CellV

Fig. 18.3. Appearance of RTL-W1 cells grown for 4 days on a Biosilon bead. Photo was kindly provided by Dr Michael Gelinsky, Max Bergmann Center of Biomaterials, University of Technology Dresden. [SEM: Zeiss DSM Gemini 982, working distance 9 mm, acceleration voltage 5 kV; the sample was critical point dried and carbon coated].

Fig. 18.3. Appearance of RTL-W1 cells grown for 4 days on a Biosilon bead. Photo was kindly provided by Dr Michael Gelinsky, Max Bergmann Center of Biomaterials, University of Technology Dresden. [SEM: Zeiss DSM Gemini 982, working distance 9 mm, acceleration voltage 5 kV; the sample was critical point dried and carbon coated].

The toximeter was evaluated for PAH sampling and combined toxi-cological and chemical analyses in several laboratory experiments and an extensive, 1-year field study at a contaminated gas works site [8]. Both laboratory experiments in semi-static exposure systems as well as the field study showed the general suitability of the Biosilon-filled toxi-meter for PAH sampling. The toximeter accumulated PAHs with a log Kow between 4.5 and 6 as predicted based on Fick's first law. Thus, it allowed the calculation of time-weighted average aqueous PAH concentrations (see also Chapter 12). The lower predictability for PAHs with log Kow values lower than 4.5 and higher than 6 was assumed to be due to lower binding affinities to Biosilon as receiving phase for the less hydrophobic PAHs (naphthalene, acenaphthene, acenaphthylene and fluorene) and hindrance in passing the ceramic membrane for the higher hydrophobic ones (dibenzo[a,h]anthracene, benzo[ghi]perylene and indeno[1,2,3-cd]pyrene). An example of accumulated masses obtained by the toximeter compared with the masses predicted based on conventional water analysis is depicted in Fig. 18.5.

The field study also confirmed the long-term stability of the toxime-ter sampling device even under extreme conditions. One of the three

Fig. 18.4. Induction of EROD activity elicited in RTL-W1 cells after 24 h of exposure to benzo[k]fluoranthene (BkF) in the bead assay. Biosilon beads were coated with BkF prior to the testing by adding methanolic BkF solutions and allowing the methanol to evaporate. Biosilon with the sorbed BkF was transferred to 96-well plates at 30 mg Biosilon per well. A total of 200 mL of an RTL-W1 cell suspension with 30,000 cells/well was then added to the beads. After 24 h of exposure, EROD induction, measured as resorufin produced per minute per cell, was determined using a fluorescence plate reader according to the method described by Ganassin et al. [13]. BkF concentrations on the X-axis were calculated based on the total amount of BkF sorbed to the beads of one well divided by the volume of the medium in one well. The response curve detected in the bead assay was similar in shape to that typically observed in the standard assay. This similarity is reflected in the bell-shape of the curve as well as in the EC50 values, which were, respectively, about 9 nM (see the figure) in the bead assay and 8 nM (Bopp, unpublished) in the standard assay.

Fig. 18.4. Induction of EROD activity elicited in RTL-W1 cells after 24 h of exposure to benzo[k]fluoranthene (BkF) in the bead assay. Biosilon beads were coated with BkF prior to the testing by adding methanolic BkF solutions and allowing the methanol to evaporate. Biosilon with the sorbed BkF was transferred to 96-well plates at 30 mg Biosilon per well. A total of 200 mL of an RTL-W1 cell suspension with 30,000 cells/well was then added to the beads. After 24 h of exposure, EROD induction, measured as resorufin produced per minute per cell, was determined using a fluorescence plate reader according to the method described by Ganassin et al. [13]. BkF concentrations on the X-axis were calculated based on the total amount of BkF sorbed to the beads of one well divided by the volume of the medium in one well. The response curve detected in the bead assay was similar in shape to that typically observed in the standard assay. This similarity is reflected in the bell-shape of the curve as well as in the EC50 values, which were, respectively, about 9 nM (see the figure) in the bead assay and 8 nM (Bopp, unpublished) in the standard assay.

boreholes investigated for 1 year was filled up to half of its depth with a tar oil phase, which occurred unexpectedly upon constructions at the test site. Despite surface discolorations of toximeters hanging directly in the tar oil phase, sampling behaviour was not impaired. This was similar to the results obtained for ceramic dosimeters [11] (see also Chapter 12), which were simultaneously employed under identical conditions. Toxicological responses elicited by the field-exposed toxi-meters in the bead assay could be explained, in part, by the chemically detected PAHs but indicated the presence of other relevant contaminants not detected in routine chemical analyses [8].

In summary, the toximeter opens new avenues in the application of passive sampling devices. Because of its immediate compatibility with cultured vertebrate cells, a solvent extraction is not required and toxicity of the sampled groundwater contaminants is assessed quickly

with minimal preparation. Rainbow trout liver cells were chosen as reporters of toxicity by PAHs in the first toximeter design because of their applicability to monitoring water quality in general. However, the cell line used could be from any species. Indeed, cell lines could be from any tissue or organ of origin, depending on the toxicological response of interest. Along these lines, we are currently applying the human breast cancer cell line, MCF-7, in order to detect cell proliferation and/or altered gene expression due to estrogenic compounds. Additional research is underway to apply the toximeter to surface- as well as pore-water.

18.2.2 Toxicological analysis of solvent extracts obtained from passive sampling devices

Passive samplers not directly designed for toxicological investigations have rarely been used to study groundwater toxicity. The methodology for using these is to obtain a solvent extract of the groundwater contaminants sampled by the chosen passive sampler and apply this extract to toxicity reporting entities. At least three interrelated aspects need to be considered in this approach. The first is that the solvent used for extraction and chemical analysis may not be compatible with application to a toxicity test. Thus, solvent exchange may comprise a necessary additional step. This may potentially lead to the loss of contaminants, particularly volatile and sparingly soluble compounds. Secondly, the extraction procedure itself may yield toxic compounds. An example was reported by Sabaliunas et al. [14] where oleic acid, a potential impurity of the triolein (the receiving phase material used in semi-permeable membrane devices (SPMDs)), was concentrated during extraction and found to be toxic toward luminescent bacteria. Thirdly, it is necessary to dilute the solvent extract in order to avoid interference with the toxi-cological reporters and/or the conditions of the toxicity tests. This may

Fig. 18.5. Comparison of accumulated amounts in the toximeters and accumulated amounts predicted from aqueous concentrations for each PAH over a 28-day exposure period in a semi-static exposure scenario. Toximeters were exposed in a semi-static system, exchanging exposure solutions every 24 h, in order to counteract sorptive losses of PAHs to test vessels. Squares depict the average accumulated mass from three toximeters. Bars represent the range of accumulated masses predicted from average aqueous concentrations with initial measured concentrations as basis for upper limits and concentrations of exchanged water after 24 h for lower limits. Abbreviations for the PAHs are as described in Ref. [3]. Indeno[1,2,3-cd]pyrene (IndP) concentrations were below the quantification limit for the toximeter samples.

limit the concentration range that can be applied. Despite these limitations, extracts of passive samplers, particularly SPMDs, have been applied successfully in a number of studies in surface water. Yet for groundwater, the potential of this approach has not been widely realized and only two examples can be provided at this point.

The first example is the use of extracts obtained from the ceramic dosimeters, exposed at a former gas works site, which were otherwise used for chemical analysis of PAHs [11]. The extracts, prepared in acetone, were applied to RTL-W1 cells at a 200-fold dilution in order to limit the acetone content to 0.5% per culture well (Bopp and Schirmer, unpublished). These extracts were found to be less potent in eliciting an EROD response when compared with the bead assay of toximeter-derived samples. One likely cause of this was the several-fold lower initial contaminant load in the acetone extract compared with that available in the bead assay. Up-concentration of the acetone extract prior to addition to the cells would have been possible but not to the same level of contaminant load as with the toximeter-derived beads. The second example is the use of extracts from SPMDs applied in groundwater in the industrial area of Bitterfeld (Germany). In addition to chemical analysis [15], SPMD extracts were applied to three organisms and one cell line-based bioassay (Altenburger and Schirmer, UFZ Centre for Environmental Research, Germany, personal communication). The cell line-based test was the induction of CYP1A in the rainbow trout liver cell line, RTL-W1. None of the extracts elicited this response, which led to the conclusion that dioxin-like compounds were not present at the site. The three organism-based toxicological tests comprised the inhibition of luminescence in Vibrio fischeri (the acute Microtox test), the Daphnia magna immobilization test as well as the inhibition of reproduction of the unicellular green alga Scenedesmus vacuolatus. All three tests showed that the samples most toxic were those from the sampling well GWM 19/91, which was the one most directly affected by seepage of spilled chemicals [15]. Toxicity of the extracts declined with increasing distance from this source, which well reflected the detected chemical load. Yet, further analysis revealed that toxicity could only partly be explained by the chemicals analysed. Thus, a combination of the bioassays and chemical fractionation and analysis could be used in the future to identify the chemicals that were sampled by the SPMDs but are not yet known to contribute to toxicity.

Taken together, these examples show that extracts obtained from passive samplers can be used in toxicity tests to investigate the potential of groundwater to be toxic. Combination with chemical analysis enables the identification of possible links between contaminants analysed chemically and the observed toxicity. A particular strength of time-integrative passive samplers is that they accumulate the bioavailable contaminants from groundwater so that they can be detected both chemically as well as biologically even at concentrations that would be too low to be identified with conventional snapshot sampling.

0 0

Post a comment