Sequential Hepatogenic Transdifferentiation of Adipose Tissue-Derived Stem Cells: Relevance of Different Extracellular Signaling Molecules,Transcription Factors Involved,and Expression of New Key Marker Genes

Abstract

Adipose tissue contains a mesenchymal stem cell (MSC) population known as adipose-derived stem cells (ASCs) capable of differentiating into different cell types. Our aim was to induce hepatic transdifferentiation of ASCs by sequential exposure to several combinations of cytokines, growth factors, and hormones. The most efficient hepatogenic protocol includes fibroblastic growth factors (FGF) 2 and 4 and epidermal growth factor (EGF) (step 1), hepatocyte growth factor (HGF), FGF2, FGF4, and nicotinamide (Nic) (step 2), and oncostatin M (OSM), dexamethasone (Dex), and insulin-tranferrin-selenium (step 3). This protocol activated transcription factors [GATA6, Hex, CCAAT/enhancer binding protein α and β (CEBPα and β), peroxisome proliferator-activated receptor-γ, coactivator 1 α (PGC1α), and hepatocyte nuclear factor 4 α (HNF4α)], which promoted a characteristic hepatic phenotype, as assessed by new informative markers for the step-by-step hepatic transdifferentiation of hMSC [early markers: albumin (ALB), α-2-macroglobuline (α2M), complement protein C3 (C3), and selenoprotein P1 (SEPP1); late markers: cytochrome P450 3A4 (CYP3A4), apolipoprotein E (APOE), acyl-CoA synthetase long-chain family member 1 (ACSL1), and angiotensin II receptor, type 1 (AGTR1)]. The loss of adipose adult stem cell phenotype was detected by losing expression of Thy1 and inhibitor of DNA binding 3 (Id3). The reexpression of phosphoenolpyruvate corboxykinase (PEPCK), apolipoprotein C3 (APOCIII), aldolase B (ALDOB), and cytochrome P450 1A2 (CYP1A2) was achieved by transduction with a recombinant adenovirus for HNF4α and finally hepatic functionality was also assessed by analyzing specific biochemical markers. We conclude that ASCs could represent an alternative tool in clinical therapy for liver dysfunction and regenerative medicine.

Keywords

Adipose tissue Mesenchymal stem cell Hepatocyte differentiation Transcription factors

Introduction

Orthotopic liver transplantation is major effective treatment for patients with fulminant and chronic liver failure. However, this clinical practice is highly limited by lack of donors and organ rejection. Applying strategies similar to those tissues use to renovate damaged cells after injury is the main idea behind new concepts of cell therapy and regenerative medicine.

Having identified stem cell plasticity (66), multiple reports raised hopes that tissue repaired by stem cell transplantation could be available in the near future. Therefore, replacing diseased hepatocytes and the stimulation of regeneration by stem cells are the main aims of liver-directed cell therapy.

Human multipotential mesenchymal stem cells (hMSC) have been reported to possibly differentiate into a variety of mesenchymal tissues (i.e., osteoblasts, chondrocytes, adipocytes, myoblasts, and cardiomyocytes), and into various types of tissue cells derived from other embryonic layers, including neural and hepatocyte-like cells (26,27,63,67,78). Although bone marrow was the first source reported to contain hMSC (bone marrow stem cells, BMSC), reports indicate that hMSC may also be isolated from different tissues, including adipose tissue (17,28,37,58,73,96,102). The stromal-vascular cell fraction of adipose tissue has increasingly become the focus of stem cell research (8,69). Adipose tissue contains a multipotent cell population called adipose-derived stem cells (ASCs) (12,37,45,87,102). This cell population can be isolated from human lipoaspirates and differentiated toward osteogenic, adipogenic, neurogenic, myogenic, and chondrogenic lineages (55,68,71,72,101). Recently, its hepatogenic differentiation potential has been demonstrated (3,4,76,85,87,94). Our previous studies demonstrated that, compared with BMSC, ASC population provides an abundant and accessible source of adult stem cells with minimal patient discomfort and, therefore, represents a rich source of MSC that also possesses the longest culture period and the highest proliferation capacity (37,87).

The ability for ASCs to give rise to hepatocye-like cells depends on the induction strategy selected. However, the studies published to date have used distinct hepatocyte differentiation media based on previous protocols applied to non-ASC progenitor cells. Consequently, analyzing the status of the derived cells is not always trivial and can lead to various conclusions. Thus, a thoroughly comparative study defining the best differentiation protocol is needed. Moreover, a battery of cell-specific markers needs to be defined and the cell-type differentiation status proven to study differentiation toward a specific cell type. In addition, because the differentiation process is gradual, it is necessary to discriminate between the genes expressed at each stage of the process. The genes expressed at the early development stages may not be representative of fully differentiated cells.

Here we have examined different hepatogenic induction assays based on the sequential exposure of ASCs to cytokines, growth factors, and hormones reflecting their temporal expression during in vivo hepatogenesis. Our differentiation protocol is based on the use of prohepatogenic signals and mediators of liver development. We have established and optimized a three-step hepatogenic induction protocol by analyzing the role of epidermal growth factor (EGF), fibroblastic growth factors 2 and 4 (FGF2 and FGF4), and bone morphogenic proteins 2 and 4 (BMP2 and BMP4) at the commitment step; hepatocyte growth factor (HGF), FGF2, FGF4, and nicotinamide (Nic) at the differentiation step; and finally HGF, oncostatin M (OSM), dexamethasone (Dex), and insulin-transferrin-selenium (ITS + premix) (28) at the maturation step. Additionally, we have conducted an extensive gene expression characterization of the ASC population at different hepatic differentiation stages, determining not only early and late hepatic markers, but also temporal activation and the role of transcription factors. Hepatic functionality was also assessed by analyzing glycogen synthesis and accumulation, plasma protein synthesis and secretion, and P450 activities as specific biochemical markers.

Materials and Methods

Isolation and Culture of ASC

Human adipose tissue was obtained from discarded tissue after suction-assisted lipectomy (i.e., abdomen liposuction) (10 patients; mean age 43.2 ± 12.1 years). Protocols were approved by the institutional Review Board. Tissues were obtained according to the hospital's Ethics Committee and after obtaining informed consents from patients. ASCs were isolated and cultured as previously described (87).

Isolation and Culture of Human Hepatocytes

Hepatocytes were isolated using a two-step perfusion technique and cultured as described in detail elsewhere (22). All liver samples (n = 5) were obtained in conformity with the rules of the hospital's Ethics Committee and all donors were not suspected of harboring any infectious disease and tested negative for human immumo-deficiency virus and hepatitis.

Cell Lines

Human hepatoma HepG2 cells (ECACC No 85011430) were cultured as previously described (87).

Hepatic Differentiation Protocol

Passage 2 ASC cultures at confluence were used to assess different hepatogenic induction assays over a 2-week period. It is estimated that at least 50 hepatocyte-like cells from each initial ASC are obtained. A three-step differentiation protocol with serum-free DMEM low-glucose medium (Gibco, Paisley, UK), supplemented with 0.3 mg/ml Matrigel (lot 354234, BD Biosciences, Madrid, Spain) and sequential addition of factors and hormones, was used (Fig. 1A). In the commitment step, cells were cultured in medium supplemented with 20 ng/ml EGF (Sigma-Aldrich Química SA, Madrid, Spain) and 10 ng/ml FGF2 (Invitrogen, Barcelona, Spain) (protocol A), or 20 ng/ml EGF, 10 ng/ml FGF2, and 10 ng/ml FGF4 (Sigma-Aldrich Química SA) (protocol B), or 20 ng/ml EGF, 10 ng/ml FGF2 and FGF4, and 50 ng/ml BMP2 and BMP4 (Sigma-Aldrich Química SA) (protocol C). Thereafter, cells were cultured until day 9 with the same media containing Nic 4.9 mmol/L (Sigma-Aldrich Química SA) and 20 ng/ml HGF (PeproTech EC Ltd., London, UK). The maturation step of hepatocyte-like cells was performed with protocol B cultures, assaying three different maturation protocols (B1, B2, and B3) (Fig. 1A). Cells were cultured with the same medium as in step 2 (protocol B1), or with 20 ng/ml HGF, 20 ng/ml OSM (PeproTech EC Ltd.), 1 μmol/L Dex (Merck Pharma, Mollét del Vallés, Spain), and ITS + premix (BD Biosciences, Madrid, Spain) (protocol B2) and finally with 20 ng/ml OSM, 1 μmol/L Dex, and ITS + premix (protocol B3).

Figure 1.

Hepatogenic induction protocols for ASC with sequential addition of growth factors, cytokines, and hormones and ASC cell morphology changes during hepatogenic induction. (A) Confluent undifferentiated ASC cultures at passage 2 were treated with different combinations of exogenous factors over a period of 16 days. Phase-contrast microscopy analysis of ASC on day 0 (B), day 9 of protocol B (C), and on day 16 of protocol B3 with (D) or without (E) supplementation of culture medium with matrigel. Original magnification of 20x for all the pictures (scale bars: 50 μm). Abbreviations: ASC, adipose-derived stem cells; EGF, epidermal growth factor; FGF, fibroblastic growth factor; BMP, bone morphogenic protein; HGF, hepatic growth factor; Nic, nicotinamide; Dex, dexamethasone; ITS, insulin-transferrin-selenium.

Transduction of ASC with the Adenoviral Vector Encoding HNF4α

ASCs on day 7 of differentiation protocol B were infected with recombinant adenovirus for the expression of HNF4α (Ad-HNF4α) for 120 min at a multiplicity of infection (MOI) of 4 plaque-forming units/cell (9,31,49). Thereafter, cells were washed with PBS and fresh medium (protocol B) was added. Cells were analyzed 48 h posttransduction.

Flow Cytometry Analysis

A FACScalibur flow cytometer (Becton Dickinson, Madrid, Spain) and CellQuest (Becton Dickinson) software program were used for flow cytometric analysis, as previously described (54). To analyze the surface marker expression, cells from the initial sample, cultured cells and nonadhered cells, were labeled with monoclonal antibodies against human antigens CD13-PE, CD34-APC, CD45-FITC, CD90-APC (Becton Dickinsonn) and CD105-PE (Serotec, Madrid, Spain). Actinomycin D (7-AAD) (Sigma-Aldrich, Madrid, Spain) was used to exclude nonviable cells from the analysis for nonspecific staining.

RNA Extraction and Real-Time Quantitative PCR

Total RNA was extracted from undifferentiated (used as a control), differentiating, and differentiated cells using Trizol reagent, following the manufacturer's recommendations. Contaminating genomic DNA was removed by incubation with DNase I amplification grade. The amount of purified RNA was estimated by ribogreen fluorescence (Molecular Probes), and its purity was assessed by the absorbance ratio 260:280 nm. RNA integrity was examined by agarose gel electrophoresis. RNA (1 μg) was reverse transcribed as previously described (64,70). Prior to the reverse transcription step, a known amount of a nonhepatic mRNA external standard was added to samples (0.01 ng of firefly luciferase mRNA per μg of total RNA). After reverse transcription, samples were amplified using the appropriate primers and were real-time quantified by SYBR Green I fluorescence measurement. The amount of luciferase cDNA in each reverse transcribed sample was estimated, and the efficiency of the RT step was determined (amount of exogenous mRNA added/amount of cDNA synthesized × 100) using a standard curve of firefly luciferase DNA. Diluted cDNA (3 μl) was amplified in a rapid thermal cycler (LightCycler Instrument, Roche Diagnostics) in 15 μl of LightCycler DNA Master SYBR Green I (Roche Molecular Biochemicals), 5 mM MgCl₂, and 0.3 μM of each oligonucleotide. Optimal MgCl₂ concentration was empirically determined for each set of primers. Gene-specific primers were designed using the OLIGO software. Whenever possible, primer sequences were chosen to span exon boundaries. The sequence of both the forward and reverse primers and the expected sizes of the PCR-amplified DNA are listed in Table 1. In parallel, we always analyzed the mRNA concentration of the human housekeeping porphobilinogen deaminase (PBGD, hydroxymethylbilane synthase) as an internal normalization control. A stable expression of PBGD was validated by comparing with the TATA box binding protein (TBP) expression as a second constitutive control gene. We found that the expression ratio of these two internal control genes was practically constant in the cell types and under the conditions investigated. A nontemplate-negative control (H₂O) was also run for each cDNA-specific primer set. PCR amplicons were confirmed to be specific by size (agarose gel electrophoresis) and melting curve analyses. After denaturing for 30 s at 95°C, amplification was performed with 40 cycles of 1 s at 94°C, 5 s at 62°C, and 15–20 s at 72°C. Reproducibility of the measurements was assessed by conducting reactions in triplicate. Real-time monitoring of the PCR reaction and the precise quantification of the products in the exponential phase of the amplification were performed with the LightCycler quantification software, according to the manufacturer's recommendations. Relative expression levels were calculated with the Relative Quantification Analysis software. Briefly, the efficiency of each PCR reaction was estimated from a serially diluted human liver cDNA standard curve. Based on the PCR efficiency values, the software calculates the relative concentration of target and reference (PBGD) cDNAs and their ratio. Moreover, a positive sample with a stable ratio of target and reference cDNAs (a calibrator) was included in each PCR run to normalize all the samples within one run and to provide a constant calibration point among several amplification runs.

Table 1.

Sequence of Primers Used for RT-PCR and Length of Fragments

Primer	Sequences	Product (bp)
ALB	Upper 5′-TGAGAAAACGCCAGTAAGTGAC-3′	265
	Lower 5′-TGCGAAATCATCCATAACAGC-3′et
hy1	Upper 5′-CACACATACCGCTCCCGAACC-3′	189
	Lower 5′-GCTGATGCCCTCACACTTGACC-3′
CYP3A4	Upper 5′-CCTTACATTACACACCCTTTGGAAGT-3′	382
	Lower 5′-AGCTCAATGCATGTACAGAATCCCCGGTTA-3′
C/EBPβ	Upper 5′-CTCGCAGGTCAAGAGCAAG-3′	271
	Lower 5′-CTAGCAGTGGCCGGAGGCGAGC-3′
HNF4α	Upper 5′-GCCTACCTCAAAGCCATCAT-3′	275
	Lower 5′-GACCCTCCCAGCAGCATCTC-3′
PGC1α	Upper 5′-AATGTGTCTCCTTCTTGTTCTT-3′	229
	Lower 5′-GGTGTCTGTAGTGGCTTGA-3′
C/EBPα	Upper 5′-GTGGAGACGCAGCAGAAG-3′	450
	Lower 5′-TTCCAAGGCACAAGGTTATC-3′
Hex	Upper 5′-CGGACGGTGAACGACTACA-3′	233
	Lower 5′-TTTGACCTGTCTCTCGCTGA-3′
GATA6	Upper 5′-AGAGCACCAATCCCGAGAA-3′	261
	Lower 5′-AAATAAAGGCACGAGAATCACG-3′
PEPCK	Upper 5′-CTGGGCGATGGGGAGTTTGTCA-3′	494
	Lower 5′-GATTGTGTTCTTCTGGATGGTCTTG-3′
APOE	Upper 5′-GTTCTGTGGGCTGCGTTGCT-3′	269
	Lower 5′-TTGTTCCTCCAGTTCCGATTT-3′
C3	Upper 5?CGCCCGTGATACACCAAG-3′	278
	Lower 5′-CTCCCAGCGGTTCTTATCTTT-3′
AGTR1	Upper 5′-CCACTATGGGCTGTCTACACA-3′	185
	Lower 5′-GCAGGTGACTTTGGCTACA-3′
ACSL1	Upper 5′-GAACTATTTCAGGTCGCAGAT-3′	158
	Lower 5′-TGGACAAGTCAAACACGAAC-3′
ID3	Upper 5′-CCTGTCGGAACGCAGTCTG-3′	156
	Lower 5′-CGATGACGCGCTGTAGGATT-3′
α2M	Upper 5′-CACCCTGTTGTCCGCAATG-3′	275
	Lower 5′-GATAAGCGAGGAGCACATAGG-3′
SEPP1	Upper 5′-CTCCATCGCCTCATTACCATC-3′	142
	Lower 5′-CCTATGCTGACCCTTGTGCTTA-3′
ALDOB	Upper 5′-TGAGGATTGCCGACCAGTGT-3′	317
	Lower 5′-CCCAGAAGAACCCGTGTGAAC-3′
CYP1A2	Upper 5′-AACAAGGGACACAACGCTGAAT-3′	431
	Lower 5′-GGAAGAGAAACAAGGGCTGAGT-3′
APOCIII	Upper 5′-ATGAAGCACGCCACCAAGAC-3′	296
	Lower 5′-TGGGGTAGGAGAGCACTGAGA-3′
hPBGD	Upper 5′-CGGAAGAAAACAGCCCAAAGA-3′	294
	Lower 5′-TGAAGCCAGGAGGAAGCACAGT-3′

Gene Expression Profiling by Microarray Analysis

RNA Preparation

Total RNA was purified from matched-pair untreated (control, day 0) and treated (day 16, protocol B3) ASCs from three different donors, and their expression profile was analyzed with the Human Genome U133 Plus 2.0 Array (Affymetrix, Santa Clara, CA). Total RNA was extracted and purified with Trizol reagent following the manufacturer's recommendations. The amount and integrity of purified RNA was estimated by microcapillary electrophoresis (2100 Bioanalyzer, Agilent Technologies, CA), and purity was assessed by the absorbance ratio 260:280 nm.

Amplification and Hybridization

Double-stranded cDNA was generated from total RNA and transcribed to obtain biotin-labeled cRNA according to the Affymetrix standard protocol. cRNA (15–20 μg) was fragmented, combined with hybridization and spike controls, and hybridized with a preequilibrated Affymetrix chip for 16 h. Following hybridization, arrays were washed and stained with streptavidin phycoerythrin conjugate using an automated GeneChip® Fluidics Station 450, and were then scanned with a GeneChip® Scanner 3000 using a 570 nm excitation wavelength laser. Target sample preparation and microarray hybridization and scanning were performed at the Gene Analysis Service, Central Research Unit (UCIM), Faculty of Medicine, University of Valencia (Spain).

Quality Control, Data Analysis, and Statistics

Microarray quality control assessment and data acquisition were performed with the GeneChip® Operating Software (Affymetrix). To permit the correct alignment of the software analysis grid, a control oligo (B2 oligo) was hybridized and used to indicate the boundaries of the probe area. The other quality control metrics used included the average background and noise (Raw Q) of the measured probe cell intensities, the ratio of expression of the 3′ probe set to the 5′ probe set of the housekeeping control gene GAPDH, and the percentage of all the probe sets in the array detected as present. Perfect match (PM) and mismatch probe intensities were used to generate a detection call (present, marginal, or absent) for each probe set, according to the microarray suite version 5 (MAS5) algorithm. Normalization among the different microarray data files and comparative data analysis were performed with DChip (Microarray Analysis Software Package, Harvard School of Public Health). The one-way analysis of variance (ANOVA) followed by Tukey's post hoc test was used to detect significant differences in gene expression between biological conditions (Bonferroni post hoc testing gave similar results). In addition, unsupervised hierarchical clustering analysis and a principal component analysis (PCA) were performed with the DChip software to establish nonforced groups of samples.

Induction of Fat Overload in HepG2 Cells and Human Hepatocytes and Determination of Lipid Content

HepG2 and 24-h cultured hepatocytes were exposed to a free-fatty acid (FFA) mixture (2:1 ratio of oleate and palmitate; Sigma) (23). A 1 mM FFA solution was prepared in culture medium containing 1% bovine serum albumin. Cells were incubated with the FFA mixture for 14 h. Lipid content was determined fluorimetrically by Nile red (Sigma) staining in HepG2 and 24-h cultured hepatocytes exposed and nonexposed to FFA and in ASC at day 16 of protocol B3, as described in detail elsewhere (23).

Glycogen Content Evaluation

Synthesis and intracellular glycogen accumulation was triggered in ASC at day 0 and at day 16 of protocol B3 by a 6-h incubation with a fresh culture medium containing 1 μM insulin and 30 mM glucose. Glycogen content was determined by an enzymatic assay, described in detail elsewhere (24).

Drug Metabolism Assessment

P450 activity assays were performed at different time points by the direct incubation of ASC at day 0 and ASC at day 16 of protocol B3 with specific substrates for P450 isozymes (42). Those metabolites that formed during the activity assays were quantified by using HPLC-MS/MS (Micromass Quattro Micro; Waters, Milford, MA, USA) in the electrospray ionization mode interfaced with an Alliance 2795 HPLC (Waters Chromatography), as described previously (43). Activity values were expressed as pmoles of product formed per hour and per milligram of protein.

Immunoblotting

Cell lysates and extracellular culture media of ASC at days 0, 2, 9, and 16 of protocol B3 were analyzed for APOE, C3, and α2M by Western blot analysis. Proteins were resolved by 7.5% and 12% SDS-PAGE (30 μg protein/lane), transferred to Immobilion membranes (Millipore Corporation, Billerica, MA, USA), and incubated with anti-APOE (1:3000, Chemicon, Millipore Corporation), anti-α2M (1:1000, DakoCytomation, Denmark), and anti-C3 (1:1000, Quidel, San Diego, CA, USA). After washing, blots were developed using an enhanced chemiluminescence kit (Amersham Biosciences UK, Ltd., Little Chalfont, Buckinghamshire, UK).

Results

Flow Cytometry Analysis of ASC From Human Adipose Tissue

To characterize the ASC population, we examined the cell surface marker expression of ASCs from 10 patients undergoing lipectomy using flow cytometry. The major cell population obtained in primary culture was positive for the mesenchymal markers CD13 (89.6 ± 13.9%), CD90 (94.1 ± 6.3%), and CD105 (77.6 ± 28.9%), whereas it was negative for the hematopoietic marker CD133 and only a small percentage of the population (10.9 ± 10.1%) was positive for the hematopoietic marker CD45. The characteristic surface stem cell marker, CD34, showed a higher percentage of positive cells in the biopsy (87.3 ± 8.4%) than in the primary culture (56.9 ± 33.7%), and CD34-positive cells decreased with culture time. Concerning the expression of embryonic markers, the primary culture was negative for SSEA-4, Oct-4, and Sox2. Briefly, our ASC isolation procedure results in a majority MSC population positive for CD90, CD105, and CD13 and negative for the hematopoietic markers CD45 and CD133 and negative for the embryonic markers SSEA-4, Oct-4, and Sox2, which coincides with our previous findings (87).

Morphological Changes of ASCs During Hepatic Transdifferentiation

The morphology of ASCs was analyzed at day 0, at the end of differentiation (day 9), and the maturation (day 16) steps. On day 0, cells exhibited a fibroblastic-like morphology that was observed until day 9 (Fig. 1B, C). OSM and Dex treatment induced a noticeable change to a polygonal shape, characteristic of mature hepatocytes (Fig. 1D). Adding Matrigel to the culture medium maintained the cell culture up to 1 month, showing a higher cellular proliferation rate and longer survival compared with Matrigel-free cultures (data not shown). Besides, Matrigel induces a more pronounced polygonal shape of ASC-derived hepatocyte-like cells (Fig. 1E).

Induction of Early Hepatogenic Differentiation of ASC by FGFs, BMPs, and HGF

To investigate which factors act at defined stages to control hepatogenic differentiation of ASCs, we assessed the expression of multiple transcription regulators. Total RNA from ASC was isolated on days 0, 2, 9, and 16 of culture. Undifferentiated cells (day 0) were used as controls. Expression levels of homeobox protein Hex (Hex) and GATA binding protein 6 (GATA6), CCAAT/enhancer binding protein β (CEBPβ) and alpha (CEBPα) and peroxisome proliferator-activated receptor gamma, co-activator 1α (PGC1α) significantly increased in the first differentiation protocol steps. As Figure 2 illustrates, similar increases in the expression levels of the five hepatic transcription factors were observed in the three differentiation media (A, B, and C), and the maximum was reached at the end of the differentiation step (day 9). However, the expression was similar or slightly lower in the maturation step (data not shown), suggesting that these transcription factors may play a role in the early transdifferentiation process of ASCs toward a hepatic phenotype.

Figure 2.

Upregulation of transcription factors in the early hepatic differentiation of ASC. Specific mRNA levels of Hex, GATA6, CEBPα, CEBPβ, and PGC1α were measured by real-time quantitative RT-PCR. The results were normalized to the housekeeping PBGD quantification values. Data were expressed as being relative to undifferentiated cells (day 0) and represent the mean ± SD (n ≥ 4). Abbreviations: ASC, adipose-derived stem cells; Hex, homeobox protein Hex; GATA6, GATA binding protein 6; CEBPβ, CCAAT/enhancer binding protein β; CEBPα, CCAAT/enhancer binding protein α; PGC1α, peroxisome proliferator-activated receptor gamma, coactivator 1α; PBGD, porphobilinogen deaminase; RT-PCR, real time-polymerase chain reaction.

To further demonstrate that these transcription factors are associated with the upregulation of typical hepatic genes, we determined the mRNA expression level of a battery of hepatic genes by RT-PCR and found that several could be selected as early markers of hepatocye differentiation: albumin (ALB), early hepatic marker regulated by CEBPα and CEBPβ, α-2-macroglobuline (α2M), an acute phase protein predominantly synthesized in the liver, complement protein C3 (C3), synthesized by the liver and involved in the complement system and acute phase response, and selenoprotein P 1 (SEPP1) secreted by the liver to act as an antioxidant. The expression levels of ALB, α2M, C3, and SEPP1 markedly increased in cells cultured with the protocols A and B, reaching maximal levels on day 9 (Fig. 3); increases of the ALB expression coincided with the maximal expression of CEBPα and CEBPβ but the cells in the presence of FGF4 (protocol B) showed a higher rate of survival and proliferation (data not shown). However, cells cultured with medium supplemented with BMPs (protocol C) generally showed a smaller increase in the mRNA level of all the hepatic markers on days 2 and 9. This probably suggests that BMPs do not play a critical role in hepatic transdifferentiation of ASC to the hepatic phenotype.

Figure 3.

Induction of early hepatic markers during the hepatogenic induction of ASC. Specific mRNA levels of ALB, α2M, C3, and SEPP1 were measured by real-time quantitative RT-PCR. Results were normalized to PBGD values. Data were expressed as being relative to undifferentiated cells (day 0) and represent the mean ± SD (n ≥ 4). Abbreviations: ASC, adipose-derived stem cells; ALB, albumin; α2M, α-2-macroglobulin; C3, complement 3; SEPP1, selenoprotein P plasma 1; PBGD, porphobilinogen deaminase; RT-PCR, real time-polymerase chain reaction.

Induction of Hepatic Maturation of ASC by HGF, OSM, and Dex

The results presented in Figure 3 show that protocols A and B seem to satisfy the requirements to induce an early hepatogenic differentiation of ASCs. However, we selected protocol B based on reports showing that several FGFs, including FGF4, are involved in the regulation of liver development. Thus, the addition of FGF4 (protocol B) would likely enhance the transdifferentiation process. Then, we investigated different conditions for hepatocyte maturation in ASCs cultured with protocol B. A battery of hepatocyte-specific genes were analyzed by quantitative RT-PCR and found that several could be selected as late markers of hepatocye differentiation: cytochrome P450 3A4 (CYP3A4), apolipoprotein E (APOE), acyl-CoA synthetase long-chain family member 1 (ACSL1), angiotensin II receptor type 1 (AGTR1), and HNF4α. As Figure 4 depicts, undifferentiated ASC (day 0) expressed very low or negligible levels of the five late hepatic markers, whereas an increase was already seen on day 9 after commitment and differentiation. ASCs cultured with the same medium until day 16 (protocol B1) showed no further increase in the expression levels of any late hepatic marker. The addition of OSM and Dex combined with HGF and ITS (protocol B2) induced a remarkable increase in ACSL1, but the other markers showed no further upregulation. The largest increase in the mRNA level of CYP3A4, APOE, AGTR1, and HNF4α occurred in cells cultured with medium supplemented with OSM, Dex, and ITS (protocol B3), but deprived of HGF (Fig. 4). These results suggest that the most potent hepatogenic induction is produced by protocol B3.

Figure 4.

Induction of late hepatic markers during the differentiation of ASC. Specific mRNA levels of ACSL1, AGTR1, APOE, CYP3A4, and HNF4α were measured by real-time quantitative RT-PCR. The results were normalized to PBGD values. Data were expressed as relative to undifferentiated cells (day 0) and represent the mean ± SD (n ≥ 4). Abbreviations: ASC, adipose-derived stem cells; ACSL1, acyl-CoA synthetase long-chain family member 1; AGTR1, angiotensin II receptor, type 1; APOE, apolipoprotein E; CYP3A4, cytochrome P450 3A4; HNF4α, hepatocyte nuclear factor 4 a; PBGD, porphobilinogen deaminase; RT-PCR, real time-polymerase chain reaction.

Loss of Adipose Adult Stem Cell Phenotype

To investigate the loss of adipose adult stem cell phenotype during the hepatogenic transdifferentiation process, we assessed the mRNA expression level of both Thy1 (CD90), and the inhibitor of DNA binding 3 (Id3), until day 16 (Fig. 5). ASC-derived hepatocyte-like cells obtained with protocol B1 expressed similar levels of Thy1 and Id3 to undifferentiated ASC (Fig. 5). However, the expression levels of both markers significantly decreased in cells cultured with HGF, OSM, Dex, and ITS (protocol B2), while a much more significant decrease was induced by medium supplemented with OSM, Dex, and ITS (protocol B3), particularly in Id3 mRNA. The results suggest that OSM and Dex are apparently critical to induce the loss of the MSC phenotype to levels as low as those in the human liver (data not shown). However, the presence of HGF caused no major effect on the downregulation of Thy 1 and Id3.

Figure 5.

Loss of undifferentiated stem cell phenotype in hepatocyte-like derived cells. Specific mRNA levels of Thy1 and Id3 were measured by real-time quantitative RT-PCR. The results were normalized to PBGD values. Data were expressed as relative to undifferentiated cells (day 0) and represent the mean ± SD (n ≥ 4). Abbreviations: Thy1, CD90; Id3, inhibitor of DNA binding 3; PBGD, porphobilinogen deaminase; RT-PCR, real time-polymerase chain reaction.

Gene Expression Patterns in ASC-Derived Hepatocyte-Like Cells

To gain deeper insight into the molecular events during the transdifferentiation of ASC, we compared three matched-pair untreated (day 0) and treated (day 16, protocol B3) ASCs by whole human genome expression analysis. The ANOVA test detected a differential gene expression between biological conditions. We selected genes whose expression varied at least twofold between undifferentiated and differentiated cells, and whose value was p < 0.05. The comparative analysis showed a total of 657 differentially expressed probes representing 498 different genes.

Hierarchical clustering of samples using the differentially expressed gene set segregated samples into two major clustering branches. All the replicate samples of the untreated group self-clustered into one branch while the differentiated samples self-clustered into another (Fig. 6A). The PCA of the whole gene profiles grouped samples into only two clusters, as in the hierarchical clustering analysis (Fig. 6B). Clustering and PCA analyses demonstrated that the transdifferentiation process leads to a cellular population with a substantially different gene expression pattern from the same cell population before treatment.

Figure 6.

Global expression microarray analysis of ASC during hepatogenic induction. Total RNA was purified from matched-pair untreated (day 0) and treated (day 16, protocol B3) ASCs from three different donors and their expression profile analyzed with the Human Genome U133 Plus 2.0 Array (Affymetrix). (A) Hierarchical clustering analysis of differentially expressed genes filtered by ANOVA with a value of p < 0.05 (light gray, downregulation genes; dark gray, upregulation genes). (B) Nonforced grouping of samples by principal component analysis (light gray, untreated cells on day 0; dark gray, hepatocyte-like cells on day 16, protocol B3). Abbreviation: ASC, adipose-derived stem cells.

Next we used the Gene Ontology Tree Machine (GOTM) function from Vanderbilt University to find overrepresented ontology categories in the lists of genes differentially expressed between our biological conditions (n ≥ 3, p > 0.01). We obtained a list of enriched gene ontology categories, which suggests that, during our sequential differentiation protocol, ASCs undergo activation and modulation of processes involved in cell differentiation and development (Table 2). Subsets of differentially expressed genes are also associated with the regulation of biological and physiological processes (positive and negative), the control of cell survival and the remodeling of extracellular matrix, cytoskeleton and cell morphology (Table 2). Several genes are also associated with diverse signal transduction mechanisms and with more specific hepatic functions, like those related to complement proteins, acute phase response proteins, lipid transport and cholesterol homeostasis (Table 2).

Table 2.

List of Gene Ontology (GO) Categories Significantly Overrepresented in ASC Hepatocyte-Like Cells

GO Level	GO Category Description	Observed Gene Number	Expected Gene Number	Ratio of Enrichment	Significance of Enrichment
2	Regulation of biological process	100	78.1	1.3	0.0018
3	Regulation of physiological process	88	69.5	1.3	0.0057
4	Positive regulation of physiological process	23	12.2	1.9	0.0024
4	Negative regulation of physiological process	27	15.4	1.8	0.0030
4	Negative regulation of cellular process	31	16.4	1.9	0.0004
5	Negative regulation of cellular physiological process	27	14.8	1.8	0.0017
4	Regulation of organismal physiological process	10	3.0	3.3	0.0009
5	Positive regulation of organismal physiological process	6	1.3	4.6	0.0018
2	Development	59	40.5	1.5	0.001
3	Regulation of development	9	1.7	5.3	0.00004
3	Anatomical structure development	47	28.8	1.6	0.0004
4	Positive regulation of development	4	0.5	7.6	0.0017
4	Cell development	9	3.0	3.0	0.0032
4	Organ development	29	12.6	2.3	0.00002
3	Dell differentiation	25	12.1	2.1	0.0005
4	Regulation of cell differentiation	6	1.2	5.0	0.0012
5	Positive regulation of cell differentiation	3	0.3	8.8	0.0042
5	Negative regulation of cell differentiation	3	0.4	7.3	0.0074
5	Regulation of cell proliferation	14	6.7	2.1	0.0075
2	Extracellular region	42	23.4	1.8	0.0001
2	Extracellular region part	30	15.3	2.0	0.0003
2	Extracellular matrix	21	6.3	3.3	0.000001
2	Extracellular matrix part	9	2.2	4.2	0.0003
3	Extracellular matrix (sensu Metazoa)	20	6.3	3.2	0.000005
3	Basement membrane	6	1.2	4.8	0.0014
4	Cell substrate adhesion	6	1.3	4.5	0.0020
5	Cell matrix adhesion	6	1.3	4.5	0.0020
4	Morphogenesis	24	13.4	1.8	0.004
5	Cell morphogenesis	14	5.9	2.4	0.002
4	Cytoskeletal protein binding	16	7.5	2.1	0.003
5	Actin binding	13	5.1	2.5	0.002
7	Actin filament-based process	10	3.7	2.7	0.004
8	Actin cytoskeleton organization and biogenesis	9	3.4	2.7	0.007
3	Death	23	13.0	1.8	0.006
4	Cell death	23	13.0	1.8	0.006
5	Programmed cell death	23	12.5	1.8	0.003
6	Apoptosis	23	12.4	1.9	0.003
5	Regulation of programmed cell death	16	8.0	2.0	0.007
6	Regulation of apoptosis	16	7.9	2.0	0.006
2	Signal transducer activity	55	37.6	1.5	0.002
3	Receptor activity	43	28.4	1.5	0.004
4	Growth factor binding	6	1.5	4.1	0.003
4	Insulin receptor substrate binding	3	0.1	30.0	0.00006
5	Positive regulation of signal transduction	7	2.2	3.2	0.006
5	Transmembrane receptor protein kinase activity	6	1.8	3.3	0.009
5	Regulation of I-κB kinase/NF-κB cascade	7	1.9	3.8	0.003
6	Positive regulation of I-κB kinase/NF-κB cascade	7	1.7	4.1	0.001
7	I-κB kinase/NF-κB cascade	8	2.5	3.2	0.004
7	Phosphoinositide-mediated signaling	6	1.7	3.5	0.008
3	Lipid transporter activity	7	1.5	4.6	0.0008
4	Cholesterol homeostasis	3	0.2	17.7	0.0005
6	Complement activation	5	0.8	6.3	0.0011
6	Complement activation/alternative pathway	3	0.3	10.3	0.0026
6	Activation of plasma proteins during acute inflammatory response	5	0.8	6.3	0.0011

In order to identify GO categories with significantly enriched gene numbers, the distribution of genes in the selected gene list in each GO category is compared to those in the reference gene list (i.e., all genes in the Human Genome U133 Plus 2.0 Array). If K out of the N genes in the reference list, and k out of the n genes in the selected list, are in category X, then, if the n selected genes are effectively a random sample uniformly selected from the reference gene list, the expected value of k is ke = (n/N)K (expected gene number). If, on the other hand, the observed k (Observed gene number) exceeds the above expected value, category X is said to be enriched, with a Ratio of enrichment R = k/ke. GOTM software reports only those enrichments that are statistically significant as determined by hypergeometric test (Significance of enrichment).

These results demonstrate that our sequential protocol triggers the adjustment of multiple pathways involved in differentiation, morphogenesis, survival, and hepatic-associated functions in ASCs.

Reactivation of Hepatic Genes by Adenovirus-Mediated Reexpression of HNF4α in ASCs

HNF4α is an essential transcription factor for the constitutive expression of key hepatic genes such as phosphoenolpyruvate corboxykinase (PEPCK), apolipoprotein C3 (APOCIII), aldolase B (ALDOB), and cytochrome P450 1A2 (CYP1A2) (49). Even though a noticeable increase of HNF4α expression was observed on day 16 of the differentiation protocol (Fig. 4), the mRNA levels of these particular HNF4α target genes were negligible (data not shown), suggesting that the expression level of HNF4α was insufficient to induce a functional transcriptional activation of all its hepatic target genes.

To investigate this possibility, ASCs were transduced with an adenoviral vector encoding the transcription factor HNF4α (Ad-HNF4α) on day 7 of the differentiation protocol B. Infection of ASCs with a control adenovirus expressing the GFP demonstrated 80–90% transduction efficiency. Moreover, transduction of ASCs with Ad-GFP or Ad-pAC (insertless adenoviral vector) neither modified the morphology nor the expression of differentiation marker genes (data not shown). Adenoviral-mediated transfection of HNF4α in ASCs caused a highly significant increase in the expression of HNF4α compared with the mRNA levels induced by protocol B3 (the most hepatogenic protocol) (2416 ± 399% vs. 520 ± 254%, percentage of day 0 undifferentiated ASCs). Concomitantly, a significant increase in the expression of its target genes, APOCIII, ALDOB, and CYP1A2, and particularly PEPCK, was observed. As a reference, we compared the levels attained with those in the differentiated hepatoma HepG2 and found comparable or, as with PEPCK, much higher values (Fig. 7). This finding suggests that the forced expression of HNF4α after early hepatogenic differentiation of ASC can effectively improve maturation and the hepatic phenotype.

Figure 7.

Transcriptional activation of liver genes by the forced expression of HNF4α ASCs was transduced with an adenoviral vector encoding the transcription factor HNF4α (Ad-HNF4α on day 7 of the hepatogenic protocol B and were collected on day 9. Specific mRNA levels of PEPCK, ALDOB, APOCIII, and CYP1A2 were measured by real-time quantitative RT-PCR and were compared with those in the differentiated hepatoma HepG2. The results were normalized to PBGD values. Data were expressed as relative to undifferentiated cells (day 0) and represent the mean ± SD (n + 4). Abbreviations: ASC, adipose-derived stem cells; HNF4α, hepatocyte nuclear factor 4 a; PEPCK, phosphoenolpyruvate carboxykinase; ALDOB, aldolase B; APOCIII, apolipoprotein C3; CYP1A2, cytocrome P450 1A2; PBGD, porphobilinogen deaminase; RT-PCR, real time-polymerase chain reaction.

Lipid Content in ASC-Derived Hepatocyte-Like Cells

Some of the identified marker genes at early and late differentiation could also be associated with adipogenesis. In order to rule out that our hepatogenic differentiation protocol is inducing concomitant adipogenesis, we investigated whether ASC-derived hepatocyte-like cells accumulate intracellular lipids. To confirm that lipid droplets did not develop while the transdifferentiation process took place, we compared the lipid content in ASC-derived hepatocyte-like cells at day 16 in culture under protocol B3, with control (non-FFA-overloaded), and FFA-overloaded HepG2 cells and human hepatocytes. Fat-overloaded HepG2 and hepatocytes showed a fivefold increase of lipid content, while differentiated 16-day ASC presented similar negligible levels to the control HepG2 cells and human hepatocytes (Fig. 8).

Figure 8.

Lipid content evaluation of ASC under protocol B3 for hepatogenic transdifferentiation. Lipid content in FFA-overloaded HepG2 and human hepatocytes was compared with that in 16-day ASC and nonoverloaded HepG2 and hepatocytes. Lipid content was estimated by Nile Red staining after a 14-h induction in overloaded samples. The results are expressed as fluorescence arbitrary units/mg protein. Abbreviations: ASC, adipose-derived stem cells; FFA, free-fatty acid.

Hepatic Functionality Assessment

The results show that glycogen synthesis, after incubating cells with insulin and high concentrations of glucose, was negligible in ASC at day 0 (1.1-fold increase), while differentiated ASC (day 16 of protocol B3) synthesized and stored glycogen at similar levels to the 24-h cultured human hepatocytes (6.3- and 4.1-fold increase, respectively). To demonstrate the functional capacity of differentiated ASC, the activity of several enzymes involved in drug metabolism was analyzed. Table 3 shows an increase in the levels for diclofenac hydroxylation, bufuralol 1′-hydroxylation, S-mephenytoin 4-hydroxilation, chlorzoxazone 6-hydroxylation, and ethoxyresorufin O-deethylation at day 16 of protocol B3. To further confirm the functionality of ASC-derived hepatocyte-like cells, we determined the APOE, α2M, and C3 protein levels in both intracellular and extracellular compartments during the differentiation protocol by immunoblotting analysis. ASC-derived hepatocytes were able to synthesize and secrete these plasma proteins in a time-dependent manner, as shown in Figure 9.

Figure 9.

Immunoblotting analysis of hepatocyte specific markers during the hepatogenic transdifferentiation of ASC. Cells were cultured under protocol B3 and monolayers were washed and fresh medium was added at the indicated times. After 6-h incubation, aliquots of the culture medium were taken and the cellular protein of the monolayers was extracted by means of a ready-to-use cell lysis solution (M-PER, Pierce). Total cell protein (30 μg) or 30 μl of extracellular medium were loaded per lane. Coomasie blue staining of membranes was performed to check for equal loading (not shown). Western blot analysis indicated that protocol B3 induced the synthesis and secretion of APOE, C3, and α2M in ASC-derived hepatocyte-like cells in a time-dependent manner. Abbreviations: ASC, adipose-derived stem cells; APOE, apolipoprotein E; α2M, α-2-macroglobulin; C3, complement 3.

Table 3.

Enzyme-Selective P450 Activities in ASC Before and After Hepatic Transdifferentiation

Activity	Day 0 (pmol/h/mg Protein)	Day 16 Protocol B3 (pmol/h/mg Protein)
Diclofenac hydroxylation	0	0.275 ± 0.012
Bufuralol 1′-hydroxylation	0.075 ± 0.012	0.225 ± 0.012
S-mephenytoin 4-hydroxylation	0	0.162 ± 0.012
Chlorzoxazone 6-hydroxylation	0.137 ± 0.012	2.45 ± 0.775
Ethoxyresorufin O-deethylation	1.087 ± 0.237	1.850 ± 0.087

Discussion

Human ASCs have been reported to transdifferentiate in vitro into several cell lineages (20,102), including hepatocyte-like cells (3,5,76,87). Therefore, the multipotent nature of ASCs, together with their easy extraction and isolation, make these cells a promising alternative for many tissue engineering strategies, autologous cell-based therapies, and preclinical drug testing.

Great effort has been made in the last decade to identify the factors regulating the onset of hepatogenesis and embryonic liver development (14,99). Based on these findings, different strategies have been attempted to induce hepatogenic transdifferentiation of ASCs into functional hepatocyte-like cells. A first approach was the use of a cocktail of exogenous factors (39,76), while other reports (3 –5,80), including ours (87), used multistep transdifferentiation protocols by sequential exposure of ASC to growth factors, cytokines, and hormones, reflecting their temporal expression during liver embryogenesis in vivo. All these reports have demonstrated the hepatogenic potentiality of ASC by the detection of classical hepatic markers and biochemical functions (3 –5,76,78,80,87). However, the hepatogenic media used and the differentiation markers determined in these previous studies differed remarkably.

Induction of hepatogenic transdifferentiation in vitro of ASC is a process that likely differs from embryonic liver development. Consequently, in vitro differentiation protocols attempting to experimentally reproduce the embryonic development of the liver may not adequately mimic the cross-lineage commitment between ASCs and hepatocytes during dedifferentiation–transdifferentiation processes.

BMPs are not only important for liver development, but also for hepatic specification of mouse and zebrafish embryonic stem cell-derived definitive endoderm (25, 79). However, and in agreement with our findings, a lack of requirement for BMP4 addition was demonstrated in the generation of hepatocytes from human embryonic stem cells (hESCs) (1). The disparity between these observations could reflect species variations. Alternatively, it has been shown that addition of HGF restored growth in mutant mouse liver explants impaired in BMP signaling via TGF-β receptors, suggesting that HGF may provide parallel signals to TGF-β-induced pathways (92).

Different FGF signals appear to initiate distinct liver development phases during mammalian organogenesis (33). The positive influence in hepatic differentiation of FGF2 (45,61,87) and, more recently FGF4, either alone or combined with EGF (16) or HGF in ASCs, have been reported (3 –5,80). FGF4 is one of the factors secreted by septum transversum mesenchyme and cardiogenic mesoderm at the early stages of endoderm formation in liver development in vivo, suggesting that related molecules probably function in a redundant capacity (77). In addition, exogenous FGF4 has been recently reported to increase the rate at which MSCs proliferate, and has no significant effect on MSC pluripotency (18). Our results show that the most consistent early hepatic transdifferentiation of ASCs is achieved with the differentiation medium supplemented with both FGF2 and FGF4 combined with EGF, HGF, and Nic (protocol B).

HGF has been reported to play an essential role in the early stages of hepatogenesis (38,75) and in liver regeneration after partial hepatectomy (60). Therefore, HGF and Nic, which enhances the in vitro differentiation of fetal liver cells (51,74), were included in our hepatogenic protocol, as in others (3,39,45,76,87). Our results point to a crucial role for HGF in the commitment and differentiation of ASCs, but it is detrimental for the last maturation step. Some studies show induction of hepatocyte-like phenotype after exposure of hMSC to HGF alone (91), or in cocktails combined with other factors (39,76,80). However, many reports like ours propose multistep hepatogenic protocols in which hMSC are exposed to HGF to induce hepatic commitment and differentiation, but not at further maturation stages of hepatocyte-like cells (3 –5,61,80,87).

Our results also demonstrate a critical role for OSM and Dex during the maturation step. This finding is consistent with the role reported for OSM to promote hepatoblasts differentiation to mature hepatocytes (56,86), that OSM in combination with glucocorticoids is required to induce maturation of liver stem and progenitor cells (34,35), and that hormones, glucocorticoids, and insulin are involved in the late maturation stage leading to increased liver-specific gene expressions (11).

Another key point investigated was the relevance of the extracellular matrix (ECM). It has been reported that ECM components contribute to hepatoblast differentiation (14,86). Our results show that supplementing culture medium with Matrigel induced a higher cellular proliferation rate and longer survival of ASC-derived hepatocyte-like cells.

The term transdifferentiation refers to the phenotypic change of one differentiated cell type to another lineage. It has been suggested that precommitted cells could change their phenotype during transdifferentiation by dedifferentiating into a primitive stem cell stage, perhaps through genome reprogramming (81). We did a whole genome expression analysis and observed a large readjustment in the expression profile of ASC during hepatogenic transdifferentiation with many genes related to development, proliferation, and differentiation undergoing important expression adjustment, which suggests a profound reprogramming event. Most of the findings agree with those previously reported by Yamamoto et al. (94).

Several transcription factors involved in liver development have been characterized, many of which were originally identified as regulators of liver-specific genes in the adult liver. GATA6 is the most critical player to regulate the differentiation of hepatoblasts (83,99). Hex, a homeobox transcription factor, is the earliest identified marker for endoderm that will form liver (7). Both are essential factors in hepatic specification and the commitment to hepatogenic cell fate (36,98,99). However, medium supplementation with FGF4 and BMPs during steps 1 and 2 led to no noticeable differences in the induction of GATA6 and Hex, thus suggesting that FGF4 and BMPs, present in protocols B and C, do not likely play a critical role in the induction of GATA6 and Hex during ASC transdifferentiation.

C/EBP factors are master regulators of liver development, and the early expression of C/EBPα and C/EBPβ exclusively in the early liver bud has been reported (89,93). It has also been reported that HGF induces the expression of C/EBP factors (86). PGC1α is an important coactivator for the control of metabolic homeostasis, and developmental regulation of PGC1α coactivation can be an essential step to build up the appropriate regulatory machinery for liver energy metabolism. In the mouse embryo liver, the expression of PGC1α is high (97) and parallels that of HNF4α and HNF6 (6). They are present in early development but their concentration steadily increases in late development up to a threshold level that is functional for some liver-specific target genes (6). Thus, synergism, interdependency, and threshold concentrations of transcription factors and coactivators could determine the time-specific expression of hepatic-specific genes during ASC transdifferentiation to hepatocytes.

Like GATA6 and Hex, CEBPα, CEBPβ, and PGC1α expressions during hepatogenic induction of ASCs revealed no significant response to FGF4 or BMPs. Moreover, the expression of these five transcription regulators increased during the first differentiation protocol steps (steps 1 and 2), but none showed further increase during step 3 (maturation). In this sense, the only transcription factor analyzed showing a significant upregulation during the maturation step was HNF4α.

HNF4α is essential for hepatocyte differentiation during later development and crucial for metabolic regulation and liver function (46). The prominent role of HNF4α during hepatocyte late development is consistent with a moderate increase of HNF4α expression at the end of the differentiation step (step 2), which becomes more noticeable in step 3 (maturation) of protocols B2 and B3. Our results are also consistent with previous observations, showing that GATA6 regulates HNF4α during development in a transcriptional cascade that controls the differentiation of the visceral endoderm (57). Moreover, C/EBPα and HNF3β are also important for sustaining the high HNF4α expression during liver development (15,41).

More recently, studies have shown that HNF4α is critical for hepatic morphogenesis. HNF4α deficiency in embryonic mouse liver results in abnormal tissue architecture and a lack of appropriate cell–cell contacts (62), while its overexpression induces cell polarity in F9 embryonic carcinoma (10), H5 hepatoma (82), and even in NIH3T3 fibroblast (62) cells. During ASC transdifferentiation, we observed a mesenchymal–epithelial transition. In steps 1 and 2, cells developed a fibroblastic-like flattened shape, but changed to a polygonal morphology in the presence of OMS and hormones in step 3 when the HNF4α expression reached the highest level.

Given the particular relevance of HNF4α in the hepatic phenotype, we decided to further investigate the level of functional activation attained in our ASC-derived hepatocytes. Despite a noticeable increase of HNF4α on day 16 of the differentiation protocol, we observed that the mRNA levels of some typical HNF4α target genes were negligible (data not shown), suggesting that the HNF4α expression level reached was likely insufficient to induce fully functional transcriptional activity. To investigate this possibility, ASCs were transduced with an adenoviral vector encoding the transcription factor HNF4α on day 7 of differentiation protocol B. This upgrading caused a highly significant increase in HNF4α mRNA compared with those attained with protocol B3 and, concomitantly, a significant increase in the expression of its target genes, APOCIII, ALDOB, and CYP1A2, and particularly PEPCK, was observed. Therefore, the adenovirus-mediated expression of HNF4α could be an effective procedure to attain full maturation of ASC-derived human hepatocytes.

After the commitment and differentiation steps, we found robust increases in the expression levels of ALB, α2M, C3, and SEPP1. Transcription of ALB, α2M, and C3 is controlled by C/EBP factors (21,32,48,52,53) and, accordingly, their induced expression correlated with the upregulation of CEBPα and CEBPβ in ASC-derived cells. SEPP1 is another hepatic-specific gene that is likely responsive to GATA binding proteins (84,95). Our previous work demonstrated that undifferentiated cells stain negatively by immunocytochemistry for albumin and AFP, while homogeneous positive staining was reported in ASC-derived hepatocyte-like cells (87). Despite the extended use of ALB as an early hepatic marker, our results demonstrate that α2M, C3, and SEPP1 are more robust markers for early hepatogenic differentiation of ASC.

During the maturation step of ASC, no further significant upregulation of ALB, α2M, C3, and SEPP1 was observed, whereas a very significant induction of CYP3A4, APOE, AGTR1, and ACSL1 mRNAs was noted, suggesting that these latter genes could be considered markers of mature hepatogenic differentiation. CYP3A4 and APOE expressions in the liver are controlled, at least in part, by HNF4α (2,50,88,90), and in agreement with these previous studies, we found a correlation between the upregulation of HNF4α and the increase in APOE and CYP3A4 during hepatogenic maturation of ASC.

ACSL1 and AGTR1 can also be considered liver-specific genes. The activation of ACSL1 in ASC could be induced by OSM, as it has been demonstrated that this cytokine can induce the expression of the homolog family members ACSL3 and ACSL5 in human hepatoma HepG2 cells (100). The mechanism responsible for the remarkable induction of AGTR in ASC deserves further investigation. The measurement of the selected late hepatic markers under different conditions demonstrated that exposure of ASC to OSM, Dex, and ITS (protocol B3) triggers the more efficient maturation toward the hepatic lineage.

The loss of the adipose MSC phenotype during the hepatogenic transdifferentiation process has not been previously investigated by others. Our results show that ASCs decrease in both the expression of the adult mesenchymal cell surface marker Thy1 and the undifferentiated status gene Id3. Thy1-positive cells were seen to decrease during early fetal rat liver development (29). Thy1 is also expressed in fetal hepatic cells throughout liver development (19), but is only expressed in oval cells in the adult liver, and not in mature hepatocytes (13,29,44,65). Similarly, Id3 has been shown to be a positive regulator of hepatoblast proliferation, but is also an inhibitor of their differentiation into hepatocytes during the development of the chick liver (59). Id3 is active in undifferentiated tissues and organs that are undergoing the differentiation process, but with proceeding development, Id3 shows a decreased expression and becomes restricted to various tissues, such as olfactory epithelium, kidney, and teeth (30). In fact, the knockdown of endogenous Id3 sensitizes epithelial cells to BMP, leading to a robust growth inhibition and induction of transdifferentiation (40). The expression of other immature markers like AFP was well characterized in ASC in a previous study from our laboratory (86). The expression of Thy1 and Id3 was elevated in undifferentiated ASC until day 9, but a significant decrease of both markers was observed in step 3 in hepatocyte-like cells obtained with protocols B2, and particularly B3, suggesting glucocorticoids and/or OSM are likely mediators of their repression during maturation of ASC-derived hepatic-like cells. Alternatively, the suppression of FGFs could trigger Thy1/Id3 downregulation.

In spite of a certain overlap in the expression profile between the adipogenic and hepatogenic lineages, the hepatocyte-like population at day 16 of protocol B3 shows very similar levels of lipid content to those detected in HepG2 and primary hepatocyte cultures, thus confirming that lipid droplet accumulation does not occur while hepatic transdifferentiation takes place.

Under hepatogenic conditions, we have shown that ASCs differentiate into cells with the characteristic morphology and phenotype of hepatocytes. In addition, the acquisition of functional characteristics of hepatocytes by transdifferentiated ASCs has been shown: the activity of several P450 enzymes involved in drug metabolism (Table 3), the capability to synthesize and store glycogen in a time-dependent manner, and synthesis and secretion of plasma proteins to culture medium (Fig. 9). These results do not completely demonstrate that our protocol triggers differentiation into fully mature hepatocytes. Comparison with data from cultured human hepatocytes indicates that the levels attained in differentiated ASC are still below those found in the adult hepatocyte (data not shown). However, it has been demonstrated that it is not necessary (or is even detrimental) to obtain fully differentiated hepatocytes when cell therapy is the final goal (76). Most in vivo models have aimed to investigate cell therapy in diseased livers that require repopulation by infused cells, which may require a less differentiated phenotype that will prove advantageous to donor cells to engraft and proliferate into the injured liver parenchyma (47).

Collectively, our experimental evidence demonstrates that the selected sequential protocol not only induces a hepatic-specific expression profile, but also leads to the loss of the mesenchymal undifferentiated phenotype. Moreover our results suggest, for the first time, the relevance of specific transcription factors (Hex, GATA6, C/EBPs, PGC1α, and HNF4α) in promoting hepatocyte-derived MSC and uncover new informative robust markers for the step-by-step transdifferentiation of hMSC into hepatocytes.

Footnotes

Acknowledgments

The authors acknowledge the financial support from the ALIVE Foundation, Instituto de Salud Carlos III, Fondo de Investigaciones Sanitarias (PI 070550), and European Commission. A.L. gratefully acknowledges Ministerio Ciencia e Innovación/Instituto de Salud Carlos III for a Miguel Server contract.

References

Agarwal

Holton

K. L.

Lanza

Efficient differentiation of functional hepatocytes from human embryonic stem cells. Stem Cells 26(5): 1117–1127; 2008.

Allan

C. M.

Taylor

J. M.

Two hepatic enhancers, HCR.1 and HCR.2, coordinate the liver expression of the entire human apolipoprotein E/C-I/C-IV/C-II gene cluster. J. Biol. Chem. 272(46): 29113–29119; 1997.

Banas

Teratani

Yamamoto

Tokuhara

Takeshita

Osaki

Kato

Okochi

Ochiya

Rapid hepatic fate specification of adipose-derived stem cells and their therapeutic potential for liver failure. J. Gastroenterol. Hepatol. 24(1): 70–77; 2009.

Banas

Teratani

Yamamoto

Tokuhara

Takeshita

Osaki

Kawamata

Kato

Okochi

Ochiya

IFATS collection: In vivo therapeutic potential of human adipose tissue mesenchymal stem cells (AT-MSCs) after transplantation into mice with liver injury. Stem Cells 26(10): 2705–2712; 2008.

Banas

Teratani

Yamamoto

Tokuhara

Takeshita

Quinn

Okochi

Ochiya

Adipose tissue-derived mesenchymal stem cells as a source of human hepatocytes. Hepatology 46(1): 219–228; 2007.

Beaudry

J. B.

Pierreux

C. E.

Hayhurst

G. P.

Plumb-Rudewiez

Weiss

M. C.

Rousseau

G. G.

Lemaigre

F. P.

Threshold levels of hepatocyte nuclear factor 6 (HNF-6) acting in synergy with HNF-4 and PGC-1alpha are required for time-specific gene expression during liver development. Mol. Cell. Biol. 26(16): 6037–6046; 2006.

Bort

Signore

Tremblay

Martinez Barbera

J. P.

Zaret

K. S.

Hex homeobox gene controls the transition of the endoderm to a pseudostratified, cell emergent epithelium for liver bud development. Dev. Biol. 290(1): 44–56; 2006.

Casteilla

Planat-Benard

Cousin

Silvestre

J. S.

Laharrague

Charriere

Carriere

Penicaud

Plasticity of adipose tissue: A promising therapeutic avenue in the treatment of cardiovascular and blood diseases? Arch. Mal. Coeur. Vaiss. 98(9): 922–926; 2005.

Castell

J. V.

Hernandez

Gomez-Foix

A. M.

Guillen

Donato

Gomez-Lechon

M. J.

Adenovirusmediated gene transfer into human hepatocytes: Analysis of the biochemical functionality of transduced cells. Gene Ther. 4(5): 455–464; 1997.

10.

Chiba

Gotoh

Kojima

Satohisa

Kikuchi

Osanai

Sawada

Hepatocyte nuclear factor (HNF)-4alpha triggers formation of functional tight junctions and establishment of polarized epithelial morphology in F9 embryonal carcinoma cells. Exp. Cell Res. 286(2): 288–297; 2003.

11.

de Juan

Benito

Alvarez

Fabregat

Differential proliferative response of cultured fetal and regenerating hepatocytes to growth factors and hormones. Exp. Cell Res. 202(2): 495–500; 1992.

12.

De Ugarte

D. A.

Morizono

Elbarbary

Alfonso

Zuk

P. A.

Zhu

Dragoo

J. L.

Ashjian

Thomas

Benhaim

Chen

Fraser

Hedrick

M. H.

Comparison of multi-lineage cells from human adipose tissue and bone marrow. Cells Tissues Organs 174(3): 101–109; 2003.

13.

Dudas

Mansuroglu

Batusic

Saile

Ramadori

Thy-1 is an in vivo and in vitro marker of liver myofibroblasts. Cell Tissue Res. 329(3): 503–514; 2007.

14.

Duncan

S. A.

Mechanisms controlling early development of the liver. Mech. Dev. 120(1): 19–33; 2003.

15.

Duncan

S. A.

Navas

M. A.

Dufort

Rossant

Stoffel

Regulation of a transcription factor network required for differentiation and metabolism. Science 281(5377): 692–695; 1998.

16.

Duret

Gerbal-Chaloin

Ramos

Fabre

J. M.

Jacquet

Navarro

Blanc

Sa-Cunha

Maurel

Daujat-Chavanieu

Isolation, characterization, and differentiation to hepatocyte-like cells of non-parenchymal epithelial cells from adult human liver. Stem Cells 25(7): 1779–1790; 2007.

17.

Erices

Conget

Minguell

J. J.

Mesenchymal progenitor cells in human umbilical cord blood. Br. J. Haematol. 109(1): 235–242; 2000.

18.

Farre

Roura

Prat-Vidal

Soler-Botija

Llach

Molina

C. E.

Hove-Madsen

Cairo

J. J.

Godia

Bragos

Cinca

Bayes-Genis

FGF-4 increases in vitro expansion rate of human adult bone marrow-derived mesenchymal stem cells. Growth Factors 25(2): 71–76; 2007.

19.

Fiegel

H. C.

Park

J. J.

Lioznov

M. V.

Martin

Jaeschke-Melli

Kaufmann

P. M.

Fehse

Zander

A. R.

Kluth

Characterization of cell types during rat liver development. Hepatology 37(1): 148–154; 2003.

20.

Filip

English

Mokry

Issues in stem cell plasticity. J. Cell. Mol. Med. 8(4): 572–577; 2004.

21.

Friedman

A. D.

Landschulz

W. H.

McKnight

S. L.

CCAAT/enhancer binding protein activates the promoter of the serum albumin gene in cultured hepatoma cells. Genes Dev. 3(9): 1314–1322; 1989.

22.

Gómez-Lechón

M. J.

Donato

M. T.

Castell

J. V.

Jover

Human hepatocytes in primary culture: The choice to investigate drug metabolism in man. Curr. Drug Metab. 5(5): 443–462; 2004.

23.

Gómez-Lechón

M. J.

Donato

M. T.

Martinez-Romero

Jimenez

Castell

J. V.

O'Connor

J. E.

A human hepatocellular in vitro model to investigate steatosis. Chem. Biol. Interact. 165(2): 106–116; 2007.

24.

Gómez-Lechón

M. J.

Ponsoda

Castell

J. V.

A microassay for measuring glycogen in 96-well-cultured cells. Anal. Biochem. 236(2): 296–301; 1996.

25.

Gouon-Evans

Boussemart

Gadue

Nierhoff

Koehler

C. I.

Kubo

Shafritz

D. A.

Keller

BMP-4 is required for hepatic specification of mouse embryonic stem cell-derived definitive endoderm. Nat. Biotechnol. 24(11): 1402–1411; 2006.

26.

Hahn

J. Y.

Cho

H. J.

Kang

H. J.

Kim

T. S.

Kim

M. H.

Chung

J. H.

Bae

J. W.

B. H.

Park

Y. B.

Kim

H. S.

Pre-treatment of mesenchymal stem cells with a combination of growth factors enhances gap junction formation, cytoprotective effect on cardiomyocytes, and therapeutic efficacy for myocardial infarction. J. Am. Coll. Cardiol. 51(9): 933–943; 2008.

27.

Heino

T. J.

Hentunen

T. A.

Differentiation of osteoblasts and osteocytes from mesenchymal stem cells. Curr. Stem Cell Res. Ther. 3(2): 131–145; 2008.

28.

Igura

Zhang

Takahashi

Mitsuru

Yamaguchi

Takashi

T. A.

Isolation and characterization of mesenchymal progenitor cells from chorionic villi of human placenta. Cytotherapy 6(6): 543–553; 2004.

29.

Isabel

Miri

Einav

Ella

B. L.

Zamir

Ran

Isolation, characterization and culture of Thy1-positive cells from fetal rat livers. World J. Gastroenterol. 12(24): 3841–3847; 2006.

30.

Jen

Manova

Benezra

Expression patterns of Id1, Id2, and Id3 are highly related but distinct from that of Id4 during mouse embryogenesis. Dev. Dyn. 207(3): 235–252; 1996.

31.

Jover

Bort

Gómez-Lechón

M. J.

Castell

J. V.

Cytochrome P450 regulation by hepatocyte nuclear factor 4 in human hepatocytes: A study using adenovirus-mediated antisense targeting. Hepatology 33(3): 668–675; 2001.

32.

Juan

T. S.

Wilson

D. R.

Wilde

M. D.

Darlington

G. J.

Participation of the transcription factor C/EBP delta in the acute-phase regulation of the human gene for complement component C3. Proc. Natl. Acad. Sci. USA 90(7): 2584–2588; 1993.

33.

Jung

Zheng

Goldfarb

Zaret

K. S.

Initiation of mammalian liver development from endoderm by fibroblast growth factors. Science 284(5422): 1998–2003; 1999.

34.

Kamiya

Gonzalez

F. J.

Nakauchi

Identification and differentiation of hepatic stem cells during liver development. Front. Biosci. 11: 1302–1310; 2006.

35.

Kamiya

Kinoshita

Ito

Matsui

Morikawa

Senba

Nakashima

Taga

Yoshida

Kishimoto

Miyajima

Fetal liver development requires a paracrine action of oncostatin M through the gp130 signal transducer. EMBO J. 18(8): 2127–2136; 1999.

36.

Keng

V. W.

Yagi

Ikawa

Nagano

Myint

Yamada

Tanaka

Sato

Muramatsu

Okabe

Sato

Noguchi

Homeobox gene Hex is essential for onset of mouse embryonic liver development and differentiation of the monocyte lineage. Biochem. Biophys. Res. Commun. 276(3): 1155–1161; 2000.

37.

Kern

Eichler

Stoeve

Kluter

Bieback

Comparative analysis of mesenchymal stem cells from bone marrow, umbilical cord blood, or adipose tissue. Stem Cells 24(5): 1294–1301; 2006.

38.

Kinoshita

Miyajima

Cytokine regulation of liver development. Biochim. Biophys. Acta 1592(3): 303–312; 2002.

39.

Koenig

Krause

Drabent

Schaeffner

Christ

Schwartz

Unthan-Fechner

Probst

The expression of mesenchymal, neural and haematopoietic stem cell markers in adult hepatocytes proliferating in vitro. J. Hepatol. 44(6): 1115–1124; 2006.

40.

Kowanetz

Valcourt

Bergstrom

Heldin

C. H.

Moustakas

Id2 and Id3 define the potency of cell proliferation and differentiation responses to transforming growth factor beta and bone morphogenetic protein. Mol. Cell. Biol. 24(10): 4241–4254; 2004.

41.

Kyrmizi

Hatzis

Katrakili

Tronche

Gonzalez

F. J.

Talianidis

Plasticity and expanding complexity of the hepatic transcription factor network during liver development. Genes Dev. 20(16): 2293–2305; 2006.

42.

Lahoz

Donato

M. T.

Castell

J. V.

Gómez-Lechón

M. J.

Strategies to in vitro assessment of major human CYP enzyme activities by using liquid chromatography tandem mass spectrometry. Curr. Drug Metab. 9(1): 12–19; 2008.

43.

Lahoz

Donato

M. T.

Picazo

Gómez-Lechón

M. J.

Castell

J. V.

Determination of major human cytochrome P450s activities in 96-well plates using liquid chromatography tandem mass spectrometry. Toxicol. In Vitro 21(7): 1247–1252; 2007.

44.

Lange

Bassler

Lioznov

M. V.

Bruns

Kluth

Zander

A. R.

Fiegel

H. C.

Liver-specific gene expression in mesenchymal stem cells is induced by liver cells. World J. Gastroenterol. 11(29): 4497–4504; 2005.

45.

Lee

R. H.

Kim

Choi

Kim

Choi

H. S.

Suh

Bae

Y. C.

Jung

J. S.

Characterization and expression analysis of mesenchymal stem cells from human bone marrow and adipose tissue. Cell. Physiol. Biochem. 14(4–6): 311–324; 2004.

46.

Ning

Duncan

S. A.

Mammalian hepatocyte differentiation requires the transcription factor HNF-4alpha. Genes Dev. 14(4): 464–474; 2000.

47.

Lysy

P. A.

Campard

Smets

Najimi

Sokal

E. M.

Stem cells for liver tissue repair: Current knowledge and perspectives. World J. Gastroenterol. 14(6): 864–875; 2008.

48.

Maire

Wuarin

Schibler

The role of cis-acting promoter elements in tissue-specific albumin gene expression. Science 244(4902): 343–346; 1989.

49.

Martinez-Jimenéz

C. P.

Gómez-Lechón

M. J.

Castell

J. V.

Jover

Underexpressed coactivators PGC1alpha and SRC1 impair hepatocyte nuclear factor 4alpha function and promote dedifferentiation in human hepatoma cells. J. Biol. Chem. 281(40): 29840–29849; 2006.

50.

Matsumura

Saito

Takahashi

Ozeki

Kiyotani

Fujieda

Yamazaki

Kunitoh

Kamataki

Identification of a novel polymorphic enhancer of the human CYP3A4 gene. Mol. Pharmacol. 65(2): 326–334; 2004.

51.

Michalopoulos

G. K.

Bowen

W. C.

Mule

Luo

HGF-, EGF-, and dexamethasone-induced gene expression patterns during formation of tissue in hepatic organoid cultures. Gene Expr. 11(2): 55–75; 2003.

52.

Milos

P. M.

Zaret

K. S.

A ubiquitous factor is required for C/EBP-related proteins to form stable transcription complexes on an albumin promoter segment in vitro. Genes Dev. 6(6): 991–1004; 1992.

53.

Milosavljevic

T. S.

Petrovic

M. V.

Cvetkovic

I. D.

Grigorov

I. I.

DNA binding activity of C/EBPbeta and C/EBPdelta for the rat alpha2-macroglobulin gene promoter is regulated in an acute-phase dependent manner. Biochemistry (Mosc) 67(8): 918–926; 2002.

54.

Mirabet

Vaca

Carbonell-Uberos

Montero

J. A.

Dalmau

M. J.

Hornero

Lequerica

J. L.

Piquer

Ischaemic heart disease: Searching for therapeutical solutions. In: Phillips

G. O.

, ed. Advances in tissue banking, vol. 6. London: Scientific Publishing Co. Pte. Ltd.; 2002: 359–374.

55.

Miranville

Heeschen

Sengenes

Curat

C. A.

Busse

Bouloumie

Improvement of postnatal neovascularization by human adipose tissue-derived stem cells. Circulation 110(3): 349–355; 2004.

56.

Miyajima

Kinoshita

Tanaka

Kamiya

Mukouyama

Hara

Role of Oncostatin M in hematopoiesis and liver development. Cytokine Growth Factor Rev. 11(3): 177–183; 2000.

57.

Morrisey

E. E.

Tang

Sigrist

M. M.

Jiang

H. S.

Parmacek

M. S.

GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev. 12(22): 3579–3590; 1998.

58.

Musina

R. A.

Bekchanova

E. S.

Sukhikh

G. T.

Comparison of mesenchymal stem cells obtained from different human tissues. Bull. Exp. Biol. Med. 139(4): 504–509; 2005.

59.

Nakayama

Matsumoto

Tatsumi

Yanai

Yokouchi

Id3 is important for proliferation and differentiation of the hepatoblasts during the chick liver development. Mech. Dev. 123(7): 580–590; 2006.

60.

Nishizaki

Takenaka

Yoshizumi

Yanaga

Soejima

Shirabe

Sugimachi

Alteration in levels of human hepatocyte growth factor following hepatectomy. J. Am. Coll. Surg. 181(1): 6–10; 1995.

61.

Ong

S. Y.

Dai

Leong

K. W.

Inducing hepatic differentiation of human mesenchymal stem cells in pellet culture. Biomaterials 27(22): 4087–4097; 2006.

62.

Parviz

Matullo

Garrison

W. D.

Savatski

Adamson

J. W.

Ning

Kaestner

K. H.

Rossi

J. M.

Zaret

K. S.

Duncan

S. A.

Hepatocyte nuclear factor 4alpha controls the development of a hepatic epithelium and liver morphogenesis. Nat. Genet. 34(3): 292–296; 2003.

63.

Pelttari

Steck

Richter

The use of mesenchymal stem cells for chondrogenesis. Injury 39(Suppl. 1): S58–65; 2008.

64.

Perez

Tabares

Jover

Gómez-Lechón

M. J.

Castell

J. V.

Semi-automatic quantitative RT-PCR to measure CYP induction by drugs in human hepatocytes. Toxicol. In Vitro 17(5–6): 643–649; 2003.

65.

Petersen

B. E.

Goff

J. P.

Greenberger

J. S.

Michalopoulos

G. K.

Hepatic oval cells express the hematopoietic stem cell marker Thy-1 in the rat. Hepatology 27(2): 433–445; 1998.

66.

Phinney

D. G.

Prockop

D. J.

Concise review: mesenchymal stem/multipotent stromal cells: the state of transdifferentiation and modes of tissue repair—current views. Stem Cells 25(11): 2896–2902; 2007.

67.

Pittenger

M. F.

Martin

B. J.

Mesenchymal stem cells and their potential as cardiac therapeutics. Circ. Res. 95(1): 9–20; 2004.

68.

Planat-Benard

Silvestre

J. S.

Cousin

Andre

Nibbelink

Tamarat

Clergue

Manneville

Saillan-Barreau

Duriez

Tedgui

Levy

Penicaud

Casteilla

Plasticity of human adipose lineage cells toward endothelial cells: Physiological and therapeutic perspectives. Circulation 109(5): 656–663; 2004.

69.

Prunet-Marcassus

Cousin

Caton

Andre

Penicaud

Casteilla

From heterogeneity to plasticity in adipose tissues: Site-specific differences. Exp. Cell Res. 312(6): 727–736; 2006.

70.

Rodriguez-Antona

C. J. R.

Gómez-Lechón

M. J.

Castell

J. V.

Quantitative RT-PCR measurement of human cytochrome P-450s: Application to drug induction studies. Arch. Biochem. Biophys. 376: 109–116; 2000.

71.

Romanov

Y. A.

Darevskaya

A. N.

Merzlikina

N. V.

Buravkova

L. B.

Mesenchymal stem cells from human bone marrow and adipose tissue: Isolation, characterization, and differentiation potentialities. Bull. Exp. Biol. Med. 140(1): 138–143; 2005.

72.

Safford

K. M.

Hicok

K. C.

Safford

S. D.

Halvorsen

Y. D.

Wilkison

W. O.

Gimble

J. M.

Rice

H. E.

Neurogenic differentiation of murine and human adipose-derived stromal cells. Biochem. Biophys. Res. Commun. 294(2): 371–379; 2002.

73.

Sakaguchi

Sekiya

Yagishita

Muneta

Comparison of human stem cells derived from various mesenchymal tissues: Superiority of synovium as a cell source. Arthritis Rheum. 52(8): 2521–2529; 2005.

74.

Sakai

Jiang

Kojima

Kinoshita

Miyajima

Enhanced in vitro maturation of fetal mouse liver cells with oncostatin M, nicotinamide, and dimethyl sulfoxide. Cell Transplant. 11(5): 435–441; 2002.

75.

Schmidt

Bladt

Goedecke

Brinkmann

Zschiesche

Sharpe

Gherardi

Birchmeier

Scatter factor/hepatocyte growth factor is essential for liver development. Nature 373(6516): 699–702; 1995.

76.

Seo

M. J.

Suh

S. Y.

Bae

Y. C.

Jung

J. S.

Differentiation of human adipose stromal cells into hepatic lineage in vitro and in vivo. Biochem. Biophys. Res. Commun. 328(1): 258–264; 2005.

77.

Serls

A. E.

Doherty

Parvatiyar

Wells

J. M.

Deutsch

G. H.

Different thresholds of fibroblast growth factors pattern the ventral foregut into liver and lung. Development 132(1): 35–47; 2005.

78.

Sgodda

Aurich

Kleist

Aurich

Konig

Dollinger

M. M.

Fleig

W. E.

Christ

Hepatocyte differentiation of mesenchymal stem cells from rat peritoneal adipose tissue in vitro and in vivo. Exp. Cell Res. 313(13): 2875–2886; 2007.

79.

Shiraki

Umeda

Sakashita

Takeya

Kume

Differentiation of mouse and human embryonic stem cells into hepatic lineages. Genes Cells 13(7): 731–746; 2008.

80.

Snykers

Vanhaecke

Papeleu

Luttun

Jiang

Vander Heyden

Verfaillie

Rogiers

Sequential exposure to cytokines reflecting embryogenesis: the key for in vitro differentiation of adult bone marrow stem cells into functional hepatocyte-like cells. Toxicol. Sci. 94(2): 330–341; 2006.

81.

Song

Tuan

R. S.

Transdifferentiation potential of human mesenchymal stem cells derived from bone marrow. FASEB J. 18(9): 980–982; 2004.

82.

Spath

G. F.

Weiss

M. C.

Hepatocyte nuclear factor 4 provokes expression of epithelial marker genes, acting as a morphogen in dedifferentiated hepatoma cells. J. Cell. Biol. 140(4): 935–946; 1998.

83.

Spear

B. T.

Jin

Ramasamy

Dobierzewska

Transcriptional control in the mammalian liver: liver development, perinatal repression, and zonal gene regulation. Cell. Mol. Life Sci. 63(24): 2922–2938; 2006.

84.

Steinert

Bachner

Flohe

Analysis of the mouse selenoprotein P gene. Biol. Chem. 379(6): 683–691; 1998.

85.

Stock

Staege

M. S.

Muller

L. P.

Sgodda

Volker

Volkmer

Lutzkendorf

Christ

Hepatocytes derived from adult stem cells. Transplant. Proc. 40(2): 620–623; 2008.

86.

Suzuki

Iwama

Miyashita

Nakauchi

Taniguchi

Role for growth factors and extracellular matrix in controlling differentiation of prospectively isolated hepatic stem cells. Development 130(11): 2513–2524; 2003.

87.

Talens-Visconti

Bonora

Jover

Mirabet

Carbonell

Castell

J. V.

Gómez-Lechón

M. J.

Hepatogenic differentiation of human mesenchymal stem cells from adipose tissue in comparison with bone marrow mesenchymal stem cells. World J. Gastroenterol. 12(36): 5834–5845; 2006.

88.

Tirona

R. G.

Lee

Leake

B. F.

Lan

L. B.

Cline

C. B.

Lamba

Parviz

Duncan

S. A.

Inoue

Gonzalez

F. J.

Schuetz

E. G.

Kim

R. B.

The orphan nuclear receptor HNF4alpha determines PXR- and CAR-mediated xenobiotic induction of CYP3A4. Nat. Med. 9(2): 220–224; 2003.

89.

Umek

R. M.

Friedman

A. D.

McKnight

S. L.

CCAAT-enhancer binding protein: A component of a differentiation switch. Science 251(4991): 288–292; 1991.

90.

Vorgia

Zannis

V. I.

Kardassis

A short proximal promoter and the distal hepatic control region-1 (HCR-1) contribute to the liver specificity of the human apolipoprotein C-II gene. Hepatic enhancement by HCR-1 requires two proximal hormone response elements which have different binding specificities for orphan receptors HNF-4, ARP-1, and EAR-2. J. Biol. Chem. 273(7): 4188–4196; 1998.

91.

Wang

P. P.

Wang

J. H.

Yan

Z. P.

M. Y.

Lau

G. K.

Fan

S. T.

Luk

J. M.

Expression of hepatocyte-like phenotypes in bone marrow stromal cells after HGF induction. Biochem. Biophys. Res. Commun. 320(3): 712–716; 2004.

92.

Weinstein

Monga

S. P.

Liu

Brodie

S. G.

Tang

Mishra

Deng

C. X.

Smad proteins and hepatocyte growth factor control parallel regulatory pathways that converge on beta1-integrin to promote normal liver development. Mol. Cell. Biol. 21(15): 5122–5131; 2001.

93.

Westmacott

Burke

Z. D.

Oliver

Slack

J. M.

Tosh

C/EBPalpha and C/EBPbeta are markers of early liver development. Int. J. Dev. Biol. 50(7): 653–657; 2006.

94.

Yamamoto

Banas

Murata

Ishikawa

Lim

C. R.

Teratani

Hatada

Matsubara

Kato

Ochiya

A comparative analysis of the transcriptome and signal pathways in hepatic differentiation of human adipose mesenchymal stem cells. FEBS J. 275(6): 1260–1273; 2008.

95.

Yasui

Hasada

Yang

J. G.

Koiwai

Structure and organization of the gene encoding human selenoprotein. Gene 175(1–2): 269–270; 1996.

96.

Young

H. E.

Steele

T. A.

Bray

R. A.

Hudson

Floyd

J. A.

Hawkins

Thomas

Austin

Edwards

Cuzzourt

Duenzl

Lucas

P. A.

Black

A.C.

Jr. Human reserve pluripotent mesenchymal stem cells are present in the connective tissues of skeletal muscle and dermis derived from fetal, adult, and geriatric donors. Anat. Rec. 264(1): 51–62; 2001.

97.

Yubero

Hondares

Carmona

M. C.

Rossell

Gonzalez

F. J.

Iglesias

Giralt

Villarroya

The developmental regulation of peroxisome proliferator-activated receptor-gamma coactivator-1alpha expression in the liver is partially dissociated from the control of gluconeogenesis and lipid catabolism. Endocrinology 145(9): 4268–4277; 2004.

98.

Zhang

Yatskievych

T. A.

Baker

R. K.

Antin

P. B.

Regulation of Hex gene expression and initial stages of avian hepatogenesis by Bmp and Fgf signaling. Dev. Biol. 268(2): 312–326; 2004.

99.

Zhao

Watt

J. A.

Luebke-Wheeler

Morrisey

E. E.

Duncan

S. A.

GATA6 is essential for embryonic development of the liver but dispensable for early heart formation. Mol. Cell. Biol. 25(7): 2622–2631; 2005.

100.

Zhou

Abidi

Kim

Chen

Huang

T. T.

Kraemer

F. B.

Liu

Transcriptional activation of hepatic ACSL3 and ACSL5 by oncostatin m reduces hypertriglyceridemia through enhanced beta-oxidation. Arterioscler. Thromb. Vasc. Biol. 27(10): 2198–2205; 2007.

101.

Zuk

P. A.

Zhu

Ashjian

De Ugarte

D. A.

Huang

J. I.

Mizuno

Alfonso

Z. C.

Fraser

J. K.

Benhaim

Hedrick

M. H.

Human adipose tissue is a source of multipotent stem cells. Mol. Biol. Cell 13(12): 4279–4295; 2002.

102.

Zuk

P. A.

Zhu

Mizuno

Huang

Futrell

J. W.

Katz

A. J.

Benhaim

Lorenz

H. P.

Hedrick

M. H.

Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng. 7(2): 211–228; 2001.