2021 Volume 46 Issue 12 Pages 601-609
Epyrifenacil, one of the protoporphyrinogen oxidase (PPO)-inhibiting herbicides, is hepatotoxic in rodents. Previous in vitro assays detected species differences in both kinetics (active hepatic uptake) and dynamics (PPO inhibitory activity) of S-3100-CA, which is a causal metabolite of the hepatotoxicity, suggesting that humans are less sensitive to the epyrifenacil-induced hepatotoxicity than are rats and mice. To elucidate the species differences in the epyrifenacil-induced hepatotoxicity between mice and humans simultaneously, this study fed epyrifenacil to chimeric mice with humanized liver with low replacement index of human hepatocytes. The distribution of S-3100-CA in the liver and subsequent protoporphyrin IX (PPIX) accumulation, an index of PPO inhibition, were compared between human and host mouse hepatocytes using mass spectrometry imaging (MSI) analysis of chimeric liver. The results showed that S-3100-CA and PPIX were significantly colocalized in regions of the liver slice containing host mouse hepatocytes, and thus it was suggested that epyrifenacil had significantly less effect on human livers than mouse livers because of the species differences in both kinetics and dynamics of S-3100-CA. Moreover, the hepatic uptake assay using cryopreserved primary hepatocytes of rats, mice and humans with inhibitors revealed that S-3100-CA is a substrate of organic anion transporting polypeptides (OATPs). These data corroborate the contribution of OATPs to hepatocellular uptake of S-3100-CA, especially in mice, and subsequent PPIX accumulation by more potent S-3100-CA-induced PPO inhibition in mice. MSI analysis of chimeric mice with humanized liver is a useful technique for elucidating species differences in pharmacokinetics and subsequent changes in toxicological biomarkers.
Epyrifenacil (ethyl[(3-{2-chloro-4-fluoro-5-[3-methyl-2,6-dioxo-4-(trifluoromethyl)-3,6-dihydropyrimidin-1(2H)-yl]phenoxy}pyridin-2-yl)oxy]acetate, S-3100) is a non-selective herbicide which works by inhibiting an essential enzyme, protoporphyrinogen oxidase (PPO), in plants. PPO catalyzes the oxidation of protoporphyrinogen IX to protoporphyrin IX (PPIX) in the plant chlorophyll biosynthesis pathway. By inhibition of PPO in plant cells, protoporphyrinogen IX accumulates in the chloroplast membrane and leaks out into the cytoplasm, where PPIX is then generated and accumulates, and in the presence of light, produces oxygen free radicals that destroy the plasma membrane (Dayan and Duke, 2010; Lee et al., 1993).
Mammals such as rodents and humans also have PPO, which catalyzes the above reaction in the heme biosynthesis pathway. It is reported that some PPO inhibitors induce anemia by blocking the production of hemoglobin and erythrocytes and/or cell damage of liver or kidney by porphyrin accumulation in rats, mice and dogs (European Food Safety Authority, 2015; United States Environmental Protection Agency, 2003, 2014). In previous subchronic toxicity studies of epyrifenacil in rats, mice, and dogs, anemia caused by inhibition of heme biosynthesis and hepatotoxicity caused by PPIX accumulation were observed (Matsunaga et al., 2021), as for the above PPO inhibitors. However, there were clear species differences especially in hepatotoxicity; hepatotoxicity was observed at lower dose in mice compared to those in rats and dogs, and hepatotoxicity in dogs was much slighter than in rats and mice. Further in vitro mechanistic studies revealed the species differences in the hepatotoxicity could be explained by the differences in the dynamics and kinetics of S-3100-CA, which is an ester-cleaved metabolite of epyrifenacil (Fig. 1). It is a major metabolite in the rat liver and in the liver microsomes of rats, mice and humans, and has the high potency in PPO inhibition (Sakurai et al., 2021; Matsunaga et al., 2021). The dynamic activity, i.e., PPO inhibitory activity of S-3100-CA, was the highest for mice, followed in order by rats, dogs and humans. About the kinetics, the hepatocyte uptake assay revealed that the active uptake rate for S-3100-CA was higher in mouse hepatocytes than in rat and human hepatocytes (Matsunaga et al., 2021). These species differences in both dynamics and kinetics of S-3100-CA in vitro were well in line with the differences in hepatotoxicity observed in animal studies. The lowest PPO inhibition and hepatic uptake in human hepatocytes suggested that humans are less sensitive to the hepatotoxicity of epyrifenacil than are rats, mice and dogs.
Structures of epyrifenacil and its major metabolite S-3100-CA.
Further investigation of how species differences in the dynamics and kinetics of S-3100-CA affect the human liver in vivo is essential to show the obvious evidence of the species differences in hepatotoxic potential. To estimate the potential of epyrifenacil-induced hepatotoxicity in humans, we planned to use the chimeric mouse with humanized liver (product name: PXB mouse®, hereafter referred as the PXB mouse) as a human model in this study. PXB mice which possess human hepatocytes have come to be used as a human model for the prediction of pharmacokinetic parameters (Sanoh and Ohta, 2014) and the evaluation of chemical-induced hepatotoxicity (Yamada, 2021). Recently, it was reported that PXB mice could be applied to evaluating the critical key events of the PPO inhibition-derived hepatotoxicity, i.e., PPIX accumulation, followed by hepatocellular injury, with subsequent regenerative cell proliferation (Eguchi et al., 2021). In addition, the evaluation of active uptake of xenobiotics via OATPs has been performed using the PXB mouse (Sanoh et al., 2020). Thus, the PXB mouse was considered a useful human model for evaluating the hepatotoxicity via the exposure to epyrifenacil.
Notably, also, PXB mice retain a small number of host mouse hepatocytes. Since the replacement index (RI) of the mouse liver with human hepatocytes is generally high (85–90%, average of supplier’s commercial production, Tateno et al., 2015), the remaining mouse hepatocytes hardly affect the results; however, they can deteriorate the result of analysis when the reactivity of the hepatocytes with a chemical is quite different between the donor human hepatocytes repopulating the mouse liver and the host mouse hepatocytes (Kakuni et al., 2012; Sanoh et al., 2015; Miyamoto et al., 2017). In the case of epyrifenacil, the residual mouse hepatocytes may have reduced the predictability of the dynamic and kinetic consequences because epyrifenacil’s effect was estimated to be much greater in mouse hepatocytes than in human hepatocytes (Matsunaga et al., 2021). To separately evaluate the reactivity of human and residual mouse hepatocytes, we intended to utilize the mass spectrometry imaging (MSI) analysis. We intended to use the MSI technique to visualize the accumulation of a hepatotoxic xenobiotic compound and its endogenous metabolite simultaneously, as the result of the dynamics and kinetics of epyrifenacil. In addition, we planned to use PXB mice with low RI of human hepatocytes in this study to accentuate the species differences in epyrifenacil-induced hepatotoxicity between human and mouse hepatocyte regions on the humanized chimeric liver in MSI analysis. Combination of MSI analysis and PXB mice with low RI enables us to elucidate the species differences in the toxicological key events between humans and mice under the same condition. A similar approach was used in only one article investigating species differences in the amiodarone metabolic rate and subsequent phospholipidosis (Sanoh et al., 2017), but no articles have reported species differences in toxicity due to variation in active hepatic uptake.
According to the previous in vitro studies (Matsunaga et al., 2021) mentioned above, there were species differences in the PPO inhibitory activity and the hepatic uptake rate of S-3100-CA. However, the species difference in the hepatic uptake of S-3100-CA was not adequately investigated. Although it was supposed that OATPs were involved in the hepatic uptake of S-3100-CA (Matsunaga et al., 2021), there were no data to support this hypothesis.
In the present study, liver sections from epyrifenacil-treated PXB mice with low RI were analyzed by MSI analysis to visualize S-3100-CA distribution in liver and subsequently PPIX accumulation for estimation of species differences in epyrifenacil-induced hepatotoxicity between humans and mice. Furthermore, the hepatic uptake assay was conducted on primary cryopreserved hepatocytes with or without OATP inhibitors to estimate the contribution of OATPs to the species difference in the hepatic uptake of S-3100-CA.
Epyrifenacil and S-3100-CA were synthesized in our laboratory. The chemical purity of these compounds was determined to be 96.8% and 99.7% (HPLC), respectively. Phenyl-14C-labeled S-3100-CA was also synthesized in our laboratory and had a specific activity of 4.26 GBq/mmol with radiochemical purity of 97.8% (radio-HPLC). The structures of these compounds are shown in Fig. 1. α-Cyano-4-hydroxycinnamic acid (α-CHCA) was supplied by Merck KGaA (Darmstadt, Germany). PPIX analytical standard was supplied by Frontier Scientific (Logan, UT, USA). Rifampicin was supplied by FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Digoxin was supplied by Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Krebs-Henseleit buffer (KHB), silicone oil, and mineral oil were purchased from Merck KGaA. The hepatocyte isolation kit was supplied by Sekisui XenoTech, LLC (Kansas City, KS, USA). Cryopreserved primary hepatocytes of male SD rats (pool of 8), female SD rats (pool of 12), male CD-1 mice (pool of 16), and female CD-1 mice (pool of 23) were from Sekisui XenoTech, LLC. Cryopreserved primary hepatocytes of humans (pool of 10 males and 10 females) were purchased from Biopredic International (Saint Grégoire, France). Other chemicals were of reagent grade.
Animals and husbandrySeven male PXB mice with low RI of human hepatocytes aged 12–14 weeks were specially prepared and supplied on request from PhoenixBio Co., Ltd. (Hiroshima, Japan). Cryopreserved human hepatocytes of BD195 (a healthy 2-year-old Hispanic female) were used as donor cells for the chimeric mice in this study. The RI values of the PXB mouse livers before shipment were estimated to be 53.0 ± 6.4% by the supplier. Ten host mice (cDNA-uPA/SCID) as a liver-PPIX-level control group were also supplied from PhoenixBio Co., Ltd.
Animals were housed in a barrier-system animal room for 7 days prior to experimentation. Throughout the study, the environmental conditions in the animal room were constant: temperature (22–26°C), relative humidity (40–70%), frequent ventilation (more than 10 times per hr), and 12-hr light-dark cycle (8 a.m. to 8 p.m. light). A commercially available powdered diet for rodents (fortified with vitamin C [300 mg/100 g], sterilized by 60Co [30 kGy] irradiation [CRF-1 with Vitamin C, Oriental Yeast Co., Ltd., Tokyo, Japan]) and filtered tap water were provided ad libitum.
All animal experiments in this study were approved by the Institutional Animal Care and Use Committee of Sumitomo Chemical Co., Ltd., and performed in accordance with The Guide for Animal Care and Use of Sumitomo Chemical Co., Ltd. All experiments using PXB mice and samples collected from PXB mice were approved in accordance with The Guide for Biosafety of Sumitomo Chemical Co., Ltd.
Dietary administration of epyrifenacilAnimals were fed diets containing 0 or 40 ppm epyrifenacil, respectively, for 7 days. Dose levels were determined based on a previously reported 90-day study in mice (Matsunaga et al., 2021). All animals were observed for mortality once a day throughout the study. Body weights were measured on Day 0 (prior to initiation of treatment) and Day 7 (on the day of necropsy). Food consumption was measured over a period of 7 days.
Calculation of RI of human hepatocytes in PXB mouse liversThe RI values in PXB mouse livers were estimated from the human albumin (hAlb) concentration in PXB mouse blood as previously described (Tateno et al., 2015).
Measurement of liver concentrations of epyrifenacil and S-3100-CAAt necropsy, samples of approximately 100 mg of PXB mouse liver were collected and added to 300 µL of methanol, respectively. After extraction using a ball mill and then centrifugation at 16,000 × g for 5 min, the supernatant was collected and subjected to LC/MS/MS analysis. The analysis was performed on a Shimadzu 20A HPLC system (Shimadzu Corporation, Kyoto, Japan) equipped with a CAPCELL PAK MGII column (50 mm × 2.0 mm, particle size 3 µm, SHISEIDO CO., LTD., Tokyo, Japan). The conditions were as follows: mobile phase, 0.1% formic acid in water (A) and acetonitrile (B); flow rate, 1.0 mL/min; gradient conditions, 30% B at 0 min, 76% B at 4.0 min, 100% B at 4.0–4.5 min, and 30% B at 4.5–5.0 min; injection volume, 5 µL; and column oven temperature, room temperature. An LTQ Orbitrap XL system (Thermo Fisher Scientific Inc., Waltham, MA, USA) was used to perform the MS/MS analysis in positive ESI mode. The selective reaction monitoring transition was m/z 518/473 for epyrifenacil and m/z 490/473 for S-3100-CA. The amounts of the analytes in the samples were calculated with the linear calibration line including the analytical standards.
Measurement of liver concentrations of PPIXSamples of approximately 200 mg of liver were collected from all animals at necropsy. PPIX was extracted from the liver sample according to the previously described method (Eguchi et al., 2021).
Tissue sectioningLiver samples were collected from PXB mice at necropsy and stored frozen (−70 to −90°C). Frozen liver tissues were cut with a cryostat (CM3050 S, Leica, Nussloch, Germany) into 10 µm thick sections at −20°C. Sections were thaw-mounted onto amino propyl silane (APS) coated glass slides (Matsunami Glass Ind., Ltd., Osaka, Japan) and dehydrated in 50 mL tubes with silica gel for approximately 30 min at room temperature. Two serial liver sections per animal were cut for MSI analysis of S-3100-CA and PPIX. Sections were stored frozen (−70 to −90°C) until MSI analysis.
Desorption electrospray ionization (DESI) -MSI analysis for S-3100-CADESI-MSI analysis was performed using the Synapt G2-Si HDMS system (Waters Corporation, Manchester, UK) equipped with the 2D DESI ion source (Prosolia, Inc., Indianapolis, IN, USA), operated in negative ion sensitivity mode. S-3100-CA distribution on the liver sections was measured at the transition m/z 488.03/488.03 using methanol/water (95:5, v/v) containing 10 mM ammonium acetate as a spray solvent delivered at 2 µL/min. The spectra were acquired at trap collision energy 10 eV, sprayer voltage 4.5 kV, nitrogen nebulizing gas pressure 4.0 bar, surface scan rate 200 µm/sec, and spatial resolution 100 µm. The acquired spectra were analyzed with the use of High Definition Imaging software (Waters Corporation).
Matrix assisted laser desorption/ionization (MALDI) -MSI analysis for PPIXInitially, α-CHCA for the MALDI matrix was sublimated using iMLayer vacuum deposition system (Shimadzu Corporation). α-CHCA was heated at 250°C and deposited on the tissue surface at a 0.7 µm thickness. MALDI-MSI analysis was performed using the MALDI Synapt G2-Si HDMS system equipped with an Nd: YAG laser, operated in positive ion sensitivity mode. PPIX distribution on the liver section was measured at the transition m/z 563.3/445.2 with trap collision energy 41 eV, scan time 0.5 sec, spatial resolution 130 µm, laser frequency 2500 Hz, and laser energy 350. The acquired spectra were analyzed with the use of High Definition Imaging software.
Tissue section stainingTo distinguish the human and mouse hepatocytes areas, hematoxylin-eosin (HE) staining was performed. After MALDI-MSI or DESI-MSI analysis, the sections were fixed with ice-cold acetone/methanol (1:1) for 10 min, dried, and finally stained with HE as previously described (Eguchi et al., 2021). The regions of both mouse and human hepatocytes were distinguished in accordance with the pathological criteria showed in the previous literatures (Tateno et al., 2004; Eguchi et al., 2021). These criteria were validated by immunohistochemical staining with anti-human STEM121 mouse monoclonal antibody using serial sections (Fig. S1, Supplemental data).
Hepatocyte uptake assayHepatocyte uptake assay was conducted to investigate the effect of the OATP inhibitors, rifampicin and digoxin, on S-3100-CA uptake using the previously described method with slight modifications (Matsunaga et al., 2021). Briefly, the uptake assay was started by adding an equal volume of [phenyl-14C]S-3100-CA solution (final concentration: 5 µM) including rifampicin or digoxin (final concentration: 0, 10, or 100 µM) in KHB containing 2% methanol and dimethyl sulfoxide. The uptake was terminated at 1.0 min by the oil spin method. The assay was conducted in triplicate.
Calculation of hepatic uptake volumeThe total radioactivity in the cell lysate (dpm/105 cells) was divided by the radioactivity concentration in culture solution (dpm/mL) to obtain the volume of substrate uptake in hepatocytes (µL/106 cells). The active uptake volume was defined as the difference between the uptake volume at 37°C and the uptake volume at 4°C.
The RI of PXB mice with low RI was estimated to be 63.6 ± 12.8% based on hAlb concentration at necropsy while that was 53.0 ± 6.4% before shipment.
Hepatic level of epyrifenacil, S-3100-CA, and PPIX in epyrifenacil-treated PXB mice with low RIThe results of LC/MS/MS analysis are summarized in Table 1. Epyrifenacil was not detected in liver samples. S-3100-CA and PPIX levels in the liver of PXB mice with low RI, after feeding 40 ppm of epyrifenacil for 7 days, were 1577 ± 500.8 ng/g wet weight liver and 3010 ± 1774 ng/g wet weight liver, respectively. After 40 ppm epyrifenacil treatment, the liver PPIX level of PXB mice was significantly increased to 28.3-fold that of the control animals (106.3 ± 18.57 ng/g wet weight liver).
| Liver concentration [ng/g wet weight liver] | Control | Epyrifenacil 40 ppm | ||||
|---|---|---|---|---|---|---|
| Epyrifenacil | N.A. | N.D. | ||||
| S-3100-CA | N.A. | 1577 | ± | 500.8 | ||
| PPIX | 106.3 | ± | 18.57 | 3010 | ± | 1774 |
Values are expressed as the mean ± standard deviation of the control group (n = 10) and epyrifenacil treatment group (n = 7).
Host mice were used as control animals for the measurement of PPIX.
N.A.: Not applicable.
N.D.: Not detected.
Histological images of liver sections and MSI images from an epyrifenacil-treated PXB mouse with low RI are shown in Fig. S2 (Supplemental data) and representative pictures are presented in Fig. 2. S-3100-CA (Fig. 2B) and PPIX (Fig. 2E) were detected in liver sections and colocalized in similar regions of serial liver slices. To distinguish regions of mouse hepatocytes from regions of human hepatocytes, the slices after MSI analysis were stained with HE (Fig. 2A and 2D) HE staining revealed that S-3100-CA and PPIX were localized in the areas of host mouse hepatocytes. The same results were obtained in all animals (Fig. S2, Supplemental data). According to the results of HE staining, the areas of human hepatocytes could be distinguished from areas of mouse hepatocytes (Fig. 2C and Fig. 2F) and the relative amounts of S-3100-CA and PPIX in each hepatocyte region were quantified. According to the quantification results, the amounts of S-3100-CA and PPIX in the human hepatocytes region were 3–4 times lower than those in the host mouse hepatocytes region (Fig. 3).
Comparison between MS and the histological images of liver sections obtained from epyrifenacil-treated PXB mice with low RI. (A) The histological image of HE-stained sections after MSI analysis of S-3100-CA: human hepatocyte regions show clearer cytoplasm compared to mouse hepatocyte regions. (Scale bar = 1 mm) (B) The MS image of S-3100-CA: the scale bar indicates the range of target compound signal intensity. (C) The human (blue) and mouse (red) hepatocyte regions used for the relative quantification of S-3100-CA in liver sections. (D) The histological image of an HE stained section obtained after MSI analysis of PPIX. (E) The MS image of PPIX: the scale bar indicates the range of target compound signal intensity. (F) The human and mouse hepatocyte regions used for the relative quantification of PPIX.
Relative amounts of the analytes in mouse (open bar) and human (shaded bar) hepatocyte regions according to MSI analysis, expressed as fold differences vs the amounts in human hepatocytes.*: p < 0.05 significantly higher compared with the value for the human hepatocyte region.**: p < 0.01 significantly higher compared with the value for the human hepatocyte region.
The results of active uptake volume and the normalized uptake volume (% of control) of [phenyl-14C]S-3100-CA into rat (each sex), mouse (each sex), and human (both genders) hepatocytes with or without the OATP inhibitors, rifampicin and digoxin, are summarized in Fig. 4. All active uptake volumes calculated by subtracting the uptake volumes of S-3100-CA at 4°C from those at 37°C produced positive values (Fig. 4A). Therefore, it was confirmed that S-3100-CA is taken up into hepatocytes by the hepatic transporters in all tested species. The largest active uptake volume in the controls (i.e. without inhibitors) was provided in male mice, followed in order by female mice, female rats, male rats, and humans. The contribution of OATP to the hepatic uptake of S-3100-CA was confirmed by the uptake assay using the inhibitors, rifampicin and digoxin. The active uptake volume was decreased by rifampicin in a concentration-dependent manner in all species tested. As shown in Fig. 4B, the normalized uptake volume with 100 µM rifampicin decreased to 17–49% of that in control. The inhibition by rifampicin was the strongest in mice (17–20% of control), followed in order by rats and humans. Although the active uptake volume was also decreased by another inhibitor of OATPs, digoxin, the inhibitory activity was lower than that of rifampicin: 65–78% by 100 µM digoxin.
Absolute (A) and normalized (B) uptake volumes of S-3100-CA into hepatocytes of mice, rats, and humans at 1.0 min with or without the OATP inhibitors rifampicin (RIF) and digoxin (DIG).
In the present study, chimeric mice with humanized liver were treated with one of the PPO inhibitors, epyrifenacil, for 7 days to estimate species differences in epyrifenacil-induced hepatotoxicity between humans and mice. The effects of epyrifenacil on human hepatocytes were distinguished from those on residual mouse hepatocytes in the chimeric liver using MSI analysis.
The PXB mouse liver is normally repopulated with human hepatocytes that exhibit a high RI (normally 85–90%), and thus host mouse hepatocytes remain at a low level (Tateno et al., 2015). To accentuate the species differences in epyrifenacil-induced hepatotoxicity between human and mouse hepatocyte regions on the humanized chimeric liver in MSI analysis, PXB mice with low RI of human hepatocytes (approx. 50–60%) were planned to be used for this study intentionally, which means host mouse hepatocytes were at a relatively high level. The results of the RI at necropsy confirmed that the host mouse hepatocytes were considered remaining at a high level during the treatment period as intended.
Previous epyrifenacil metabolism study in rats (Sakurai et al., 2021) reported that epyrifenacil is mainly metabolized to S-3100-CA in the plasma and the liver. In addition, other in vitro studies (Matsunaga et al., 2021) revealed that S-3100-CA is also predominant in liver microsomes of rats, mice and humans, and it causes PPO inhibition in the liver mitochondrial fraction. From these results, it was concluded that the main causal metabolite for epyrifenacil-induced hepatotoxicity is S-3100-CA. It was also reported that the inhibition of PPO induced the accumulation of PPIX, which is an endogenous heme biosynthesis-related compound, and subsequently cell damage in the liver (Smith and Foster, 2018; Holsapple et al., 2006). Therefore, we considered that S-3100-CA and PPIX levels are suitable indices of epyrifenacil-induced hepatotoxicity. After epyrifenacil administration to PXB mice with low RI, epyrifenacil, S-3100-CA and PPIX levels in liver homogenates were measured by LC/MS/MS analysis. In all treated animals, no epyrifenacil was detected. This supports the results from in vitro metabolism test that S-3100-CA is a predominant Phase I metabolite also in human livers as in rats and mice (Matsunaga et al., 2021). In the epyrifenacil-treated chimeric liver homogenates, there was significant PPIX accumulation compared with the livers of non-treated control mice (28.3-fold, Table 1). It is indicated that 7-day epyrifenacil treatment induced hepatic PPO inhibition in PXB mice with low RI. However, the S-3100-CA level and PPIX accumulation ratio in chimeric livers were significantly lower than wild-type (CD-1) mice fed epyrifenacil at the same dosage level (S-3100-CA; 11,764.2 ng/g wet weight liver, PPIX; 43-fold increase compared to non-treated animals, data not shown). It is considered that the decreases in liver exposure to S-3100-CA and PPIX accumulation compared to the wild-type mice reflected the species differences in the hepatic uptake of S-3100-CA (i.e., kinetics) and the PPO inhibitory activity (i.e., dynamics) as reported in the previous in vitro studies (Matsunaga et al., 2021).
As above, LC/MS/MS analysis of epyrifenacil-treated PXB mice with low RI revealed the chimeric liver was less sensitive to epyrifenacil than the liver of wild-type mice. To separately estimate the effect on the amounts of S-3100-CA and PPIX in human or mouse hepatocytes in the chimeric liver sections of epyrifenacil-treated PXB mice with low RI, MSI analysis was conducted. As PXB mice with low RI have livers with both human and mouse hepatocytes, the MSI analysis enabled us to simultaneously compare the effects of epyrifenacil on toxicological key events between human and mouse hepatocytes under the same conditions. From the results of HE staining after MSI analysis, we concluded that the hepatocytes of each species could be distinguished clearly and were localized in a mottled pattern (Fig. 2A and 2D). This showed that human and host mouse hepatocytes are mixed in the liver sections to the degree indicated by the RI value calculated from the hAlb measurement. MSI analysis revealed that S-3100-CA and PPIX on serial HE-stained liver slices were colocalized in the host mouse hepatocyte regions (Fig. 2B and 2E). The relative quantification results confirmed that the each compound was significantly localized in mouse hepatocyte areas rather than human hepatocyte areas (Fig. 3). It is considered that the high hepatic uptake of S-3100-CA and PPIX accumulation in epyrifenacil-treated PXB mice with low RI was due mainly to mouse hepatocytes rather than human hepatocytes, because mouse hepatocytes have a higher uptake activity for S-3100-CA and higher sensitivity to S-3100-CA-induced PPO inhibition, as reported previously (Matsunaga et al., 2021). Although clear species differences in the main two key events of epyrifenacil-induced hepatotoxicity were investigated in this study, it is also important to assess whether there are quantitative differences in the subsequent key events (i.e. hepatocyte injury followed by regenerative proliferation) between humans and mice. In previous research, it is reported that PXB mouse is a useful humanized in vivo animal model to evaluate these key events in humans (Eguchi et al., 2021), and therefore further investigation using PXB mice with higher RI (more than 80–85%) as a human model, is highly expected.
In the present and previous studies (Matsunaga et al., 2021), species differences in the kinetics and dynamics of S-3100-CA were indicated. The species difference in the PPO inhibitory activity was considered to be caused by the variance in the binding affinity of the substrate to PPO as previously reported (Arakawa et al., 2017). On the other hand, there was no evidence explaining the active hepatic uptake. S-3100-CA contains a carboxylic acid moiety (Fig. 1), which is highly likely to be ionized and exist as a COO- anion under physiological conditions. It is known that such anions are sometimes taken up into organs via organic anion transporters (OATs) or OATPs (Li et al., 2019). Moreover, a previous study reported that the S-3100-CA uptake rate in hepatocytes of male mice was significantly higher than that in hepatocytes of female mice (Matsunaga et al., 2021). In a previous article, the hepatic mRNA levels of Oat1, Oat2, and Oat3 were reported to be similar between male and female in mice (Buist and Klaassen, 2004). On the other hand, the mRNA expression of Oatp1a1 is more than 3 times higher in male mice than in female mice (Fu et al., 2012). Therefore, we focused on the OATPs that cause species and gender differences in the hepatic uptake of S-3100-CA.
To evaluate the contribution of OATPs to the hepatic uptake of S-3100-CA in each sex and/or species hepatocyte population, the hepatic uptake assay was conducted with the known OATP inhibitors, rifampicin and digoxin. The results of the inhibition assay showed that both OATP inhibitors decreased the S-3100-CA uptake in a concentration-dependent manner in all species tested, suggesting that S-3100-CA is a substrate of OATPs in rats, mice, and humans (Fig. 4A). The inhibitory activity of rifampicin was higher than that of digoxin in all species tested. Rifampicin is known to inhibit OATP family-mediated uptake extensively (Vavricka et al., 2002) while digoxin selectively inhibits Oatp1a4-mediated uptake in rodents (Ishida et al., 2018). Therefore, it was suggested that the contribution of Oatp1a4 to hepatic uptake of S-3100-CA is not predominant in rodents. Comparing the active uptake volumes for rifampicin among the species, the inhibition by 100 µM rifampicin was the largest in mice, followed in order by rats and humans, as shown in Fig. 4B. The extent of inhibition by rifampicin corresponds to the species differences in the hepatic uptake rate of S-3100-CA in each hepatocyte population (i.e., mice > rats > humans). It is reported in some of the literature that the mRNA expression level of Oatp1a1, one of the major OATPs in the liver, is higher in the livers of male than female mice (Buist and Klaassen, 2004; Fu et al., 2012), which is consistent with the gender difference in the hepatic uptake rate of S-3100-CA in mice observed in this and previous studies. Some papers have described the expression levels of liver OATPs in humans and mice. A study comparing the protein expression levels of OATPs between humans and mice reported that the plasma membrane protein level of Oatp1a1 in male mouse liver (Kamiie et al., 2008) was 7 times or 8 times higher than that of the predominant OATPs in the human liver, OATP1B1, OATP1B3, and OATP2B1 (Ohtsuki et al., 2012). The substrate specificity of the liver OATPs for S-3100-CA should be further investigated, but on the basis of currently available data, it was suggested that the hepatic uptake of S-3100-CA by OATPs which was derived from the expression levels of OATPs in livers might contributes to the species differences in the liver exposure of S-3100-CA between rodents and humans.
In conclusion, MSI analysis when applied to samples from epyrifenacil-treated PXB mice with low replacement index revealed that S-3100-CA and PPIX were mainly distributed in host mouse hepatocytes, which corroborates the obvious species differences in hepatocellular uptake of S-3100-CA and PPIX accumulation due to PPO inhibition between humans and mice. Furthermore, the hepatic uptake assay with the inhibitors of OATPs revealed that S-3100-CA is the substrate of OATPs in humans and rodents and one of the reasons for the species difference in liver exposure to S-3100-CA. This is the first report to evaluate the species differences in toxicity due to variation in active hepatic uptake using the combination of chimeric animals and mass spectrometry imaging technologies.
The authors are grateful to Dr. Chise Tateno and other contributors from PhoenixBio Co., Ltd. for scientific advice and assistance. We would like to extend our thanks to Ms. Keiko Tanaka for technical support and Ms. Naoko Fujita for research advice. We also thank the other contributors to this research project from Sumitomo Chemical Company, Ltd.
Conflict of interestThe authors declare that there is no conflict of interest.
