Background and Aims. Embryonic stem (ES) cells have a pluripotent ability to differentiate into a variety of cell lineages in vitro. We have recently found the emergence of cell clusters that show the cellular uptake of indocyanine green (ICG) in the culture of differentiated ES cells. ICG is clinically used as a test substance to evaluate liver function because it is eliminated exclusively by hepatocytes. The aim of the present study was to investigate the hepatic characteristics of ICG-stained cells.
Methods. Embryoid bodies (EBs), formed by a 5-day hanging drop culture of ES cells, were allowed to outgrow in the placed culture. Gene expression of hepatocyte markers was analyzed by reverse transcriptase-polymerase chain reaction, and albumin production was examined immunohistochemically. Morphology and cellular components were investigated by electron microscopy. ICG-stained cells were further transplanted into the portal vein of mice.
Results. ICG-stained cells appeared around 14 days of the EB culture and formed distinct three-dimensional structures. They were immunoreactive to albumin and expressed mRNAs such as albumin, alpha-fetoprotein, transthyretin, hepatocyte nuclear factor 3 beta, alpha-1-antitrypsin, tryptophan-2,3-dioxygenase, urea cycle enzyme, gluconeogenic enzyme, and liver-specific organic anion transporter-1. An ultrastructural analysis revealed a well-developed system of organelles such as mitochondria, lysosomes, Golgi apparatus, and rough and smooth endoplasmic reticulum. The transplantation of ICG-positive cells into the portal vein resulted in the incorporation into mice livers, where they were morphologically indistinguishable from neighboring hepatocytes.
Conclusions. ES cell-derived ICG-positive cells possess characteristics of hepatocytes, and ICG-staining is a useful marker to identify differentiated hepatocytes from EBs in vitro.
Mouse embryonic stem (ES) cells are clonal cell lines derived from the inner cell mass of developing blastocysts [1,, 2]. The distinguishing feature of these cells is their capacity to differentiate into a broad spectrum of derivatives of all three embryonic germ layers [3,, 4]. Recently, human ES cell lines were also isolated  and the mutilineage differentiating ability of human ES cell lines was reviewed in this journal . This ability has drawn attention to ES cells as a novel source of cell populations for new therapeutic strategies such as cell transplantation and tissue engineering. When ES cells are allowed to differentiate in a suspension culture, they form spherical multicellular aggregates, termed embryoid bodies (EBs), that have been shown to contain a variety of cell populations [3,, 4]. Developmental programs associated with cell lineage commitment in EBs show remarkable similarities to those found in normal embryos [3,, 4], indicating that this model system provides access to particular types of cell populations that develop in a normal fashion. To date, some success has been achieved in inducing mouse ES cells to differentiate into particular types of cells such as hematopoietic cells [7,, 8], cardiomyocytes , smooth muscle cells [10–, 12], and neurons [12–, 16], within an EB environment. However, in vitro differentiation of ES cells into hepatocytes has not yet been shown except for a recent report .
To obtain hepatocytes from ES cells, it is essential to establish not only an appropriate culture system to induce ES cells to differentiate into hepatocytes, but also a useful method for identifying differentiated hepatocytes. One of the major functions of the liver is elimination of various endogenous and exogenous compounds from the circulation. This clearance process involves basolateral (sinusoidal and lateral) membrane transport systems that mediate the hepatocellular uptake of bile acids, organic anions, and organic cations [18,, 19]. Indocyanine green (ICG) is an organic anion that is clinically used as a test substance to evaluate liver function since it is nontoxic and eliminated exclusively by hepatocytes [20–, 23]. In the present study, we examined the cellular uptake of ICG to identify differentiated hepatocytes within developing EBs. Using an EB culture system, we have found that mouse ES cells can give rise to a functional hepatocyte-like cell cluster, which forms a three-dimensional structure and demonstrates cellular uptake of ICG. To clarify the characteristic features of ICG-positive cells, we analyzed the gene expression of hepatocyte markers such as albumin, α-fetoprotein (AFP), transthyretin (TTR), hepatocyte nuclear factor 3β (HNF3β), α-1-antitrypsin (α-1-AT), tryptophan-2,3-dioxygenase (TDO), carbamoyl phosphate synthetase I (CPSase I), and phosphoenolpyruvate carboxykinase (PEPCK). In addition, we examined the gene expression of liver-specific organic anion transporter-1 (LST-1), which is expressed exclusively in the liver [24–, 26]. We also investigated the ultrastructural characteristics of ICG-positive cell clusters by electron microscopy. Finally, we transplanted ICG-positive cells into mice livers through the portal vein, and examined their grafting ability as well as immunoreactivity for albumin. Our results provide evidence of a novel source of hepatocytes for new therapeutic strategies such as cell transplantation and tissue engineering.
Materials and Methods
ES Cell Culture
Undifferentiated ES cells (G4-2) were maintained on gelatin-coated dishes without feeder cells in Dulbecco's modified Eagle's medium (DMEM; Sigma; St. Louis, MO; http://www.sigmaaldrich.com) supplemented with 10% fetal bovine serum (FBS; GIBCO/BRL; Grand Island, NY; http://www.tmc.tulane.edu/sif/tulgib.htm), 0.1 mM 2-mercaptoethanol (Sigma), 0.1 mM nonessential amino acids (GIBCO/BRL), 1 mM sodium pyruvate (Sigma), and 1,000 U/ml of leukemia inhibitory factor (LIF; GIBCO/BRL). The G4-2 cells (a kind gift from Dr. Hitoshi Niwa, Osaka University) carried the blasticidin S-resistant selection marker gene driven by the Oct-3/4 promoter (active under undifferentiated status) and were maintained in medium containing 10 μg/ml blasticidin S to eliminate differentiated cells . The G4-2 ES cells were derived from EB3 ES cells and carried the enhanced green fluorescent protein (EGFP) gene under the control of the CAG expression unit. The EB3 cells were a subline derived from E14tg2a ES cells  and generated by a targeted integration of Oct-3/4-IRES-BSD-pA vector into the Oct-3/4 allele . To induce EB formation, dissociated ES cells were cultured in hanging drops [3,, 4] with minor modifications. The cell density of one drop was 500 cells per 15 μl of ES cell medium in the absence of LIF. After 5 days in a hanging drop culture, the resulting EBs were plated onto plastic 100-mm gelatin-coated dishes and allowed to attach for the outgrowth culture.
Isolation and Culture of Hepatocytes
Hepatocytes were isolated from male 129/SVJ mice (The Jackson Laboratory; Bar Harbor, ME; http://www.jax.org) by a two-step collagenase perfusion method . Briefly, mouse livers were perfused with calcium- and magnesium-free Hanks' balanced salt solution (HBSS), and then with HBSS containing 0.05% collagenase. Hepatocytes were separated from the resulting cell suspension by differential centrifugation at 50 g for 1 minute, and then by centrifugation through a 45% Percoll gradient at 50 g for 10 minutes. Isolated hepatocytes were plated onto 100-mm collagen-coated cell culture dishes. The medium consisted of DMEM/F12 supplemented with 5% FBS.
ICG Uptake Study
ICG (25 mg; Dai-ichi Pharm. Co., Ltd; Tokyo, Japan) was dissolved in 5 ml of solvent in a sterile vial and then added to 20 ml of DMEM containing 10% FBS. The final concentration of the resulting ICG solution was 1 mg/ml. Preliminary experiments indicated that there was no adverse effect on cell viability at this concentration. The ICG solution was added to the cell culture dish and incubated at 37°C for 15 minutes. After the dish was rinsed three times with phosphate-buffered saline (PBS), the cellular uptake of ICG was examined with a stereomicroscope. After the examination, the dish was refilled with DMEM containing 10% FBS. ICG was completely eliminated from the cells 6 hours later.
Cells were fixed with 99.5% ethanol at –30°C and stained by an avidin-biotin coupling immunoperoxidase technique using a commercial kit (Vectastain Elite ABC™; Vector Laboratories; Burlingame, CA; http://www.vectorlabs.com), according to the instructions of the manufacturer. The primary antibody was a rabbit immunoglobulin G (IgG) against mouse albumin (1:100) (Cappel; Aurora, OH). After quenching endogenous peroxidase activity, cells were incubated with the primary antibody at room temperature for 1 hour. After three washes with PBS, cells were further reacted with a biotinylated goat anti-rabbit IgG as the secondary antibody at room temperature for 30 minutes. Immunostaining of albumin was visualized with 3,3′-diamino-benzidine tetrahydrochloride containing 0.01% hydrogen peroxide.
RNA Extraction and Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) Analysis
Total RNA was extracted from the cells using the acid guanidinium thiocyanate-phenol-chloroform method. One microgram of DNase-treated total RNA was used for the first-strand cDNA. This reaction was performed using Super Script II and random hexanucleotide (GIBCO/BRL), following the protocol of the manufacturer. cDNA samples were subjected to PCR amplification with specific primers under linear conditions, in order to reflect the original amount of the specific transcript. The cycling parameters were as follows: denaturation at 94°C for 1 minute; annealing at 55-60°C for 1 minute (depending on the primer); and elongation at 72°C for 1 minute (40 cycles). The PCR primers and the length of the amplified products were as follows: β-actin, (TGAACTG GCTGACTGCTGTG and CATCCTTGGCCTCAGCATAG, 174 bp); albumin, (TGAACTGGCTGACTGCTGTG and CATCCTTGGCCTCAGCATAG, 718 bp); AFP, (CCACC CTTCCAGTTTCCAG and GGGCTTTCCTCGTGT AACC, 609 bp); TTR, (AGTCCTGGATGCTGTCCGAG and TTC CTGA GCTGCTAACACGG, 440 bp); HNF3β, (TATTG GCTGCA GCTAAGCGG and GACTCGGACTCAGGT GAGGT, 508 bp); α-1-AT, (TGGGGTCTACTGCTTCTGG and TCATGGGCACCTTC ACCGT, 693 bp); TDO, (AGAGCCAGCAAAGGAGGAC and CTGTCTGC TCCT GCTCTGAT, 500 bp); CPSase I, (ATGACGAGGATTTT GACAGC and CTTCACAGAAAGGAGCCTGA, 126 bp); PEPCK, (TCTGCCAAGGTCATCCAGG and GTTTT GGG GATGGGCACTG, 290 bp); and LST-1, (AGCTACACCGAC CAAAGCTG and GTTGGCCTGCGATGCTGTC, 604 bp).
Tissues were fixed with 2.5% glutaraldehyde, 1.25 mM CaCl2, and 3% sucrose in a 0.05 M cacodylate buffer (pH 7.4) at room temperature for 3-4 hours. They were post-fixed with 1% OsO4, and then stained en bloc with uranyl acetate and embedded in epoxy resin. Ultrathin sections were double stained with uranyl acetate and lead citrate, before being examined with an electron microscope (H-7100; Hitachi; Tokyo, Japan; http://www.hitachi.com).
Transplantation Procedures and Observation of EGFP
ICG-positive clusters were isolated from the EB outgrowth cultures on day 21 under a stereomicroscope, and then trypsinized at 37°C for 5 minutes with 0.25% trypsin. The dissociated ICG-positive cells were resuspended in DMEM without FBS. A total of 2 × 103 ICG-positive cells in 0.3 ml of suspension was transplanted into male 129/SVJ mice (n = 10, 13 weeks old) (The Jackson Laboratory) through the portal vein. Four weeks after cell transplantation, the whole liver was removed from the mice, excised, and embedded in OCT compound (Tissue-TEK; Miles; Elkhart, IN), and then frozen in liquid nitrogen. Sections were cut into 10 μm thick slices with the use of a cryostat and placed on 3-aminopropyltriethoxysilane-coated slides. Specimens were examined without fixation under a fluorescent microscope (Olympus; Tokyo, Japan; http://www.olympus.com), with a 460-490 nm band-pass excitation filter and a 510 nm long-pass emission filter to detect the distribution of cells with EGFP fluorescence. All animal procedures were in accordance with our institutional guidelines as well as those of the National Institute of Health.
The same specimens were fixed with 99.5% ethanol and then incubated overnight with a rabbit IgG against mouse albumin (1:100) (Cappel). This was followed by incubation with a Rhodamine-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch; Baltimore, PA) at room temperature for 30 minutes. Immunoreactivity to albumin was examined by immunofluorescence of Rhodamine using fluorescence microscopy with a 520-550 nm band-pass excitation filter and a 580 nm long-pass emission filter. The same specimens were then stained with hematoxylin and eosin to examine the histology of the ICG-positive cells that had integrated into the native livers.
ICG Uptake within the Adhesive Outgrowth of EBs
We first examined the cellular uptake of ICG in hepatocytes isolated from adult mice livers. Most of them were positive for ICG (Fig. 1A), while undifferentiated ES cells were negative (Fig. 1B). ES cells were cultured for 5 days in a suspension culture system and allowed to form EBs. The resulting EBs were attached to gelatin-coated dishes and then spread around. Five days after beginning the outgrowth culture of EBs, various types of cells, including cardiac beating muscle cells, emerged from the outgrowths. At this stage, ICG-positive cells were not observed within the EBs (Fig. 1C). On approximately day 14 of EB culture, ICG-positive cells were apparent as clusters with a three-dimensional structure, most of which were adjacent to the cardiac beating muscles (Fig. 1D).
Immunocytochemistry for Albumin
To determine in vitro albumin production by the ICG-positive cells, we examined albumin immunoreactivity within the EB outgrowths. On approximately day 21, ICG-positive cell clusters proliferated to form more prominent three-dimensional structures (Fig. 2A and 2B). The ICG-positive clusters were intensely stained for albumin, although the albumin-immunoreactivity was observed in the surrounding cells (Fig. 2C).
Gene Expression in ICG-Positive Cells
To clarify the characteristic features of the ICG-positive cells, we analyzed the gene expression of a variety of hepatocyte markers by RT-PCR analysis (Fig. 3). RNA samples were obtained from undifferentiated ES cells, cells isolated from the ICG-negative area (ICG-negative cells), cells isolated from ICG-positive cell clusters (ICG-positive cells), and an adult liver. For the isolation of ICG-negative cells and ICG-positive cells, the outgrowths on day 21 of EB culture were used. Liver-specific serum protein genes including albumin, AFP, and TTR, as well as the endoderm-specific transcription factor gene, HNF3β mRNA, were expressed in the ICG-negative cells (lane 2), ICG-positive cells (lane 3), and adult liver (lane 4), though the levels of albumin and TTR mRNA in ICG-negative cells were much lower than those in the ICG-positive cells and adult liver. Undifferentiated ES cells expressed none of the hepatocyte markers (lane 1). The gene of α-1-AT, a glycoprotein synthesized chiefly in hepatocytes , was highly expressed in the ICG-positive cells and adult liver, whereas it was not detected in the ICG-negative cells. ICG-positive cells slightly expressed the genes of a hepatic cytosolic hemoprotein tryptophan catabolizer, TDO , and CPSase I, which is the first enzyme of the urea cycle pathway in the liver . A key regulatory enzyme of hepatic gluconeogenesis, PEPCK , and LST-1 were also expressed in the ICG-positive cells, though their mRNA levels were lower than those in the adult liver. In the ICG-negative cells, the expression of TDO, CPSase I, PEPCK, and LST-1 mRNA was not detected.
Ultrastructural Characteristics of ICG-Positive Cells
The ICG-positive cell clusters consisted of a trabecula-like cell arrangement, however, epithelial cells forming sinusoids and Kupffer cells in Space of Disse were not observed between trabeculae. The absence of both blood vessels and bile ducts was striking, as shown in a semi-thin section stained with toluidine blue (Fig. 4A). Binucleate cells were frequently observed in the cluster (inset in Fig. 4A). Electron microscopy revealed that ICG-positive cell clusters were similar to hepatocytes from a morphological viewpoint with numerous mitochondria, lysosomes, rough and smooth endoplasmic reticulum, and well-developed Golgi apparatus. Bile canaliculi were common between adjacent cells and sealed with tight junctions (Fig. 4B and 4C).
ICG-Positive Cell Transplantation
ES cell-derived ICG-positive cells were transplanted into mice livers through the portal vein. In all the mice that received graft cells, transplanted ICG-positive cells were detected by EGFP fluorescence 4 weeks after implantation (Fig. 5A and 5B), and no development of teratomas was observed. The neighboring slice was examined for albumin-immunoreactivity after fixation with ethanol. The transplanted ICG-positive cells seemed to retain an albumin-producing ability because their immunoreactivity for albumin was similar to that seen in native hepatocytes (Fig. 5C). They looked to be normally incorporated into liver parenchymal structure, morphologically indistinguishable from neighboring hepatocytes (Fig. 5D), although the double fluorescence staining for both EGFP and albumin was the only way to evaluate the albumin production of the grafted cells.
Here we showed that mouse ES cells can give rise to a hepatocyte-like cell cluster in vitro, which forms a three-dimensional structure and demonstrates cellular uptake of ICG. The ICG-positive cells were characterized by their gene expression of a variety of hepatocyte markers. Furthermore, they showed distinctive ultrastructural features characteristic of hepatocytes. Our results indicate that ES cell-derived ICG-positive cells possess genetic and morphological properties characteristic of hepatocytes.
The hepatocellular uptake of ICG is mainly mediated by Na+-independent basolateral (sinusoidal and lateral) membrane transport systems in vivo [18,, 19]. Initially, the organic anion transporting polypeptide (Oatp) was considered to be exclusive to the Na+-independent transporting mechanisms in the liver, however, it has been found to be expressed in the brain and kidney as well [34–, 38]. Recently, a novel organic anion transporter, LST-1, has been found to be expressed exclusively in the liver, even when the gene expression of Oatp is not detected [24–, 26], and is considered to be an essential molecule for transporting bile acids and organic anions in hepatocytes. The present study demonstrated that LST-1 mRNA was expressed in ICG-positive cells. Our results are consistent with a suspected role for LST-1 in the hepatocellular uptake of ICG in the liver. We consider that this ICG-staining method is easy and useful to identify highly possible candidates as hepatocytes, although further clarification of the intracellular transfer and excretion of ICG is required to verify the mechanisms of ICG-staining thoroughly.
Molecular analyses revealed that liver-specific serum protein genes such as albumin, AFP, and TTR, as well as the endoderm-specific transcription factor gene, HNF3β, were expressed in ICG-positive cells. However, the expression of these genes can be found in the yolk sac, a derivative of the extraembryonic endoderm, as well as in the fetal liver [39–, 41]. Thus, these markers are insufficient to prove in vitro differentiation into hepatocytes, though they can be used to show early endoderm commitment in EBs derived from ES cells [42–, 44]. In the present study, we indeed observed albumin immunoreactivity and mRNA expression for albumin, AFP, TTR, and HNF3β in cells other than ICG-positive clusters. In contrast, the gene of α-1-AT, a glycoprotein synthesized chiefly in hepatocytes , was highly expressed in ICG-positive cells. Furthermore, ICG-positive cells expressed the genes related to hepatic metabolism, including TDO (a tryptophan catabolizer) , CPSase I (the first enzyme of the urea cycle pathway) , and PEPCK (a key regulatory enzyme of hepatic gluconeogenesis) , whereas these genes were not expressed in ICG-negative cells. These results indicate that ICG-positive cell differentiation is specified toward hepatocytes and that these cells possess the genetic characteristics of almost terminally differentiated hepatocytes.
Our ultrastructural observations provided more convincing evidence that the ES cell-derived ICG-positive cells are characterized as hepatocytes. ICG-positive cells contained well-developed organelles such as mitochondria, lysosomes, Golgi apparatus, and rough and smooth endoplasmic reticulum. Notably, a bile canaliculus was observed in the intercellular space of adjacent two cells. Furthermore, we found many binuclear cells, which were commonly found in vivo. ICG-positive cells connected to each other by tight junctions that demarcate the limits of a bile canalicular space. Adjacent hepatocytes isolated from the liver usually form a primary unit of bile canalicular function including bile formation and secretion, since the apical canalicular polarity is retained between the two adjacent hepatocytes [45–, 47]. Although ICG-positive cell clusters lacked some components such as blood vessels and bile ducts, they possessed morphological properties of hepatocytes.
In the present study, we showed that ES cells can differentiate into “hepatocytes” within developing EBs without the addition of growth or differentiation factors, suggesting that EBs themselves provide an appropriate microenvironment that supports hepatogenesis in vitro. It is well known that EBs can generate all three cellular lineages indicative of endoderm, mesoderm, and ectoderm within the first few days of differentiation, and mimic a developing embryo [3,, 4]. Interestingly, most of the cellular clusters showing the uptake of ICG were adjacent to differentiated cardiac beating muscles. This finding may support the recent work of Zaret and coworkers [48–, 51], who have reported that hepatic differentiation requires interaction with the adjacent cardiac mesoderm producing multiple fibroblast growth factors in the mouse embryo.
Finally, we found that ES cell-derived ICG-positive cells transplanted through the portal vein were normally incorporated into mice liver parenchymal structure, without forming a teratoma. They were morphologically indistinguishable from neighboring hepatocytes and seemed to retain an albumin-producing ability for at least 4 weeks. These results raise the possibility that ES cells can provide a novel source of hepatocytes for cell transplantation. In light of cell transplantation therapy, the present results make an important contribution to attempts to isolate large numbers of living hepatocytes from EBs, which may circumvent the problem of a lack of donor material.
We thank Dr. Hitoshi Niwa for his critical review of the manuscript, Dr. Hiroshige Nakano for his helpful discussion, and Shigeko Torihashi, Masashi Ando, Sanehito Ogawa, and Yae Hanesaka for their expert technical assistance. This work was supported by a Research Grant-in-Aid from the Ministry of Education, Science, and Culture of Japan.