Human umbilical cord blood (HUCB) contains stem/progenitor cells, which can differentiate into a variety of cell types. In this study, we investigated whether HUCB cells differentiate into hepatocytes in vitro and in vivo. We also examined whether CD34 could be the selection marker of stem cells for hepatocytes. HUCB cells were obtained from normal full-term deliveries, and CD34+/− cells were further separated. For in vitro study, HUCB cells were cultured for 4 wk, and expressions of liver-specific genes were examined. For the in vivo study, nonobese diabetic/severe combined immunodeficient mice were subjected to liver injury by a Fas ligand-carried adenoviral vector or only radiated. Mice were treated simultaneously with or without cell transplantation of HUCB, CD34+, or CD34− cells. After 4 wk, human-specific gene/protein expression was examined. In the in vitro study, human liver-specific genes were positive after 7 days of culture. The immunofluorescent study showed positive staining of α-fetoprotein, cytokeratin 19, and albumin in round-shaped cells. In the in vivo study, immunohistochemical analysis showed human albumin-positive, hepatocyte-specific antigen-positive cells in mouse livers of the Fas ligand/transplantation group. Fluorescence in situ hybridization analysis using the human Y chromosome also showed positive signals. However, no difference between transplanted cell types was detected. In contrast, immunopositive cells were not detected in the irradiated/transplantation group. The RT-PCR result also showed human hepatocyte-specific gene expressions only in the Fas ligand/transplantation group. HUCB cells differentiated into hepatocyte-like cells in the mouse liver, and liver injury was essential during this process. The differences between CD34+ and CD34− cells were not observed in human hepatocyte-specific expression.
- cell transplantation
- Fas ligand
- human umbilical cord blood
- liver injury
stem cells are defined as undifferentiated long-lived cells that are capable of multiple rounds of division. Human umbilical cord blood (HUCB) is reported to contain stem/progenitor cells in greater numbers than the bone marrow or adult peripheral blood (4, 14, 39). Recently, successful HUCB cell transplantation for various blood diseases has resulted in a lower incidence of graft versus host disease than conventional therapy (32, 44). In these clinical applications, hematopoietic stem cells can differentiate into mature blood cells. CD34+ cells, which carry the well-known hematopoietic stem cell marker, have the capacity to differentiate into all blood cell lineages (7, 14, 35). HUCB cells have also been reported to differentiate into a variety of cell types such as bone (9, 13), cartilage (9), muscle (9) and nerve cells (5, 13) and hepatocytes (15, 17, 27, 40).
Recent studies have shown that male recipients of female liver allografts have a small number of male hepatocytes within the graft in long-term observations (20, 33). Other recent studies also have shown that bone marrow cells could differentiate into hepatocytes in the damaged liver (1, 3, 6, 21, 30). These results suggested that the bone marrow contains a certain number of hepatic progenitor cells (3, 30). Oval cells in the canal of Herring may proliferate after hepatic injury, when the growth of mature hepatocytes is suppressed (34, 37). In addition, oval cells can differentiate into cholangiocytes in response to various types of stress or liver injury. From all these studies, it remains unclear which cell type actually differentiates into the hepatocyte during liver injury. Although hepatic progenitor cells exist in the bone marrow and HUCB, it is still unclear whether these cells differentiate into hepatocytes or fuse with host hepatocytes (2, 15, 25, 36, 41, 42). Recent studies have shown that oval cells express CD34, Thy-1, and c-Kit, all of which are also found in hematopoietic stem cells (30). For this reason, we postulated that CD34+ cells in HUCB could differentiate into hepatocytes in the hepatocyte-specific niche.
From clinical application to cell therapy, there are many problems with bone marrow cell transplantation, such as the limitation of cell number, the donor’s physical responsibility, and the prevention of rejection. HUCB, however, can be collected without any harm to the newborn infant and has been found to be multipotent and cause fewer episodes of rejection because of the immaturity. Therefore, the applications of HUCB cells should be further expanded and may serve as an excellent alternative to bone marrow-derived stem cells for cell transplantation in various diseases. In this study, we investigated whether HUCB cells differentiate into hepatocytes in vivo and whether CD34 can be the selection marker of stem cells for the hepatocyte.
MATERIALS AND METHODS
Samples of male HUCB were obtained during normal full-term deliveries with informed consent, which was approved by the institutional Ethics Review Committee. HUCB was centrifuged at 1,500 rpm for 35 min at 25°C on a layer of Ficoll-Paque (Amersham-Pharmacia Biotech; Little Chalfont, UK). The mononuclear cell layer was suctioned, followed by a reaction with fluorescence-activated cell sorting (FACS) lysing solution (Becton Dickinson; San Jose, CA). After being washed with PBS, cells were first dissolved with PBS, FcR, and CD34 MicroBeads (Miltenvi Biotec; Auburn, CA), according to the manufacturer’s instructions. The cells were incubated for 20 min at 4°C and washed with PBS. The isolation of CD34+ cells was performed using Auto MACS (Miltenvi Biotech). Isolated CD34+ cells were stained with both CD34-PE antibody and lineage cocktail-FITC antibody (Becton Dickinson) and analyzed with FACS (Becton Dickinson). All the following studies were performed when the purity of CD34+ and lineage cocktail− cells was >80%.
HUCB cells and CD34+ or CD34− cells were cultured in DMEM supplemented with 15% FBS, 100 U/ml penicillin, 100 U/ml streptomycin, and growth factors including hepatocyte growth factor (20 ng/ml) and fibroblast growth factor (10 ng/ml). HUCB cells were plated on collagen-coated dishes and maintained at 37°C in a 5% CO2 atmosphere. Medium change was carried out twice weekly thereafter.
Eight- to ten-week-old nonobese diabetic (NOD)/severe combined immunodeficient (SCID) female mice purchased from Clea Japan (Tokyo, Japan) were used. All mice were maintained under specific pathogen-free conditions in the animal center with free access to water and food. Animal care was in accordance with the guidelines of the National Research Institute for Child Health and Development. The mice were first divided into two groups (liver injury group and irradiation group). The liver injury group was injected with an adenoviral vector encoding Fas ligand (FasL) (AxCALNFasL) and Cre (AxCALNCre) via the tail vein (10, 28). The mice were simultaneously treated with cell transplantation (FasL/TP), and mice without transplantation were used as a control (FasL). The irradiated group was irradiated 250 cGy before the cell transplantation (Ra/TP). Cell transplantation was performed with 5×104 mononuclear cells, either CD34+ or CD34− cells. The recipient mice were intraperitoneally injected with 20 μl of anti-asialo GM1 antibody (Wako; Osaka, Japan) to delete the natural killer cells before the cell transplantation. Subsequent antibody treatments were performed on days 11 and 22. Four weeks after the cell transplantation, animals were killed, and peripheral blood, liver tissues, the spleen, the tibia, and the femur were collected.
Measurement of Alanine Aminotransferase
Alanine aminotransferase (ALT) was assayed by test paper using the Reflotron checker (Reflotron, Roche; Basel, Switzerland). A normal value for ALT was <40 U/l.
Mouse peripheral blood samples were centrifuged at 3,000 rpm for 20 min at 25°C on a layer of Ficoll-Paque. Collected mononuclear cells were incubated with 20 μl human CD45-PE antibody (Becton Dickinson) and processed for FACS analysis.
Bone marrow cells were collected by tibial and femural aspiration with PBS using a syringe. Cell suspensions were then filtered through a cell strainer and processed for FACS analysis. The spleen was rubbed with slides to liberate cells. Cell suspensions were then filtered through a cell strainer and analyzed by FACS.
In addition, purified HUCB cells were incubated with human anti-Fas antibody (Becton Dickinson) to examine Fas expression on HUCB cells and processed for FACS analysis. Human peripheral blood from healthy volunteers was used as a positive control.
Total RNA was extracted from frozen mouse liver tissue and cultured cells using ISOGEN (Nippon Gene; Tokyo, Japan), according to the manufacturer’s instructions. Normal human liver tissue obtained from normal areas during the surgical resection of hepatocellular carcinoma was used as a control. RNA was purified to remove contaminating DNA according to the DNA-free protocol (Ambion; Austin, TX) and quantified using a spectrophotometer. Equal amounts of RNA were subjected to cDNA synthesis using oligo dT (Invitrogen; Carlsbad, CA). The obtained cDNA was amplified with Zymoreactor (ATTO; Tokyo, Japan) according to the conditions described in Table 1. The sequences and product sizes for human albumin, α-fetoprotein (α-FP), glutamine synthetase (GS), transferrin, cytokeratin 19 (CK-19), β-actin, and mouse GAPDH are shown in Table 1. The amplified products were electrophoresed on 1.5% agarose gels and stained with ethidium bromide.
Immunofluorescent Staining Analysis of Cultured Cells
Cultured HUCB cells were fixed with acetone for 10 min. After being washed with PBS, cells were incubated with primary antibodies at room temperature for 1 h. For primary antibodies, anti-human albumin-FITC (DAKO; Kyoto, Japan), CK-19 (DAKO), and α-FP (DAKO) antibodies were used. The slides were then washed with PBS and incubated for 30 min at room temperature with anti-mouse IgG TRITC conjugate (Sigma) or phycoprobe R-PE anti-goat IgG (Biomeda). After being washed, slides were immediately applied with 4,6-diamidino-2-phenylindole (DAPI, Vector Laboratories) and then analyzed by fluorescent microscope (model AX80T, Olympus; Tokyo, Japan).
To evaluate the number of immunostained cells, photographs of 10 random fields (×100) per slide were analyzed by NIH Image software (National Institutes of Health; Bethesda, MD).
Horseradish peroxidase-conjugated sheep anti-human albumin antibody (Serotec; Oxford, UK) was used to avoid a nonspecific reaction with the mouse liver as follows: NOD/SCID mice were injected with antibody via tail vein, and peripheral blood was collected in 15 min. After centrifugation, the serum was inactivated at 56°C for 30 min.
Mononuclear cells collected from HUCB were transferred to slides by the cytospin method. The slides of 5-μm frozen liver sections were fixed with acetone followed by formol calcium for 2 min. Endogenous peroxidase was blocked for 10 min using 0.3% H2O2 in PBS. After being washed with PBS, sections were incubated with purified human albumin antibody for 3 h at room temperature. The staining was visualized with diaminobenzidine (DAB). Anti-human hepatocyte antibody (DAKO), which specifically recognizes human hepatocytes, was used as the primary antibody. Immunostaining was visualized using a biotin-conjugated secondary antibody according to the manufacturer’s instruction. To evaluate the number of immunostained cells and the immunopositive area, photographs of five random fields (×100) per slide were analyzed by NIH Image software.
Fluorescence In Situ Hybridization
Human Y chromosome DNA probe CEP Y (Vysis) was used. The slides of 5-μm frozen liver sections were fixed with methanol for 30 min. After being washed with PBS, slides were incubated with 2× SSC-0.1% Nonidet P-40 (Nacalai Tesque; Kyoto, Japan) for 30 min at 37°C and dehydrated with ethanol. The sections were denatured with 70% formamide-2× SSC at 73°C for 5 min and then hybridized with the denatured probe overnight at 37°C. After being washed, slides were immediately applied with DAPI and then analyzed by fluorescent microscope (model AX80T).
Results are expressed as means ± SD, and a statistical analysis was performed with Student’s t-test. P < 0.05 was considered as a significant difference between groups. A statistical evaluation for graft survival was performed using the Kaplan-Meier test. P values <0.05 were considered statistically significant.
In Vitro Studies
Cultured HUCB cells.
HUCB cells were differentiated into both round and spindle-shaped cells by day 7 (Fig. 1A). The cytoplasm of the round cells was enlarged, whereas the spindle-shaped cells had differentiated into fibroblast-like cells during the time course (Fig. 1B). However, when CD34+ or CD34− cells were cultured, living cells were decreased gradually during the time course. The rate of round cells in all cultured cells was about 70%, which did not change during incubation (Fig. 1C).
Immunofluorescent staining analysis.
The immunofluorescent study showed positive expression of human α-FP, albumin, and CK-19 in the round cells but not in the fibroblast-like cells (Fig. 2). α-FP-positive staining was observed with a peak on day 7 that then decreased gradually until day 28. In contrast, the rate of albumin- and CK-19-positive cells increased during the longer culture. Albumin-positive cells accounted for about 50% of all cultured cells on days 21 and 28 (Fig. 3). Cultured CD34+ and CD34− cells were also positive for human α-FP, albumin, and CK-19, although the numbers of living cells decreased during incubation (data not shown).
RT-PCR was performed to detect human-specific expression of albumin, α-FP, GS, transferrin, and CK-19 (Fig. 4). The normal human liver was used as a positive control. Freshly isolated HUCB cells did not express these genes. However, HUCB cells expressed all these genes by day 14.
In Vivo Studies
Low-level Fas expression in HUCB cells.
We first examined the expression of Fas antigen on the surface of HUCB cells. As shown in Fig. 5, very low expression of Fas antigen on HUCB cells compared with human peripheral blood cells, which strongly express Fas, was observed. These results indicated that the transplanted HUCB cells would not be damaged by the Fas-mediated liver injury model used in this study.
Fas-mediated liver injury.
To evaluate the amount of recombinant adenovirus vector AxCALNFasL administration, mice were administered 4.2 × 108, 4.8 × 108, or 5.3 × 108 adenoviral plaque-forming units (pfu), respectively. The amount of recombinant adenoviral vector AxCALNCre was 1.0 × 109 pfu. The peak ALT level after three doses was on day 3, and the values were 480.8 ± 120.8 IU/l (Fig. 6). Hepatic apoptosis and necrosis were observed in all groups as detected by hematoxylin-eosin (HE) staining, although the degree was almost the same in each group. Vascular architecture was intact in all groups. The liver damage was transient, and histology was nearly normal on day 10 in the surviving mice (Fig. 6C). Because the hepatic necrosis areas shown by the HE stain were equal, mice were administered 4.8 × 108 pfu AxCALNFasL to induce liver injury.
Mice were killed 1 mo after HUCB cell transplantation, and the population of HUCB-derived leukocytes was studied using anti-human CD45 antibody on the spleen, bone marrow, and peripheral blood in both the FasL/TP and Ra/TP groups by FACS analysis (Fig. 7). In the FasL/TP group, CD45+ leukocytes were observed only in the peripheral blood. In contrast, CD45+ cells were detected in the spleen, bone marrow, and peripheral blood in the Ra/TP group.
Mononuclear HUCB cells were transferred to the slide by cytospin, stained with human albumin antibody, and confirmed to be negative (Fig. 8A). The NOD/SCID mouse liver was also negative for human albumin (Fig. 8B). In the Ra/TP group, human albumin-positive cells were not observed, although three different sections in each mouse were examined (n = 15; data not shown). In contrast, in the FasL/TP group, human albumin-positive cells were easily detected (Fig. 8, C and D), as follows: 0.39% (0 ∼ 0.98, n = 5), 0.40% (0 ∼ 0.73, n = 4), and 0.39% (0 ∼ 0.68, n = 5) in the transplanted group with CD34+, CD34−, and mononuclear cells, respectively. The area of reactivity was as follows: 0.24% (0 ∼ 0.76, n = 5), 0.21% (0 ∼ 0.42, n = 4), and 0.11% (0 ∼ 0.24, n = 5) in the transplanted group with CD34+, CD34−, and mononuclear cells, respectively. The difference, however, was not statistically significant (Fig. 9). The cells that reacted to the anti-albumin antibody were sparsely detected. These cells were located in the parenchyma and the circumference of vessels in each group. Clusters of positive cells were detected, although very rarely (date not shown). In a long-term study, the positive cells were still observed 5 mo after cell transplantation, and the rate decreased during the time course (data not shown).
Immunostaining of human hepatocyte-specific antigen was also positive in the FasL/TP group, and reactivity against anti-human hepatocytes was similar in all groups (Fig. 8, E and F). FasL and Ra/TP groups were negative (data not shown).
FISH for human Y chromosome showed positive signals on sections of the FasL/TP group (Fig. 8, G and H). However, FasL and Ra/TP groups were negative (data not shown). The positive cells existed in the parenchyma (Fig. 8G) and in the circumference of vessels (Fig. 8H) in the FasL/TP group.
RT-PCR was performed to detect human-specific expression of albumin, α-FP, GS, transferrin, and CK-19 (Fig. 10). The normal human liver was used as a positive control. Freshly isolated HUCB cells and the mouse liver of the FasL group did not express these genes. The human liver demonstrated these expressions. In the FasL/TP group, these genes were all expressed except for CK-19, and the expression was not different in each transplantation group of CD34+, CD34−, and mononuclear cells. These results suggested that transplanted cells differentiated into hepatocyte-like cells, and these three groups showed similar results.
In our experiment, some mice of the FasL group died within 6 days, and the final survival rate was 70% (Fig. 11). However, none of the FasL/TP group mice died, and the rate was significantly higher, with a final result of 100% (P < 0.05).
In the present study, we showed cultured CD34+, CD34−, and HUCB cells could differentiate into the cells expressing human liver-specific genes and proteins by RT-PCR assay and immunofluorescent staining. In addition, we showed that transplanted CD34+, CD34−, and HUCB cells expressed human hepatocyte-specific genes and proteins by an RT-PCR assay and immunohistochemical study in the injured mouse liver, respectively. FISH analysis also confirmed the presence of human Y chromosome in the female mouse liver. HUCB cells and mouse hepatocytes did not express these human genes and proteins; therefore, the transplanted CD34+, CD34−, and HUCB cells differentiated into hepatocyte-like cells in this Fas-mediated liver injury model. In addition, the hepatocyte-like cells may be hepatic progenitor cells or mature hepatocyte-like cells, because α-FP is known to be expressed in hepatic progenitor cells, including oval cells (10, 31), and GS is a marker of mature hepatocytes (12, 26). Previous reports of HUCB cells have shown that albumin-producing cells have been detected at 21 days in a primary culture system containing growth and differentiation factors (17) and at 4 wk after transplantation in vivo (17), respectively, which is consistent with our study.
Wagers et al. (38) showed that single transplanted hematopoietic stem cells differentiated at a very low level of into hepatocytes in mice without liver dysfunction and that transplanted cells did not express albumin. The present study has also shown that transplanted HUCB cells could differentiate into hepatocyte-like cells only, but not cholangiocytes, in the Fas-mediated injured liver. Because cultured HUCB cells could differentiate into cholangiocyte-like cells, which express CK-19, this might be the reason that this model did not damage bile ducts. Liver injury and the surrounding environment may be essential for transplanted cells to be engrafted in the liver and to differentiate into hepatocytes (15, 17, 27, 38, 40). Recent studies have suggested that stromal cell-derived factor and its receptor CXCR4 might be important for stem cell recruitment to the liver (18, 29). In addition, CK-19, which is a differentiation marker of liver progenitor cells into bile duct cells, was not detectable in this study. This finding may explain why the bile ducts were not damaged in our liver injury model.
Recent studies have further shown that transplanted HUCB cells could differentiate into cells that express human albumin in carbon tetrachloride-induced (17) or 2-acetylaminofluorene-induced (15) liver injury models. Our animal model induced rapid and massive apoptosis in the liver by the FasL recombinant adenoviral expresson vector AxCALNFasL. FasL expression induces hepatocyte apoptosis both by a direct interaction with Fas in hepatocytes and by Fas-positive inflammatory cells (28). The primary cause of liver damage used in this study is the Fas/FasL pathway, and this liver injury may be similar to human viral hepatitis rather than drug-induced injury, such as that by carbon tetrachloride or 2-acetylaminofluorene. In that context, the present study demonstrates clinical applications worth considering.
We observed that the Ra/TP group expressed human CD45+ leukocytes not only in peripheral blood but also in the spleen and bone marrow, which confirmed that the transplantation was successful, because CD34+ cells differentiate into blood cells (14, 35, 44). The FasL/TP group expressed CD45+ cells only in peripheral blood. The mechanism of this phenomenon is that transplanted cells enter the bone marrow and differentiate into blood cells but do not survive long term and are excluded because the bone marrow is not damaged by radiation treatment in this liver injury model.
Although we postulated that HUCB-derived CD34 (a hematopoietic stem cell marker)-expressing cells may be an important marker for hepatocyte differentiation, a difference was not observed between the CD34+ and CD34− cells that differentiated into human hepatocyte-specific expression cells in our study. Even though the number of selected cells was not high, at 100% human albumin-positive cells did not increase in number in several experiments using 98% purified cells, which suggested that CD34+ or CD− cells exhibit no difference in terms of differentiation to hepatocytes.
Wang et al. (40) showed that both CD34+ and CD34+-CD38+-CD7− cells from HUCB and bone marrow differentiat into hepatocyte-like cells by liver injury. Normal adult bone marrow was also reported to contain a small number of hepatic progenitor cells, which can differentiate into hepatocytes when the growth of mature hepatocytes is suppressed by hepatic injury (1, 3, 6, 21). Although it is still unclear which cell types of bone marrow and HUCB cells differentiate into hepatocytes, CD34 (26, 30), CD38 (40), c-Kit (11, 30), Thy-1 (11, 30, 31), and Liv-8 (43) may be candidates. However, differentiation into hepatocytes may be multifactorial. Further study is needed to determine the cell marker for liver cell differentiation. However, Lee et al. (22) showed that human mesenchymal stem cells derived from bone marrow and cord blood were effectively differentiated into hepatocyte-like cells in vitro. Because stem cells have plasticity, cells may differentiate into hepatocytes under conditions of hepatic injury without selecting HUCB by cell markers.
Recent studies have suggested that bone marrow cells are pluripotent and are able to become mature hepatocytes. Several groups have attributed this apparent plasticity to transdifferentiation. Willenbring et al. and others (2, 25, 36, 41, 42), however, showed that bone marrow-derived hepatocytes emerged from fusion between donor bone marrow-derived cells and host hepatocytes, and they suggested that cell fusion could explain these results. They also suggested that hepatocytes derived from HUCB cell transplantation might arise from cell fusion. Newsome et al. (27), however, showed that HUCB cells could differentiate into hepatocytes in the absence of cell fusion. We have not investigated whether hepatocytes derived from HUCB cells arise from cell fusion. Further study is necessary to evaluate the role of cell fusion in our study.
In our experiment, the survival rate of FasL group mice was 70%, whereas that of FasL/TP group mice was 100% (P < 0.05). FasL expression induced hepatocyte apoptosis by a direct interaction with Fas of hepatocytes and by Fas-positive inflammatory cells, which caused the death of the mice in the FasL group within 6 days. Transplanted cells may play a role in the survival of the mice in early-phase Fas-mediated liver injury. One possibility is that transplanted cells secrete some anti-inflammatory factors that modify the immunological state so as to improve the survival rate (8, 19, 24). The other possibility is that transplanted HUCB cells may heal liver injury in our model, although only a small number of albumin-expressing cells were observed in our study. A recent study (16) in which hematopoietic stem cells were transplanted into CCl4 injured mice demonstrated that 7.6% of the total liver cells were human albumin-positive cells, and serum ALT, prothrombin time, and fibrinogen recovered as early as 2 days after transplantation; however, the precise mechanism has not been clarified. Similar to this report, it is possible that HUCB cell transplantation can suppress liver injury in our previous model. However, the reasons why the survival rate improved in Fas-mediated liver injury mice is still unclear, and further study is needed to elucidate this question.
In summary, the present study shows that CD34+, CD34−, and HUCB cells differentiate into hepatocyte-like cells in this Fas-mediated liver injury model and that the survival rate of transplanted mice was improved.
This study was supported in part by Ministry of Education, Culture, Sports, Science and Technology of Japan Grant-In-Aid 14657293 and a grant for the Organized Research Combination System.
The authors are grateful to Dr. S. Enosawa for useful suggestions and comments. They also thank M. Wakabayashi and N. Takahashi for the technical assistance.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society