Flow cytometric isolation and clonal identification of self-renewing bipotent hepatic progenitor cells in adult mouse liver

Flow Cytometric Isolation and Clonal Identification ofSelf-Renewing Bipotent Hepatic Progenitor Cells in

Adult Mouse LiverAtsushi Suzuki,1,2 Sayaka Sekiya,1 Makiko Onishi,1 Naoko Oshima,3 Hiroshi Kiyonari,3

Hiromitsu Nakauchi,4 and Hideki Taniguchi2,5

The adult liver progenitor cells appear in response to several types of pathological liverinjury, especially when hepatocyte replication is blocked. These cells are histologicallyidentified as cells that express cholangiocyte markers and proliferate in the portal area of thehepatic lobule. Although these cells play an important role in liver regeneration, the precisecharacterization that determines these cells as self-renewing bipotent primitive hepatic cellsremains to be shown. Here we attempted to isolate cells that express a cholangiocyte markerfrom the adult mouse liver and perform single cell-based analysis to examine preciselybilineage differentiation potential and self-renewing capability of these cells. Based on theresults of microarray analysis and immunohistochemistry, we used an antibody againstCD133 and isolate CD133� cells via flow cytometry. We then cultured and propagatedisolated cells in a single cell culture condition and examined their potential for proliferationand differentiation in vitro and in vivo. Isolated cells that could form large colonies (LCs) inculture gave rise to both hepatocytes and cholangiocytes as descendants, while maintainingundifferentiated cells by self-renewing cell divisions. The clonogenic progeny of an LC-forming cell is capable of reconstituting hepatic tissues in vivo by differentiating into fullyfunctional hepatocytes. Moreover, the deletion of p53 in isolated LC-forming cells resultedin the formation of tumors with some characteristics of hepatocellular carcinoma and chol-angiocarcinoma upon subcutaneous injection into immunodeficient mutant mice. Thesedata provide evidence for the stem cell-like capacity of isolated and clonally culturedCD133� LC-forming cells. Conclusion: Our method for prospectively isolating hepaticprogenitor cells from the adult mouse liver will facilitate study of their roles in liver regen-eration and carcinogenesis. (HEPATOLOGY 2008;48:1964-1978.)

Abbreviations: Alb, albumin; DAPI, 4’,6-diamidino-2-phenylindole; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; ES, embryonic stem; FAH, fumarylacetoacetatehydrolase; LC, large colony; NOD/SCID, nonobese diabetic/severe combined immunodeficient; NTBC, 2-(2-nitro-4-trifluoromethylbenzoyl)-1,3-cyclohexanedione; qPCR,quantitative polymerase chain reaction; RT-PCR, reverse-transcription polymerase chain reation; SC, small colony.

From the 1Division of Organogenesis and Regeneration, Post-Genome Science Center, Medical Institute of Bioregulation, Kyushu University, Fukuoka, Japan; the2Research Unit for Organ Regeneration and the 3Laboratory for Animal Resources and Genetic Engineering, Center for Developmental Biology, RIKEN, Hyogo, Japan; the4Laboratory of Stem Cell Therapy, Center for Experimental Medicine, Institute of Medical Science, University of Tokyo, Tokyo, Japan; and the 5Department of RegenerativeMedicine, Graduate School of Medicine, Yokohama City University, Kanagawa, Japan

Received March 31, 2008; accepted July 28, 2008.Supported in part by the Program for Improvement of Research Environment for Young Researchers from Special Coordination Funds for Promoting Science and

Technology commissioned by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan; Grant-in-Aids for Scientific Research from the MEXTof Japan; and a grant from the Leading Project in Japan.

*Address reprint requests to: Atsushi Suzuki, Division of Organogenesis and Regeneration, Post-Genome Science Center, Medical Institute of Bioregulation, KyushuUniversity, 3-1-1 Maidashi, Higashi-ku, Fukuoka, Fukuoka 812-8582, Japan. E-mail: [email protected]; fax: (81)-92-642-6793; or Hideki Taniguchi,Department of Regenerative Medicine, Graduate School of Medicine, Yokohama City University, 3-9 Fuku-ura, Kanazawa-ku, Yokohama, Kanagawa 236-0004, Japan.E-mail: [email protected]; fax: (81)-45-787-8963.

Copyright © 2008 by the American Association for the Study of Liver Diseases.Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/hep.22558Potential conflict of interest: Nothing to report.Additional Supporting Information may be found in the online version of this article.

1964

The liver can regenerate in two distinct ways, de-pending on the cellular compartments undergo-ing proliferation. After partial hepatectomy or

chemical injury, cells in the remaining liver tissue, espe-cially hepatocytes, proliferate rapidly to restore lost cellswithout any contribution from hepatic stem/progenitorcells.1 When hepatocyte proliferation is impaired in somechronic injury responses, however, a regenerative signalinduces emergence of small epithelial cells known as ovalcells in the portal area of the hepatic lobule. Oval cells areconsidered to be transit-amplifying hepatic progenitorcells, due to their enormous growth capacity for repopu-lation of damaged livers and their potential to differenti-ate into hepatocytes and cholangiocytes.2 This oval cell-mediated style of liver regeneration is likely related to anas-yet-undefined stem cell system operating in both liverregeneration and carcinogenesis.2-3

Extensive studies of oval cells have been performed in acommonly used rat model in which rats are treated with2-acetylaminofluorene followed by a growth stimulusarising from partial hepatectomy or chemical injury.4-7

The cells considered to be oval cells are also found in thelivers of mice fed 3,5-diethoxycarbonyl-1,4-dihydrocolli-dine (DDC).8-9 These oval cells in rat and mouse liver aremarked by the expression of cholangiocyte markers andcontained in cells that form primitive ductules withpoorly defined lumens in the portal areas.4-9 A number ofhistological studies demonstrated that oval cells in rat livercould migrate toward the central veins from the portalareas and gradually differentiate into hepatocytes to re-place the damaged liver tissues.2,10-12 The lack of amethod for tracing oval cell fate in vitro and in vivo, how-ever, resulted in insufficient evidence for the hepatocytedifferentiation from oval cells.

Fresh oval cells have been isolated using centrifugalelutriation and histochemical analyses for various featuresof liver parenchymal and nonparenchymal cells.13 Bycombining a gradient centrifugation method with cellsorting by panning or flow cytometry, further enrichmentof oval cells has been performed to examine their growthand differentiation potential both in vitro and in vivo.9,14

The cell populations isolated by these protocols containeda relatively large number of oval cells. The subsequentanalyses of isolated cells, however, have been conductedonly with other lineage cells that were still contaminatedwith oval cells. Therefore, the precise characterizationthat determines oval cells as self-renewing bipotent prim-itive hepatic cells could be provided only after tracing cellfate in single cell-based analyses of isolated cells.

Others and we have reported strategies for isolation ofbipotent hepatic progenitor cells known as hepatoblastsfrom the developing mouse liver.15-17 Prospective identi-

fication and isolation of a highly enriched progenitor cellpopulation, followed by culture of one cell in each well of96-well plates, allows precise analysis of the nature of eachcell; this approach ultimately led to experimental proof ofthe existence of self-renewing bipotent hepatic progenitorcells in the developing mouse liver.16 In the present study,as in the case of hepatic progenitor cells in liver develop-ment, we attempted to isolate hepatic progenitor cellsfrom the adult mouse liver by combining flow cytometryand fluorescence-conjugated antibodies against cell sur-face antigens expressing on cells that formed primitiveductules in DDC-treated mouse liver. Screening of genesup-regulated in DDC-treated livers revealed that CD133(also known as prominin-1), a five-transmembrane cellsurface glycoprotein, is one of the markers for this celltype. The detailed investigation of CD133� cells isolatedfrom DDC-treated livers in in vitro and in vivo clonalanalyses allowed a precise characterization of isolated cellsand provided evidence for the stem cell-like capacity ofLC-forming cells within a portion of the CD133� cellpopulation.

Materials and Methods

Flow Cytometry. Single cell suspensions of liver cellswere prepared from C57BL/6 wild-type or p53�/�

mice,18 treated with or without a DDC (Sigma-Aldrich,St Louis, MO) diet (0.1%) for 2 weeks, by the followingdual-protease digestion protocol. After a standard two-step perfusion with liver perfusion medium (Invitrogen,Carlsbad, CA) and liver digest medium (Invitrogen), theundigested tissue was further digested with dispase (1,000protease units/mL; Godoshusei, Japan) for 30 to 60 min-utes at 37°C with shaking. Cell viability after treatmentexceeded 90% as assessed by Trypan blue dye exclusion.Cells were washed with staining medium (phosphate-buffered saline containing 3% fetal bovine serum), andthen incubated with fluorescence-conjugated antibodiesas described.16 We used phycoerythrin-Cy7–conjugatedanti-CD45, Ter119 monoclonal antibodies (Pharmin-gen, San Jose, CA) and fluorescein isothiocyanate–conju-gated anti-CD133 monoclonal antibodies (eBioscience,San Diego, CA). The fluorescence-labeled cells were ana-lyzed and separated with FACS Aria (BD Biosciences, SanJose, CA) after excluding residual erythrocytes, debris,doublets, and dead cells by forward scatter, side scatter,and propidium iodide gating.

In Vitro Colony Assay. For low-density culture anal-ysis, sorted cells were plated on type IV collagen-coated6-well plates (BD Biosciences) at a density of 1,000 cells/well and cultured in our fresh standard medium.15 Forsingle cell culture analysis, cells identified on clone-sort-

HEPATOLOGY, Vol. 48, No. 6, 2008 SUZUKI ET AL. 1965

Fig. 1.

1966 SUZUKI ET AL. HEPATOLOGY, December 2008

ing via FACS Aria were cultured in individual wells oftype IV collagen-coated 96-well plates (BD Biosciences)in standard medium 50% supplemented with mediumconditioned by 7-day culture of embryonic day 13.5 livercells. Both culture media included human recombinanthepatocyte growth factor (50 ng/mL) (Sigma-Aldrich)and epidermal growth factor (20 ng/mL) (Sigma-Al-drich). The number of LCs and small colonies (SCs) wasdetermined after 8 days of culture.

Gene Expression Analysis. Reverse-transcriptionpolymerase chain reaction (RT-PCR) and quantitativePCR (qPCR) were conducted as described.19-20 The in-formation of PCR primers and probes were presented inour previous papers,15-16,19 except for RT-PCR primersfor CK7 (5�-ATC CGC GAG ATC ACC ATC AAT-3�and 5�-ATG TGT CTG AGA TCT GCG ACT-3�) andqPCR primers/probes for CD133 (TaqMan Gene Ex-pression Assay ID: Mm00477115_m1) (Applied Biosys-tems, Japan) and CK7 (TaqMan Gene Expression AssayID: Mm00466676_m1) (Applied Biosystems). ForqPCR analysis of CD34, c-kit, and Thy-1 expression, weused primers for CD34 (ID: MA040159), c-kit (ID:MA072931), and Thy-1 (ID: MA077324) from PerfectReal Time Support System (Takara, Japan) andSYBRPremixExTaqII (Takara) according to the manu-facturer’s instructions.

Immunostaining. Tissue sections and cultured cellswere fixed and incubated with primary antibodies againstA6, CD133 (eBioscience), keratin (Dako, Carpinteria,CA), albumin (Alb) (tissue sections, Bethyl, Montgom-ery, TX; cultured cells, Biogenesis, Poole, UK), CK7(Chemicon, Temecula, CA), CK19,15 and fumarylaceto-acetate hydrolase (FAH). The antibodies against A6 andFAH were generous gifts from V. Factor and R. M. Tan-guay, respectively. After washing, the sections and/or cellswere incubated with horseradish peroxidase–conjugatedsecondary antibodies specific to the appropriate species

(1:500; Dako) for immunohistochemistry or Alexa 488-and/or Alexa 555-conjugated secondary antibodies(1:200; Molecular Probes, Eugene, OR) with 4’,6-dia-midino-2-phenylindole (DAPI) for immunofluorescencestaining.

Microarray Analysis. Total RNA was prepared fromno-DDC and 2 weeks DDC-treated livers using RNeasyMini Kit (Qiagen, Japan) according to the manufacturer’sinstructions. Multiple gene expression in no-DDC andDDC-treated livers was analyzed using One-Cycle TargetLabeling and Control Reagents (Affymetrix, Japan) andGeneChip Mouse Genome 430 2.0 Arrays (Affymetrix)according to GeneChip Expression Analysis TechnicalManual (Affymetrix).

Cell Transplantation. Clonogenic progeny propa-gating in culture from a single LC-forming CD133�

CD45� Ter119� cell (CD133� large colony-forming cell[LCC]) isolated from DDC-treated liver weretrypsinized, washed, and resuspended in 100 �L of stan-dard medium and then injected intrasplenically into theliver of FAH-deficient (FAH�/�) recipient mice (4 � 106

cells/mouse). In 12 clones that we established from eachCD133� LCC, we used representative 3 clones as donorcells for transplantation (recipient mice: n � 6 for eachclone). The FAH-deficient mice are normally maintainedon drinking water containing 7.5 mg/L 2-(2-nitro-4-trif-luoromethylbenzoyl)-1,3-cyclohexanedione (NTBC) (akind gift from K. Z. Travis and J. Doe),21 but this treat-ment was stopped just after cell transplantation. For anal-ysis of tumor formation, clonogenic progeny propagatingin culture from a single CD133� LCC isolated fromDDC-treated wild-type and p53�/� livers weretrypsinized, washed, and resuspended in 150 �L of stan-dard medium with 150 �L matrigel (BD Biosciences) andthen injected subcutaneously into nonobese diabetic/se-vere combined immunodeficient (NOD/SCID) mice(3.5 � 107 cells/mouse; n � 5).

Generation of FAH Mutant Mouse. We performedtargeted disruption of mouse FAH by homologous re-combination in TT2 embryonic stem (ES) cells as de-scribed.22-25 PCR was conducted to amplify the 5�-arm(8.2kb) and the 3�-arm (5.1kB) sequences of the targetingvector using primers for the 5�-arm (5�-GTT TGA GTCGAC CCA ACA AGG ATT ACA TGA GAC CGC C-3�and 5�-TAA GGC GGC CGC GGA TGC TCT TGCCTC CTT CCA TG-3�) and the 3�-arm (5�-TGC CTCGTC GAC GCG GCC GCG CTG TGA TTG CATGTG TGA CCT TCC-3� and 5�-CAC CCT CGA GTTAGA CCT GCA GAT GGT AGC CGC C-3�). Germ-line chimeras were produced via injection of targeted EScells into CD-1 host blastocysts.23 Genotyping via PCRwas conducted using primers (FAHwt-F, 5�-AGG CCT

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™Fig. 1. Specific expression of CD133 in cholangiocytes and DDC-

induced primitive ductal cells. (A) Immunohistochemical staining of A6 inno-DDC (upper panel) and DDC-treated (lower panel) livers. DDC treat-ment (2 weeks) induced proliferation of cells that formed primitiveductules in the portal areas. Both cholangiocytes and DDC-inducedprimitive ductal cells expressed A6. Scale bars � 500 �m. (B) Geneexpression analysis via qPCR for no-DDC and DDC-treated (1 to 4 weeks)livers. All data were normalized to the values of a no-DDC liver, and folddifferences are shown. Data are expressed as the mean � standarddeviation (n � 3). (C) Western blotting of protein samples obtained fromno-DDC and DDC-treated (1 to 4 weeks) livers. Representative data fromthree independent experiments are shown. (D,E) Co-immunofluorescencestaining of CD133 and A6 (D) or keratin (E) for no-DDC and DDC-treated(2 weeks) livers. DNA is stained with DAPI. Scale bars � 100 �m. (F)Co-immunofluorescence staining of CD133 and Keratin was also ob-served by using a confocal microscopy. Scale bars � 25 �m.


AAC CTC TTG CTT CAT TCA-3�; FAHmut-F, 5�-CCA GCT CAT TCC TCC CAC TC-3�; FAHwt/mut-R, 5�-ATC GGG GTT CCA GAT ACC AC-3�) forwild-type (FAHwt-F and FAHwt/mut-R; 809bp) and formutant allele (FAHmut-F and FAHwt/mut-R; 433bp).The accession number of FAH mutants in the RIKEN isCDB0201K.

Results

Expression of CD133 in Normal Cholangiocytesand DDC-Induced Primitive Ductal Cells. Expansionof cells that formed primitive ductules containing a pop-ulation of oval cells in mouse liver occurred after 2 weeksof DDC feeding, as confirmed on immunohistochemicalanalysis of A6, an antigen for normal cholangiocytes andDDC-induced primitive ductal cells (Fig. 1A). To iden-tify cell surface markers expressed on cells forming prim-itive ductules, we compared gene expression in the normal(no-DDC) and 2-week DDC-treated livers via microar-ray analysis. This analysis revealed an up-regulation ofCK19, �-fetoprotein (AFP), and CD34 expression, butnot of c-kit or Thy-1 expression, in DDC-treated livers;each of these genes has been previously identified as anoval cell marker.9,26-30 In addition to these genes, the ex-pression of CD133 was also up-regulated more than 4.5-fold in DDC-treated livers (Supplementary Table 1).Subsequent qPCR analysis confirmed the increase inCD133 expression during the entire course of DDC treat-ment, as well as increases in CK19, CK7, AFP, and CD34expression, while the expression of c-kit was not up-regu-lated and expression of Thy-1 increased after 3 weeks ofDDC feeding (Fig. 1B). The CD133 protein also accu-mulated in DDC-treated livers (Fig. 1C). Moreover,immunofluorescence analysis revealed that both cholan-giocytes in no-DDC livers and primitive ductal cells pro-liferating in DDC-treated livers were positive for CD133(Fig. 1D,E). The confocal microscopic analysis alsoshowed that CD133 was expressed on the surface of cellsthat form primitive ductules in DDC-treated mouse liver(Fig. 1F). Thus, CD133 could be a marker for both nor-mal cholangiocytes and DDC-induced primitive ductalcells.

Prospective Isolation of CD133� Cells Using FlowCytometric Cell Sorting. To isolate CD133� hepaticcells, single cells obtained from no-DDC and DDC-treated livers were fractionated via flow cytometry basedon their expression of CD133, CD45 (leukocyte com-mon antigen), and Ter119 (a molecule resembling glyco-phorin and exclusively expressing on immature erythroidcells). As expected, the percentage of CD133� cells in the

nonhematopoietic CD45� Ter119� cell population wasincreased more than 5 times in DDC-treated livers incomparison with no-DDC livers (Fig. 2A). The increaseof the percentage of CD45� and Ter119� hematopoieticlineage cells in DDC-treated liver may be related to theenhanced inflammatory response caused by DDC.Immunofluorescence analysis for CD133� CD45�

Ter119� cells immediately after isolation from no-DDCand DDC-treated livers revealed expression of CK7 andCK19, but not that of albumin, a marker of hepatocytes(Fig. 2B,C). Moreover, messenger RNA expression of he-patocyte marker genes, such as Alb, glucose-6-phosphatase,tryptophan-2, 3-dioxygenase, was not detected in CD133�

CD45� Ter119� cells isolated from DDC-treated livers(Fig. 2D). These results demonstrated little or no contam-ination of hepatocytes in isolated CD133� CD45�

Ter119� cells.Next, to characterize CD133� CD45� Ter119� cells

isolated from no-DDC and DDC-treated livers, we cul-tured these cells at low density. Plating 1,000 cells intoeach well of type IV collagen-coated 6-well plates enabledsemiclonal analysis in which cells with potential to prolif-erate could form colonies. In this condition, CD133�

CD45� Ter119� cells isolated from no-DDC liversformed only SCs that contained 10 to 50 cells following8-day culture of cells (Fig. 2E). On the other hand,CD133� CD45� Ter119� cells isolated from DDC-treated livers were capable of forming not only SCs butalso LCs comprising several hundred cells (Fig. 2E).These two types of colonies were easily distinguished inculture by their size. Both SCs and LCs were frequentwhen CD133� CD45� Ter119� cells were isolated andcultured, whereas few or no colonies were formed by cellsisolated from CD45�, Ter119�, and CD133� CD45�

Ter119� cell subpopulations (Fig. 2F). The appearanceof cells that could form LCs was strongly dependent onDDC treatment, and the number of SCs was also in-creased more than 5 times in DDC-treated livers (Fig.2F). Moreover, LC formation by DDC-treated liver-de-rived CD133� CD45� Ter119� cells critically requiredgrowth factors (hepatocyte growth factor and/or epider-mal growth factor) and extracellular matrix components(type IV collagen or laminin) in the culture conditions(Supplementary Fig. 1).

Single Cell Culture and Clonal Analysis of CD133�

CD45� Ter119� Cells. To examine the potential ofeach CD133� CD45� Ter119� cell, we attempted toculture one cell in each well of type IV collagen-coated96-well plates, following clone-sorting of cells via flowcytometry. As in the semiclonal low-density culture con-dition, cells isolated from CD133� CD45� Ter119� cellsubpopulation could form SCs or LCs in each well, ac-


Fig. 2. In vitro semiclonal colony assay of CD133� CD45� Ter119� cells following isolation via flow cytometry. (A) No-DDC and DDC-treated livercells were fractionated via flow cytometry based on the expression of cell surface markers. Initially, CD45� and Ter119� cells were gated and removedfrom no-DDC and DDC-treated liver tissue specimens. The CD45� Ter119� cells, whose ratios are shown in each left panel, were then furtherfractionated based on CD133 expression. For the in vitro colony assay, the sorting gates were set for the CD133� CD45� Ter119� and CD133�

CD45� Ter119� cell subpopulations. The ratios of CD133� cells in CD45� Ter119� cells and in unfractionated total cells are shown as percentageoutside and inside parentheses in each right panel, respectively. The small panel within the right lower panel shows the result of reanalysis of freshlyisolated CD133� CD45� Ter119� cells via flow cytometry. The Y-axes of the dot plots show the level of phycoerythrin fluorescence. No-DDC livercells stained with isotype control antibodies were used as a control (upper panels). (B) Co-immunofluorescence staining of Alb and CK7 for CD133�

CD45� Ter119� cells adhered to glass slides, soon after isolation from no-DDC and DDC-treated livers via flow cytometry. DNA is stained with DAPI.Scale bars � 50 �m. (C) The percentage of cells being positive for Alb, CK7, and CK19, which was determined after co-immunofluorescence stainingof these markers for freshly isolated CD133� CD45� Ter119� cells. We examined 1,638 cells (DDC) and 1,098 cells (no-DDC) for Alb/CK7 stainingand 1,599 cells (DDC) and 1,279 cells (no-DDC) for Alb/CK19 staining in three independent experiments (mean � standard deviation). Note thatno Alb� cell was contained in isolated CD133� CD45� Ter119� cells. (D) RT-PCR analysis revealed that messenger RNA expression of hepatocytemarker genes was detected in no-gated and CD133� CD45� Ter119� cells, but not in CD133� CD45� Ter119� cells isolated from DDC-treatedlivers. (E) Semiclonal colony formation from CD133� CD45� Ter119� cells isolated from no-DDC and DDC-treated livers after 8 days of culture. LCsappeared to form only from CD133� CD45� Ter119� cells obtained from DDC-treated liver. Representative colonies are shown. Scale bars � 500�m (upper panels) and 100 �m (lower panels). (F) Numbers of SCs and LCs per 1,000 cells in each cell subpopulation derived from no-DDC andDDC-treated livers. The graph shows the average of 12 dishes for each cell subpopulation in three independent experiments (mean � standarddeviation).


cording to whether the liver has been pretreated withDDC (Fig. 3A-F). To ascertain that single cells have beendeposited, we always examine each well to confirm thepresence of a single cell under the microscope after clone-sorting. Once a cell sorter is adjusted for the optimalsetting prior to the experiment, we seldom find wells withmore than two cells after clone-sorting. When we foundthese wells, we excluded them from samples for analysis.

Clonal colony formation in the single cell culture ofCD133� CD45� Ter119� cells enabled us to investigatethe types of cells that differentiated from original cells.Immunofluorescence analysis using antibodies againstAlb and CK7 revealed that most SCs derived from no-DDC and DDC-treated livers consisted of cells express-ing only CK7 at days 6 and 18 of culture (Fig. 4A,B). Incontrast, at day 6 of culture, half of LCs contained CK7�

cells, Alb� CK7� cells, and a small number of Alb� cells,whereas the other half of these colonies were formed ex-clusively by CK7� cells (Fig. 4A,B). After 18 days, how-ever, cells in more than 70% of LCs eventually gave rise toAlb� cells, CK7� cells, and a small number of Alb� CK7�

cells (Fig. 4A,B). Regardless of colony size and length ofculture, no colony formed by only Alb� cells was observed(Fig. 4B). RT-PCR also showed that cells in LCs ex-pressed multiple genes encoding markers for hepatocytesand cholangiocytes, including genes detected in maturehepatocytes, such as tryptophan-2,3-dioxygenase, glutathi-one S-transferase, and glutamine synthetase (Table 1).Moreover, in LCs cultured for more than 3 weeks, manybinucleate Alb� cells with abundant glycogen stores, re-

sembling functionally mature hepatocytes, emerged nearthe center of colonies (Fig. 4C,D). Similar results werealso obtained from single cell culture of CD133� CD45�

Ter119� cells deposited onto each well of 96-well platesby micromanipulation, following isolation from DDC-treated liver (Supplementary Fig. 2). Because CD133�

CD45� Ter119� cells isolated from DDC-treated liverswere initially positive for CK7 and CK19, but not for Alb(Fig. 2B-D), we conclude that cells capable of formingLCs gave rise to progeny that differentiated into Alb�

mature hepatocytes in the process of culture.Self-Renewal Capability of CD133� LCCs in Cul-

ture. We next performed subcloning experiments to testthe self-renewal potential of CD133� CD45� Ter119�

cells. Cells in individual clonal colonies formed in eachwell of type IV collagen-coated 96-well plates weretrypsinized, replated into each well of type IV collagen-coated 6-well plates, and cultured as independent clones.We used SCs or SCs/LCs obtained from no-DDC orDDC-treated livers, respectively. After 14 days of culture,many secondary LCs (5 to 10 colonies from each clone)arose from progeny of a CD133� LCC isolated fromDDC-treated liver, but not from progeny of SC-formingCD133� CD45� Ter119� cells (CD133� small colony-forming cell [SCC]) isolated from no-DDC and DDC-treated livers (Fig. 5A and data not shown). Cellspropagating by forming secondary LCs were thentrypsinized again and underwent clone-sorting and singlecell culture in each well of type IV collagen-coated 96-wellplates. Although the percentage of resorted cells that

Fig. 3. Single-cell culture andclonal colony formation of CD133�

CD45� Ter119� cells. (A) Single-cell culture of no-DDC liver-derivedCD133� CD45� Ter119� cells. Arepresentative SC that was formedin a well of type IV collagen-coated96-well plates at day 8 of culture isshown. (B,C) Magnified pictures ofthe SC presented in panel A. Scalebar � 500 �m (B) and 100 �m(C). (D) Single cell culture of DDC-treated liver-derived CD133�

CD45� Ter119� cells. A representa-tive LC that was formed in a well oftype IV collagen-coated 96-wellplates at day 8 of culture is shown.(E,F) Magnified pictures of the LCpresented in panel D. Scale bar �500 �m (E) and 100 �m (F).


Fig. 4. In vitro bilineage colony formation from DDC-treated liver-derived CD133� LCCs. (A) Co-immunofluorescence staining of Alb and CK7 wasconducted for clonal colonies upon single-cell culture of CD133� CD45� Ter119� cells, isolated from no-DDC or DDC-treated livers, for 6 and 18days. Representative colonies are shown. DNA is stained with DAPI (lower panels). Scale bars � 100 �m. (B) The percentage of clonal coloniesconstituted by Alb� cells and CK7� cells, Alb� cells and CK7� cells, Alb� cells and CK7� cells, or Alb� cells and CK7� cells was determined afterco-immunofluorescence staining of Alb and CK7 for these colonies formed from no-DDC or DDC-treated liver-derived CD133� CD45� Ter119� cellsat days 6 and 18 of culture. Note that the data for SCs formed from no-DDC and DDC-treated liver-derived CD133� SCCs were similar. This graphshows the average of three independent experiments (mean � standard deviation). The number of colonies examined in each experiment is 16 to62. (C) PAS staining for clonal colonies revealed that a DDC-treated liver-derived CD133� LCC, but not a no-DDC liver-derived CD133� SCC, gaverise to functionally mature hepatocytes, containing abundant glycogen stores, after 25 days of culture. Scale bars � 100 �m. (D) Immunofluo-rescence staining of Alb for clonal colonies revealed that binucleate Alb� mature hepatocytes were differentiated from a DDC-treated liver-derivedCD133� LCC at day 30 of culture. DNA is stained with DAPI. Scale bar � 100 �m.


formed LCs was low (2.4 � 0.6% [three independentclones were analyzed by using each 5 plate]), cells in morethan 80% of LCs differentiated into both Alb� cells andCK7� cells by day 18 of culture (Fig. 5B,C). When weperformed single-cell culture of clonogenic progeny of aCD133� LCC by putting cells in each well of 96-wellplates with a micromanipulator, cells capable of reform-ing LCs could also differentiate into both Alb� cells andCK7� cells (data not shown). These data indicate thatCD133� LCCs isolated from DDC-treated livers notonly produced both Alb� hepatocytes and CK7� cholan-giocytes as descendants, but also maintained cells harbor-ing bilineage differentiation potential by self-renewingcell divisions in culture.

In Vivo Liver Repopulation by Clonogenic Progenyof a CD133� LCC. The reconstitution and functionalrepair of injured tissues is one of the specific roles fororgan stem/progenitor cells. It has been previously re-ported that an oval cell population isolated from DDC-treated mouse livers could reconstitute hepatic tissuesafter transplantation into the livers of FAH-deficientmice.9 To determine whether CD133� LCCs isolatedfrom DDC-treated livers could also reconstitute hepatictissues, we intrasplenically injected clonally propagatingcells derived from a CD133� LCC into FAH�/� mouselivers. The FAH�/� mice used here were generated fromES cells with targeted disruption of FAH (Figure 6), andthese mice were phenotypically indistinguishable fromanother FAH mutant mouse line.31

In this study, we used clonogenic progeny of aCD133� LCC isolated from a DDC-treated liver as do-nor cells, which contained bipotent primitive hepatic cellsand differentiating hepatocytes and cholangiocytes, fol-lowing in vitro propagation for 4 to 5 weeks. At 1 and 2months posttransplantation, FAH� donor cells engrafteddiffusively and reconstituted hepatic nodules in all recip-ient mice, thus these mice could survive even in long-termNTBC withdrawal (Fig. 7A-C). In addition, immunoflu-orescence staining and periodic acid-Schiff staining re-vealed that FAH� donor-derived cells expressed Alb andcontained abundant glycogen stores (data not shown).Moreover, to examine the possibility of cell fusion, weconducted dual Y chromosome fluorescence in situ hy-bridization with FAH immunofluorescence staining onFAH�/� male mouse liver after transplantation of cellsprepared from culture of clonogenic progeny of aCD133� LCC derived from DDC-treated female mouseliver. In FAH� donor-derived cells, we did not detect anyY chromosome, indicating that donor cells have not fusedwith recipient cells after transplantation (SupplementalFig. 3). These data demonstrated that clonogenic progenyof a CD133� LCC isolated from a DDC-treated liver hadthe potential to reconstitute hepatic tissues by giving riseto morphologically and functionally mature hepatocytes.

Hepatocarcinoma Formation by Clonogenic Prog-eny of a p53-Deficient CD133� LCC. In addition toDDC treatment, oval cells also emerge in several experi-mental models for rodent liver carcinogenesis using

Table 1. Expression of Lineage Marker Genes in Colonies Formed After Clone-Sorting

Colony number 1 2 3 4 5 6 7 8 9 10 11 12

Hepatocyte markersAlbumin � � � � � � � � � � � ��-fetoprotein � � � � � � � � � � � ��-1-antitrypsin � � � � � � � � � � � �Glucose-6-phosphatase � � � � � � � � � � � �Dipeptidylpeptidase IV � � � � � � � � � � � �Tryptophan-2,3-dioxygenase � � � � � � � � � � � �Glutathione S-transferase � � � � � � � � � � � �Glutamine synthetase � � � � � � � � � � � �

Cholangiocyte markersCK19 � � � � � � � � � � � �CK7 � � � � � � � � � � � �Thymosin �4 � � � � � � � � � � � �Billary glycoprotein � � � � � � � � � � � ��-glutamyl transpeptidase � � � � � � � � � � � �Vinculin � � � � � � � � � � � �

MiscellaneousCK18 � � � � � � � � � � � �CK8 � � � � � � � � � � � �TTR � � � � � � � � � � � �c-met � � � � � � � � � � � �HPRT � � � � � � � � � � � �

Twelve individual LCs formed upon single-cell culture of DDC-treated liver-derived CD133� CD45� Ter119� cells were examined via RT-PCR at day 18.


Dipin32 or the choline-deficient plus ethionine diet,33 aswell as in human liver pathologies (hepatitis C virus,hemochromatosis, and alcoholic liver disease34) that de-velop more efficiently to hepatocarcinoma.35-37 Thus, it ispossible to speculate that transformed oval cells are can-didates for the origin of both rodent and human hepato-carcinoma. This idea is also supported by recent evidencesuggesting the presence of putative cancer-initiating cellsor cancer stem cells in many types of cancers.38-46 Consis-tent with this idea, oval cells derived from p53-deficientmice that were treated with choline-deficient and ethi-onine-supplemented diet could produce immortalizedcell lines that could generate tumors after injection intonude mice.47

To examine whether CD133� CD45� Ter119� cellscould also initiate tumor formation, we isolated these cellsfrom no-DDC and DDC-treated p53�/� livers. Tumorswere not observed in the livers of p53�/� mice when we

isolated cells. The CD133� CD45� Ter119� cells iso-lated from DDC-treated p53�/� livers, but not thosefrom no-DDC p53�/� livers, formed LCs and propagatedin culture, similar to wild-type liver-derived CD133�

CD45� Ter119� cells (Fig. 8A). In contrast, anyCD133� CD45� Ter119� cells isolated from no-DDCand DDC-treated p53�/� livers could not form LCs norexpand in culture. The progeny of a CD133� LCC ob-tained from DDC-treated p53�/� livers displayed smallercell size and had a significantly higher growth capacitythan that from wild-type livers (Fig. 8A,B), while thepotential to differentiate into hepatocytes and cholangio-cytes was not affected (Fig. 8C,D). At 2 months after thesubcutaneous injection of clonogenic cells expanded froma DDC-treated p53�/� liver-derived CD133� LCC intoNOD/SCID mice, tumors containing Alb� hepatocytesand bile duct–like structure-forming keratin� cholangio-cytes were generated in all animals, while those from wild-

Fig. 5. Bipotency of secondary clone-sorted progeny of a CD133� LCC isolated from DDC-treated liver. (A) Following subculture of cells formingindividual colonies for 14 days, secondary colonies formed from progeny of a DDC-treated liver-derived CD133� LCC, but not from that of a no-DDCliver-derived CD133� SCC. These secondary colonies were stained with Tuerk’s solution and photoimaged. To analyze the self-renewal potential ofDDC-treated liver-derived CD133� LCCs, cells in secondary colonies underwent secondary clone-sorting via flow cytometry and single-cell culture. (B)Co-immunofluorescence staining of Alb and CK7 was conducted for LCs, upon 18 days single cell culture of resorted progeny of a DDC-treatedliver-derived CD133� LCC. A representative colony is shown. DNA is stained with DAPI. Scale bars � 100 �m. (C) The percentage of clonal LCsconstituted by Alb� cells and CK7� cells, Alb� cells and CK7� cells, Alb� cells and CK7� cells, or Alb� cells and CK7� cells was determined afterco-immunofluorescence staining of Alb and CK7 for these colonies formed from resorted progeny of a DDC-treated liver-derived CD133� LCC at day18 of culture. The graph shows the average of three independent experiments (mean � standard deviation). The number of colonies examined ineach experiment is 12 to 21.


type livers never formed tumors (Fig. 8E-K). Therefore,upon the loss of p53 function, clonogenic progeny of aCD133� LCC isolated from DDC-treated livers was ca-pable of forming tumors with some characteristics of bothhepatocellular carcinoma and cholangiocarcinoma inNOD/SCID mice, while it is still difficult to determinewhether these tumors were derived from bipotent primi-tive hepatic cells that were maintained in culture or fromtransformed hepatocytes and cholangiocytes that werenewly generated from CD133� LCCs in the process ofculture.

DiscussionRecently, CD133 was identified as a marker not only

for organ/tissue stem cells or progenitor cells,48-52 but alsofor putative cancer-initiating cells or cancer stemcells.38-46 Although the role of CD133 in these cells isunknown, this antigen might be exploited as a commoncell surface marker for isolation of immature cells fromboth normal and malignant tissues. In the case of the liver,

however, the identity of CD133� cells has been contro-versial. The expression of CD133 was used to isolate he-patic stellate cells from adult rat liver,53 but in anotherreport, its expression was increased during induction ofrat oval cells.54 On the other hand, CD133 was not de-tectable in putative human hepatic progenitor cells.55

More recently, Rountree et al.56 reported that CD133�

CD45� cells isolated from DDC-treated mouse livers ex-pressed both Alb and CK19, as assessed with RT-PCR,and proliferated in culture with expression of thesemarker genes. Although these results suggested expressionof CD133 in putative oval cells, the contamination ofother cell types in isolated CD133� cells could not beruled out, and the potential for bilineage differentiation,self-renewing cell division, and hepatic tissue reconstitu-tion was not examined using single cell-based assays. Inthis study, we demonstrated that, at least in mouse livers,CD133 is a marker for normal cholangiocytes and DDC-induced primitive ductal cells. In single-cell cultures ofCD133� LCCs isolated from DDC-treated livers, both

Fig. 6. Targeting strategy for pro-ducing the FAH null allele. (A) Tar-geted disruption of the mouse FAHby homologous recombination in EScells. (B) Genomic southern blottingof parental and targeted ES cellsusing 5� and 3� probes. (C) PCRgenotyping using genomic DNA fromtail snips obtained from FAH�/�

wild-type, FAH�/� heterozygous,and FAH�/� mice. M, marker.


hepatocytes and cholangiocytes appeared to differentiateas culture advanced, with maintenance of undifferenti-ated cells by self-renewing cell divisions. Following trans-plantation of clonogenic progeny propagating from aDDC-treated liver-derived CD133� LCC into the liverof FAH�/� mice, these cells could reconstitute hepatictissues by differentiating into fully functional hepato-cytes, and were able to rescue liver function in these mice.Moreover, after subcutaneous injection into NOD/SCIDmice, the progeny of a DDC-treated p53�/� liver-derivedCD133� LCC generated tumors containing hepatocytesand bile duct–like structures. Taken together, these dataindicate that CD133� LCCs in DDC-treated livers fulfillthe criteria for hepatic oval cells: potential for bilineagedifferentiation, self-renewing cell division, hepatic tissuereconstitution, and tumor formation upon loss of p53.Thus, CD133 is a useful marker for isolation of progeni-tor cells from the adult liver, as well as from other organsand tissues. Our recent study also showed that CD133 isexpressed in pancreatic ductal cells that contained multi-potent progenitor cells in pancreas development,52 sug-gesting that CD133 could be a common marker for stem/progenitor cells within endoderm-derived digestiveorgans.

In the rat 2-acetylaminofluorene plus liver injurymodel, oval cells proliferate and gradually differentiateinto hepatocytes marked by the expression of liver-en-riched transcription factors and Alb.10-12 As reported here,CD133� LCCs obtained from DDC-treated mouse livers

could also differentiate into mature hepatocytes followingculture for several weeks with hepatocyte growth factor,epidermal growth factor, and extracellular matrix compo-nents, and after transplantation into FAH-deficient mice.After the treatment with DDC, however, hepatocyte nu-clear factor 4�, a liver-enriched transcription factor, andAlb were never detectable in primitive ductal cells appear-ing in liver tissue (data not shown). Furthermore,CD133� CD45� Ter119� cells isolated from no-DDCand DDC-treated livers were initially negative for theexpression of hepatocyte markers. Thus, DDC-inducedprimitive ductal cells proliferate, but do not differentiateinto hepatocytes without any additional cues for hepato-cyte proliferation and differentiation. This is consistentwith a fact that a hepatic injury required for inducinghepatocyte proliferation and differentiation is not in-cluded in the DDC mouse model. Therefore, the DDCmodel would be a powerful tool for the isolation of un-differentiated progenitor cells and the investigation oftheir potential for proliferation and differentiation.

The CD133� CD45� Ter119� cells isolated fromDDC-treated livers could form LCs that contained bothhepatocytes and cholangiocytes in a single-cell culturecondition. In addition to LCs, however, SCs comprisingonly CK7� cells were also formed from these CD133�

CD45� Ter119� cells. Because CD133 was expressed incholangiocytes, and because CD133� CD45� Ter119�

cells isolated from no-DDC livers formed only SCs con-sisting of CK7� cholangiocytes, we conclude that these

Fig. 7. Clonogenic progeny prop-agating in culture from a CD133�

LCC isolated from a DDC-treatedliver can reconstitute hepatic tissuesin vivo. (A) Immunohistochemicalstaining of FAH was conducted foran NTBC-supplemented FAH�/�

mouse liver (upper panel) and for aFAH�/� liver at 2 months aftertransplantation of cells preparedfrom culture of clonogenic progenyof a DDC-treated liver-derivedCD133� LCC (lower panel). Manydonor cell–derived FAH� hepaticnodules were found in the recipientmouse liver. Insets denote hematox-ylin liver staining of each slide. (B,C)Magnified pictures of donor cell-de-rived FAH� hepatic nodules in therecipient FAH�/� liver at 1 month(B) or 2 months (C) after transplant.Scale bars � 100 �m.


Fig. 8. Hepatocarcinoma formation by clonogenic progeny of a p53-deficient CD133� LCC isolated from DDC-treated liver. (A) Regardless of thelack of p53, clonal LCs and SCs were formed from CD133� CD45� Ter119� cells isolated from DDC-treated and no-DDC liver, respectively, similarto the result of wild-type (p53�/�) liver-derived CD133� CD45� Ter119� cells. Cells proliferating from a DDC-treated p53�/� liver-derived CD133�

LCC, however, are smaller than those from a DDC-treated wild-type liver-derived CD133� LCC. Scale bars � 100 �m. (B) Clonogenic progeny ofa CD133� LCC isolated from DDC-treated p53�/� liver proliferated more vigorously than that from a DDC-treated wild-type liver. Initially, 1,000 cellswere plated in each well of type IV collagen-coated 6-well plates (n � 3; mean � standard deviation). (C,D) Co-immunofluorescence staining ofAlb and CK7 was conducted for LCs upon single cell culture of DDC-treated p53�/� liver-derived CD133� LCCs for 18 days. A representative colonyis shown. DNA is stained with DAPI (D). Scale bar � 50 �m. (E) Clonogenic progeny of a CD133� LCC isolated from DDC-treated p53�/� liver,but not from DDC-treated wild-type liver, formed tumors at 2 months after subcutaneous injection of cells into NOD/SCID recipient mice. (F) The tumordissected from a recipient mouse shown in panel E. Scale bar � 1 cm. (G,H) Hematoxylin-eosin staining of areas of tumor with (H) or without (G)duct-like structures. Scale bars � 100 �m. (I-K) Areas of tumor containing many Alb� cells and few keratin� cells (I), many keratin� cells that formedduct-like structures and a few Alb� cells (J), or both Alb� cells and duct-like structure-forming keratin� cells (K). The data shown in G-K were obtainedfrom the tumor presented in F. Scale bars � 100 �m.


SCs arose from lineage-committed cholangiocytes. Theincrease of the number of CD133� CD45� Ter119�

cells in DDC-treated livers may thus occur as a result ofcholangiocyte proliferation, and bipotent hepatic ovalcells may appear as a small population in proliferatingcholangiocytes that form primitive ductules, suggestingthat oval cells might be derived from cholangiocytes. Todistinguish oval cells from cholangiocytes, oval cell–spe-cific or cholangiocyte-specific antigens should be identi-fied and used in combination with CD133.

During liver development, hepatic progenitor cellscontribute to several steps of organogenesis, using theirpotential for proliferation and differentiation. In the adultliver, however, these cell classes may be dormant and mayonly be activated in the context of some unavoidable cri-sis. Therefore, oval cells are considered the equivalent ofhepatic progenitor cells that transiently emerge and pro-liferate under specific conditions. While the origin of ovalcells remains to be determined, it is still promising that wehave now identified a marker to isolate these cells andhave succeeded in single-cell culture analysis. Thisprogress will facilitate a fundamental understanding ofthe characteristics of oval cells: the signals that determinetheir lineage commitment, the genes that control theirdifferentiation and self-renewing cell divisions, theirsource and progeny lineages, and the manner in whichoval cells are associated with liver regeneration and tumorformation. Cell populations enriched in prospectivelyidentified oval cells will also provide a tool for developingtherapeutic strategies, such as gene therapy, cell therapybased on cell transplantation, and artificial livers.

Acknowledgment: We thank Valentina Factor, KimZ. Travis, John Doe, and Robert M. Tanguay for sharingreagents; Shinichi Aizawa for providing p53-deficientmice (CDB0001K); Kenichiro Uno and Junko Nishio formicroarray analysis; and Keiko Sueyoshi and Setsuko Fu-jii for excellent technical assistance.

References1. Michalopoulos GK, DeFrances MC. Liver regeneration. Science 1997;

276:60-66.2. Fausto N, Campbell JS. The role of hepatocytes and oval cells in liver

regeneration and repopulation. Mech Dev 2003;120:117-130.3. Lee JS, Heo J, Libbrecht L, Chu IS, Kaposi-Novak P, Calvisi DF, et al. A

novel prognostic subtype of human hepatocellular carcinoma derived fromhepatic progenitor cells. Nat Med 2006;12:410-416.

4. Tatematsu M, Ho RH, Kaku T, Ekem JK, Farber E. Studies on the pro-liferation and fate of oval cells in the liver of rats treated with 2-acetylami-nofluorene and partial hepatectomy. Am J Pathol 1984;114:418-430.

5. Petersen BE, Zajac VF, Michalopoulos GK. Hepatic oval cell activation inresponse to injury following chemically induced periportal or pericentraldamage in rats. HEPATOLOGY 1998;27:1030-1038.

6. Paku S, Schnur J, Nagy P, Thorgeirsson SS. Origin and structural evolu-tion of the early proliferating oval cells in rat liver. Am J Pathol 2001;158:1313-1323.

7. Sell S. Heterogeneity and plasticity of hepatocyte lineage cells. HEPATOL-OGY 2001;33:738-750.

8. Preisegger KH, Factor VM, Fuchsbichler A, Stumptner C, Denk H, Thor-geirsson SS. Atypical ductular proliferation and its inhibition by trans-forming growth factor beta1 in the 3,5-diethoxycarbonyl-1,4-dihydrocollidine mouse model for chronic alcoholic liver disease. LabInvest 1999;79:103-109.

9. Wang X, Foster M, Al-Dhalimy M, Lagasse E, Finegold M, Grompe M.The origin and liver repopulating capacity of murine oval cells. Proc NatlAcad Sci U S A 2003;100 Suppl 1:11881-11888.

10. Evarts RP, Nagy P, Nakatsukasa H, Marsden E, Thorgeirsson SS. In vivodifferentiation of rat liver oval cells into hepatocytes. Cancer Res 1989;49:1541-1547.

11. Nagy P, Bisgaard HC, Thorgeirsson SS. Expression of hepatic transcrip-tion factors during liver development and oval cell differentiation. J CellBiol 1994;126:223-233.

12. Paku S, Nagy P, Kopper L, Thorgeirsson SS. 2-acetylaminofluorene dose-dependent differentiation of rat oval cells into hepatocytes: confocal andelectron microscopic studies. HEPATOLOGY 2004;39:1353-1361.

13. Yaswen P, Hayner NT, Fausto N. Isolation of oval cells by centrifugalelutriation and comparison with other cell types purified from normal andpreneoplastic livers. Cancer Res 1984;44:324-331.

14. Germain L, Noel M, Gourdeau H, Marceau N. Promotion of growth anddifferentiation of rat ductular oval cells in primary culture. Cancer Res1988;48:368-378.

15. Suzuki A, Zheng YW, Kondo R, Kusakabe M, Takada Y, Fukao K, et al.Flow cytometric separation and enrichment of hepatic progenitor cells inthe developing mouse liver. HEPATOLOGY 2000;32:1230-1239.

16. Suzuki A, Zheng YW, Kaneko S, Onodera M, Fukao K, Nakauchi H, et al.Clonal identification and characterization of self-renewing pluripotentstem cells in the developing liver. J Cell Biol 2002;156:173-184.

17. Tanimizu N, Nishikawa M, Saito H, Tsujimura T, Miyajima A. Isolationof hepatoblasts based on the expression of Dlk/Pref-1. J Cell Sci 2003;116:1775-1786.

18. Tsukada T, Tomooka Y, Takai S, Ueda Y, Nishikawa S, Yagi T, et al.Enhanced proliferative potential in culture of cells from p53-deficientmice. Oncogene 1993;8:3313-3322.

19. Suzuki A, Iwama A, Miyashita H, Nakauchi H, Taniguchi H. Role forgrowth factors and extracellular matrix in controlling differentiation ofprospectively isolated hepatic stem cells. Development 2003;130:2513-2524.

20. Suzuki A, Nakauchi H, Taniguchi H. Glucagon-like peptide 1 (1-37)converts intestinal epithelial cells into insulin-producing cells. Proc NatlAcad Sci U S A 2003;100:5034-5039.

21. Overturf K, Al-Dhalimy M, Tanguay R, Brantly M, Ou CN, Finegold M,et al. Hepatocytes corrected by gene therapy are selected in vivo in a murinemodel of hereditary tyrosinaemia type I. Nat Genet 1996;12:266-273.

22. Yagi T, Ikawa Y, Yoshida K, Shigetani Y, Takeda N, Mabuchi I, et al.Homologous recombination at c-fyn locus of mouse embryonic stem cellswith diphtheria toxin A fragment gene in negative selection. Proc NatlAcad Sci U S A 1990;87:9918-9922.

23. Yagi T, Nada S, Watanabe N, Tamemoto H, Kohmura N, Ikawa Y, et al.A novel negative selection for homologous recombinants using diphtheriatoxin A fragment gene. Anal Biochem 1993;214:77-86.

24. Yagi T, Tokunaga T, Furuta Y, Nada S, Yoshida M, Tsukada T, et al. Anovel ES cell line, TT2, with high germline-differentiating potency. AnalBiochem 1993;214:70-76.

25. Murata T, Furushima K, Hirano M, Kiyonari H, Nakamura M, Suda Y, etal. ang is a novel gene expressed in early neuroectoderm, but its null mutantexhibits no obvious phenotype. Gene Expr Patterns 2004;5:171-178.

26. Lemire JM, Shiojiri N, Fausto N. Oval cell proliferation and the origin ofsmall hepatocytes in liver injury induced by D-galactosamine. Am J Pathol1991;139:535-552.

27. Fujio K, Evarts RP, Hu Z, Marsden ER, Thorgeirsson SS. Expression ofstem cell factor and its receptor, c-kit, during liver regeneration fromputative stem cells in adult rat. Lab Invest 1994;70:511-516.


28. Omori N, Omori M, Evarts RP, Teramoto T, Miller MJ, Hoang TN, et al.Partial cloning of rat CD34 cDNA and expression during stem cell-depen-dent liver regeneration in the adult rat. HEPATOLOGY 1997;26:720-727.

29. Petersen BE, Goff JP, Greenberger JS, Michalopoulos GK. Hepatic ovalcells express the hematopoietic stem cell marker Thy-1 in the rat. HEPA-TOLOGY 1998;27:433-445.

30. Matsusaka S, Tsujimura T, Toyosaka A, Nakasho K, Sugihara A, Oka-moto E, et al. Role of c-kit receptor tyrosine kinase in development of ovalcells in the rat 2-acetylaminofluorene/partial hepatectomy model. HEPA-TOLOGY 1999;29:670-676.

31. Grompe M, al-Dhalimy M, Finegold M, Ou CN, Burlingame T, Ken-naway NG, et al. Loss of fumarylacetoacetate hydrolase is responsible forthe neonatal hepatic dysfunction phenotype of lethal albino mice. GenesDev 1993;7:2298-2307.

32. Factor VM, Radaeva SA. Oval cells-hepatocytes relationships in Dipin-induced hepatocarcinogenesis in mice. Exp Toxicol Pathol 1993;45:239-244.

33. Shinozuka H, Lombardi B, Sell S, Iammarino RM. Early histological andfunctional alterations of ethionine liver carcinogenesis in rats fed a choline-deficient diet. Cancer Res 1978;38:1092-1098.

34. Lowes KN, Brennan BA, Yeoh GC, Olynyk JK. Oval cell numbers inhuman chronic liver diseases are directly related to disease severity. Am JPathol 1999;154:537-541.

35. Prior P. Long-term cancer risk in alcoholism. Alcohol Alcohol 1988;23:163-171.

36. Tsukuma H, Hiyama T, Tanaka S, Nakao M, Yabuuchi T, Kitamura T, etal. Risk factors for hepatocellular carcinoma among patients with chronicliver disease. N Engl J Med 1993;328:1797-1801.

37. Deugnier YM, Guyader D, Crantock L, Lopez JM, Turlin B, Yaouanq J, etal. Primary liver cancer in genetic hemochromatosis: a clinical, pathologi-cal, and pathogenetic study of 54 cases. Gastroenterology 1993;104:228-234.

38. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as ahierarchy that originates from a primitive hematopoietic cell. Nat Med1997;3:730-737.

39. Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF.Prospective identification of tumorigenic breast cancer cells. Proc NatlAcad Sci U S A 2003;100:3983-3988.

40. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, et al.Identification of human brain tumour initiating cells. Nature 2004;432:396-401.

41. Collins AT, Berry PA, Hyde C, Stower MJ, Maitland NJ. Prospectiveidentification of tumorigenic prostate cancer stem cells. Cancer Res 2005;65:10946-10951.

42. Ricci-Vitiani L, Lombardi DG, Pilozzi E, Biffoni M, Todaro M, Peschle C,et al. Identification and expansion of human colon-cancer-initiating cells.Nature 2007;445:111-115.

43. O’Brien CA, Pollett A, Gallinger S, Dick JE. A human colon cancer cellcapable of initiating tumour growth in immunodeficient mice. Nature2007;445:106-110.

44. Prince ME, Sivanandan R, Kaczorowski A, Wolf GT, Kaplan MJ, DalerbaP, et al. Identification of a subpopulation of cells with cancer stem cellproperties in head and neck squamous cell carcinoma. Proc Natl Acad SciU S A 2007;104:973-978.

45. Zucchi I, Sanzone S, Astigiano S, Pelucchi P, Scotti M, Valsecchi V, et al.The properties of a mammary gland cancer stem cell. Proc Natl Acad SciU S A 2007;104:10476-10481.

46. Li C, Heidt DG, Dalerba P, Burant CF, Zhang L, Adsay V, et al. Identi-fication of pancreatic cancer stem cells. Cancer Res 2007;67:1030-1037.

47. Dumble ML, Croager EJ, Yeoh GC, Quail EA. Generation and character-ization of p53 null transformed hepatic progenitor cells: oval cells give riseto hepatocellular carcinoma. Carcinogenesis 2002;23:435-445.

48. Bussolati B, Bruno S, Grange C, Buttiglieri S, Deregibus MC, Cantino D,et al. Isolation of renal progenitor cells from adult human kidney. Am JPathol 2005;166:545-555.

49. Uchida N, Buck DW, He D, Reitsma MJ, Masek M, Phan TV, et al.Direct isolation of human central nervous system stem cells. Proc NatlAcad Sci U S A 2000;97:14720-14725.

50. Richardson GD, Robson CN, Lang SH, Neal DE, Maitland NJ, CollinsAT. CD133, a novel marker for human prostatic epithelial stem cells.J Cell Sci 2004;117:3539-3545.

51. Lee A, Kessler JD, Read TA, Kaiser C, Corbeil D, Huttner WB, et al.Isolation of neural stem cells from the postnatal cerebellum. Nat Neurosci2005;8:723-729.

52. Oshima Y, Suzuki A, Kawashimo K, Ishikawa M, Ohkohchi N, TaniguchiH. Isolation of mouse pancreatic ductal progenitor cells expressing CD133and c-Met by flow cytometric cell sorting. Gastroenterology 2007;132:720-732.

53. Kordes C, Sawitza I, Muller-Marbach A, Ale-Agha N, Keitel V,Klonowski-Stumpe H, et al. CD133� hepatic stellate cells are progenitorcells. Biochem Biophys Res Commun 2007;352:410-417.

54. Yovchev MI, Grozdanov PN, Joseph B, Gupta S, Dabeva MD. Novelhepatic progenitor cell surface markers in the adult rat liver. HEPATOLOGY

2007;45:139-149.55. Herrera MB, Bruno S, Buttiglieri S, Tetta C, Gatti S, Deregibus MC, et al.

Isolation and characterization of a stem cell population from adult humanliver. Stem Cells 2006;24:2840-2850.

56. Rountree CB, Barsky L, Ge S, Zhu J, Senadheera S, Crooks GM. A CD133expressing murine liver oval cell population with bi-lineage potential. StemCells 2007;25:2419-2429.


Documents

Flow cytometric isolation and clonal identification of self-renewing bipotent hepatic progenitor cells in adult mouse liver