,098,573 and 5,395,426 and Canadians Patent #2,037,320) were used for a minimum of four weeks to a maximum of six weeks in two points of the Turin drinking water plant. The design of commercially available SPMDs consists of a specified length, typically 91.4 cm, of 2.5 cm wide layflat LDPE tubing, containing 1 ml of triolein (95% purity). SPMD sampling was performed according to recommended good SPMD practice: immersed in hexane to remove monomers and others impurities for 24 h, then placed in clean airtight steel cans and transported to sampling places with transport-trip and SPMD field blanks. Both Po river surface water and drinking water, obtained at the end of the treatment process, were sampled in tanks with a continuous water change using a vertical perforated stainless steel container to protect the membranes against mechanical damage and to restrict water flow velocity at the membrane. Water temperatures were measured automatically in continuity. Number of exposed SPMDs per one site were
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
given to tested parameters and QA/QC aspect; in this research two membranes were used per site. Surface water (INPUT) and drinking water (OUTPUT) were monitored with SPMDs from October 2001 to January 2004 (24 samplings, for a total of 96 SPMDs). Field blank SPMDs, for the quality control (a minimum of one for each sampling), had accompanied the sampler during the transport, deployment, retrieval and are subsequently processed and analyzed exactly as deployed SPMDs. After being sampled, each sampler was rinsed by clean water and placed in a clean airtight steel can. Periphyton, minerals and rough particulates were then removed from membrane surface with clean cloth and then rinsed by clean water. Exposed membranes were preserved frozen at 20 C until analyzed. One SPMD per site was utilized for chemical analysis and the other was cut in three fragments (30 cm): two for the biological tests and one to be stored; for this step a standardized procedure (sterilized scissor and forceps) was utilized. 2.3. Biological assays Fragments of SPMDs were solubilized with 20 ml of an acetone:DMSO mixture (1:1, v/v) for 24 h. Then, the acetone was evaporated by a gentle stream of nitrogen. If it is not otherwise specified all chemicals were purchased from Sigma, USA. 2.3.1. MicrotoxTM test SPMDs extracts dissolved in DMSO (diluted to achieve a 1% DMSO concentration) were tested for acute toxicity with Microtox bioassay as decribed in MicrotoxTM manual (1995). The MicrotoxTM test uses bioluminescent sea-bacteria (Vibrio fischeri) as biological matrix. The principle of the system is based on the evaluation of the luminous energy naturally emitted by these bacteria. A toxic substance will cause changes to the cellular state, which are rapidly reflected in a decrease of bioluminescence. This light reduction is proportional to the sample toxicity. Briefly, the luminosity values registered at different times are used to calculate the brightness loss (H%). Luminescence was measured at time zero and after 5 0 , 15 0 , 30 0 and compared to the control. The final expression of the samples toxic potentialities is expressed as ‘‘Effective Concentration’’ at 30 min, EC50, showing the sample concentration factor which has caused a decrease of the bacteria brightness in the 50% of the organisms population. EC50s were subsequently converted in toxic units (TU) that are proportional to toxicity: TU ¼ ð1=EC50Þ 100 Considering the empiric toxicity scale (expressed in TU) adopted in Belgium and submitted to the evaluation of the European Community Commission (ACE 89/BE2/
1693
Table 1 Toxicity scale (expressed in TU) and judgement of toxicity (ACE 89/BE2/D3) Toxic units
Toxicity judgement
>100 100–11 10–1 <1 No toxic effect
Extremely toxic Very toxic Toxic Weakly toxic Acute toxicity absence
D3), the toxicity judgement depends on the values as shown in Table 1. The samples toxicity can also be expressed as Vtox (Kocı´ et al., 2003), which represents a volume of media that is theoretically needed for dilution of all toxicants absorbed in one membrane during one average day of deployment to obtain a solution with a chosen effective concentration, for example EC50. The higher Vtox reflects the larger volume of toxicants absorbed and thus the higher contamination of the sampled site. The following formula defines Vtox, where (m) is the concentration of extracted membranes in solvent mixture expressed as number of membranes in ml of solvent mixture (pcs/ml), (d) is duration of deployment of membrane during a sampling (days) and ECxx is an effective concentration of extract on the chosen organism, for example EC50 (ml/l): Vtoxð50Þ ¼ 1=ðm EC50 dÞ For each test, as quality control, we analyzed a MicrotoxTM reference toxicant, phenol, DMSO and a field blank SPMD. 2.3.2. Ames test The mutagenic activity of the SPMDs extracts, dissolved in 10 ml of DMSO, was determined using the Salmonella/microsome assay according to the standard plate method of Maron and Ames (1983). Each sample was evaluated with and without metabolic activation (10% S9 mix), using TA98 and TA100 Salmonella typhimurium strains, which are sensitive to the majority of water mutagens (Albaladejo et al., 1995). Solvent mutagenicity and triolein assays were included to check sample preparation interferences. Sodium azide (SA) and 2-nitrofluorene (2-NF), 1 lg/plate, were used as positive control for TA100 and TA98 without S9, respectively; 2-aminofluorene (2-AF), 2 lg/plate, was used to assess microsomal fraction efficiency. Mutagenic activity was expressed as a mutagenicity ratio (MR), excluding the contribution of the pre-existing revertants (Georghiou et al., 1989). MR ¼
net revertants ðpre-existing revertantsÞ spontaneous revertants ðpre-existing revertantsÞ
1694
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
Samples were considered mutagenic when net revertants values were at least twice as much as the frequency of spontaneous revertants, and when a linear doseresponse relationship, for the four doses employed, was indicated. 2.4. Chemical analysis (GC/MS) Dialysis was performed in hexane on the second SPMD fragment in the dark for 24-h. The solvent was then rinsed and the sample dialysed for further 48 h. The dialysed sample was combined and the volumes were reduced using rotary evaporation. The sample was then re-suspended in 500 ll of hexane. The extract obtained contains a substantial amount of material different from the PAHs, which may interfere with the analytical determination. In order to purify the sample the extracts were loaded on 500 mg SPE silica cartridges, PAHs fraction was collected eluting with cyclohexane. GC analyses of PAHs were accomplished with a Varian 3800 gas-chromatograph coupled with a Varian Saturn 2000 IT mass detector. The GC/MS was operating in the single ion storage (SIS) acquisition mode. The separation of PAHs was performed injecting in split mode 2 ll of purified samples extract into Varian 1079 PTV split/splitless injector. Separation was achieved with a 30 m capillary column. 250 · 0.25 lm Varian Facto Four VF5 Column by means of a thermal gradient from 80 C to 300 C at 10 C/min. The injection port temperature, transfer line temperature and line trap temperature were held isothermally at 280 C, 270 C and 220 C, respectively. Quantification was performed by mean of a multi level internal standard calibration. Internal standard of labelled PAHÕs used for quantification purposes were naphthalene-d8, phenanthrene-d10, chrysene-d12. Three replications for each standard calibration level were performed during analysis in order to obtain instrumental calibration.
3. Results 3.1. Biological assays A segment extract of SPMD per site, at four different dilutions (in triplicate), for each assay, was tested. 3.1.1. MicrotoxTM test Table 2 gives EC50s to determine the potential toxicity of SPMDs extracts, including controls (solvents and field blank SPMD), for the analysis of replicates samples the RSDs were <15%. The surface water was classified as weakly toxic (TU range: 0.03–0.92) for V. fischeri. The drinking water was no toxic except for January, August, September, November and December 2002 when it was weakly toxic (TU range: 0.50–0.87) and only during
Table 2 Mean EC50s of SPMDs extracts and controls, assayed with MicrotoxTM Sample type SPMDs Surface water (input) Drinking water (output) Quality control Field blank SPMDs Triolein DMSO Microtox phenol reference toxicant (lg/ml H2O)
Mean MicrotoxTM toxicity EC50 (ds) 4.92 (7.20) 4.99 (11.92) 23.81 (4.27) nd nd 17.95 (3.84)
nd: none detected.
March 2002 it was classified as toxic (TU = 1). We had taken into account that field blank SPMDs had shown 0.04 ± 0.01 TU. Fig. 1 shows the toxicity data, expressed in TU/mg triolein and in Vtox(50) ml/day obtained for SPMDs samples from surface and drinking water. The mean toxic units were similar for drinking water (TU/ mg triolein = 1.82 ± 2.34) and for surface water (TU/ mg triolein = 1.67 ± 1.55), there is no statistical significative difference. Considering the Vtox(50), the means were similar for drinking water (Vtox(50) ml/day = 1.62 ± 2.09) and for surface water (Vtox(50) ml/day = 1.41 ± 1.31), also for this parameter there is no statistical significative difference. There is a positive and significative correlation between toxic units and Vtox(50) both for surface water (r = 0.967, p < 0.01) and drinking water (r = 0.928, p < 0.01). 3.1.2. Ames test Spontaneous reversion of the tester strains to histidine independence is measured in each Ames test and is expressed as the number of spontaneous revertants per plate; spontaneous reversion for each strain is at a frequency that is characteristic of the strain. The data are expressed as the average revertants number per plate from the triplicates; Table 3 summarizes the means results obtained from the study tests. The results indicate that neither surface water nor drinking water increase the revertants numbers in all two tester strains at all four concentrations tested in comparison to the spontaneous controls. MR values were always below 2, so any samples were considered mutagenic. Any sample was considered mutagenic because the data generated in these assays are acceptable, because the spontaneous revertants number is within the normal range reported in the literatureÕs data, the positive mutagens (both in activated and in non-activated system) yielded a significant increase (2–9 fold increase) in the number of revertants as compared to control. Thus, at the tested concentrations, we did not report the problem of cytotoxicity.
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
1695
Fig. 1. Toxicity data (expressed both in toxic units and in Vtox) obtained for SPMDs samples deployed in surface (INPUT) and in drinking water (OUTPUT).
Table 3 The means results of mutagenicity assays (Ames test) expressed as the average revertants number per plate from triplicates Mean revertants/plate (DS) Sample type Spontaneous SPMDs Water surface 10 ll 20 ll 50 ll 100 ll Drinking water 10 ll 20 ll 50 ll 100 ll Quality control Field blank SPMDs Triolein DMSO (100 ll) SA (1 lg/plate) 2-NF (1 lg/plate) 2-AF (2 lg/plate)
TA98 23 (6)
TA100 131 (14)
TA98 + S9 38 (7)
TA100 + S9 137 (13)
26 28 26 24
(10) (9) (8) (11)
130 125 124 142
(25) (22) (18) (20)
39 35 45 42
(9) (11) (7) (13)
132 145 139 135
(15) (19) (22) (17)
25 28 29 33
(11) (9) (11) (8)
138 148 152 150
(19) (15) (20) (17)
34 33 36 40
(8) (11) (10) (10)
130 140 125 135
(16) (18) (10) (19)
155 120 133 286
(10) (10) (12) (34)
40 (8) 37 (7)
150 (8) 141 (9)
348 (29)
292 (40)
26 (7) 22 (4) 24 (5) 57 (10)
3.2. Chemical analysis Fig. 2 represents both standards and sample processed by assaying one SPMD extract with GC/MS.
The GC/MS analyses were performed on ten samples, three of which were inputs and seven outputs. PHAs concentration showed to be higher in the input samples than outputs (Table 4), where concentrations were about
1696
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
Fig. 2. PAHs analysis: standards (down) and sample (up) processed by assaying one SPMD extract with GC/MS.
Table 4 PAHs concentrations in surface (Input) and in drinking water (Output), LOD and RDS PAHs
LOD
RDS Input (lg/mg triolein)
Output (lg/mg triolein)
Oct-02 Nov-02 Dec-02 Jul-02
Aug-02 Sep-02 Oct-02 Nov-02 Dec-02 Feb-03
Naphthalene 1-Methylnaphthalene 2-Methylnaphthalene 1-Chloranapthalene Acenaphtylene Acenaphthalene Phenanthrene Anthracene Pyrene Phluoranthene Chrysene Benz[a]anthracene Benzo[j]phluoranthene Benzo[k]phluoranthene Benzo[a]pyrene Indeno[1,2,3-cd]pyrene Dibenz[a,h]anthracene Benzo[g,h,i]perylene
<0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±5% ±5% ±6% ±5% ±8% ±7% ±7%
1.489 2.965 1.816 <0.006 0.212 0.350 0.687 <0.006 0.123 0.170 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.339 1.355 0.775 <0.006 0.142 0.253 0.502 <0.006 0.129 0.153 <0.006 <0.006 <0.006 0.198 <0.003 <0.006 <0.002 <0.002
15.129 25.885 17.682 <0.006 2.677 1.525 27.443 2.983 28.815 38.022 19.335 27.492 15.291 24.033 10.731 <0.006 <0.002 <0.002
0.248 0.216 0.314 <0.006 <0.006 <0.006 0.096 <0.006 0.075 <0.006 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.467 1.109 0.863 <0.006 <0.006 <0.006 0.232 <0.006 0.113 0.031 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.186 0.322 0.326 <0.006 <0.006 0.234 0.116 <0.006 0.073 0.033 <0.006 0.036 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.201 0.165 <0.006 <0.006 <0.006 <0.006 0.082 <0.006 0.037 0.013 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.190 0.238 <0.006 <0.006 0.094 0.206 0.099 <0.006 0.038 <0.006 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 0.062
0.100 0.076 <0.006 <0.006 <0.006 <0.006 0.130 0.016 0.111 0.101 <0.006 0.075 0.103 0.067 0.042 <0.006 <0.002 <0.002
0.121 0.070 <0.006 <0.006 0.080 <0.006 0.098 0.052 <0.006 <0.006 <0.006 <0.006 <0.006 0.027 <0.003 <0.006 <0.002 <0.002
Total PHAs
<0.006 ±3%
7.811
3.864
257.043
0.948
2.816
1.326
0.498
0.928
0.821
0.449
one or two orders of magnitude lower. In our study, 1methylnaphthalene, 2-methylnaphthalene and naphthalene are the most represented PAH compounds. Fig. 3
shows that there is no relation between acute toxicity and PHAs (SpearmanÕs r = 0.030; p > 0.05). Analytical recoveries for known analytes are determined by adding
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
1697
Fig. 3. Acute toxicity (TU) and PAHs concentration in surface water (black bars) and in drinking water (grey bars) SPMDs extracts.
to the triolein of non-exposed SPMDs appropriate levels of the analytes. Generally, such SPMDs are spiked with chemicals at levels equivalent to the midpoint of the calibration curve. Recoveries are generally >70%, with good precision (<15% RSD). The RSDs for analysis of replicate samples are often <8% (Table 4), thus, these data are confirmed in the literature (Petty et al., 2000a,b). The water temperatures, in the tanks used for the SPMDs samplings, were 14.2 C (±4.6, range 7.3– 23.2 C) for surface water while for drinking water was 14.7 (±4.7, range 7.8–23.7 C). There are not significative correlations between temperature and TUs or PAHs, neither for surface water nor for drinking water, infact for field application of SPMDs, the results imply that temperature is not a key factor that controls uptake rates (Booij et al., 2003).
4. Discussion and conclusions With respect to PAH compounds, data obtained with surface water samples have shown to be higher when compared to those in the literature (McCarthy and Gale, 1999; Louch et al., 2003). On the other hand, to our knowledge, there have been only few reports on such analysis for drinking water. The acute toxicity of the samples is similar in drinking and surface water. Such data were probably ob-
tained because each membrane was used for about thirty days sampling about 150 l and so the toxicity data were referred to this volume of drinking water. Therefore, the presence of by-products in the chlorinated drinking water has probably influenced the toxicity. All input membranes presented a reddish brown colour biofilm: there is the possibility that the uptake of contaminants by SPMDs could have been severely reduced under fouling condition as reported by Richardson et al. (2002). For both drinking and surface water, we always found low acute toxicity cases (range between 0 and 11.6 TU/mg triolein) and a part of these are also due to the SPMD oleic acid as confirmed by Sabaliunas et al. (1999). The absence of correlation between the acute toxicity results and the PAHs concentrations highlights that toxicity could be due to more than one class of pollutants. This finding is in line with literature data (Wang et al., 2003) confirming the great importance of the biological assays to evaluate the effects of complex mixture like water. The results indicate that drinking water treatment process is able to reduce the presence of the evaluated PAHs. Although the effect of single PAHs on Salmonella activated strains is reported (i.e. Bostrom et al., 1998), our data indicate that SPMD extracts are not mutagenic in the Ames test; we had to consider that the highest tested concentration of the extracts corresponds to 10 lg of triolein in which the concentration of total PHAs is low: it ranged from 0.004 to 0.028 lg for drinking
1698
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
water and from 0.039 to 2.570 lg for surface water (Table 4). SPMDs showed to be an excellent tool in assessing the presence of hazardous compounds, therefore this method could be of wide interest in assessing toxic contaminants in input and output drinking water treatment plants for their capability in determining pollutant sources and relative levels at different locations. Furthermore, the results of this study have highlighted the usefulness of MicrotoxTM toxicity test, in addition to the conventional chemical examination of water quality, to evaluate the concentration of ambient bioavailable chemicals and their possible additive effects. It would be of great interest to carry out additional studies on the evaluation of genotoxic properties by applying another extraction method from the membranes and, furthermore, to determine the ability of the SPMDs to capture by-products present in drinking water. Acknowledgement This research was funded by Societa` Metropolitana Acque Torino s.p.a. References Albaladejo, V., Villanueva, O., Ortega, M., 1995. The evaluation of the mutagenic activity of public drinking water by the Ames test. Rev. Esp. Salud. Publica. 5 (69), 393–408. Booij, K., Hofmans, H.E., Fischer, C.V., Van Weerlee, E.M., 2003. Temperature-dependent uptake rates of nonpolar organic compounds by semipermeable membrane devices and low-density polyethylene membranes. Environ. Sci. Technol. 37 (2), 361–366. Bostrom, E., Engen, S., Eide, I., 1998. Mutagenicity testing of organic extracts of diesel exhaust particles after spiking with polycyclic aromatic hydrocarbons (PAH). Arch. Toxicol. 72 (10), 645–649. Carraro, E., Bugliosi, E.H., Meucci, L., Baiocchi, C., Gilli, G., 2000. Biological drinking water treatment processes, with special reference to mutagenicity. Water Res. 34, 3042–3054. De Marini, D.M., Abu-Shakra, A., Felton, C.F., Patterson, K.S., Shelton, M.L., 1995. Mutation spectra in Salmonella of chlorinated, chloramminated, or ozonated drinking water extracts: comparison to MX. Environ. Mol. Mutagen. 26 (4), 270–285. Georghiou, P.E., Bladgen, P.A., Winsor, L., Williams, D.T., 1989. Spontaneous revertants in modified S. typhimurium mutagenicity test employing elevated numbers of the tester strain. Mut. Res. 225, 33–39. Huckins, J.N., Tubergen, M.W., Manuweera, G.K., 1990. Semipermeable membrane devices containing model lipid: a new approach to monitoring the availability of lipophilic contaminants and estimating their bioconcentration potential. Chemosphere 20, 533–552.
IPSW, 2004. In: International Passive Sampling Workshop and Symposium, Conference Proceedings, 14-18/4/2004, Ceske Budejovice. Kocı´, V., Mlejneky, M., Kocha´nkova´z, L., 2003. Toxicological evaluation of exposed SPMD membranes. Central European Journal of Chemistry 1, 28–34. Kraybill, H.F., 1981. Carcinogenesis of synthetic organic chemicals in drinking water. JAWWA 73, 370–372. Louch, J., Allen, G., Erickson, C., Wilson, G., Schmedding, D., 2003. Interpreting results from field deployments of SPMDs, 2003. Environ. Sci. Technol. 37, 1202–1207. Maron, D.M., Ames, B., 1983. Revised method for the Salmonella mutagenicity test. Mut. Res. 113, 172–215. McCarthy, K.A., Gale, R.W., 1999. Investigation of the distribution of organochlorine and PAH compounds in the lower Columbia river using SPMDs. USGS, Water resource investigations report 99-4051. MICROTOX, 1995. MicrotoxTM acute toxicity basic test. Azur Environmental. Opperhuizen, A., Van der Velde, E.W., Gobas, F.A., Liem, A.K., Van der Stehen, J., Hutzinger, O., 1985. Relationship between bioconcentration in fish and steric factors of hydrophobic chemicals. Chemosphere 14, 1871–1896. Peters, R.J.B., De Leer, E.W.B., De Galan, L., 1990. Dihaloacetonitriles in Dutch drinking water. Water Res. 24, 797– 800. Petty, J.D., Jones, S.B., Huckins, J.N., Cranor, W.L., Parris, J.T., McTague, T.B., Boyle, T.P., 2000a. An approach for assessment of water quality using SPMDs and bioindicator tests. Chemosphere 41, 311–321. Petty, J.D., Orazio, C.E., Huckins, J.N., Gale, R.W., Lebo, J.C., Meadows, J.C., Echols, K.R., Cranor, W.L., 2000b. Consideration involved with use of semipermeable membrane devices for monitoring environmental contaminants. J. Chromatogr. A 879, 83–95. Rehena, Z., Malik, A., Ahmad, M., 1996. Genotoxicity of the Ganges water at Narora (UP), India. Mut. Res. 367, 187– 193. Richardson, B., Lam, P.K.S., Zheng, G., McClellan, K.E., De Luca-Abbott, S., 2002. Biofouling confounds the uptake of trace organic contaminants by SPMDs. Marine Pollut. Bull. 44, 1372–1379. Richardson, B.J., Zheng, G.J., Tse, E., Lam, P.K.S., 2003. A comparison of mussels (Perna viridis) and SPMDs for monitoring chlorinated trace organic contaminants in Hong Kong coastal waters. Chemosphere 45, 1201– 1208. Sabaliunas, D., Ellington, J., Sabaliuniene, I., 1999. Screening bioavailable hydrophobic toxicants in surface waters with SPMDs: role of inherent oleic acid in toxicity evaluations. Ecotoxicol. Environ. Saf. 44, 160–167. Sadiq, R., Rodriguez, M.J., 2004. Disinfection by-products (DBPs) in drinking water and predictive models for their occurrence: a review. Sci. Total Environ. 321, 21–46. Sodergren, A., 1990. Monitoring of persistent, lipophilic pollutants in water and sediment by solvent-filled dialysis membranes. Ecotoxicol. Environ. Saf. 19, 143–149. Verweij, F., Booij, K., Satumalay, K., Van der Molen, N., Van der Oost, R., 2004. Assessment of bioavailable PAH, PCB
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699 and OCP concentrations in water, using semipermeable membrane devices (SPMDs), sediments and caged carp. Chemosphere 54 (11), 1675–1689. Wang, C., Wang, Y., Kiefer, F., Yediler, A., Zijian, W., Kettrup, A., 2003. Ecotoxicological and chemical charac-
1699
terization of selected treatment process effluents of municipal sewage treatment plant. Ecotoxicol. Environ. Saf. 56, 211–217. WHO, 2004. Guidelines for drinking water quality, Third edition.
Application of semipermeable membrane device for assessing toxicity in drinking water G. Gilli
b
a,*
, T. Schiliro` a, C. Pignata a, D. Traversi a, E. Carraro b, C. Baiocchi c, R. Aigotti c, D. Giacosa d, E. Fea a
a Department of Public Health and Microbiology, University of Turin, Via Santena, 5bis, 10126 Turin, Italy Department of Science and Advanced Technology, University of Eastern Piedmont A. Avogadro, Alessandria, Italy c Department of Analytical Chemistry, University of Turin, Italy d SMAT, Societa` Metropolitana Acque Torino, Italy
Received 18 October 2004; received in revised form 14 March 2005; accepted 23 March 2005 Available online 12 May 2005
Abstract Semipermeable membrane devices (SPMDs) mimic passive diffusive transport of bioavailable hydrophobic organic compounds through biological membranes and their partitioning between lipids and environmental levels. Our study was developed on a surface water treatment plant based in Turin, Northern Italy. The investigated plant treats Po River surface water and it supplies about 20% of the drinking water required by Turin city (about one million inhabitants). Surface water (input) and drinking water (output) were monitored with SPMDs from October 2001 to January 2004, over a period of 30 days. The contaminant residues, monthly extracted from SPMDs by dialysis in organic solvent, were tested with the MicrotoxTM acute toxic test and with the Ames mutagenicity test. Same extracts were also analyzed with gaschromatography—mass spectrometry technique in order to characterise the organic pollutants sampled, especially Polycyclic Aromatic Hydrocarbons (PAHs). Although the PAHs mean concentration is about one hundred times lower in the output samples, the mean toxic units are similar in drinking and surface water. Our data indicate that the SPMD is a suitable tool to assess the possible toxicity in drinking water. 2005 Elsevier Ltd. All rights reserved. Keywords: Semipermeable membrane devices (SPMDs); Toxicity; Genotoxicity; Polycyclic aromatic hydrocarbon (PAH)
1. Introduction The need of drinking water, due to the growth of urbanization and industrialization of the urban areas, has determinated a rise of surface resources for drinking
* Corresponding author. Tel.: +39 011 6705810; fax: +39 011 6705872. E-mail address: [email protected] (G. Gilli).
water. This evidence has caused an increased risk related to microbiological and chemical contaminations of the aquatic environment (WHO, 2004). Many studies have reported the presence of a variety of hazardous compounds in treated drinking water (Kraybill, 1981; Peters et al., 1990; Rehena et al., 1996; Sadiq and Rodriguez, 2004). These chemicals result sometimes from raw water pollution and eutrophication, but they can also be produced during water treatment: chemical oxidants, reacting with natural organic matter in raw water (humic and
0045-6535/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2005.03.085
1692
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
fulvia acids), can lead to the formation of disinfection by-products (De Marini et al., 1995; Sadiq and Rodriguez, 2004). The level of exposure to such toxicants or mutagens through drinking water is lower than other pathway, but this is maintained throughout life and can represent a risk factor for human population (Carraro et al., 2000). Chemical and biological evaluations through reliable sampling and analysing methods are compelled. The resources for the protection of the environment are limited, so the implementation of lowtech and cost-effective monitoring methods is often feasible. Such methods should allow to monitor chemicals fate and concentration in the environment, to evaluate their effects and to assess the potential hazard for both the ecosystem and human health (Sabaliunas et al., 1999). Semipermeable membrane devices (SPMDs) can be used as indicator of bioavailability of chemical pollutants, they are commonly used as a time integrated measure of aqueous concentration of persistent hydrophobic chemicals (Huckins et al., 1990), including PAHs, pesticides dioxins and PCBs. SPMDs accomplish two tasks simultaneously: provide a highly reproducible sampling matrix, largely unaffected by most environmental stressors that affect biomonitoring organisms; mimic the bioconcentration of organic contaminants in organismsÕ fatty tissues, thus reflect the true extent of exposure of living organism to chemical contaminants in the aquatic environment. The uptake will reflect the relative composition and concentration of pollutants in the matrix. In these devices, the uptake of chemicals is based on the process of passive partitioning of compounds between water and the synthetic lipid triolein that is enclosed in thin semipermeable polymeric membrane. The molecular size–exclusion limit of the polyethylene membrane is similar to that of biological membranes, while triolein constitutes a significant fraction of the lipidic pool of most aquatic organisms (Opperhuizen et al., 1985). Furthermore, they can be exposed to harsh environmental condition for long time and still remain operative, in contrast to most living organisms (Sodergren, 1990). In many cases, they proved to be better than aquatic organisms in terms of available chemicals for analysis (Sabaliunas et al., 1999). Some authors used SPMDs for in situ passive monitoring of aquatic contaminants (Sabaliunas et al., 1999; Wang et al., 2003; Richardson et al., 2003; Verweij et al., 2004). Most of these studies have focused on the identification and quantification of chemicals accumulated by SPMDs, but these analysis would be integrated with standard bioassay to measure toxic and mutagenic effects of accumulated pollutants in order to identify the possible hazard for environment and public health. Despite these promising results and numerous attractive qualities of SPMDs (i.e., their long-term stability, low cost, and ease of utilization), there are only limited published data pertaining to these passive sampling tools used in drinking water monitoring (IPSW, 2004).
In this study, we explored the potential of SPMDs as passive water samplers in surface and drinking water for assessing their toxic and mutagenic properties and the presence of PAHs. The results obtained by applying this method are presented, the aim was to assess both biological tests (acute toxicity test and genotoxicity test) and chemical analyses (GC–MS) on the same water sample.
2. Materials and methods 2.1. Characteristics of the surface treatment plant This study was developed with the co-operation of the Turin metropolitan water company (SMAT s.p.a.—Societa` Metropolitana Acque Torino) and undertaken in a Turin water treatment plant. The investigated plant treats up to 130 000 m3 per day of Po river surface water and it supplies about 20% of the drinking water required by Turin (about one million inhabitants). Briefly, after sedimentation and ozonation the water treatment process is divided into a ‘‘biological section’’, where the ammonia removal is performed in activated carbon filters, and a ‘‘chemical section’’, where the ammonia removal is performed using chlorine. In both sections, the treatment process is based on a clarification in Cyclofloc basin followed by two-beds biological activated carbon (BAC) or granular activated carbon (GAC) filtration. Water effluents from the two sections are collected in a tank and chlorinated with chlorine dioxide (0.2–0.4 mg/l) before entering a second tank, from which it is pumped in the distribution network. 2.2. Water sampling Standard SPMDs (Environmental Sampling Technologies Laboratories, US Patents 5,098,573 and 5,395,426 and Canadians Patent #2,037,320) were used for a minimum of four weeks to a maximum of six weeks in two points of the Turin drinking water plant. The design of commercially available SPMDs consists of a specified length, typically 91.4 cm, of 2.5 cm wide layflat LDPE tubing, containing 1 ml of triolein (95% purity). SPMD sampling was performed according to recommended good SPMD practice: immersed in hexane to remove monomers and others impurities for 24 h, then placed in clean airtight steel cans and transported to sampling places with transport-trip and SPMD field blanks. Both Po river surface water and drinking water, obtained at the end of the treatment process, were sampled in tanks with a continuous water change using a vertical perforated stainless steel container to protect the membranes against mechanical damage and to restrict water flow velocity at the membrane. Water temperatures were measured automatically in continuity. Number of exposed SPMDs per one site were
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
given to tested parameters and QA/QC aspect; in this research two membranes were used per site. Surface water (INPUT) and drinking water (OUTPUT) were monitored with SPMDs from October 2001 to January 2004 (24 samplings, for a total of 96 SPMDs). Field blank SPMDs, for the quality control (a minimum of one for each sampling), had accompanied the sampler during the transport, deployment, retrieval and are subsequently processed and analyzed exactly as deployed SPMDs. After being sampled, each sampler was rinsed by clean water and placed in a clean airtight steel can. Periphyton, minerals and rough particulates were then removed from membrane surface with clean cloth and then rinsed by clean water. Exposed membranes were preserved frozen at 20 C until analyzed. One SPMD per site was utilized for chemical analysis and the other was cut in three fragments (30 cm): two for the biological tests and one to be stored; for this step a standardized procedure (sterilized scissor and forceps) was utilized. 2.3. Biological assays Fragments of SPMDs were solubilized with 20 ml of an acetone:DMSO mixture (1:1, v/v) for 24 h. Then, the acetone was evaporated by a gentle stream of nitrogen. If it is not otherwise specified all chemicals were purchased from Sigma, USA. 2.3.1. MicrotoxTM test SPMDs extracts dissolved in DMSO (diluted to achieve a 1% DMSO concentration) were tested for acute toxicity with Microtox bioassay as decribed in MicrotoxTM manual (1995). The MicrotoxTM test uses bioluminescent sea-bacteria (Vibrio fischeri) as biological matrix. The principle of the system is based on the evaluation of the luminous energy naturally emitted by these bacteria. A toxic substance will cause changes to the cellular state, which are rapidly reflected in a decrease of bioluminescence. This light reduction is proportional to the sample toxicity. Briefly, the luminosity values registered at different times are used to calculate the brightness loss (H%). Luminescence was measured at time zero and after 5 0 , 15 0 , 30 0 and compared to the control. The final expression of the samples toxic potentialities is expressed as ‘‘Effective Concentration’’ at 30 min, EC50, showing the sample concentration factor which has caused a decrease of the bacteria brightness in the 50% of the organisms population. EC50s were subsequently converted in toxic units (TU) that are proportional to toxicity: TU ¼ ð1=EC50Þ 100 Considering the empiric toxicity scale (expressed in TU) adopted in Belgium and submitted to the evaluation of the European Community Commission (ACE 89/BE2/
1693
Table 1 Toxicity scale (expressed in TU) and judgement of toxicity (ACE 89/BE2/D3) Toxic units
Toxicity judgement
>100 100–11 10–1 <1 No toxic effect
Extremely toxic Very toxic Toxic Weakly toxic Acute toxicity absence
D3), the toxicity judgement depends on the values as shown in Table 1. The samples toxicity can also be expressed as Vtox (Kocı´ et al., 2003), which represents a volume of media that is theoretically needed for dilution of all toxicants absorbed in one membrane during one average day of deployment to obtain a solution with a chosen effective concentration, for example EC50. The higher Vtox reflects the larger volume of toxicants absorbed and thus the higher contamination of the sampled site. The following formula defines Vtox, where (m) is the concentration of extracted membranes in solvent mixture expressed as number of membranes in ml of solvent mixture (pcs/ml), (d) is duration of deployment of membrane during a sampling (days) and ECxx is an effective concentration of extract on the chosen organism, for example EC50 (ml/l): Vtoxð50Þ ¼ 1=ðm EC50 dÞ For each test, as quality control, we analyzed a MicrotoxTM reference toxicant, phenol, DMSO and a field blank SPMD. 2.3.2. Ames test The mutagenic activity of the SPMDs extracts, dissolved in 10 ml of DMSO, was determined using the Salmonella/microsome assay according to the standard plate method of Maron and Ames (1983). Each sample was evaluated with and without metabolic activation (10% S9 mix), using TA98 and TA100 Salmonella typhimurium strains, which are sensitive to the majority of water mutagens (Albaladejo et al., 1995). Solvent mutagenicity and triolein assays were included to check sample preparation interferences. Sodium azide (SA) and 2-nitrofluorene (2-NF), 1 lg/plate, were used as positive control for TA100 and TA98 without S9, respectively; 2-aminofluorene (2-AF), 2 lg/plate, was used to assess microsomal fraction efficiency. Mutagenic activity was expressed as a mutagenicity ratio (MR), excluding the contribution of the pre-existing revertants (Georghiou et al., 1989). MR ¼
net revertants ðpre-existing revertantsÞ spontaneous revertants ðpre-existing revertantsÞ
1694
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
Samples were considered mutagenic when net revertants values were at least twice as much as the frequency of spontaneous revertants, and when a linear doseresponse relationship, for the four doses employed, was indicated. 2.4. Chemical analysis (GC/MS) Dialysis was performed in hexane on the second SPMD fragment in the dark for 24-h. The solvent was then rinsed and the sample dialysed for further 48 h. The dialysed sample was combined and the volumes were reduced using rotary evaporation. The sample was then re-suspended in 500 ll of hexane. The extract obtained contains a substantial amount of material different from the PAHs, which may interfere with the analytical determination. In order to purify the sample the extracts were loaded on 500 mg SPE silica cartridges, PAHs fraction was collected eluting with cyclohexane. GC analyses of PAHs were accomplished with a Varian 3800 gas-chromatograph coupled with a Varian Saturn 2000 IT mass detector. The GC/MS was operating in the single ion storage (SIS) acquisition mode. The separation of PAHs was performed injecting in split mode 2 ll of purified samples extract into Varian 1079 PTV split/splitless injector. Separation was achieved with a 30 m capillary column. 250 · 0.25 lm Varian Facto Four VF5 Column by means of a thermal gradient from 80 C to 300 C at 10 C/min. The injection port temperature, transfer line temperature and line trap temperature were held isothermally at 280 C, 270 C and 220 C, respectively. Quantification was performed by mean of a multi level internal standard calibration. Internal standard of labelled PAHÕs used for quantification purposes were naphthalene-d8, phenanthrene-d10, chrysene-d12. Three replications for each standard calibration level were performed during analysis in order to obtain instrumental calibration.
3. Results 3.1. Biological assays A segment extract of SPMD per site, at four different dilutions (in triplicate), for each assay, was tested. 3.1.1. MicrotoxTM test Table 2 gives EC50s to determine the potential toxicity of SPMDs extracts, including controls (solvents and field blank SPMD), for the analysis of replicates samples the RSDs were <15%. The surface water was classified as weakly toxic (TU range: 0.03–0.92) for V. fischeri. The drinking water was no toxic except for January, August, September, November and December 2002 when it was weakly toxic (TU range: 0.50–0.87) and only during
Table 2 Mean EC50s of SPMDs extracts and controls, assayed with MicrotoxTM Sample type SPMDs Surface water (input) Drinking water (output) Quality control Field blank SPMDs Triolein DMSO Microtox phenol reference toxicant (lg/ml H2O)
Mean MicrotoxTM toxicity EC50 (ds) 4.92 (7.20) 4.99 (11.92) 23.81 (4.27) nd nd 17.95 (3.84)
nd: none detected.
March 2002 it was classified as toxic (TU = 1). We had taken into account that field blank SPMDs had shown 0.04 ± 0.01 TU. Fig. 1 shows the toxicity data, expressed in TU/mg triolein and in Vtox(50) ml/day obtained for SPMDs samples from surface and drinking water. The mean toxic units were similar for drinking water (TU/ mg triolein = 1.82 ± 2.34) and for surface water (TU/ mg triolein = 1.67 ± 1.55), there is no statistical significative difference. Considering the Vtox(50), the means were similar for drinking water (Vtox(50) ml/day = 1.62 ± 2.09) and for surface water (Vtox(50) ml/day = 1.41 ± 1.31), also for this parameter there is no statistical significative difference. There is a positive and significative correlation between toxic units and Vtox(50) both for surface water (r = 0.967, p < 0.01) and drinking water (r = 0.928, p < 0.01). 3.1.2. Ames test Spontaneous reversion of the tester strains to histidine independence is measured in each Ames test and is expressed as the number of spontaneous revertants per plate; spontaneous reversion for each strain is at a frequency that is characteristic of the strain. The data are expressed as the average revertants number per plate from the triplicates; Table 3 summarizes the means results obtained from the study tests. The results indicate that neither surface water nor drinking water increase the revertants numbers in all two tester strains at all four concentrations tested in comparison to the spontaneous controls. MR values were always below 2, so any samples were considered mutagenic. Any sample was considered mutagenic because the data generated in these assays are acceptable, because the spontaneous revertants number is within the normal range reported in the literatureÕs data, the positive mutagens (both in activated and in non-activated system) yielded a significant increase (2–9 fold increase) in the number of revertants as compared to control. Thus, at the tested concentrations, we did not report the problem of cytotoxicity.
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
1695
Fig. 1. Toxicity data (expressed both in toxic units and in Vtox) obtained for SPMDs samples deployed in surface (INPUT) and in drinking water (OUTPUT).
Table 3 The means results of mutagenicity assays (Ames test) expressed as the average revertants number per plate from triplicates Mean revertants/plate (DS) Sample type Spontaneous SPMDs Water surface 10 ll 20 ll 50 ll 100 ll Drinking water 10 ll 20 ll 50 ll 100 ll Quality control Field blank SPMDs Triolein DMSO (100 ll) SA (1 lg/plate) 2-NF (1 lg/plate) 2-AF (2 lg/plate)
TA98 23 (6)
TA100 131 (14)
TA98 + S9 38 (7)
TA100 + S9 137 (13)
26 28 26 24
(10) (9) (8) (11)
130 125 124 142
(25) (22) (18) (20)
39 35 45 42
(9) (11) (7) (13)
132 145 139 135
(15) (19) (22) (17)
25 28 29 33
(11) (9) (11) (8)
138 148 152 150
(19) (15) (20) (17)
34 33 36 40
(8) (11) (10) (10)
130 140 125 135
(16) (18) (10) (19)
155 120 133 286
(10) (10) (12) (34)
40 (8) 37 (7)
150 (8) 141 (9)
348 (29)
292 (40)
26 (7) 22 (4) 24 (5) 57 (10)
3.2. Chemical analysis Fig. 2 represents both standards and sample processed by assaying one SPMD extract with GC/MS.
The GC/MS analyses were performed on ten samples, three of which were inputs and seven outputs. PHAs concentration showed to be higher in the input samples than outputs (Table 4), where concentrations were about
1696
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
Fig. 2. PAHs analysis: standards (down) and sample (up) processed by assaying one SPMD extract with GC/MS.
Table 4 PAHs concentrations in surface (Input) and in drinking water (Output), LOD and RDS PAHs
LOD
RDS Input (lg/mg triolein)
Output (lg/mg triolein)
Oct-02 Nov-02 Dec-02 Jul-02
Aug-02 Sep-02 Oct-02 Nov-02 Dec-02 Feb-03
Naphthalene 1-Methylnaphthalene 2-Methylnaphthalene 1-Chloranapthalene Acenaphtylene Acenaphthalene Phenanthrene Anthracene Pyrene Phluoranthene Chrysene Benz[a]anthracene Benzo[j]phluoranthene Benzo[k]phluoranthene Benzo[a]pyrene Indeno[1,2,3-cd]pyrene Dibenz[a,h]anthracene Benzo[g,h,i]perylene
<0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±3% ±5% ±5% ±6% ±5% ±8% ±7% ±7%
1.489 2.965 1.816 <0.006 0.212 0.350 0.687 <0.006 0.123 0.170 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.339 1.355 0.775 <0.006 0.142 0.253 0.502 <0.006 0.129 0.153 <0.006 <0.006 <0.006 0.198 <0.003 <0.006 <0.002 <0.002
15.129 25.885 17.682 <0.006 2.677 1.525 27.443 2.983 28.815 38.022 19.335 27.492 15.291 24.033 10.731 <0.006 <0.002 <0.002
0.248 0.216 0.314 <0.006 <0.006 <0.006 0.096 <0.006 0.075 <0.006 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.467 1.109 0.863 <0.006 <0.006 <0.006 0.232 <0.006 0.113 0.031 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.186 0.322 0.326 <0.006 <0.006 0.234 0.116 <0.006 0.073 0.033 <0.006 0.036 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.201 0.165 <0.006 <0.006 <0.006 <0.006 0.082 <0.006 0.037 0.013 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 <0.002
0.190 0.238 <0.006 <0.006 0.094 0.206 0.099 <0.006 0.038 <0.006 <0.006 <0.006 <0.006 <0.006 <0.003 <0.006 <0.002 0.062
0.100 0.076 <0.006 <0.006 <0.006 <0.006 0.130 0.016 0.111 0.101 <0.006 0.075 0.103 0.067 0.042 <0.006 <0.002 <0.002
0.121 0.070 <0.006 <0.006 0.080 <0.006 0.098 0.052 <0.006 <0.006 <0.006 <0.006 <0.006 0.027 <0.003 <0.006 <0.002 <0.002
Total PHAs
<0.006 ±3%
7.811
3.864
257.043
0.948
2.816
1.326
0.498
0.928
0.821
0.449
one or two orders of magnitude lower. In our study, 1methylnaphthalene, 2-methylnaphthalene and naphthalene are the most represented PAH compounds. Fig. 3
shows that there is no relation between acute toxicity and PHAs (SpearmanÕs r = 0.030; p > 0.05). Analytical recoveries for known analytes are determined by adding
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
1697
Fig. 3. Acute toxicity (TU) and PAHs concentration in surface water (black bars) and in drinking water (grey bars) SPMDs extracts.
to the triolein of non-exposed SPMDs appropriate levels of the analytes. Generally, such SPMDs are spiked with chemicals at levels equivalent to the midpoint of the calibration curve. Recoveries are generally >70%, with good precision (<15% RSD). The RSDs for analysis of replicate samples are often <8% (Table 4), thus, these data are confirmed in the literature (Petty et al., 2000a,b). The water temperatures, in the tanks used for the SPMDs samplings, were 14.2 C (±4.6, range 7.3– 23.2 C) for surface water while for drinking water was 14.7 (±4.7, range 7.8–23.7 C). There are not significative correlations between temperature and TUs or PAHs, neither for surface water nor for drinking water, infact for field application of SPMDs, the results imply that temperature is not a key factor that controls uptake rates (Booij et al., 2003).
4. Discussion and conclusions With respect to PAH compounds, data obtained with surface water samples have shown to be higher when compared to those in the literature (McCarthy and Gale, 1999; Louch et al., 2003). On the other hand, to our knowledge, there have been only few reports on such analysis for drinking water. The acute toxicity of the samples is similar in drinking and surface water. Such data were probably ob-
tained because each membrane was used for about thirty days sampling about 150 l and so the toxicity data were referred to this volume of drinking water. Therefore, the presence of by-products in the chlorinated drinking water has probably influenced the toxicity. All input membranes presented a reddish brown colour biofilm: there is the possibility that the uptake of contaminants by SPMDs could have been severely reduced under fouling condition as reported by Richardson et al. (2002). For both drinking and surface water, we always found low acute toxicity cases (range between 0 and 11.6 TU/mg triolein) and a part of these are also due to the SPMD oleic acid as confirmed by Sabaliunas et al. (1999). The absence of correlation between the acute toxicity results and the PAHs concentrations highlights that toxicity could be due to more than one class of pollutants. This finding is in line with literature data (Wang et al., 2003) confirming the great importance of the biological assays to evaluate the effects of complex mixture like water. The results indicate that drinking water treatment process is able to reduce the presence of the evaluated PAHs. Although the effect of single PAHs on Salmonella activated strains is reported (i.e. Bostrom et al., 1998), our data indicate that SPMD extracts are not mutagenic in the Ames test; we had to consider that the highest tested concentration of the extracts corresponds to 10 lg of triolein in which the concentration of total PHAs is low: it ranged from 0.004 to 0.028 lg for drinking
1698
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699
water and from 0.039 to 2.570 lg for surface water (Table 4). SPMDs showed to be an excellent tool in assessing the presence of hazardous compounds, therefore this method could be of wide interest in assessing toxic contaminants in input and output drinking water treatment plants for their capability in determining pollutant sources and relative levels at different locations. Furthermore, the results of this study have highlighted the usefulness of MicrotoxTM toxicity test, in addition to the conventional chemical examination of water quality, to evaluate the concentration of ambient bioavailable chemicals and their possible additive effects. It would be of great interest to carry out additional studies on the evaluation of genotoxic properties by applying another extraction method from the membranes and, furthermore, to determine the ability of the SPMDs to capture by-products present in drinking water. Acknowledgement This research was funded by Societa` Metropolitana Acque Torino s.p.a. References Albaladejo, V., Villanueva, O., Ortega, M., 1995. The evaluation of the mutagenic activity of public drinking water by the Ames test. Rev. Esp. Salud. Publica. 5 (69), 393–408. Booij, K., Hofmans, H.E., Fischer, C.V., Van Weerlee, E.M., 2003. Temperature-dependent uptake rates of nonpolar organic compounds by semipermeable membrane devices and low-density polyethylene membranes. Environ. Sci. Technol. 37 (2), 361–366. Bostrom, E., Engen, S., Eide, I., 1998. Mutagenicity testing of organic extracts of diesel exhaust particles after spiking with polycyclic aromatic hydrocarbons (PAH). Arch. Toxicol. 72 (10), 645–649. Carraro, E., Bugliosi, E.H., Meucci, L., Baiocchi, C., Gilli, G., 2000. Biological drinking water treatment processes, with special reference to mutagenicity. Water Res. 34, 3042–3054. De Marini, D.M., Abu-Shakra, A., Felton, C.F., Patterson, K.S., Shelton, M.L., 1995. Mutation spectra in Salmonella of chlorinated, chloramminated, or ozonated drinking water extracts: comparison to MX. Environ. Mol. Mutagen. 26 (4), 270–285. Georghiou, P.E., Bladgen, P.A., Winsor, L., Williams, D.T., 1989. Spontaneous revertants in modified S. typhimurium mutagenicity test employing elevated numbers of the tester strain. Mut. Res. 225, 33–39. Huckins, J.N., Tubergen, M.W., Manuweera, G.K., 1990. Semipermeable membrane devices containing model lipid: a new approach to monitoring the availability of lipophilic contaminants and estimating their bioconcentration potential. Chemosphere 20, 533–552.
IPSW, 2004. In: International Passive Sampling Workshop and Symposium, Conference Proceedings, 14-18/4/2004, Ceske Budejovice. Kocı´, V., Mlejneky, M., Kocha´nkova´z, L., 2003. Toxicological evaluation of exposed SPMD membranes. Central European Journal of Chemistry 1, 28–34. Kraybill, H.F., 1981. Carcinogenesis of synthetic organic chemicals in drinking water. JAWWA 73, 370–372. Louch, J., Allen, G., Erickson, C., Wilson, G., Schmedding, D., 2003. Interpreting results from field deployments of SPMDs, 2003. Environ. Sci. Technol. 37, 1202–1207. Maron, D.M., Ames, B., 1983. Revised method for the Salmonella mutagenicity test. Mut. Res. 113, 172–215. McCarthy, K.A., Gale, R.W., 1999. Investigation of the distribution of organochlorine and PAH compounds in the lower Columbia river using SPMDs. USGS, Water resource investigations report 99-4051. MICROTOX, 1995. MicrotoxTM acute toxicity basic test. Azur Environmental. Opperhuizen, A., Van der Velde, E.W., Gobas, F.A., Liem, A.K., Van der Stehen, J., Hutzinger, O., 1985. Relationship between bioconcentration in fish and steric factors of hydrophobic chemicals. Chemosphere 14, 1871–1896. Peters, R.J.B., De Leer, E.W.B., De Galan, L., 1990. Dihaloacetonitriles in Dutch drinking water. Water Res. 24, 797– 800. Petty, J.D., Jones, S.B., Huckins, J.N., Cranor, W.L., Parris, J.T., McTague, T.B., Boyle, T.P., 2000a. An approach for assessment of water quality using SPMDs and bioindicator tests. Chemosphere 41, 311–321. Petty, J.D., Orazio, C.E., Huckins, J.N., Gale, R.W., Lebo, J.C., Meadows, J.C., Echols, K.R., Cranor, W.L., 2000b. Consideration involved with use of semipermeable membrane devices for monitoring environmental contaminants. J. Chromatogr. A 879, 83–95. Rehena, Z., Malik, A., Ahmad, M., 1996. Genotoxicity of the Ganges water at Narora (UP), India. Mut. Res. 367, 187– 193. Richardson, B., Lam, P.K.S., Zheng, G., McClellan, K.E., De Luca-Abbott, S., 2002. Biofouling confounds the uptake of trace organic contaminants by SPMDs. Marine Pollut. Bull. 44, 1372–1379. Richardson, B.J., Zheng, G.J., Tse, E., Lam, P.K.S., 2003. A comparison of mussels (Perna viridis) and SPMDs for monitoring chlorinated trace organic contaminants in Hong Kong coastal waters. Chemosphere 45, 1201– 1208. Sabaliunas, D., Ellington, J., Sabaliuniene, I., 1999. Screening bioavailable hydrophobic toxicants in surface waters with SPMDs: role of inherent oleic acid in toxicity evaluations. Ecotoxicol. Environ. Saf. 44, 160–167. Sadiq, R., Rodriguez, M.J., 2004. Disinfection by-products (DBPs) in drinking water and predictive models for their occurrence: a review. Sci. Total Environ. 321, 21–46. Sodergren, A., 1990. Monitoring of persistent, lipophilic pollutants in water and sediment by solvent-filled dialysis membranes. Ecotoxicol. Environ. Saf. 19, 143–149. Verweij, F., Booij, K., Satumalay, K., Van der Molen, N., Van der Oost, R., 2004. Assessment of bioavailable PAH, PCB
G. Gilli et al. / Chemosphere 61 (2005) 1691–1699 and OCP concentrations in water, using semipermeable membrane devices (SPMDs), sediments and caged carp. Chemosphere 54 (11), 1675–1689. Wang, C., Wang, Y., Kiefer, F., Yediler, A., Zijian, W., Kettrup, A., 2003. Ecotoxicological and chemical charac-
1699
terization of selected treatment process effluents of municipal sewage treatment plant. Ecotoxicol. Environ. Saf. 56, 211–217. WHO, 2004. Guidelines for drinking water quality, Third edition.