Human decidua-derived mesenchymal stromal cells differentiate into hepatic-like cells and form functional three-dimensional structures

Cytotherapy, 2012; 14: 1182–1192

Human decidua-derived mesenchymal stromal cells differentiate into hepatic-like cells and form functional three-dimensional structures RAFAEL BORNSTEIN1,2∗, MARIA I. MACIAS2†, PAZ DE LA TORRE2†, JESUS GRANDE3,2 & ANA I. FLORES2 1Madrid

Cord Blood Bank, Hospital Universitario 12 de Octubre, Madrid, Spain, 2Regenerative Medicine Group, Research Center, Instituto de Investigacion Hospital 12 de Octubre, Madrid, Spain, and 3Department of Obstetrics and Gynaecology, Hospital Universitario 12 de Octubre, Madrid, Spain

Abstract Background aims. Previously, we have shown that human decidua-derived mesenchymal stromal cells (DMSC) are mesenchymal stromal cells (MSC) with a clonal differentiation capacity for the three embryonic layers. The endodermal capacity of DMSC was revealed by differentiation into pulmonary cells. In this study, we examined the hepatic differentiation of DMSC. Methods. DMSC were cultured in hepatic differentiation media or co-cultured with murine liver homogenate and analyzed with phenotypic, molecular and functional tests. Results and Conclusions. DMSC in hepatic differentiation media changed their fibroblast morphology to a hepatocyte-like morphology and later formed a 3-dimensional (3-D) structure or hepatosphere. Moreover, the hepatocyte-like cells and the hepatospheres expressed liver-specific markers such as synthesis of albumin (ALB), hepatocyte growth factor receptor (HGFR), α-fetoprotein (AFP) and cytokeratin-18 (CK-18), and exhibited hepatic functions including glycogen storage capacity and indocyanine green (ICG) uptake/secretion. Human DMSC co-cultured with murine liver tissue homogenate in a non-contact in vitro system showed hepatic differentiation, as evidenced by expression of AFP and ALB genes. The switch in the expression of these two genes resembled liver development. Indeed, the decrease in AFP and increase in ALB expression throughout the co-culture were consistent with the expression pattern observed during normal liver organogenesis in the embryo. Interestingly, AFP and ALB expression was significantly higher when DMSC were co-cultured with injured liver tissue, indicating that DMSC respond differently under normal and pathologic micro-environmental conditions. In conclusion, DMSC-derived hepatospheres and DMSC co-cultured with liver homogenate could be suitable in vitro models for toxicologic, developmental and pre-clinical hepatic regeneration studies. Key Words: decidua-derived mesenchymal stromal cells, hepatic differentiation, hepatic-specific functions, hepatospheres, placenta, pre-clinical hepatic regeneration studies

Introduction An alternative to orthotopic liver transplantation is hepatocyte cell transplantation. However, transplantation of isolated hepatocytes is not feasible for human clinical applications because of the large number of cells required and the dedifferentiation of these cells in the in vitro expansion systems (1). Therefore, an important goal in human liver therapy is to obtain unlimited numbers of clinically useful hepatocytes. Embryonic and adult stem cells differentiate into hepatic-like cells representing valuable sources for cell liver transplantation (2,3). Mesenchymal stromal cells (MSC) are adult stem cells that have emerged recently as attractive ∗Current

candidates for tissue regeneration because they are easily isolated and have a high proliferation and differentiation potential, as well as immunosuppressive and immunomodulatory properties (4,5). Human MSC from different sources, such as bone marrow (BM) (6), adipose tissue (7), umbilical cord blood (8), umbilical cord tissue (9), menstrual blood (10) and placenta (11–14), have been differentiated into hepatocyte-like cells. Cell-to-cell contact is important for hepatocyte proliferation, differentiation and function, suggesting that three-dimensional (3-D) structures resembling the normal liver tissue are the best microenvironment for growing and maintaining these cells

address: Department of Hematology, Hospital Central de la Cruz Roja, Avda. Reina Victoria, 26, 28003 Madrid, Spain. authors contributed equally to this work. Correspondence: Ana I. Flores, PhD Centro de Investigacion, Grupo de Medicina Regenerativa, Instituto de Investigacion Hospital 12 Octubre, Avda. Cordoba s/n, 28041 Madrid, Spain. E-mail: afl[email protected]

†These

(Received 29 November 2011; accepted 19 June 2012) ISSN 1465-3249 print/ISSN 1477-2566 online © 2012 Informa Healthcare DOI: 10.3109/14653249.2012.706706

Hepatic differentiation of decidua-derived MSCs 1183 in vitro (1,15–17). There is a growing interest in 3-D liver devices as transient supports for patients waiting for a suitable donor or their own liver regeneration (2,18). Nowadays, BM is the principal source of MSC for experimental and clinical settings (19). However, placenta represents a source of MSC with several advantages. Placenta-derived MSC are isolated by non-invasive procedures, are adult stem cells without ethical concerns, can be cryopreserved, have a low risk of viral infection, a high expansion and differentiation capacity and a low or no immune response (20,21). Recently, we isolated and characterized a population of MSC from the maternal part of human placenta (22). These cells, called decidua-derived MSC (DMSC), showed a high proliferation and differentiation capacity into cells from the three embryonic layers. Undifferentiated DMSC expressed GATA-4, a zing finger transcription factor required for endoderm specification and involved in hepatocyte differentiation and in the early stages of liver development (23,24). GATA-4 expression in DMSC could be an indicator of hepatic differentiation capacity for DMSC. DMSC were differentiated efficiently into endodermal type II alveolar cells, although their differentiation into liver tissue was not evaluated (22). In this study, we assayed the in vitro hepatic differentiation capacity of DMSC in a highly defined and serumfree hepatocyte medium, as well as their response to a liver microenvironment. Hepatic in vitro culture medium induced the differentiation of DMSC into hepatocyte-like cells that finally formed a 3-D spheroid structure or hepatosphere. Both the hepatocytelike cells and the hepathospheres exhibited hepatic functions, including glycogen storage capacity and indocyanine green (ICG) uptake/secretion. DMSC co-cultured with murine liver tissue homogenate differentiated into hepatocyte-like cells and expressed markers of early and mature hepatic cells, such as α-fetoprotein (AFP) and albumin (ALB), through a process resembling the embryonic hepatic development. The DMSC response level to healthy and damaged liver was different in potency, suggesting a higher potential of DMSC to engraft and differentiate into hepatocytes after transplantation into injured livers. Additionally, hepatospheres could be an important in vitro model for toxicologic, developmental and regeneration studies.

Methods Isolation and culture of placental cells Human placentas from healthy mothers were obtained from the Department of Obstetrics and Gynecology with written informed consent

approved by the ethics committee from Hospital 12 Octubre (Madrid, Spain). DMSC were isolated and cultured from fetal membranes as described previously (22). Briefly, tissue was digested with trypsin-versene (Lonza, Barcelona, Spain) and cells were seeded at 1.16 ⫻ 105 cells/cm2 and cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 2 mM glutamine, 0.1 mM sodium pyruvate, 55 μM β-mercaptoethanol, 1% non-essential amino acids, 1% penicillin/streptomycin, 10% fetal bovine serum and 10 ng/mL epidermal growth factor (EGF) (Sigma-Aldrich Quimica SA, Madrid, Spain). Non-adherent cells were removed and adherent cells were grown to confluence and passaged at a density of 4–5 ⫻ 104 cells/cm2. Hepatic differentiation mediated by growth factors DMSC at confluence were cultured in hepatocyte complete medium (HCM medium bullet kit; Lonza Iberica SA, Barcelona, Spain) supplemented with hepatocyte growth factor (HGF; 40 ng/mL), basic fibroblast growth factor (b-FGF; 20 ng/mL), fibroblast growth factor-4 (FGF-4; 20 ng/mL) and stem cell factor (SCF; 40 ng/mL) (all from SigmaAldrich). Cultures were fed every 3–4 days. Morphologic changes were followed up with phase-contrast microscopy using a Leica DMIL microscope. The expression of liver-specific markers was assessed by immunofluorescence staining and reverse transcription (RT)–polymerase chain reaction (PCR) analysis. Adult liver RNA was used as the RT-PCR-positive control. Normal human dermal fibroblasts (Lonza) were used as a negative control for the differentiation studies. Inmunostaining Cells were fixed with 10% formalin for 20 min at room temperature and permeabilized with 0.3% Triton-X 100 for 10 min. Non-specific binding was blocked with 5% horse serum. Primary antibodies [ALB, hepatocyte growth factor receptor (HGFR) and AFP, purchased from Sigma-Aldrich, and cytokeratin-18 (CK-18), from Millipore Iberica SAU, Madrid, Spain] were incubated overnight at 4°C, rinsed with phosphate-buffered saline (PBS) twice and incubated with fluorescein isothiocyanate (FITC)- or tetramethyl rhodamine isothiocyanate (TRITC)-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories, Vitro SA, Madrid, Spain). Nuclei were counterstained with 0.2 μg/mL 4′,6′ diamidino-2-phenylindole (DAPI; Sigma-Aldrich) for 1 min. Images were visualized using a Leica DMIL fluorescence microscope (Leica Microsistemas SLU, Barcelona, Spain) or a Spectral

1184 R. Bornstein et al. Confocal Microscopy LSM510 META ConfoCor 3 (Carl Zeiss AG, Jena, Germany). Periodic acid–Schiff staining for glycogen Intracellular glycogen was evaluated using the periodic acid–Schiff (PAS) staining system (SigmaAldrich) according to manufacturer’s instructions, but without the final step of counterstaining in hematoxylin solution. Briefly, cells were fixed with formalin–ethanol fixative solution, incubated in periodic acid solution, rinsed and stained with Schiff ’s reagent. Semi-quantitative analysis of PAS immunoreactivity was performed at day 10 in differentiation media. PAS-stained cells were examined in bright-field using a Leica DMIL fluorescence microscope (Leica Microsistemas SLU, Barcelona, Spain) coupled to a Nikon COOLPIX 995 camera. To quantify PAS-positive cells, we counted the number of red Schiff cells on three 5 ⫻ 5 cm2 randomly selected areas of three digitally captured images. Positive cells were normalized with the total number of cells. ICG uptake and excretion assay At different time-points during the hepatogenic differentiation, the cells were stained with 1 mg/mL ICG (ICG-PULSION; PULSION Medical Systems, Feldkirchen, Germany) for 15 min at 37°C. The cells were rinsed three times with PBS to remove the organic anion and dishes were refilled with hepatic differentiation media. At this point, ICG uptake by undifferentiated DMSC, hepatocytelike cells and hepatospheres was observed using an inverted Leica DMIL microscope. ICG elimination was confirmed 8 h later. ICG uptake was quantified as described previously with some modifications (25). Briefly, after ICG staining for 15 min at 37°C, cells were washed three times with PBS and subsequently lysed with 250 μL 1% Triton-X 100 in PBS. ICG concentrations were determined spectrophotometrically (Enspire 2300; PerkinElmer, Tres Cantos, Madrid, Spain) at 805 nm with the use of a standard curve. Values were normalized with a protein concentration. Protein concentrations were determined with the Lowry protein assay (Bio-Rad Laboratories SA, Alcobendas, Madrid, Spain) using bovine serum albumin (BSA) as the standard. ALB secretion The content of human ALB secreted to the differentiation media was determined with a AssayMax human ALB enzyme-linked immunosorbent assay (ELISA) kit (Assaypro, Deltaclon SL, Madrid, Spain)

according the manufacturer’s instructions for cell culture supernatants. The ALB content in each supernatant media was calculated from the standard curve generated each time the assay was performed. Hepatic differentiation induced by co-culture with murine liver tissue in a transwell system Healthy and injured liver tissues were obtained from Balb/c mice. Animal experiments were approved by the ethical and animal care committee of the Hospital 12 de Octubre. Liver damage was induced with a single intraperitoneal injection of 10 μL/g body weight 10% carbon tetrachloride (CCl4) in mineral oil, and mice were killed 48 h later. After that, healthy and injured livers were surgically removed, weighed, finely minced and then mechanically homogenized with a glass Dounce homogenizer in 4 mL coculture medium (DMEM supplemented with 2 mM glutamine, 0.1 mM sodium pyruvate, 55 μM βmercaptoethanol, 1% non-essential amino acids and 1% penicillin/streptomycin; all from Lonza). This process mechanically disrupts all the cells present in the liver tissue, releasing the organelles and the cytoplasm. With the homogenate technique, we obtained surviving groups of proteins, enzymes and molecules without the loss of their in vivo properties. Finally, we added 1 mL of the same medium and stored it at –80°C until use. Co-culture experiments were carried out according to the protocol described by Jang et al. (26) for hematopoietic stem cells, except for minor modifications. The co-culture model included a Boyden chamber and a transwell insert (pore size 0.4 μm; Corning Costar, Amsterdam, The Netherlands). Human DMSC were seeded at 4–5 ⫻ 104 cells/cm2 into the transwell insert. At confluence, the co-culture medium with or without 50 mg minced healthy or damaged murine liver was added to the lower chamber. Control co-cultures were set up with medium without tissue. The viability of the cells in the transwell membrane after 48 and 72 h of incubation was analyzed by DAPI staining. Then the transwell inserts were washed and frozen at –80°C until used for RNA isolation. Expression of liver-specific genes by RT-PCR and quantitative real-time PCR Total RNA extraction was performed using a RNeasy mini kit (Qiagen, IZASA Distribuciones Tecnicas SA, Madrid, Spain). To reverse transcribe the RNA, we used the Transcriptor High Fidelity cDNA Synthesis kit (Roche, Roche España, Madrid, Spain). Human ALB and glyceraldehyde 3-phosphate dehydrogenase (GAPDH)-specific primers and PCR conditions are shown in Table I. PCR products were

Hepatic differentiation of decidua-derived MSCs 1185 Table I. Primers and conditions for the PCR. Human ALB-specific primers do not amplify the mouse cDNA sequences. Marker ALB GAPDH

Temperature (°C)

Size (bp)

Forward primer (5′→3′)

Reverse primer (5′→3′)

58,3 57,3

272 512

ctcaagtgtgccagtctcca gagtcaacggatttggtcgt

tgggatttttccaacagagg tgtggtcatgagtccttcca

separated on a 2% agarose gel and visualized with SYBR® gold staining (Invitrogen SA, Barcelona, Spain). Real-time PCR was performed using an ABI 7,500 sequence detection system (Applied Biosystems, Life Technologies SA, Madrid, Spain) with inventoried Taqman-specific probes (ALB, Hs00910225_m1; AFP, Hs00173490_m1; TATA box binding protein (TBP), Hs00427620_m1). Data were analyzed using the 2⫺ΔΔCt method of relative gene expression, with the target genes values normalized to the housekeeping gene TBP (TATA box binding protein). cDNA from adult and fetal human livers was used as positive controls for ALB and AFP expression, respectively.

Statistical analysis Data represent means ⫾ SEM. Statistical differences between groups were analyzed using the Student’s t-test. Statistical significance was established at P ⬍ 0.05.

Results DMSC cultured in expansion media retained their fibroblast-like morphology (Figure 1A). After 8 days in hepatic differentiation medium, the cells adopted a polygonal shape characteristic of hepatocyte-like cells (Figure 1B). Undifferentiated DMSC expressed very low levels of ALB and HGFR (Supplementary Figure 2 to be found online at http://informahealthcare.com/doi/abs/10.3109/14653249.2012.706706; and Figure 3). Around days 8–10, human ALB and HGFR protein levels increased (data not shown). The monolayer of polygonal cells underwent a dramatic architectural transformation at around day 14 (Figure 1C). This morphologic change was observed in all cells present in the tissue culture plate and was reproduced with cells obtained from multiple placentas. From that moment on, cells started to retract and form clusters (Figure 1C–E). At around day 21–30, most of the cells became compactly encapsulated into a fibrous layer smoothly attached

Figure 1. DMSC formed 3-D structures during hepatic differentiation. (A) Undifferentiated DMSC. (B) Differentiated cells at day 8 showing hepatocyte-like morphology. (C) The cells started to aggregate at around day 14. (D, E) Formation of 3-D structures. (F) Hepatosphere smoothly attached by a fibrous layer (arrows) at around days 21–30. Scale bar: 200 μm.

1186 R. Bornstein et al. to the culture plate (Figure 1F, arrows). Several 3-D dimensional spheroid structures or ‘hepatospheres’ were obtained in each culture plate from all isolated DMSC lines tested. The hepatospheres contained positive cells for several human hepatocyte-associated markers, such as AFP (see Supplementary Figure 1 to be found online at http://informahealthcare.com/doi/abs/ 10.3109/14653249.2012.706706), ALB, HGFR and CK-18 (Figure 2B,E,H). Hepathospere sections revealed the expression of these hepaticspecific markers inside the 3-D structure. This analysis confirmed hepatic differentiation at the cell level inside the hepatosphere and ruled out the possibility of artifacts (Figure 2C,F,I). In addition, positive cells for each staining inside the hepatosphere showed hepatic epithelial morphology (Figure 2C,F,I, inserts). Expression of the ALB gene was also confirmed by RT-PCR analysis in hepatospheres obtained from three different placentas (Figure 3A, lanes 2–4) and quantitative real-time PCR (Figure 3B). Undifferentiated DMSC expressed this marker at a low level, as assessed by inmunocytochemistry (see Supplementary Figure 2B to be found online

at http://informahealthcare.com/doi/abs/10.3109/ 14653249.2012.706706), RT-PCR (Figure 3, lane 1) and real-time PCR (Figure 3B). Interestingly, the hepatosphere showed an expression eight times higher, which confirmed the hepatic differentiation. As a negative control, dermal human fibroblasts were cultured in the hepatic differentiation medium but did not show either morphologic changes or immunostaining for ALB (data not shown). In addition, secondary antibody staining of the hepatospheres was negative, pinpointing the specificity of the staining procedure (see Supplementary Figure 2F to be found online at http://informahealthcare.com/doi/ abs/10.3109/14653249.2012.706706). Hepatocyte-like cells and hepatospheres have liver biologic functions because they have the capacity to synthesize and store glycogen. The storage of glycogen was detected clearly in hepatocyte-like cells after 10 days in differentiation media (Figure 4C), and these levels increased as the cells aggregated and formed the hepatosphere (Figure 4D). At day 10 in differentiation media, 80% of the total cells were already PAS-positive (P ⬍ 0.0001). Interestingly, 68% of the PAS-positive cells presented an

Figure 2. Hepatospheres expressed liver-specific proteins. (A, D, G) Phase–contrast of three different hepatospheres. (B, C) ALB protein expression. (E, F) Hepatocyte growth factor receptor protein expression (c-Met). (H, I) CK-18 protein expression. (B, E, H) Immunofluorescent staining of 3-D dimensional hepatospheres visualized using a Leica DMIL fluorescence microscope. (C, F, I) Hepatosphere cross-sectional areas visualized using a Spectral Confocal Microscopy LSM510 META ConfoCor 3 (Carl Zeiss AG, Jena, Germany). Inserts show one positive cell inside the hepatosphere with hepatic epithelial morphology. Scale bar: 200 μm.

Hepatic differentiation of decidua-derived MSCs 1187

Figure 3. ALB gene expression on hepatospheres differentiated from three different placentas. (A) ALB gene expression was analyzed on DMSC-derived hepatospheres (lanes 2–4). Undifferentiated cells did not express this marker (lane 1). Reverse Transcription (RT) control (lane 5). Adult liver was used as a positive control (lane 6). GAPDH was used as a housekeeping gene. (B) ALB gene expression was analyzed by quantitative realtime PCR on a DMSC-derived hepatosphere and expressed relative to undifferentiated cells. The results were normalized to the TBP values used as a housekeeping gene.

intensive red Schiff color, indicative of high levels of glycogen. Glycogen staining intensity was very evident in a cross-sectional area of a hepatosphere (Figure 4D, insert). Undifferentiated DMSC were negative for PAS staining (Figure 4B). Furthermore, ICG uptake by undifferentiated DMSC, hepatocytelike cells and hepatospheres was evaluated, confirming that the differentiated cells functioned like

hepatocytes. Hepatocyte-like cells and hepatospheres were positive for ICG (Figure 5C,E), while undifferentiated DMSC were negative (Figure 5A), indicating that the differentiated cells took up ICG from the medium. Quantification of ICG uptake demonstrated that the cells differentiated for 15 days had a significantly higher ICG uptake rate compared with DMSC (198.54 ⫾ 18.10 versus. 64.03 ⫾ 1.27 nmol/mg protein, respectively; P ⬍ 0.0001), which correlated with the differentiation stage. In addition, ICG was eliminated from the hepatocyte-like cells and hepatospheres 8 h later (Figure 5D,F), suggesting that the differentiated cells were metabolically active because they secreted completely the absorbed ICG. We attempted to study an additional hepatic biologic function of the hepatospheres, such as ALB secretion into the culture supernatant. Unfortunately, the high levels of this bovine factor in the differentiation medium prevented its reliable quantification by ELISA assay. We evaluated the role of liver tissue microenvironment in the fate of DMSC using a non-contact co-culture assay, and examined the expression of hepatic-specific genes in DMSC by quantitative realtime PCR in response to normal and damaged tissue. Human DMSC were co-cultured in the presence of either healthy or injured mouse liver homogenate for

Figure 4. DMSC-derived hepatospheres produced and stored glycogen. (A) Phase–contrast of undifferentiated DMSC. (B) Negative PAS staining in undifferentiated DMSC. (C, D) DMSC after 10 days in differentiation media, and hepatospheres, were positive for PAS staining. Glycogen storage was visualized by the pink to red color indicative of the glycol-containing cellular elements. Insert show positive PAS staining inside the hepatosphere. Scale bar: 200 μm.

1188 R. Bornstein et al.

Figure 5. Cellular uptake and secretion of ICG by DMSC-derived hepatospheres. (A, C, E) ICG uptake by undifferentiated DMSC, hepatocyte-like cells at day 15 and a hepatosphere. (B, D, F) ICG was eliminated 8 h later. Scale bar: 200 μm.

48 and 72 h, and AFP and ALB gene expression was analyzed. The viability of the cells in the transwell membrane was higher than 85% at both time-points, as demonstrated by phase–contrast and DAPI nuclear staining (data not shown). AFP is a hepatoblast and liver regeneration-associated marker, while ALB is a marker of early hepatic cells and its expression is restricted to the liver. Expression of both markers was very low in DMSC in both proliferation and co-culture media (Figure 3 and data not shown). In DMSC co-cultured for 48 h in the presence of healthy or injured liver homogenate, AFP gene was induced 1.6-fold and 2.2-fold, respectively,

calculated as fold-induction with respect to the gene level of DMSC co-cultured without tissue homogenate (Figure 6A). AFP expression showed a marked decrease from 48 h to 72 h in cells cocultured with healthy (0.6-fold) and injured liver (1.6-fold). The difference between AFP levels at 72 h in DMSC co-cultured with healthy and injured liver homogenate was statistically significant (P ⬍ 0.025). ALB gene expression showed an increase from 48 h to 72 h in the presence of healthy and injured liver during co-culture (Figure 6B). The differences between ALB levels at 48 h and 72 h in DMSC

Hepatic differentiation of decidua-derived MSCs 1189

Figure 6. DMSC co-cultured with murine liver tissue homogenate induced hepatic differentiation. AFP and ALB gene expression in DMSC co-cultured with healthy and injured murine liver was determined by quantitative real-time PCR. Values are expressed as fold induction with respect to gene levels in DMSC co-cultured without liver homogenate. DMSC showed a decrease from 48 h to 72 h on AFP gene expression (A) and an increase on ALB gene expression (B). AFP and ALB expression was significantly higher in the presence of injured tissue.

co-cultured with healthy and injured liver homogenate were statistically significant (P ⬍ 0.0001 and P ⬍ 0.0025, respectively). These data suggested that liver homogenate induced DMSC hepatic differentiation. A significantly stronger response in AFP and ALB expression was obtained in the presence of injured liver with respect to healthy tissue. Discussion DMSC have several advantages over other adult stem cells, including the high numbers obtained without any risk to the donor, prolonged stemness in vitro, a high differentiation capacity, and hypoimmunogenicity because they do not express either the major histocompatibility complex (MHC) class II or T-cell co-stimulatory molecules (22). Here, we show the differentiation of DMSC into hepatocyte-like cells and the formation of 3-D spheroid structures known as hepatospheres (27,28). Cell confluence favors the change from fibroblast to hepatocyte-like morphology, suggesting that cell–cell contact is an important issue for DMSC hepatic differentiation (29). Adult stem cells, including DMSC, do not express AFP (8,29). DMSC-derived hepatospheres expressed hepatic markers, such as HGFR, ALB, AFP and CK-18. Moreover, differentiated hepatocyte-like cells and hepatospheres have functional properties of hepatocytes, such as glycogen storage (30) and ICG uptake and secretion. ICG is an organic anion clinically used to evaluate liver function because it is specifically taken up and eliminated by the hepatocytes (31). During embryonic liver development and in adult liver, hepatocytes need a sophisticated 3-D microenvironment to carry out their physiologic functions (15,32). The hepatosphere is a structure with a

similar architecture to native tissue (28). Hepathosphere formation allows cell–cell and cell–matrix interactions, preserving the hepatocytes’ function in vitro and after transplantation in vivo (1,2,27,28,32,33). Although the differentiation of adult MSC into hepatocyte-like cells has been described already (6,7,9–14,29,34–36), a few studies have reported 3-D structures of adult stem cells cultured under hepatic differentiation media (10,37). The formation of hepatospheres could rely on the intrinsic characteristics of MSC from different sources, the different culture conditions used and/or the diverse end-points of the experiments. Several methods have been used to obtain 3-D cell cultures, including different combinations of growth factors, rotary or rocking systems and/or artificial matrices (1,2,16,28,33,38–40). Different in vitro culture systems and scaffolds trying to mimic the in vivo liver physiology have been developed in order to obtain large numbers of functional hepatocytes (1,27,30,36,40–42). For human liver therapies, it would be necessary to obtain approximately 1013 DMSC-derived hepatocyte-like cells. In previously reported results, we observed an exponential growth for DMSC between the second and 22nd passages (22). From an initial cell number of 6.88 ⫾ 5.23 ⫻ 106, we obtained a cumulative cell number of 2.54 ⫾ 4.97 ⫻ 1014 at a median of 20 total passages. Because we observed that all the cells were able to differentiate into hepatocyte-like cells and express hepatocyte functional markers, we suggest that with a single placenta there will be enough cells for transplantation for 25–50 patients. However, in order to use DMSC in regenerative medicine in humans, it may be worth improving the initial yield and optimizing their in vitro growth to obtain enough cells for transplantation in a low number of passages (43). Some authors have reported a significant correlation between yield and

1190 R. Bornstein et al. low plating density for BM MSC (44,45) and firsttrimester chorionic villi MSC (46). Whether DMSC would give the same or a larger output under a low plating density needs to be ascertained. The use of bioreactors is a promising alternative for increasing the number of undifferentiated DMSC in few passages and additionally to induce and maintain the differentiation longer in order to improve the maturation without losing viability (47,48). Bioreactors increase the cell viability and proliferation of cells because the cells have not limited access to nutrients at any time because of the perfusion flow component of the bioreactors. This constant flow also prevents de-differentiation and an increase in the number of differentiated cells with a continuous supplementation of growth factors to expand the cells. In pre-clinical studies for liver tissue engineering, the bioreactors most frequently used are the hollow-fiber 3-D perfusion bioreactors. This type of bioreactor allows the supply of oxygen to the cells, increasing the cell proliferation and differentiation into mature hepatocytes, which even acquire a structural liver organization (16,48,49). Undifferentiated MSC possess a low integration capacity into damaged liver. However, preconditioned adipose and BM-derived MSC to hepatocyte-like cells increases their engraftment significantly (7,50). We hypothesize that hepatospheres could further increase this engraftment capacity and could provide an alternative approach to cell devices in pre-clinical trials for liver transplantation. We are now testing this hypothesis in an animal model of partial hepatectomy. As in vitro toxicology assays using primary hepatocytes or immortalized cell lines do not reproduce the response occurring in animal models, hepatospheres could be an alternative screening system for studying the effects of hormones, drugs, toxins and carcinogens, reducing the number of animals needed for those experiments. In addition, DMSC-derived hepatospheres is a simple in vitro system based on differentiated human cells that would be useful for studying liver physiology and diseases difficult to address using animal models (41). In vitro DMSC liver differentiation is induced under closely monitored conditions. However, in vivo transplanted cells will be exposed to local and systemic factors that will modulate their response and will be decisive for the transplant outcome. In vitro models mimicking the local in vivo microenvironment allow the observation of stem cell behavior and/or differentiation capacity to anticipate their response in vivo (3,26,51). Here, we have used an in vitro co-culture system to study the response of DMSC in the presence of a liver microenvironment. AFP and ALB expression was up-regulated in

DMSC co-cultured with murine liver homogenate, indicating that the liver microenvironment induced the differentiation of DMSC. In addition, the level of AFP expression decreased whereas ALB expression increased during the co-culture. This switch is similar to the gene expression pattern observed during normal liver organogenesis in the embryo (52,53). The hepatic differentiation of DMSC in response to a healthy liver microenvironment is dependent on soluble factors because tissue homogenate and cells are co-cultured in a non-contact system. Interestingly, in response to injured liver homogenate, we observed a similar switch in AFP and ALB expression, although the expression levels of both genes was significantly higher. CCl4 liver damage is a reproducible and widely used in vivo animal model for hepatic injury (54). This model shows rapid endogenous hepatocyte regeneration when low dose and/or short-term CCl4 administration is used (55). We hypothesize that the higher DMSC hepatic differentiation in response to CCl4-treated liver could be a response to factors secreted by the regenerating liver or to molecules released by the damaged hepatocytes and/or both mechanisms working simultaneously. Recently, a study of BM-derived MSC cocultured with normal or injured liver showed similar results (6). Furthermore, AFP gene expression is an indicator of both endodermal embryonic differentiation and liver regeneration, because its expression is increased in adult liver after partial hepatectomy or acute intoxication (52,56,57). The co-culture data suggest that the liver microenvironment would have an effect on transplanted DMSC in vivo to integrate efficiently and differentiate to repair the damaged liver tissue; DMSC could be an alternative source of stem cells for future liver cell therapy. In summary, we hypothesized that DMSC would have a high-affinity, engrafting and differentiation capacity to repair damaged liver after transplantation. Additionally, DMSC-derived hepatospheres and co-culture studies using murine and human liver could be very useful in vitro systems for studying liver physiology and diseases, providing alternative screening systems for toxicologic assays and an alternative to cellular devices in pre-clinical trials for liver transplantation. The results obtained from these experiments should be very useful for the future design of human clinical trials using DMSC. Acknowledgements The authors thank to Dr I. Vegh for critical reading of the manuscript and English grammar revision, and to the veterinarian Montserrat Grau for her assistance with the animal protocols (Research Centre, Research Institute Hospital 12

Hepatic differentiation of decidua-derived MSCs 1191 de Octubre). We also thank to J. C. Gonzalez for his help with the confocal microscopy analysis and J. C. Rubio for his contribution to the real time PCR determinations. Disclosure of interests: The authors declare that no conflict of interests exists. This work was supported by grants to Dr Flores from the Spanish Institute of Health Carlos III (FIS 05/2288; CP03/00004), the MMA Foundation (2004-029; 2005-088), the Neurosciences and Aging Foundation, and a contract associated with Dr Flores’ project 05/2288 from the Hospital 12 Octubre Biomedical Research Foundation. References 1. Thomas RJ, Bhandari R, Barrett DA, Bennett AJ, Fry JR, Powe D, et al. The effect of three-dimensional co-culture of hepatocytes and hepatic stellate cells on key hepatocyte functions in vitro. Cells Tissues Organs. 2005;181:67–79. 2. Soto-Gutierrez A, Yagi H, Uygun BE, Navarro-Alvarez N, Uygun K, Kobayashi N, et al. Cell delivery: from cell transplantation to organ engineering. Cell Trans. 2010;19: 655–65. 3. Snykers S, De Kock J, Rogiers V, Vanhaecke T. In vitro differentiation of embryonic and adult stem cells into hepatocytes: state of the art. Stem Cells. 2009;27:577–605. 4. Lin H, Xu R, Zhang Z, Chen L, Shi M, Wang FS. Implications of the immunoregulatory functions of mesenchymal stem cells in the treatment of human liver diseases. Cell Mol Immunol. 2010;8:19–22. 5. Kode JA, Mukherjee S, Joglekar MV, Hardikar AA. Mesenchymal stem cells: immunobiology and role in immunomodulation and tissue regeneration. Cytotherapy. 2009;11: 377–91. 6. Mohsin S, Shams S, Ali Nasir G, Khan M, Javaid Awan S, Khan SN, et al. Enhanced hepatic differentiation of mesenchymal stem cells after pretreatment with injured liver tissue. Differentiation. 2011;81:42–8. 7. Aurich H, Sgodda M, Kaltwasser P, Vetter M, Weise A, Liehr T, et al. Hepatocyte differentiation of mesenchymal stem cells from human adipose tissue in vitro promotes hepatic integration in vivo. Gut. 2009;58:570–81. 8. Lee KD, Kuo TK, Whang-Peng J, Chung YF, Lin CT, Chou SH, et al. In vitro hepatic differentiation of human mesenchymal stem cells. Hepatology. 2004;40:1275–84. 9. Anzalone R, Lo Iacono M, Corrao S, Magno F, Loria T, Cappello F, et al. New emerging potentials for human Wharton’s jelly mesenchymal stem cells: immunological features and hepatocyte-like differentiative capacity. Stem Cells Devel. 2010;19:423–38. 10. Meng X, Ichim TE, Zhong J, Rogers A, Yin Z, Jackson J, et al. Endometrial regenerative cells: a novel stem cell population. J Trans Med. 2007;5:57. 11. Tamagawa T, Oi S, Ishiwata I, Ishikawa H, Nakamura Y. Differentiation of mesenchymal cells derived from human amniotic membranes into hepatocyte-like cells in vitro. Hum Cell. 2007;20:77–84. 12. Chien CC, Yen BL, Lee FK, Lai TH, Chen YC, Chan SH, et al. In vitro differentiation of human placenta-derived multipotent cells into hepatocyte-like cells. Stem Cells. 2006; 24:1759–68.

13. Miki T, Lehmann T, Cai H, Stolz DB, Strom SC. Stem cell characteristics of amniotic epithelial cells. Stem Cells. 2005;23: 1549–59. 14. De Coppi P, Bartsch G Jr, Siddiqui MM, Xu T, Santos CC, Perin L, et al. Isolation of amniotic stem cell lines with potential for therapy. Nature Biotech. 2007;25:100–6. 15. Tanimizu N, Miyajima A. Molecular mechanism of liver development and regeneration. Int Rev Cytol. 2007;259:1–48. 16. Miki T, Ring A, Gerlach J. Hepatic differentiation of human embryonic stem cells is promoted by three-dimensional dynamic perfusion culture conditions. Tissue Eng Part C Methods. 2011;17:557–68. 17. Yamamoto T, Navarro-Alvarez N, Soto-Gutierrez A, Yuasa T, Iwamuro M, Kubota Y, et al. Treatment of acute liver failure in mice by hepatocyte xenotransplantation. Cell Trans. 2010; 19:799–806. 18. Uygun BE, Soto-Gutierrez A, Yagi H, Izamis ML, Guzzardi MA, Shulman C, et al. Organ reengineering through development of a transplantable recellularized liver graft using decellularized liver matrix. Nature Med. 2010;16:814–20. 19. Vaananen HK. Mesenchymal stem cells. Ann Med. 2005; 37:469–79. 20. Parolini O, Alviano F, Bagnara GP, Bilic G, Buhring HJ, Evangelista M, et al. Concise review. Isolation and characterization of cells from human term placenta: outcome of the first international Workshop on Placenta Derived Stem Cells. Stem Cells. 2008;26:300–11. 21. Marcus AJ, Woodbury D. Fetal stem cells from extraembryonic tissues: do not discard. J Cell Mol Med. 2008;12: 730–42. 22. Macias MI, Grande J, Moreno A, Dominguez I, Bornstein R, Flores AI. Isolation and characterization of true mesenchymal stem cells derived from human term decidua capable of multilineage differentiation into all 3 embryonic layers. Am J Obstet Gynecol. 2010;203:495.e9–495.e23. 23. Watt AJ, Zhao R, Li J, Duncan SA. Development of the mammalian liver and ventral pancreas is dependent on GATA4. BMC Devel Biol. 2007;7:37. 24. Holtzinger A, Evans T. Gata4 regulates the formation of multiple organs. Development. 2005;132:4005–14. 25. de Graaf W, Hausler S, Heger M, van Ginhoven TM, van Cappellen G, Bennink RJ, et al. Transporters involved in the hepatic uptake of (99m)Tc-mebrofenin and indocyanine green. J Hepatol. 2011;54:738–45. 26. Jang YY, Collector MI, Baylin SB, Diehl AM, Sharkis SJ. Hematopoietic stem cells convert into liver cells within days without fusion. Nature Cell Biol. 2004;6:532–9. 27. Mitaka T, Sato F, Ikeda SI, Sugimoto SI, Higaki N, Hirata K, et al. Expression of carbamoylphosphate synthetase I and glutamine synthetase in hepatic organoids reconstructed by rat small hepatocytes and hepatic nonparenchymal cells. Cell Tissue Res. 2001;306:467–71. 28. van Zijl F, Mikulits W. Hepatospheres: three dimensional cell cultures resemble physiological conditions of the liver. World J Hepatol. 2010;27:1–7. 29. Hong SH, Gang EJ, Jeong JA, Ahn C, Hwang SH, Yang IH, et al. In vitro differentiation of human umbilical cord bloodderived mesenchymal stem cells into hepatocyte-like cells. Biochem Biophys Res Comm. 2005;330:1153–61. 30. Roelandt P, Sancho-Bru P, Pauwelyn K, Verfaillie C. Differentiation of rat multipotent adult progenitor cells to functional hepatocyte-like cells by mimicking embryonic liver development. Nature Protocols. 2010;5:1324–36. 31. Yamada T, Yoshikawa M, Kanda S, Kato Y, Nakajima Y, Ishizaka S, et al. In vitro differentiation of embryonic stem cells into hepatocyte-like cells identified by cellular uptake of indocyanine green. Stem Cells. 2002;20:146–54.

1192 R. Bornstein et al. 32. Sugiyama Y, Koike T, Shiojiri N. Immunohistochemical analyses of cell–cell interactions during hepatic organoid formation from fetal mouse liver cells cultured in vitro. Histochem Cell Biol. 2007;128:521–31. 33. Flouriot G, Vaillant C, Salbert G, Pelissero C, Guiraud JM, Valotaire Y. Monolayer and aggregate cultures of rainbow trout hepatocytes: long-term and stable liver-specific expression in aggregates. J Cell Sci. 1993;105:407–16. 34. Lee OK, Kuo TK, Chen WM, Lee KD, Hsieh SL, Chen TH. Isolation of multipotent mesenchymal stem cells from umbilical cord blood. Blood. 2004;103:1669–75. 35. Ong SY, Dai H, Leong KW. Hepatic differentiation potential of commercially available human mesenchymal stem cells. Tissue Eng. 2006;12:3477–85. 36. Talens-Visconti R, Bonora A, Jover R, Mirabet V, Carbonell F, Castell JV, et al. Hepatogenic differentiation of human mesenchymal stem cells from adipose tissue in comparison with bone marrow mesenchymal stem cells. World J Gastroenterol. 2006;12:5834–45. 37. van de Ven C, Collins D, Bradley MB, Morris E, Cairo MS. The potential of umbilical cord blood multipotent stem cells for nonhematopoietic tissue and cell regeneration. Exp Hematol. 2007;35:1753–65. 38. Deutsch G, Jung J, Zheng M, Lora J, Zaret KS. A bipotential precursor population for pancreas and liver within the embryonic endoderm. Development. 2001;128:871–81. 39. Shi XL, Zhang Y, Gu JY, Ding YT. Coencapsulation of hepatocytes with bone marrow mesenchymal stem cells improves hepatocyte-specific functions. Transplantation. 2009;88: 1178–85. 40. Wang Y, Cui CB, Yamauchi M, Miguez P, Roach M, Malavarca R, et al. Lineage restriction of human hepatic stem cells to mature fates is made efficient by tissue-specific biomatrix scaffolds. Hepatology. 2011;53:293–305. 41. Griffith LG, Swartz MA. Capturing complex 3D tissue physiology in vitro. Nature Rev. 2006;7:211–24. 42. Price JA, Caldwell J, Hewitt NJ. The effect of EGF and the comitogen, norepinephrine, on the proliferative responses of fresh and cryopreserved rat and mouse hepatocytes. Cryobiology. 2006;53:182–93. 43. Tateishi K, Ando W, Higuchi C, Hart DA, Hashimoto J, Nakata K, et al. Comparison of human serum with fetal bovine serum for expansion and differentiation of human synovial MSC: potential feasibility for clinical applications. Cell Transplant. 2008;17:549–57. 44. Sessarego N, Parodi A, Podesta M, Benvenuto F, Mogni M, Raviolo V, et al. Multipotent mesenchymal stromal cells from

Supplementary material available online Supplementary Figures 1 and 2.

45.

46.

47.

48.

49.

50.

51.

52. 53. 54.

55.

56.

57.

amniotic fluid: solid perspectives for clinical application. Haematologica. 2008;93:339–46. Pappa KI, Anagnou NP. Novel sources of fetal stem cells: where do they fit on the developmental continuum? Regen Med. 2009;4:423–33. Poloni A, Rosini V, Mondini E, Maurizi G, Mancini S, Discepoli G, et al. Characterization and expansion of mesenchymal progenitor cells from first-trimester chorionic villi of human placenta. Cytotherapy. 2008;10:690–7. Naing MW, Williams DJ. Three-dimensional culture and bioreactors for cellular therapies. Cytotherapy. 2011;13: 391–9. Nussler AK, Zeilinger K, Schyschka L, Ehnert S, Gerlach JC, Yan X, et al. Cell therapeutic options in liver diseases: cell types, medical devices and regulatory issues. J Mat Sci. 2011;22:1087–99. Ring A, Gerlach J, Peters G, Pazin BJ, Minervini CF, Turner ME, et al. Hepatic maturation of human fetal hepatocytes in four-compartment three-dimensional perfusion culture. Tissue Eng. Part C Methods. 2010;16:835–45. Aurich I, Mueller LP, Aurich H, Luetzkendorf J, Tisljar K, Dollinger MM, et al. Functional integration of hepatocytes derived from human mesenchymal stem cells into mouse livers. Gut. 2007;56:405–15. Avital I, Inderbitzin D, Aoki T, Tyan DB, Cohen AH, Ferraresso C, et al. Isolation, characterization, and transplantation of bone marrow-derived hepatocyte stem cells. Biochem Biophys Res Comm. 2001;288:156–64. Lazarevich NL. Molecular mechanisms of alpha-fetoprotein gene expression. Biochemistry. 2000;65:117–33. Zhao R, Duncan SA. Embryonic development of the liver. Hepatology. 2005;41:956–67. Jang JH, Kang KJ, Kim YH, Kang YN, Lee IS. Reevaluation of experimental model of hepatic fibrosis induced by hepatotoxic drugs: an easy, applicable, and reproducible model. Transplant Proc. 2008;40:2700–3. Iredale JP. Models of liver fibrosis: exploring the dynamic nature of inflammation and repair in a solid organ. J Clin Invest. 2007;117:539–48. Kubota H, Storms RW, Reid LM. Variant forms of alphafetoprotein transcripts expressed in human hematopoietic progenitors. Implications for their developmental potential towards endoderm. J Biol Chem. 2002;277:27629–35. Suzuki A, Zheng YW, Fukao K, Nakauchi H, Taniguchi H. Liver repopulation by c-Met-positive stem/progenitor cells isolated from the developing rat liver. Hepato-gastroenterology. 2004;51:423–6.