Mobilization of hepatic mesenchymal stem cells from human liver grafts

ORIGINAL ARTICLE

Mobilization of Hepatic Mesenchymal Stem CellsFrom Human Liver GraftsQiuwei Pan,1 Suomi M. G. Fouraschen,2 Fatima S. F. Aerts Kaya,3 Monique M. Verstegen,3,4

Mario Pescatori,5 Andrew P. Stubbs,5 Wilfred van IJcken,6 Antoine van der Sloot ,2 Ron Smits,1

Jaap Kwekkeboom,1 Herold J. Metselaar,1 Geert Kazemier,2 Jeroen de Jonge,2 Hugo W. Tilanus,2

Gerard Wagemaker,3 Harry L. A. Janssen,1 and Luc J. W. van der Laan2

1Department of Gastroenterology and Hepatology, 2Laboratory of Experimental Transplantation andIntestinal Surgery, Department of Surgery, 3Department of Hematology, 4Department of Pediatrics;5Department of Bioinformatics; and 6Center for Biomics, Erasmus MC–University Medical Center Rotterdam,Rotterdam, the Netherlands

Extensive studies have demonstrated the potential applications of bone marrow–derived mesenchymal stem cells (BM-MSCs) asregenerative or immunosuppressive treatments in the setting of organ transplantation. The aims of the present study were toexplore the presence and mobilization of mesenchymal stem cells (MSCs) in adult human liver grafts and to compare their func-tional capacities to those of BM-MSCs. The culturing of liver graft preservation fluids (perfusates) or end-stage liver disease tissuesresulted in the expansion of MSCs. Liver-derived mesenchymal stem cells (L-MSCs) were equivalent to BM-MSCs in adipogenicand osteogenic differentiation and in wingless-type-stimulated proliferative responses. Moreover, the genome-wide gene expres-sion was very similar, with a 2-fold or greater difference found in only 82 of the 32,321 genes (0.25%). L-MSC differentiation into ahepatocyte lineage was demonstrated in immunodeficient mice and in vitro by the ability to support a hepatitis C virus infection.Furthermore, a subset of engrafted MSCs survived over the long term in vivo and maintained stem cell characteristics. Like BM-MSCs, L-MSCs were found to be immunosuppressive; this was shown by significant inhibition of T cell proliferation. In conclusion,the adult human liver contains an MSC population with a regenerative and immunoregulatory capacity that can potentially contrib-ute to tissue repair and immunomodulation after liver transplantation. Liver Transpl 17:596-609, 2011. VC 2011 AASLD.

Additional Supporting Information may be found in the online version of this article.

Abbreviations: BM, bone marrow; BM-MSC, bone marrow–derived mesenchymal stem cell; c-Met, mesenchymal epithelialtransition factor; CCl4, carbon tetrachloride; cDNA, complementary DNA; CFSE, carboxyfluorescein diacetate succinimidyl ester;CK, cytokeratin; CyB, cytochrome B; DMEM, Dulbecco’s modified Eagle’s medium; FITC, fluorescein isothiocyanate; Flk-1, fetalliver kinase 1; G5, generation 5; G6, generation 6; G7, generation 7; G8, generation 8; G9, generation 9; GAPDH, glyceraldehyde3-phosphate dehydrogenase; HCV, hepatitis C virus; HGF, hepatocyte growth factor; HLA, human leukocyte antigen; IRES,internal ribosome entry site; JFH1, Japanese fulminant hepatitis 1; L-CM, L cell conditioned medium; L-MSC, liver-derivedmesenchymal stem cell; Lgr5, leucine-rich repeat-containing G protein-coupled receptor 5; mRNA, messenger RNA; MSC,mesenchymal stem cell; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; ND, not detectable; NOD, nonobesediabetic; OD490, optical density at 490 nm; PBMC, peripheral blood mononuclear cell; PHA, phytohemagglutinin; PI, proliferationindex; RT-PCR, reverse-transcription polymerase chain reaction; Sca-1, stem cell antigen 1; SCID, severe combinedimmunodeficient; TAZ, tafazzin; UW, University of Wisconsin; Wnt, wingless-type; Wnt3a-CM, Wnt3a conditioned medium.

Qiuwei Pan designed and performed the experiments, analyzed the data, and wrote the manuscript. Suomi M. G. Fouraschen,Fatima S. F. Aerts Kaya, Monique M. Verstegen, and Antoine van der Sloot performed the research and edited the manuscript.Mario Pescatori, Wilfred van IJcken, and Andrew P. Stubbs performed the gene array analysis. Ron Smits provided the reagentsand edited the manuscript. Jaap Kwekkeboom, Herold J. Metselaar, Geert Kazemier, Jeroen de Jonge, Hugo W. Tilanus, GerardWagemaker, and Harry L. A. Janssen provided clinical input and samples and edited the manuscript. Luc J. W. van der Laandesigned and supervised the research, analyzed the data, and wrote the manuscript.

This study was supported financially by the Erasmus MC Translational Research Fund and the Liver Research Foundation(Rotterdam, the Netherlands). The authors thank ZonMw for supporting the In Vivo Imaging System (an animal imaging system)through project 91104017.

Address reprint requests to Luc J. W. van der Laan, Ph.D., Laboratory of Experimental Transplantation and Intestinal Surgery, Department ofSurgery, Erasmus MC–University Medical Center Rotterdam, Room L458, sGravendijkwal 230, 3015 CE Rotterdam, the Netherlands.Telephone: þ31 10 4632759; FAX: þ31 10 4632793; E-mail: [email protected]

DOI 10.1002/lt.22260View this article online at wileyonlinelibrary.com.LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

LIVER TRANSPLANTATION 17:596-609, 2011

VC 2011 American Association for the Study of Liver Diseases.

Received June 10, 2010; accepted December 22, 2010.

The adult liver harbors a population of facultativeprogenitors (oval cells in rodents and hepatic progen-itor cells or hepatoblasts in humans) that respond tospecific injuries and can differentiate into hepato-cytes and biliary cholangiocytes.1,2 These liver pro-genitor cells are quiescent in the healthy liver but areactivated when certain liver diseases impair the re-generative capacity of mature hepatocytes, cholan-giocytes, or both.3 However, oval cells/hepatic pro-genitors do not constitute a homogeneouspopulation, and their precise origin and the signalsgoverning their activation are not entirely clear.4,5

Previous studies have indicated the presence of astem cell niche at the proximal biliary tree (thecanals of Hering) that contains hepatic stem cellsserving as precursors to hepatic progenitor cells.6,7

Recent studies have further characterized these he-patic stem cells, which are abundant in human fetaland adult livers and have been proposed to be pre-cursors of hepatic progenitors.8 This population islocated in ductal plates in fetal and neonatal liversand in the canals of Hering in pediatric and adultlivers.9

The mesenchymal stem cell (MSC) is one of a fewcell types on the brink of being used clinically in dif-ferent areas of therapeutic application, includingorgan transplantation.10 The bone marrow (BM) com-partment harbors resident MSCs with multilineagedifferentiation potential and anti-inflammatory andimmunomodulatory properties, which have been pro-posed to play a role in the response to liverinjury.11,12 Encouragingly, various studies have dem-onstrated the therapeutic potential of bone marrow–derived mesenchymal stem cells (BM-MSCs) in differ-ent liver disease models, such as liver resection, ful-minant hepatic failure, and, in particular, liver trans-plantation.13-15 Besides their hepatic differentiationpotential, MSCs produce trophic factors that havebeen shown to provide paracrine support for hepato-cyte proliferation, angiogenesis, tissue repair, andimmunomodulation.16-18

In contrast to the experimental significance ofBM-MSCs in liver injury responses and in contrastto the reports on the presence of MSCs in fetalhuman livers and on the presence of mesenchymal-like stem cells in adult rat livers,19-22 sufficientstudies describing MSCs in the adult human liverare lacking. In this study, we first investigated thepresence of MSCs in adult human liver tissue andtheir mobilization during graft cold storage at thetime of liver transplantation. Secondly, phenotypicand functional analyses were performed to evaluatethe biological characteristics and therapeuticpotential of isolated liver-resident MSCs. Gene arrayanalysis revealed a high degree of similarity betweengene expression profiles of BM-MSCs and liver-

derived mesenchymal stem cells (L-MSCs). Further-more, Wnt responsiveness and hepatic differentia-tion in vitro and in mice confirm that L-MSCsrepresent a bona fide stem cell/progenitor popula-tion in the adult human liver.

MATERIALS AND METHODS

Isolation and Culture of MSCs From Liver

Tissue and Liver Perfusate Solutions

End-stage liver disease tissue samples were obtainedfrom the explanted livers of liver transplant recipients.Patient and liver tissue characteristics are shown inSupporting Table 1.

Liver graft preservation fluids (perfusates) were col-lected from human liver grafts at the time of trans-plantation (Supporting Table 2), as described previ-ously.23 Cells were cultured in Dulbecco’s modifiedEagle’s medium (DMEM; Lonza, Verviers, Belgium)supplemented with 10% fetal bovine serum (Hyclone,Logan, UT), 100 IU/mL penicillin, and 100 lg/mLstreptomycin. The use of liver tissues and perfusateswas approved by the medical ethics committee ofErasmus MC.

Flow Cytometry

MSCs were stained for 30 minutes at 4�C with directlylabeled mouse monoclonal antibodies directed againstCD90 and CD105 (R&D Systems, Abingdon, UnitedKingdom), CD34 (Miltenyi Biotec, Bergisch Gladbach,Germany), CD45 (Beckman Coulter, Inc., Fullerton,CA), human leukocyte antigen DR (HLA-DR; BD Phar-mingen, San Diego, CA), and CD166 (BD Pharmin-gen). Flow cytometry analysis was performed withFACSCalibur and CellQuest Pro software (BD Bio-sciences, San Jose, CA).

Adipogenic and Osteogenic Differentiation

For adipogenic differentiation, MSCs were cultured inDMEM supplemented with 10% fetal bovine serum, 1lM dexamethasone, 500 lM isobutyl methyl xanthine,5 lg/mL insulin, and 60 lM indomethacin (Sigma-Aldrich) for 3 weeks. Oil Red O staining (Sigma-Aldrich) was used for the detection of adipocytes. Forosteogenic differentiation, cells were cultured inDMEM with 10% fetal bovine serum supplementedwith 0.2 mM ascorbic acid, 100 nM dexamethasone,and 10 mM b-glycerol phosphate (Sigma-Aldrich) for 3weeks. Alizarin Red S staining (Sigma-Aldrich) wasperformed to detect deposited calcium phosphates.

Cell Proliferation Assays

L-MSCs (5 � 103) were plated onto 96-well plates andwere treated with a Wnt3a conditioned medium

LIVER TRANSPLANTATION, Vol. 17, No. 5, 2011 PAN ET AL. 597

(Wnt3a-CM) and a control L cell conditioned medium(L-CM), as described previously.24 At the indicatedtimes, the number of metabolically active cells wasquantified with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (0.5 mg/mL).

L-MSCs (5 � 104) were stained with 0.2 lM carboxy-fluorescein diacetate succinimidyl ester (CFSE; Molec-ular Probes, Leiden, the Netherlands) for 5 minutes at37�C and were plated onto 6-well plates. At differenttime points, cells were harvested and stained with 7-amino-actinomycin D (BD Pharmingen) and weremeasured with flow cytometry. Generation analysiswas performed with ModFit LT version 3.0 software(Verity Software House, Topsham, ME) and was gatedto exclude 7-amino-actinomycin D–positive dead cells.The proliferation index (PI), which is the sum of thecells in all generations divided by the computed num-ber of original parent cells, was used to indicate theextent of cell proliferation.

Gene Expression Profiling by Microarray

The total RNA of 3 independent L-MSC cultures atpassages 2 to 5 (1 culture derived from a liver tissuebiopsy sample and 2 cultures derived from liver perfu-sates), 3 BM-MSC cultures at passages 2 to 5 (fromdifferent donors), and 3 hepatoma cell line (Huh7) cul-tures was used for genome-wide microarray analysiswith the Affymetrix GeneChip HuGene 1.0 ST.v1 array(Affymetrix, Santa Clara, CA) according to the manu-facturer’s procedures. Transcript-level expressionmeasures were generated with the robust multi-arrayaverage procedure as implemented in the AffymetrixGene Expression Console, and probe set annotationswere retrieved from NetAffx with the same software.Probe sets that differentially expressed between condi-tions were identified with the class comparison toolimplemented in BRB-ArrayTools (National Institutesof Health, Bethesda, MD). Principal component analy-sis was performed with Partek (Partek, Inc., SaintLouis, MO). Hierarchical clustering was performedwith Spotfire (Spotfire, Inc., Somerville, MA). Pathwayanalysis was performed with Ingenuity Pathway Anal-ysis (Ingenuity Systems, Redwood City, CA).

In Vitro Hepatogenic Differentiation

In vitro hepatic differentiation was performed in 3steps, as reported previously,25 but the final step wasmodified. For the final maturation step, cultures wereincubated with infectious Japanese fulminant hepati-tis 1 (JFH1)–derived hepatitis C virus (HCV) par-ticles.26 Hepatogenic differentiation was determinedby quantitative reverse-transcription polymerasechain reaction (RT-PCR) detection of albumin andHCV internal ribosome entry site (IRES) RNA.

Real-Time RT-PCR

Confluent monolayers of MSCs or liver graft tissue bi-opsy samples were lysed with TRIzol (Invitrogen-Gibco), and RNA was precipitated with 75% ethanoland captured with a Micro RNeasy silica column (Qia-gen, Venlo, the Netherlands). RNA was quantified witha NanoDrop ND-1000 spectrophotometer (NanoDropTechnologies, Wilmington, DE). Complementary DNA(cDNA) was prepared from 1 lg of total RNA with aniScript cDNA synthesis kit from Bio-Rad Laboratories(Stanford, CA). The cDNA of human and mouse albu-min, CD90, CD105, cytokeratin 18 (CK18), CK19,hepatocyte growth factor (HGF), mesenchymal epithe-lial transition factor (c-Met), leucine-rich repeat-con-taining G protein-coupled receptor 5 (Lgr5), HCVIRES, cytochrome B (CyB), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was quantifiedwith real-time polymerase chain reaction (MJResearch Opticon, Hercules, CA), which was per-formed with SYBR Green (Sigma-Aldrich) according tothe manufacturer’s instructions. CyB or GAPDH wasused as a reference gene to normalize the geneexpression, which was calculated with the delta-deltacycle threshold method.

Periodic Acid–Schiff Staining

Hepatically differentiated and undifferentiated MSCswere fixed in 4% paraformaldehyde for 20 minutes,and then intracellular glycogen was stained with theperiodic acid–Schiff method. Briefly, the fixed cellswere oxidized in a 0.5% periodic acid solution for 5minutes. After they were rinsed in distilled water, theywere placed in the Schiff reagent for 15 minutes.Then, they were washed in lukewarm tap water for 5minutes and counterstained with Mayer’s hematoxylinfor 1 minute.

MSC Transplantation in Mice

Immunodeficient nonobese diabetic (NOD)/severecombined immunodeficient (SCID) mice (from Eras-mus MC institutional breeding) that were 6 to 8 weeksold were intraperitoneally injected with 100 lL of oliveoil [containing 10 lL of carbon tetrachloride (CCl4)]per 20 g of body weight. After 24 hours, 1 � 106 L-MSCs (n ¼ 5) or BM-MSCs (n ¼ 2) suspended in 0.2mL of phosphate-buffered saline were injected intothe spleen. Untreated (n ¼ 2) or nonengrafted CCL4-treated NOD/SCID mice (n ¼ 3) served as negativecontrols. Four weeks after engraftment, the mice weresacrificed, and their livers were removed. Second, lu-ciferase-labeled L-MSCs (5 � 105 cells) were subcuta-neously injected into NOD/SCID mice (n ¼ 2), and theluciferase activity was measured at the indicated timepoints with the In Vivo Imaging System camera.

Fluorescent Immunohistochemistry

Mouse liver tissue was dissected and cryoprotected in30% sucrose for the generation of frozen sections. The

598 PAN ET AL. LIVER TRANSPLANTATION, May 2011

sections were incubated with fluorescent-labeled anti-bodies at the dilution of 1:100 for 30 minutes. After 3washes, nuclear staining was achieved by incubationwith 40,6-diamidino-2-phenylindole (Sigma-Aldrich) atthe dilution of 1:50 for 5 minutes; 20 to 30 images ofeach condition were captured by confocal microscopy.

T Cell Proliferation/Suppression Assay

The effect of L-MSCs on the proliferation of T cellswas determined with the mixed lymphocyte responseassay. Briefly, peripheral blood mononuclear cells(PBMCs; 4 � 105) in the presence or absence of MSCswere seeded onto 96-well, round-bottom plates. Irra-diated allogeneic PBMCs (2 � 105) or phytohemagglu-tinin (PHA; Murex Biotech, United Kingdom) was usedfor stimulation. After 5 days, proliferation wasassessed by the determination of the incorporation of0.5 lCi (0.0185 MBq) of [3H]-thymidine (Radiochemi-cal Center, Amersham, United Kingdom) for 18 hours.

Statistical Analysis

Statistical analysis was performed with either thepaired nonparametric test (the Wilcoxon signed-ranktest) or the unpaired nonparametric test (the Mann-Whitney test) with GraphPad InStat software (Graph-Pad Software, Inc., San Diego, CA). P values lowerthan 0.05 were considered statistically significant.

RESULTS

Mobilization of Hepatic MSCs From Adult

Human Liver Grafts

Graft perfusion, procurement, and cold storage areassociated with ischemia and tissue injury. Previ-ously, we have found that the washout of the graftpreservation solution (perfusate) collected at the timeof liver transplantation contains high numbers ofmononuclear cells that detach from the liver; theseinclude lymphocytes, natural killer cells, antigen-pre-senting cells,23,27 and hematopoietic stem cells.28

Flow cytometry analysis of perfusate mononuclearcells revealed the presence of a small but consistentfraction of cells double-positive for MSC surfacemarkers CD90 and CD105 (0.09% 6 0.07%, n ¼ 8)and CD90 and CD166 (0.02% 6 0.02%; Fig. 1A). Pro-spectively, fresh perfusates from 15 consecutive livertransplants were collected, and the mononuclear cellswere isolated and cultured for the presence of MSCs(Supporting Table 1). Fibroblast-like cells wereobserved in the initial cultures of all perfusates (Fig.1B). In a majority of the cultures, the number of thesecells rapidly increased (Fig. 1C). These cells could beexpanded and passaged for several months undernormal nonhypoxic culture conditions; they wereclearly distinct from a previously described albu-minþCD105� population of hepatic stem cells.29 Flowcytometry analysis of expanded cells at passages 4 to9 revealed a surface marker profile typical for MSCs(Fig. 1D). A high percentage of cells stained positive

for CD90 (59% 6 18%, n ¼ 11), CD105 (55% 6 14%),and CD166 (44% 6 16%), and they were mostly nega-tive for the hematopoietic stem cell marker CD34(0.8% 6 0.7%) and the leukocyte lineage markersCD45 (0.7% 6 0.7%) and HLA-DR (1.9% 6 1%). Afunctional analysis showed that the expanded liver-derived cells had a multilineage potential with acapacity for adipogenic (Fig. 1E) and osteogenic differ-entiation (Fig. 1F) that was similar to the capacity ofBM-MSCs.

To confirm the presence of MSCs in the adulthuman liver, tissue samples of explant livers from avariety of patients with end-stage liver disease (Sup-porting Table 2) were dissociated, and the unfractio-nated cell suspensions were cultured. After 4 to 10days, in a majority of the cultures, clusters of cellswith a fibroblast-like morphology were observed. LikeMSCs from perfusates, these cells were highly positivefor CD90, CD105, and CD166, were negative for themarkers CD34, CD45, and HLA-DR, and had equiva-lent adipogenic and osteogenic differentiation capacity(not shown). L-MSCs could also be expanded fromdisease-free liver graft tissue obtained from postmor-tem organ donors (data not shown). Cultures ofPBMCs from end-stage liver disease patients, brain-dead multiorgan donors, and healthy controls did notshow any MSCs. Therefore, it is unlikely that L-MSCsin perfusates are directly mobilized from the BM com-partment and derived from residual donor blood.

Wnt Signaling Promotes L-MSC Proliferation

Wnt signaling has been shown to modulate the growthof human BM-MSCs30 and plays an important role inliver homeostasis and pathology.31 As shown in Fig.2A, L-MSCs (from both tissue and perfusate) thatwere stimulated with Wnt3a exhibited a significantincrease in viable (metabolically active) cells in com-parison with control-treated MSCs as measured bythe MTT assay (mean increase ¼ 44% 6 17% on day6, P < 0.001). In order to confirm that this increasewas related to enhanced cell proliferation and was notdue to enhanced cell survival, a CFSE fluorescence-based proliferation assay was used. As shown in Fig.2B, the Wnt3a-CM treatment accelerated CFSE dilu-tion of labeled L-MSCs, and this was indicative ofenhanced cell proliferation. On culture day 6, the per-centage of cells that underwent 8 rounds of cell divi-sion (ninth generation) was 10% for L-CM–treatedcells and 50% for Wnt3a-CM–treated cells. Similarproliferative responses were seen with BM-MSCs (notshown).

Gene Expression Profiles of L-MSCs and

BM-MSCs

In order to gain further insight into the molecularphenotype of L-MSCs, we performed genome-wideexpression profiling on early-passage MSC cultures(passages 2-5). The expression profile of these L-MSCs was compared to that of BM-MSCs (3 cultures


Figure 1. Characterization of MSCs mobilized from the human liver at the time of graft cold storage. (A) The percentage of cellsdouble-positive for CD90þCD105þ (0.09% 6 0.07%, n ¼ 8) and CD90þCD166þ (0.02% 6 0.02%, n ¼ 8) was low in the liver perfusatesbefore culturing (Pre), but the cells rapidly expanded with culturing (44% 6 8% and 37% 6 6%, respectively) at passages 4 to 9 (Post;n ¼ 6, P < 0.001). (B) In the majority of the cultures, cells with a fibroblast-like morphology appeared within 10 days. (C) Fibroblast-like cells rapidly proliferated and could be subcultured and expanded for 10 to 20 passages. (D) Flow cytometry analysis of surfacemarkers showed that the expanded cells exhibited a typical MSC-like phenotype positive for CD90 (59% 6 18%, n ¼ 11), CD105 (55%6 14%), and CD166 (44% 6 16%), and they were mostly negative for the hematopoietic stem cell marker CD34 (0.8% 6 0.7%) and forthe leukocyte lineage markers CD45 (0.7% 6 0.7%) and HLA-DR (1.9% 6 1%). The black lines in the histograms represent the specificstaining, and the gray lines show the background staining of an isotype-matched control antibody. (E) Adipogenic differentiation of L-MSCs was detected by Oil Red O staining for lipid droplets (red). (F) Osteogenic differentiation of these cells was evaluated through thedetection of deposited calcium phosphates with Alizarin Red S staining (red). Representative stains of 4 independent cultures areshown (magnification �100). Similar adipogenic and osteogenic differentiation was observed with BM-MSCs and MSCs obtained fromliver biopsy samples (data not shown).


at passages 2-5 from different donors). Cultures ofHuh7 hepatoma cells (n ¼ 3) served as generic con-trols of a replicating cell population of hepatic origin.The principal component analysis of their genome-wide expression profiles grouped the 3 cell types into3 separate clusters on a 3-dimensional scatter plot(Fig. 3A). Notably, liver-derived Huh7 cells clusteredfar apart from both L-MSCs and BM-MSCs, regardlessof their hepatic or extrahepatic origin. Accordingly, incomparison with L-MSCs, more than 20% of theHuh7 genes were differentially expressed. A direct

comparison of the gene expression of liver tissuesobtained from grafts at the time of transplantationshowed that L-MSCs highly expressed CK19 andHGF, whereas the expression of CK18, c-Met, andLgr5 was lower, and albumin messenger RNA (mRNA)was not detectable (ND; Fig. 3B). A comparative analy-sis of L-MSCs and BM-MSCs showed comparableexpression levels of most of the known MSC-associ-ated genes (Fig. 3C).32 However, with identical analy-sis settings, less than 1% of the genes were differen-tially expressed between L-MSCs and BM-MSCs (311

Figure 2. Wnt3a promotes L-MSCproliferation. (A) The MTT assayshowed that the stimulation ofMSCs with Wnt3a-CM significantlyincreased the number of cells incomparison with the control L-CMtreatment on days 2 (16% 6 6%increase), 4 (20% 6 3%), and 6(44% 6 17%). The means andstandard deviations of 6independent experiments areshown (*P < 0.001). (B) IncreasedWnt3a-induced cell proliferationwas confirmed by CFSE dilutionof labeled L-MSCs. A markedincrease in the percentage of cellswas observed in G5 and G6 on day3 (47% for L-CM versus 66% forWnt3a-CM) and in G8 and G9 onday 6 (19% for L-CM versus 50%for Wnt3a-CM); this was alsoreflected in the marked increase inthe PI seen in the Wnt3a-stimulated cells at both time points(11.56 and 72.03 for Wnt3a-CMversus 8.63 and 32.05 for L-CM).


Figure 3. Gene expression profiling of L-MSCs, BM-MSCs, and Huh7 hepatoma cells. (A) Principal component analysis of genome-wideexpression profiles was used to visualize correlation relationships between samples. The 3 independent L-MSC preparations clusteredinto a separate grouping apart from both the BM-MSCs and the Huh7 cells. The short distance between the L-MSC and BM-MSCclusters in the 3-dimensional correlation space suggests that the 2 MSC populations were characterized by similar patterns of geneexpression at variance with Huh7. (B) Real-time RT-PCR analysis of the L-MSC gene expression levels relative to GAPDH and normalizedto levels in the donor graft liver tissue. MSCs highly expressed CK19 and HGF, whereas the expression of CK18, c-Met, Lgr5, andalbumin was lower or ND. The means and standard deviations of 1 representative experiment in triplicate are shown. (C) A gene arrayanalysis of L-MSCs and BM-MSCs showed comparable gene expression of known MSC markers, whereas Huh7 generally exhibited lowexpression of these genes. (D) Although the expression profiles of the 2 cell types appeared very similar, L-MSCs and BM-MSCs could bedistinguished by the consistent differential expression of a small proportion of their transcriptomes (<1%, P < 0.001).


of 32,321 genes, P < 0.001; Fig. 3D). Overall, theexpression of only 45 genes was more than 2-foldhigher in L-MSCs, and the expression of 37 genes wasmore than 2-fold higher in BM-MSCs (see SupportingTable 3). Among the genes differentially expressed inMSCs were matrix metallopeptidase 1, actin gamma 2smooth muscle enteric, and interleukin-33 (>20-foldhigher in L-MSCs) and microfibrillar associated pro-tein 5, insulin-like growth factor binding protein 3,and retinol binding protein 4 plasma (>10-fold higherin BM-MSCs, P < 0.001). However, no significant dif-ferences in biological or functional pathways of geneexpression were observed between L-MSCs and BM-MSCs with Ingenuity Pathway Analysis (data notshown). To distinguish L-MSCs from hepatic stellatecells, a gene array analysis was performed for 10known stellate cell–associated genes and 2 recentlyreported markers, CD133 and octamer 4.33,34 The dif-ferential gene expression of these 12 markers wascomparable between L-MSCs and BM-MSCs (data notshown). Together, these data indicate that in terms oftranscriptome composition, L-MSCs are very similarto BM-MSCs and, like BM-MSCs, appear to be distinctfrom hepatic stellate cells.

Long-Term Survival of L-MSCs In Vivo

To evaluate the fate of MSCs in vivo, L-MSCs or BM-MSCs were engrafted into the livers of NOD/SCID micesubjected to CCl4-induced liver toxicity. As shown inFig. 4A, RT-PCR analysis at 4 weeks showed theexpression of the human MSC markers CD90 andCD105 in the livers of MSC-engrafted mice but not inthe livers of sham-treated controls. Fluorescent immu-nohistochemistry confirmed the presence of humanCD90-positive cells in the mouse liver tissue (Fig. 4B),although these cells did not express the cell prolifera-tion marker proliferating cell nuclear antigen. Dissoci-ated mouse liver tissue was cultured for 5 days.Human MSCs with a typical fibroblast-like morphologyrapidly expanded in culture (Fig. 4C). No such cellswere observed in control mice not engrafted withhuman MSCs (Fig. 4D). Flow cytometry analysisrevealed that the majority of the explanted liver cells,after culture expansion, expressed human CD90(60.8% 6 18.2%; Fig. 4E) and CD105 (61.6% 6 20.8%,n ¼ 3; Fig. 4F). This was further confirmed by RT-PCRanalysis showing human-specific CD90 and CD105gene expression in cultured cells. No CD90 or CD105mRNA was detected in cultures of sham-treated con-trols (data not shown). For further evaluation of the invivo survival, luciferase-labeled L-MSCs were subcuta-neously engrafted into NOD/SCID mice. As shown inFig. 4G, a luciferase signal was clearly visible up to 25days after engraftment; this confirmed longer termMSC survival in vivo, although the signal graduallydeclined over time.

Hepatic Differentiation of L-MSCs

In Vitro and In Vivo

The hepatocyte lineage differentiation potential of L-MSCs was determined in vitro with an establishedhepatogenic culture procedure. After 30 days of cul-ture, morphological changes in most cells wereobserved in all cultures (n ¼ 6). In contrast to thefibroblast-like morphology of undifferentiated MSCs(Fig. 5A), hepatogenic differentiation induced a polyg-onal morphology at 15 (Fig. 5B) and 25 days of cul-ture (Fig. 5C). Fluorescent immunohistochemistrystaining showed the expression of human albuminprotein with differentiated MSCs (Fig. 5D) but notwith undifferentiated MSCs (Fig. 5E). Similarly, glyco-gen storage was detected in differentiated MSCs (Fig.5F) but not in undifferentiated MSCs (Fig. 5G). Quan-titative RT-PCR analyses confirmed the specificexpression of albumin mRNA in differentiated BM-MSCs and L-MSCs (Fig. 5H).

To further investigate the functionality of MSC-derived, hepatocyte-like cells, we challenged differen-tiated and undifferentiated L-MSCs with infectiousHCV particles (JFH1-derived, cell culture–producedHCV); this virus has a hepatic tropism produced byHuh7.5 cells. RT-PCR analyses of the HCV-specificIRES sequence clearly showed that, unlike undifferen-tiated MSCs, MSC-derived, hepatocyte-like cells per-mitted HCV infection (Fig. 5I). The levels of HCV RNAwere approximately 100-fold lower than those of high-replicating HCV replicon cells.

The hepatogenic differentiation potential of L-MSCswas further evaluated in vivo in NOD/SCID mice withCCl4-induced liver injury. Four weeks after MSCengraftment, mouse livers were harvested, and RT-PCR analyses displayed detectable levels of human-specific albumin gene expression (Fig. 6A). Immunohis-tochemical staining of mouse liver tissue showed thepresence of human albumin–positive cell clusters in allMSC-engrafted mice (Fig. 6B), although the frequencyof these clusters was generally low. No human albuminpositivity was observed in untreated or sham CCL4-treated control mice (Fig. 6D). Flow cytometry analysesof dissociated mouse livers confirmed the presence ofhuman hepatocyte–like cells (Fig. 6E) with a mean con-centration of albumin-positive cells of 1.09% 6 0.39%(n ¼ 5). No positive cells were detected in livers fromcontrol mice (Fig. 6F). Overall, these results indicatethat L-MSCs (or a subset) have hepatogenic potentialin vitro and in vivo and that MSC-derived, hepatocyte-like cells can be infected with HCV.

MSCs Effectively Suppress T Cell Proliferation

It is well established that MSC populations from vari-ous tissues, including the BM, spleen, heart, and fat,have immunoregulatory and suppressive properties,including the inhibition of T cell proliferation.16,35 Theeffect of adult human L-MSCs on the proliferation ofmitogen- and alloantigen-stimulated T cells was inves-tigated in vitro through the measurement of [3H]-


Figure 4. Human L-MSCs retain stem cellcharacteristics after their engraftment into mice.L-MSCs or BM-MSCs were engrafted into thelivers of NOD/SCID mice subjected to CCL4-induced liver injury. Four weeks after the MSCadministration, mouse liver tissue was analyzedfor the presence of human MSCs. (A) RT-PCRanalysis showed the expression of the human-specific MSC markers CD90 and CD105 in theliver tissue from transplanted mice but not in thetissue from sham-treated controls (�). (B)Immunofluorescent staining of mouse liver tissueshowed human-specific CD90-positive cells (red)present in the MSC-engrafted mice but not in thecontrol mice (not shown). 40,6-Diamidino-2-phenylindole nuclear staining is shown in blue(bar ¼ 5 lm). (C) In a majority of the cultures ofdissociated mouse liver cells, MSC-like cells with atypical fibroblast-like morphology rapidlyexpanded (magnification �400). (D) No such cellswere observed in cultures of control mouse livers(magnification �400). (E,F) Flow cytometryanalysis of engrafted mice confirmed that a highpercentage of the cells in the culture were positivefor human MSC surface markers CD90 (60.8% 618%) and CD105 (61.6% 6 21%), respectively.Gray lines indicate isotype-matched controlstaining. (G) Survival of subcutaneously engraftedL-MSCs in NOD/SCID mice. MSCs expressed theluciferase reporter gene, and the luciferase signalwas measured at different time points afterengraftment. The luciferase signal graduallydeclined over time but was clearly detectable forat least 25 days, and this confirmed the long-termsurvival of viable MSCs in vivo.

Figure 5. L-MSCs differentiate into hepatocyte-like cells that permit HCV infection. (A) In contrast to the fibroblast-like spreading ofundifferentiated L-MSCs, (B,C) hepatogenic differentiation of L-MSCs induced a polygonal morphology with granular cytoplasm at 15and 25 days of differentiation, respectively (magnification �400). Fluorescent immunocytochemistry showed human albumin staining(green) (D) in hepatically differentiated MSCs on day 30 but not (E) in undifferentiated MSCs. Glycogen storage (pink) was seen (F) inhepatically differentiated MSCs but not (G) in undifferentiated MSCs (magnification �100). (H) A gene expression analysis of culturedMSCs showed clear expression of the hepatocyte-specific albumin gene after hepatogenic differentiation (MSC-hepa) by quantitativereal-time RT-PCR. Albumin mRNA was ND in undifferentiated BM-MSCs or L-MSCs, and the Huh7 hepatoma cell line served as apositive control. (I) For further characterization of the MSC-derived, hepatocyte-like cells, differentiated and undifferentiated cells wereincubated with a conditioned medium containing infectious HCV particles (cell culture–produced HCV). Real-time RT-PCR analysis ofthe HCV IRES sequence showed that differentiated MSCs could be infected by HCV, whereas undifferentiated BM-MSCs and L-MSCsdid not permit infection. Huh7 replicon cells (Huh7-ET) with high-level HCV replication served as positive controls.


Figure 5.


thymidine incorporation. PBMCs or purifiedCD4þCD25� T cells stimulated with mitogenic PHA orirradiated allogeneic PBMC cells were cocultured withMSCs at different ratios. As shown in Fig. 7A, signifi-cant inhibition of alloantigen-stimulated proliferationwas observed with MSC/PBMC ratios of 1:2 (97% in-hibition, P < 0.001), 1:4 (92%, P < 0.001), 1:8 (76%, P< 0.001), and 1:16 (53%, P < 0.05, n ¼ 6). Compara-ble inhibition of proliferation was observed with PHA-stimulated PBMCs (Fig. 7B). Similar results wereobtained when BM-MSCs were used and when puri-fied CD4þCD25� T cells were used as responder cells(not shown). These findings indicate that L-MSCs, likeother MSC populations, are potent inhibitors of T cellproliferation and may contribute to alloimmune regu-lation after liver transplantation.

DISCUSSION

Currently, the therapeutic potential of stem cells suchas MSCs are being explored at an incredible pace forthe treatment of liver disease as well as many otherdiseases.4,14 In vitro hepatic differentiation has beendescribed for MSCs derived from different sources,including the BM, fat, lungs, cord blood, and amnioticfluid.25,36-38 In the current study, we found evidence

of the presence of MSCs in the adult human liveritself. Gene expression profiling showed a high degreeof similarity between L-MSCs and BM-MSCs (Fig. 3).Overall, only 82 genes (0.25% of the transcriptome)were more than 2-fold different in expression betweenL-MSCs and BM-MSCs (Supporting Table 3). Theirhigh similarity, which included the expression profil-ing of most of these MSC markers (Fig. 3C), supportsthe idea that these isolated cells constitute a bonafide MSC population. In addition, the distinct genomicsignature (Supporting Table 3) provides an indicationof their distinct origin, which in turn provides poten-tial markers for distinguishing these 2 MSCpopulations.

In order to determine the hepatic differentiation ofL-MSCs in vitro, we combined the conventional meth-ods with a novel approach of infecting the differenti-ated cells with HCV (Fig. 5). We found that hepatic dif-ferentiation occurred in a subpopulation of L-MSCsthat not only changed in their morphology andexpressed albumin but also supported HCV infection.Because HCV infection is largely restricted to maturehepatocytes, the results strongly indicate that at leastsome of the MSC-derived hepatocytes were fully

Figure 6. MSC differentiation into hepatocyte-like cells in vivo. L-MSCs or BM-MSCs were engrafted into the livers of NOD/SCID micesubjected to CCL4-induced liver toxicity. Four weeks after the MSC administration, the mouse livers were harvested and analyzed forevidence of hepatic differentiation. (A) RT-PCR analysis for human- and mouse-specific albumin RNA. Human albumin expression wasobserved in all MSC-engrafted livers (lanes 3-5) and was not observed in the control mouse liver (lane 2). Human liver RNA served as apositive control (lane 1), and mouse albumin was detected in all mouse livers (lanes 2-7) but not in the human liver. (B)Immunohistochemical staining of the mouse livers confirmed the presence of human albumin–positive cell clusters (indicated byarrows) in mice engrafted with BM-MSCs or L-MSCs (n ¼ 7). (C) Control mouse and (D) human liver tissue served as negative andpositive controls, respectively, for the albumin staining (bar ¼ 20 lm). The concentration of human-specific, albumin-positive cells, asquantified by flow cytometry, was (E) 1.09% 6 0.39% in the dissociated mouse livers (n ¼ 5) and (F) ND (<0.01%) in the sham-treatedcontrol mice (n ¼ 3).


differentiated and functionally supported viral entryand replication.

Consistently, the hepatic differentiation of MSCshas been reported in vivo as well. Several studieshave mentioned the occurrence of MSC-oriented he-patic differentiation in models of liver injury andregeneration.14,25,37,39 Although hepatic differentia-tion generally occurs at a relatively low frequency(also shown in Fig. 6) and concerns have been raisedabout the fate of stem cells after transplantation inimmune-competent hosts,40 these findings still bring

new hope for MSC-based cell therapies for regenera-tive liver diseases. Notably, the hepatic differentiationof MSCs appears to be less robust than that seen witha different, more committed population of hepaticstem cells identified from human fetal and postnatallivers.8 But the in vivo hepatic differentiation capacityof MSCs could be further improved by pre-differentia-tion in vitro before engraftment.41 If MSCs are used totreat liver diseases, both hepatic differentiation and invivo survival of the stem cells are crucial. In thisstudy, we found that L-MSCs not only can differenti-ate toward hepatocytes but also can be retrieved 4weeks after transplantation into NOD/SCID mice (Fig.4); this indicates their long-term survival properties invivo.

Notably, MSCs can migrate to injured tissue andcontribute to tissue repair and wound healing. Thismobilization is likely regulated by specific danger sig-nals and chemotactic factors.42,43 MSCs have a pro-foundly greater capacity to survive under conditionsof ischemia because in the absence of oxygen, MSCscan survive by anaerobic adenosine triphosphate pro-duction.44 Our previous studies have shown that sev-eral types of liver-derived hematopoietic cells aremobilized during perfusion of the graft and are contin-uously released into the recipient after liver trans-plantation.23,27 The continuous migration of donorleukocytes into the recipient’s circulation, which leadsto chimerism, occurs more often in liver transplanta-tion than other organ transplantation procedures andhas been associated with graft acceptance.45 Wehypothesize that ex vivo vascular ischemic perfusionof liver grafts may stimulate MSC mobilization at thetime of transplantation. Indeed, substantial numbersof L-MSCs were isolated from liver perfusates by cul-turing of the cell fraction, and this demonstrated themigration of graft MSCs during cold storage and per-fusion (Fig. 1). Thus, liver preservation fluid can beconsidered a novel source of MSCs that is particularlyimportant because normal healthy human liver tissueis usually not available. Because L-MSCs possesspotent immunomodulatory properties (Fig. 7), wespeculate that like many liver leukocytes, graft-derived MSCs may also migrate to the recipient afterliver transplantation and subsequently play a role inimmunoregulation, tissue repair, and regeneration.

In conclusion, this study provides evidence for thepresence of MSCs in the adult human liver. These cellshave characteristics similar to those of BM-MSCs withrespect to phenotype and function. The migration ofgraft MSCs at the time of liver transplantation providesan alternative source of L-MSCs, and this also suggeststhat a continuous release of graft MSCs and a systemiccontribution may occur after transplantation in a recip-ient. We believe that our observations have paved theway for further study of the role of these cells in physi-ological and particular pathological conditions.

ACKNOWLEDGMENTThe authors thank Professor Takaji Wakita (National

Figure 7. Suppression of T cell proliferation by L-MSCs. (A) Tcells were stimulated with irradiated allogeneic PBMCs and werecocultured with allogeneic MSCs in 1:2, 1:4, 1:8, and 1:16 MSC/PBMC ratios for 5 days. (B) T cells were stimulated with PHAand were cocultured with MSCs for 5 days in different ratios.Significant inhibition of cell proliferation was observed withdifferent ratios. Similar results were observed when BM-MSCswere used and when purified CD4þCD25� T cells were used asresponders (data not shown). The means and standarddeviations of 6 experiments are shown. *P < 0.05, **P < 0.01,and ***P < 0.001.


Institute of Infectious Diseases, Tokyo, Japan) for gen-erously providing the full-length, JFH1-derived infec-tious HCV replicon; Professor Ralf Bartenschlager andDr. Volker Lohmann (University of Heidelberg, Heidel-berg, Germany) for providing the Huh7 subgenomicHCV replicon cells; Dr. Thanyalak Tha-In, Dr. ScotHenry, and Dr. Wendy van Veelen for providing techni-cal support; Dr. Yuedan Li and Dr. Martijn Brugmanfor performing the gene array data analysis; and Dr.Martin Hoogduijn for critically reading the manuscript.

REFERENCES

1. Fausto N, Campbell JS. The role of hepatocytes and ovalcells in liver regeneration and repopulation. Mech Dev2003;120:117-130.

2. Forbes S, Vig P, Poulsom R, Thomas H, Alison M.Hepatic stem cells. J Pathol 2002;197:510-518.

3. Dolle L, Best J, Mei J, Al Battah F, Reynaert H, vanGrunsven LA, Geerts A. The quest for liver progenitorcells: a practical point of view. J Hepatol 2010;52:117-129.

4. Oertel M, Shafritz DA. Stem cells, cell transplantationand liver repopulation. Biochim Biophys Acta 2008;1782:61-74.

5. Dorrell C, Erker L, Lanxon-Cookson KM, Abraham SL,Victoroff T, Ro S, et al. Surface markers for the murineoval cell response. HEPATOLOGY 2008;48:1282-1291.

6. Paku S, Schnur J, Nagy P, Thorgeirsson SS. Origin andstructural evolution of the early proliferating oval cells inrat liver. Am J Pathol 2001;158:1313-1323.

7. Kuwahara R, Kofman AV, Landis CS, Swenson ES, Bare-ndswaard E, Theise ND. The hepatic stem cell niche:identification by label-retaining cell assay. HEPATOLOGY

2008;47:1994-2002.

8. Schmelzer E, Zhang L, Bruce A, Wauthier E, Ludlow J,Yao HL, et al. Human hepatic stem cells from fetal andpostnatal donors. J Exp Med 2007;204:1973-1987.

9. Zhang L, Theise N, Chua M, Reid LM. The stem cellniche of human livers: symmetry between developmentand regeneration. HEPATOLOGY 2008;48:1598-1607.

10. Reinders ME, Fibbe WE, Rabelink TJ. Multipotent mes-enchymal stromal cell therapy in renal disease and kid-ney transplantation. Nephrol Dial Transplant 2010;25:17-24.

11. Petersen BE, Bowen WC, Patrene KD, Mars WM, SullivanAK, Murase N, et al. Bone marrow as a potential sourceof hepatic oval cells. Science 1999;284:1168-1170.

12. Bird TG, Lorenzini S, Forbes SJ. Activation of stem cellsin hepatic diseases. Cell Tissue Res 2008;331:283-300.

13. Inoue S, Popp FC, Koehl GE, Piso P, Schlitt HJ, GeisslerEK, Dahlke MH. Immunomodulatory effects of mesen-chymal stem cells in a rat organ transplant model.Transplantation 2006;81:1589-1595.

14. Kuo TK, Hung SP, Chuang CH, Chen CT, Shih YR, FangSC, et al. Stem cell therapy for liver disease: parametersgoverning the success of using bone marrow mesenchy-mal stem cells. Gastroenterology 2008;134:2111-2121.

15. Parekkadan B, van Poll D, Suganuma K, Carter EA, Ber-thiaume F, Tilles AW, Yarmush ML. Mesenchymal stemcell-derived molecules reverse fulminant hepatic failure.PLoS One 2007;2:e941.

16. Uccelli A, Moretta L, Pistoia V. Mesenchymal stem cellsin health and disease. Nat Rev Immunol 2008;8:726-736.

17. Yan Y, Xu W, Qian H, Si Y, Zhu W, Cao H, et al. Mesen-chymal stem cells from human umbilical cords amelio-rate mouse hepatic injury in vivo. Liver Int 2009;29:

356-365.

18. van Poll D, Parekkadan B, Cho CH, Berthiaume F, Nah-mias Y, Tilles AW, Yarmush ML. Mesenchymal stem cell-derived molecules directly modulate hepatocellular deathand regeneration in vitro and in vivo. HEPATOLOGY 2008;47:1634-1643.

19. Campagnoli C, Roberts IA, Kumar S, Bennett PR, Bellan-tuono I, Fisk NM. Identification of mesenchymal stem/progenitor cells in human first-trimester fetal blood,liver, and bone marrow. Blood 2001;98:2396-2402.

20. O’Donoghue K, Chan J. Human fetal mesenchymal stemcells. Curr Stem Cell Res Ther 2006;1:371-386.

21. Gotherstrom C, West A, Liden J, Uzunel M, Lahesmaa R,Le Blanc K. Difference in gene expression betweenhuman fetal liver and adult bone marrow mesenchymalstem cells. Haematologica 2005;90:1017-1026.

22. Tarnowski M, Koryciak-Komarska H, Czekaj P, SebestaR, Czekaj TM, Urbanek K, et al. The comparison of mul-tipotential for differentiation of progenitor mesenchymal-like stem cells obtained from livers of young and oldrats. Folia Histochem Cytobiol 2007;45:245-254.

23. Demirkiran A, Bosma BM, Kok A, Baan CC, MetselaarHJ, Ijzermans JN, et al. Allosuppressive donorCD4þCD25þ regulatory T cells detach from the graftand circulate in recipients after liver transplantation.J Immunol 2007;178:6066-6072.

24. Shibamoto S, Higano K, Takada R, Ito F, Takeichi M,Takada S. Cytoskeletal reorganization by soluble Wnt-3aprotein signalling. Genes Cells 1998;3:659-670.

25. Campard D, Lysy PA, Najimi M, Sokal EM. Native umbil-ical cord matrix stem cells express hepatic markers anddifferentiate into hepatocyte-like cells. Gastroenterology2008;134:833-848.

26. Pan Q, Henry SD, Metselaar HJ, Scholte B, KwekkeboomJ, Tilanus HW, et al. Combined antiviral activity of inter-feron-alpha and RNA interference directed against hepa-titis C without affecting vector delivery and genesilencing. J Mol Med 2009;87:713-722.

27. Bosma BM, Metselaar HJ, Mancham S, Boor PP, KustersJG, Kazemier G, et al. Characterization of human liverdendritic cells in liver grafts and perfusates. LiverTranspl 2006;12:384-393.

28. Pan Q, Aerts-Kaya F, Kazemier G, Kwekkeboom J, Met-selaar HJ, Tilanus HW, et al. Human liver hematopoieticstem cells have a multi-potent and self-renewal capacityand may function as local progenitors for liver residentleukocytes. Liver Transpl 2009;15:S99.

29. Herrera MB, Bruno S, Buttiglieri S, Tetta C, Gatti S,Deregibus MC, et al. Isolation and characterization of astem cell population from adult human liver. Stem Cells2006;24:2840-2850.

30. De Boer J, Wang HJ, Van Blitterswijk C. Effects of Wntsignaling on proliferation and differentiation of humanmesenchymal stem cells. Tissue Eng 2004;10:393-401.

31. Thompson MD, Monga SP. WNT/beta-catenin signalingin liver health and disease. HEPATOLOGY 2007;45:1298-1305.

32. da Silva Meirelles L, Caplan AI, Nardi NB. In search ofthe in vivo identity of mesenchymal stem cells. StemCells 2008;26:2287-2299.

33. Van Rossen E, Vander Borght S, van Grunsven LA, Rey-naert H, Bruggeman V, Blomhoff R, et al. Vinculin andcellular retinol-binding protein-1 are markers for quies-cent and activated hepatic stellate cells in formalin-fixedparaffin embedded human liver. Histochem Cell Biol2009;131:313-325.

34. Kordes C, Sawitza I, Haussinger D. Hepatic and pancre-atic stellate cells in focus. Biol Chem 2009;390:1003-1012.

35. Bartholomew A, Polchert D, Szilagyi E, Douglas GW,


Kenyon N. Mesenchymal stem cells in the induction oftransplantation tolerance. Transplantation 2009;87:S55-S57.

36. Lee KD, Kuo TK, Whang-Peng J, Chung YF, Lin CT,Chou SH, et al. In vitro hepatic differentiation of humanmesenchymal stem cells. HEPATOLOGY 2004;40:1275-1284.

37. Banas A, Teratani T, Yamamoto Y, Tokuhara M, Take-shita F, Quinn G, et al. Adipose tissue-derived mesen-chymal stem cells as a source of human hepatocytes.HEPATOLOGY 2007;46:219-228.

38. Zheng YB, Gao ZL, Xie C, Zhu HP, Peng L, Chen JH,

et al. Characterization and hepatogenic differentiation ofmesenchymal stem cells from human amniotic fluid andhuman bone marrow: a comparative study. Cell Biol Int2008;32:1439-1448.

39. Sato Y, Araki H, Kato J, Nakamura K, Kawano Y,Kobune M, et al. Human mesenchymal stem cells xen-ografted directly to rat liver are differentiated intohuman hepatocytes without fusion. Blood 2005;106:756-763.

40. Brulport M, Schormann W, Bauer A, Hermes M, ElsnerC, Hammersen FJ, et al. Fate of extrahepatic humanstem and precursor cells after transplantation intomouse livers. HEPATOLOGY 2007;46:861-870.


Mobilization of hepatic mesenchymal stem cells from human liver grafts

Documents

Transcript of Mobilization of hepatic mesenchymal stem cells from human liver grafts