In vitro bioassays for monitoring drinking water quality of tap water, domestic filtration and bottled water

Background Location-specific patterns of regulated and non-regulated disinfection byproducts (DBPs) were detected in tap water samples of the Barcelona Metropolitan Area. However, it remains unclear if the detected DBPs together with undetected DPBs and organic micropollutants can lead to mixture effects in drinking water. Objective To evaluate the neurotoxicity, oxidative stress response and cytotoxicity of 42 tap water samples, 6 treated with activated carbon filters, 5 with reverse osmosis and 9 bottled waters. To compare the measured effects of the extracts with the mixture effects predicted from the detected concentrations and the relative effect potencies of the detected DBPs using the mixture model of concentration addition. Methods Mixtures of organic chemicals in water samples were enriched by solid phase extraction and tested for cytotoxicity and neurite outgrowth inhibition in the neuronal cell line SH-SY5Y and for cytotoxicity and oxidative stress response in the AREc32 assay. Results Unenriched water did not trigger neurotoxicity or cytotoxicity. After up to 500-fold enrichment, few extracts showed cytotoxicity. Disinfected water showed low neurotoxicity at 20- to 300-fold enrichment and oxidative stress response at 8- to 140-fold enrichment. Non-regulated non-volatile DBPs, particularly (brominated) haloacetonitriles dominated the predicted mixture effects of the detected chemicals and predicted effects agreed with the measured effects. By hierarchical clustering we identified strong geographical patterns in the types of DPBs and their association with effects. Activated carbon filters did not show a consistent reduction of effects but domestic reverse osmosis filters decreased the effect to that of bottled water. Impact statement Bioassays are an important complement to chemical analysis of disinfection by-products (DBPs) in drinking water. Comparison of the measured oxidative stress response and mixture effects predicted from the detected chemicals and their relative effect potencies allowed the identification of the forcing agents for the mixture effects, which differed by location but were mainly non-regulated DBPs. This study demonstrates the relevance of non-regulated DBPs from a toxicological perspective. In vitro bioassays, in particular reporter gene assays for oxidative stress response that integrate different reactive toxicity pathways including genotoxicity, may therefore serve as sum parameters for drinking water quality assessment.


Table of Content (in order of occurrence)
Table S1.Physical parameters of the water (data reprinted from Redondo-Hasselerharm et al. [1]. Table S2.Analytical results (data reprinted from Redondo-Hasselerharm et al. [1]).
Text S1.Additional information on the experimental method for direct neurotoxicity measurement with unenriched water.Table S3.Inhibitory concentration IC10 for 10 % reduction of cell viability in the AREc32 cell line and effect concentration ECIR1.5 for activation of oxidative stress response of the water samples.
Text S2. Results of the direct neurotoxicity measurement with unenriched water.Table S4.Inhibitory concentration IC10 for 10 % reduction of cell viability in the neuronal cell line SH-SY5Y and effect concentration EC10 for reduction of neurite length by 10% for selected samples.Table S5.Effect concentrations ECIR1.5 and cytotoxicity IC10 for the detected chemicals in AREc32.

Text S1. Additional information on the experimental method for direct neurotoxicity measurement with unenriched water
Aliquots of the sampled water were filtered, aliquoted and frozen until their use.Cell death was assessed by the uptake of the fluorescent exclusion dye propidium iodide.This method relies upon the fact that an intact nuclear membrane has low permeability for propidium iodide.However, when cell integrity becomes compromised, it gains access to the nucleus where it complexes with DNA rendering the nucleus highly fluorescent.SH-SY5Y cells (2x10 4 cells/well) on 48-well plates were incubated with tap water samples media for 24 h, then stained with propidium iodide (PI, Sigma, St. Louis, MO, USA) solution at final concentration of 20 µg/mL for 5 min in the dark.Immediately thereafter microscope images were obtained using a fluorescence microscopy Olympus BX61 (Olympus America Inc., NY, USA) with excitation 540 nm and emission 620 nm.Cell death was determined as the percentage of red fluorescent cells per total cells counted.
To determine the effect of tap water samples on neuronal morphological differentiation of SH-SY5Y cells, tap water samples media was supplemented with 1% PS, 1% FBS and with 10 µM all-trans retinoic acid (Sigma Aldrich).After incubation for 3 days, the cells were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 3 min and washed with PBS.
The morphological changes in the cells were observed under a phase-contrast microscope Olympus BX61 (Olympus America Inc., NY, USA).Briefly, the neurite length was determined by measuring the total length of the outgrowing neurites per soma using Image J software (v. 1.53p).Those cells whose cell body diameters longer than twice of the diameter of cell body were considered as neurite-bearing cells.The average of neurites length was determined by counting at least 200 cells in each sample.
Statistical analysis was performed using the software Statistical Package for the Social Sciences (SPSS v.27).The homogeneity of variances was analysed with Levene's test.If variances were homogenous, ANOVA was then used followed by Bonferroni's test in order to analyse all dose groups simultaneously.The Kruskal-Wallis test was used when variances were not homogeneous.Significance was set at p < 0.05.Cell viability assay did not show significant (p>0.05)changes on treated cells with tap water samples media over 24 h, respect to control (Fig. S3).In addition, tap water samples media treatment also showed no significance differences in LDH leakage after 24 h of exposition, respect to control (Fig. S3).These results suggest that contaminants present in tap water samples had not high enough concentration to cause effect on cell viability and cytotoxicity on SH-SY5Y cells.Staining with PI did not show significant (p>0.05)increase of membrane damage or apoptotic cells on treated cells with tap water samples media over 24 h, respect to control (Fig. S2A).
Only treated cells with 15 % of DMSO (positive control) showed a significant diminution (p < 0.05) respect to control.These results suggest that DBPs and other contaminants present in tap water samples were not sufficiently concentrations induce apoptotic cell death or damage in membrane on SH-SY5Y cells.
Differentiation of SH-SY5Y with 10 µM RA in presence of tap water samples media did not show any significant (p>0.05)differences on average of neurite length after 72 h of exposition, respect to control (Fig. 2SB).These results suggest that contaminants present in tap water samples had not enough potency to disrupt neuronal differentiation of SH-SY5Y cells and the growing neurite length.SH-SY5Y cells were stained with PI (left y-axis) after 24 h of exposition to tap water samples media.The cells were also differentiated (right y-axis) with 10 µM retinoic acid in tap water samples media over 72 h.The percentage of the positive cells stain, respect to the total cells, and the average of neurite length data are expressed as the mean ± SD of four independent experiments.Significant differences relative to the control (DEMEM/F12) were analyzed by one-way ANOVA followed by Bonferroni's post hoc test: *p < 0.05 versus control.

Figure S1 .
Figure S1.pH shift towards lower pH appears to be related to the sum concentrations of haloacetic acids (HAA).

Figure S5 .
Figure S5.Concentration-response curves in the neurotoxicity assay dosed with enriched water extracts.

Figure S6 .
Figure S6.Effect of acidity of some of the samples in the neurotoxicity assay.

Figure S7 .
Figure S7.Comparison of the effect concentrations of tap water effect data of this study with literature of drinking water and other water types.

Figure S8 .
Figure S8.Comparison of cytotoxicity, oxidative stress response and total organic carbon (TOC) of the tap water.

Figure S10 .
Figure S10.Hierarchical clustering after scaling of measured effects.

Figure S1 .Figure S2 .
Figure S1.pH shift towards lower pH appears to be related to the sum of molar concentrations of haloacetic acids (HAAs).

Figure S3 .
Figure S3.Cytotoxicity of SH-SY5Y dosed directly with water.SH-SY5Y cell viability quantified with MTT (left y-axis) and LDH leakage (right y-axis) measured after 24h of exposure to tap water samples media.Cell viability and LDH leakage data are expressed as the mean ± SD of four independent experiments.Significant differences relative to the control (medium DEMEM/F12) were analyzed by one-way ANOVA followed by Bonferroni's post hoc test: *p < 0.05 versus control.

Figure S6 .
Figure S6.Effect of acidity of some of the samples in the neurotoxicity assay.Phase-contrast images of the differentiated SH-SY5Y cells show the cell bodies as dark grey areas and the neurites as pink lines connecting the cells in (a) unexposed cells, (b) cells exposed to tap water 08019, (c) artefacts observed in tap water 08008 due to cell debris on the background.

Figure S7 .
Figure S7.Comparison of the effect concentrations of tap water effect data of this study with literature of drinking water and other water types.(a) Oxidative stress response detected with AREc32 compared with drinking water from literature (Hebert et al. 2018 [3] and Neale et al. 2020 [4]), surface water during rain events [5] and wastewater treatment plant effluent (WWTP) [5].(a) EC10 in the neurite outgrowth inhibition assay compared with EC10 of surface water during rain events [5] and wastewater treatment plant effluent (WWTP) [5].

Figure S8 .
Figure S8.Comparison of cytotoxicity, oxidative stress response and total organic carbon (TOC) of the tap water; (a) Oxidative stress response expresses as DBAN-EQbio compared to cytotoxicity toxic units TUbio; (b) TUbio compared with TOC; (c) DBAN-EQbio compared to TOC.

Table S2 .
[1]lytical results: concentrations of DBPs in units of µg/L (data reprinted from Redondo-Hasselerharm et al.[1]).MBAA, transferred to a new 96-well assay plate in triplicate wells.Then, 50 μL of reaction mixture solution was added to each well.Plates were incubated with gentle shaking on an orbital shaker for 30 min at room temperature and 50 µL of stop solution was added to each sample well.The absorbance was read at 490 nm and 680 nm using a microplate reader (BioTek, Power Wave XS, USA).To determine LDH activity, the 680-nm absorbance value (background) was subtracted from the 490-nm absorbance before calculation of % cytotoxicity using the spontaneous LDH activity of the solvent control and the maximum LDH activity of the positive control in relation to the LDH activity of the water sample: to induce their differentiation.DMEM/F12 media served as negative control (Gibco BRL, Life Technologies, Paisley, UK).The human SH-SY5Y neuroblastoma cell line was purchased from American Type Culture Collection (Manassas, VA, USA).Cells were maintained in complete culture medium consisting of Dulbecco's modified Eagle's medium (DMEM)/F12 pH 7.4 (Gibco BRL, Life Technologies, Paisley, UK) supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin/streptomycin (PS, Gibco BRL, Life Technologies, Paisley, UK).The cells were maintained in a humidified incubator with 5% CO 2 and 95% air at 37°C.The medium was refreshed every 2-3 days.Cell viability was quantified using the metabolic dye 3-[4,5-dimethylthiazol-2-yl]−2,5diphenyltetrazolium bromide (MTT, Sigma, St. Louis, MO, USA).When cultures were confluent (70-80%), cells were seeded in 48-well tissue culture plates at density of 2×10 4 cells/well in complete medium.After 24 h of cell attachment, cells were exposed to tap water samples media for 24 h.MTT was added to the cells 3 h prior to the end of the experiment at a final concentration of 0.5 mg mL −1 .During that time, MTT was reduced to produce a dark blue formazan product.Formazan production was measured after removal of the medium and extraction with DMSO by the change in absorbance at 560 nm using a microplate reader (BioTek Power Wave XS).The viability results were expressed as a percentage of control.37°C.The cells were incubated with tap water samples media for 24 hours.To determinate maximum LDH activity, 20 µL of 10X Lysis buffer was added to no treated well for 45 min before elapsed time.Plates were then centrifuged at 400 x g for five minutes and 50 μL of cell supernatant was

Table S4 .
Inhibitory concentration IC10 for 10 % reduction of cell viability in the neuronal cell line SH-SY5Y and effect concentration EC10 for reduction of neurite length by 10% for selected samples.Artefacts due to cell debris observed in neurite outgrowth measurement at REF 150 and 300; no cell debris observed below REF 150. b