March 9, 2019
ScitoVation will be attending the 2019 Society of Toxicology (SoT) conference in Baltimore on March 10-14. Visit us at booth 3218 in the exhibit hall, or see one of our presentations.
ScitoVation will be co-presenting six different presentations and posters this year, addressing topics from transcriptomics to cell-based assays.
Presentations include:
Zonal-specific transcriptional programs associated with PPAR activation in the rat liver and their role in liver cancer in rodents
Abstract Number / Poster Number: 1985 / P372
Session Title: Carcinogenesis II
Presentation Date/Time: March 12, 2019; 9:15am – 4:30pm; author attended time block: 9:15 – 10:45am
Previously, we compared whole genome gene expression following exposures to multiple doses of GW7647 – a ligand with high selectivity for PPARa – in rat and human hepatocytes and in the intact rat liver. Using ChIP methods, we found that the PPARa peroxisomal proliferator response element (PPREa) binding sites in human and rat hepatocytes differed qualitatively for down regulated genes in the rat. In addition, responses in the intact rat were more diverse than in rat hepatocytes with many more gene families significantly enriched in the intact liver, including pathways for cell proliferation which were absent in vitro. Here, using laser capture microdissection, we examined differential responses in periportal (PP), midzonal (MZ) and centrilobular (CL) hepatocytes using 6 doses from 0.1 to 20 mg GW/kg/d for 5 consecutive days plus a control. Gene expression analysis of PP and CL subpopulations of cells showed that there were many more genes differentially expressed in PP than in CL cells (2837 versus 432 at 20 mg/kg/d). The magnitude of the transcriptional changes was larger and the dose at which the cells responded was lower (AC50 = 0.38 versus 1.94 mg/kg/day) in the PP compared to the CL cells. Analysis of canonical PPARa target genes and pathway enrichment showed that all three zones responded with alterations in fatty acid metabolic machinery. However, there were a number of other processes – including cell cycle, apoptosis, cell-cell adhesion, citric acid cycle, respiratory electron transport, and ATP production – that were exclusively affected in PP cells. Differential PPARa binding to up and down regulated genes in rat and human hepatocytes appears to have a role in rodent-human differences in the capacity of a PPARa to cause proliferation, while the differential periportal responses affected by PPARa agonists may be a key step in cell proliferation and carcinogenesis associated with PPARa agonists in rodents.
Extending the PLETHEM platform for PBPK modeling: batch mode processing, dermal routes of exposure, and integration with mode-of-action tools
Abstract Number / Poster Number: 1838 / P219
Session Title: Biological Modeling
Presentation Date/Time: March 12, 2019; 9:15am – 4:30pm; author attended time block: 10:45am – 12:15pm
The EPA Office of Research and Development’s 2003 framework for computational toxicology emphasized the need for computational methods to bridge the source-to-outcome continuum. This goal can be achieved by linking exposure estimation methods, physiologically based pharmacokinetic (PBPK) modeling, and computational systems biology pathway modeling tools into a standardized framework. Over the last year we implemented support for exposure estimation tools under our PLETHEM framework. We now have further extended the modeling capabilities in PLETHEM by adding a batch mode for running multiple chemicals at once and incorporating a dermal route of exposure. Using HT-IVIVE workflows within PLETHEM we can now extrapolate in vitro point of departures (POD) to estimate steady state plasma concentrations and associated complementary in vivo doses. The steady-state concentrations can also be used for route-to-route extrapolation of exposures. Gene expression analysis, informed by dictionaries of cellular processes, is used to establish a mode of action for the chemical of interest. We have developed another tool (MoAviz) that makes use of interactive visualizations and machine learning to better interpret results from these analyses. As a first step towards extending PLETHEM into the domain of systems biology, we are incorporating support for results obtained by MoAviz. The PODs for different modes of action can be directly imported within PLETHEM. The HT-IVIVE workflow can then be used to extrapolate these PODs to exposures and establish a margin of exposure using exposure estimation tools. Integrating support for MoAviz, exposure estimation tools and multiple routes of exposure will meet the need for a unified computational tool that can be used along the source-to-outcome continuum. This research is funded by the ACC-LRI and is conducted under an MoU with USEPA NCCT.
Development of a multi-functional fit-for-purpose rat liver co-culture assay for hepatotoxicity testing
Abstract Number / Poster Number: 2446 / P861
Session Title: Alternatives to Mammalian Models I
Presentation Date/Time: March 12, 2019; 9:15am – 4:30pm; author attended time block: 10:45am-12:15pm
Extensive in vivo experiments for toxicity testing are cost-prohibitive and not always predictive of human responses. In both the regulatory and the industrial arenas, the goal is to move away from in-life rodent studies and towards safety assessment strategies that rely on testing species-relevant cells in vitro. The liver has been a major focus of these efforts, yet there are currently no in vitro alternatives for hepatotoxicity testing accepted by regulators, and the assays that do exist typically utilize hepatocyte monolayer culture or focus on a single phenotypic endpoint. While hepatocytes have been the primary component of in vitro toxicity assay development, it has become increasingly clear that the non-parenchymal cells (NPCs) (i.e., hepatic stellate cells, Kupffer cells, and liver sinusoidal endothelial cells) play a critical role in the progression of liver pathologies. The goal of this study was to develop a multi-functional organotypic co-culture system, in which primary rat hepatocytes are cultured in the presence of NPCs. We have been optimizing conditions for co-culture of hepatocytes with NPCs in 2D monolayers. We found that health of hepatocytes and liver sinusoidal endothelial cells, as indicated by marker expression and morphology, could be substantially improved by culture in a blend of William’s E and endothelial growth media vs. William’s E medium alone. We also examined the effect of altering the ratio of hepatocytes to NPCs, and obtained good results using a ratio of 36:12. We have begun to extend this co-culture system into three dimensions by encapsulating the hepatocytes and NPCs into alginate beads. Whereas the 2D cultures typically expire within 7 days, the 3D cultures consistently remain viable to at least 21 days. We are beginning to investigate the potential for these long-lived cultures in long term chemical safety studies.
Development of a fit-for-purpose in vitro model of lung toxicity
Abstract Number / Poster Number: 2465 / P880
Session Title: Alternatives to Mammalian Models II
Presentation Date/Time: March 12, 2019; 9:15am – 4:30pm; author attended time block: 1:30 – 3:00pm
A shift in toxicity testing from in vivo animal models to in vitro human cell-based models is a major focus of current research, due to the high cost, low throughput, ethical concerns, and questionable human relevance of in vivo testing. While many high throughput in vitro assays for biological perturbations do exist (as illustrated by the ToxCast effort), it is clear that more physiologically relevant cellular models are required to reproduce adverse outcomes. The lung epithelium is a frequent target of chemicals under regulation for toxic effects, yet efforts to develop in vitro lung models fit for the purpose of toxicity testing have been relatively few. To reproduce such adverse effects as proliferation and fibrosis, it will be necessary to utilize cellular models of the lung that incorporate multiple differentiated cell types growing at the air-liquid interface (ALI). Furthermore, to ensure interpretable dosimetry, exposure of cells to test compounds must be via the air. We are exploring testing approaches that meet these criteria. As a proof-of-principle experiment, we have treated primary human bronchial-tracheal epithelial cells (HBTECs), grown at ALI, with chloroform vapor using the Vitrocell inhalation exposure system. Chloroform is metabolized by CYP2E1 to form reactive intermediates, including phosgene, which react with glutathione (GSH). Metabolically competent cells are expected in deplete GSH through conjugation as well as oxidation to GSSG; therefore, we chose to monitor oxidized and total GSH after exposure. To identify exposures likely to span the point-of-departure for this endpoint, we first performed a range-finding experiment with A549 human alveolar carcinoma cells treated with chloroform in solution for 24 hours. 1 μM had little effect on total and oxidized GSH levels compared to vehicle, whereas 25 uM greatly increased oxidation. Chloroform concentrations bracketing this POD (2.5, 10, and 40 uM) were selected and converted to equivalent 1 h vapor-based exposures, (69, 267, and 1069 ppm, respectively), by matching the nominal AUCs of the media and vapor-based exposure using a blood:air partition coefficient. HBTECs grown at ALI were then treated with the chosen concentrations of chloroform vapor for 1 or 3 hours. At both timepoints, GSH levels were significantly reduced, and GSSG:GSH ratios were increased, at 1000 ppm chloroform compared to clean air treatment. This result demonstrates the promise of this system for in vitro inhalation safety testing.
Addressing systematic inconsistencies between in vitro and in vivo transcriptomic mode of action signatures
Abstract Number / Poster Number: 1803 / P182
Session Title: Bioinformatics
Presentation Date/Time: March 12, 2019; 9:15am – 4:30pm; author attended time block 1:30-3:00pm
Because of their broad biological coverage and increasing affordability transcriptomic technologies have increased our ability to evaluate cellular response to chemical stressors, providing a potential means of evaluating chemical response without requiring traditional long-term studies with apical endpoints. Dose-response modeling of transcriptomic data is being rapidly incorporated into risk assessment frameworks as a means of approximating points of departure from short-term in-life studies. However, identification of mode of action from transcriptomics lacks a similar systematic framework. To this end, we developed a web-based interactive browser—MoAviz—that allows visualization of perturbed pathways. We populated this browser with expression data from a large public toxicogenomic database (TG-GATEs), providing the groundwork for performing “biological read-across” between compounds based on their transcriptomic fingerprints. We evaluated the extent to which gene expression changes from in-life exposures could be associated with mode of action by developing a novel similarity index—the Modified Jaccard Index (MJI)—that provides a quantitative description of genomic pathway similarity (rather than gene level comparison). While typical compound-compound similarity is low (MJI = 0.026), clustering of the TG-GATES compounds identifies groups of similar chemistries. Some clusters aggregated compounds with known similar modes of action, including PPARa agonists (MJI = 0.330) and NSAIDs (MJI = 0.327). Analysis of paired in vitro (hepatocyte)-in vivo (liver) experiments revealed systematic patterns in the responses of model systems to chemical stress. Accounting for these model- specific—but chemical-independent—differences improved pathway concordance by 36% between in vivo and in vitro models. This approach, which leverages recent developments in computational approaches and cell culture and sequencing technologies, will improve early decision-making about chemical safety.
Fit-for-purpose in vitro assays in risk-based decision making—example with estrogenic compounds
Abstract Number / Poster Number: 1868 / P251
Session Title: Risk Assessment I
Presentation Date/Time: March 12, 2019; 9:15am – 4:30pm; author attended time block: 3:00 – 4:30pm
Chemical risk assessment relies on expensive in vivo studies, and the current scale of animal testing is insufficient to address chemical safety concerns as regulatory agencies require more complete toxicity data. We demonstrate a practical tiered testing approach to reduce animal use by employing multiple levels of risk prioritization. A fit-for- purpose uterine estrogen assay was coupled with in silico and high-throughput in vitro to in vivo extrapolation (HT- IVIVE) to improve prioritization of estrogenic compounds and support in vitro hazard characterization in a tissue- specific context. Initial prioritization was based on activity in estrogen receptor (ER) dependent ToxCast assays. To assess uterine-specific reactivity, 112 ToxCast-prioritized compounds were screened in the uterine estrogen response assay for ER-mediated proliferation in Ishikawa cells, a human uterine cell line. Estimates of activity (AC50s) were based on dose-dependent proliferation, and quantitative in vitro-to-in vivo extrapolation (Q-IVIVE) was used to predict a human equivalent dose (HED) for each compound. Estrogen response assay-derived HEDs were compared with in vivo points of departure from existing uterotrophic assay and two-generation reproductive toxicity studies. Pearson correlation (in log-space) between in vitro HED and in vivo points of departure was 0.47. Estrogen response assay-derived HEDs were lower than in vivo points of departure in 37 (93%) of 40 compounds active in both models. Of 56 compounds active in in vivo studies, 16 (29%) were inactive in the estrogen response assay, possibly due to differences in metabolism in the two test systems. Our positive control, the synthetic estrogen 17alpha-ethinylestradiol, produced mean HEDs of 0.56, 0.030, and 0.0030 μg/kg-day in the ToxCast assays, estrogen response assay, and in vivo studies, respectively. These results support the use of fit-for-purpose assays to refine chemical risk prioritization based on the estrogen response assay’s range of detection and characterization of a potent ER agonist, which are improved over ToxCast assays. Developing well-accepted and more efficient in vitro assays and IVIVE methods is crucial for the proper interpretation of in vitro hazard data for health risk assessment.