Validation of reference gene stability for APAP hepatotoxicity studies in different in vitro systems and identification of novel potential toxicity biomarkers

Liver cell lines and primary hepatocytes are becoming increasingly valuable for in vitro toxicogenomic studies, with RT-qPCR enabling the analysis of gene expression profiles following exposure to potential hepatotoxicants. Supporting the accurate normalisation of RT-qPCR data requires the identification of reference genes which have stable expression during in vitro toxicology studies. Therefore, we performed a comprehensive analysis of reference gene stability in two routinely used cell types, (HepG2 cells and primary rat hepatocytes), and two in vitro culture systems, (2D monolayer and 3D scaffolds). A robust reference gene validation strategy was performed, consisting of geNorm analysis, to test for pair wise variation in gene expression, and statistical analysis using analysis of variance. This strategy identified stable reference genes with respect to acetaminophen treatment and time in HepG2 cells (GAPDH and PPIA), and with respect to acetaminophen treatment and culture condition in primary hepatocytes (18S rRNA and α-tubulin). Following the selection of reference genes, the novel target genes E2F7 and IL-11RA were identified as potential toxicity biomarkers for acetaminophen treatment. We conclude that accurate quantification of gene expression requires the use of a validated normalisation strategy for each species and experimental system employed.     

Advantages of in vitro cytotoxicity testing by using primary rat hepatocytes in comparison with established cell lines

We investigated and compared the cytotoxicity of 16 reference compounds in four in vitro systems: primary cultured rat hepatocytes, hepatoma HepG2 cell line, non-hepatic HeLa and Balb/c 3T3 cell lines. After 24 hr of exposure to the test compounds, the water-soluble tetrazolium salts WST-1 assay was used as an endpoint to evaluate cytotoxicity. Acetaminophen, diclofenac sodium cyclophosphamide and disulfiram displayed from 2 to more than 10 times higher IC50 values in three cell lines than in rat primary cultured hepatocytes. The cytotoxic effects of aspirin, amiodarone, clorfibiric acid, chlorpromazine, erythomycin, lithocholic acid, cisplatin and quinidine in rat hepatocytes were similar or 2 times stronger than those observed in cell lines. Ketoconazole resulted in the lowest IC50 value in the HeLa cell line. The data suggested that the compounds which are known to be metabolism-mediated liver toxicants have a differential hepatotoxicity in vitro and that primary cultured rat hepatocytes could represent a valuable tool for both screening and study of the effects of bio-transformation on the cytotoxicity of new chemical entities and xenobiotics in vitro.     

Immortalized hepatocytes as in vitro model systems for toxicity testing: the comparative toxicity of menadione in immortalized cells, primary cultures of hepatocytes and HTC hepatoma cells

Immortalized differentiated liver cell lines capable of continuous proliferation, and expressing stable liver-specific functions, would be valuable for in vitro toxicity testing in the pharmaceutical, chemical, food and cosmetics industries. Immortalized rat hepatocyte cell lines have been produced by transfection of SV40 DNA by electroporation or calcium phosphate precipitation. Their utility has been assessed by studying the toxicity of a model compound, menadione, and by measuring the activities of DT-diaphorase and NADPH cytochrome c reductase. Enzyme activities and toxicity were compared in freshly isolated hepatocytes, the immortalized cell lines and dedifferentiated HTC hepatoma cells. In HTC cells DT-diaphorase activity was 100-fold elevated compared with freshly isolated hepatocytes. In only one cell line, C2.1.2. (produced by calcium phosphate precipitation), was DT-diaphorase activity increased (twofold) compared with freshly isolated hepatocytes. Menadione caused loss of viability at similar concentrations (40–80 μ ) in the immortalized cell lines and 24-hr primary cultures of hepatocytes, whereas HTC cells showed loss of viability only with menadione concentrations above 200 μ . The immortalized lines therefore appear to have potential for predicting toxicity and for menadione this can be correlated with the expression of DT-diaphorase.     

Species‐specific differences in peroxisome proliferation,catalase, and SOD2 upregulation as well as toxicity in human,mouse, and rat hepatoma cells induced by the explosive and environmental pollutant 2,4,6‐trinitrotoluene

2,4,6‐Trinitrotoluene (TNT) has been widely used as an explosive substance and its toxicity is still of interest as it persisted in polluted areas. TNT is metabolized in hepatocytes which are prone to its toxicity. Since analysis of the human liver or hepatocytes is restricted due to ethical reasons, we investigated the effects of TNT on cell viability, reactive oxygen species (ROS) production, peroxisome proliferation, and antioxidative enzymes in human (HepG2), mouse (Hepa 1‐6), and rat (H4IIEC3) hepatoma cell lines. Under control conditions, hepatoma cells of all three species were highly comparable exhibiting identical proliferation rates and distribution of their cell cycle phases. However, we found strong differences in TNT toxicity with the lowest IC₅₀ values (highest cell death rate) for rat cells, whereas human and mouse cells were three to sevenfold less sensitive. Moreover, a strong decrease in cellular dehydrogenase activity (MTT assay) and increased ROS levels were noted. TNT caused peroxisome proliferation with rat hepatoma cells being most responsive followed by those from mouse and human. Under control conditions, rat cells contained fivefold higher peroxisomal catalase and mitochondrial SOD2 activities and a twofold higher capacity to reduce MTT than human and mouse cells. TNT treatment caused an increase in catalase and SOD2 mRNA and protein levels in human and mouse, but not in rat cells. Similarly, human and mouse cells upregulated SOD2 activity, whereas rat cells failed therein. We conclude that TNT induced oxidative stress, peroxisome proliferation and mitochondrial damage which are highest in rat cells rendering them most susceptible toward TNT. © 2016 Wiley Periodicals, Inc. Environ Toxicol 32: 989–1006, 2017.     

Involvement of enniatins-induced cytotoxicity in human HepG2 cells

Enniatins (ENNs) are mycotoxins found in Fusarium fungi and they appear in nature as mixtures of cyclic depsipeptides. The ability to form ionophores in the cell membrane is related to their cytotoxicity. Changes in ion distribution between inner and outer phases of the mitochondria affect to their metabolism, proton gradient, and chemiosmotic coupling, so a mitochondrial toxicity analysis of enniatins is highly recommended because they host the homeostasis required for cellular survival. Two ENNs, ENN A and ENN B on hepatocarcinoma cells (HepG2) at 1.5 and 3 μM and three exposure times (24, 48 and 72 h) were studied. Flow cytometry was used to examine their effects on cell proliferation, to characterize at which phase of the cell cycle progression the cells were blocked and to study the role of the mitochondrial in ENNs-induced apoptosis. In conclusion, apoptosis induction on HepG2 cells allowed to compare cytotoxic effects caused by both ENNs, A and B. It is reported the possible mechanism observed in MMP changes, cell cycle analysis and apoptosis/necrosis, identifying ENN B more toxic than ENN A.     

Comparative analysis of eight cytotoxicity assays evaluated within the ACuteTox Project

A comparative analysis of eight cytotoxicity assays [the 3T3 and normal human keratinocytes Neutral Red Uptake (NRU) assay, the primary rat hepatocytes, human HepG2 and 3T3 MTT assay, and the human A.704, SH-SY5Y and HepG2 cells propidium iodide (PI) assay] included in several work packages of the EU Integrated Project ACuteTox, has been carried out. The aim was to evaluate whether cells originating from liver, kidney and brain provided different in vitro acute toxicity results, and the influence of primary liver cells versus cell lines originated from the same tissue. Spearman rank correlation analysis and Hierarchical Cluster Analysis were performed based on the IC₅₀ (50% inhibitory concentrations for the endpoint measured) values generated for 57 chemicals. A relatively large number of neurotoxicants and hepatotoxicants were included which allowed to examine the impact of chemicals with specific tissue toxicity on the results. Our analyses confirmed the similarity between the NRU assays and between the two hepatic cell systems related MTT assays. The type of assay appears to have the greatest influence upon the clustering result regardless of the origin of the cells used. The information provided by the NRU and MTT assays differed from that provided by the PI assay. This approach did not allow to show tissue specific toxicity but it does reveal the effectiveness of the clustering methodology for choosing assays for a testing program for predicting e.g. acute oral toxicity in humans.     

The Fusarium mycotoxins enniatins and beauvericin cause mitochondrial dysfunction by affecting the mitochondrial volume regulation,oxidative phosphorylation and ion homeostasis

The mechanisms of cell toxicity of mycotoxins of the enniatin family produced by Fusarium sp. enniatin B, a mixture of enniatin homologues (3% A, 20% A₁, 19% B, 54% B1) and beauvericin, were investigated. In isolated rat liver mitochondria, exposure to submicromolar concentrations of the enniatin mycotoxins depleted the mitochondrial transmembrane potential, uncoupled oxidative phosphorylation, induced mitochondrial swelling and decreased calcium retention capacity of the mitochondria. The mitochondrial effects were strongly connected with the potassium (K⁺) ionophoric activity of the enniatins. The observed enniatins induced K⁺ uptake by mitochondria. This shows that the enniatins acted as ionophores highly selective for potassium ions. The effects were observed in potassium containing media whereas less or no effect remained to be observed when K⁺ was partially or totally replaced by isomolar concentrations of Na⁺. The rank order of enniatin induced mitochondrial impairment was beauvericin > enniatin mixture > enniatin B. Exposure to the enniatins depleted the mitochondrial membrane potential also in intact human neural (Paju), murine insulinoma (Min-6) cells as well as boar spermatozoa. Exposure to enniatin B in media with physiological (4 mM) or low (<1 mM) but not in high (60 mM) external concentration of K⁺ induced hyperpolarization of the spermatozoal plasma membrane indicating enniatin that catalysed efflux of the cytosolic K⁺ ions. These results indicate that the cellular toxicity targets of the enniatin mycotoxins are the mitochondrion and the homeostasis of potassium ions.     

Use of the γH2AX assay for assessing the genotoxicity of polycyclic aromatic hydrocarbons in human cell lines

The development of in vitro genotoxic assays as an alternative method to animal experimentation is of growing interest in the context of the implementation of new regulations on chemicals. However, extrapolation of toxicity data from in vitro systems to in vivo models is hampered by the fact that in vitro systems vary in their capability to metabolize chemicals, and that biotransformation can greatly influence the experimental results. Therefore, much attention has to be paid to the cellular models used and experimental conditions. Polycyclic aromatic hydrocarbons (PAHs) are carcinogenic ubiquitous pollutants. Human exposure to PAHs is mainly from food origin. In this study, a detailed analysis of the biotransformation capabilities of three human cell lines commonly used for in vitro testing (HepG2, ACHN and Caco-2) was undertaken using 3 model PAHs (benzo(a)pyrene [B(a)P], fluoranthene [FLA] and 3-methylcholanthrene [3-MC]). Concomitantly the genotoxicity of these PAHs was investigated in different cell lines, using a new genotoxic assay (H2AX) in 96-well plates. The metabolic rates of B(a)P, FLA and 3-MC were similar in HepG2 and Caco-2 cell lines, respectively, though with the production of different metabolites. The ACHN cell line was shown to express very limited metabolic capabilities. We demonstrated that the PAHs having a high metabolic rate (B(a)P and 3-MC) were genotoxic from 10⁻⁷ molar in both HepG2 and Caco-2 cells. The present study shows that H2AX measurement in human cell lines competent for the metabolism, is an efficient and sensitive genotoxic assay requiring less cells and time than other currently available tests.     

Assessment of α-amanitin toxicity and effects of silibinin and penicillin in different in vitro models

Silibinin (Sil) is used as hepatoprotective drug and is approved for therapeutic use in amanitin poisoning. In our study we compared Sil-bis-succinate (Sil_BS), a water-soluble drug approved for i.v.-administration, with Sil solved in ethanol (Sil_EtOH), which is normally used in research.We challenged monocultures or 3D-microtissues consisting of HepG2 cells or primary hepatocytes with α-amanitin and treated with SIL_BS, SIL_EtOH, penicillin and combinations thereof. Cell viability and the integrity of the microtissues was monitored. Finally, the expression of the transporters OATP1B1 and B3 was analyzed by qRT-PCR.We demonstrated that primary hepatocytes were more sensitive to α-amanitin compared to HepG2. Primary hepatocytes cultures were protected by Sil_BS and Sil_EtOH independent of penicillin from the cytotoxic effects of α-amanitin. Subsequent studies of the expression profile of the transporters OATP1B1/B3 revealed that primary hepatocytes do express both whereas in HepG2 cells they were hardly detectable.Our study showed that Sil_BS has significant advantage over Sil_EtOH with no additional benefit of penicillin. Moreover, HepG2 cells may not represent an appropriate model to investigate Amanita phalloides poisoning in vitro with focus on OATP transporters since these cells are lacking sensitivity towards α-amanitin probably due to missing cytotoxicity-associated transporters suggesting that primary hepatocytes should be preferred in this context.     

Cylindrospermopsin effects on protein profile of HepG2 cells

Human hepatoma cells (HepG2) were exposed to purified cylindrospermopsin (CYN), a potent toxicant for eukaryotic cells produced by several cyanobacteria. Exposure to 10?μg?l^?1 of CYN for 24?h resulted in alteration of expression of 48 proteins, from which 26 were identified through mass spectrometry. Exposure to 100?μg?l^?1 of CYN for 24?h affected nuclear area and actin filaments intensity, which can be associated with cell proliferation and toxicity. The proteins are implicated in different biological processes: protein folding, xenobiotic efflux, antioxidant defense, energy metabolism and cell anabolism, cell signaling, tumorigenic potential, and cytoskeleton structure. Protein profile indicates that CYN exposure may lead to alteration of glucose metabolism that can be associated with the supply of useful energy to cells respond to chemical stress and proliferate. Increase of G protein-coupled receptors (GPCRs), heterogeneous nuclear ribonucleoproteins (hnRNP), and reactive oxygen species (ROS) levels observed in HepG2 cells can associate with cell proliferation and resistance. Increase of MRP3 and glutathione peroxidase can protect cells against some chemicals and ROS. CYN exposure also led to alteration of the expression of cytoskeleton proteins, which may be associated with cell proliferation and toxicity.     

Genomics and proteomics analysis of cultured primary rat hepatocytes.

The use of animal models in pharmaceutical research is a costly and sometimes misleading method of generating toxicity data and hence predicting human safety. Therefore, in vitro test systems, such as primary rat hepatocytes, and the developing genomics and proteomics technologies, are playing an increasingly important role in toxicological research. Gene and protein expression analysis were investigated in a time series (up to 5 days) of primary rat hepatocytes cultured on collagen coated dishes. Especially after 24h, a significant down-regulation of many important Phase I and Phase II enzymes (e.g., cytochrome P450's, glutathione-S-transferases, sulfotransferases) involved in xenobiotic metabolism, and antioxidative enzymes (e.g., catalase, superoxide dismutase, glutathione peroxidase) was observed. Acute-phase-response enzymes were frequently up-regulated (e.g., LPS binding protein, alpha-2-macro-globulin, ferritin, serine proteinase inhibitor B, haptoglobin), which is likely to be a result of cellular stress caused by the cell isolation procedure (perfusion) itself. A parallel observation was the increased expression of several structural genes (e.g., beta-actin, alpha-tubulin, vimentin), possibly caused by other proliferating cell types in the culture, such as fibroblasts or alternatively by hepatocyte dedifferentiation. In conclusion, the careful interpretation of data derived from this in vitro system indicates that primary hepatocytes can be successfully used for short-term toxicity studies up to 24h. However, culturing conditions need to be further optimized to reduce the massive changes of gene and protein expression of long-term cultured hepatocytes to allow practical applications as a long-term toxicity test system.     

Primary culture of rat hepatocytes in 96-well plates: effects of extracellular matrix configuration on cytochrome P450 enzyme activity and inducibility, and its application in in vitro cytotoxicity screening.

Basal level enzyme activities and enzyme inducibility were compared for rat hepatocytes that were cultured in 96-well plates with three different extracellular matrix configurations: single layer (SL) collagen type I, SL Matrigel, and collagen/Matrigel (C/M) sandwich. Overall, C/M sandwich and SL Matrigel plates were both superior to SL collagen type I plates in maintaining enzyme activities and inducibility and C/M sandwich plates had higher induced activity for CYP3A enzymes than SL Matrigel plates did. Cytotoxicity of nine reference compounds to rat hepatocytes (C/M sandwich configuration), rat hepatoma H4IIE and mouse fibroblast Balb/c 3T3 (3T3) cells was evaluated in 96-well plates using neutral red uptake (for 3T3) and tetrazolium salt MTS assays (for H4IIE and rat hepatocytes). For compounds chlorpromazine, quinidine, trichlorfon, thiopental, and antipyrine, the absolute differences in cytotoxicity LogIC(50) values obtained from different cell types were relatively small and without an obvious trend. The DeltaLogIC(50) values between cultured hepatocytes and the cell lines were much larger for acetaminophen and cyclophosphamide (1.35 < or =/DeltaLogIC(50)/ < or = 3.40), and for clofibrate and thioacetamide (not cytotoxic in hepatocytes at their highest dose levels). These large differences were likely the result of metabolism of these compounds in rat hepatocytes. The relationship between in vitro cytotoxicity LogIC(50) values and in vivo mouse or rat oral acute LogLD(50) values showed that compared to the cell lines, cultured rat hepatocytes improved correlation for acetaminophen and cyclophosphamide. The potential benefit of conducting in vitro cytotoxicity screening using a combination of permanent cell lines and cultured hepatocytes would allow us to obtain mechanistic insight on bioactivation, as well as improve the predictability of metabolism-mediated toxicity.     

Involvement of endoplasmic reticulum and autophagy in microcystin-LR toxicity in Vero-E6 and HepG2 cell lines

This work investigates the involvement of the endoplasmic reticulum (ER) and autophagy in microcystin-LR (MCLR) toxicity in Vero-E6 and HepG2 cell lines. Additionally, morphological alterations induced by MCLR in lysosomes and mitochondria were studied. Cytotoxicity evaluation showed that pure MCLR and MCLR from LMECYA110 extract induce concentration dependent viability decays after 24 h exposure. HepG2 cells showed an increased sensitivity to MCLR than Vero cells, with lower cytotoxic thresholds and EC₅₀ values. Conversely, LC3B immunofluorescence showed that autophagy is triggered in both cell lines as a survival response to low MCLR concentrations. Furthermore, MCLR induced a MCLR concentration-dependent decrease of GRP94 expression in HepG2 cells while in Vero cells no alteration was observed. This suggests the involvement of the ER in HepG2 apoptosis elicited by MCLR, while in Vero cells ER destructuration could be a consequence of cytoskeleton inflicted damages. Additionally, in both cell lines, lysosomal destabilization preceded mitochondrial impairment which occurred at high toxin concentrations. Although not an early cellular target of MCLR, mitochondria appears to serve as central mediators of different signaling pathways elicited by the organelles involved in MCLR toxicity. As a result, kidney and hepatic cell lines exhibit cell type and dose-dependent mechanisms to overcome MCLR toxicity.