Genetic toxicity assessment using liver cell models: past,present, and future

ABSTRACT

Genotoxic compounds may be detoxified to non-genotoxic metabolites while many pro-carcinogens require metabolic activation to exert their genotoxicity in vivo. Standard genotoxicity assays were developed and utilized for risk assessment for over 40 years. Most of these assays are conducted in metabolically incompetent rodent or human cell lines. Deficient in normal metabolism and relying on exogenous metabolic activation systems, the current in vitro genotoxicity assays often have yielded high false positive rates, which trigger unnecessary and costly in vivo studies. Metabolically active cells such as hepatocytes have been recognized as a promising cell model in predicting genotoxicity of carcinogens in vivo. In recent years, significant advances in tissue culture and biological technologies provided new opportunities for using hepatocytes in genetic toxicology. This review encompasses published studies (both in vitro and in vivo) using hepatocytes for genotoxicity assessment. Findings from both standard and newly developed genotoxicity assays are summarized. Various liver cell models used for genotoxicity assessment are described, including the potential application of advanced liver cell models such as 3D spheroids, organoids, and engineered hepatocytes. An integrated strategy, that includes the use of human-based cells with enhanced biological relevance and throughput, and applying the quantitative analysis of data, may provide an approach for future genotoxicity risk assessment.     

Semipermeable Hollow Fiber Membranes in Hepatocyte Bioreactors: A Prerequisite for a Successful Bioartificial Liver?

Abstract: Recent studies have shown that liver support systems based on viable hepatocytes can prolong life in animal models of acute liver failure. Now the time has come to elucidate the design characteristics that are essential to construct an efficient bioreactor. The gold standard remains the intact liver. Despite the very high cell density in this organ, individual cell perfusion is guaranteed resulting in low diffusional gradients which are essential for optimal mass transfer. These conditions are not met in bioreactors based on hollow fiber membranes. Moreover, the semipermeable membranes can foul and act as a diffusional barrier between the hepatocytes and the blood or plasma of the recipient. We devised a novel bioreactor for use as a bioartificial liver that does not include hollow fiber membranes for blood or plasma perfusion. The device is based on an integral oxygenator and a nonwoven polyester matrix material for hepatocyte culture as small aggregates. The efficacy of this original design was tested in rats with liver ischemia. Preliminary results show statistically significantly improved survival; life was prolonged 100% compared to the control experiments.     

Cytotoxicity of propyl gallate and related compounds in rat hepatocytes

 The cytotoxic effects of propyl gallate (PG), its related gallates and gallic acid have been studied in freshly isolated rat hepatocytes. Addition of PG (0.5–2.0 mM) to hepatocyte suspension elicited concentration-dependent cell death accompanied by losses of intracellular ATP, adenine nucleotide pools, glutathione (GSH) and protein thiols. The rapid loss of intracellular ATP preceded the onset of cell death caused by PG. In the comparative toxic effects of PG and related gallates at concentration of 1 mM, octyl gallate (OG), dodecyl gallate (DG) and butyl gallate (BG) elicited an abrupt depletion of ATP, followed by an acute cell death. These gallates were more toxic than PG; the toxic effects of PG were similar to those of methyl gallate (MG) and ethyl gallate (EG). In mitochondria isolated from rat liver, PG caused a concentration-dependent increase in the rate of state 4 oxygen consumption, indicating an uncoupling effect. The rate of state 3 oxygen consumption was inhibited by OG and DG. According to the respiratory control index, the order of impairment potency to mitochondria was OG>BG, DG>PG>EG, MG>gallic acid. These results indicate that PG and related gallates are toxic to hepatocytes and that the acute cytotoxicity may be due to mitochondrial dysfunction. Received: 16 May 1994 / Accepted: 15 August 1994     

Animal model for liver cell banking from non-heart beating donors after prolonged ischaemia time

BACKGROUND AND AIM: Although there is a growing interest on the use of non-heart beating donors to enlarge the liver donor pool, livers with prolonged warm ischaemia time are not currently considered for organ transplantation. We hypothesised that these organs may represent a source of hepatocytes for cell transplantation and/or use in bioartificial liver devices. Thus, we investigated if prolonged ischaemia could influence the recovery and viability of functional hepatocytes dissociated from rat livers. METHODS: Hepatocytes were isolated from the liver within 15 min after death (t=15 min) and after 4, 8 and 12h of ischaemia. Cells were either maintained in culture or cryopreserved. In all products, we evaluated cell recovery and viability, hepatocyte markers and cellular functions, including albumin and urea production. RESULTS: The number of cells per gram of tissue was similar at 15 min, 4 and 8h, while it was significantly decreased at 12h. About 0.2 x 10(6) viable cells expressing hepatocyte markers and producing albumin and urea were isolated up to 8h of ischaemia per gram of tissue. CONCLUSIONS: Recovery of viable and functional hepatocytes seems possible after prolonged ischaemia time. These data warrant the evaluation of hepatocyte isolation from human livers of non-heart beating donors.     

Correlation between in vivo and in vitro toxic effects of foreign compounds

Studies in rats in vivo and in isolated hepatocytes from the same species and strain of animal in vitro with the hepatotoxicant hydrazine have shown that despite measuring the same parameters in each system, the effects do not always show a quantitative or qualitative correlation. For example depletion of glutathione and ATP occurred in both systems but required a much higher concentration of hydrazine in vitro. The effects on triglyceride levels, citrulline synthesis and taurine levels in vivo were not observed in vitro and the inhibition of urea synthesis and cytotoxicity in vitro were not observed in vivo.Inhibition of protein synthesis proved to be the marker which showed the best correlation, occurring at a similar hydrazine concentration in vitro and in vivo although not to the same extent.The situation with other toxicants is variable, in some cases correlation is good, in others modification of conditions in vitro are required.     

Inhibition of MAP kinase activity moderates changes in expression and function of Cx32 but not claudin-1 during DNA synthesis in primary cultures of rat hepatocytes

The signal transduction pathways and activation of the MAP kinase or PI3 kinase signaling cascade regulate a variety of cellular processes, including proliferation and differentiation in hepatocytes. To elucidate the mechanisms of signal transmission required for the regulation of gap and tight junctions during DNA synthesis in rat hepatocytes, we determined changes of expression and function of gap and tight junctions of cells grown in primary culture, using inhibitors of signaling pathways for MAP kinase (PD98059) and PI3 kinase (LY294002). During the stimulation of DNA synthesis induced by epidermal growth factor (EGF), immunoreactivity and mRNAs of gap junction protein Cx32 and of tight junction protein claudin-1 markedly decreased with reduction of gap junctional intercellular communication (GJIC) and the fence function of tight junctions. In Western blots, whole-cell lysate of claudin-1 protein decreased and phosphorylated Cx32 protein in the insoluble fraction of Triton X-100 increased during the stimulation of DNA synthesis. During reinhibition of DNA synthesis, the changes of Cx32 and claudin-1 returned to control levels, as did both functions. In treatment with the inhibitors before DNA synthesis, PD98059 inhibited the changes of expression and function of Cx32, but not claudin-1, without inhibition of cell growth, whereas LY294002 completely inhibited cell growth. These findings indicate that the PI3 kinase pathway rather than the MAP kinase pathway plays an important role for EGF-induced proliferation of rat hepatocytes, and that changes of Cx32 in hepatocytes during the stimulation of DNA synthesis may be in part controlled through MAP kinase. Furthermore, Cx32, but not claudin-1, protein may be a target of activated MAP kinase in hepatocytes.     

Microtubules are required for cytokeratin aggresome (Mallory body) formation in hepatocytes: an in vitro study

Mallory bodies are cytokeratin-ubiquitin aggresomes that form in hepatocytes in many different chronic liver diseases. One of the key components in aggresome formation, not yet investigated in Mallory body formation, is the role of microtubules. An in vitro tissue culture assay is required to test for microtubule involvement in Mallory body formation so that Mallory body formation can be observed in the presence or absence of microtubule-disrupting agents. In this report, a new model of in vitro Mallory body formation was developed, which uses cultured hepatocytes isolated from drug-primed mice. When hepatocytes were incubated in the presence of antimicrotubule agents, they failed to form Mallory bodies. It is concluded that intact microtubules are required for Mallory body formation.     

Effect of cyclosporin A,azathioprine, and prednisolone on carbohydrate metabolism of rat hepatocytes

The effect of different immunosuppressive drugs (prednisolone, azathioprine, cyclosporin A) on liver carbohydrate metabolism in the rat was investigated. Daily administration of prednisolone (3 mg/kg body weight) and azathioprine (2 mg/kg body weight) intraperitoneally for 2 weeks caused significantly lower liver glycogen content than that in NaCl-treated controls. Liver glucose and lactate content, as well as plasma glucose, glucagon, and serum insulin concentration of these animals, remained unchanged. There were no differences in any of these parameters between cyclosporin A (15 mg/kg body weight)-treated and vehicle (olive oil/ethanol)-treated animals. Prednisolone caused significantly lower glucose production in isolated rat hepatocytes using Na-pyruvate as the substrate, whereas glucose production was unchanged in hepatocytes of azathioprine-treated rats using pyruvate or l-serine as substrates. Glucose production from pyruvate or serine was significantly inhibited by cyclosporin A compared to the vehicle, but did not differ from the effects of azathioprine and prednisolone. Lactate production was significantly lower in cyclosporin-treated animals than in those given either the vehicle or azathioprine. Cyclosporin A completely reversed the inhibition of hepatocyte glycogen consumption caused by the vehicle. However, glycogen production in the presence of cyclosporin A was comparable to the effects of prednisolone and azathioprine. Finally, hepatocyte ketone body production using pyruvate as the substrate was higher in the presence of all immunosuppressive drugs. In the presence of serine, acetoacetate production increased in rats treated with 50 mg/kg body weight cyclosporin A, and -hydroxybutyrate production in animals receiving 15 and 50 mg/kg body weight cyclosporin A.This article is dedicated to Professor K. Kochsiek, Chief of the Medical Department, University of Würzburg, FRG, on the occasion of his 60th birthday     

The mycotoxin ochratoxin A is a substrate for phenylalanine hydroxylase in isolated rat hepatocytes and in vivo

Ochratoxin A (OTA), is a mycotoxin contaminating food and feed stuffs, consisting of a chlorinated dihydroisocoumarin linked through a 7-carboxyl group tol-phenylalanine by an amide bond. When OTA (0.12–1.4 mM) is incubated with freshly isolated rat hepatocytes, it inhibits both the hydroxylation of phenylalanine (0.05 mM) to tyrosine, catalyzed by phenylalanine hydroxylase and the subsequent metabolism of tyrosine as measured by homogentisate oxidation. The IC₅₀ of OTA for phenylalanine hydroxylation is 0.43 mM. OT, (0.5–1.0 mM), the dihydroisocoumarin moiety of OTA, does not inhibit phenylalanine hydroxylase activity under these conditions. During incubations of hepatocytes with uniformly labelled [³H]-OTA and unlabelled phenylalanine, tyrosine-ochratoxin A is formed (up to 6% of the total mycotoxin added), indicating that ochratoxin can act as a substrate for phenylalanine hydroxylase. In vivo tyrosine-OTA is also found in liver of poisoned animals.     

Thymidine transport in hepatocytes

Thymidine transport and phosphorylation were investigated in isolated rat hepatocytes and AS 30 D hepatoma cells. In contrast to hepatoma cells, hepatocytes exhibited a minimum of thymidine phosphorylation due to a 100-fold smaller thymidine kinase activity. In hepatocytes thymidine is transported by two transport systems: a specific concentrative high affinity system and an unspecific non-concentrative low affinity system. In hepatoma cells only the low affinity system could be detected. A single dose of 20 or 50 mg diethylnitrosamine/kg body weight induced in hepatocytes a remarkable increase of thymidine kinase activity and a decrease of the transport by the high affinity system. Thymidine transport and phosphorylation by hepatocytes are considered to be sensitive markers for early recognition of toxin-induced liver regeneration.

---

Zusammenfassung Thymidintransport und -phosphorylierung wurden in isolierten Rattenhepatozyten und AS 30 D-Hepatomzellen untersucht. Hepatozyten wiesen im Gegensatz zu Hepatomzellen aufgrund einer 100fach niedrigeren Thymidinkinaseaktivität eine äußerst niedrige Thymidinphosphorylierungsrate auf. In Hepatozyten wurde Thymidin durch zwei Transportsysteme aufgenommen: ein spezifisches, konzentratives high affinity System und ein unspezifisches, nichtkonzentratives low affinity System. In ATP-freien Hepatomzellen konnte nur das low affinity System nachgewiesen werden. Eine einmalige Dosis von 20 bzw. 50 mg Diäthylnitrosamin/kg Körpergewicht bewirkte in Hepatozyten einen Anstieg der Thymidinkinaseaktivität und eine Verminderung des Thymidintransports über das high affinity System. Thymidintransport und -phosphorylierung in Hepatozyten erwiesen sich als sensitives System zur frühen Erkennung beginnender Leberregeneration nach toxischer Vorschädigung.