Human plasma concentrations of herbicidal carbamate molinate extrapolated from the pharmacokinetics established in in vivo experiments with chimeric mice with humanized liver and physiologically based pharmacokinetic modeling

Highlights

• We developed a simple pharmacokinetic (PBPK) model of molinate.

• Rat or mouse PBPK model of molinate was scaled to human model to predict in humans.

• Molinate elimination in humans was estimated to be slow compared with those in rats.

• Concentrations of molinate were evaluated by chimeric mice with humanized liver.

• The simplified human PBPK modeling is useful for risk evaluation of molinate.

Abstract

To predict concentrations in humans of the herbicidal carbamate molinate, used exclusively in rice cultivation, a forward dosimetry approach was carried out using data from lowest-observed-adverse-effect-level doses orally administered to rats, wild type mice, and chimeric mice with humanized liver and from in vitro human and rodent experiments. Human liver microsomes preferentially mediated hydroxylation of molinate, but rat livers additionally produced molinate sulfoxide and an unidentified metabolite. Adjusted animal biomonitoring equivalents for molinate and its primary sulfoxide from animal studies were scaled to human biomonitoring equivalents using known species allometric scaling factors and human metabolic data with a simple physiologically based pharmacokinetic (PBPK) model. The slower disposition of molinate and accumulation of molinate sulfoxide in humans were estimated by modeling after single and multiple doses compared with elimination in rodents. The results from simplified PBPK modeling in combination with chimeric mice with humanized liver suggest that ratios of estimated parameters of molinate sulfoxide exposure in humans to those in rats were three times as many as general safety factor of 10 for species difference in toxicokinetics. Thus, careful regulatory decision is needed when evaluating the human risk resulting from exposure to low doses of molinate and related carbamates based on data obtained from rats.

Introduction

Complex, specific, multi-compartment physiologically based pharmacokinetic (PBPK) models for predicting chemical concentrations in various biological fluids in animals and humans can be found in the literature (Edwards and Preston, 2008, McLanahan et al., 2012); however, simple, easy, inexpensive, and/or reliable methods are needed for accurately evaluating chemical toxic risks for humans (McLanahan et al., 2012). We proposed a simple and reliable PBPK model capable of both forward and reverse dosimetry approaches using a three-compartment PBPK model for acrylonitrile (Takano et al., 2010). The developed PBPK model consisted simply of the gut as the chemical absorption compartment, the liver as the metabolizing compartment, and the general circulation as the central compartment.

Furthermore, it was possible to validate the estimates obtained from simplified human PBPK modeling by comparing their results with in vivo experimental results from humanized mice transplanted with human liver cells (Hasegawa et al., 2011, Higuchi et al., 2014, Yamazaki et al., 2012). Recently developed TK–NOG (Hasegawa et al., 2011) mice were treated to express a herpes simplex virus type 1 thymidine kinase (HSVtk) within the livers of severely immunodeficient NOG (non-obese diabetes-severe combined immunodeficiency- interleukin-2 receptor gamma chain-deficient) mice and induced by a non-toxic dose of ganciclovir, and human liver cells were transplanted in the absence of ongoing drug treatment. We recently described the use of humanized mice in combination with PBPK modeling for risk assessment of melengestrol acetate (Tsukada et al., 2013). When relevant and reliable estimates of the internal dose of a compound or a key metabolite are available, the results of toxicology studies can often be better understood and evaluated in terms of the internal dose.

Molinate (S-ethyl hexahydro-1H-azepine-1-carbothioate) is a thiocarbamate herbicide widely used on rice fields; it has been reported previously to result in testicular toxicity after metabolic activation via sulfoxidation (Jewell et al., 1998). A preliminary seven-compartment PBPK model for molinate, intended to extrapolate the reproductive risk to humans, has been reported (Campbell, 2009); this model was validated in the rat and then extrapolated to humans. In Campbell’s model, the pharmacologically active metabolite sulfoxide is generated in the liver and can circulate in the blood. However, the complicated multiple compartments and equations found in traditional PBPK modeling cause severe difficulties when applying the model for many researchers. Simple and reliable methods are still needed to explore the biological significance of a wide range of chemicals, including thiocarbamate herbicides.

The present study established a simplified PBPK model for molinate in humans; the model was based on physiological parameters derived from the literature, coefficients derived in silico, metabolic parameters determined in vitro using relevant liver microsomes, and in vivo experiment-supported PBPK modeling in rats, wild type mice, and mice with humanized liver. The model was able to estimate some accumulation of molinate sulfoxide after multiple molinate doses in humans for the purpose of evaluating the different molinate exposure levels in humans and rodents in comparison with general safety factor of 10 for species difference in toxicokinetics.