Experimental Exposure Assessment of Ionizable Organic Chemicals in In Vitro Cell-Based Bioassays.

Exposure assessment in in vitro cell-based bioassays is challenging for ionizable organic chemicals (IOCs), because they are present as more than one chemical species in the bioassay medium. Furthermore, compared to neutral organic chemicals, their binding to medium proteins and lipids is driven by more complex molecular interactions. Total medium concentrations (Ctotal,medium) and/or freely dissolved medium concentrations (Cfree,medium) were determined for one neutral chemical and 14 IOCs (acids, bases, multifunctional) at concentrations relevant for determination of cytotoxicity and effect. Cfree,medium was measured in two in vitro bioassays at the time of dosing and after 24 h of incubation using solid-phase microextraction. Cfree,medium was maximally 1.7 times lower than the nominal concentrations (Cnom) for the hydrophilic chemicals (caffeine and lamotrigine). For the organic acids (naproxen, ibuprofen, warfarin, and diclofenac), Cfree,medium was by a factor of 4 lower than Cnom at high concentrations, but the ratio was much higher at low concentrations, indicating a nonlinear binding behavior. The experimental Cfree,medium was also compared with Cfree,medium predicted with a mass balance model accounting for binding to medium proteins and lipids. The mass balance model performed well for five of the test chemicals (within a factor of 10), but it underestimated Cfree,medium by up to a factor of 1200 for chemicals that showed nonlinear binding to medium components. These findings emphasize that experimental exposure assessment is required for improved understanding of in vitro toxicity data.

Cellular Metabolism in High-Throughput In Vitro Reporter Gene Assays and Implications for the Quantitative In Vitro-In Vivo Extrapolation.

High-throughput in vitro reporter gene assays are increasingly applied to assess the potency of chemicals to alter specific cellular signaling pathways. Genetically modified reporter gene cell lines provide stable readouts of the activation of cellular receptors or transcription factors of interest, but such reporter gene assays have been criticized for not capturing cellular metabolism. We characterized the metabolic activity of the widely applied AREc32 (human breast cancer MCF-7), ARE-bla (human liver cancer HepG2), and GR-bla (human embryonic kidney HEK293) reporter gene cells in the absence and in the presence of benzo[a]pyrene (BaP), an AhR ligand known to upregulate cytochrome P450 in vitro and in vivo. We combined fluorescence microscopy with chemical analysis, real-time PCR, and ethoxyresorufin-O-deethylase activity measurements to track temporal changes in BaP and its metabolites in the cells and surrounding medium over time in relation to the expression and activity of metabolic enzymes. Decreasing BaP concentrations and formation of metabolites agreed with the high basal CYP1 activity of ARE-bla and the strong CYP1A1 mRNA induction in AREc32, whereas BaP concentrations were constant in GR-bla, in which neither metabolites nor CYP1 induction was detected. The study emphasizes that differences in sensitivity between reporter gene assays may be caused not only by different reporter constructs but also by a varying biotransformation rate of the evaluated parent chemical. The basal metabolic capacity of reporter gene cells in the absence of chemicals is not a clear indication because we demonstrated that the metabolic activity can be upregulated by AhR ligands during the assay. The combination of methods presented here is suitable to characterize the metabolic activity of cells in vitro and can improve the interpretation of in vitro reporter gene effect data and extrapolation to in vivo human exposure.

Experimental Validation of Mass Balance Models for in Vitro Cell-Based Bioassays.

The freely dissolved concentration in the assay medium (Cfree) and the total cellular concentration (Ccell) are essential input parameters for quantitative in vitro-to-in vivo extrapolations (QIVIVE), but available prediction tools for Cfree and Ccell have not been sufficiently validated with experimental data. In this study, medium-water distribution ratios (DFBS/w) and cell-water distribution ratios (Dcell/w) for four different cells lines were determined experimentally for 12 neutral and five ionizable chemicals. Literature data for seven organic acids were added to the dataset, leading to 24 chemicals in total. A mass balance model based on bovine serum albumin-water (DBSA/w) and liposome-water distribution ratios (Dlip/w) of the chemicals was used to calculate DFBS/w and Dcell/w. For all neutral and basic test chemicals, the mass balance model predicted DFBS/w and Dcell/w within a factor of 3 and 3.4, respectively, indicating that existing models can reliably predict Cfree and Ccell for these chemicals. For organic acids, a further refinement of the model will be required as large deviations between modeled and measured binding to assay medium and cells of up to a factor of 370 were found. Furthermore, saturation of medium proteins should be further explored for organic acids and neutral chemicals with moderate hydrophobicity.

C18-Coated Solid-Phase Microextraction Fibers for the Quantification of Partitioning of Organic Acids to Proteins, Lipids, and Cells.

The effects measured with in vitro cell-based bioassays are typically reported as nominal effect concentrations ( Cnom), but the freely dissolved concentration in the exposure medium ( Cw) and the total cellular concentration ( Ccell) are considered more quantitative dose metrics that allow extrapolation to the whole-organism level. To predict Cw and Ccell, the partitioning of the test chemicals to medium proteins and lipids and cells has to be known. In this study, we developed a solid-phase microextraction (SPME) method based on C18-coated fibers to quantify the partitioning of diclofenac, 2,4-dichlorophenoxyacetic acid (2,4-D), ibuprofen, naproxen, torasemide, warfarin, and genistein to bovine serum albumin (BSA), phospholipid liposomes, fetal bovine serum (FBS), and cells. For ibuprofen, 2,4-D, naproxen, and warfarin, the partitioning to the SPME fibers was found to be concentration dependent, which had to be considered for the calculation of distribution ratios to biological materials. The sorption isotherms to FBS were nonlinear for diclofenac, 2,4-D, ibuprofen, naproxen, and warfarin. The FBS isotherms could be described by assuming that the total amount of chemical bound to FBS is the sum of the amount specifically bound to the binding sites of albumin and nonspecifically bound to all medium proteins and lipids. The determined cell-water distribution ratios ( Dcell/w) differed considerably between four different cell lines (up to 1.83 log-units) and also between different batches of the same cell line (up to 0.48 log-units). The relative importance of protein and lipid content for Dcell/w was evaluated with a mass balance model and different types of cellular proteins and lipids as input parameters. Existing in vitro mass balance models may underestimate Cw because they do not account for saturable protein binding and overestimate Ccell for organic acids, if BSA is used as surrogate for cellular proteins.

Combined Ion-Trapping and Mass Balance Models To Describe the pH-Dependent Uptake and Toxicity of Acidic and Basic Pharmaceuticals in Zebrafish Embryos ( Danio rerio).

The aim of the current study was to understand and develop models to predict the pH-dependent toxicity of ionizable pharmaceuticals in embryos of the zebrafish Danio rerio. We found a higher uptake and toxicity with increasing neutral fraction of acids (diclofenac, genistein, naproxen, torasemide, and warfarin) and bases (metoprolol and propranolol). Simple mass balance models accounting for the partitioning to lipids and proteins in the zebrafish embryo were found to be suitable to predict the bioconcentration after 96 h of exposure if pH values did not differ much from the internal pH of 7.55. For other pH values, a kinetic ion-trap model for the zebrafish embryo explained the pH dependence of biouptake and toxicity. The total internal lethal concentrations killing 50% of the zebrafish embryos (ILC50) were calculated from the measured BCF and LC50. The resulting ILC50 were independent of external pH. Critical membrane concentrations were deduced by an internal mass balance model, and apart from diclofenac, whose specific toxicity in fish had already been established, all pharmaceuticals were confirmed to act as baseline toxicants in zebrafish.

Quantification of freely dissolved effect concentrations in in vitro cell-based bioassays.

Improved understanding of chemical exposure in in vitro bioassays is required for quantitative in vitro-in vivo extrapolation (QIVIVE). In this study, we quantified freely dissolved concentrations in medium sampled from in vitro cell-based bioassays (Cfree,medium) for nine chemicals with different hydrophobicity and speciation at the time point of dosing and after an incubation period of 24 h using solid-phase microextraction. The chemicals were tested in two reporter gene assays, the AREc32 assay indicative of the oxidative stress response and the PPARγ-GeneBLAzer assay that responds to chemicals which bind to the peroxisome proliferator-activated receptor gamma. For seven of the nine chemicals, Cfree,medium did not change significantly over time in both assays and the experimentally determined Cfree,medium generally agreed well with predictions of a mass balance model that describes the partitioning between proteinaceous and lipidous medium constituents, cells and the aqueous phase. Two chemicals showed a decrease of Cfree,medium in the AREc32 assay over time that was probably caused by cellular metabolism. Furthermore, Cfree,medium of the acidic chemical diclofenac deviated from the model predictions by more than a factor of 10 at higher concentrations, which indicates nonlinear binding and saturation of the medium proteins. Bioassay results are typically reported as nominal effect concentrations (ECnom), although it is established that freely dissolved effect concentrations (ECfree) are a better measure for the bioavailable dose and the method developed here provides a simple experimental approach to measure and model ECfree in in vitro bioassay for improved QIVIVE models.

Trout and Human Plasma Protein Binding of Selected Pharmaceuticals Informs the Fish Plasma Model.

Concerns are increasing that pharmaceuticals released into the environment pose a risk to nontarget organism such as fish. The fish plasma model is a read-across approach that uses human therapeutic blood plasma concentrations for estimating likely effects in fish. However, the fish plasma model neglects differences in plasma protein binding between fish and humans. Because binding data for fish plasma are scarce, the binding of 12 active pharmaceutical ingredients (APIs; acidic, basic, and neutral) to rainbow trout (Oncorhynchus mykiss) and human plasma was measured using solid-phase microextraction (SPME). The plasma/water distribution ratios (D plasma/w ) of neutral and basic APIs were similar for trout and human plasma, differing by no more than a factor of 2.7 for a given API. For the acidic APIs, the D plasma/w values of trout plasma were much lower than for human plasma, by up to a factor of 71 for naproxen. The lower affinity of the acidic APIs to trout plasma compared with human plasma suggests that the bioavailability of these APIs is higher in trout. Read-across approaches like the fish plasma model should account for differences in plasma protein binding to avoid over- or underestimation of effects in fish. For the acidic APIs, the effect ratio of the fish plasma model would increase by a factor of 5 to 60 if the unbound plasma concentrations were used to calculate the effect ratio. Environ Toxicol Chem 2022;41:559-568. © 2020 The Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC.

Optimization of a pre-metabolization procedure using rat liver S9 and cell-extracted S9 in the Ames fluctuation test.

Many environmental pollutants pose a toxicological hazard only after metabolic activation. In vitro bioassays using cell lines or bacteria have often no or reduced metabolic activity, which impedes their use in the risk assessment. To improve the predictive capability of in vitro assays, external metabolization systems like the liver S9 fraction are frequently combined with in vitro toxicity assays. While it is typical for S9 fractions that samples and testing systems are combined in the same exposure system, we propose to separate the metabolism step and toxicity measurement. This allows for a modular combination of metabolic activation by enzymes isolated from rat liver (S9) or a biotechnological alternative (ewoS9R) with in vitro bioassays that lack metabolic capacity. Benzo(a)pyrene and 2-aminoanthracene were used as model compounds to optimize the conditions for the S9 metabolic degradation/activation step. The Ames assay with Salmonella typhimurium strains TA98 and TA100 was applied to validate the set-up of decoupling the S9 activation/metabolism from the bioassay system. S9 protein concentration of 0.25 mgprotein/mL, a supplement of 0.13 mM NADPH and a pre-incubation time of 100 min are recommended for activation of samples prior to dosing them to in vitro bioassays using the regular dosing protocols of the respective bioassay. EwoS9R performed equally well as Moltox S9, which is a step forward in developing true animal-free in vitro bioassays. After pre-incubation with S9 fraction, chemicals induced bacteria revertants in both the TA98 and the TA100 assay as efficiently as the standard Ames assay. The pre-incubation of chemicals with S9 fraction could serve for a wide range of cellular in vitro assays to efficiently combine activation and toxicity measurement, which may greatly facilitate the application of these assays for chemical hazard assessment and monitoring of environmental samples.

Thermal proteome profiling allows quantitative assessment of interactions between tetrachloroethene reductive dehalogenase and trichloroethene.

Thermal proteome profiling (TPP) is increasingly applied in eukaryotes to investigate protein-ligand binding through protein melting curve shifts induced by the presence of a ligand. In anaerobic bacteria, identification of protein-substrate interactions is a major challenge. We applied TPP to Sulfurospirillum multivorans, which is able to use trichloroethene as electron acceptor for growth, to investigate the interaction of its tetrachloroethene reductive dehalogenase PceA with trichloroethene. Several modifications in the protocol (e.g., incubation under anaerobic conditions; increasing the temperature range up to 97 °C) extended the protein detection range and allowed the investigation of oxygen-sensitive proteins. Enzymatic reductive dehalogenation was prevented by omitting the electron donor during incubations. This enabled detecting the interaction of PceA with trichloroethene and confirmed that trichloroethene is a substrate of this enzyme. Interestingly, a putative response regulator showed a similar trend, which is the first biochemical hint for its proposed role in trichloroethene respiration. We proved that our TPP approach facilitates the identification of protein-substrate interactions of strictly anaerobic reductive dehalogenases and probably their regulators. This strategy can be used to identify yet unknown substrate specificities and possible signal-sensing proteins, and therefore has the potential to elucidate one of the unresolved fields in research on organohalide-respiring bacteria. SIGNIFICANCE: The assessment of enzyme-substrate or protein-ligand interactions in organohalide-respiring bacteria is a fundamental challenge. Thermal proteome profiling (TPP) allows elucidating proteome-wide thermal stability changes relying on the sensitivity of modern mass spectrometry. This gives access to the identification of interactions not detectable with other methods. In this TPP study, we demonstrate the interactions of a chlorinated substrate with a reductive dehalogenase and potentially with a response regulator, thereby supporting the response regulator's function in organohalide respiration. The strategy might also be applied to identify yet unknown substrates of other enzymes in bacteria which are difficult to investigate or for which only low amounts of biomass are available. The assessment of enzyme-substrate interactions, which might enable conclusions about enzyme specificities, represents a new application for TPP.