Abstract
Fetal rat anemia from flumioxazin, an N-phenylimide herbicide, is caused by suppression of heme synthesis resulting from inhibition of protoporphyrinogen oxidase (PPO). A series of studies to investigate the effects of flumioxazin have revealed that developmental toxicity is caused in rats but not in rabbits, and the adverse effects are not likely to occur in humans. In this study, as a final weight-of-evidence approach for assessing the human safety of flumioxazin, we compared the toxic potential of inhibition of heme synthesis leading to anemia between human and rat embryonic erythroid cells, which were degenerated as the target of flumioxazin in the rat developmental toxicity. To obtain embryonic erythroid cells, we established respective differentiation methods for embryonic erythroid cells from both human and rat pluripotent stem cells. Derived human and rat embryonic erythroid cells were treated with flumioxazin or dihydroartemisinin (DHA), an anti-malarial drug that causes reduction of embryonic erythroid cells and leads to anemia without species differences. In the human embryonic erythroid cells, DHA inhibited cell proliferation and heme synthesis, whereas there were no effects on heme content or cell proliferation with flumioxazin. In the rat embryonic erythroid cells, however, a dose-related reduction in heme synthesis occurred with treatment of flumioxazin and of DHA. These results confirmed that flumioxazin has no effect on heme synthesis in human embryonic erythroid cells. The present data were in accordance with the results of previous studies and demonstrated that there are no concerns in humans regarding the developmental toxicity of flumioxazin observed in rats.
INTRODUCTION
Flumioxazin is an N-phenylimide herbicide that is widely used for controlling annual broadleaf weeds. This herbicide has developmental toxicity in rats, but not in rabbits, including embryonic lethality and teratogenicity (mainly ventricular septal defects [VSDs] and wavy ribs). The mechanism of its effects has been well elucidated. Flumioxazin inhibits protoporphyrinogen oxidase (PPO), which is a key enzyme in chlorophyll and heme biosynthesis and is also an herbicidal target. PPO inhibition causes embryonic anemia by suppressing heme production and producing erythroblastic degeneration in rat embryos. As a compensatory reaction to anemia, stroke volume is increased, which leads to the enlargement of heart, and then VSDs are produced by mechanical distortion of the heart or abnormal cardiac hemodynamics (Kawamura et al., 1995, 1996a, 1996b, 2014, 2016; Abe et al., 2018).
To investigate the human relevancy of the toxicity in rats, several studies were reported. In vitro PPO inhibition assays using rat, rabbit, monkey, and human hepatocytes revealed that the rat cells were more sensitive to PPO inhibition by flumioxazin than the human cells, while those of the rabbits and monkeys were almost insensitive (Abe et al., 2018). Additionally, cell-based assays with human erythroleukemia (K562), human cord blood (CD36+), and rat erythroleukemia (REL) cells demonstrated that heme synthesis in two types of human cells was invulnerable to flumioxazin, while heme production in rat cells with hemoglobin-synthesizing was reduced (Kawamura et al., 2021). These studies suggest that flumioxazin seems not to cause developmental toxicity in humans, unlike in rats. However, it has remained unclear whether the herbicide affects the heme synthesis in human embryonic erythroid cells, which is equivalent to the cells degenerated by flumioxazin in rat embryo, due to a lack of access to normal human cells in early development.
In mammals, hematopoietic development comprises “primitive” and “definitive” phases (Palis and Segel, 1998). Regarding erythroids, the major difference between the two phases lies in the globin subtypes. In “primitive” erythropoiesis, the yolk sac transiently generates embryonic erythroid cells. Primitive erythroid cells mainly express embryonic-specific ε-globin with ζ-globin. Definitive erythropoiesis then occurs in the aorta-gonad-mesonephros (AGM) region. During “definitive” hematopoiesis, the main region of erythropoiesis moves to the fetal liver and finally to bone marrow. The fetal definitive erythroids express fetal γ-globin chains with α-globin. After birth, adult erythroid cells generated in bone marrow express adult β-globin with α-globins (Demirci and Tisdale, 2018). In humans, the availability of primitive erythroid cells is largely limited, as they are difficult to obtain from a spontaneous or therapeutic abortus. Although the differentiated human K562 cells used in the latest study (Kawamura et al., 2021) were reported to express embryonic ε-globin, K562 cells are derived from definitive origin and leukemia patients, namely tumor cells (Cioe et al., 1981). Whereas the human CD36+ cells are non-cancer cells derived from human cord blood, the cells are not likely to be embryonic, because they were reported to express fetal γ-globin and adult β-globin (Scicchitano et al., 2003). Therefore, whether erythroid cells in human embryos are susceptible to flumioxazin has not been fully demonstrated.
Recently, pluripotent stem cells (PSCs), such as embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs), have become a promising source of normal human cells for in vitro tests, including phototoxicity assay (Takahashi et al., 2007; Mori et al., 2017). PSCs have potential to proliferate infinitely and to differentiate any types of cells under specific differentiation conditions. PSCs also enable us to obtain the cells at a developmental stage, because the differentiation process of PSCs reflects normal development. Therefore, PSC-derived cells can be a powerful tool to elucidate the adverse outcome pathway of chemicals that are harmful to the cells in developmental stage or in the inaccessible organs.
Several methods for the erythroid differentiation from human PSCs have been reported. These methods were mainly based on the coculture with feeder cells (Kaufman et al., 2001; Vodyanik et al., 2005; Olivier et al., 2006; Qiu et al., 2008) such as stromal cells of the OP9 murine bone marrow stromal line, or the formation of embryoid bodies (EBs) (Zambidis et al., 2005; Chang et al., 2006; Olivier et al., 2016). These methods were used to elucidate the mechanism of human organogenesis in embryology (Qiu et al., 2008); however, to our knowledge, no study has assessed developmental toxicity using human PSC-derived embryonic erythroid cells.
In this study, to evaluate the human developmental toxicity of flumioxazin, we established respective methods for differentiating both human iPSCs (hiPSCs) and rat ESCs (rESCs) into embryonic erythroid cells expressing embryonic-specific ε-globin. We then compared the effects on heme synthesis between species when treated with flumioxazin or the anti-malarial medicine dihydroartemisinin (DHA), which was used as a positive control of inhibition of heme synthesis preceding anemia and VSDs in rats due to its similar effects with flumioxazin (Clark et al., 2004, 2008; Finaurini et al., 2010, 2012; Kawamura et al., 2021).
MATERIALS AND METHODS
Test chemicals
Flumioxazin was supplied by Sumitomo Chemical Co., Ltd. (Tokyo, Japan). Dihydroartemisinin (DHA) was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Dimethyl sulfoxide (DMSO) was purchased from Sigma-Aldrich Co., Ltd. (St. Louis, MO, USA).
Generation of embryonic erythroid cells from hiPSCs
The hiPSCs (HC-6#10 cell line; Ishida et al., 2016) were kindly provided by Dr. Muguruma (RIKEN Center for Developmental Biology, Hyogo, Japan). The experimental protocols for handling of human subjects were approved by the institutional committees of Sumitomo Chemical Co., Ltd. The hiPSCs were maintained on iMatrix-511 (Nippi, Incorporated, Tokyo, Japan)-coated 6-well plates in StemFit AK02N medium (Ajinomoto Co., Inc., Tokyo, Japan) in accordance with the published protocol with some minor modifications. Briefly, we used Accumax (Innovative Cell Technologies, Inc., San Diego, CA, USA) for the dissociation of undifferentiated hiPSCs instead of 0.5 × TrypLE Select (Thermo Fisher Scientific, Inc., Waltham, MA, USA). For erythroid differentiation, the serum-free floating culture of embryoid body-like aggregates with quick reaggregation (SFEBq) culture method (Nakano et al., 2012; Kuwahara et al., 2015) was used with modifications. The hiPSCs were dissociated into single cells with Accumax containing 20 μM Y-27632 (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) and then placed in low-cell-adhesion 96-well plates with V-bottomed conical wells (Sumilon PrimeSurface plate; Sumitomo Bakelite Co., Ltd., Tokyo, Japan) at 3000 cells/well, quickly re-aggregated in STEMdiff Hematopoietic Basal Medium containing 0.5% STEMdiff Hematopoietic Supplement A (STEMCELL Technologies Inc., Vancouver, Canada) and 20 μM Y-27632 (EDM-A; erythroid differentiation medium A), and centrifuged (400 × g, 5 min) to enhance the formation of cell aggregates and subsequent EBs. On day 3 of differentiation, the EBs were transferred from 96-well plates to 6-well plates (Corning Inc., Corning, NY, USA) coated with Matrigel (Growth factor reduced; Corning Inc.) at 5 EBs/well in STEMdiff Hematopoietic Basal Medium containing 0.5% STEMdiff Hematopoietic Supplement B (STEMCELL Technologies Inc.), human recombinant erythropoietin (hEPO; 30 ng/mL, GenScript USA Inc., Piscataway, NJ, USA), and 20 μM Y-27632 (hEDM-B; human erythroid differentiation medium B) for the adherent culture. On day 7, fresh hEDM-B was added at 0.5 mL per well. On day 10, non-adherent floating erythroid cells were collected and used in the experiments that followed.
Generation of embryonic erythroid cells from rESCs
The rESCs (D6 cell line) were obtained from DS Pharma Biomedical Co., Ltd. (Osaka, Japan). The cells were established according to their published methods (Ueda et al., 2008; Kawamata and Ochiya, 2010a, 2010b). The rESCs were maintained on a feeder layer of mouse embryonic fibroblasts (MEFs) inactivated by mitomycin C treatment (Reprocell Inc., Kanagawa, Japan) in StemMedium (DS Pharma Biomedical Co., Ltd.) supplemented with A-83-01 (0.5 μM, FUJIFILM Wako Pure Chemical Corporation), CHIR99021 (3 μM, Cayman Chemical Company, Ann Arbor, MI, USA), Y-27632 (20 μM, FUJIFILM Wako Pure Chemical Corporation), 2-mercaptoethanol (0.1 mM, FUJIFILM Wako Pure Chemical Corporation), Antibiotic-Antimycotic (1%, Thermo Fisher Scientific, Inc.), and rat LIF (ESGRO; 2000 U/mL, Millipore, Burlington, MA, USA). For erythroid differentiation, non-adherent colonies of undifferentiated rESCs were collected and dissociated into single cells with Accutase (Innovative Cell Technologies, Inc., San Diego, CA, USA) and seeded to 96-well V-bottomed plates in EDM-A as described above. On day 3, the EBs were transferred onto feeder layers of OP9 cells, a murine bone marrow stromal line obtained from American Type Culture Collection (ATCC), with Matrigel-coated 6-well plates in rat erythroid differentiation medium B (rEDM-B; hEDM-B supplemented with Y-27632 [20 μM], rat stem cell factor [rSCF; 30 ng/mL, R&D Systems, Inc., Minneapolis, MN, USA], and rat erythropoietin [rEPO; 30 ng/mL, R&D Systems, Inc.]). OP9 cells were maintained and prepared as previously described (Fujita et al., 2016) and mitotically inactivated by mitomycin C treatment. On day 5, the medium was completely changed to fresh rEDM-B supplemented with hydrocortisone (1 μM, FUJIFILM Wako Pure Chemical Corporation), rat recombinant interleukin-3 (IL-3; 10 ng/mL, R&D Systems, Inc.), human recombinant IL-11 (IL-11; 20 ng/mL, R&D Systems, Inc.), human recombinant insulin-like growth factor-1 (IGF-1; 20 ng/mL, R&D Systems, Inc.), and rEPO (60 ng/mL). On day 8, non-adherent floating erythroid cells were collected and used in the experiments that followed.
Cell cultures
K562 cells, human erythroleukemia cells, were obtained from the Human Science Research Resources Bank (Osaka, Japan). Rat erythroleukemia (REL) cells were kindly gifted by Dr. Sonja Pavlović (University of Belgrade, Belgrade, Republic of Serbia; Pavlović et al., 1999). K562 and REL cells were maintained in Roswell Park Memorial Institute 1640 (RPMI) medium supplemented with 10% fetal bovine serum (FBS; Corning Inc.) and 100 IU/mL penicillin, 100 μg/mL streptomycin (1% P/S; Sigma-Aldrich Co., Ltd.). The induction to heme-producing cells from K562 cells and REL cells was conducted as previously described (Kawamura et al., 2021). Briefly, K562 cells were cultured in the maintenance medium supplemented with 1.0 mM sodium butyrate (NaB; Sigma-Aldrich Co., Ltd.) and REL cells were differentiated in RPMI medium containing 40% FBS, 1% P/S, and 1.0 mM hexamethylene bisacetamide (HMBA) on low-cell-adhesion 60-mm dishes (Sumitomo Bakelite Co., Ltd.) at 5.0 × 105 cells/dish. On days 0, 2, 4, and 8, the cells were harvested and analyzed.
Treatment of PSCs-derived erythroid cells with flumioxazin or DHA
The effects of flumioxazin and DHA were tested in a heme synthesis assay system (Fig. 7, Fig. 8). DHA is an antimalarial drug and also the major active metabolite of antimalarial drug artemisinin derivatives. DHA was used as a positive control of heme synthetic inhibition, since DHA causes embryonic red cell depletion without species difference as tested (Clark et al., 2004; 2008). The hiPSCs-derived erythroid cells were collected on day 10 and resuspended in hEDM-B containing either 0.1% DMSO, flumioxazin (1 μM or 5 μM), or DHA (0.5 μM or 1 μM). For the rESCs-derived erythroid cells, cells were harvested on day 8 and resuspended in Stemline II (Sigma-Aldrich Co., Ltd.) supplemented with rEPO (60 ng/mL), rSCF (30 ng/mL), and Y-27632 (20 μM) (rEDM-D) containing either 0.1% DMSO, flumioxazin (1 μM or 5 μM), or DHA (0.5 μM). The cells were plated onto 6-well cell culture plates (Corning Inc.) at 2.0 × 105 cells/1.5 mL/well. On day 14 of hiPSCs differentiation or day 12 of rESCs differentiation, 4 days after exposure, the floating cells were harvested. The number of living cells (i.e., cells not stained by trypan blue solution; Thermo Fisher Scientific, Inc.) was counted using a hemocytometer (Erma, Inc., Saitama, Japan), and cell viability was calculated. The pellets of 0.3 to 1.0 × 105 cells were obtained by centrifugation and used for the analyses that followed.
Immunohistochemistry
For immunofluorescence analyses, the floating cells were fixed in 4% paraformaldehyde (FUJIFILM Wako Pure Chemical Corporation) for 15 min at room temperature and centrifuged onto coated-glass slides (Thermo Fisher Scientific, Inc.) using a Cytospin4 centrifuge (Thermo Fisher Scientific, Inc.). The cells were permeabilized with 0.2% Triton X-100 (FUJIFILM Wako Pure Chemical Corporation) in tris-buffered saline (pH 7.6; Takara Bio, Inc., Shiga, Japan) for 15 min and blocked in Superblock Blocking Buffer (Thermo Fisher Scientific, Inc.) for 30 min or more. Thereafter, the cells were incubated overnight at 4°C with primary antibodies diluted with Antibody Diluent OP Quant (Thermo Fisher Scientific, Inc.). The primary antibodies were anti-ε-globin (rabbit/1:750/GeneTex, Inc., Taipei, Taiwan) and anti-Glycophorin A (GPA; goat/1:100/Novus Biologicals, LLC, Centennial, CO, USA). The cells were then incubated for 1 hr with two secondary antibodies: anti-rabbit IgG (H+L) Alexa Fluor 488 (donkey/1:1000/Thermo Fisher Scientific, Inc.) and anti-goat IgG (H+L) Alexa Fluor 488 (donkey/1:1000/Thermo Fisher Scientific, Inc.). Counter nuclear staining was performed with Hoechst 33342 staining dye solution (Dojindo Laboratories, Kumamoto, Japan). Images were acquired with a fluorescence microscope (Axioimager M2, Carl Zeiss AG, Oberkochen, Germany).
Quantitative RT-PCR
The mRNA expression levels of the β-like globin gene (embryonic ε-globin, fetal γ-globin, adult β-globin) were analyzed using quantitative RT-PCR (qRT-PCR). Total RNA was extracted from the cells using the RNeasy micro kit and cDNA was synthesized using Superscript III Reverse Transcriptase. qRT-PCR analyses were performed with TaqMan Fast Advanced Master Mix and gene-specific TaqMan Gene Expression Assays using the StepOnePlus Realtime PCR system. For comparisons of the expression levels of β-like globin mRNA, absolute quantification was performed. The copy number of each mRNA was calculated from standard curves generated using dilutions of PCR amplicons that were amplified with primers designed to be outer region of the target sequence of each TaqMan assay primer. All materials were obtained from Thermo Fisher Scientific, Inc., and experiments were performed according to the manufacturer’s instructions. The product numbers of TaqMan assays and the sequences of primers for amplicons are listed in Supplemental Table 1.
Extraction and purification of PPIX and heme from cells
The protoporphyrin IX (PPIX) accumulation can be used as an indicator of PPO inhibition, since its substrate protoporphyrinogen IX leaks out from the mitochondria and is converted to PPIX by cytosolic enzymes when the PPO activity is inhibited (Kawamura et al., 1996a; 2014). The inhibition of heme was measured as an indicator of anemia. PPIX and heme were extracted according to previous methods (Kawamura et al., 1996a; Abe et al., 2018). Internal standards for PPIX (10 ng/mL of protoporphyrin IX-d4 [PPIX-d4]) and heme (500 ng/mL of deuteroporphyrin IX [DP]) in a total volume of 700 µL of basic methanol (methanol/0.1 M ammonia solution = 9/1, v/v) were added to the tubes. The cell suspensions were then vortexed for 30 sec, sonicated for 30 sec, vortexed again, and centrifuged at 20,400 × g, at 4°C for 15 min. Obtained supernatants were purified using the Oasis MAX µElution Plate (Waters Corporation, Milford, MA, USA). The purification plate was equilibrated sequentially with 400 µL each of methanol and water, and then the prepared samples were loaded into the wells. After being washed with 400 µL of 5% ammonia solution, 200 µL of methanol, and 200 µL of hexane, the wells were treated with 100 µL of methanol containing 2% formic acid to elute the PPIX, heme, and internal standards. The eluted samples were analyzed using LC/MS.
Quantification of PPIX and heme in the cell extracts
Concentrations of PPIX and heme in the purified samples were analyzed using LC/MS in accordance with a previously described method (Moulin et al., 2008; Abe et al., 2018) with modifications. Details of the LC/MS analytical conditions are described in Supplemental Table 2. PPIX and heme were analyzed and quantitated with MS and UV, respectively. Calibration samples were also analyzed. These samples were prepared by adding a small volume of the serial dilutions of PPIX and heme into untreated cell extracts, and purifying in the same manner. After LC/MS analysis of the samples, peak areas were computed and quantified using Thermo Xcalibur 2.2 (Thermo Fisher Scientific, Inc.). The concentrations of PPIX and heme were calculated using the regression line obtained from a plot of the calibration sample results.
Statistical analysis
All statistical analyses were performed using GraphPad Prism 5 software (GraphPad Software Inc., San Diego, CA, USA). The averages under various conditions were evaluated by Dunnett’s test. Standard errors of the mean are shown as error bars in all figures. The sample sizes (n) were defined by biological replicates. Significance was assumed with *p < 0.05, **p < 0.01, ***p < 0.001.
RESULTS
Generation of embryonic erythroid cells from hiPSCs
To obtain embryonic erythroid cells for the in vitro heme synthesis assay, we developed a 2-step protocol that combines the 3-dimensional (3D) embryoid body (EB) method with subsequent adherent 2-dimensional (2D) culture using commercially available hematopoietic differentiation medium and erythropoietin (Fig. 1a). On day 3 of differentiation, EBs were formed (Fig. 1b, left), and the hematopoietic-like non-adherent cells were observed to grow out from the adherent EBs within 7 days of differentiation (Fig. 1b, middle). A large number of suspended cells was generated on day 10 (Fig. 1b, right) and the cells were collected and transferred to new 6-well plates with fresh hEDM-B medium on day 10. The pellets obtained by harvesting the spherical cells showed increasingly intense red color from day 10, which indicated that most of these hiPSC-derived cells were erythroid cells that continuously produced a considerable amount of heme (Fig. 1c).
Next, we determined the level of erythroid cell differentiation. From immunofluorescence analyses, most erythroid cells expressed Glycophorin A (GPA; an erythroid marker) and embryonic ε-globin on days 10, 14, and 18 (Fig. 2a-l). Moreover, qRT-PCR assay revealed that the hiPSC-derived erythroid cells expressed a high percentage (65 ± 3.6%, 61 ± 4.2%, and 65 ± 2.4% on day 10, 14, and 18, respectively) of human embryonic ε-globin in the total β-like globin mRNA expression from day 10, and kept the expression levels until day 18 (Fig. 3a). In contrast, in heme-producing cells induced from K562 cells using NaB, the content of ε-globin was up to 30% until day 8, and it was lower than that in the hiPSC-derived erythroid cells (Fig. 3b). These data confirmed that the hiPSC-derived erythroid cells could be considered embryonic erythroid cells and are developmentally more primitive than induced K562 cells.
Generation of embryonic erythroid cells from rESCs
We also differentiated rESCs into erythroid cells for in vitro assays. ESCs were used since availability of rat iPSCs was limited. At first, we attempted to apply the hiPSCs-to-rESCs differentiation method. However, the yield of hematopoietic-like cells and the reproducibility of differentiation were insufficient for the assay that followed (data not shown). Therefore, we optimized the method for rESCs differentiation (Fig. 4a). EBs were formed and transferred onto OP9 feeder cells (Fig. 4b left) placed in Matrigel-coated 6-well plates (5 EBs/well) on day 3, in differentiation medium (rEDM-B) containing differentiation supplement B, rat erythropoietin (rEPO), and rat stem cell factor (rSCF). On day 5, the medium was changed to rEDM-B medium additionally supplemented with hydrocortisone, IL-3, IL-11, and IGF-1. Hematopoietic-like non-adherent cells began to be generated around day 5 of differentiation (Fig. 4b middle). On day 8, a large number of hematopoietic-like cells emerged (Fig. 4b, right) and the cells were harvested, then transferred to new 6-well plates with the third differentiation medium (rEDM-D), which was StemlineII supplemented with rEPO and rSCF. Similar to the results of hiPSCs analysis, the pellets of obtained cells increased in redness from day 8, indicating heme synthesis (Fig. 4c). In addition, immunostaining revealed that most of the floating cells expressed ε-globin on days 8, 10, and 12 (Fig. 5a-f). Furthermore, it was shown that rESC-derived erythroid cells expressed a high percentage (62 ± 1.6% and 62 ± 3.4% on day 10 and 12, respectively) of rat embryonic ε-globin in the total β-like globin mRNA expression until day 12, whereas adult β-globin expression was also detected on day 12 (Fig. 6a). In contrast, almost all β-like globins expressed (> 98%) in the heme-synthesizing erythroid-like cells induced from REL cells using HMBA were adult β-globin (Fig. 6b). These results confirmed that, like hiPSC-derived erythroid cells, rESC-derived erythroid cells could be considered as embryonic erythroid cells and developmentally more primitive than the available cultured cell-derived ones.
To investigate effects of flumioxazin on heme synthesis in human and rat embryonic erythroid cells, we first established an exposure regimen for in vitro heme synthesis assay using human embryonic erythroid cells derived from hiPSCs (Fig. 7a). Day 10 as the starting point of the exposure period was set based on the high expression of ε-globin and availability of an adequate number of cells. The erythroid cells were collected and resuspended in hEDM-B medium containing test chemicals. After that, the cells were collected on day 14 and analyzed to determine the viable cell number and the concentrations of PPIX and heme.
As results, DHA, used as a positive control of heme synthetic inhibition, caused significant reduction in cell proliferation at 0.5 μM and 1 μM (Fig. 7b; 0.35-fold and 0.26-fold vs control, respectively) and heme synthesis at 1 μM (Fig. 7d; 0.26-fold vs control) but had no effect on the accumulation of PPIX (Fig. 7c; 0.72-fold and 0.44-fold vs control, respectively). This confirmed that DHA could deplete embryonic erythroid cells and shut off heme production, and that the exposure regimen for this in vitro system enabled detection of developmental toxicants that affect heme synthesis and/or cell viability, leading to anemia. In contrast, treatment with flumioxazin had no effect on cell proliferation (Fig. 7b; 0.86-fold and 0.95-fold vs control at 1 μM and 5 μM, respectively) or heme content (Fig. 7d; 1.2-fold and 1.0-fold vs control at 1 μM and 5 μM, respectively), although a significant accumulation of PPIX was observed in a dose-dependent manner (Fig. 7c; 2.1-fold and 5.7-fold vs control at 1 μM and 5 μM, respectively). In detail, the concentrations of heme when treated with control or 5 μM of flumioxazin were 2705 vs 2558 ng/106 cells (run 1), 1664 vs 1632 ng/106 cells (run 2), and 2494 vs 2709 ng/106 cells (run 3), respectively. These results confirmed that flumioxazin did not inhibit heme production in the human embryonic erythroid cells, whereas it did inhibit PPO activity, since PPIX accumulated in the cells.
Studies of effects of flumioxazin on heme synthesis in rESCs-derived rat embryonic erythroid cells
For a comparative study to the human in vitro system, we also constructed an equivalent assay system for the evaluation of heme synthesis using rat embryonic erythroid cells derived from rESCs (Fig. 8a). Based on the analysis of globin expression showing that the rESC-derived erythroid cells rapidly differentiated into cells expressing adult β-globin (Fig. 6a), we adopted day 8 as the starting point for the exposure period, to ensure collection of a satisfactory amount of embryonic erythroid cells. On day 8, cells were collected and transferred to a new plate with rEDM-D medium containing DHA or flumioxazin. After 4 days exposure in the same manner as for the hiPSCs-derived human embryonic erythroid cells, the numbers of cells were counted and the concentrations of PPIX and heme were evaluated in each sample. The concentration of 0.5 μM DHA was adopted because it was the highest dose within the acceptable level of the cell viability in a preliminary study (data not shown). As a result, DHA was found to reduce the proliferation (Fig. 8b; 0.30-fold vs control) and heme synthesis (Fig. 8d; 0.42-fold vs control) in the rESC-derived embryonic erythroid cells, but to show no accumulation of PPIX (Fig. 8c; 0.67-fold vs control). This confirmed that DHA worked as a positive control with this exposure regimen for the in vitro system using rat embryonic erythroid cells. In contrast to the results for the hiPSC-derived cells, flumioxazin showed tendency to reduce heme synthesis with dose-dependency in the rESC-derived embryonic erythroid cells (Fig. 8d; 0.83-fold and 0.66-fold vs control at 1 μM and 5 μM, respectively), while increased PPIX accumulation was observed (Fig. 8c). To be more specific, 5 μM of flumioxazin consistently reduced the heme synthesis compared with the control (run 1: 5231 vs 3757 ng/106 cells, run 2: 3699 vs 2785 ng/106 cells, and run 3: 5792 vs 3151 ng/106 cells in control and 5 μM of flumioxazin, respectively). Comparing the results from assays using the hiPSC-derived erythroid cells, flumioxazin had no effect on heme synthesis in the human embryonic cells, but inhibited heme production in the rat embryonic cells.
DISCUSSION
In this study, we established a highly reproducible and simple differentiation system specific to embryonic erythroid cells for both humans and rats. The previous protocol for differentiation into embryonic erythroid cells was modified into a 2-step protocol, namely, floating culture by formation of EBs and subsequent adherent culture by transferring EBs to Matrigel-coated plates (Fig. 1a, Fig. 4a). This 2-step method enabled differentiation into erythroid cells under high cell-to-cell density conditions, while ensuring enough space existed between EBs and controlling the number of cells. In this system, floating hematopoietic-like cells were successfully produced from both hiPSC- and rESC-derived adherent EB colonies within 8 days (Fig. 1b, Fig. 4b). Moreover, the red color of non-adherent cell pellets was intensified, which indicated high level of continuous heme synthesis (Fig. 1c, Fig. 4c). These results demonstrated that the method of erythroid cell generation established in this study was robust, reproducible, and applicable to PSCs from both species.
The rat as an experimental animal has been used for over 150 years in many biologic sciences, including toxicology and physiology (Jacob, 1999). Comparing data from in vitro systems using rat cells with the in vivo data obtained from rats would enable us to validate in vitro methods, and directly comparing in vitro systems using rat cells with in vitro systems using human cells would enable us to evaluate species difference. Although rESCs are useful to generate required normal cells, rESCs were not established until 2008, whereas mouse ESCs were first prepared in 1981 (Buehr et al., 2008; Li et al., 2008). The delay of the establishment of rESCs had hindered the application of rESCs to rat in vitro assay systems. In addition, some differences between rESCs and mouse/human PSCs have been reported. For example, it was shown that extrinsic addition of small molecule inhibitors that can shield ESCs from inductive differentiation cues (such as glycogen synthase kinase 3 [GSK3] inhibitor, mitogen-activated protein kinase kinase [MEK] inhibitor, and fibroblast growth factor [FGF] receptor inhibitor) are needed for the maintenance of rESCs. In addition, it was reported that the majority of cells were shown to die within 2 days, instead of forming EBs without signal inhibitors (Li et al., 2008). Because of these differences, an optimization of the differentiation method for rESCs is needed along with a robust protocol to differentiate both human and rat PSCs into desired cells. We found that centrifuging the plates after seeding rESCs on low-adherence plates enabled rESCs to form EBs with high viability, even if cultured without additional inhibitors (Fig. 4b left).
In our differentiation protocol, most erythroid cells derived from hiPSCs and rESCs expressed embryo-specific ε-globin protein in humans (on days 10, 14, and 18) and rats (on days 10 and 12). Moreover, quantitative mRNA expression analyses revealed that these erythroid cells expressed a high level of ε-globin relative to total β-like globin gene expression (> 60%; Fig. 3a, Fig. 6a). It is known that γ-globin is expressed not only in fetal erythroid cells, but also in embryonic cells, and that the percentage of ε-globin expression among all the types of globin in the human yolk sac is about 30% (Palis and Segel, 1998; Demirci and Tisdale, 2018). The amount of 30% of ε-globins among all the types of globin could be regarded as 60% among the β-like globins, since hemoglobin consists of two α-like globins and two β-like globins. The expression levels of ε-globins among the β-like globins of the hiPSC- and rESC-derived erythroid cells were consistent with the reported data until day 18 (65%; Fig. 3a) and 12 (62%; Fig. 6a), respectively. These results indicated the hiPSC- and rESC-derived erythroid cells in this study could be considered to be mostly embryonic at least until day 18 and 12, respectively.
When embryonic erythroid cells were treated with DHA, which was reported to cause impairment of primitive erythroblasts in yolk sac in vitro and in vivo (Longo et al., 2006a, 2006b), we demonstrated that cell proliferation and heme production were inhibited in both humans and rats in in vitro systems (Fig. 7b, d, Fig. 8b, 8d). Although the reduction of the proliferation in DHA in rat embryonic erythroid cells was not significant, the ratio of the reduction was of the same degree at the highest concentration to hiPS-derived cells (0.26-fold and 0.30-fold vs control in hiPSC- and rESC-derived cells, respectively). In addition, flumioxazin promoted the accumulation of PPIX by inhibiting the activity of PPO, and tended to suppress heme production in rESC-derived embryonic erythroid cells (Fig. 8c, d). Although the reduction of heme synthesis was not statistically significant, the suppression was observed dose-dependently (Fig. 8d) and the degree of suppression at 5 μM of flumioxazin was consistent in each trial when compared with the concurrent control. The agreement between these results from rESC-derived cells and those from rat in vivo study (Kawamura et al., 1996a, 1996b, 2014, 2016) indicated that the in vitro heme synthesis assay using PSC-derived erythroid cells could be reasonably used to evaluate the effect of inhibition of heme synthesis, leading to embryonic anemia. As the effects of DHA were also equivalent in human and rat erythroid cells, we concluded that the assay system using human embryonic erythroid cells could properly be used to evaluate toxicity in humans.
In contrast to its effects in rats, however, there were no effects on heme content or cell proliferation in the human embryonic erythroid cells even when cells were treated with 5 μM of flumioxazin (Fig. 7b, d), which is close to its limit of solubility. In the previous study, the analysis using a physiologically based pharmacokinetic model demonstrated that the simulated maximum concentration of flumioxazin in the embryo and fetus of pregnant humans during a theoretical accidental intake (1000 mg/kg) was 1.92 μM (Takaku et al., 2014). Therefore, it was considered to be reasonable that we used the top dose of 5 μM of flumioxazin for the evaluation in this study, since the dose exceeded the simulated maximum concentration in the human embryo.
Since PPIX accumulation was observed both in human and rat erythroid cells (Fig. 7c, Fig. 8c), flumioxazin was shown to inhibit human PPO as well as rat PPO. However, heme synthesis was inhibited only in the rat erythroid cells, and not in the human embryonic erythroid cells. This variance in the effects on heme production between the human and rat cells might be due to the difference in the potency of PPO inhibition and the relative activities of heme synthetic enzymes. In fact, our previous study demonstrated based on in vivo and in vitro metabolism experiments that human PPO was less sensitive to flumioxazin than rat PPO (Abe et al., 2018). Moreover, 3D molecular dynamics simulations of human and rat PPO-flumioxazin complexes revealed that these differences were caused by the dynamics of the 107-120 loop region derived from amino acid sequence variants (Arakawa et al., 2017). It was also suggested that the human PPO activity in erythroid cells is much higher than other heme synthesis enzymes such as ALAS2, which is the rate-limiting enzyme in the heme biosynthesis pathway, although rat PPO activity seemed to be close to the rate-limiting enzyme in erythroid cells (Kawamura et al., 2021). Since our data obtained from human and rat embryonic erythroid cells were comparable to those of a previous report that used human K562 cells, human CD36+ cells, and REL cells (Kawamura et al., 2021), it was confirmed that flumioxazin has no effect on the heme synthesis in either primitive or definitive erythroid cells in humans. Taken together, flumioxazin does not inhibit heme synthesis in any of 3 different types of human cells, whereas it does in two types of rat cells, seeming to reveal a fundamental qualitative difference between humans and rats. The consistency of the results from these in vitro assays strengthened the evidence that flumioxazin is safe for humans.
In conclusion, we assessed the developmental toxicity of flumioxazin by employing a comparative in vitro assay system using hiPSCs and rESCs. Our results indicated that there could be a species difference in the effects of flumioxazin on heme production between humans and rats; therefore, developmental toxicity via embryonic anemia shown in rats is unlikely to be caused in human embryos. Additionally, our new approach using hiPSCs and rESCs for evaluating toxicity of chemicals would be a powerful tool for human safety assessment.
ACKNOWLEDGMENTS
We thank Asami Sakai and Takako Kobayashi for their technical assistance; and Tomoya Yamada, Kaori Miyata, Koichi Saito, Noriyuki Suzuki, Kumiko Kobayashi, and Tokushige Nakano for their invaluable comments.
Conflict of interest
All authors are employees of Sumitomo Chemical Co. Ltd.
REFERENCES
- Abe, J., Isobe, N., Mikata, K., Nagahori, H., Naito, Y., Saji, H., Ono, M. and Kawamura, S. (2018): Flumioxazin metabolism in pregnant animals and cell-based protoporphyrinogen IX oxidase (PPO) inhibition assay of fetal metabolites in various animal species to elucidate the mechanism of the rat-specific developmental toxicity. Toxicol. Appl. Pharmacol., 339, 34-41.
- Arakawa, A., Otani, M., Iwashita, K. and Yamazaki, K. (2017): Molecular dynamics mechanism to generate species differences in inhibition of protoporphyrinogen oxidase by flumioxazin. Comput. Toxicol., 1, 12-21.
- Buehr, M., Meek, S., Blair, K., Yang, J., Ure, J., Silva, J., McLay, R., Hall, J., Ying, Q.L. and Smith, A. (2008): Capture of authentic embryonic stem cells from rat blastocysts. Cell, 135, 1287-1298.
- Chang, K.H., Nelson, A.M., Cao, H., Wang, L., Nakamoto, B., Ware, C.B. and Papayannopoulou, T. (2006): Definitive-like erythroid cells derived from human embryonic stem cells coexpress high levels of embryonic and fetal globins with little or no adult globin. Blood, 108, 1515-1523.
- Clark, R.L., Arima, A., Makori, N., Nakata, Y., Bernard, F., Gristwood, W., Harrell, A., White, T.E. and Wier, P.J. (2008): Artesunate: developmental toxicity and toxicokinetics in monkeys. Birth Defects Res. B Dev. Reprod. Toxicol., 83, 418-434.
- Clark, R.L., White, T.E., Clode, S., Gaunt, I., Winstanley, P. and Ward, S.A. (2004): Developmental toxicity of artesunate and an artesunate combination in the rat and rabbit. Birth Defects Res. B Dev. Reprod. Toxicol., 71, 380-394.
- Cioe, L., McNab, A., Hubbell, H.R., Meo, P., Curtis, P. and Rovera, G. (1981): Differential expression of the globin genes in human leukemia K562(S) cells induced to differentiate by hemin or butyric acid. Cancer Res., 41, 237-243.
- Demirci, S. and Tisdale, F.T. (2018): Definitive Erythropoiesis from Pluripotent Stem Cells: Recent Advances and Perspectives. In: Cell Biology and Translational Medicine, Volume 3 (Turksen, K. ed.), pp.1-13, Springer.
- Finaurini, S., Basilico, N., Corbett, Y., D’Alessandro, S., Parapini, S., Olliaro, P., Haynes, R.K. and Taramelli, D. (2012): Dihydroartemisinin inhibits the human erythroid cell differentiation by altering the cell cycle. Toxicology, 300, 57-66.
- Finaurini, S., Ronzoni, L., Colancecco, A., Cattaneo, A., Cappellini, M.D., Ward, S.A. and Taramelli, D. (2010): Selective toxicity of dihydroartemisinin on human CD34+ erythroid cell differentiation. Toxicology, 276, 128-134.
- Fujita, A., Uchida, N., Haro-Mora, J.J., Winkler, T. and Tisdale, J. (2016): β-Globin-Expressing Definitive Erythroid Progenitor Cells Generated from Embryonic and Induced Pluripotent Stem Cell-Derived Sacs. Stem Cells, 34, 1541-1552.
- Ishida, Y., Kawakami, H., Kitajima, H., Nishiyama, A., Sasai, Y., Inoue, H. and Muguruma, K. (2016): Vulnerability of Purkinje Cells Generated from Spinocerebellar Ataxia Type 6 Patient-Derived iPSCs. Cell Rep., 17, 1482-1490.
- Jacob, H.J. (1999): Functional genomics and rat models. Genome Res., 9, 1013-1016.
- Kaufman, D.S., Hanson, E.T., Lewis, R.L., Auerbach, R. and Thomson, J.A. (2001): Hematopoietic colony-forming cells derived from human embryonic stem cells. Proc. Natl. Acad. Sci. USA, 98, 10716-10721.
- Kawamata, M. and Ochiya, T. (2010a): Establishment of embryonic stem cells from rat blastocysts. Methods Mol. Biol., 597, 169-177.
- Kawamata, M. and Ochiya, T. (2010b): Generation of genetically modified rats from embryonic stem cells. Proc. Natl. Acad. Sci. USA, 107, 14223-14228.
- Kawamura, S., Kato, T. and Fantel, A.G. (2014): Close link between protoporphyrin IX accumulation and developmental toxicity induced by N-phenylimide herbicides in rats. Birth Defects Res. B Dev. Reprod. Toxicol., 101, 429-437.
- Kawamura, S., Kato, T., Matsuo, M., Katsuda, Y. and Yasuda, M. (1996a): Species difference in protoporphyrin IX accumulation produced by an N-phenylimide herbicide in embryos between rats and rabbits. Toxicol. Appl. Pharmacol., 141, 520-525.
- Kawamura, S., Kato, T., Matsuo, M., Sasaki, M., Katsuda, Y., Hoberman, A.M. and Yasuda, M. (1995): Species Difference in Developmental Toxicity of an N‐Phenylimide Herbicide between Rats and Rabbits and Sensitive Period of the Toxicity to Rat Embryos. Congenit. Anom. (Kyoto), 35, 123-132.
- Kawamura, S., Otani, M., Miyamoto, T., Abe, J., Ihara, R., Inawaka, K. and Fantel, A.G. (2021): Different effects of an N-phenylimide herbicide on heme biosynthesis between human and rat erythroid cells. Reprod. Toxicol., 99, 27-38.
- Kawamura, S., Yoshioka, T., Kato, T., Matsuo, M. and Yasuda, M. (1996b): Histological changes in rat embryonic blood cells as a possible mechanism for ventricular septal defects produced by an N-phenylimide herbicide. Teratology, 54, 237-244.
- Kawamura, S., Yoshioka, T., Mito, N., Kishimoto, N., Nakaoka, M. and Fantel, A.G. (2016): Mechanism of Developmental Effects in Rats Caused by an N-Phenylimide Herbicide: Transient Fetal Anemia and Sequelae during Mid-to-Late Gestation. Birth Defects Res. B Dev. Reprod. Toxicol., 107, 45-59.
- Kuwahara, A., Ozone, C., Nakano, T., Saito, K., Eiraku, M. and Sasai, Y. (2015): Generation of a ciliary margin-like stem cell niche from self-organizing human retinal tissue. Nat. Commun., 6, 6286.
- Li, P., Tong, C., Mehrian-Shai, R., Jia, L., Wu, N., Yan, Y., Maxson, R.E., Schulze, E.N., Song, H., Hsieh, C.L., Pera, M.F. and Ying, Q.L. (2008): Germline competent embryonic stem cells derived from rat blastocysts. Cell, 135, 1299-1310.
- Longo, M., Zanoncelli, S., Manera, D., Brughera, M., Colombo, P., Lansen, J., Mazué, G., Gomes, M., Taylor, W.R. and Olliaro, P. (2006a): Effects of the antimalarial drug dihydroartemisinin (DHA) on rat embryos in vitro. Reprod. Toxicol., 21, 83-93.
- Longo, M., Zanoncelli, S., Torre, P.D., Riflettuto, M., Cocco, F., Pesenti, M., Giusti, A., Colombo, P., Brughera, M., Mazué, G., Navaratman, V., Gomes, M. and Olliaro, P. (2006b): In vivo and in vitro investigations of the effects of the antimalarial drug dihydroartemisinin (DHA) on rat embryos. Reprod. Toxicol., 22, 797-810.
- Mori, T., Higashi, K., Nakano, T., Ando, S., Kuwahara, A., Suzuki, N. and Saito, K. (2017): Novel phototoxicity assay using human embryonic stem cell-derived retinal pigment epithelial cells. Toxicology, 378, 1-9.
- Moulin, M., McCormac, A.C., Terry, M.J. and Smith, A.G. (2008): Tetrapyrrole profiling in Arabidopsis seedlings reveals that retrograde plastid nuclear signaling is not due to Mg-protoporphyrin IX accumulation. Proc. Natl. Acad. Sci. USA, 105, 15178-15183.
- Nakano, T., Ando, S., Takata, N., Kawada, M., Muguruma, K., Sekiguchi, K., Saito, K., Yonemura, S., Eiraku, M. and Sasai, Y. (2012): Self-formation of optic cups and storable stratified neural retina from human ESCs. Cell Stem Cell, 10, 771-785.
- Olivier, E.N., Marenah, L., McCahill, A., Condie, A., Cowan, S. and Mountford, J.C. (2016): High-Efficiency Serum-Free Feeder-Free Erythroid Differentiation of Human Pluripotent Stem Cells Using Small Molecules. Stem Cells Transl. Med., 5, 1394-1405.
- Olivier, E.N., Qiu, C., Velho, M., Hirsch, R.E. and Bouhassira, E.E. (2006): Large-scale production of embryonic red blood cells from human embryonic stem cells. Exp. Hematol., 34, 1635-1642.
- Palis, J. and Segel, G.B. (1998): Developmental biology of erythropoiesis. Blood Rev., 12, 106-114.
- Pavlović, S., Mitrović, T., Nikcević, G., Grujicić, N., Lazić, D., Glišin, V. and Popović, Z. (1999): The rat β (b miny)-globin promoter: nuclear protein factors and erythroid-specific induction of transcription. Cell. Mol. Life Sci., 56, 871-881.
- Qiu, C., Olivier, E.N., Velho, M. and Bouhassira, E.E. (2008): Globin switches in yolk sac-like primitive and fetal-like definitive red blood cells produced from human embryonic stem cells. Blood, 111, 2400-2408.
- Scicchitano, M.S., McFarland, D.C., Tierney, L.A., Narayanan, P.K. and Schwartz, L.W. (2003): In vitro expansion of human cord blood CD36+ erythroid progenitors: temporal changes in gene and protein expression. Exp. Hematol., 31, 760-769.
- Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K. and Yamanaka, S. (2007): Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell, 131, 861-872.
- Takaku, T., Nagahori, H. and Sogame, Y. (2014): Metabolism and physiologically based pharmacokinetic modeling of flumioxazin in pregnant animals. Toxicol. Appl. Pharmacol., 277, 242-249.
- Ueda, S., Kawamata, M., Teratani, T., Shimizu, T., Tamai, Y., Ogawa, H., Hayashi, K., Tsuda, H. and Ochiya, T. (2008): Establishment of rat embryonic stem cells and making of chimera rats. PLoS One, 3, e2800.
- Vodyanik, M.A., Bork, J.A., Thomson, J.A. and Slukvin, I.I. (2005): Human embryonic stem cell-derived CD34+ cells: efficient production in the coculture with OP9 stromal cells and analysis of lymphohematopoietic potential. Blood, 105, 617-626.
- Zambidis, E.T., Peault, B., Park, T.S., Bunz, F. and Civin, C.I. (2005): Hematopoietic differentiation of human embryonic stem cells progresses through sequential hematoendothelial, primitive, and definitive stages resembling human yolk sac development. Blood, 106, 860-870.
