Response to Letter to the Editor from Drs. van Kesteren, Pronk, Heusinkveld, Luijten and Hakkert concerning Yamada et al. (2021): Critical evaluation of the human relevance of the mode of action for rodent liver tumor formation by activators of the constitutive androstane receptor (CAR). Crit. Rev. Toxicol. Vol. 51: 373-394.

To the Editor,

We are pleased to have the opportunity to respond to the letter to the Editor by Drs. van Kesteren, Pronk, Heusinkveld, Luijten and Hakkert concerning our published paper entitled “Critical evaluation of the human relevance of the mode of action for rodent liver tumor formation by activators of the constitutive androstane receptor (CAR)” (Yamada et al. 2021). We should like to thank Drs. van Kesteren and colleagues for the interest that they have shown in our paper and for the comments made in their letter. As described below, we have carefully evaluated their comments in order to clarify the points made and to deal with some possible misconceptions concerning the established mode of action (MOA) for rodent liver tumor formation by nongenotoxic CAR activators and its human relevance.

The framework for MOA analysis and the assessment of the human relevance of an animal MOA for tumor formation MOA is now well established, a fundamental consideration being the comparison of effects in experimental animals with those in humans (Boobis et al. 2006). Based on studies conducted in a number of laboratories over many years, the pivotal species difference is that CAR activators are mitogenic agents in rodent liver (i.e. mice and rats) but not in human hepatocytes (Elcombe et al. 2014; Lake 2018; Yamada 2018; Yamada et al. 2021). In contrast, some other effects of CAR activators, including receptor activation, liver hypertrophy and induction of cytochrome P450 (CYP) enzymes, can be demonstrated in both rodent and human liver. In their letter Drs. van Kesteren and colleagues note that phenobarbital (PB) is an “archetypical CAR activator”. PB is indeed a model compound which has been extensively studied both in experimental animals and in humans (Whysner et al. 1996; IARC 2001); with data from investigations in mouse and rat liver providing robust data with which to evaluate if a new chemical (as in Table 1 of our paper) may be classified as a rodent CAR activator and hepatocyte mitogen (Elcombe et al. 2014; Yamada et al. 2021).

In our paper we noted the marked species similarities between the hepatic effects of nongenotoxic CAR activators and those of nongenotoxic activators of the peroxisome proliferator-activated receptor alpha (PPARα). Analysis of a large amount of available literature clearly demonstrates a pivotal species difference in that while both CAR and PPARα activators are mitogenic agents in the mouse and rat, they are not mitogenic agents in other species including the Syrian hamster, guinea pig and humans (Corton et al. 2014, 2018; Lake 2018; Yamada et al. 2021). Indeed, the MOA for rodent liver tumor formation by PPARα activators is considered not to be relevant for humans (ECHA 2017).

In their letter Drs. van Kesteren and colleagues question the relevance of the two experimental systems used, namely cultured hepatocytes and chimeric mice with human hepatocytes, to evaluate species differences in effects on hepatocyte replicative DNA synthesis (RDS) and also the available human epidemiology data. With respect to the experimental systems, it is stated that information on the positive predictive value with human hepatocytes for known human carcinogens is not presented. The rodent liver CAR activator MOA has been established for nongenotoxic agents, with alternative MOAs having been excluded. It is thus not necessary to consider known human carcinogens operating by other MOAs, which would certainly include genotoxic agents. Our responses to the points made on the experimental systems and the epidemiology data are described below.

Studies in cultured hepatocytes

Issues raised by Drs. van Kesteren and colleagues include that the use of cultured hepatocytes to determine species differences in effects on RDS is not sufficiently substantiated and that the preparation of human hepatocytes is complex, donor-dependent and that functionality can be compromised. In addition, they query points concerning the number and type of chemicals which do not produce RDS in human hepatocytes, the minimum number of donors to be used together with the age range and sex of the donors.

We consider that primary hepatocyte cultures are an established test system, which has been employed by a number of laboratories over many years to screen chemicals for effects on RDS and to evaluate species differences in response. For example, a comparison of the effects of PB and some other rodent liver mitogenic agents in rat and human hepatocytes was reported over 30 years ago (Parzefall et al. 1991). An evaluation of the literature demonstrates that the observed species difference in response of human hepatocytes to CAR activators has been consistently observed in a number of laboratories employing different hepatocyte culture conditions and analytical techniques. Studies with human hepatocyte preparations have been performed with both freshly isolated cells obtained by liver perfusion and more recently, by what is now readily commercially available, using preparations of cryopreserved plateable hepatocytes. In terms of hepatocyte viability, a typical experimental protocol would include assays of cell viability (e.g. morphology, mitochondrial function, enzyme leakage) in order to confirm the viability of the hepatocyte preparation and to ensure that the test chemical is evaluated at up to cytotoxic concentrations in order to clearly demonstrate a negative or positive response. In terms of functional viability of cultured rodent and human hepatocyte preparations, many studies have also included measurement of CYP enzyme mRNA levels and/or enzyme activities. Thus for human hepatocytes, as would be expected from in vivo studies with PB and some other therapeutic agents, treatment with CAR activators has been demonstrated to induce CYP enzymes such as CYP2B6 and CYP3A4.

Table 4 of our paper presents data where PB and several other known rodent liver CAR activators of diverse chemical structure stimulate RDS in cultured mouse and/or rat hepatocytes, but not in human hepatocytes. The data described in Table 4 clearly addresses the points concerning the numbers, sex, and age range of the donors of human hepatocytes, with these studies being conducted in many human hepatocyte preparations from both male and female donors aged 0.8-80 years. The data shown in Table 4 are supported by several other published investigations where PB, other rodent liver mitogenic agents, and the potent human CAR activator 6-(4-chlorophenyl)imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4- dichlorobenzyl)oxime (CITCO) have been shown not to induce RDS in cultured human hepatocytes (Parzefall et al. 1991; Soldatow et al. 2016; Haines et al. 2018; Lake 2018).

In terms of a suitable positive control for RDS studies in both cultured hepatocytes and chimeric mice with human hepatocytes, Drs. van Kesteren and colleagues question the use of either epidermal growth factor (EGF) and/or hepatocyte growth factor (HGF) in the absence of a suitable chemical. As demonstrated in Table 4 of our paper and in many other studies (Lake 2018; Yamada 2018, 2021), this is a fundamental species difference in that while PB and other CAR activators induce RDS in mouse and/or rat hepatocytes, CAR activators (including the potent human CAR activator CITCO) do not induce RDS in human hepatocytes. As such, EGF and HGF are suitable available positive controls to evaluate the responsiveness of the two experimental systems to a mitogenic stimulus.

Studies in chimeric mice with human hepatocytes

In this in vivo model of human hepatocyte function, transplanted human hepatocytes replace most of the host mouse hepatocytes, with the human hepatocytes exhibiting nearly normal morphology and expressing most genes at similar levels to those expressed by normal human liver (Tateno and Kojima 2020). This in vivo model has been employed for many applications, including studies of xenobiotic metabolism and toxicity (Tateno and Kojima 2020; Yamada 2021). In one study chimeric mice with human hepatocytes were treated with PB at up to maximum tolerated dose levels, as some lethality was observed (Yamada et al. 2014). While PB treatment had no effect on hepatocyte RDS, significant increases were observed in human CYP2B6 and CYP3A4 mRNA levels. The lack of effect of PB on RDS in chimeric mice with human hepatocytes was supported by the results of gene array studies where increases in genes associated with cell proliferation were not observed (Yamada et al. 2014, 2020; Ohara et al. 2017). These findings are supported by a recent proteomic study (Sprenger et al. 2022), where PB induced CYP enzymes in human hepatocytes of chimeric mice but there was no effect on hepatocyte RDS. In another study, two other nongenotoxic rodent CAR activators were shown not to increase RDS in hepatocytes from three separate human donors (Okuda et al. 2017). The validity of this test system for evaluating species differences in cell proliferation was demonstrated by transplanting rat hepatocytes into chimeric mice where, as would be expected from previous in vivo rat studies, subsequent PB treatment resulted in hepatocyte hypertrophy, increased RDS and induction of CYP2B1/2 mRNA levels (Yamada et al. 2020). Other studies have also established the validity of this in vivo model of human hepatocyte function. As would be expected from many species difference studies with PPARα activators, the hypolipidemic drug fenofibrate did not stimulate RDS in human hepatocytes (Tateno et al. 2015; Corton et al. 2018). In an investigation into porphyria-mediated hepatocellular injury, subsequent regenerative hyperplasia was observed in chimeric mice with human hepatocytes (Eguchi et al. 2021), and in a study with an anti-human tumor necrosis factor-related apoptosis-inducing ligand receptor 2 monoclonal antibody, an upregulation of cell cycle-related functions likely to represent cellular regeneration was observed (Nihira et al. 2019). This in vivo model of human hepatocyte function has been employed for many studies of xenobiotic-induced toxicity, including effects on cellular regeneration, but clearly does not respond to the mitogenic effects of rodent liver CAR and PPARα activators (Yamada 2021).

Epidemiology studies

It is stated that the available epidemiology data mostly concerns PB and that human relevance analysis should be conducted taking into account both toxicodynamic and kinetic data. PB has been used as a sedative, hypnotic and antiepileptic drug in humans for several decades in thousands of patients and as such there is now a large epidemiology data set confirming lack of tumor formation in humans (La Vecchia and Negri 2014; Stritzelberger et al. 2021; Yamada et al. 2021). In terms of the toxicodynamic and kinetic data point made by Drs. van Kesteren and colleagues, this is surely more than adequately addressed by PB where many human subjects have received therapeutic doses of PB for extended periods which provide similar blood levels in humans to those which produce liver tumors in mice (Monro 1993). In addition to PB, epidemiological data are also available for oxazepam and carbamazepine, which are known rodent liver CAR activators. Like PB, these two therapeutic agents have been administered over extended periods to many human subjects without any evidence of increased liver tumor formation (Yamada et al. 2021).

Conclusions

We consider that the data presented in our paper (Yamada et al. 2021) is an accurate reflection of the available literature. There are sufficient data to confirm the validity of either cultured human hepatocytes or chimeric mice with human hepatocytes as suitable test systems for the evaluation of species differences in the effects of rodent liver CAR activators. The literature contains many examples of both in vitro and in vivo studies where CAR and PPARα activators have been shown to stimulate RDS in mouse and rat hepatocytes, but not in hepatocytes from other species including humans. There is a clear species difference in rodent CAR and PPARα receptor regulation, where increased cell proliferation is only observed in mouse and rat liver. The established rodent liver tumor MOAs are thus not applicable to humans, this conclusion being supported by substantial epidemiology data for PB, some other CAR activators and hypolipidemic agents (Corton et al. 2018; Yamada et al. 2021).

Overall, we consider that the MOA for rodent liver tumor formation by CAR activators is not qualitatively plausible for humans on the basis of established species differences in hepatocyte proliferation. Rodent liver CAR activators thus do not provide any carcinogenic hazard for humans. Indeed, where appropriate data sets have been obtained, a number of known rodent CAR activators, including pharmaceuticals, agrochemicals, food ingredients, natural substances, and consumer products, have not been classified as carcinogenic agents by regulatory agencies, concluding that this MOA is not relevant to human cancer risk (Yamada 2018; Terry et al. 2021). In addition to this, for many rodent CAR activators, likely human exposure is orders of magnitude below doses which increase hepatocyte proliferation and liver tumor formation in rodents and hence the established MOA for rodent liver tumor formation is also quantitatively not plausible for humans on the basis of likely human exposure.

Acknowledgements

The authors thank members of Environmental Health Science Laboratory, Sumitomo Chemical Company, Ltd. for their encouragement to write this response. All authors gave final approval and agreed to be accountable for all aspects of work in ensuring that questions relating to the accuracy or integrity of any part of the work are appropriately investigated and resolved. An additional review of the response for internal approval purposes was conducted by Kyoko Odawara, a Research Director of Environmental Health Science Laboratory, Sumitomo Chemical Company, Ltd., and approved without any changes to what we had written.

Declaration of interest

Dr. Tomoya Yamada is employed by Sumitomo Chemical Company, Ltd. Prof. Samuel M. Cohen and Prof. Brian G. Lake consult for Sumitomo Chemical Company, Ltd. regarding research on CAR activators as well as on other matters. The response was written as part of the authors’ normal employment, but the authors have sole responsibility for the writing and content of this response. The views presented in this response are those of the authors based on many years of research in the respective areas of investigations reported in this response. This work was supported by the Sumitomo Chemical Company, Ltd. (https://www.sumitomo-chem.co.jp/eng-lish/company/). This company is interested in the human relevance of the CAR-mediated MOA for rodent liver tumor formation, because some company products (e.g. pesticides) have CAR activation activity in rodent liver. Some of the research on CAR activators produced by Sumitomo Chemical Company, Ltd. has been utilized by the company in regulatory submissions to provide MOA information for interpreting the relevance to humans of the liver tumors in rodents. None of the authors has appeared in any legal or regulatory proceedings related to the contents of this response. No potential conflict of interest was reported by the authors.