Comparative Evaluation of Differentiation Potential of Menstrual Blood- Versus Bone Marrow- Derived Stem Cells into Hepatocyte-Like Cells

Menstrual blood has been introduced as an easily accessible and refreshing stem cell source with no ethical consideration. Although recent works have shown that menstrual blood stem cells (MenSCs) possess multi lineage differentiation capacity, their efficiency of hepatic differentiation in comparison to other stem cell resources has not been addressed so far. The aim of this study was to investigate hepatic differentiation capacity of MenSCs compared to bone marrow-derived stem cells (BMSCs) under protocols developed by different concentrations of hepatocyte growth factor (HGF) and oncostatin M (OSM) in combination with other components in serum supplemented or serum-free culture media. Such comparison was made after assessment of immunophenotye, trans-differentiation potential, immunogenicity and tumorigeicity of these cell types. The differential expression of mature hepatocyte markers such as albumin (ALB), cytokeratin 18 (CK-18), tyrosine aminotransferase and cholesterol 7 alpha-hydroxylase activities (CYP7A1) at both mRNA and protein levels in differentiating MenSCs was significantly higher in upper concentration of HGF and OSM (P1) compared to lower concentration of these factors (P2). Moreover, omission of serum during differentiation process (P3) caused typical improvement in functions assigned to hepatocytes in differentiated MenSCs. While up-regulation level of ALB and CYP7A1 was higher in differentiated MenSCs compared to driven BMSCs, expression level of CK-18, detected level of produced ALB and glycogen accumulation were lower or not significantly different. Therefore, based on the overall comparable hepatic differentiation ability of MenSCs with BMSCs, and also accessibility, refreshing nature and lack of ethical issues of MenSCs, these cells could be suggested as an apt and safe alternative to BMSCs for future stem cell therapy of chronic liver diseases.


Introduction
Cell therapy, using human hepatocytes, is being regarded worldwide as an alternative approach to organ transplantation for liver failure. Major obstacles, including donor organ shortage have encouraged scientists and physicians to take advantage of stem cells for cell therapy of liver disorders [1]. Adult bone marrow has commonly been known as the most conventional stem cell source in the field of regenerative medicine and tissue engineering, including liver tissue engineering, because bone marrow-derived mesenchymal stem cells (BMSCs) do not present the ethical issues of embryonic stem cells (ESCs), have hepatogenic differentiation ability in vitro and in vivo [2,3] and exhibit immunosuppressive capabilities [4,5]. However, problems such as less availability, invasive methods for sample collection and lower proliferation capacity in comparison with ESCs limit applicability of BMSCs for clinical therapy of liver diseases. Pertaining to other sources of stem cells and regardless of great achievements in generating terminally-differentiated hepatocyte-like cells from human induced pluripotent stem (iPS) cells, limitations such as the risk of tumor formation are yet to be addressed in this stem cell type [6,7].
Several studies have reported that menstrual blood (MB) contains a unique population of cells with properties similar to adult stem cells [8][9][10][11]. It has been proposed that MB contains circulating BMSCs, which contribute to endometrial regeneration [12]. Menstrual blood-derived stem cells (MenSCs) exhibit a long term self-renewal ability, greater proliferation capacity compared to BMSCs and have minimal risk of karyotypic abnormalities [8][9][10][11]13]. In addition, recent studies have showed that reprogramming efficiency for generation of iPS cells could be increased using MenSCs as a cell source even in the absence of ectopic expression of c-Myc [14,15]. These characteristics, as well as the ease of access and the possibility of cyclic sample collection, make MB an appropriate stem cell supply for tissue engineering and regenerative medicine.
We have previously presented evidence on the potential of MenSCs to generate hepatocyte-like cells [16]. In the present study, we sought to compare for the first time the hepatogenic differentiation potential of MenSCs to that of BMSCs.
To date, different concentrations of hepatocyte growth factor (HGF) and oncostatin M (OSM), as main factors involved in hepatocyte development and maturation, in combination with other cytokines and growth factors, such as epidermal growth factor (EGF), basic fibroblast growth factor (b-FGF) and insulinlike growth factor (IGF), have been used to generate both biochemically-and metabolically-active hepatocyte-like cells from BMSCs. Indeed, chemical compounds such as dexamethasone (Dexa), retinoic acid (RA), sodium butyrate, nicotinamide (NTA), norepinephrine, and dimethylsulfoxide (DMSO) are known to promote hepatic differentiation of BMSCs [17][18][19][20]21]. Moreover, much effort is underway to induce hepatic differentiation of BMSCs under serum-free conditions for human trial applications [22][23][24][25]26]. In this study, we have presented some data about: 1) unique immunophenotypic characteristics and trans-differentiation capability of MenSCs compared to BMSCs, and 2) chromosomal stability, and non-immunogenic and non-tumorigenic nature of MenSC. As a main goal, differentiation potential of MenSCs into hepatocyte-like cells has been compared to BMSCs under three protocols developed using combinations of HGF and OSM with other components in both serum-supplemented and serum-free culture media.

Isolation of MenSCs and BMSCs
Menstrual blood was collected from healthy females aged 25-35 years using a Diva cup (Diva International Co., Lunette, Finland) on day 2 of menstrual cycle. All donors were given a complete description of the study and provided written informed consent before sampling. This study was approved by the Medical Ethics Committee of Avicenna Research Institute.
The contents of Diva cup were transferred into tubes containing 2.5 mg/ml fungizone (Gibco, Scotland, UK), 100 mg/mL streptomycin, 100 U/mL penicillin (Sigma-Aldrich, MO, USA) and 0.5 mM EDTA in phosphate buffered saline (PBS) without Ca 2+ or Mg 2+ . Isolation of stem cells from menstrual blood was performed as described previously [27,28].
Bone marrow aspirates (5-10 mL) were obtained from iliac crests of human donors at the Bone Marrow Transplantation Center, Shariati Hospital, Tehran, Iran. Samples were harvested after getting signed informed consent. This work was approved by the Medical Ethics Committee, Ministry of Health, Iran. Bone marrow stem cells (BMSCs) were isolated using a combination of density gradient centrifugation and plastic adherence as described in our previous studies [29].
Indeed, cells were fixed in acetone at 220uC for 5 min and then subjected to immunofluorescent staining for OCT-4, vimentin and GFAP. In brief, the fixed cells were permeabilized with 0.4% triton X-100 for 20 min. After washing step, cells were incubated for 1 h at room temperature (RT) with rabbit anti-human OCT-4 polyclonal antibody (Abcam), mouse anti-human vimentin monoclonal antibody (clone V9, 1:200; Sigma) or rabbit anti-human GFAP monoclonal antibody (clone name:EP672Y, 1:250). As reagent negative control, the cells were treated in parallel with the same concentrations of normal rabbit irrelevant IgG for OCT-4 and GFAP and mouse irrelevant IgG1 for vimentin.
Subsequently, the cells were washed three times with PBS and incubated with FITC-labeled goat anti-rabbit IgG (Sigma) or FITC-labeled sheep anti-mouse IgG (Avicenna Research Institute) at RT for 45 minutes in the dark. Thereafter, cells were incubated with 49, 6 diamidino-2-phenylindole (DAPI; 1:1000) (Sigma-Aldrich) for nuclear staining. The cells were visualized and photomicrographed using an epifluorescence microscope (Olympus BX51 microscope, Tokyo, Japan) connected to digital camera (Olympus DP71, Tokyo, Japan).

Multi-lineage differentiation potential of MenSCs and BMSCs
To further characterization of isolated MenSCs in comparison with BMSCs, we assessed differentiation ability of these cells into osteoblasts, chondrocytes and adipocytes as described previously (27,28). The differentiated cells into osteoblasts were identified by specific histochemical staining for calcium with Alizarin red staining (Sigma-Aldrich). Chondrogenesis was assessed by immunofluorescence staining using primary monoclonal mouse antihuman Collagen type II (clone 5B2.5, 1:500; Abcam) and FITClabeled goat anti-mouse IgG (Abcam). Adipogenic-induced cells were stained for fat vacuoles using the Oil red O staining. Control cultures without the differentiation stimuli were maintained in parallel to the differentiation experiments and stained in the same manner.

Multiplex Ligation-dependent Probe Amplification (MLPA)
To investigate chromosomal stability of MenSCs during passages, MLPA analysis was performed on genomic DNA of cells at passages 2 and 12 using the SALSA MLPA kit P036-E1 Human telomer3 (MRC-Holland, Netherlands) according to the manufacturer's protocol. Briefly, a total of 100 ng of genomic DNA in a final volume of 5 ml was denatured and hybridized to SALSA probe mix, followed by incubation at 60uC for 18 hr. Subsequently, the annealed probes were ligated using provided Ligase-65 mix at 54uC for 15 min. In the next step, 10 ml of ligated products, as template, were used for DNA amplification. The PCR amplicons were run on a Genetic Analyzer 3130 (Applied Biosystems, USA), and the results were analyzed by GeneMarker software version 2.4 (SoftGenetics, USA). The normal pattern was expected to produce a normalized signal value ratio of 1:1; any value out of the ranges ,0.75 or .1.35 was considered as abnormal and corresponded to a deletion or duplication, respectively. In each MLPA reaction, whole blood of adult people with no evidence of genetic anomalies, cancerous tissue of colorectal cancer with chromosomal abnormality and aborted fetus with monosomy 21 were simultaneously used as controls. In addition, all control samples were screened for chromosomal abnormality through karyotyping and a normal 550 GTG banding protocol [30].

Examination of MenSCs immunogenicity and tumorigenicity
Testing of in vivo immune response to MenSCs was carried out in five immunocompetent male C57BL/6 mice aged 6 to 8 weeks. Moreover, five nude mice were used to evaluate MenSCs tumorigenicity. To this end, MenSCs at passages 2-4 were harvested using trypsin-EDTA and cell density was adjusted to 1610 6 cells/ml in serum-free medium. After that, 0.2 ml of cell suspension was subcutaneously injected in the dorsolateral flank of C57BL/6 mice and nude mice. After two weeks, 20-30 ml of sera was obtained from C57BL/6 mice and sera immunoreactivity to cultured MenSCs was evaluated using immunofluorescence staining.
In order to test tumorigenicity potential, the treated nude mice were scarified and autopsied 12 weeks after cells injection. The inoculation site from the deep aspect was inspected and excised. Moreover, to evaluate the metastatic potential of the MenSCs, the ipsilateral axillary lymph node, spleen, liver, lung, kidneys and heart were excised for histological examination using hematoxylin and eosin (H & E) staining. All procedures were approved by the animal care and ethics committee of Avicenna Research Institute.

Hepatogenic differentiation
MenSCs and BMSCs were plated at a density of 1.5610 4 cells/ cm 2 in 1% ECM gel (derived from Engelbreth-Holm-Swarm mouse sarcoma, Sigma-Aldrich)-coated T-25 flasks. In order to induce hepatogenic differentiation, sequential exposures to three different combinations of cytokines, growth factors and hormones were examined. In the commitment step of protocol 1 (P1), 60-70% confluent MenSCs cultured in DMEM-F12 medium were supplemented with 10% fetal bovine serum (FBS) and fortified with 20 ng/ml EGF and 10 ng/ml b-FGF (Sigma-Aldrich) for 2 days prior to differentiation step. Differentiation was induced by treating cells with 5% FBS-supplemented media containing 10 27 M Dexa, 1% (insulin, transferrin, selenium pre-mix) ITS+1, 50 mg/ml NTA and 40 ng/ml HGF (all from Sigma-Aldrich). After 14 days, differentiation step was followed by maturation step using a substituted media containing 5% FBS, 10 27 M Dexa, 1% ITS +1 and 20 ng/ml OSM (Sigma-Aldrich) for an additional 14 days. Protocol 2 (P2) was done in a similar manner as P1, but with half concentration of HGF (20 ng/ml) and OSM (10 ng/ml). In protocol 3 (P3), cells at 100% confluence were serum-deprived during the treatment period. In this protocol, cells were committed in serum-free DMEM medium supplemented with 20 ng/ml EGF and 10 ng/ml b-FGF for 2 days. Differentiation and maturation steps were conducted as in P1, but in serum-free media. Media in all culture conditions were changed every two days and each protocol lasted up to 30 days. Control cultures without the differentiation stimuli were maintained in parallel to the differentiation experiments in the same manner. The isolated hepatocytes from aborted fetuses (8-12 weeks) or HepG2 cell line were used as positive control. The summary of the differentiation protocols is shown in Figure 1.

Immunofluorescence staining of albumin (ALB), cytokeratin 18 (CK-18) and Alpha-fetoprotein (AFP)
On the day 30 of differentiation, cells were harvested enzymatically, using 0.25% trypsin-EDTA solution. Cytospins were prepared by centrifugation of the cell suspension onto poly-L-lysine coated glass slides. The cells (10 4 cells per slide) were fixed in acetone at 220uC for 5 min and incubated overnight at 4uC with primary monoclonal mouse anti-human ALB (clone HAS-11, 1:100; Sigma-Aldrich), AFP (clone C3, 1:200; Sigma-Aldrich) or CK-18 (clone RGE53, 1:100; Chemicon International, Temecula, CA) antibodies. As negative control, cells were treated in parallel with the mouse irrelevant IgG2a for ALB and AFP and IgG1 for CK-18 at the same dilutions. Subsequently, the cells were washed three times with PBS and incubated with FITC-labeled sheep antimouse IgG (Avicenna Research Institute) at RT for 45 min in the dark. After washing with PBS, the cells were processed as for OCT-4 immunostaining. Human HepG2 hepatoma cells (National Cell Bank, Pasteur Institute, Tehran, Iran) were simultaneously stained for ALB and CK-18 and considered as positive control. The green fluorescence intensity of 30 protein marker/DAPI positive cells was determined using ImageJ software (ImageJ, NIH, Bethesda, Maryland, USA, http://imagej.nih.gov/ij/, 1997-2012).
To assess expression of CYP7A1, 1 mL of cDNA was admixed with 12.5 mL reaction master mix (Amplicon) and 1 mL of each primer (Table 1). After initial denaturation at 94uC for 3 min, PCR amplification was continued at 94uC for 30 s, 60uC for 30 s, and 72uC for 30 s for total cycles of 35, and a final extension was performed at 72uC for 7 min. GAPDH amplification was used as an internal standard. The amplified DNA fragments were electrophoresed on a 2% agarose gel and visualized by an ultraviolet transilluminator (Uvitec-USA). For semi-quantitative determination, specific band density was normalized to that of the corresponding GAPDH using AlphaEase software (Genetic Technologies, Inc.).

Albumin secretion
We quantified ALB in the media to assess ALB secretion ability of differentiated cells, a reliable assay to evaluate hepatic metabolic function [2,32]. Sample culture media was collected on days 0, 15, 20 and 25 of differentiation and measured by human ALB ELISA quantitative Kit (Immunology Consultant Lab, OR, USA) according to the manufacturer's instructions. The measurements were performed in triplicate.

Glycogen storage
Intracellular glycogen was analyzed by Periodic Acid-Schiff (PAS) staining. The cells were seeded on poly L-lysine mounted slides and fixed in 4% paraformaldehyde. The slides were oxidized in 1% periodic acid for 5 min and rinsed three times in deionized

Statistical analysis
All experiments were performed using cells at passages 2-4 from 3-6 donors. All measurements were performed in triplicate. The results of flow cytometry were presented as median 6 range. ALB quantification using ELISA was analyzed using Mann-Whitney Utest and the data were expressed as mean 6 standard deviation (SD). For the statistical analysis, the SPSS 13 software was used and P value,0.05 was considered significant.
Statistical analysis of relative gene expression results in real time PCR was performed using RESTß freeware according to formula presented by Pfaffl et al [37].

In vitro and in vivo characterization of cultured cells
The MenSCs like BMSCs had a same spindle-shaped and fibroblast-like morphology (Fig. 2A). The immunophenotyping analysis of the cells by flow cytometry and immunofluorescence staining revealed that MenSCs in a similar manner with BMSCs typically expressed CD105, CD166, CD146 and vimentin, while unlike BMSCs failed to express CD106 and GFAP. In contrast, expression of OCT-4 was only observed in MenSCs. Lack of CD133 and CD45 in cultured cells reflected a non-hematopoietic stem cell phenotype (Fig. 2B and 2C).
MenSCs differentiated into chondrocytes had strong immunoreactivity with monoclonal antibody against Collagen type II in a similar manner with differentiated BMSCs. Mineralization was also pronounced in both osteoblast-differentiated MenSCs and BMSCs as showed by Alizarin red staining. However, the mineralization potential of MenSCs was much lower than that of BMSCs. Meanwhile, unlike formation of oil droplets in differentiated BMSCs, Oil red O staining of the differentiated MenSCs into adipocytes was negative (Fig. 3A).
MLPA analysis showed that MenSCs, in contrast to positive controls, maintained diploid phenotype without chromosomal aberrations during passages (Fig. 3C).
According to histological examination of nude mice injected with MenSCs (Fig. 4A & B), no evidence of tumor growth was found in inoculation site and the examined tissues had no morphological characteristics of tumor as judged by H & E staining. Moreover, analysis of immunoreactivity of mice sera to cultured MenSCs revealed no positive reaction in immunofluorescence staining (Fig. 4C).

Phenotypic characterization of differentiated cells into hepatocyte-like cells
Morphological assessment of the cells using phase contrast microscopy (Fig. 5) showed that under all differentiation conditions, the fibroblast-like morphology of MenSCs and BMSCs did not change significantly during the commitment step. The hepatocyte-like morphology, evidenced by binucleation and cytoplasmic granulation, was observed when both cell types were exposed to differentiation media under protocols 1 and 3 (P1 and P3), with more polyhedral contours and binuclation under P3 compared to P1. These morphological changes were more obvious in differentiated MenSCs compared to differentiated BMSCs. The MenSCs and BMSCs cultured under P2 did not exhibit significant changes in cell morphology during differentiation.

Immunofluorescent staining of Albumin, CK18 and AFP
Albumin (ALB) was not detectable in undifferentiated MenSCs or BMSCs, whereas there was a significant level of ALB expression at the end of differentiation process in both MenSCs and BMSCs under all conditions. The intensity of ALB expression in MenSCs was dependent on differentiation protocol, with the highest level under P3 (44.86618.34) and lowest under P2 (18.8564.15). The mean intensity of ALB expression was 27.45610 in differentiated cells under P1 (Fig. 6A). Nonetheless, the differentiated MenSCs showed a considerably higher level of ALB expression compared to differentiated BMSCs under P1 and P3. The mean intensity of ALB expression was highest when BMSCs differentiated under P3 (25.26615.2) and lowest (12.8563.2) in differentiated BMSCs under P2.
Expression of CK-18 in differentiated MenSCs under P1 and P2 was not different with undifferentiated cells, but, the treated MenSCs using P3 showed significantly higher expression level of this protein compared with other groups (Fig. 6B). In contrast, relatively the same levels of AFP expression were observed in MenSCs differentiated by either of protocols and undifferentiated cells (Fig. 6C).

Hepatic gene expression
As shown in figure 7A, the MenSCs showed a significant upregulation of ALB gene following differentiation under P1  Although the expression level of ALB gene in both differentiated BMSCs and MenSCs under all protocols was lower than those observed in isolated fetal hepatocytes, the expression levels of TAT, CK-18 and CK-19 genes in both differentiated cells were significantly higher than those of isolated fetal hepatocytes in a protocol-dependent manner (Fig. 7B).

Functionality assay
Glycogen storage and ALB secretion were evaluated to determine whether the hepatocyte-like cells also had functional features related to hepatocytes. Our analysis for ALB secretion indicated that undifferentiated MenSCs and BMSCs did not secrete ALB, whereas after hepatic induction under all protocols, a time-dependent increase in the level of secreted ALB was observed (Fig. 8A) Based on the findings of PAS staining (Fig. 8B), all differentiation protocols resulted in glycogen accumulation in derived MenSCs or BMSCs, however the intensity of PAS staining varied dependent on the differentiation protocol and cell type. The highest PAS intensity was in both cell types was observed in P3 (4+). While the intensity of PAS staining had no gross difference between BMSCs induced by P1 and P2 (both 2+), the differentiated MenSCs under P1 exhibited higher level of glycogen deposits compared to those induced by P2 (3+ versus 1+). Furthermore, mRNA expression of Cytochrome P450 7A1 (CYP7A1) was examined for further functional characterization of differentiated hepatocyte-like cells. As shown in figure 8C, MenSCs in undifferentiated state did not express the CYP7A1 gene, however, expression of this marker was detectable on the last day of differentiation process under all protocols. Unlike differentiated BMSCs that expressed high levels of CYP7A1 under all protocols, differentiated MenSCs exhibited a protocol-dependent variation in CYP7A1 expression. Interestingly, in consistence with data obtained by other functionality tests, differentiated MenSCs under P3 expressed the highest level of CYP7A1 gene.

Discussion
Menstrual blood is an interesting stem cell source in the field of regenerative medicine and tissue engineering, because it is abundant and easy to access. Furthermore, MenSCs isolated from MB can be expanded in quantities relevant to clinical applications without chromosomal abnormalities. Non-immunogenic nature and lack of tumor formation capability, as shown in our study, are among other outstanding advantages that make MenSCs a suitable candidate for cell therapy. While current reports of animal studies on MenSCs-based therapy for some diseases such as cardiac [38], muscular [39] and nervous system disorders [40,41] hold promise for cell-based therapy using these stem cells, there is very limited information about MenSCs potential to generate hepatocytes [8,16], particularly compared with other known stem cells such as BMSCs. In this study, some protocoldependent differences were beheld between hepatogenic differentiation potential of MenSCs and BMSCs.
We showed recently that MenSCs possess some markers of mesenchymal stem cells such as CD44, CD73, CD29 and CD90 but fail to express STRO-1 [27,28]. To further characterization of these cells, in this study we evaluated expression of other mesenchymal and embryonic stem cell markers. The high expression level of BMSCs markers such as CD146, CD166, CD105 and vimentin in parallel to OCT-4, the later being an embryonic stem cell marker, suggests a dual characteristic for MenSCs. On the other hand, lack of some other mesenchymal (STRO-1, CD106, and GFAP) and embryonic (NANOG and SSEA-4) stem cell markers distinguishes MenSCs from these stem cell types [27,28].
Our preliminary evaluation indicated a trans-differentiation ability of MenSC population into chondrogenic and osteogenic lineages in a similar manner with BMSCs. However, differentiation capability of MenSCs to adipocytes was significantly lower than that of BMSCs. Considering aforesaid differences, comparative evaluation of MenSCs and BMSCs in terms of differentiation into other lineages including hepatic lineage would help us to understand other potential distinctive characteristics that they may present. To our best knowledge, such data has not been reported so far. In the present study, to determine hepatogenic differentiation potential of MenSCs especially compared with BMSCs, we developed three protocols (P1-P3) using HGF and OSM, in two combinations with other components in both serum-supplemented and serum-free culture media. All differentiation processes were entirely examined on ECM-seeded cells, considering the well- known growth and hepatogenic differentiation promoting effect of ECM on stem cells [42,43].
Here we showed that all induced MenSCs using three protocols acquired some features of functional mature hepatocytes such as mRNA and/or protein expression of ALB, CK-18, TAT, CYP7A1 and also ALB secretion and glycogen accumulation (table 2). Moreover, the differentiated MenSCs showed a slight expression of oval cell (mature hepatocyte precursor) markers, CK-19 and AFP [33,34,35], suggesting presence of a small portion of these cells in mature hepatocyte population.
Based on our results, degree of differentiation was protocoldependent. In comparison with differentiated MenSCs in inducing media containing 20 ng/ml HGF and 10 ng/ml OSM (P2), cells differentiated under higher concentration of these factors (P1) displayed upper level of ALB at the protein level. In addition, the mean intensity of ALB signal in differentiated MenSCs was higher than differentiated BMSCs under P1 and P3. These differences were also reflected in the results of secreted ALB levels. In consistent with these results, expression level of CYP7A1 and ALB mRNA was significantly higher in differentiated MenSCs under P1 in reference to those developed by P2. Moreover, differentiated hepatocyte-like cells were able to accumulate more glycogen in differentiated cells induced by P1 compared to those by P2.
Therefore, although expression level of CK-18 mRNA had no significant difference between induced MenSCs by P1 and P2, accumulative data showed P1 was more efficient to differentiate MenSCs into hepatocyte-like cells compared to P2.
In a similar manner with MenSCs, differentiated BMSCs under P1 showed a higher level of mature hepatocyte markers including ALB mRNA, CYP7A1 mRNA and produced ALB protein   to those induced by P2. While some parameters such as CYP7A1 and CK-18 and secreted albumin showed no significant differences between differentiated MenSCs and BMSCs, the expression level of ALB and CK-18 at mRNA and/or protein level was significantly different between these two cell types. The greatest expression of hepatic markers in both MenSCs and BMSCs induced by P1 compared to those differentiated by P2 presents an additional evidence of critical role of HGF and OSM in stem cell differentiation which has also been reported by others [29,26,44]. On the other hand, it sounds that the source of stem cells serves a fundamental role on expression pattern of mature hepatic markers. In consistent with the this conclusion, significant differences in hepatic differentiation potential were reported between stem cells derived from various origins including placental-derived stem cells and BMSCs [45].
In sharp contrast to the main role of serum in stem cell expansion, serum-free conditions have been applied on a routine basis for hepatic differentiation of embryonic or adult stem cells [46,47,26]. In our knowledge, there are no reports available on hepatogenic fate of stem cells in serum-free media compared with that in serum-fortified media. Surprisingly, we found that omission of FBS during three steps of differentiation process (P3) causes typical improvement in functions assigned to hepatocytes. Some morphological characteristics of the differentiated MenSCs under P3, such as binucleation, granulation of the cytoplasm, and formation of polyhedral contours, resembled native hepatocytes. In addition to the morphological appearance, the expression of ALB protein, TAT, CK-18 and CYP7A1 in mRNA and/or protein levels under P3 was typically superior compared to those of MenSCs differentiated under both P1 and P2. Although ALB mRNA and secreted level of ALB showed no significant differences between MenSCs treated under P3 and P1, the further up-regulation of other mentioned parameters suggested more efficacy of P3 than P1 and P2 to develop hepatocyte-like cells from MenSCs. In line with this conclusion, P3-induced MenSCs accumulated more glycogen than those primed by P1 or P2. Although molecular mechanisms governing the higher efficiency of supplemented serum-free in comparison with serum-fortified media for MenSC differentiation remain to be determined, data suggest that FBS deprivation from differentiation media may serve as a triggering factor for MenSCs differentiation.
Compared to differentiated MenSCs, driven BMSCs exhibited more variable expression pattern during differentiation under P1 and P3. While up-regulation level of ALB and CYP7A1 was higher in BMSCs induced by P1 compared to P3, the opposite pattern was found in the case of CK-18 expression and glycogen accumulation. Indeed, BMSCs differentiated under P1 and P3 showed the same levels of ALB secretion. Thus, based on these results superiority of P3 over P1 in BMSCs is not as certain as MenSCs and further studies are needed to counsel a robust notion. In addition, with respect to high costs and safety problems with FBS, serum-free composition of P3 would be advantageous over P1 that contains FBS during differentiation of both MenSCs and BMSCs.
Differentiated MenSCs and BMSCs showed upper level of CK-18, CK-19 and TAT and lower level of ALB compared with isolated fetal hepatocytes. Such inconsistency has also been reported for hepatocyte-like cells derived from iPS cells [48,49]. It indicates that the other supplementary strategies, such as threedimensional cultures or co-culture may be required to further induce the acquisition of native mature hepatocyte nature.
Taken together, the evidence presented in this study indicates MenSCs are a stem cell population with unique immunophenotype and differentiation potential and no risk of tumor formation and immunological reaction. We proved MenSCs have capability to generate functional hepatocyte-like cells, but with different pattern compared to BMSCs depending on critical growth factor concentration and culture media condition. P3 enriched in vitro production of functional MenSC-derived hepatocyte-like cells more than P1 and especially P2. However, in the case of BMSCs, P1 and P3 exhibited rather same efficacy. Differences in protocoldependent expression pattern of hepatogenic markers between these cells may be attributed to different immunophenotypic features and signaling machinery.
On the other hand, regarding overall expression pattern of hepatic markers in differentiated MenSCs with BMSCs, it sounds that hepatic differentiation potential of these two stem cells had no significant predominance to each other. Here again, considering the refreshing nature, accessibility and lack of ethical issues, MenSCs are superior over BMSCs. Further studies using in vivo models and more functional tests are needed before a firm conclusion on applicability of MenSCs-derived hepatocyte-like cells in the clinic or could be made.