Failure of hepatocyte marker-expressing hematopoietic progenitor cells to efficiently convert into hepatocytes in vitro

Experimental Hematology 34 (2006) 348–358

Failure of hepatocyte marker-expressing hematopoietic progenitor cells to efficiently convert into hepatocytes in vitro Gewei Lian*, Chengyan Wang*, Chunbo Teng, Cong Zhang, Liying Du, Qian Zhong, Chenglin Miao, Mingxiao Ding, and Hongkui Deng Department of Cell Biology and Genetics, College of Life Sciences, Peking University, Beijing, China (Received 11 January 2005; revised 31 October 2005; accepted 2 December 2005)

Objective. Whether bone marrow (BM) hematopoietic stem/progenitor cells can directly differentiate into nonhematopoietic cells remains controversial. The aim of this study is to further investigate the potentiality of BM hematopoietic progenitor cells to convert into hepatocytes in vitro. Materials and Methods. Different subsets of BM cells from C57/BL6 mice were isolated using markers of hematopoietic stem cells by magnetic cell sorting and by flow cytometry. These cells were induced to transdifferentiate to hepatocytes in vitro in the presence of various cytokines or of hepatocytes (or tissue) from damaged liver, which have been reported to stimulate the conversion. Hepatic gene markers in freshly isolated or cultured BM cells were determined by reverse transcriptase polymerase chain reaction and immunofluorescence. Results. Freshly isolated hematopoietic progenitor cells (HPC) expressed a low level of messenger RNAs of hepatic cell-specific markers including albumin and a-fetoprotein (AFP), but did not significantly upregulate expression of these markers, even in the presence of cytokines or cocultured hepatocytes (or tissue). HPCs induced in vitro did not express the message of aanti-trypsinda mature hepatocyte marker. At protein level, the specific staining of AFP was not detected in the HPCs, either freshly isolated or in vitro induced. Albumin protein was detected in freshly isolated albumin mRNA-positive and -negative BM cell subpopulations. Albumin-stained BM cells disappeared after being induced for 5 days, but restained if mouse serum was supplemented in medium for a 24-hour extended culture, suggesting that albumin was absorbed by BM cells instead of de novo expression. Conclusions. HPCs expressed mRNAs of hepatic cell markers, but could not efficiently convert into hepatocytes in vitro under our experimental conditions. Our observation raises a cautionary note in determining whether in vitro transdifferentiation of BM cells to hepatocytes can actually take place. Ó 2006 International Society for Experimental Hematology. Published by Elsevier Inc.

Whether hematopoietic stem cells (HSCs) can directly differentiate into nonhematopoietic cells remains controversial [1–16]. Many previous studies have reported that the bone marrow (BM) cells or HSCs possess the ability to ‘‘transdifferentiate’’ into other cell types, including hepatocytes [17– 21]. Subsequent studies suggest that the transdifferentiation could result from fusion of donor cells with receptor cells in vivo [2,3,22–28]; few HSCs are capable of directly transdifOffprint requests to: Mingxiao Ding, Ph.D., and Hongkui Deng, Ph.D., Department of Cell Biology and Genetics, College of Life Sciences, Peking University, Beijing, 100871, China; E-mail: hongkui_deng@pku. edu.cn (H.D.) *Drs. Lian and Wang contributed equally to this article.

ferentiating into other tissue lineage cells like hepatocytes [4]. However, recent studies have shown that conversion of BM-derived cells to other cells can occur without cellto-cell fusion [5]. Particularly, cultured BM cells can directly transdifferentiate into hepatocyte-like cells in vitro [29–37]. There are several possible mechanisms that can explain the phenomena of transdifferentiation. One explanation is that HSCs dedifferentiate and redifferentiate, or transdifferentiate directly to other cell types, representing true transdifferentiation. A second one is that the fusion of donor HSC progeny with recipient cells occurs without transdifferentiation. A third, more likely, scenario is that multipotent adult progenitor cells (MAPCs), and various tissue-restricted

0301-472X/06 $–see front matter. Copyright Ó 2006 International Society for Experimental Hematology. Published by Elsevier Inc. doi:10.1016/j.exphem.2005.12.004

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

stem cells exist in an isolated HSC population (or BM cells). Significant evidence has shown it possible that two or more mechanisms may exist in the transdifferentiation process [5,6,23,24]. In this study, we investigated in vitro transdifferentiation of different subsets of BM cells into hepatocytes by cytokine stimulation and by coculture with hepatocytes or damaged liver tissue. In contrast to earlier results [29–37], our novel findings provided another alternative explanation that the presence of gene ectopic expression (and protein absorption) in BM cells might also be responsible for the phenomena of transdifferentiation, and raised a cautionary note about whether in vitro transdifferentiation can actually occur. Materials and methods Isolation of BM nucleated cells Tibias and femurs were collected from pathogen-free 4- to 8week-old male C57/BL6 mice (Bred in the Animal Center of Peking University), flushed with phosphate-buffered saline (PBS) containing 2% fetal bovine serum (FBS) (GIBCO BRL, Grand Island, NY, USA) at 4
C and BM cells were filtered through a 70mm nylon mesh to remove cell clumps. Erythrocytes were depleted from BM cells by preincubating with rat anti-mouse erythroid cell maker TER-119 antibody (Pharmingen, San Diego, CA, USA);

349

subsequent labeling with anti-rat IgG microbeads followed by magnetic cell sorting (MACS; Miltenyi Biotech, Auburn, CA, USA) was carried out according to manufacturer’s protocol. Then, BM nucleated cells were counted and subjected to preparation of RNA, immunocytochemistry, and cell culture. Isolation of lin2 c-kitD sca-1D hematopoietic progenitor cells BM cells were obtained by flushing the tibias and femurs of C57/ BL mice as described previously [19,38] and incubated with rat monoclonal antibodies specific for lineage markers, which included B220 for the B cell marker; CD2, CD4, CD5, and CD8a for T-cell markers; Gr-1 and CD11b for myeloid markers; and Ter-119 for erythrocytes (all lineage marker antibodies are from Pharmingen). Labeled cells were then incubated with goat antirat IgG microbeads (Dynal, Lake Shearer, NY, USA); mature lymphoid and myeloid cells were removed by exposure to a magnetic field and lineage-depleted cells were collected in flow-through. Purity of MACS-sorted lin2 cells was about 80% to 90% as checked by using fluorescein-activated cell sorting (FACS) (Fig. 1c). These lineage-depleted cells were stained with phycoerythrin (PE)-conjugated goat anti-rat IgG (HDL) antibody (Pharmingen) to relabel linD cells, washed with PBS containing 2% FBS to remove any unlinked antibodies and then were incubated with rat isotypic IgG to block the remaining binding sites of the goat anti-rat IgG antibody. These cells were again incubated with biotinylated rat anti-sca-1 antibody (Pharmingen), and subsequently stained with streptavidin-conjugated PE-Texas and allophycocyanin (APC)-conjugated anti-CD117 (c-kit) antibody

Figure 1. Sorting of different subsets of bone marrow (BM) cells and representative culture conditions. (a,b) c-kitD or c-kit2 cell sorting. (a) Control for sorting c-kitD cells. BM nucleated cells were stained with allophycocyanin (APC)-conjugated isotype antibody. (b) c-kitD or c-kit2 cell sorting. BM nucleated cells were stained with APC-conjugated anti-c-kit antibody. R1: c-kit2 cells, R2: c-kitD cells. (c,d) sorting of lin2 and lin2 c-kitD sca-1D cells. (c) lin2 cells sorted by magnetic cell sorting were again sorted by fluorescein-activated cell sorting, showing that 80290% cells (inside R2 gate) were lineage-negative. (d) lin2 c-kitD sca-1D are sorted from R2 in (2) and R4 gates, are about 0.1% of BM cells. (e) lin2 cells cultured in six kinds of different culture conditions for 7 days. Line 1-6 represented six kinds of culture conditions. aFGF 5 acidic fibroblast growth factor; bFGF 5 basic fibroblast growth factor; DMEM 5 Dulbecco’s modified Eagle’s medium; FBS 5 fetal bovine serum; HGF 5 hepatocyte growth factor; LPS 5 lipopolysaccharide; OSM 5 Oncostatin M; IL 5 interleukin; SCF 5 stem cell factor.

350

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

(Pharmingen). At the same time, BM cells or lineage-depleted cells stained with APC-conjugated isotypic control MoAb (Pharmingen), or streptavidin-conjugated PE-Texas were used as background controls. After the final wash, sca-1Dc-kitDlin2 cells were sorted using a fluorescence-activated cell sorter (Mflo). All sorts and analyses were repeated at least two times, the purity of sorted lin2sca-1Dc-kitD cells checked by FACS reanalysis is about 98.5%. For isolation of lin2 cells, PE-labeled cells described above were directly sorted using FACS. For isolation of c-kitD and/or c-kit2 cells, erythrocyte-free BM cells were stained with APC-conjugated anti-CD117 antibody (Pharmingen) and sorted into both c-kitD fraction and c-kit2 one using FACS. Liver injury and hepatocyte separation Livers of C57/BL6 mice (8–12 weeks) were damaged three times by intraperitoneal injections of CCl4/earthnut oil (1:10) at doses of 1.0 mg CCl4/kg body weight. Hepatotoxicity was confirmed by serum transaminase increase and liver pathological damage. One day after the third injection, hepatocytes were isolated from the damaged liver using a standard two-step ethylenediaminetetraacetic acid-collagenase (Roche Applied Science, Indianapolis, IN, USA) liver perfusion [39]. This procedure routinely yielded approximately 5 3 107 hepatocytes per mouse with viabilities ranging from 60% to 80%. Cells were stored on ice for coculture. Cell culture For differentiation of BM cells into hepatocytes, different subsets of BM cells, including BM nucleated cells, lin2 cells or sca-1D c-kitD lin2 progenitor cells were respectively inoculated in cell culture plates with culture mediums consisting of Dulbecco’s modified Eagle’s medium (DMEM)/F12 (1:1), 10% FBS, 100 U/ mL penicillin/streptomycin, 1025 M 2-mercaptoethanol (Sigma, St Louis, MO, USA), 20 to 100 ng/mL hepatocyte growth factor (HGF; Peprotech Inc, Rocky Hill, NJ, USA), 25 ng/mL acidic fibroblast growth factor (aFGF; Peprotech Inc), 25 ng/mL basic fibroblast growth factor (bFGF; R&D Systems, Minneapolis, MN, USA), 20 ng/mL FGF-4 (Peprotech Inc), 20 ng/mL Oncostatin M (OSM; Peprotech Inc), and 1027 M dexamethasone (Sigma); 10 ng/mL stem cell factor (SCF; R&D systems), 5 ng/mL interleukin (IL)-3 (R&D Systems), 5 ng/mL IL-6 (R&D Systems), and/or 100 ng/mL lipopolysaccharide (LPS; Sigma) were also selectively supplemented into the culture medium to investigate their influence on the transdifferentiation of BM cells. As a control, these cells were also cultured in a basal culture medium, which consisted of DMEM/F12 (1:1), 10% FBS and 100 U/mL penicillin/ streptomycin. To investigate the possibility that BM cells could take up serum albumin in medium, cultured cells in the basal medium had to be treated with DMEM/F12 (1:1) containing 5% mouse serum and 5% FBS for 24 hours before immunofluorescence staining. For coculture of BM cells with hepatocytes (or liver tissue), each well of six-well culture plates was divided into two chambers by transwell membrane (Transwell; Corning Coaster, Cambridge, MA, USA; pore size 0.4 mm); hepatocytes (3–5 3 105 cell/cm2) and lin2 cells (1–3 3 106 cells/well) were inoculated in the upper and lower chambers, respectively, with DMEM/F12 (1:1) medium containing 10% FBS, 20 ng/mL HGF, 20 ng/mL epidermal growth factor (EGF; R&D systems), 10 ng/mL SCF, 20 ng/mL OSM, 1027M Dex and ITS (GIBCO BRL). Cells were grown at 37
C in humidified 5% CO2/95% air for 3 to 7 days.

RNA extraction and reverse transcriptase polymerase chain reaction (RT-PCR) Total RNA was usually prepared from freshly isolated or cultured cells using NucleoSpin RNAII (Cat No.740955.50; MachereyNagel, Germany), but for more than 2 3 106 cells or about 50 mg liver tissue, total RNA was prepared using TRIzol Reagent (Life Technologies, Rockville, MD, USA). Extracted RNA was dissolved in 20 to 30 mL diethylpyrocarbonate (DEPC)-treated H2O, and RNA samples were reverse transcribed using Superscript II First-strand Synthesis System (Invitrogen, Carlsbad, CA, USA). Resulting cDNA was amplified using Platinum Tag DNA Polymerase (Invitrogen, Waltham, MA, USA) on Peltier Thermal Cycler (MJ Research) under the following conditions: initial denaturation at 94
C for 3 minutes, and then 25 cycles for b-actin or 40 cycles for liver-specific markers (94
C for 30 seconds, 54
C to 59
C for 40 seconds, 72
C for 40 seconds), and final extension at 72
C for 10 minutes. Primers for PCR to detect the various liver-specific genes were as follows: albumin (55
C, 389 bp), 50 GATGAAACATATGT CCCCAAAGA-30 and 50 -TGTGTTCCTAGGGTGTTGATTTTA-30 ; a-fetoprotein (AFP; 54.2
C, 324 bp), 50 -CCTATGCCCCTCCCCCATTC-30 and 50 -CTCACACCAAAGCGTCAA CACATT-30 ; Tyrosine aminotansferase (TAT) (56
C, 210 bp), 50 -AGCGACTCATG AGCCCTTTA-30 and 50 -TGTGTCAGCTCGAGGAAATG-30 ; a-1-anti-trypsin (AAT) (55
C, 484 bp), 50 -AATGGAAGAAGCCATTCGAT-30 and 50 -AAGACTGTAGCTG CTGCAGC-30 ; cytokeratin 18 (CK18) (57
C, 274 bp), 50 -GCTGGAGACAGAAATCGAGG-30 and 50 -C TTGGTGGTGACAACTGTGG-30 ; CK19 (56
C, 570 bp), 50 -GTCCTACAGATT GAC AATGC-30 and 50 -CACGCTCTGGATCTGTGACAG-30 ; HNF1 (58.7
C, 418 bp), 50 - GTAAGGTCCACGGTGTACG GTA-30 and 50 -GAGGCTGCTGATAGGAGGGAT G-30 ; HNF3b (58
C, 345 bp) 50 -ACCTGAGTCCGAGTCTGACC-30 and 50 GGCA CCTTGAGAAAGCAGTC-30 ; HNF-4a (55
C, 270 bp), 50 -ACACGTCCCCATCTG AAG-3 and 50 -CTTCCTTCTTCATGCCAG-30 ; b-actin (58
C, 572 bp), 50 -GGTGGG AATGGGTCAGAAGG-30 and 50 -AGGAAGAGGATGCGGCAGTG-30 . All primers were selected from two different exons, including at least one intron, to avoid false signals from genomic DNA. PCR product was kept at 4
C and then separated in 1% agarose gel. In order to determine the specificity of PCR product, all of the amplified gene fragments were linked into T-vector for sequencing analysis. All experiments were repeated at least three times and they included two negative controls: one in which no reverse transcriptase was added in the RT reaction and another in which no template was used in the RT-PCR reaction, no gene expression was detected in any of the negative control samples. Real time RT-PCR Real-time PCR was conducted with SYBR Green on an ABI PRISMs 7000 Sequence Detection System (ABI, Foster City, CA, USA). Lin2 cells were induced by a representative combination of cytokines (condition 5 of Fig. 1e) or cocultured with damaged liver tissue for 7 days. mRNA was reverse-transcribed with Superscript II First-strand Synthesis System (Invitrogen). PCR reaction was performed in a 25-mL volume containing 45 ng cDNA, 13 SYBR Green PCR master mix, 300 nM of the forward and the reverse primers (AFP: 50 -CCTATGCCCCTCCCCCATTC-30 and 50 -CTCACACCAAAGCGTCAACACATT-30 ; b-actin: 50 -TTT CCAGCCTTCCTTCTTG-30 and 50 -TGGCATAGAGGTCTTTACGG-30 ). PCR profile was as follows, 2 minutes at 50
C,

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

10 minutes at 95
C, followed by 41 cycles of 15 seconds at 95
C, 20 seconds at 58
C, and 20 seconds at 72
C. AFP mRNA quantity was determined by comparative CT method. The relative quantitation value of AFP gene, normalized to an endogenous control bactin gene and relative to a calibrator, is expressed as 2-DDCt. Immunocytochemistry Different subsets of BM cells, including BM nucleated cells, linD and lin2 cells, c-kitD and c-kit2 cells and sca-1D c-kitD lin2 progenitor cells were washed twice in Dulbecco’s PBS, then centrifugated onto poly-lysine-coated microscope slides and fixed immediately with 4% paraformaldehyde at 4
C for 20 minutes. These cells were then washed three times in PBS and permeabilized with 0.3% Triton X-100 (Fisher Chemicals, Fair Lawn, NJ, USA) for 5 minutes at room temperature. Following blocking in 10% preimmune goat serum in PBS for 2 hours, cells were analyzed by standard immunocytochemistry methodology with primary antibodies including rabbit anti-mouse albumin antibody (1:150 dilution; Biogenesis, Poole, UK) and rabbit anti-human AFP antibody (1:150 dilution, DAKO Copenhagen, Denmark), mouse monoclonal anti-human CK18 (1:200, Progen, Heidelberg, Germany), mouse monoclonal anti-CK19 (1:200; Santa Cruz Biotechnology Inc., Santa Cruz, CA) and second antibodies fluorescein isothiocyanate (FITC)-labeled goat anti-rabbit IgG (1:100 dilution, Jackson ImmunoResearch, West Grove, PA, USA), FITC-labeled goat anti-mouse IgG (1:100, Jackson ImmunoResearch). Finally, cells were washed six times and then counterstained with 40 , 6-diamidino-2-phenylindole in PBS for staining of cell nuclei. FITC fluorescence was observed under a fluorescence microscope (Olympus IX-71, Japan). In order to reveal profile of albumin existence in BM cells, confocal laser-scanning immunofluorescence was performed using Leica confocal systems (model: A4P2) on BM cells stained using anti-albumin antibody and propidium iodide. Serial confocal images of BM cells were collected at an interval of 1.2 mm along the Z-axis.

Results Cytokine treatments failed to induce the transdifferentiation of BM cells into hepatocytes. To test the BM cells’ ability to transdifferentiate into hepatocytes in vitro, we isolated BM nucleated cells, lin2 cells and lin2 c-kitD sca-1D hematopoietic progenitor cells, from adult mouse bone marrow (Fig. 1a–d) and cultured them in a DMEM/F12 (1:1) medium containing 10% FBS and various combinations of cytokines including HGF, EGF, aFGF, bFGF, OSM, Dex, platelet-derived growth factor, SCF, IL-3, and LPS. These cytokines have been reported to either induce transdifferentiation of BM cells into hepatocytes, or to upregulate expression of albumin and c-met, a receptor of HGF [33–38,41,42]. On the 7th day after these lin2 cells were treated with some representative combination of cytokines (Fig. 1e), these cells were found to contain, as detected by RT-PCR, messages for hepatocyte-specific markers including albumin, AFP, CK18, and TAT, but not mRNA for AATda marker for mature hepatocytes (Fig. 2a). These cells also possessed

351

mRNA for cytokeratin 19 (CK19), a cholangiocyte-specific marker, as well as hepatocyte transcription factors, including hepatocyte nuclear factor (HNF)1, HNF3b, and HNF4a, which were required for hepatocyte or endoderm differentiation. These data were consistent with earlier reports [29–37] and suggested that the lin2 cells could transdifferentiate to some extent into hepatocytes or cholangiocytes when induced by cytokines. Unexpectedly, however, we found that most of these hepatic cell-specific markers, including AFP, albumin, and TAT (lane 6 of Fig. 2a), were already detectable at comparable levels in lin2 cells grown in a control culture medium without added cytokines (Fig. 1e, line 6). Similar results were obtained when BM nucleated cells and lin2 c-kitD sca-1D cells were cultured with and without cytokines (data not shown). These results showed that under in vitro conditions, many hematopoietic progenitor cells had already possessed hepatocyte markers and cytokine treatments did not significantly raise the expression levels of such markers. Hepatocyte marker-expressing BM cells are enriched in freshly isolated lin2c-kitDsca-1D hematopoietic progenitor cells. The finding that lin2 cells in the control medium expressed hepatocyte markers was significant, as it suggested that perhaps some native BM cells might already express such markers in vivo. To test this possibility, we isolated the erythroid cell-free BM cells from the tibias and femurs of C57/BL6 mice. RT-PCR analysis of such cells indeed detected hepatocyte-specific messages for albumin, AFP, TAT, and CK18, the cholangiocyte-specific gene CK19, and hepatocyte transcription factors HNF1 and HNF4a (Fig. 3a); the identities of the PCR products including albumin, AFP, and TAT were confirmed by sequencing. In contrast, messages for AAT, a mature hepatocyte marker, were not detected. These results indicated that at least some of the BM cells indeed expressed hepatocyte-specific markers in vivo. To determine which subpopulation of BM cells expressed hepatic markers, we analyzed different subsets of BM cells using RT-PCR. Albumin markers were detectable by PCR in lin2 cells rather than in linD cells (despite use of threefold excess of the linD cells). mRNA levels for AFP and CK18 in lin2 cells were also higher than those in linD cells when normalized to an endogenous control b-actin (Fig. 3b). Likewise, albumin was also expressed only in c-kitD cells rather than c-kit2 cells while both c-kitD and ckit2 cells expressed AFP and CK18 (Fig. 3b). Results indicated that the hepatocyte marker-expressing cells were enriched in the lin2 and the c-kitD populations. We then analyzed the expression of hepatocyte-specific markers in freshly isolated lin2 c-kitD sca-1D hematopoietic progenitor cells. The lin2 c-kitD sca-1D cells were sorted from BM cells by FACS (Fig. 1c and d), and were found to express hepatocyte markers including AFP, CK18, and albumin (Fig. 3c). Consistent with the results of the lin2 or the c-kitD cells as shown in Figure 3b, the

352

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

Figure 2. Analysis of gene expression of lin2 cells cultured in different conditions for 7 days. (a) Reverse-transcriptase polymerase chain reaction analysis (PCR); 0.5–1 3 106 lin2 cells were cultured in different conditions (lane 1–6) for 7days, 0.5–2.0-mg fractions (equivalent to 0.5–2 3 105 cells) of total RNAs extracted from the induced cells were reverse-transcribed in a total 20-mL volume, and then PCR was performed using 2.0 mL cDNA products for each gene as described in Materials and Methods. Lane 1–6, respectively, represented expression of hepatic cell-specific genes for lin2 cells in six different kinds of culture conditions corresponding to line 1–6 of Figure 1e, ‘‘L’’ represented liver tissue as positive control; 25 cycles of PCR for b-actin and 40 cycles for other hepatic cell markers. (b) Immunofluorescence detection of albumin (fluorescein isothiocyanate [FITC]) for lin2 cells cultured in condition 5 (Fig. 1e, line 5). Albumin was not detectable at the protein level, and albumin expression in other culture conditions was similar to that in condition 5. (c) Positive control for albumin immunostaining. Mouse hepatocytes were stained together with lin2 cells as described in Materials and Methods. Scale bar: 20 mm, original magnification: 340. AAT 5 a-1 anti-trypsin; AFP 5 a-fetoprotein; CK 5 cytokeratin; DAPI 5 40 , 6-diamidino-2-phenylindole; HNF 5 hepatocyte nuclear factor; TAT 5 tyrosine aminotansferase.

mRNAs of both AFP and CK18 were enriched in hematopoietic progenitor cells, and albumin expression was detected only in these cells. These results confirmed that the hepatocyte-marker2expressing cells are further enriched in the lin-c-kitDsca-1D hematopoietic progenitor cells, and suggested that cultured HSC normally express low levels of the mRNAs of hepatic cell markers. Expression of hepatic cell marker proteins in BM cells We then immunostained the freshly isolated BM nucleated cells to examine the expression of albumin and AFP at the protein level. We found that about 5% to 10% of BM nucleated cells were albumin-positive, but few cells were AFPpositive (Fig. 4 a and b); preimmune rabbit sera and rabbit anti-mouse albumin antibody preincubated with mouse se-

rum were negative (Fig. 4c and d). We found that some lin2 cells were CK18-positive staining. However, the same staining was also observed even without use of murine primary antibodies (data not shown), thus suggesting this immunostaining might be due to the contamination of lin2 cells by a small number of IgG-secreting lymphoid cells. Taken together, these results indicated that, although we could not determine with certainty the presence of keratin antigen (CK18), subsets of BM cells indeed expressed the albumin protein. Albumin in freshly isolated BM cells originated from absorption instead of de novo expression To clarify which population of BM cells can make the albumin protein, we examined by immunofluorescence the protein in linD vs lin2 cells, and c-kitD vs c-kit2 cells.

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

353

Figure 3. Reverse transcriptase polymerase chain reaction (RT-PCR) analyses of hepatic cell-specific genes in different subsets of freshly isolated bone marrow (BM) cells. Total RNAs were prepared from (a) 1 3 107 BM nucleated cells and 50 mg liver tissue; (b) 1 3 106 linD and 3 3 105 lin2 cells, 4 3 105 ckitD and 1.3 3 106 c-kit2 cells; (c) 3 3 105 lin2 c-kitD sca-1D cells (lane 2) and 1 3 106 lin2 c-kitD sca-1D cell-deficient BM nucleated cells (lane 1). RTPCR was performed as described in Materials and Methods with 5–20% of total RNA amounts of these samples. All PCR products were linked to T-vector and subjected to sequencing analysis; sequencing results showed that the fragment sequences were identical to those reported in GenBank. ‘‘L’’ represents liver tissue as positive control; ‘‘Con’’ represents negative control; 25 cycles of PCR for b-actin and 40 cycles for other hepatic cell markers. The figure was a representative of at least three-times2repeated experiment results. AAT 5 a-1 anti-trypsin; AFP 5 a-fetoprotein; CK 5 cytokeratin; HNF 5 hepatocyte nuclear factor; TAT 5 tyrosine aminotansferase.

Our results were unexpected because: 1) it was found that the number of albumin-positive cells in c-kitD group was almost identical to that in the c-kit2 group (Fig. 4e and f); similar results were obtained in the linD and lin2 cells by anti-albumin immunostaining (data not shown); 2) despite the anti-albumin immunostaining signals, we did not detect any albumin mRNA in the linD or c-kit2 cells by RT-PCR (Fig. 3b); 3) although albumin protein could be detected in freshly isolated lin2 cells, it was undetectable in these cells after they had been cultured for 7 days (Fig. 2b). Taken together, these results raised the possibility that BM cells could absorb albumin from the blood. This possibility was supported by our data. First, we found that the amount of albumin in cultured lin2 cells, when they were grown in 10% serum medium for 4 days, decreased and finally disappeared (Fig. 5a). Second, if these albumin-depleted BM cells were fed with a medium containing mouse serum, almost 50% of the cultured cells reacquired the albumin (Fig. 5b). Third, no BM cells were found to be able to synthesize albumin even when they were cultured in a hepatic progenitor growth medium for 5 days (Fig. 5c). Finally, confocal microscopy revealed that albumin was mainly membrane-associated in most albumin-positive BM cells (Fig. 5d), while only a few cells showed cytoplasmic staining (Fig. 5e). Altogether, these results suggested that BM cells could take up albumin in culture, and that albumin-positive staining of BM cells was due to absorption.

Effect of damaged liver tissue on the in vitro transdifferentiation of the lin2 hematopoietic progenitor cells into hepatocytes To examine the effects of hepatocytes and damaged liver tissues on conversion of hematopoietic progenitor cells into hepatocytes, we cultured the lin2 cells under three conditions: alone, in the presence of hepatocytes from normal liver, or hepatocytes from damaged liver. The lin2 cells were physically separated from the hepatocytes by a transwell membrane. In all three conditions, no evident difference in expression of hepatocyte markers including AFP, Alb, and AAT was detected by RT-PCR (Fig. 6) and by real time RT-PCR (analyzing AFP expression; Fig. 7). Similar results were also obtained when the lin2 c-kitD sca-1D hematopoietic progenitor cells were cultured with hepatocytes from normal or damaged livers (not shown). These data indicated that the damaged liver tissues did not have a significant inductive effect on the in vitro conversion of the lin2 (or lin2 sca-1D c-kitD) hematopoietic progenitor cells into hepatocytes.

Discussion Although our results showed that hematopoietic progenitor cells after induction expressed some hepatic cell markers as reported previously [29–37] at RNA and protein levels, we do not think that transdifferentiation had actually taken

Figure 4. Detection of hepatocyte-specific markers for freshly isolated bone marrow (BM) nucleated cells, c-kitD, and c-kit2 cells by immunofluorescence. (a,b) BM nucleated cells. (a) Expression of a-fetoprotein (AFP) (fluoroscein isothiocyanate [FITC]) was hardly detected. (b) About 5–10% BM cells displayed albumin staining (FITC). (c) No staining (FITC) was found when BM nucleated cells were stained with preimmune rabbit IgG instead of rabbit antialbumin IgG. (d) Specificity of albumin staining. Mouse serum can efficiently blocked albumin staining when anti-albumin antibody was preincubated with 10% mouse serum for 5 minutes prior to immunostaining. (e,f) albumin staining of c-kitD(e) and c-kit2 (f) cells. No apparent difference in the amount of albumin-positive cells between c-kitD cells and c-kit2 cells was observed. Scale bar: 10 mm, original magnification: 340. DAPI 5 40 , 6-diamidino-2phenylindole.

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

355

Figure 5. Identification of albumin absorption of bone marrow (BM) cells. (a–c) albumin staining (fluoroscein isothiocyanate [FITC]) of cultured lin2 cells. (a) No albumin was detected in cells cultured in Dulbecco’s modified Eagle’s medium/F12 containing 10% fetal bovine serum for 4 days. (b) More than 50% albumin-positive cells were observed when these cells continue to be cultured in medium containing 5% mouse serum for 24 hours. (c) Albumin could hardly be observed in cells cultured in a hepatocyte growth medium for 5 days. Scale bar: 20 mm, original magnification: 340. (d,e) Representative confocal images of albumin staining for freshly isolated BM cells. (d) Albumin located on membrane of most of positive BM cells (e) Albumin existed in cytoplasm and membrane of some positive BM cells. Scale bar: 8 mm, original magnification: 3240. DAPI 5 40 , 6-diamidino-2-phenylindole; PI 5 propidium iodide.

356

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

Figure 6. Reverse transcriptase polymerase chain reaction (RT-PCR) analyses of hepatic cell-specific markers in lin2 cells cocultured with hepatocytes from either normal or damaged liver. About 1 3 106 lin2 cells were cultured alone or cocultured with hepatocytes from either normal or damaged liver for 7 days, and then RT-PCR was performed with 10% of total RNA amounts of the samples. Expression of a-fetoprotein (AFP), albumin, and c-met mRNAs could be detected from lin2 cells cultured alone (lane 1), cocultured with hepatocytes from damaged livers (lane 2) and from normal livers (lane 3). No obvious differences in mRNA amount of hepatocyte markers were observed among the cells cultured in those three different conditions (lane 1–3). AAT 5 a-1 anti-trypsin; L 5 liver tissue as positive control.

place because of the following observations: 1) freshly isolated BM cells and hematopoietic progenitor cells have expressed some hepatic cell markers including albumin, AFP, CK18, and CK19; 2) it was not observed that hepatocyte marker expression was significantly altered (or upregulated) before and after hematopoietic progenitor cells were induced in vitro by exposing them to selected cytokines and damaged liver tissues; 3) such cells could take up albumin from the blood and the culture medium, suggesting albumin in BM cells cannot be regarded as a proof for the commitment of a BM cell to a hepatocyte fate; and 4) AAT mRNA could not be detected even after the lin2 sca-1D c-kitD cells were treated with selected cytokines for 7 days, suggesting that a complete set of hepatocyte markers, including some hepatocyte functions such as detoxification and metabolism, should be used to verify transdifferentiation of BM cells to hepatocytes. Several studies showed that BM cells/HSCs could directly transdifferentiate into hepatocytes when they were cultured in the induction condition [29–37]. There are two possible explanations for the discrepancy between our results and the reported ones [29–30]. One was the different mode of liver injury, which could cause the different induction results. Another may result from the different host cell population. In our study, we separated HSCs by FACS with several marker antigens (Lin2, Sca1D and

Figure 7. Real-time reverse transcriptase polymerase chain reaction analyses of expression difference of hepatocyte-marker a-fetoprotein (AFP) between freshly isolated lin2 cells and cultured lin2 cells. Sample a: hepatocytes as positive control; sample b: freshly isolated lin2 cells; sample c: lin2 cells induced by a representative combination of cytokines (condition 5 of Fig. 1e); and sample d: lin2 cells cocultured with damaged liver tissue. No obvious differences in AFP mRNA amount were observed among these lin2 cells.

c-kitD), but Jang and colleagues used a quite different method to separate HSCs [29]. The fact that BM cells can express hepatic markers may involve two mechanisms: 1) a few bona fide hepatic stem cells existing in bone marrow and 2) illegitimate (or ectopic) transcription in BM cells. Because no hepatic stem (progenitor) cell-like colonies were formed when BM cells were placed in a culture medium designed for hepatic progenitor cell growth, our data were inconsistent with the first mechanism. Illegitimate transcription, i.e., a low level of transcription of tissue-specific genes in ectopic cells, is known to occur in many cell types [40–42]. Our data favors the possibility that ectopic transcription is responsible for expression of hepatic markers in hematopoietic progenitor cells. Our finding that hepatocyte-specific markers, including albumin, AFP, and CK19, are expressed in freshly isolated BM cells is consistent with several studies [43–46]. In addition, we detected the messenger RNAs of several other hepatocyte markers that have not been detected in BM cells previously; these markers include CK18, TAT, and HNF1 (Fig. 3). Given the recent suggestions that BM (and hematopoietic progenitor) cells can be converted to hepatocytes in vitro [29–37], based on the expression of albumin and other hepatocyte markers, our findings raise a cautionary note regarding whether such an in vitro transdifferentiation of BM (or hematopoietic progenitor) cells actually take place.

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

Acknowledgments We thank Dr. Tung-Tien Sun of New York University School of Medicine for critical reading of the manuscript and Dr. Huachun Sang of Peking University for assistance in confocal microscopy. The study was supported by grants (no. 2001CB510106, no. 1999053900) from Ministry of Science and Technology, by an Outstanding Young Scientist Award from National Natural Science Foundation of China (30125022) and Bill and Melinda Gates Foundation Grant (37871) to H. Deng, and by a Chinese Postdoctoral Science Foundation (2003033068) to G. Lian.

References 1. Almeida-Porada G, Porada CD, Chamberlain J, Torabi A, Zanjani ED. Formation of human hepatocytes by human hematopoietic stem cells in sheep. Blood. 2004;104:2582–2590. 2. Terada N, Hamazaki T, Oka M, et al. Bone marrow cells adopt the phenotype of other cells by spontaneous cell fusion. Nature. 2002; 416:542–545. 3. Ying QL, Nichols J, Evans EP, Smith AG. Changing potency by spontaneous fusion. Nature. 2002;416:545–548. 4. Wagers AJ, Sherwood RI, Christensen JL, Weissman IL. Little evidence for developmental plasticity of adult hematopoietic stem cells. Science. 2002;297:2256–2259. 5. Harris RG, Herzog EL, Bruscia EM, Grove JE, Van Arnam JS, Krause DS. Lack of a fusion requirement for development of bone marrowderived epithelia. Science. 2004;305:90–93. 6. Newsome PN, Johannessen I, Boyle S, et al. Human cord bloodderived cells can differentiate into hepatocytes in the mouse liver with no evidence of cellular fusion. Gastroenterology. 2003;124: 1891–1900. 7. Theise ND, Krause DS, Sharkis S. Comment on ‘Little evidence for developmental plasticity of adult hematopoietic stem cells.’ Science. 2003;299:1317. 8. Nygren JM, Jovinge S, Breitbach M, et al. Bone marrow-derived hematopoietic cells generate cardiomyocytes at a low frequency through cell fusion, but not transdifferentiation. Nat Med. 2004;10:494–501. 9. Kanazawa Y, Verma IM. Little evidence of bone marrow-derived hepatocytes in the replacement of injured liver. Proc Natl Acad Sci U S A. 2003;100(suppl 1):11850–11853. 10. Raff M. Adult stem cell plasticity: fact or artifact? Ann Rev Cell Dev Biol. 2003;19:1–22. 11. Wagers AJ, Weissman IL. Plasticity of adult stem cells. Cell. 2004; 116:639–648. 12. Herzog EL, Chai L, Krause DS. Plasticity of marrow-derived stem cells. Blood. 2003;102:3483–3493. 13. Cogle CR, Yachnis AT, Laywell ED, et al. Bone marrow transdifferentiation in brain after transplantation: a retrospective study. Lancet. 2004;363:1432–1437. 14. Ianus A, Holz GG, Theise ND, Hussain MA. In vivo derivation of glucose-competent pancreatic endocrine cells from bone marrow without evidence of cell fusion. J Clin Invest. 2003;111:843–850. 15. Lechner A, Yang YG, Blacken RA, Wang L, Nolan AL, Habener JF. No evidence for significant transdifferentiation of bone marrow into pancreatic beta-cells in vivo. Diabetes. 2004;53:616–623. 16. Castro RF, Jackson KA, Goodell MA, Robertson CS, Liu H, Shine HD. Failure of bone marrow cells to transdifferentiate into neural cells in vivo. Science. 2002;297: 1299–1299. 17. Ferrari G, Cusella-De Angelis G, Coletta M, et al. Muscle regeneration by bone marrow–derived myogenic progenitors. Science. 1998; 279:1528–1530. 18. Petersen BE, Bowen WC, Patrene KD, et al. Bone marrow as a potential source of hepatic oval cells. Science. 1999;284:1168–1170.

357

19. Lagasse E, Connors H, Al-Dhalimy M, et al. Purified hematopoietic stem cells can differentiate into hepatocytes in vivo. Nat Med. 2000; 6:1229–1234. 20. Krause DS, Theise ND, Collector MI, et al. Multi-organ, multi-lineage engraftment by a single bone marrow derived stem cell. Cell. 2001; 105:369–377. 21. Theise ND, Nimmakayalu M, Gardner R, et al. Liver from bone marrow in humans. Hepatology. 2000;32:11–16. 22. Alvarez-Dolado M, Pardal R, Garcia-Verdugo JM, et al. Fusion of bone-marrow-derived cells with Purkinje neurons, cardiomyocytes and hepatocytes. Nature. 2003;425:968–973. 23. Wang X, Willenbring H, Akkari Y, et al. Cell fusion is the principal source of bone-marrow-derived hepatocytes. Nature. 2003;422: 897–901. 24. Vassilopoulos G, Wang PR, Russell DW. Transplanted bone marrow regenerates liver by cell fusion. Nature. 2003;422:901–904. 25. Spees JL, Olson SD, Ylostalo J, Lynch PJ, et al. Differentiation, cell fusion, and nuclear fusion during ex vivo repair of epithelium by human adult stem cells from bone marrow stroma. PNAS. 2003;100: 2397–2402. 26. Camargo FD, Finegold M, Goodell MA. Hematopoietic myelomonocytic cells are the major source of hepatocyte fusion partners. J Clin Invest. 2004;113:1266–1271. 27. Willenbring H, Bailey AS, Foster M, et al. Myelomonocytic cells are sufficient for therapeutic cell fusion in liver. Nat Med. 2004;10: 744–748. 28. Shi D, Reinecke H, Murry CE, Torok-Storb B. Myogenic fusion of human bone marrow stromal cells, but not hematopoietic cells. Blood. 2004;104:290–294. 29. Jang YY, Collector MI, Baylin SB, Diehl AM, Sharkis SJ. Hematopoietic stem cells convert into liver cells within days without fusion. Nat Cell Biol. 2004;6:532–539. 30. Jang YY, Sharkis SJ. Metamorphosis from bone marrow derived primitive stem cells to functional liver cells. Cell Cycle. 2004;3: 980–982. 31. Okumoto K, Saito T, Hattori E, et al. Differentiation of bone marrow cells into cells that express liver-specific genes in vitro: implication of the Notch signals in differentiation. Biochem Biophys Res Commun. 2003;304:691–695. 32. Avital I, Inderbitzin D, Aoki T, et al. Isolation, characterization, and transplantation of bone marrow-derived hepatocyte stem cells. Biochem Biophys Res Commun. 2001;288:156–164. 33. Wang PP, Wang JH, Yan ZP, et al. Expression of hepatocyte-like phenotypes in bone marrow stromal cells after HGF induction. Biochem Biophys Res Commun. 2004;320:712–716. 34. Hong H, Chen JZ, Zhou F, Xue L, Zhao GQ. Influence of serum from liver-damaged rats on differentiation tendency of bone marrowderived stem cells. World J Gastroenterol. 2004;10:2250–2253. 35. Oh SH, Miyazaki M, Kouchi H, et al. Hepatocyte growth factor induces differentiation of adult rat bone marrow cells into a hepatocyte lineage in vitro. Biochem Biophys Res Commun. 2000;279: 500–504. 36. Miyazaki M, Akiyama I, Sakaguchi M, et al. Improved conditions to induce hepatocytes from rat bone marrow cells in culture. Biochem Biophys Res Commun. 2002;298:24–30. 37. Yamazaki S, Miki K, Hasegawa K, Sata M, Takayama T, Makuuchi M. Sera from liver failure patients and a demethylating agent stimulate transdifferentiation of murine bone marrow cells into hepatocytes in coculture with nonparenchymal liver cells. J Hepatol. 2003; 39:17–23. 38. Spangrude GJ, Heimfeld S, Weissman IL. Purification and characterization of mouse hematopoietic stem cells. Science. 1988;241: 58–62. 39. Seglen PO. Preparation of isolated rat liver cells. Methods Cell Biol. 1976;13:29–83.

358

G. Lian et al./ Experimental Hematology 34 (2006) 348–358

40. Chelly J, Concordet JP, Kaplan JC, Kahn A. Illegitimate transcription: transcription of any gene in any cell type. Proc Natl Acad Sci U S A. 1989;86:2617–2621. 41. Trifunovic A, Zaric J, Lusic M, Mijatovic M, Glisin V, Popovic Z. Induced synthesis of albumin-like protein in damaged rat reticulocytes. Br J Haematol. 2001;115:205–212. 42. Toran-Allerand CD. Coexistence of a-fetoprotein, albumin and transferrin immunoreactivity in neurones of the developing mouse brain. Nature. 1980;286:733–735. 43. Muller C, Petermann D, Pfeffel F, Oesterreicher C, Fugger R. Lack of specificity of albumin-mRNA-positive cells as a marker of circulating hepatoma cells. Hepatology. 1997;25:896–899.

44. Iguchi T, Sogo S, Hisha H, et al. HGF activates signal transduction from EPO receptor on human cord blood CD34D/CD45D cells. Stem Cells. 1999;17:82–91. 45. Ratajczak MZ, Kucia M, Reca R, Majka M, Janowska-Wieczorek A, Ratajczak J. Stem cell plasticity revisited: CXCR4-positive cells expressing mRNA for early muscle, liver and neural cells ‘hide out’ in the bone marrow. Leukemia. 2004;18:29–40. 46. Kubota H, Storms RW, Reid LM. Variant forms of alpha-fetoprotein transcripts expressed in human hematopoietic progenitors. Implications for their developmental potential towards endoderm. J Biol Chem. 2002;277:27629–27635.