Characterization of a commercially available line of iPSC hepatocytes as models of hepatocyte function and toXicity for regulatory purposes

Hisham Qosa, Alexandre J.S. Ribeiro, Neil R. Hartman 1, Donna A. Volpe *
United States Food and Drug Administration, Center for Drug Evaluation and Research, Office of Translational Sciences, Office of Clinical Pharmacology, Division of Applied Regulatory Science, 10903 New Hampshire Ave, Silver Spring, MD 20993-0002, United States of America

A R T I C L E I N F O

* Corresponding author at: 10903 New Hampshire Ave., WO64-2034, Silver Spring, MD 20993-0002, USA.
E-mail address: [email protected] (D.A. Volpe).
1 Deceased.

https://doi.org/10.1016/j.vascn.2021.107083

Received 25 January 2021; Received in revised form 11 May 2021; Accepted 1 June 2021
Available online 5 June 2021
1056-8719/Published by Elsevier Inc.
Keywords: CYP enzyme Canaliculi Glucuronide Hepatocyte
iPSC hepatocyte Maturation Sulfation Spheroids ToXicity

A B S T R A C T

It has recently become possible to produce hepatocytes from human induced pluripotent stem cells (iPSC-heps), which may offer some advantages over primary human hepatocytes (Prim-heps) in the regulatory environment. The aim of this research was to assess similarities and differences between commercially available iPSC-heps and Prim-heps in preliminary assays of drug metabolism, hepatotoXicity, and drug transport. Hepatocytes were either cultured in collagen-coated 96-well plates (Prim-heps and 2d-iPSC-heps) or in ultra-low adhesion plates as spheroids (3d-iPSC-heps). 3d-iPSC-heps were used to enhance physiological cell-cell contacts, which is essential to maintain the phenotype of mature hepatocytes. Cytochrome P450 (CYP) 3A4, CYP1A2, and CYP2B6 activity levels were evaluated using fluorescent assays. Phase II metabolism was assessed by HPLC measurement of formation of glucuronides and sulfates of 4-methylumbelliferone, 1-naphthol, and estradiol. The toXicity of acetaminophen, amiodarone, aspirin, clozapine, tacrine, tamoXifen, and troglitazone was monitored using a luminescent cell viability assay. Canaliculi formation was monitored by following the fluorescence of 5,6-carboXy-2′,7′-dichlorofluorescein diacetate. All culture models showed similar levels of basal CYP3A4, CYP1A2 and
CYP2B6 activity. However, while Prim-heps showed a vigorous response to CYP inducing agents, 2d-iPSC-heps showed no response and 3d-iPSC-heps displayed an inconclusive response. 2d-iPSC-heps showed reduced, yet appreciable, glucuronide and sulfate formation compared to Prim-heps. All culture models showed similar activity in tests of hepatotoXicity, with Prim-heps generally being more sensitive. All models formed canaliculi capable of transporting carboXy-2′,7′-dichlorofluorescein. The iPSC-heps appear to be useful for toXicity and transport studies, but metabolic activity is not optimum, and metabolism studies would benefit from a more mature model.

1. Introduction

Current hepatic preclinical testing using animal models cannot fully recapitulate drug metabolism seen in human subjects and can result in unforeseen clinical hepatic toXicity (Navarro & Senior, 2006). To address this limitation, studies have been conducted with human he- patic tissue (e.g., hepatocytes, liver slices). A better understanding of drug effects related to human-specific metabolism can also be obtained by using human liver enzymes, either from liver subcellular fractions (e. g., microsomes) or from purified recombinant enzymes expressed in other cell systems (Zelasko, Palaria, & Das, 2013). However, these ap- proaches may not account fully for the relative proportion of enzymes seen in intact cells nor can they reveal unexpected or complex metabolic mechanisms. Whole cell approaches using liver slices (Parker, Collins, & Strong, 1996) or isolated hepatocytes (Go´mez-Lecho´n, Castel, & Donato, 2010) often give a better representation of human liver metabolic function, but these tissues are in short supply and tissues from different donors may have substantially different activities and responses. Whole cell systems can also assess the potential of a drug to cause hepatotoX- icity (Go´mez-Lecho´n et al., 2010). Human primary hepatocytes are the optimal source for whole cells systems, although of limited supply. Human hepatic immortalized cell lines (e.g., HepG2, HepaRG™) are also used for hepatic toXicity studies (Gerets et al., 2012), and while they are available in much larger quantities, they tend to express drug metabo- lizing enzymes poorly, and thus are not as useful for evaluating drugs that might need to be metabolically activated to a toXic form.
Recently, to avoid these difficulties, tissues of various types have been generated from human induced pluripotent stem cells (iPSC) carboXy-2′,7′-dichlorofluorescein diacetate (CDFA), 6-(4-chlorophenyl) imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4-dichlorobenzyl) (Scott, Peters, & Dragan, 2013). These cells are reprogrammed from oXime (CITCO), estradiol, estradiol-3-glucuronide, ketoconazole, 4- somatic cells that can be isolated from several tissue types and can be grown continuously in culture to whatever number necessary, given adequate quality control criteria to avoid genetic abnormalities. These iPSCs can then be differentiated to form hepatocyte-like cells (iPSC- hepatocytes) (Cotovio & Fernandes, 2020). These iPSC-hepatocytes have many similarities to primary human hepatocytes including meta- bolism and toXicity assays appropriate to drug discovery (Lu et al., 2015; Kvist et al., 2018; Sirenko, Hesley, Rusyn, & Cromwell, 2014; Sirenko et al., 2016; Sjogren et al., 2014). It is also possible, in theory, to develop iPSC-hepatocytes from subjects with genetic traits and variabilities that potentially allow pre-clinical study of individuals or groups who might have unusual genetically based reactions to medications (Williams, 2018). Moreover, the use of 3d-iPSC-heps cultures has been demon- strated to enhance physiological cell-cell contacts, which is essential to maintain the phenotype of mature hepatocytes such as CYP enzymes activity and canaliculi formation (Meier et al., 2017; Takayama et al., 2013). However, there are differences between primary and iPSC he- patocytes, the general pattern being that iPSC-hepatocytes behave as less developmentally mature cells than primary hepatocytes. For instance, it is difficult to show cytochrome P450 (CYP) induction in iPSC-hepatocytes (Lu et al., 2015; Kvist et al., 2018). This has motivated the development of several procedures to increase the apparent maturity of iPSC-hepatocytes (Agarwal, Popovic, Martucci, Fraunhoffer, & Soto- Gutierrez, 2019). This field is progressing rapidly (Cotovio & Fer- nandes, 2020) and continued ongoing evaluation of new data is neces- sary to update the documented capabilities of iPSC-hepatocytes.
As an alternative to reprogramming iPSCs and differentiating cul- tures in-house, it has recently been possible to obtain ready-to-use iPSC- hepatocytes from commercial sources (Kvist et al., 2018; Lu et al., 2015; Sirenko et al., 2014). This increases the likelihood that iPSC-hepatocytes preparations will find use in preclinical drug development studies. Moreover, iPSC cells can provide a continuous supply of cells with he- patocyte characteristics that can be used as a reliable model for meta- bolic, transporters and toXicity testing of candidate drugs in preclinical studies at a lower cost and less effort in cell maintenance compared to the primary hepatocytes, which would provide substantial savings in drug development time and costs.
To address the present utility of iPSC-hepatocytes in drug develop- ment, this laboratory has compared a commercially available iPSC he- patocyte line with commercially available primary human hepatocytes in an initial set of assays to compare and contrast their respective merits. Metabolic, toXicity, and transport assays, in 2- or 3-dimensions (2d, 3d), were selected to encompass hepatocytes activities that are used in early drug development. Metabolic and transporter activities of in vitro he- patocytes models are important to predict the potential drug-drug in- teractions (DDIs) during early drug development, while toXicity studies are used to evaluate the risk of drug induced liver injury (DILI). Therefore, optimizing hepatocytes models for the purpose of metabolic and toXicity studies is an integral part of early drug development and can help in the reduction of the risk of DDIs and DILI.

2. Materials and methods

2.1. Materials
Most cell culture reagents and supplies were obtained from Ther- moFisher (Pittsburg, PA) with exceptions as noted. Oncostatin M was obtained from R&D Systems (Minneapolis, MN). GravityTRAP™ ULA plates were obtained from InSphero Inc. (Brunswick ME). Amiodarone, aspirin, clozapine, estradiol-17-glucuronide, tacrine, tamoXifen and troglitazone were purchased from Cayman Chemical (Ann Arbor, MI). Naphthol glucuronide and estradiol-17-sulfate were acquired from Toronto Research Chemicals (Toronto, ON). Acetaminophen, 5(6)- methylumbelliferone, 4-methylumbelliferone glucuronide, 4-methyl- umbelliferone sulfate, 1-naphthol, omeprazole and fetal bovine serum (FBS) were purchased from Millipore Sigma (Saint Louis, MO).

2.1.1. Cell culture
A single lot of cryopreserved human hepatocytes (Prim-heps) were obtained from Thermo Fisher and cultured in supplemented Williams
Medium E on collagen-coated 96-well plates according to the supplier’s instructions (ThermoFisher, 2020). Cells were plated at 3.5 × 105 cells/ well and overlaid with Geltrex™ (Thermo Fisher) after being allowed to adhere for four hours. Investigations commenced 24 h after plating.
Human iPSC-hepatocytes (2d-iPSC-heps) were obtained from Cellular Dynamics International (iCell® hepatocytes, Madison, WI) and cultured in supplemented RPMI 1640 medium on collagen-coated 96- well plates according to the supplier’s instructions (Cellular Dynamics International, 2020a). Cells were plated at 3.5 105 cells/well and cultured for five days in medium supplemented with oncostatin M to allow for cell maturation. No Geltrex™ overlay was necessary for these cultures. Investigations commenced after the 3 days maturation period in medium without oncostatin M.iPSC-hepatocyte spheroids (3d-iPSC-heps) were produced from Cellular Dynamics International iCell® hepatocytes. Cells were cultured in supplemented DMEM/F12 medium. Cells were initially cultured in collagen-coated 12-well plates according to the supplier’s instructions (Cellular Dynamics International, 2020b). After five days the cells were dissociated with Accutase® and re-cultured at 2000 cells/well in 96 well ultra-low adhesion plates (GravityTRAP™ ULA). Spheroids formed spontaneously over four days and investigations began immediately thereafter.

2.1.2. CYP measurements
Activity levels of CYP3A4, CYP1A2 and CYP2B6 were determined using the Promega P450-Glo™ 3A4, 1A2, and 2B6 test systems (Madi- son, WI) according to the manufacturer’s directions (Promega, 2020b). Inhibition of basal CYP3A4 activity in Prim-heps and 2d-iPSC-heps was performed by adding 1 μM ketoconazole to the fresh medium containing a luminogenic CYP substrate to the cells followed by incubation for 30–60 min before measuring CYP3A activity. Induction of CYP enzymes was investigated by treatment with 20 μM rifampicin (CYP3A4), 5 μM omeprazole (CYP1A2), or 1 μM CITCO (CYP2B6) in the appropriate medium for 72 h prior to evaluation. Medium with inducers was changed daily.

2.1.3. Conjugate studies
Prim-heps and 2d-iPSC-heps were exposed to 150 μM 4-methylum- belliferone (4MU), 50 μM 1-naphthol, or 100 μM estradiol in 120 μL of the appropriate culture medium in 96-well plates. Probe substrates were added as media solutions with ethanol as the solvent at a con- centration below 0.5% in all cases. The complete culture medium was collected at 0, 1, 2, 4, 7, and 24 h and protein was precipitated with the addition of 240 μL of acetonitrile. The samples were centrifuged, the supernatant aspirated, and acetonitrile was evaporated under a stream of dry nitrogen. The samples were reconstituted with 120 μL 5% acetonitrile in water and 25 μL samples were subjected to HPLC analysis. Conjugates were identified and quantified by comparing the retention time and spectrum of samples to authentic standards. Insufficient ma- terial was available to allow for identification of conjugates in the spheroid culture system. The liquid chromatography assay was con- ducted with a Agilent 1200 HPLC system using a Phenomenex Luna® C18–2 column (5 μ, 150 4.6 mm) with a 1 mL/min solvent flow rate. Other parameters for the substrates are as shown below:

• 4-Methylumbelliferone: gradient of 95% 10 mM potassium phos- phate buffer pH 6.5–5% acetonitrile progressing to 55% acetonitrile over 15 min; detection by UV absorbance at 314 nm.
• 1-Naphthol: gradient of 95% 10 mM potassium phosphate buffer pH 6.5–5% acetonitrile progressing to 30% acetonitrile over 15 min;
detection by UV absorbance at 295 nm.
• Estradiol: gradient of 100% 10 mM potassium phosphate buffer pH 6.5–0% acetonitrile progressing to 30% acetonitrile over 8 min, holding at 30% acetonitrile for 2 min; detection by fluorescence, excitation 280 nm, emission 310 nm.

2.1.4. Toxicity studies
Multiple concentrations of siX drugs toXic to hepatocytes, plus aspirin as a negative control, were made in the medium appropriate to the cell type (see above). Most drugs were added as a media solution in DMSO except for acetaminophen, aspirin, and tamoXifen which were added in ethanol. Concentrations of the co-solvents were below 0.5% in all cases. Blank medium with appropriate concentrations of co-solvents were used to establish background activity. Concentrations were as follows:
• Acetaminophen: 250, 500, 1000, 5000, 10,000, 50,000 μM
• Amiodarone: 1.8, 3.7, 7.3, 14.7, 29.3 μM
• Aspirin: 55, 111, 222, 444, 888 μM
• Clozapine: 0.04, 0.4, 4, 38, 76, 100 μM
• Tacrine: 5, 11, 21, 43, 426, 1000 μM
• TamoXifen: 0.01, 0.1, 1.0, 11. 108, 150 μM
• Troglitazone: 1.7, 3.4, 6.8, 13.6, 135.9 μM

Cells were incubated for 48 or 72 h as indicated. ATP levels were measured using the Promega CellTiter-Glo® luminescent cell viability assay per the manufacturer’s directions (Promega, 2020a). Data was normalized to the background reading. IC50 values were determined by fitting the data to a variable slope four parameter log(inhibitor) vs. response model with the bottom parameter constrained to zero (GraphPad Prism 8, San Diego, CA).

2.1.5. Canaliculi visualization
For staining functional bile canaliculi, cultures were washed 3 with culture medium, incubated at 37 ◦C with 2 mg/mL CDFA prepared in HBSS buffer with or without 5 mM CaCl2 for 20 min, and washed 3×using HBSS buffer with or without 5 mM CaCl2 again prior to fluores- cence microscopy. Imaging was conducted for four consecutive days starting at day 2, 7 and 9 of Prim-heps, 2d-iPSC-heps and 3d-iPSC-heps plating, respectively, with different culture wells used on each imaging day. Imaging was by means of a Zeiss PALM instrument (White Plains, NY).

2.2. Statistical analysis
Unless otherwise indicated, the data were expressed as mean standard deviation. The experimental results were statistically analyzed for significant difference using two-tailed Student’s t-test for two groups, and one-way analysis of variance (ANOVA) for more than two group analysis. Values of p < 0.05 were considered statistically significant.

3. Results

3.1. Oxidative metabolism
The enzyme activity studies were conducted on day 5, 10 and 12 of Prim-heps, 2d-iPSC-heps and 3d-iPSC-heps plating, respectively. Both Prim-heps and 2d-iPSC-heps demonstrated baseline CYP activity that was stable over several days, as shown for CYP3A4 in Fig. 1. The amount of CYP3A4 activity was comparable between the two cell sources for a similar number of cells and this activity was significantly reduced by addition of CYP3A4 inhibitor, 1 μM ketoconazole (Fig. 1).
Induction experiments were conducted on the Prim-heps, 2d-iPSC- heps and 3d-iPSC-heps after 2, 7 and 9 days of culture, respectively. As can be seen in Fig. 2A, while rifampicin treatment of Prim-heps showed vigorous induction of CYP3A4 activity, there was little evi- dence of CYP3A4 induction by rifampicin in 2d-iPSC-heps. Similar re- sults were seen for CYP1A2 induction by omeprazole and CYP2B6 induction by CITCO. Studies extending the length of culture of 2d-iPSC- heps (7 days maturation period in medium without oncostatin M) did not lead to an improvement of induction activities of 72-h treatment with the respective inducer (Fig. 2B). Studies evaluating 3d-iPSC-heps showed minimal induction for CYP1A2 and CYP2B6, however this did not rise to the level of statistical significance (Fig. 2A).

3.2. Conjugation
Both Prim-heps and 2d-iPSC-heps were capable of forming conju- gates of 4MU (Fig. 3). The rate (pmol/103 cells/min) of sulfation (S) of 4MU was comparable between the two cell types, however the rate (pmol/103 cells/min) of glucuronidation (G) of 2d-iPSC-heps was about 30% of Prim-heps. The formation of conjugates from naphthol was very rapid in Prim-heps, with complete metabolism of the parent compound observed by 4 h and therefore a plateau formed in the metabolite traces. While the metabolic rate (pmol/103 cells/min) of naphthol in the 2d-iPSC-heps was slower, metabolism was complete by 24 h. Naphthol- glucuronide and naphthol-sulfate formation at four hours by 2d-iPSC- heps was 21% and 30% of Prim-heps, respectively. Estradiol conjugation was also much more rapid in Prim-heps, with a reduction in rate (pmol/103 cells/min) noted at about 7 h likely due to depletion of the parent compound. The formation of estradiol-3-glucuronide by 2d-iPSC-
Fig. 1. Basal CYP3A4 activity of the same cultures of Prim-heps and iPSC-heps with increased time in culture in the presence and absence of ketoconazole (Ketoc). Determinations represent the mean ± standard error of three replicates. *** p ≤ 0.001.
Fig. 2. (A) EXtent of CYP induction for Prim-heps, 2d-iPSC-heps and 3d-iPSC-heps induction is expressed as the percent increase over basal levels. Cells were induced for three days with 20 μM rifampicin (CYP3A4), 5 μM omeprazole (CYP1A2), or 1 μM CITCO (CYP2B6). Determinations represent the mean and standard error of three replicates; (B) 2d-iPSC-heps activities for CYP3A4, CYP1A2 and CYP2B6 did not show induction after treatment with their respective inducers for 24 h at early maturation (day 3 of maturation period in medium without oncostatin M or day 8 of plating) or extended maturation (day 7 of maturation period in medium without oncostatin M or day 12 of plating). ** p ≤ 0.01, * p ≤ 0.05, ns not significant.
Fig. 3. Time course of conjugate formation for Prim-heps and 2d-iPSC-heps incubated with 150 μM 4MU, 50 μM 1-naphthol, or 100 μM estradiol. Plots depict results from a single experiment. Prim = Prim-heps, iPSC = 2d-iPSC-heps, 4MU-G = 4MU-glucuronide, 4MU-S = 4MU-sulfate, Naph-G = naphthol-glucuronide, Napth-S = naphthol-sulfate, E2-3G = estradiol-3-glucuronide, E2-17G = estradiol-17-glucuronide, E2-SO4 = estradiol-17-sulfate.
heps at 4 h was only 4% of that of primary hepatocytes; no estradiol-17- glucuronide was detected. Estradiol-17-sulfate was undetectable at early timepoints in 2d-iPSC-heps, however it reached 50% of the Prim-heps level by 24 h.

3.3. Toxicity
The effect of siX drugs known to be toXic to hepatocytes, with aspirin as a negative control, was assessed on Prim-heps, 2d-iPSC-heps and 3d- iPSC-heps. Cell viability was determined using ATP content. The results of studies were generally similar between the three preparations, with Prim-heps generally being more sensitive than the 2d-iPSC preparations. This can be seen in significantly lower IC50 values for acetaminophen at 48-h incubation and acetaminophen, clozapine, tacrine and tamoXifen at 72-h incubation (Table 1 and Fig. 4). The 72-h incubations generally demonstrated lower IC50 concentrations than the 48-h preparations. 2d- iPSC-heps and 3d-iPSC-heps were similar with the notable exception of amiodarone, where no toXicity was observed over the tested range with the 3d-iPSC-heps.

3.4. Canaliculi visualization
Prim-heps developed easily visualized canaliculi over the course of four days (Fig. 5A). These same structures were present in 2d-iPSC-heps on the first day of testing, however these cells had been plated three days previously to allow for maturation (Fig. 5B). In the absence of calcium, the qualitative degree of dye localization was substantially reduced for both preparations. 3d-iPSC-heps also formed easily recognizable canal- iculi, however these were not reduced upon treatment with calcium free medium (Fig. 5C).

4. Discussion

In this preliminary study, the basal levels of CYP3A4, CYP1A2, and CYP2B6 were highly similar between Prim-heps and 2d-iPSC-heps, with consistent demonstration of metabolism which was observed for the same cells over several days. These metabolic isozymes were chosen due to their importance in drug metabolism (Zanger & Schwab, 2013) and the availability of luminescence-based assays suitable for whole cell preparations. Studies evaluating the activity of other prominent iso- zymes (e.g., CYP2D6, CYP2C9, CYP2C19) would be valuable for future evaluations of 2d and 3d iPSC hepatocytes. The advantage of spheroid iPSC hepatocytes was demonstrated by Takayama et al., in that spher- oids had higher expression, activity and induction of CYP2C9 or CYP3A4 than monolayer cultures (Takayama et al., 2013).
Kvist et al., also measured the CYP activity in two commercial iPSC hepatocyte sources (including iCell™ hepatocytes) finding low of CYP1A2, CYP2B6, CYP2C8, and CYP2D6 activity with no detection of CYP2C19 activity (Kvist et al., 2018). CYP3A4 activity was 20–35% lower in the iPSC hepatocytes as compared to Prim-heps (Kvist et al., 2018). iPSC hepatocytes differentiated from fibroblasts had lower IC50 values for siX toXicants plus aspirin as negative control.
CYP1A2, CYP2C9 and CYP3A4 activities than cryopreserved Prim-heps (Sjogren et al., 2014). In another study, the basal activity was higher in iCell™ hepatocytes for CYP1A1, CYP2E1 and CYP3A4, but lower for CYP2B6, CYP2C8, CYP2C9, CYP2C19 and CYP2D6, in comparison to Prim-heps (Lu et al., 2015). However, Kratochwil et al., found day 4 iCell™ hepatocyte cultures had lower CYP3A4, CYP2D6, CYP2C9, CYP2B6 and CYP1A2 activity than primary hepatocyte suspension cul- tures (Kratochwil et al., 2017). In hepatocytes derived from an iPSC cell line, CYP1A, CYP2B6, CYP2C9 and CYP3A activities were lower than Prim-heps (Ulvestad et al., 2013). These studies demonstrate the vari- ability of iPSC hepatocyte CYP activity based on cell origin in compar- ison to Prim-heps under different assay conditions.
Production of conjugative metabolites was comparable in Prim-heps and 2d-iPSC-heps although 2d-iPSC-heps generally produced lower levels of conjugative metabolites than Prim-heps. 4-MU is a relatively non-specific substrate in regard to glucuronidation isozymes, whereas 1- naphthol is relatively specific for UGT1A6. UGT1A1 is responsible for estradiol 3-glucuronidation whereas the 17-conjugation is performed by the UGT2B isozymes (Donato, Montero, Castell, Go´mez-Lecho´n, & Lahoz, 2010). The results seen in this study suggest that while 2d-iPSC- heps show relatively efficient non-specific glucuronidation, some iso- zymes may be in present in much lower levels compared to Prim-heps. Because no estradiol-17-glucuronide was detected from the 2d-iPSC- heps, it is likely that the UGT2B isozymes are poorly represented. In another study with.
Sulfation enzymes have not been characterized as well as those for glucuronidation (Coughtrie, 2002), however the model substrates used in this investigation can be sulfated as well as glucuronidated (Cole et al., 2009; Mouelhi & Kauffman, 1986; Sugahara et al., 2018). This study shows that 2d-iPSC-heps have good sulfation activity. While there is some evidence of induction of conjugation enzymes in primary he- patocytes (Donato et al., 2010; Li, Hartman, Lu, Collins, & Strong, 1999), the increases tend to be modest and highly dependent on the source material. Future studies to evaluate induction of phase II enzymes would be valuable in characterizing iPSC-heps. Another study with iCell™ hepatocytes cultures exhibited lower phase II activity for UGT1A1 (SN-38) and nonspecific glucuronidation of 7-hydroXycou- marin than in Prim-hep suspension cultures or HepaRG™ cells (Kra- tochwil et al., 2017). Sulfation of 7-hydroXycoumarin was greater in iCell™ hepatocytes than for the HepaRG™ cells or Prim-hep suspension cultures (Kratochwil et al., 2017).
While basal levels of CYP metabolism were comparable in our study, there was no evidence of CYP induction in the iPSC-heps. A comparison found mRNA induction but activity
48 h exposure and Prim-Heps after both 24 and 48 h exposures (Lu et al., 2015). In contrast to the results with the commercial iCell™ hepato- cytes, a study with mature iPSC-derived hepatocyte like cells showed a 3.3-fold induction of CYP3A4 after rifampin exposure along with an elevation of PXR expression levels (Bulutoglu et al., 2020).
While this investigation measured the formation of oXidized and conjugated metabolites, quantification of phase I and II metabolism genes provide further information on the variability of metabolic capability in different iPSC-hepatocyte sources in comparison to Primheps (Gao & Liu, 2017; Kvist et al., 2018; Ma et al., 2013; Sjogren et al., 2014; Smutný et al., 2018). Meier et al., compared 2d and microspheroid cultures of hiPSC-derived hepatocyte-like cells finding the spheroid cultures had higher expression and activity of CYP3A4 than 2d cultures, although it was less than that of Prim-heps (Meier et al.,
Fig. 4. ToXicity (as expressed by ATP content) of siX different toXicants plus aspirin as a negative control. Blue lines represent Prim-heps, red lines 2d-iPSC-heps, green lines 3d-iPSC-heps. Determinations represent the mean and standard error of three replicates. Note: acetaminophen concentrations are expressed as milli- molar while all other concentrations are expressed as micromolar. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
2017). Additionally, information on the expression variability of hepatic uptake and effluX transporters between the different iPSC-hepatocyte cells in comparison to Prim-Heps will be beneficial in their further characterization as models of drug transport (Gao & Liu, 2017; Kvist et al., 2018; Sjogren et al., 2014; Ulvestad et al., 2013).
A discouraging number of drugs have been shown to be hepatotoXic in which the mechanism of this toXicity is frequently unknown. The hepatotoXic examples used here were chosen for their varied chemical structures and to the extent known for their differing mechanisms of action. Acetaminophen is a classic hepatotoXicant with an established mechanism of action requiring metabolic activation to the reactive species, N-acetyl-p-benzoquinone imine (NAPQI) (Bessems & Vermeu- len, 2001). However, concentrations much higher than known toXic blood levels are necessary to cause toXicity in cell culture (Jemnitz, Veres, Monostory, Kobori, & Vereczkey, 2008), suggesting that other factors, such as other hepatic cell types, may be involved (Nelson et al., 2015). Differing mechanisms of toXicity have been suggested for amio- darone (Wu et al., 2016), clozapine (Rowe et al., 2018), tacrine (Meng et al., 2007), tamoXifen (Yokoyama et al., 2018), and troglitazone (Shen, Meng, & Zhang, 2012), although metabolic activation has been impli- cated in most mechanisms. The current investigation indicates that 2d- iPSC-heps react to hepatotoXicants in the same rank order as Prim-heps while being incrementally less sensitive to a given compound. This suggests that hepatotoXicants act on Prim-heps and 2d-iPSC-heps in a similar manner and that 2d-iPSC-heps are capable of generating reactive metabolites. It also supports the use of 2d-iPSC-heps as a potential model for in vitro hepatotoXicity studies. These studies evaluated toXicity by measuring cellular ATP levels. Similar studies should be performed with other toXicity assays, such as those specific for cell lysis and caspase activation. ToXicity of compounds has been evaluated in other studies in comparison to Prim-heps iPSC-hepatocytes (Kang et al., 2016; Smutný et al., 2018) including high-content assays (Grimm, Iwata, Sirenko, Bittner, & Rusyn, 2015; Sirenko et al., 2014).
This study examines preliminary aspects for the determination of transporter function. CDFA enters the hepatocytes passively, where it is then de-acetylated and excreted into the canaliculi by the multidrug resistance transporter proteins MRP2 and MRP3 (Zamek-Gliszczynski et al., 2003). This allows a facile way to image the canalicular spaces that form between cultured hepatocytes. This localization is inhibited in calcium-free medium due to the formation of gaps in the canalicular junctions. By comparing the localization of appropriate substrates with and without calcium, it is possible to estimate canalicular transport of various transport proteins (Abe, Bridges, & Brouwer, 2008). The results shown here suggest that this process works in 2d-iPSC-heps and that it could be a useful way to investigate drug transport in this cell system.
As has been noted by other researchers, 2d-iPSC-heps tend to have metabolic activity that more closely resembles fetal hepatocytes than hepatocytes from adults (Lu et al., 2015). To address this, culture techniques which more closely resemble the structure in the intact liver, such as culture in three dimensions or co-culture with other cell types,Fig. 5. Phase contrast and fluorescent micrographs of (A) Prim-heps with and without 5 mM calcium for 20 min (40× magnification); (B) 2d-iPSC-heps with and without 5 mM calcium for 20 min (40× magnification); (C) 3d-iPSC-heps with and without 5 mM calcium.
have been suggested. The technique suggested by the iPSC-hep source, Cellular Dynamics International, involves culture as spheroids. This consisted of placing a fiXed number of cells in plates treated for ultra-low adhesion; the cells then aggregate spontaneously into small spheroids. In our hands, 3d-iPSC-heps produced a similar amount of basal metabolism as 2d-iPSC-heps for given cell number. While there was possible evi- dence for induction of CYP1A2 and CYP2B6 activity, this was marginal and did not reach statistical significance. There was no evidence of CYP3A4 induction in these experiments. The smaller amount of material available did not allow quantification of phase II metabolism for the 3d- iPSC-hep preparation.
The results of the toXicity studies were remarkably consistent be- tween the 2d and 3d iPSC cultures. The main difference was for amio- darone, which showed no toXicity over the tested concentration for the 3d-iPSC-heps. This was consistent over several experiments and is an observation which deserves further investigation. An advantage of the difficulty in depleting calcium in the inner portion of the spheroid. For this reason, it is unlikely this preparation will be useful in drug trans- porter assays taking advantage of this feature (Abe et al., 2008), how- ever more investigation is justified. The spheroid assay in its current form would be difficult to use for transport studies because multiple rinses are generally needed to perform these studies (Alluri et al., 2014), and since the cell mass is not adherent there are many opportunities to lose the material. Sjogren et al., compared the IC50 values between 2d and 3d (spheroid) cultures iCell™ hepatocytes after exposure to various compounds for 72 h resulting in the majority of the IC50 values not statistically significantly different between the two culture types. (Sjogren et al., 2014). This was similar to the results in the present experiments.
In summary, when considering the utility of iPSC-hepatocyte models for preclinical drug studies, this preliminary investigation supports the further investigations. An advantage of this study was the comparison of spheroid system for hepatotoXicity studies is that fewer cells are monolayer and spheroids iPSC-heps in drug metabolism, toXicity andnecessary to generate the spheroids than to produce a 96-well plate of cell monolayers, allowing the same number of cells to produce ten times the number of tests for 3d vs 2d. Both the 2d- and 3d-iPSC cultures showed good formation of canaliculi, however the 3d-iPSC-heps did not lose the signal in response to low calcium medium. This is likely due to transport assays. This study was limited by the use of a single lot of Prim- heps and iPSC-heps. Further studies with different sources and lots of hepatic iPSCs ought to be conducted in metabolism, transporter and toXicity assays to examine interindividual variability which is seen in primary human hepatocytes. While 2d-iPSC-heps show comparable basal metabolic activity to Prim-heps, there is little evidence that 2d- iPSC-heps allow induction of oXidative metabolism, which is one of the main reasons to use whole-cell preparations for metabolic studies. Studies with microsomes or recombinant enzymes can usually provide information on basal metabolic activity more specifically and at lower cost and effort. 2d-iPSC-heps form glucuronide and sulfate conjugates efficiently, but generally at a lower level than Prim-heps. On the other hand, 2d-iPSC-heps are probably very suitable for toXicity studies. They behave very similarly to Prim-heps across several different toXicants and conditions and are more consistent and available in larger amounts than Prim-heps. 2d-iPSC-heps may also be useful for transporter studies, however more research with functional assays will be necessary to fully evaluate this potential.

Disclaimer
The findings and conclusions in this article have not been formally disseminated by the Food and Drug Administration and should not be construed to represent any Agency determination or policy. The mention of commercial products, their sources, or their use in connection with material reported herein is not to be construed as either an actual or implied endorsement of such products by the U.S. Department of Health and Human Services.

Funding
This research was supported by an internal grant from the Center for Drug Evaluation and Research.

Declaration of Competing Interest
The authors declare that they have no known competing financial hepatoma cell lines. Cell Biology and Toxicology, 33(4), 407–421. https://doi.org/ 10.1007/s10565-017-9383-z. 28144825.

References

Abe, K., Bridges, A. S., & Brouwer, K. L. R. (2008). Use of sandwich-cultured human hepatocytes to predict biliary clearance of angiotensin II receptor blockers and HMG- CoA reductase inhibitors. Drug Metabolism and Disposition, 37(3), 247–252.
Agarwal, N., Popovic, B., Martucci, N. J., Fraunhoffer, N. A., & Soto-Gutierrez, A. (2019). Biofabrication of autologous human hepatocytes for transplantation: How do we get there? Gene Expression, 19, 89–95.
Alluri, R. V., Ward, P., Kunta, J. R., Ferslew, B. C., Thakker, D. R., & Dallas, S. (2014). In vitro characterization of intestinal and hepatic transporters: MRP2. In G. W. Caldwell, & Z. Yan (Eds.), Optimization in drug discovery: In vitro methods, methods in pharmacology and toxicology (pp. 369–404). New York: Springer Science Business Media.
Bessems, J. G. M., & Vermeulen, N. P. E. (2001). Paracetamol (acetaminophen)-induced toXicity: Molecular and biochemical mechanisms, analogues and protective approaches. Critical Reviews in Toxiciology, 31(1), 55–138.
Bulutoglu, B., Rey-Bedo´n, C., Mert, S., Tian, L., Jang, Y. Y., Yarmush, M. L., & Usta, O. B. (2020). A comparison of hepato-cellular in vitro platforms to study CYP3A4 induction. PLoS One, 15(2), Article e0229106. https://doi.org/10.1371/journal. pone.0229106. 32106230.
Cellular Dynamics International. (2020a). iCell® hepatocytes 2.0 user's guide. Retrieved from https://fujifilmcdi.com/assets/CDI_iCellHepatocytes2_UG.pdf.
Cellular Dynamics International. (2020b). Modeling 3D liver tissue: 3D hepatocyte spheroids in low attachment plates. Retrieved from https://dj2jug4sd9jom.cloudfron t.net/assets/CDI_iCellHepatocytes_3DSpheroids_AP.pdf.
Cole, G. B., Keum, G., Liu, J., Small, G. W., Satyamurthy, N., Kepe, V., & Barrio, J. R. (2009). Specific estrogen sulfotransferase (SULT1E1) substrates and molecular imaging probe candidates. Procedings of the National Acadamy of Sciences, 107(14), 6022–6027.
Cotovio, J. P., & Fernandes, T. G. (2020). Production of human pluripotent stem cell- derived hepatic cell lineages and liver organoids: Current status and potential applications. Bioengineering, 7(2). https://doi.org/10.3390/bioengineering7020036.
Coughtrie, M. (2002). Sulfation through the looking glass—Recent advances in sulfotransferase research for the curious. The Pharmacogenomics Journal, 2, 297–308.
Donato, M. T., Montero, S., Castell, J. V., Go´mez-Lecho´n, M. J., & Lahoz, A. (2010). Validated assay for studying activity profiles of human liver UGTs after drug exposure: Inhibition and induction studies. Analytical and Bioanalytical Chemistry, 396, 2251–2263.
Gerets, H. H. J., Tilmant, K., Gerin, B., ChanteuX, H., Depelchin, B. O., Dhalluin, S., & Atienzsar, F. A. (2012). Characterization of primary human hepatocytes, HepG2 cells, and HepaRG cells at the mRNA level and CYP activity in response to inducers and their predictivity for the detection of human hepatotoXins. Cell Biology and Toxicology, 28, 69–87.
Go´mez-Lecho´n, M. J., Castel, J. V., & Donato, M. T. (2010). The use of hepatocytes to investigate drug toXicity. Methods Tamoxifen in Molecular Biology, 640, 389–415.
Grimm, F. A., Iwata, Y., Sirenko, O., Bittner, M., & Rusyn, I. (2015). High-content assay multiplexing for toXicity screening in induced pluripotent stem cell-derived cardiomyocytes and hepatocytes. ASSAY and Drug Development Technologies., 13(9), 529–546. https://doi.org/10.1089/adt.2015.659. 26539751.
Jemnitz, K., Veres, Z., Monostory, K., Kobori, L., & Vereczkey, L. (2008). Interspecies differences in acetaminophen sensitivity of human, rat, and mouse primary hepatocytes. Toxicology In Vitro, 22, 961–967.
Kang, S. J., Lee, H. M., Park, Y. I., Yi, H., Lee, H., So, B., … Kang, H. G. (2016).Chemically induced hepatotoXicity in human stem cell-induced hepatocytes compared with primary hepatocytes and HepG2. Cell Biology and Toxicology., 32(5), 403–417. https://doi.org/10.1007/s10565-016-9342-0. 27287938.
Kratochwil, N. A., Meille, C., Fowler, S., Klammers, F., Ekiciler, A., Molitor, B., … Parrott, N. J. (2017). Metabolic profiling of human long-term liver models and hepatic clearance predictions from in vitro data using nonlinear miXed-effects modeling. AAPS Journal., 19(2), 534–550. https://doi.org/10.1208/s12248-016-0019-7. 28050713.
Kvist, A. J., Kanebratt, K. P., Walentinsson, A., Palmgren, H., O’Hara, M., Bjo¨rkbom, A. Andersson, T. B. (2018). Critical differences in drug metabolic properties of human hepatic cellular models, including primary human hepatocytes, stem cell derived hepatocytes, and hepatoma cell lines. Biochemical Pharmacology, 155, 124–140. https://doi.org/10.1016/j.bcp.2018.06.026. 29953844.
Li, A. P., Hartman, N. R., Lu, C., Collins, J. M., & Strong, J. M. (1999). Effects of cytochrome P450 inducers on 17μ-ethinylestradiol (EE2) conjugation by primary human hepatocytes. British Journal of Clinical Pharmacology, 48, 733–742.
Lu, J., Einhorn, S., Venkatarangan, L., Miller, M., Mann, D. A., Watkins, P. B., & LeCluyse, E. (2015). Morphological and functional characterization and assessment of iPSC-derived hepatocytes for in vitro toXicity testing. Toxicological Sciences, 147 (1), 39–54.
Ma, X., Duan, Y., Tschudy-Seney, B., Roll, G., Behbahan, I. S., Ahuja, T. P., … Zern, M. A. (2013). Highly efficient differentiation of functional hepatocytes from human induced pluripotent stem cells. STEM CELLS Translational Medicine., 2(6), 409–419. https://doi.org/10.5966/sctm.2012-0160. 23681950.
Meier, F., Freyer, N., Brzeszczynska, J., Kno¨spel, F., Armstrong, L., Lako, M., interests or personal relationships that could have appeared to influence the work reported in this paper.
Gao, X., & Liu, Y. (2017). A transcriptomic study suggesting human iPSC-derived hepatocytes potentially offer a better in vitro model of hepatotoXicity than most
Zellinger, K. (2017). Hepatic differentiation of human iPSCs in different 3D models: A comparative study. International Journal of Molecular Medicine, 40(6), 1759–1771.
Meng, Q., Ru, J., Zhang, G., Shen, C., Schmitmeier, S., & Bader, A. (2007). Re-evaluation of tacrine hepatotoXicity using gel entrapped hepatocytes. Toxicology Letters, 168, 140–147.
Mouelhi, M. E., & Kauffman, F. C. (1986). Sublobular distribution of transferases and hydrolases associated with glucuronide, sulfate and glutathione conjugation in human liver. Hepatology, 6(3), 450–456.
Navarro, V. J., & Senior, J. R. (2006). Drug-related hepatotoXicity. New England Journal of Medicine, 354, 731–739.
Nelson, L. J., Navarro, M., Treskes, P., Samuel, K., Tura-Ceide, O., Morley, S. D., … Plevris, J. N. (2015). Acetaminophen cytotoXicity is ameliorated in a human liver organotypic co-culture model. Scientific Reports, 5, 17455. https://doi.org/10.1038/ srep17455.
Parker, R. J., Collins, J. M., & Strong, J. M. (1996). Identification of 2,6-Xylidine as a major lidocaine metabolite in human liver slices. Drug Metabolism and Disposition, 24, 1167–1173.
Promega. (2020a). CellTiter-Glo® luminescent cell viability assay. Retrieved from https://www.promega.com/Products/Cell-Health-Assays/Cell-Viability-and-Cy totoXicity-Assays/CellTiter_Glo-Luminescent-Cell-Viability-Assay/?catNum G7570.
Promega. (2020b). P450-Glo™ assays. Retrieved from https://www.promega. com/Products/Cell-Health-Assays/ADME-Assays/P450-Glo-CYP3A4-Assay-and-Screening-System/? catNum V8801#protocols.
Rowe, C., Shaeri, M., Large, E., Cornforth, T., Robinson, A., Kostrzewski, T., … Hughes, D. (2018). Perfused human hepatocyte microtissues identify reactive metabolite forming and mitochondria-perturbing hepatotoXins. Toxicology In Vitro, 46, 29–38.
Scott, C. W., Peters, M. F., & Dragan, Y. P. (2013). Human induced pluripotent stem cells and their use in drug discovery for toXicity testing. Toxicology Letters, 219(1), 49–58. https://doi.org/10.1016/j.toXlet.2013.02.020.
Shen, C., Meng, Q., & Zhang, G. (2012). Species-specific toXicity of troglitazone on rats and human by gel entrapped hepatocytes. Toxicology and Applied Pharmacology, 258, 19–25.
Sirenko, O., Hancock, M. K., Hesley, J., Hong, D., Cohen, A., Gentry, J., … Mann, D. A. (2016). Phenotypic characterization of toXic compound effects on liver spheroids derived from iPSC using confocal imaging and three-dimensional image analysis. ASSAY and Drug Development Technologies., 14(7), 381–394. https://doi.org/ 10.1089/adt.2016.729. 27494736.
Sirenko, O., Hesley, J., Rusyn, I., & Cromwell, E. F. (2014). High-content assays for hepatotoXicity using induced pluripotent stem cell-derived cells. ASSAY and Drug Development Technologies., 12(1), 43–54. https://doi.org/10.1089/adt.2013.520. 24229356.
Sjogren, A. K., Liljevald, M., Glinghammar, B., Sagemark, J., Li, X. Q., Jonebring, A., … Andersson, T. B. (2014). Critical differences in toXicity mechanisms in induced pluripotent stem cell-derived hepatocytes, hepatic cell lines and primary hepatocytes. Archives of Toxicology, 88(7), 1427–1437. https://doi.org/10.1007/ s00204-014-1265-z (PMID: 24912781).
Smutný, T., Harjuma¨ki, R., Kanninen, L., Yliperttula, M., Pa´vek, P., & Lou, Y. R. (2018). A feasibility study of the toXic responses of human induced pluripotent stem cell- derived hepatocytes to phytochemicals. Toxicology In Vitro., 52, 94–105. https://doi. org/10.1016/j.tiv.2018.06.012. 29902661.
Sugahara, S., Fukuhara, K., Tokunaga, Y., Tsutsumi, S., Ueda, Y., Ono, M., … Yasuda, S. (2018). Radical scavenging effects of 1-naphthol, 2-naphthol, and their sulfate- conjugates. The Journal of Toxicological Sciences, 43(3), 213–221.
Takayama, K., Kawabata, K., Nagamoto, Y., Kishimoto, K., Tashiro, K., Sakurai, F., … Mizuguchi, H. (2013). 3D spheroid culture of hESC/hiPSC-derived hepatocyte-like cells for drug toXicity testing. Biomaterials., 34(7), 1781–1789. https://doi.org/ 10.1016/j.biomaterials.2012.11.029. 23228427.
ThermoFisher. (2020). Thawing and use of plateable and suspension cryopreserved hepatocytes. Retrieved from https://www.thermofisher.com/document-connect/document-connect.html?url
Ulvestad, M., Nordell, P., Asplund, A., Rehnstro¨m, M., Jacobsson, S., Holmgren, G.,Andersson, T. B. (2013). Drug metabolizing enzyme and transporter protein profiles of hepatocytes derived from human embryonic and induced pluripotent stem cells.Biochemical Pharmacology, 86(5), 691–702. https://doi.org/10.1016/j.bcp.2013.06.029. 23856292.
Williams, D. P. (2018). Application of hepatocyte-like cells to enhance hepatic safety risk assessment in drug discovery. Philosophical Transactions B, 373. https://doi.org/ 10.1098/rstb.2017.0228.
Wu, Q., Ning, B., Xuan, J., Ren, Z., Guo, L., & Bryant, M. S. (2016). The role of CYP 3A4 and 1A1 in amiodarone-induced hepatocellular toXicity. Toxicology Letters, 253, 55–62.
Yokoyama, Y., Sasaki, Y., Terasaki, N., Kawataki, T., Takekawa, K., Iwase, Y., … Ohtab, S. (2018). Comparison of drug metabolism and its related hepatotoXic effects in HepaRG, cryopreserved human hepatocytes, and HepG2 cell cultures. Biological and Pharmceutical Bulletin, 41, 722–732.
Zamek-Gliszczynski, M. J., Xiong, H., Patel, N. J., Turncliff, R. Z., Pollack, G. M., & Brouwer, K. L. R. (2003). Pharmacokinetics of 5 (and 6)-carboXy-2′,7′-dichlorofluorescein and its diacetate promoiety in the liver. The Journal of Pharmacology and Experimental Therapeutics, 304(2), 801–809.
Zanger, U. M., & Schwab, M. (2013). Cytochrome P450 enzymes in drug metabolism: Regulation of gene expression, enzyme activities, and impact of genetic variation. Pharmacology & Therapeutics, 138(1), 103–141. https://doi.org/10.1016/j. pharmthera.2012.12.007.
Zelasko, S., Palaria, S., & Das, A. (2013). Optimizations to achieve high-level expression of cytochrome P450 proteins using Escherichia coli expression systems. Protein Expression and Purification, 92(1), 77–87.