Application of Live Mitochondria Staining in Cell-Sorting to Purify Hepatocytes Derived from Human Induced Pluripotent Stem Cells

Embryonic and induced pluripotent stem cells (ES and iPS, respectively) are considered promising cell sources for regenerative therapies. However, the efficiency of differentiating these cells into specific target cell types can vary, even when using the same cell line, protocol, and experimenter^1,2,3,4. This variability may be attributed to the high sensitivity of human ES/iPS cells to their environment. Therefore, it is currently difficult to consistently obtain pure target cells. To achieve highly safe regenerative medicine, it is crucial to eliminate proliferative cells and undifferentiated stem cells in therapeutic cells, and advanced purification technology for target cells is essential^5,6,7.

A cell sorter is a device that instantaneously analyzes individual cells and sorts live cells of interest based on the fluorescent signal strengths, offering a promising solution. This can be accomplished through antibody staining of cell type-specific surface markers or by utilizing cell type-specific reporter gene expressions. Using this technique, there are several reports on methods for purifying pluripotent stem cell-derived cardiomyocytes^8,9,10 and hepatocytes^11,12. Hattori et al. developed an innovative mitochondrial purification method using a cell sorter¹³. Taking advantage of the fact that cardiomyocytes have high energy demands through mitochondrial activity, staining the cells with the live mitochondria-indicative dye TMRM (tetramethylrhodamine methyl ester) can be used to label and highly purify cardiomyocytes by FACS from human ES cell-derived embryoid bodies containing various cell types. The absence of tumorigenicity was confirmed by teratoma formation assays with the purified cardiomyocytes. Furthermore, Yamashita et al. unexpectedly discovered a method to purify hepatocytes from human ES cell-derived embryoid bodies by isolating fractions with high mitochondrial activity and ALCAM-positive expression¹⁴. The rationale for this method is that hepatocytes also have a relatively high number of mitochondria due to their high consumption of ATP for nutrient metabolism and detoxification¹⁵, and hepatocytes express ALCAM, a member of the immunoglobulin superfamily, which plays a role in cell adhesion and migration¹⁶.

Previous highly purifying methods for pluripotent stem cell-derived hepatocytes required genetic modifications, and non-genetic purification methods had low efficiency¹⁷. The mitochondrial non-genetic method holds merit for achieving high purity. When hepatic progenitor cells are needed, CD133 and CD13- or Dlk1-based methods^18,19 can be chosen. Although the accuracy of genome editing technology has advanced, the potential risk of unforeseen genomic changes (e.g., carcinogenesis) cannot be entirely eliminated. Methods based on mitochondrial activity, without involving genetic modification, can be free from such risks.

As a result of examining various mitochondrial indicators in neonatal rat cardiomyocytes, TMRM was observed to disappear completely within 24 h, whereas other dyes remained for at least 5 days¹³. Moreover, it is important to note that TMRM and JC-1 did not impact cell viability when assessed with the 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, while other dyes demonstrated varying effects on cell viability¹³. TMRM demonstrates higher safety.

To date, several two-dimensional (2D) differentiation methods have been developed as more efficient approaches for inducing differentiation compared to three-dimensional (3D) embryoid body formation. This is because step-wise differentiation-inducing compounds or cytokines can be uniformly administered to cells on a 2D plane, rather than in a 3D space. In this study, the previously reported method²⁰ was modified to induce differentiation into human iPS cell-derived hepatocytes. Here, the details of the procedures for the up-to-date 2D differentiation and purification of hepatocytes derived from human iPS cells are described.

This study used commercially obtained human iPS cells (253G1 strain) (see Table of Materials).

1. Maintenance of human iPS cells

1. Maintain human iPS cells in feeder-free conditions in AK02 medium on culture dishes coated with 0.25 µg/cm² iMatrix-511 (see Table of Materials).

2. Hepatic differentiation of human iPS cells in 2D cultures

NOTE: Refer to Figure 1A for the daily schedule of hepatic differentiation. Throughout the process, use 6-well culture plates and add 2 mL of the designated wash or culture medium to each well. Maintain the cell culture condition at 100% humidity, 5% CO₂, and 20% O₂ at 37 °C.

1. Coat each well of a 6-well plate with 1 mL of basement membrane matrix (diluted 1/25 in PBS (-)), and incubate at 37 °C for 1 h.
2. Passage human iPS cells onto the coated dish at a density of 20,000-30,000 cells/cm² using AK02N medium supplemented with 10 µM ROCK inhibitor (Y-27632) (see Table of Materials). Replace the medium without Y-27632 the following day and culture for an additional 2 days.
3. Wash the cells once with RPMI-1640 medium. Initiate endoderm differentiation by replacing the medium with RPMI-B27-Glu (RPMI-1640 + 2% B27 Insulin + 1% Dipeptide L-Alanyl-L-Glutamine) supplemented with 3-6 µM CHIR-99021 and 100 ng/mL Activin A for 24 h (see Table of Materials).
4. Replace the medium with RPMI-B27-Glu supplemented with 50 ng/mL Activin A for 24 h.
5. To induce hepatic progenitor differentiation, replace the medium with RPMI-B27-Glu supplemented with 1% dimethyl sulfoxide (DMSO) for 5 days.
6. To promote hepatocyte maturation, switch to hepatocyte maturation medium: Leibovitz's L-15 medium containing 10 µM hydrocortisone 21-hemisuccinate, 8.3% tryptose phosphate broth, 100 nM dexamethasone (DEX), 50 µg/mL sodium-L-ascorbate, 0.58% insulin-transferrin-selenium (ITS), 8.3% fetal bovine serum (FBS), 2 mM Dipeptide L-Alanyl-L-Glutamine, and 100 nM N-Hexanoic-Try-Ile-(6)-amino Hexanoic amide (Dihexa) (see Table of Materials).
7. Change the medium every 2 days for 20 days. Sort the cells approximately 27 days after initiating differentiation.

3. Cell preparation for purification of human iPS cells-derived hepatocytes

1. Wash hepatic differentiated cells with 2 mL PBS (-) per well.
2. Add 1 mL enzyme solution per well, which includes 0.1% collagenase, 0.083% trypsin, 10 µM ROCK inhibitor (Y-27632), 20 nM ciclosporin A, and 50 µg/mL sodium-L-ascorbate in Ads buffer (116 mM NaCl, 12.5 mM NaH₂PO₄, 20 mM HEPES, 5.4 mM KCl, 5.6 mM glucose, and 0.8 mM MgSO4; pH 7.35) (see Table of Materials). Detach cells from the plates by horizontally rotating at 37 °C for about 2 h.
3. After confirming that cells are detached and rounded under a microscope, disperse them into single cells by gentle pipetting, and collect them in a 50 mL tube.
4. Centrifuge cells at 200 x g for 5 min at 18 °C to pellet them. Remove the supernatant using an aspirator.
5. Stain the mitochondria of the cells by dispersing them in 5 mL of hepatocyte maturation medium containing 100 nM TMRM (tetramethylrhodamine, methl ester) (see Table of Materials). Incubate for 30 min at 37 °C while protecting from light by wrapping in aluminum foil.
6. Centrifuge the cells at 200 x g for 5 min at 18 °C to pellet them. Remove the supernatant.
7. Resuspend the cells with 1-5 mL of cold Ads buffer containing 2% FBS.
8. Count the cell number using a 0.4% Trypan blue solution and a cell count plate.
9. Centrifuge the cells at 200 x g for 5 min at 4 °C to pellet them. Remove the supernatant.
10. Dilute an anti-ALCAM antibody (primary antibody, see Table of Materials) at a 1:50 ratio in Ads buffer, and add 100 µL of the diluted antibody per 10⁶ cells. Stain the cells for 50 min on ice. Prepare cells that are not stained with the primary antibody as a negative control.
11. Wash the cells with cold Ads buffer containing 2% FBS twice.
12. Dilute an Alexa Fluor 488 donkey anti-goat IgG (secondary antibody, see Table of Materials) at a 1:100 ratio in Ads buffer, and add 100 µL of the diluted antibody per 10⁶ cells. Stain the cells for 30 min on ice. Stain cells that were not stained with the primary antibody using the secondary antibody.
13. Wash the cells with cold Ads buffer containing 2% FBS twice.
14. Resuspend the cells in Ads buffer and filter them through a snap cap cell strainer (35 µm mesh) just before sorting by FACS to remove large cell clumps and debris.

4. FACS analyses and sorting of human iPS cells-derived hepatocytes

1. Set up the FACS machine for cell sorting following the manufacturer's instructions (see Table of Materials).
   NOTE: Use an 85 µm or 100 µm nozzle size and an ND filter of 2.0 for hepatocytes. Use the FITC channel for detecting Alexa Flour 488 for ALCAM and the PE channel for detecting TMRM fluorescence for mitochondria.
2. Adjust the FSC (forward scatter) and SSC (side scatter) voltages to properly visualize the cell population.
3. Adjust the voltages of the FITC and PE fluorescent channels. Start with unlabeled cells and center the cell population at the bottom of each channel plot. Ensure labeled cells are appropriately positioned within the plot.
4. Use a common gating strategy to eliminate doublets: Gate the major FSC-A vs. SSC-A population (Figure 2A), followed by FSC-H vs. -W, and then SSC-H vs. -W (Figure 2B,C).
5. Gate the TMRM^hi and ALCAM⁺ population (P2 in Figure 2E) to sort hepatocytes. Gate the TMRM^hi and ALCAM^- population (P1 in Figure 2E) to sort non-hepatocytes. Collect sorted hepatocytes and non-hepatocytes in 15 mL collection tubes containing hepatocyte maturation medium.
6. Centrifuge the collection tube containing the sorted cells at 200 x g for 5 min at 18 °C.
7. Remove the supernatant and resuspend the cells in hepatocyte maturation medium.
8. Seed the cells on mitomycin C-treated mouse embryonic fibroblasts (MEFs) or gel-like basement membrane matrix with hepatocyte maturation medium supplemented with 10 µM Y-27632 and 20 nM ciclosporin A. Culture the cells in a 37 °C CO₂ incubator for 5 days.

5. Immunocytochemistry for detection of human iPS cells-derived hepatocytes

1. Wash the cells cultured for 5 days after sorting with PBS (-).
2. Fix the cells with 4% paraformaldehyde for 5 min at room temperature.
3. Wash the cells with TBS-T buffer (1x TBS with 1% Tween 20) twice to remove the paraformaldehyde.
4. Incubate the cells with 0.1% Triton X-100 diluted TBS-T buffer for 5 min at room temperature.
5. Wash the cells with TBS-T buffer twice to remove Triton X-100.
6. Add commercially available blocking solution (see Table of Materials) to the cells for 30 min at room temperature.
7. Add primary antibodies against albumin (hepatocyte marker) and human nuclear antigen (hNA, see Table of Materials), diluted 1:50 in blocking solution.
8. Incubate the cells at 4 °C overnight. Wash the cells with TBS-T buffer twice.
9. Add secondary antibodies: Alexa Fluor 488 donkey anti-rabbit IgG and Alexa Fluor 546 donkey anti-mouse IgG, diluted 1:100 in blocking solution (see Table of Materials).
10. Incubate the cells at room temperature for 30 min. Wash the cells with TBS-T buffer twice.
11. Evaluate hepatocyte purity by counting albumin-positive and hNA-positive cells under a fluorescence microscope. Calculate the percentage of albumin-positive cells per hNA-positive cell for both the P1 and P2 populations.

The timeline of the processes inducing human iPS cells to differentiate into hepatocytes through 2D culture (Figure 1A) and representative cell features (Figure 1B) are shown. Approximately on differentiation day 12, cells began to exhibit polygonal cell shapes and round nuclei, characteristic of hepatocytes. Some hepatocytes also displayed multinucleation.

FACS analysis was performed using the cells on differentiation day 27. To identify the ALCAM⁺ population, comparative plots with and without the first anti-ALCAM antibody are shown in Figure 2D,E, respectively. Inducing hepatic differentiation through 2D culture, the population of hepatocytes (P2) was found to be 10.1% in this experiment, representing a significant increase from the 0.8% (data not shown) found using the 3D embryoid body formation method.

P1 and P2 cells that had been sorted were cultured on mouse embryonic fibroblasts treated with mitomycin C. P2 cells exhibited a compact, spherical colony formation, whereas P1 cells were dispersed and had a flatter morphology. Immunohistochemical staining for human nuclear antigen (hNA) and albumin confirmed the enrichment of hepatocyte-like cells in P2, as depicted in Figure 3. To assess the purity of hepatocytes in P1 and P2 cells, the ratio of albumin-positive cells to hNA-positive cells was calculated (n = 5). The results revealed that 0% of P1 cells and 97% ± 0.38% of P2 cells were derived from human iPSCs and displayed hepatocyte-like characteristics (Figure 3).

To confirm the successful sorting of hepatocytes during the FACS experiment, it is recommended to co-stain ASGR1 with its specific antibody conjugated with different fluorophores and confirm that P2 contains ASGR1-positive cells.

Figure 1: 2D Hepatic differentiation of human iPSCs. (A) The timeline for hepatic differentiation. This figure is adapted from Yamashita et al.¹⁴. (B) Morphological changes during hepatic differentiation. Scale bars = 200 µm. Please click here to view a larger version of this figure.

Figure 2: Gating strategy for sorting hepatocytes and non-hepatocytes in hepatic differentiated cells (day 27). (A) Gating the target cell population using the FSC-A vs. SSC-A plot. (B) Expanding the gated population from step (A) using the FSC-H vs. FSC-W plot and exclude doublet cells. (C) Further expanding the gated population from step (B) using the SSC-H vs. SSC-W plot and excluding doublet cells. (D) and (E) Expanding the gated population from step (C) using the FITC-A (ALCAM) vs. PE-A (TMRM) plot. (D) Stained cells with TMRM and the 2^nd antibody. (E) Stained cells with TMRM and the 1^st and 2^nd antibodies. Please click here to view a larger version of this figure.

Figure 3: Immunohistochemistry results. Immunohistochemical detection of albumin and human nuclear antigen (hNA) in sorted P1 and P2 cells, and quantification of albumin (+) cell-fraction in hNA (+) cells. Scale bars = 100 µm. Data are represented as mean ± standard deviation (SD) (n = 5). Statistical analysis was conducted using the Student's t-test. *P < 0.05. This figure is adapted from Yamashita et al.¹⁴. Please click here to view a larger version of this figure.

Due to their functions in nutrient metabolism and detoxification, hepatocytes possess a relatively large number of mitochondria compared to other cell types¹⁵. ALCAM is a member of the immunoglobulin superfamily and plays a role in cell adhesion and migration. It is expressed in various cell types, including hepatic, epithelial, lymphocytic, myeloid, fibroblast, and neuronal cells¹⁶. By utilizing a combination of the mitochondria-based method and the ALCAM antibody, human iPS cell-derived hepatocytes were successfully identified.

While the assays to characterize the secretory function or enzyme activity of the purified hepatocytes were not performed, the mRNA expression of the CYP3A4 gene in human iPS cell-derived hepatocytes purified were verified. Furthermore, hepatocytes population with TMRM (high) and ALCAM (+) isolated from embryoid bodies derived from human ES cells, that were observed higher expressions of hepatocyte-associated gene sets (VTN, SERPINA1, CYP1A1, FGB, FGA, FGG, AFP, ALB, and APOB) compared to the ES cell-derived cardiomyocytes¹⁴.

A critical step within the protocol is to gently detach 2D hepatic differentiated cells from the culture dishes with minimal damage while ensuring the preservation of mitochondrial activity for TMRM staining. Depending on the state of differentiation, hepatocytes may exhibit resistance to detachment from the dish even with enzyme treatment, a challenge that can be effectively addressed by the addition of 0.25 mM EDTA.

Existing methods for the purification of hepatocytes via FACS involve the introduction of reporter genes downstream of promoters specific to hepatocyte-expressed genes. However, this approach carries the potential risk of side effects associated with genetic modifications. The mitochondrial method, on the other hand, eliminates such concerns. Several cell surface markers of adult or fetal liver stem/progenitor cells have been identified¹⁸, and hepatic progenitor cells derived from human pluripotent stem cells could be sorted using the CD13+CD133+ population¹⁹. On the other hand, among pluripotent stem cell-derived hepatocytes, a small fraction of the matured hepatocytes could be sorted using the ASGR1 (asialoglycoprotein receptor 1) antibody¹⁷, which is the only cell surface marker specific to hepatocytes. It was demonstrated that the population of cells exhibiting high mitochondrial and ALCAM positivity contained both populations marked with CD13+CD133+ cells and ASGR1+ cells. This suggests that high mitochondrial and ALCAM-positive populations may contain a wide range of hepatocytes with various maturities¹⁴. Consequently, this method may enable the purification of a hepatocyte population with little loss.

Analysis of mouse fetal whole viscera at embryonic day 14.5 using this method revealed the expression of ALCAM in both myocytes and hepatocytes¹⁴. Other previous studies have reported ALCAM expression in murine myocardial progenitor cells²¹. In the case of human pluripotent stem cell-derived cells, the cardiomyocytes were identified in ALCAM negative population. These data show that ALCAM expression in cardiac and hepatic lineages differs depending on both the species and the cell maturity stages. However, due to the higher mitochondrial fraction in cardiomyocytes compared to hepatocytes, it was actually succeeded that to separate the two cell types by FACS even from mixed cells derived from the mouse whole embryo. This suggests that the proposed method could be useful for purifying hepatocytes and cardiomyocytes from other animal species.

The hepatocyte purification technology described herein is anticipated to play a pivotal role in advancing the field of regenerative medicine by mitigating the potential risk of tumor formation associated with the transplantation of human iPS cell-derived hepatocytes. However, negation of a teratoma-forming ability in the hepatocyte population purified by this technique has not been done. Further studies for safety assurance have to be done.

Large and adherent cells such as hepatocytes are highly vulnerable to mechanical damage during FACS sorting, resulting in significant loss of viable cells. For clinical applications, it is essential to develop a new purification method that establishes a culture environment enabling the survival of hepatocytes exclusively such as lactate method for cardiomyocyte purification²².