Establishment of "in vitro" and "in vivo" models with underlying mitochondrial dysfunction for testing idiosyncratic drug toxicity

Idiosyncratic hepatotoxicity is an adverse reaction that can be evoked by certain drugs. The characteristic of this rare event is its occurrence at therapeutic doses that are usually well tolerated by the patients, and most importantly, this toxicity frequently becomes manifested in a specific population of subjects. These motifs point towards underlying or acquired genetically and/or environmental risk factors rendering some patients more susceptible to idiosyncratic adverse reactions. Although the mechanisms of drug-induced liver injuries are not completely understood, mitochondrial dysfunction and/or formation of toxic metabolites have been recognized as major contributors in many cases of idiosyncratic hepatotoxicity. Moreover, predisposing mitochondrial dysfunction has been suspected to be one of the factors enhancing the risk for idiosyncratic toxicity. Therefore, severe inhibition of cellular metabolic functions, maintained by the mitochondria, an organelle crucial for proper energy homeostasis and delivery for the cell, may induce oxidative stress that in turn can result in cell death and ultimately to liver dysfunction and death. The lack of reliable methods to identify substances with the risk for idiosyncratic mitochondrial toxicity represents a major issue for the pharmaceutical industry, since unpredicted idiosyncratic toxicity can consequence in withdrawal of the drug from the market or in warnings by health authorities. It is thus an important need to characterize drugs that are related to idiosyncratic toxicity for a better understanding of implicated mechanisms and also to develop suitable in vitro and in vivo systems that can be applied to test drugs for their idiosyncratic potential.
This thesis includes three papers that have been published or submitted for publication and a short report as pilot project. The first two projects and the short report present the development of an in vitro and in vivo system with underlying mitochondrial defects that could serve as tool to uncover drugs with an increased risk for idiosyncratic toxicity, whereas the third project addresses the molecular mechanisms of hepatotoxicity in vitro of the antimitotic agents of azoles.
Our first paper describes a newly developed in vitro model in HepG2 cells induced by the chemical hydroxy-cobalamin [c-lactam] (HCCL), a vitamin B12 analog. The model exhibits subtle changes in the respiratory chain function and slight increase of oxidative stress leading to mitochondrial swelling. HCCL is known to competitively inhibit proper processing of substrates of the propionate pathway and causes similar defects as seen for the inherited metabolic disorder methylmalonic aciduria. We showed that subtoxic concentrations of HCCL could impair mitochondrial function and thus used this model that should reflect preexisting mitochondrial diseases of patients rendering them more susceptible to mitochondrial toxicants. In this context, we demonstrated that known mitochondrial toxicants applied on our HCCL model exert increased toxicity, which supports the hypothesis that risk factors such as underlying mitochondrial defects predispose for idiosyncratic toxicity of compounds that have a known adverse mechanisms on mitochondrial function such as dronedarone, benzbromarone, and ketoconazole.
Our second paper aimed to expand the results obtained from the HCCL in vitro study into a mouse model. We showed that daily i.p. treatment of HCCL over three weeks induced an elevation of methylmalonic acid, a marker of methylmalonic aciduria. Moreover, slight decrease of complex I of the electron transport chain in liver homogenates were observed. Histopathological analysis revealed enhanced glycogen content and fat accumulation in the liver explaining the observed elevation of liver weight. Besides that, we did not find any increased oxidative stress or other dysfunctions of the mitochondria due to HCCL exposure. Because of the metabolic changes found in this study were minor, we did not further pursue the exploration of this model in order to test idiosyncratic drugs on it.
The third paper focused on the discovery of the molecular mechanisms of hepatotoxicity of antifungal azoles. Cytotoxicity experiments in HepG2 cells uncovered dose-dependent toxicity of ketoconazole and posaconazole, whereas fluconazole and voriconazole did not affect the cells. Toxicity studies in HepaRG cells did not point at any implication of metabolites for all four tested azoles. Our functional examination on mitochondrial function of ketoconazole and posaconazole depicted a strong inhibitory potential on the respiratory chain function, accompanied by oxidative stress and apoptosis, suggesting mitochondrial dysfunction responsible for azole-induced hepatotoxicity. The exposure of these two mitochondrial toxins on our previously created HCCL in vitro model evidenced an increased toxicity already at therapeutic concentrations, strongly supporting the assumption that preexisting mitochondrial defects can lower the threshold level of a drug’s toxicity and thereby a manifestation of idiosyncratic hepatotoxicity.
Lastly, the short report readopts the subject of the establishment of in vitro models with mitochondrial dysfunction, implementing it as system for testing idiosyncratic toxicity. With siRNA we depleted two different essential mitochondrial enzymes, SIRT3 or SOD2, in HepG2 cells. We hypothesized to find an increase in cytotoxicity and enhanced oxidative stress after exposure to known mitochondrial toxicants if cells either present depletion in the deacetylation of mitochondrial proteins or a silencing of one of the major antioxidant defense systems. However, our results did not reflect this assumption and did not reveal any enhanced toxicity of HepG2 cells lacking specific mitochondrial functions compared to normal HepG2 cells. Furthermore, induction of a second stress factor such as the removal of glutathione in SOD2 siRNA cells did not increase the susceptibility of these cells to mitochondrial toxicants. Nevertheless, with this study we presented interesting basic approaches how to establish models to test idiosyncratic toxicity in vitro.