Development of a two-layer transwell co-culture model for the in vitro investigation of pyrrolizidine alkaloid-induced hepatic sinusoidal damage

Regulatory role of oxidative stress in retrorsine - Induced apoptosis and autophagy in primary rat hepatocytes

Pyrrolizidine alkaloids (PAs) are a group of naturally occurring alkaloids widely present in plants. PAs are highly hepatotoxic and have been documented to cause many incidents of human and animal poisoning. Retrorsine (RTS) is a pyrrolizidine alkaloid (PA) derived from the Compositae Senecio, which has been shown to cause hepatotoxicity. Human liver poisoning occurs through the consumption of RTS-contaminated food, and animals can also be poisoned by ingesting RTS-containing toxic plants. The mechanism of RTS-induced liver toxicity is not fully understood. In this study, we demonstrated that RTS-induced oxidative stress plays a pivotal role in RTS-induced liver toxicity involving apoptosis and autophagy. The results showed that RTS treatment in the cultured Primary rat hepatocytes caused cytotoxicity and release of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) in a time- and dose-dependent manner. Our study showed that treatment of RTS induced ROS and MDA (malondialdehyde, a lipid peroxidation marker) in the hepatocytes, and reduced antioxidant capacity (GSH content, SOD activity), suggesting RTS treatment caused oxidative stress response in the hepatocytes. Furthermore, we found that RTS induced apoptosis and autophagy in the hepatocytes, and RTS-induced apoptosis and autophagy could be alleviated by ROS scavenger N-acetylcysteine (NAC) and the MAPK pathway inhibitors suggesting ROS/MAPK signaling pathway plays a role in RTS induced apoptosis and autophagy. Collectively, this study reveals the regulatory mechanism of oxidative stress in RTS-induced apoptosis and autophagy in the hepatocytes, providing important insights of molecular mechanisms of hepatotoxicity induced by RTS and related pyrrolizidine alkaloids in liver. This mechanism provides a basis for the prevention and treatment of PA poisoning in humans and animals.

Utilization of a novel human hepatocyte-endothelial cell coculture model to determine differential toxicities of pyrrolizidine alkaloid food contaminants

Pyrrolizidine alkaloids (PA) are comprised of a family of hundreds of metabolites, produced by plants as a mechanism to protect against herbivory. Upon ingestion and metabolism, dehydropyrrolizidine alkaloids are formed, which are known to generate DNA adducts and subsequently double-strand DNA breaks. Within the liver, the most sensitive cell type to PA exposure is the sinusoidal endothelial cell, as evidenced by the generation of veno-occlusive disease in the human population. PAs are a common crop contaminant and have been regulated by some agencies, using the precautionary principle; each equally potent and genotoxic. Therefore, as a proof of principle we have established a human in vitro coculture model system, utilizing the metabolically active HepaRG hepatocyte and the SK-Hep-1 endothelial cell, to determine differential potencies of different PAs commonly found in crops and food products, notably cell death, targeting of endothelial cells, and genotoxicity comparing the micronucleus assay versus γH2AX assay. Our results demonstrate differential potencies of the PAs used, which encompass three esterification states (monoester, cyclic diester, and open-chain diester). The results suggest that a more nuanced approach to the regulation of PAs may be more appropriate in the regulatory decision-making process.

Three-Dimensional Cell Co-Culture Liver Models and Their Applications in Pharmaceutical Research

As the primary site for the biotransformation of drugs, the liver is the most focused on organ type in pharmaceutical research. However, despite being widely used in pharmaceutical research, animal models have inherent species differences, while two-dimensional (2D) liver cell monocultures or co-cultures and three-dimensional (3D) liver cell monoculture in vitro liver models do not sufficiently represent the complexity of the human liver’s structure and function, making the evaluation results from these tools less reliable. Therefore, there is a pressing need to develop more representative in vitro liver models for pharmaceutical research. Fortunately, an exciting new development in recent years has been the emergence of 3D liver cell co-culture models. These models hold great promise as in vitro pharmaceutical research tools, because they can reproduce liver structure and function more practically. This review begins by explaining the structure and main cell composition of the liver, before introducing the potential advantages of 3D cell co-culture liver models for pharmaceutical research. We also discuss the main sources of hepatocytes and the 3D cell co-culture methods used in constructing these models. In addition, we explore the applications of 3D cell co-culture liver models with different functional states and suggest prospects for their further development.

The native liver as inspiration to create superior in vitro hepatic models

Drug induced liver injury (DILI) is one of the major reasons of drug withdrawal during the different phases of drug development. The later in the drug development a drug is discovered to be toxic, the higher the economical as well as the ethical impact will be. In vitro models for early detection of drug liver toxicity are under constant development, however to date a superior model of the liver is still lacking. Ideally, a highly reliable model should be established to maintain the different hepatic cell functionalities to the greatest extent possible, during a period of time long enough to allow for tracking of the toxicity of compounds. In the case of DILI, toxicity can appear even after months of exposure. To reach this goal, an in vitro model should be developed that mimics the in vivo liver environment, function and response to external stimuli. The different approaches for the development of liver models currently used in the field of tissue engineering will be described in this review. Combining different technologies, leading to optimal materials, cells and 3D-constructs will ultimately lead to an ideal superior model that fully recapitulates the liver.

Genotoxicity of pyrrolizidine alkaloids in metabolically inactive human cervical cancer HeLa cells co-cultured with human hepatoma HepG2 cells

Pyrrolizidine alkaloids (PAs) are secondary plant metabolites, which can be found as contaminant in various foods and herbal products. Several PAs can cause hepatotoxicity and liver cancer via damaging hepatic sinusoidal endothelial cells (HSECs) after hepatic metabolization. HSECs themselves do not express the required metabolic enzymes for activation of PAs. Here we applied a co-culture model to mimic the in vivo hepatic environment and to study PA-induced effects on not metabolically active neighbour cells. In this co-culture model, bioactivation of PA was enabled by metabolically capable human hepatoma cells HepG2, which excrete the toxic and mutagenic pyrrole metabolites. The human cervical epithelial HeLa cells tagged with H2B-GFP were utilized as non-metabolically active neighbours because they can be identified easily based on their green fluorescence in the co-culture. The PAs europine, riddelliine and lasiocarpine induced micronuclei in HepG2 cells, and in HeLa H2B-GFP cells co-cultured with HepG2 cells, but not in HeLa H2B-GFP cells cultured alone. Metabolic inhibition of cytochrome P450 enzymes with ketoconazole abrogated micronucleus formation. The efflux transporter inhibitors verapamil and benzbromarone reduced micronucleus formation in the co-culture model. Furthermore, mitotic disturbances as an additional genotoxic mechanism of action were observed in HepG2 cells and in HeLa H2B-GFP cells co-cultured with HepG2 cells, but not in HeLa H2B-GFP cells cultured alone. Overall, we were able to show that PAs were activated by HepG2 cells and the metabolites induced genomic damage in co-cultured HeLa cells.

A TMT-based shotgun proteomics uncovers overexpression of thrombospondin 1 as a contributor in pyrrolizidine alkaloid-induced hepatic sinusoidal obstruction syndrome

Hepatic sinusoidal obstruction disease (HSOS) is a rare but life-threatening vascular liver disease. However, its underlying mechanism and molecular changes in HSOS are largely unknown, thus greatly hindering the development of its effective treatment. Hepatic sinusoidal endothelial cells (HSECs) are the primary and essential target for HSOS. A tandem mass tag-based shotgun proteomics study was performed using primary cultured HSECs from mice with HSOS induced by senecionine, a representative toxic pyrrolizidine alkaloid (PA). Dynamic changes in proteome were found at the initial period of damage and the essential role of thrombospondin 1 (TSP1) was highlighted in PA-induced HSOS. TSP1 over-expression was further confirmed in human HSECs and liver samples from patients with PA-induced HSOS. LSKL peptide, a known TSP1 inhibitor, protected mice from senecionine-induced HSOS. In addition, TSP1 was found to be covalently modified by dehydropyrrolizidine alkaloids in human HSECs and mouse livers upon senecionine treatment, thus to form the pyrrole-protein adduct. These findings provide useful information on early changes in HSECs upon PA treatment and uncover TSP1 overexpression as a contributor in PA-induced HSOS.

Constructing vascularized hepatic tissue by cell-assembled viscous tissue sedimentation method and its application for vascular toxicity assessment

In vitro Construction of the liver sinusoidal structure using artificial tissue is an important but worthwhile challenge, particularly for assessing the risk of diseases such as sinusoidal obstruction syndrome (SOS). Current models are unsuitable for evaluating the toxicity because of lacking sinusoidal capillary. In this study, we developed a vascularized hepatic tissue (VHT) using a unique tissue engineering technique, the cell assembled viscous tissue by sedimentation (CAViTs) method. The "viscous bodies" created using the CAViTs method exhibited significant self-assembly within 6 h after seeding, promoting cell-cell interaction. The level of albumin secreted by the VHT was four times higher than that of 2D-coculture and maintained for 1 month. The gene expression pattern of the VHT was closer to that of total human liver, compared with the 2D system. Quantitative evaluations of the vascular structure of VHT treated with two typical SOS-inducing compounds, monocrotaline and retrorsine, revealed higher sensitivity (IC₅₀ = 40.35 µM), 19.92 times higher than the cell-viability assay. Thus, VHT represents an innovative in vitro model that mimics the vessel network structure and could become a useful tool for the early screening of compounds associated with a risk of vascular toxicity. STATEMENT OF SIGNIFICANCE: Mimicking sinusoidal structures in in vitro liver model is important to consider from the perspective of predicting hepatotoxicity such like sinusoidal obstruction syndrome (SOS). However, it was difficult to reconstruct the vascular structure within the hepatocyte-rich environment. In this study, we constructed a vascularized hepatic tissue in a high-throughput manner by a unique method using collagen and heparin, and evaluated its applicability to toxicity assessment. Vessel morphology analysis of the model treated by monocrotaline, which is a well-known SOS-inducing compound, could predict the toxicity with higher sensitivity. To the best of our knowledge, this is the first report to provide vascularized hepatic tissues using sinusoidal endothelial cells at least for demonstrating applicability to the evaluation of SOS induction risk.

Dietary alcohol exacerbates the hepatotoxicity induced by pyrrolizidine alkaloids: Hazard from food contamination

Pyrrolizidine alkaloids (PAs) are the most common plant-derived toxins with emerging evidence to contaminate soil, water, nearby plants and derived food products. Outbreaks of human poisoning cases, due to the ingestion of PA-contaminated food, have been reported in various countries including Ethiopia. This study first investigated the contamination of PAs in retail honey in Ethiopia. A striking 77% of honey samples (27/30) were found to contain PAs with the content ranging over 1.5-323.4 μg/kg. Notably, these PAs were also found as contaminants in mead, an alcoholic beverage made from local honey, indicating the transfer of PAs from the primarily contaminated honey into mead. Further toxicological examinations revealed that long-term PA exposure caused vasculature damage, fibrosis, and steatosis in mouse livers, and co-exposure to dietary alcohol exacerbated the PA-induced chronic hepatotoxicity. Furthermore, the study revealed that moderate alcohol intake did not affect the initiation mechanism (hepatic cytochrome P450-mediated bioactivation) of PA-induced hepatotoxicity but significantly disturbed hepatic glutathione homeostasis, thereby increasing oxidative stress in mouse liver and enhancing PA-induced hepatotoxicity. Our findings exemplify the carry-over of PA contamination through the food chain. Precautionary interventions are warranted on the hazardous effects of dietary exposure to PAs, particularly with concomitant alcohol consumption.

Fasting augments pyrrolizidine alkaloid-induced hepatotoxicity

Pyrrolizidine alkaloids (PAs) are phytotoxins widely present in various natural products and foodstuffs. The present study aims to investigate the effects of fasting on PA-induced hepatotoxicity and the underlying biochemical mechanisms. The results of hepatotoxic study showed that 15-h overnight fasting significantly exacerbated the hepatotoxicity of retrorsine (RTS, a representative toxic PA) in fasted rats compared to fed rats, as indicated by remarkably elevated plasma ALT and bilirubin levels and obvious liver histological changes. Further toxicokinetic studies revealed that fasting significantly enhanced cytochromes P450 enzymes (CYPs)-mediated metabolic activation of RTS leading to increased formation of pyrrole-protein adducts and thus decreased the in vivo exposure and excretion of both parent RTS and its N-oxide metabolite. Metabolic studies demonstrated that fasting induced enzyme activities of CYP1A2, CYP2B6 and CYP2E1 that participated in catalyzing RTS to its reactive pyrrolic metabolites. Moreover, fasting also dramatically decreased hepatic glutathione (GSH) content, which restricted the detoxification of GSH by neutralizing the reactive pyrrolic metabolite of RTS, further contributing to the enhanced hepatotoxicity. The present findings may have an impact on future PA toxicity tests with different dietary styles and/or risk assessment of metabolite-mediated toxins by considering the profound effects of fasting.

Modeling nonalcoholic fatty liver disease on a liver lobule chip with dual blood supply

Nonalcoholic fatty liver disease (NAFLD) has emerged as a public health concern. To date, the mechanism of NAFLD progression remains unclear, and pharmacological treatment options are scarce. Traditional animal NAFLD models are limited in helping address these problems due to interspecies differences. Liver chips are promising for modeling NAFLD. However, pre-existing liver chips cannot reproduce complex physicochemical microenvironments of the liver effectively; thus, NAFLD modeling based on these chips is incomplete. Herein, we develop a biomimetic liver lobule chip (LC) and then establish a more accurate on-chip NAFLD model. The self-developed LC achieves dual blood supply through the designed hepatic portal vein and hepatic artery and the microtissue cultured on the LC forms multiple structures similar to in vivo liver. Based on the LC, NAFLD is modeled. Steatosis is successfully induced and more importantly, changing lipid zonation in a liver lobule with the progression of NAFLD is demonstrated for the first time on a microfluidic chip. In addition, the application of the induced NAFLD model has been preliminarily demonstrated in the prevention and reversibility of promising drugs. This study provides a promising platform to understand NAFLD progression and identify drugs for treating NAFLD. STATEMENT OF SIGNIFICANCE: Liver chips are promising for modeling nonalcoholic fatty liver disease. However, on-chip replicating liver physicochemical microenvironments is still a challenge. Herein, we developed a liver lobule chip with dual blood supply, achieving self-organized liver microtissue that is similar to in vivo tissue. Based on the chip, we successfully modeled NAFLD under physiologically differentiated nutrient supplies. For the first time, the changing lipid zonation in a single liver lobule with the early-stage progression of NAFLD was demonstrated on a liver chip. This study provides a promising platform for modeling liver-related diseases.

Transforming a well into a chip: A modular 3D-printed microfluidic chip

Organ-on-a-Chip platforms provide rich opportunities to observe interactions between different cell types under in vivo-like conditions, i.e., in the presence of flow. Yet, the costs and know-how required for the fabrication and implementation of these platforms restrict their accessibility. This study introduces and demonstrates a novel Insert-Chip: a microfluidic device that provides the functionality of an Organ-on-a-Chip platform, namely, the capacity to co-culture cells, expose them to flow, and observe their interactions-yet can easily be integrated into standard culture systems (e.g., well plates or multi-electrode arrays). The device is produced using stereolithograpy 3D printing and is user-friendly and reusable. Moreover, its design features overcome some of the measurement and imaging challenges characterizing standard Organ-on-a-Chip platforms. We have co-cultured endothelial and epithelial cells under flow conditions to demonstrate the functionality of the device. Overall, this novel microfluidic device is a promising platform for the investigation of biological functions, cell-cell interactions, and response to therapeutics.

Metabolism-mediated cytotoxicity and genotoxicity of pyrrolizidine alkaloids

Pyrrolizidine alkaloids (PAs) and PA N-oxides are common phytotoxins produced by over 6000 plant species. Humans are frequently exposed to PAs via ingestion of PA-containing herbal products or PA-contaminated foods. PAs require metabolic activation to form pyrrole-protein adducts and pyrrole-DNA adducts which lead to cytotoxicity and genotoxicity. Individual PAs differ in their metabolic activation patterns, which may cause significant difference in toxic potency of different PAs. This review discusses the current knowledge and recent advances of metabolic pathways of different PAs, especially the metabolic activation and metabolism-mediated cytotoxicity and genotoxicity, and the risk evaluation methods of PA exposure. In addition, this review provides perspectives of precision toxicity assessment strategies and biomarker development for the risk control and translational investigations of human intoxication by PAs.

Metabolic Pattern of Hepatotoxic Pyrrolizidine Alkaloids in Liver Cells

Contamination with 1,2-unsaturated pyrrolizidine alkaloids (PAs) is a serious problem for certain phytomedicines, foods, and animal feeds. Several of these PAs are genotoxic and carcinogenic, primarily in the liver, upon cytochrome P450 (CYP)-catalyzed activation into reactive (pyrrolic and pyrrole-like) metabolites. Here we investigated the metabolism of selected PAs (echimidine, europine, lasiocarpine, lycopsamine, retrorsine, and senecionine) in rat hepatocytes in primary culture and in human CYP3A4-transfected HepG2 cells. The open-chained diesters echimidine and lasiocarpine and the cyclic diester senecionine were extensively metabolized in rat hepatocytes into a broad spectrum of products released into the medium. A large portion of unidentified, possibly irreversibly bound, products remained in the cells while detectable amounts of reactive and other metabolites were found in the incubation media. In HepG2-CYP3A4 cells, lasiocarpine was more extensively metabolized than echimidine and senecionine which also gave rise to the release of pyrrolic metabolites. In human cells, no pyrrolic metabolites were detected in retrorsine or lycopsamine incubations, while no such metabolites were detected from europine in both cell types. Other types of metabolic changes comprised modifications such as side chain demethylation or oxygenation reactions like the formation of N-oxides. The latter, considered as a detoxification step, was a major pathway with cyclic diesters, was less distinctive for echimidine and lycopsamine and almost negligible for lasiocarpine and europine. Our data are in agreement with previously published cyto- and genotoxicity findings and suggests that the metabolic pattern may contribute substantially to the specific toxic potency of a certain congener. In addition, marked differences were found for certain congeners between rat hepatocytes and transfected human HepG2 cells, whereby a high level of bioactivation was found for lasiocarpine, whereas a very low level of bioactivation was observed for monoesters, in particular in human cells.

Lung injury induced by pyrrolizidine alkaloids depends on metabolism by hepatic cytochrome P450s and blood transport of reactive metabolites

Pyrrolizidine alkaloids (PAs) are common phytotoxins with both hepatotoxicity and pneumotoxicity. Hepatic cytochrome P450 enzymes are known to bioactivate PAs into reactive metabolites, which can interact with proteins to form pyrrole-protein adducts and cause intrahepatic cytotoxicity. However, the metabolic and initiation biochemical mechanisms underlying PA-induced pneumotoxicity remain unclear. To investigate the in vivo metabolism basis for PA-induced lung injury, this study used mice with conditional deletion of the cytochrome P450 reductase (Cpr) gene and resultant tissue-selective ablation of microsomal P450 enzyme activities. After oral exposure to monocrotaline (MCT), a pneumotoxic PA widely used to establish animal lung injury models, liver-specific Cpr-null (LCN) mice, but not extrahepatic Cpr-low (xh-CL) mice, had significantly lower level of pyrrole-protein adducts in the serum, liver and lungs compared with wild-type (WT) mice. While MCT-exposed LCN mice had significantly higher blood concentration of intact MCT, compared to MCT-exposed WT or xh-CL mice. Consistent with the MCT in vivo bioactivation data, MCT-induced lung injury, represented by vasculature damage, in WT and xh-CL mice but not LCN mice. Furthermore, reactive metabolites of MCT were confirmed to exist in the blood efflux from the hepatic veins of MCT-exposed rats. Our results provide the first mode-of-action evidence that hepatic P450s are essential for the bioactivation of MCT, and blood circulating reactive metabolites of MCT to the lung causes pneumotoxicity. Collectively, this study presents the scientific basis for the application of MCT in animal lung injury models, and more importantly, warrants public awareness and further investigations of lung diseases associated with exposure to not only MCT but also different PAs.

Establishment of a novel CYP3A4-transduced human hepatic sinusoidal endothelial cell model and its application in screening hepatotoxicity of pyrrolizidine alkaloids

Pyrrolizidine alkaloids (PAs) are extensively distributed in plants and are known to damage hepatic sinusoidal endothelial cells (HSECs) via metabolic activation mediated by hepatic cytochrome P450 enzymes (CYPs), particularly the CYP3A4 isozyme. Different PAs have distinct toxic potencies and their toxic effects on HSECs are difficult to be determined in cultured cells, because HSECs lack the key CYP3A4 isozyme for metabolic activation. This study aims to establish a novel, convenient and reliable CYP3A4-expressing HSEC model using human HSECs transduced with lentivirus carrying CYP3A4-ires-eGFP, for evaluating the hepatotoxicity of different PAs on their target HSECs. The developed CYP3A4-expressing HSEC (HSEC-CYP3A4) model was verified by the expression of GFP and CYP3A4 and by the ability to metabolize nifedipine, a classic CYP3A4 substrate. Treated with retrorsine, a representative toxic PA, HSEC-CYP3A4 cells showed significantly reduced cell viability, depletion of GSH, and increased formation of pyrrole-protein adducts. Furthermore, this newly developed cell model successfully discriminated the cytotoxic potency of different PAs evidenced by their IC₄₀ values. In conclusion, the established HSEC-CYP3A4 cell model can be used as a rapid screening platform for assessing the relative potencies of individual PAs on their target HSECs and for investigating the mechanisms underlying PA-induced hepatic sinusoidal damage.