Differentiation of putative hepatic stem cells derived from adult rats into mature hepatocytes in the presence of epidermal growth factor and hepatocyte growth factor

C Blackwell Verlag 2003

Differentiation (2003) 71:281–290

ORIGINAL ARTICLE

Z. P. He ¡ W. Q. Tan ¡ Y. F. Tang ¡ M. F. Feng

Differentiation of putative hepatic stem cells derived from adult rats into mature hepatocytes in the presence of epidermal growth factor and hepatocyte growth factor

Accepted in revised form: 28 March 2003

Abstract Oval cells, putative hepatic stem cells, can differentiate into a wide range of cell types including hepatocytes, bile epithelial cells, pancreatic cells and intestinal epithelial cells. In this study, we used different growth factor combinations to induce oval cells to differentiate into mature hepatocytes. We isolated and purified oval cells utilizing selective enzymatic digestion and density gradient centrifugation. Oval cells were identified by their morphological characteristics and the strong expressions of OV-6, albumin, cytokeratin (CK)19 and CK-7. Using a 2-step induction protocol, we demonstrated that oval cells first changed into small hepatocytes, then differentiated into mature hepatocytes. Small hepatocytes were distinguished from oval cells by their morphological features (e.g. round shape and nuclei) and the lack of CK-19 mRNA expression. Mature hepatocytes were identified by their ultrastructural traits and their expressions of albumin, CK-18, tyrosine aminotransferase (TAT), and alpha-1-antitrypsin (a-1-AT). Differentiated cells acquired the functional attributes of hepatocytes in that they secreted albumin and synthesized urea at a high level throughout differentiation. Oval cells can thus differentiate into cells with the morphological, phenotypic and functional characteristics of hepatocytes. This 2-step induction procedure could provide an abundant source of hepatocytes for cell transplantation and tissue engineering. Key words oval cells ¡ differentiation ¡ small hepatocytes ¡ mature hepatocytes ¡ epidermal growth

✉

Z. P. He ¡ W. Q. Tan ¡ Y. F. Tang ¡ M. F. Feng ( ) State Key Lab of Biomembrane and Membrane Biotechnology Institute of Zoology, Chinese Academy of Sciences Beijing 100080, China e-mail: fengmf/panda.ioz.ac.cn Tel: π8610 6262 8740, Fax: π8610 6257 1017 U. S. Copyright Clearance Center Code Statement:

factor ¡ hepatocyte growth factor ¡ hepatic stimulator substance

Introduction Oval cells have a potential to differentiate into a variety of cell lineages in vitro, including hepatocytes and bile epithelial cells (Golding et al., 1995; Lazaro et al., 1998; Crosby et al., 2001). Although oval cells are thought to be bipotential stem cells, they can transdifferentiate into intestinal epithelial cells and pancreatic cells when transplanted into a duodenal wall or pancreas (Tatematsu et al., 1985; Suzuki et al., 2002; Yang et al., 2002). This indicates that oval cells can be used to generate different cell types for use in cell therapy and regenerative tissue engineering. On the other hand, the shortage of hepatocytes limits the wider application of hepatocyte transplantation and tissue engineering such as bioartificial liver. One attractive solution to this problem would be to use the ability of oval cells to proliferate and differentiate into mature and functional hepatocytes (Faris et al., 2001; Strain and Neuberger, 2002; He et al., 2003). Of great concern is the finding that hepatocellular carcinoma and cholangiocarcinoma may have arisen from the abnormal differentiation of oval cells or their maturation arrest (Sell and Pierce, 1994; Hixson et al., 1997). Oval cells are believed to play a key role in the development of hepatocellular carcinoma and cholangiocarcinoma and act as tumor progenitor cells in human hepatocarcinogenesis (Faris et al., 1991; Ruck et al., 1996; Dumble et al., 2002). Furthermore, it has been demonstrated that hepatocellular carcinoma does not result from the dedifferentiation of hepatocytes but from the abnormal differentiation of oval cells (Sell, 1990). It is therefore essential to induce oval cells to differentiate into mature hepatocytes in order to minimize the possibility of aberrant maturation and malignant transfor-

0301–4681/2003/7104–281 $ 15.00/0

282

mation. This paper describes the induction of mature and functional hepatocytes from oval cells isolated from adult rat livers via a 2-step procedure.

acquired on the flow cytometer (Becton Dickinson, San Jose, CA, USA), and the data were analyzed with Cell-Quest software (Becton Dickinson, San Jose, CA, USA). Replacement of the primary antibody with PE-labeled anti-rabbit or anti-mouse IgG served as a negative control.

Methods

In vitro differentiation of oval cells into hepatocytes

Animals and oval cell activation

Oval cells were plated at a density of 1 ¿ 105 cells per 35-mm dish with a feeder layer composed of fibroblasts treated with mitomycin C overnight at 37 æC. DMEM medium was supplemented with 10 % fetal calf serum, 0.5 mg/mL hydrocortisone (Sigma, St. Louis, MO, USA), 100 U/mL penicillin G and 100 mg/mL streptomycin. Oval cells were induced to differentiate into hepatocytes by the 2step induction procedure illustrated in Fig. 1. Each experiment was replicated 4 times. Fresh medium was provided every day and cell growth was carefully examined under a phase-contrast microscope. For the 2-step induction, oval cells were first cultured in the medium supplemented with 50 ng/mL epidermal growth factor (EGF) or 50 mg/mL hepatic stimulator substance (HSS) for 5 days. Thereafter, 100 ng/mL hepatocyte growth factor (HGF) was added to the medium for another 5-day induction. For the 1-step induction, oval cells were treated with 100 ng/mL HGF alone. The purification and identification of HSS were performed according to previously published methods (Fleig and Hoss, 1989; Liu et al., 1998).

Male Wistar rats (120 – 150 g) obtained from the specific pathogenfree (SPF) laboratory animal-breeding center (Institute of Zoology, Chinese Academy of Sciences) were provided with standard rat chow and water ad libitum. The model of 2-acetylaminofluorene (AAF) treatment with partial hepatectomy was used to activate oval cells, as described previously (Thorgeirsson et al., 1993). Oval cell isolation The isolation and purification of oval cells were performed according to the protocol of Pack et al. with some modifications (Pack et al., 1993). Briefly, liver tissue was minced in Hank’s balanced salt solution (HBSS) (GIBCO, BRL, NY, USA) and digested in Dulbecco’s minimum essential medium (DMEM) (GIBCO, BRL, NY, USA) containing 0.10 % w/v collagenase IV (Sigma, St. Louis, MO, USA) and 0.025 % w/v EDTA (Sigma, St. Louis, MO, USA) for 15 min at 37 æC in a shaking bath while a 2-step collagenase perfusion was accomplished. After centrifugation at 500 rpm for 5 min, the supernatant was recentrifuged at 1500 rpm for 5 min. The pellets were suspended in 50 mL of DMEM containing 0.10 % w/v Pronase E (Sigma) and 0.005 % w/v DNase I (Sigma, St. Louis, MO, USA), and incubated for 30 min at 37 æC. Thereafter, the suspension was centrifuged at 1500 rpm for 5 min at 4 æC, and the pellets were resuspended in 2 mL HBSS. The Percoll density gradient was as follows: 1.20 g/mL Percoll, 1.12 g/mL Percoll, cell suspension and HBSS. After centrifugation at 1500 rpm for 20 min at 4 æC, the interface between 1.20 g/mL and 1.12 g/mL Percoll was decanted and recentrifuged at 1500 rpm for 5 min. The viability of oval cells was estimated by the staining of 0.25 % w/v trypan blue. Transmission electron microscopy (TEM) Freshly isolated cells, less-differentiated cells, and fully differentiated cells were fixed in 2.5 % w/v glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4). After extensive washing in phosphate-buffered saline (PBS), the cells were post-fixed in 1 % w/v OsO4 for 30 min, dehydrated in a graded solution of ethanol and embedded in Epon. Ultrathin sections were cut and examined under a Phillips EM 301 electron microscope after staining with uranyl acetate and lead citrate. Flow cytometric analysis Freshly isolated cells were fixed in cold acetone for 10 min at 4 æC. After centrifugation at 1500 rpm for 5 min at 4 æC, pellets were suspended in 0.1 mL PBS and stained with the following antibodies respectively: albumin (rabbit anti-human polyclonal), cytokeratin (CK)-19 (mouse anti-rat monoclonal), CK-7 (mouse anti-rat monoclonal), CD34 (mouse anti-human monoclonal) and OV-6 (mouse anti-rat monoclonal) (a kind gift from Prof. Stewart Sell, Albany Medical College) for 60 min at 37 æC. All the primary antibodies except for OV-6 were purchased from DAKO Corporation (Kyoto, Japan). Centrifugation was performed at 1500 rpm for 15 min and followed by extensive washes with PBS. The pellets were suspended in 0.1 mL PBS to prepare for cell suspension. Binding of the primary antibody was detected by a phycoerythrin (PE)labeled IgG (Becton Dickinson, San Jose, CA, USA). Cells were

Scanning electron microscopy (SEM) For SEM studies, less-differentiated cells generated by first-step induction were grown on coverslips rinsed with pre-warmed PBS without Ca2π and Mg2π, and fixed in 3.5 % v/v glutaraldehyde (0.1 M phosphate buffer, 2 % v/v sucrose, pH 7.3) for 2 h at 4 æC. After several washes in distilled water, the cells were dehydrated in alcohol and dimethoxypropane, and sputter-coated with 10 nm gold. SEM examination was carried out with a JSM-5600LV electron microscope. Immunofluorescence analysis Immunofluorescence analysis of fully differentiated cells derived from colonies was performed at day 11. The cells were fixed in cold acetone for 5 min, blocked with normal goat serum after extensive washes with PBS (pH 7.4), and incubated with primary antibodies including albumin, CK-18 (mouse anti-human monoclonal), CK19 and CK-7 respectively overnight at 4 æC. After three washes in PBS, the cells were incubated with PE-labeled IgG for 30 min at 37 æC, washed again 3 times with PBS, and observed microscopically for epifluorescence. Replacement of the primary antibody with PE-labeled anti-mouse and anti-rabbit IgG served as a negative control. RNA extraction and reverse transcriptase polymerase chain reaction (RT-PCR) Total RNA was extracted from oval cells, less-differentiated cells, fully differentiated cells and bile epithelial cells using TRIzol (Life Technologies, Rockville, MD, USA). To eliminate genomic DNA contamination, mRNA was purified using oligo(dT) cellulose (MicroFastTrack 2.0 kit, Invitrogen, California, USA) according to the manufacturer’s protocol. One microgram of total RNA was used for the first-strand cDNA, and cDNA samples were subjected to PCR amplification with specific primers under linear conditions. The cycling parameters were as follows: denaturation at 94 æC for 1 min; annealing at 60 æC for 1 min; and elongation at 72 æC for 1 min (35 cycles). The PCR primers used for amplification were as follows: albumin (forward 5ø-CATGACACCATGCCTGCTGAT3ø, and reverse 5ø-CTCTGATCTTCAGGAAGTGTA-3ø, 618 bp);

283

Fig. 1 Flow diagram of induction procedure.

CK-19 (forward 5ø-GTCCTACAGATTGACATTGC-3ø, and reverse 5ø-CACGCTCTGGATCTGTGACAG-3ø, 425 bp); alpha-1antitrypsin (a-1-AT) (forward 5ø-ATGGATTACCTGGGCAACGC-3ø, and reverse 5ø-TTTTCCCACAAAGAGGGGGC-3ø, 398 bp); and tyrosine aminotransferase (TAT) (forward 5ø-TACTCAGTTCTGCTGGAGCC-3ø, and reverse 5ø-GCAAAGTCTCTAGAGAGGCC-3ø, 471 bp). The PCR products were separated by electrophoresis in 1.5 % agarose gels.

mitochondria and rough endoplasmic reticulum (Fig. 2C,D). Compared with hepatocytes, oval cells were relatively small, with a median diameter ranging from 6 to 8 mm.

Albumin and urea assay

We analyzed the phenotypic characteristics of freshly isolated cells in order to confirm their identity. Flow cytometric analysis showed that 86.48 % of the cells expressed OV-6, an antigen specific to rodent oval cells (Fig. 3D). This provided conclusive evidence that most of the cells were oval cells. A single staining assay revealed that 84.71 % of the cells were positive for albumin (Fig. 3C), and more than 85 % of the cells were CK19 and CK-7 positive (Fig. 3E,F). To further clarify the characteristic phenotypes of freshly isolated cells, we also examined the gene expressions of hepatocytic and cholangiocytic markers by RT-PCR analysis. Freshly isolated cells were co-expressing albumin and CK-19 mRNA at high levels (Fig. 7, lane 1), thereby demonstrating their bipotency. Taken together, these tests showed that the freshly isolated cells expressed both hepatocyte and cholangiocyte markers, indicating that they retained the bipotential nature of hepatic stem cells. In addition, 15.65 % of the cells expressed CD34 (Fig. 3B). These CD34-positive cells may have been derived from an extrahepatic source such as bone marrow or peripheral blood (Petersen et al., 1999; Theise et al., 2000; Avital et al., 2001; Petersen, 2001).

Albumin production from cultured medium was measured to evaluate the function of hepatocytes using the ELISA method, as described previously (Lazaro et al., 1998). Urea production was determined by colorimetric assay (measure absorbance at 340 nm). Rat hepatocytes grown in monolayer were used as a positive control and culture medium was used as a negative control.

Results Morphological characteristics of freshly isolated cells We obtained approximately 2 ¿ 107 cells/rat with a purity of 90 %, as evaluated by a phase-contrast microscope (Fig. 2A). The freshly isolated cells with a high nuclei/cytoplasm ratio showed a cobblestone-like appearance after they attached to the dish (Fig. 2B). The viability of the freshly isolated cells was more than 95 %, as assessed by their ability to exclude trypan blue. Transmission electron microscopy revealed that these cells had an ovoid nucleus with condensed chromatin, scant cytoplasm, and few organelles including immature

Phenotypic properties of freshly isolated cells

284

Fig. 2 The morphological characteristics of oval cells. Freshly isolated cells (A) assumed an oval-shaped appearance when they attached to culture dish (B). Note the ovoid nuclei (Nu), condensed

chromatin (Ch), a large nuclei/cytoplasm ratio, few organelles including mitochondria (Mi) and rough endoplasmic reticulum (Er) (C, D). Magnification: A, B ¿200; C, D ¿12000.

The characterization of oval cells’ differentiation

differentiated cells at day 10 (Fig. 4E). Similar results were observed when oval cells were treated with HSS and HGF. Oval cells first turned into small hepatocytes after treatment with HSS (Fig. 4F). These small hepatocytes then enlarged their cytoplasm in the presence of HGF (Fig. 4G). Interestingly, when cultured with HGF, they fused to form a cord-like structure by day 8 (Fig. 4H). In the medium with the addition of HGF alone, oval cells differentiated into bile epithelial cells (Fig. 4I), which were identified by the RT-PCR analysis of the expression for only CK-19 mRNA but not albumin, TAT or a-1-AT mRNA (Fig. 7, lane 4).

Oval cells assumed a cobblestone appearance when attaching to the culture plates (Fig. 4A). By day 2, in the presence of EGF, oval cells changed into less-differentiated cells with a high proliferation ability, confirmed by a remarkable increase in the number of round cells with round nuclei (Fig. 4B). We named these less-differentiated cells ‘‘small hepatocytes’’ because they possessed hepatocyte traits, which are described later. It is worth noting that these small hepatocytes became enlarged and showed a polygonal appearance with the addition of HGF at day 7 (Fig. 4C). The alteration of cellular morphology might indicate the reconstruction of small hepatocytes. Small hepatocytes possessed an ability to congregate to form clusters of hundreds of cells (Fig. 4D). With time in culture, some of them integrated into a hepatic plate-like structure that consisted of fully

Morphological characteristics of small hepatocytes Morphological characteristics were an important means of distinguishing small hepatocytes from oval cells.

285

Transmission electron microscopy revealed that small round-shaped hepatocytes had round nuclei, abundant mitochondria and rough endoplasmic reticulum (Fig. 5A). The median diameter of small hepatocytes ranged from 10 to 15 mm, approximately half that of fully differentiated hepatocytes. Scanning electron microscopy confirmed the results of the light microscope, showing that these small hepatocytes were round in shape. In addition, there were many protuberances on the surface of these cells (Fig. 5B,C). Of great interest was the finding that small hepatocytes could form a three-dimensionallike cluster consisting of a dozen cells (Fig. 5D).

The phenotypic properties of small hepatocytes We also distinguished small hepatocytes from oval cells on the basis of differences in their gene expression patterns. RT-PCR analysis showed that freshly isolated oval cells expressed mRNA of both albumin and CK-19. By contrast, small hepatocytes did not express mRNA for CK-19, a marker for bile epithelial cells (Fig. 7, lane 2). However, a high level of mRNA expression for albumin, a marker for hepatocytes, was observed in small hepatocytes (Fig. 7, lane 2). These data indicate that small hepatocytes possess certain characteristics of hepatocytes.

Immunofluorescence analysis and TEM examination of fully differentiated cells Examination of the phenotypic characteristics of fully differentiated cells derived from colonies using immunofluorescence analysis revealed that all the cells were positive for albumin (Fig. 4J) and CK-18 (a late marker for hepatocytes) (Fig. 4K). In contrast, these cells did not express CK-19 or CK-7 (data not shown). The expression of albumin and CK-18, along with the lack of expressions of CK-19 and CK-7, suggests that oval cells differentiated along the hepatocytic, rather than bile epithelial cell lineage. In addition, our ultrastructural observations provided more convincing evidence that the fully differentiated cells were, in fact, mature hepatocytes. Fully differentiated cells contained well-developed organelles such as mitochondria, rough and smooth endoplasmic reticulum, lysosome and Golgi apparatus (Fig. 6A). Notably, we found that many cells had two nuclei (Fig. 6B), a characteristic feature of mature hepatocytes.

Expression of mRNA for CK-19, albumin, TAT and a-1-AT To further evaluate the phenotypic properties of fully differentiated cells, we also examined the expressions of

Fig. 3 Phenotypic characteristics of oval cells as indicated by flow cytometric analysis. A total of 1¿105 freshly isolated cells were incubated with primary antibodies including CD34 (B), albumin (C), OV-6 (D), CK-19 (E) and CK-7 (F) respectively, followed by anti-rabbit or anti-mouse PE-labeled IgG as a second antibody. The majority of cells were positive for OV-6, albumin, CK-19 and CK-7. We also observed few cells expressing anti-CD34. Replacement of the primary antibodies with PE-labeled IgG served as a negative control (A).

differentiation markers including CK-19, albumin, AAT and a-1-AT. a-1-AT is a glycoprotein synthesized chiefly in hepatocytes, and TAT is a marker of terminally differentiated hepatocytes. RT-PCR analysis revealed that fully differentiated cells expressed mRNA for albumin, a-1-AT and TAT at high levels (Fig. 7, lane 3). However, we did not detect CK-19 mRNA in fully differentiated cells. These results indicate that oval cells are specified toward hepatocytes and that fully differentiated cells possess the genetic characteristics of mature hepatocytes.

Functional hepatocyte activity in differentiated cells We measured albumin and urea production at various time points to determine whether cells with the morphological and phenotypic characteristics of hepatocytes

286

287 Fig. 4 The morphological characterization of oval cell differentiation into hepatocytes and bile epithelial-like cells. Freshly isolated cells with a high nuclei/cytoplasm ratio were ovoid in shape after attaching to the plate (A). Oval cells grew into relatively less differentiated cells that proliferated to generate more round cells with round nuclei in the presence of EGF by day 2 (B). Round cells enlarged their cytoplasm (denoted by arrows) with the addition of HGF (C). Some of these round cells aggregated to form clusters consisting of hundreds of fully differentiated cells (D). Differentiated cells in clusters fused to form a hepatic plate-like structure (indicated by arrow) by day 9 (E). The same was true when oval cells were treated with HSS and HGF. Oval cells first changed into less differentiated cells with round nuclei in the presence of HSS (F). These round cells also became enlarged (denoted by arrows) in their shape (G). A hepatic cord-like structure (indicated by arrow) developed after the addition of HSS and HGF (H). Bile epitheliallike cells were observed in the medium with HGF alone by day 8 (I). Immunofluorescence analysis of the fully differentiated cells derived from colonies by day 11. The cells were incubated with anti-albumin, CK-18, CK-19 and CK-7 respectively, followed by anti-mouse or anti-rabbit PE-labeled IgG to detect the primary antibodies. Fully differentiated cells strongly expressed albumin (J) and CK-18 (K), which was confirmed by the observation that the cytoplasm of these cells was stained red. No fluorescence was detected in cells incubated with CK-19 or CK-7 (data not shown). Magnification: A ¿100; B ¿100; C ¿200; D ¿100; E ¿400; F ¿100; G ¿100; H ¿200; I ¿200; J, K ¿200.

also had the functional attributes of hepatocytes. Albumin production is a specific test for the presence and metabolic activity of hepatocytes. In the presence of HGF alone, albumin concentration remained at a low level throughout differentiation. The slight increase in albumin production during the first 4 days of induction could be due to the mitogenic effects of HGF on oval cells. In contrast, the addition of both HSS and HGF was followed by a massive increase in albumin production that peaked on day 7. Combining EGF with HGF further enhanced albumin production, the maximum of which was 520 ng/106 cells within 24 h (Fig. 8A). Urea production is another characteristic of hepatocyte activity. Urea synthesis by differentiated cells was measured throughout their differentiation. Oval cells treated with HGF alone did not produce urea. Following treatment with EGF and HGF, urea production by differentiated cells increased in a time-dependent manner during the first 7 days of induction, and then remained at a stable level (Fig. 8B). For oval cells treated with HSS and HGF, urea production increased, reaching peak level on day 7, similar to our observations on oval cells after treatment with EGF and HGF.

Discussion The results show that putative hepatic stem cells (oval cells) can differentiate in vitro into cells that are morphologically, phenotypically and functionally hepatocytes via a 2-step induction protocol. Clearly, since the use of HGF alone induced the production of bile epithelial

Fig. 5 The morphological characteristics of small hepatocytes. Note the round nuclei, abundant mitochondria (Mi) and rough endoplasmic reticulum (Er) (A). SEM examination revealed that small hepatocytes were round (B, C), and that they aggregated to form a tri-dimensional cluster containing a dozen cells (D). Magnification: A ¿10000; B ¿9000; C, D ¿8500.

cells, the success of this procedure depends on the use of EGF or HSS. EGF and HSS probably initiate a stable hepatic phenotype and play a key role in oval cells’ specification toward hepatocytes, as indicated by the fact that both these substances induced oval cells to change into small hepatocytes with hepatic characteristics. Another distinguishing feature of small hepatocytes is their extensive proliferation potential, which may reflect the mitogenic effects of EGF and HSS on them. EGF, a mitogen for oval cell lines, has been shown to induce hepatocytes to dedifferentiate and proliferate (Block et al., 1996; Isfort et al., 1997), whereas HSS can strongly stimulate hepatic DNA, RNA and protein synthesis (LaBrecque, 1991; Zhou et al., 1992). It is worth noting that the proliferation of small hepatocytes occurred before hepatocyte differentiation. This cellular proliferation caused by EGF and HSS may play an important role in hepatocyte differentiation. HGF may be associated with the ontogenesis of the liver, which can be confirmed by the observation that transgenic mice lacking gene expression for HGF have abnormal liver development (Schmidt et al., 1995). Although HGF has been demonstrated to expand bile epithelial cells derived from human livers (Strain et al., 1995), it can also induce hepatocyte differentiation of stem cells when combined with other growth factors. In studies of hepatic stem cells derived from mid-phase fetal liver tissues, HGF together with Flt-3 ligand and stem cell factor increased the frequencies of hepatocytes in explant culture of fetal livers (Monga et al., 2001). Our results showed that the addition of EGF and HGF

288

Fig. 6 The morphological characteristics of fully differentiated cells. TEM examination showed a dense and large cytoplasm with well-developed organells including mitochondria (Mi), smooth and

rough endoplasmic reticulum (Er) Lysosome (Ly) and Golgi apparatus (Go) (A). Many binuclear cells that are commonly found in vivo were also observed (B). Magnification: A, B ¿7000.

led to the formation of a hepatic plate-like structure by day 9, and that the addition of HSS and HGF resulted in hepatic cord-like formation. In contrast, oval cells in medium with HGF alone differentiated into bile epithelial cells. Based on these results, we therefore conclude that the advantage of a 2-step induction procedure utilizing HGF with either EGF or HSS outweighs that of a 1-step induction method using HGF alone.

We previously proposed that the hepatocyte differentiation of oval cells was comprised of several intermediate stages. These results confirm this hypothesis by showing that oval cells first changed into small hepatocytes, and then differentiated into mature hepatocytes. Small hepatocytes have been characterized as proliferating cells with hepatic characteristics. Morphologically and phenotypically different from oval cells, small hepatocytes are round, have round nuclei, and fail to express CK-19. Meanwhile, small hepatocytes expressed albumin mRNA but not mRNA for a-1-AT and TAT. These results suggest that small hepatocytes may represent a novel cell population. It is noteworthy that no specific marker is available for small hepatocytes. The next goal, therefore, is to develop new strategies to produce molecular probes and specific antibodies for small hepatocytes. Similarly, we hypothesized that the generation of the hepatic plate-like structure might consist of three phases. First, oval cells change into small hepatocytes. Secondly, small hepatocytes proliferate and congregate to form colonies. Finally, small hepatocytes differentiate into mature hepatocytes that fuse and integrate to form the hepatic plate-like structure. The establishment of an in vitro hepatic plate-like assay as described here will offer a possibility for us to develop techniques for liver tissue or organ in vitro culture. This 2-step induction strategy may provide a novel source of hepatocytes for cell transplantation, bioartificial liver development, and gene therapy for the treatment of liver-related diseases.

Fig. 7 RT-PCR analysis of albumin, CK-19, TAT, and a-1-AT mRNA expression. Lane 1 – Expression in freshly isolated cells; Lane 2 – less differentiated cells derived from cells treated with EGF at day 6; Lane 3 – fully differentiated cells obtained from oval cells treated with EGF and HGF by the end of induction; Lane 4 – bile epithelial cells derived from cells treated with HGF alone by day 11. Freshly isolated cells strongly expressed both albumin and CK-19 mRNA, thereby demonstrating their dual differentiation potential. Less differentiated cells expressed albumin mRNA but lacked expression of mRNA for CK-19, indicating that oval cells differentiated along the hepatocyte lineage. The genes for albumin, TAT, and a-1-AT were highly expressed in fully differentiated cells at 11 days after initiation of culture. In contrast, bile epithelial-like cells only expressed CK-19 mRNA at a high level.

Acknowledgements We would like to thank Prof. Stewart Sell, Department of Pathology and Lab Medicine, Albany Medical College, for his kind provision of OV-6 antibody and helpful suggestions on preparing this manuscript. The authors are supported by

289

Fig. 8 Albumin and urea production by differentiated cells at various time points. Albumin production (A) and urea production (B) were expressed per 106 cells and per hour. Tests were replicated four times and repeated three times for each assay. Combination of EGF with HGF led to massive albumin production over a tenday induction, which was similar to that by oval cells treated with HSS and HGF. Albumin production induced by the addition of HGF alone decreased to an undetectable level by day 10. Urea production by oval cells treated with EGF and HGF increased to a peak on day 7, which was similar to that seen in oval cells treated with HSS and HGF. Urea production by oval cells and bile epithelial-like cells was not detected.

grants from the High-tech Project of the Chinese Ministry of Science and Technology (No. 2001AA216051) and the Natural Science Foundation of Beijing (No. 7022023).

References Avital, I., Inderbitzin, D., Aoki, T., Tyan, D.B., Cohen, A.H., Ferraresso, C., Rozga, J., Arnaout, W.S. and Demetriou, A.A. (2001) Isolation, characterization, and transplantation of bone marrow-derived hepatocyte stem cells. Biochem Biophys Res Commun 288:156–164. Block, G.D., Locker, J., Bowen, W.C., Petersen, B.E., Katyal, S., Strom, S.C., Riley, T., Howard, T.A. and Michalopoulos, G.K.

(1996) Population expansion, clonal growth and specific differentiation patterns in primary cultures of hepatocytes induced by HGF/SF, EGF and TGF alpha in a chemically defined (HGM) medium. J Cell Biol 132:1133–1149. Crosby, H.A., Kelly, D.A. and Strain, A.J. (2001) Human hepatic stem like cells isolated using c-kit or CD34 can differentiate into biliary epithelium. Gastroenterology 120:534–544. Dumble, M.L., Croager, E.J., Yeoh, G.C. and Quail, E.A. (2002) Generation and characterization of p53 null transformed hepatic progenitor cells: oval cells give rise to hepatocellular carcinoma. Carcinogenesis 23:435–445. Faris, R.A., Konkin, T. and Halpert, G. (2001) Liver stem cells: a potential source of hepatocytes for the treatment of human liver disease. Artif Organs 25:513–521. Faris, R.A., Monfils, B.A., Dunsford, H.A. and Hixson, D.C. (1991) Antigenic relationship between oval cells and a subpopulation of hepatic foci, nodules, and carcinomas induced by the ‘‘resistant hepatocytes’’ model system. Cancer Res 51:1308–1317. Fleig, W.E. and Hoss, G. (1989) Partial purification of rat hepatic stimulator substance and characterization of its action on hepatoma cells and normal hepatocytes. Hepatology 9:240–248. Golding, M., Sarraf, C.E., Lalani, E.N., Anilkumar, T.V., Edwards, R.J., Nagy, P., Thorgeirsson, S.S. and Alison, M.R. (1995) Oval cell differentiation into hepatocytes in the acetylaminofluorenetreated regenerating rat liver. Hepatology 22:1243–1253. He, Z.P., Tang, Y.F., Liu, Y.B. and Feng, M.F. (2003) Advances in studies on hepatic stem cells. Prog Nat Sci 13:166–172. Hixson, D.C., Chapman, L., McBride, A., Faris, R. and Yang, L. (1997) Antigenic phenotypes common to rat oval cells, primary hepatocellular carcinomas and developing bile ducts. Carcinogenesis 18:1169–1175. Isfort, R.J., Cody, D.B., Stuard, S.B., Randall, C.J., Miller, C., Ridder, G.M., Doersen, C.J., Richards, W.G., Yoder, B.K., Wilkinson, J.E. and Woychik, R.P. (1997) The combination of epidermal growth factor and transforming growth factor-beta induces novel phenotypic changes in mouse liver stem cell lines. J Cell Sci 10:3117–3129. LaBrecque, D.R. (1991) Hepatic stimulator substance: discovery, characteristics and mechanism of action. Dig Dis Sci 36:669– 673. Lazaro, C.A., Rhim, J.A., Yamada, Y. and Fausto, N. (1998) Generation of hepatocytes from oval precursors in culture. Cancer Res 58:5514–5522. Liu, X.J., An, W., Lei, T.G., Rong, Y. and Du, G.G. (1998) Regulator effect of hepatic stimulator substance on the proliferation of human hepatoma cell. Acta Physiol Sin 50:543–550. Monga, S.P., Tang, Y., Candotti, F., Rashid, A., Wildner, O., Mishra, B., Iqbal, S. and Mishra, L. (2001) Expansion of hepatic and hematopoietic stem cells utilizing mouse embryonic liver explants. Cell Transplant 10:81–89. Pack, R., Heck, R., Dienes, H.P., Oesch, F. and Steinberg, P. (1993) Isolation, biochemical characterization, long-term culture, and phenotype modulation of oval cells form carcinogen-fed rats. Exp Cell Res 204:198–209. Petersen, B.E. (2001) Hepatic ‘‘stem cells’’ coming full circle. Blood Cell Mol Dis 27:590–600. Petersen, B.E., Bowen, W.C., Patrene, K.D., Mars, W.M., Sullivan, A.K., Murase, N., Boggs, S.S., Greenberger, J.S. and Goff, J.P. (1999) Bone marrow as a potential source of hepatic oval cells. Science 284:1168–1170. Ruck, P., Xiao, J.C. and Kaiserling, E. (1996) Small epithelial cells and histogenesis of hepato-blastoma: electron microscopic, immunoelectron microscopic, and immunohistochemical findings. Am J Pathol 148:321–329. Schmidt, C., Bladt, F., Goedecke, S., Brinkmann, V., Zschiesche, W., Sharpe, M., Gherardi, E. and Bichmeir, C. (1995) Scatter factor/hepatocyte growth factor is essential for live development. Nature 373:699–702. Sell, S. (1990) Is there a liver stem cell? Cancer Res 50:3811–3815.

290 Sell, S. and Pierce, G.B. (1994) Maturation arrest of stem cell differentiation is a common pathway for the cellular origin of teratocarcinomas and epithelial cancers. Lab Invest 70:6–21. Strain, A.J. and Neuberger, J.M. (2002) A bioartificial liver-state of the art. Science 295:1005–1009. Strain, A.J., Wallace, L., Joplin, R., Daikuhara, Y., Ishii, T., Kelly, D.A. and Neuberger, J.M. (1995) Characterization of biliary epithelial cells isolated from needle biopsies of human liver in the presence of hepatocyte growth factor. Am J Pathol 146:537–545. Suzuki, A., Zheng, Y.W., Kaneko, S., Onodera, M., Fukao, K., Nakauchi, H. and Taniguchi, H. (2002) Clonal identification and characterization of self-renewing pluripotent stem cells in the developing liver. J Cell Biol 156:173–184. Tatematsu, M., Kaku, T., Medline, A. and Farber, E. (1985) Intestinal metaplasia as a common option of oval cells in relation to

cholangiofibrosis in liver of rats exposed to 2-acetylaminofluorene. Lab Invest 52:354–362. Theise, N.D., Nimmakayalu, M., Gardner, R., Illei, P.B., Morgan, G., Teperman, L., Henegariu, O. and Krause, D.S. (2000) Liver from bone marrow in humans. Hepatology 32:11–16. Thorgeirsson, S.S., Evarts, R.P., Bisgaard, H.C., Fujio, K. and Hu, Z. (1993) Hepatic stem cell compartment: activation and lineage commitment. Proc Soc Exp Bio Med 204:253–260. Yang, L., Li, S., Hatch, H., Ahrens, K., Cornel, J.G., Petersen, B.E. and Peck, A.B. (2002) In vitro trans-differentiation of adult hepatic stem cells into pancreatic endocrine hormone-producing cells. Proc Natl Acad Sci U S A 99:8078–8083. Zhou, Z., Tan, R., Su, S. and Zhou, C. (1992) Effects of human HSS on hepatocyte and hepatoma cell proliferation and D-GAL induced acute liver failure. Chin Med Sci J 7:197–200.