Unlike many cell lines, HepaRG™ cells are able to regulate glycogenolysis and gluconeogenesis at comparable rates to primary human hepatocytes. Therefore, they represent a relevant model with which to investigate the regulation of carbohydrate pathways and how they are altered in associated metabolic diseases. In HepaRG™ cultures, the presence of glycogen, visualised using periodic acid staining (PAS), can be detected in the hepatocytes but not in surrounding cholangiocytes (Figure 1).
In addition to other regulatory pathways, a novel regulatory mechanism modulating glucose and insulin sensitivity has recently been shown (Caron et al., Mol Cell Biol, 2013). This pathway involves the farnesoid X receptor (FXR, a nuclear bile acid receptor) which negatively regulates the expression of several glucose-regulated genes.
Another important endogenous pathway in the liver is lipid metabolism. The responsiveness to stimuli that alter lipid pathways is lost in most cell lines; however, HepaRG™ cells have been shown to retain response mechanisms involving FXR, PPAR and LXR, as well as associated lipid metabolising enzymes. For example, Madec et al. (DMD, 1997) demonstrated that the lipid metabolising enzyme, CYP4F3B, and ω-hydroxylation of fatty acids are equivalent in HepaRG™ cells and human liver. Likewise, Samanez et al. (Arch Physiol Biochem, 2012) analysed the expression and regulation of genes involved in lipid (and glucose) metabolism and showed that the lipogenic pathway could be regulated by altering the glucose concentration.
Therefore, HepaRG™ cells can be used to study the interlinking pathways involved in carbohydrate homeostasis and lipid metabolism.
The hepatotoxicity of many compounds is due to their metabolism to reactive metabolite(s), rather than the parent compound per se. As such, any cell type intended for hepatic-specific assays should ideally have a metabolic enzyme profile as close to primary human hepatocytes as possible. Figure 2 shows that metabolically competent HepaRG™ cells are highly sensitive to the CYP3A4-mediated cytotoxicity of aflatoxin B1. By contrast, there is a lack of sensitivity of HepG2 cells to this potent hepatotoxic compound.
The ease of adaptation of HepaRG™ cells to multi-well formats, high throughput screening involving single measurements or high content screening, as well as lower throughput assays for more investigative research into toxicity pathways, makes them ideal for both early and mechanistic drug development.
HepaRG™ cells can be used to predict chronic disease states in patients treated with drugs, such as steatosis resulting from the accumulation of triglycerides. For some steatotic drugs, the pathology is detected only after chronic dosing and even primary human hepatocytes would not be an appropriate model to identify this effect due to their limited lifespan in culture and the fact that they tend to accumulate triglycerides spontaneously when they age, even if they are cultured in a sandwich format. By contrast, HepaRG™ cells can be treated for up to 4 weeks in order to capture the chronic effects of compounds. In addition, specific genes involved in metabolising saturated and poly-unsaturated fats (CYP4F3B and CYP4A11) are markedly higher in HepaRG™ than in other cell types e.g. HepG2, Huh7, making them a suitable model for investigating lipid metabolism.
An example of how steatosis is visualised using Oil red staining to detect lipids is shown in Figure 3. There is a moderate amount of lipid accumulation after 24 h of treatment with amiodarone; however, a severe chronic effect is only evident after 14 days. The pathologies caused by this compound have been associated with the up-regulation of related genes e.g. lipid droplet formation is related to the genes, ADFP and PLIN4, thus this toxicity can be monitored visually and at the gene level.
Using a high throughput metabolomics approach, Brown et al. (Obesity, 2014) showed that HepaRG™ cells were able to mimic the distinct metabolic signatures including amongst others, an elevated level of lipid metabolites that is also observed in non-alcoholic fatty liver disease (NAFLD) patients.
In addition to detecting steatotic drugs, HepaRG™ cells can also be used to identify potential drug candidates for their anti-steatotic effects. A recent report demonstrated how HepaRG™ cells, pre-treated with oleic acid to induce steatosis, were used to investigate the time-dependent effects of PPAR agonists on lipid levels and metabolism (Rogue et al., TAP, 2014). Co-treatment with PPAR agonists decreased lipid vesicles by up to 50%, while fatty acid oxidation was induced after 2 weeks. The reversal of steatosis demonstrated in HepaRG™ cells was reflected in-vivo clinical findings, thus supporting their use as a predictive and sensitive model for steatosis.
Phospholipidosis (accumulation of phospholipids and formation of lamellar bodies) is not itself overtly toxic but is indicative of accumulation of the drug and/or its metabolites, which in turn are responsible for the ensuing liver injury. It is therefore important to be able to predict phospholipidosis, especially as a result of chronically administered drugs. HepaRG™ cells have been shown to accumulate phospholipids after short-term treatment with amiodarone (Anthérieu et al., Hepatology, 2011 and TIV, 2012).
The appearance of phospholipids is an acute effect evident after only short-term exposure but it also accumulates over time (Figure 4). The development of phospholipidosis was concomitant with an up-regulation of genes responsible for lipid synthesis and the formation of vesicles.
Cholestatic drugs can be identified in HepaRG™ cells by measuring their effects on bile acid efflux transporters (via direct inhibition or down-regulation) and bile canaliculi structure. An example of an effect of a drug on the morphology of the bile canaliculus that can be captured and quantified is shown in Figure 5 and in the time lapse video below. Chlorpromazine causes constriction of the canalicular lumen, thus resulting in cholestasis.
In addition to the direct actions of cholestatic compounds, the influence of proinflammatory cytokines such as IL-6 and IL-1β, thought to be involved in drug-induced-liver injury (DILI), can be assessed. For example, cytotoxic and cholestatic effects induced by the idiosyncratic drug, chlorpromazine, is increased in HepaRG™ cells. The combination of the suppression of CYPs caused by cytokines, together with increased cholestatic and cytotoxic impact of compounds could be important when interpreting in-vivo effects in the presence of high inflammatory state (Bachour-El Azzi et al. DMD, 2014).
HepaRG™ cells have been applied to a number of in-vitro genotoxicity and carcinogenicity assays. For genotoxicity, the current in-vitro battery (Ames and micronucleus test (MNT)) results in a high number of positive outcomes which are negative in in-vivo studies i.e. “false positives”. One way in which the specificity (correctly identifying a true negative) of the MNT has been improved is to use the most appropriate cell type (Fowler et al., Mutat Res, 2012). HepaRG™ cells naturally lend themselves to the MNT because they are metabolically competent and thus bioactivate pro-genotoxicants; and can proliferate, which is needed to detect micronuclei (MN). For this reason, the IWGT workshop considered HepaRG™ cells to be a “promising” cell type for use in the MNT (Pfuhler et al., Mutat Res, 2011). By contrast, other cell types used, although proliferating, are metabolically incompetent and require exogenous metabolic system (S9). An advantage of HepaRG™ cells is that they can be used for repeat treatments over 7 days to increase their sensitivity to metabolically activated genotoxins such as benzo[a]pyrene (BaP), aflatoxin B1 (AFB1) and 2-nitrofluorene (2NF), compared to a single one-day treatment (Figure 6, Jossé et al., Mutagenesis, 2012). The longer treatment duration allows for the formation of genotoxic metabolites from the pro-genotoxins. At the same, the formation of micro-mononucleated (MNMN) in cells treated with the non-genotoxic non-carcinogenic chemical, nocodazole, and the non-genotoxic carcinogen, dichlorodiphenyldichloroethylene (DDE) was unaffected by the longer repeat treatment (and therefore the high specificity of the assay is retained, Figure 6).
A recent comprehensive investigation (Le Hegarat et al., Tox Sci 2014) of the use of HepaRG™ cells in the MNT and the Comet assay (which detects DNA single strand breaks) using a panel of 16 chemicals from the list recommended by ECVAM showed that these cells were able to detect direct and bioactivated genotoxins with a high sensitivity (ability to detect true positives), specificity and accuracy compared to those reported by Kirkland et al. (Mut Res 2005) for standard in-vitro test assays, which included the use of rodent cell lines such as CHO and V79 cells.
The carcinogenic potential of 10 prototypical non-carcinogens and 20 hepatocarcinogens has been studied in an EU FP6 project called “carcinoGENOMICS”. The method involved the use of a combination of toxicogenomics and in-vitro models, including HepaRG™ cells. The findings from this project revealed that HepaRG™ cells were a better model for carcinogenesis than rat hepatocytes, hepG2 cells, as well as hESC derived hepatocytes. There were a number of gene signatures identified that were specific to either genotoxic or non-genotoxic carcinogens and were able to correctly identify genotoxins as a result of the resulting classification model (Doktorova et al. EXCLI Journal 2014). In addition to pathways related to DNA damage, independent biostatistical approaches were used to identify a number of genes and signalling pathways regulated by p53 that are altered in HepaRG™ cells and primary human hepatocytes by carcinogens.
These studies therefore support the use of whole genome expression analysis of HepaRG™ cells to discriminate between genotoxic and non-genotoxic compounds and provide information on compound- and time-dependent effects on cell cycle and apoptosis signalling pathways (Jossé et al., 2012).
There are a number of other mechanisms of toxicity that have been investigated using HepaRG™ cells, although the extent of research in these areas is not as comprehensive as for the endpoints described above. These include:
Many compounds causing liver toxicity (DILI) do so by causing mitochondrial dysfunction. Therefore, there is an increasing need to identify such compounds early in drug development. The fluorescent dye, JC-1, has been successfully used as to detect mitochondrial membrane depolarization associated with permeability transition (MPT) pore opening in mitochondria after treatment of HepaRG™ cells with acetaminophen (McGill et al., Hepatology, 2011). Moreover, this method has been adapted to an automated, image-based, in situ toxicity test that can efficiently screen the hepatotoxicity of multiple chemicals (Pernelle et al., TAP, 2011).
Mitologics S.A.S. is specialized in detection of mitochondrial alterations and uses HepaRG™ differentiated cells to evaluate long term and/or metabolites mitochondrial toxicity as cells can be treated almost during two weeks.
Read Mitolocigs's presentation on the assessment of drug-induced mitochrondrial toxicity on HepaRG™ cells >>
HepaRG™ cells undergo apoptosis, which can occur via a number of pathways. Gene profiles altered in HepaRG™ cells in response to toxic compounds, such as phenobarbital, indicate many of these pathways are present (Lambert et al., TIV, 2009), although the JNK activation pathway is not activated in these cells (Xie et al., TAP, 2014). HepaRG™ cells have been used to screen for hepatocellular carcinoma-selective cytotoxins, thus providing a key tool to select and develop therapies for this common cancer (Cuconati et al., PLOSone, 2013).
Figure 1. Staining of glycogen in hepatocytes in HepaRG™ cultures using periodic acid staining (PAS).
Figure 2. Cytotoxicity of aflatoxin B1 after 24 h incubation in HepaRG™ (•) and HepG2 cells (▴).
Adapted from Aninat et al., DMD 2006.
Figure 3. Lipid accumulation in HepaRG™ cells after acute (24 h (A)) and chronic (14 days (B)) treatment with amiodarone.
Taken from Anthérieu et al., Hepatology, 2011.
|(A) 24 h treatment
||(B) 14 days treatment
Figure 4. Appearance of lamellar bodies (LB) and lipid droplets (LD) in HepaRG™ cells, visualised by electron microscopy.
HepaRG™ cells were treated with vehicle control (A) and 20 µM amiodarone for 14 days (B, C). (BC=bile canaliculus; N=nucleus).
Adapted from Anthérieu et al., Hepatology, 2011.
Figure 5. Effect of the cholestatic drug, chlorpromazine, on the structure of bile canaliculi in HepaRG™ cells.
Arrows indicate the canaliculi.
Taken from Anthérieu et al., Hepatology, 2013
||(B) 2 h after chlorpromazine treatment (>40μM)
Figure 6. Increased sensitivity of HepaRG™ cells to bioactivated genotoxins using repeat dosing over 7 days (black bars).
For bioactivated genotoxins only, the number of MNMN cells was increased compared to that in cells treated for only one day (grey bars).
Taken from Jossé et al., Mutagenesis, 2012.
The use of HepaRG™ cells to investigate the influence of inflammation (cytokines) on the toxicity of compounds is described above with respect to cholestasis. Another aspect of inflammation - sepsis - has also been investigated in which lipopolysaccharide (LPS) and the catecholamines, epinephrine and norepinephrine, have been shown to provoke a down-regulation of CYP3A4 and a strong increase in levels of protein reactive C in primary human hepatocytes and HepaRG™ cells by stimulating IL-6 production (Aninat et al., Crit Care med, 2008). This is of importance because these catecholamines are naturally released in response to septic shock and are also administered as treatment to re-establish blood pressure. Therefore, HepaRG™ cells may be used to investigate mechanisms involved in sepsis and the choice of suitable drugs for its treatment.