SEARCH

SEARCH BY CITATION

Keywords:

  • bone marrow stromal stem cells;
  • valproic acid;
  • differentiation;
  • hepatocyte;
  • epigenetic modification

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Bone marrow stromal stem cells (BMSSCs) may have potential to differentiate in vitro and in vivo into hepatocytes. Here, we investigated the effects of valproic acid (VPA) involved in epigenetic modification, a direct inhibitor of histone deacetylase, on hepatic differentiation of mouse BMSSCs. Following the treatment of 2.5 mM VPA for 72 hrs, the in vitro expanded, highly purified and functionally active mouse BMSSCs from bone marrow were either exposed to some well-defined cytokines and growth factors in a sequential way (fibroblast growth factor-4 [FGF-4], followed by HGF, and HGF + OSM + ITS + dexamethasone, resembling the order of secretion during liver embryogenesis) or transplanted (caudal vein) in mice submitted to a protocol of chronic injury (chronic i.p. injection of CCl4). Additional exposure of the cells to VPA considerably improved the in vitro differentiation, as demonstrated by a more homogeneous cell population exhibited epithelial morphology, increasing expression of hepatic special genes and enhanced hepatic functions. Further more, in vivo results indicate that the pre-treatment of VPA significantly increased the homing efficiency of BMSSCs to the site of liver injury and, additionally, for supporting hepatic differentiation as well as in vitro. We have demonstrated the usefulness of VPA in the transdifferentiation of BMSSCs into hepatocytes both in vitro and in vivo, and regulation of fibroblast growth factor receptors (FGFRs) and c-Met gene expression through post-translational modification of core histones might be the primary initiating event for these effects. This mode could be helpful for liver engineering and clinical therapy.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Acute liver failure is a severe liver disease with 60–90% mortality. The only therapeutic option, orthotopic liver transplantation, is limited because of the shortage of suitable donor organs. Recently, hepatocyte transplantation has emerged as a promising and less aggressive method to treat liver disease. Nevertheless, difficulty to obtain freshly isolated hepatocytes and barriers in cell managing encumbered application of this procedure. Accordingly, it will be greatly beneficial if we could efficiently induce functional hepatocytes from autologous non-hepatic sources.

Bone marrow stromal stem cells (BMSSCs, also known as mesenchymal stem cells [1]) are readily accessible from bone marrow and are capable of differentiate into mesoderm cell lineages such as osteoblasts, adipocytes and chondrocytes [2–4]. In recent years, a number of studies also demonstrated that BMSSCs encompass a plasticity of multiple cell lineages, such as epidermal-like cells, neurons and hepatocyte both in vivo and in vitro[3, 5–9]. The autologous nature of BMSSCs, together with their multi-potentiality, and their usage might sidestep obstacles, such as ethical concerns and risks of rejection, makes these cells an excellent choice for future tissue engineering and cell based therapies [10–14]. Besides clinical therapy, efficient protocols for directing the hepatic differentiation of stem cells in vitro will provide a model that is more amenable to study hepatogenesis and liver metabolism. In the passed decade, several hepatic differentiation protocols from bone marrow cells of different origin have been established [3, 5–7, 9]. However, the incidence of bone marrow-derived hepatocytes was low. Moreover, a long culture period is needed in most cases. Obviously, a more rapid and efficient response of bone marrow-derived liver cells are required before bone marrow cells become a therapeutic option for patients with liver disease.

Histone deacetylase (HDAC) inhibitors, such as valproic acid (VPA), sodium butyrate and trichostatin A, have recently exhibited profound therapeutic activities in a pre-clinical tumour test, which was mediated by their ability to regulate the expression of specific proliferative and apoptotic genes [15–18]. In present study, we aimed to investigate the role of VPA on hepatic differentiation of BMSSCs. The results indicated that additional exposure of the cells to VPA considerably increased the hepatic differentiation in vitro and improved the liver injury by increased homing efficiency of cells to the damaged site in vivo.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Isolation, culture and characterization of BMSSCs

Mice BMSSCs were prepared as described previously [6]. Briefly, the marrow was extruded by clipping of the epiphysial ends of the bones and flushing with Iscove’s Modified Dulbecco’s Medium (IMDM) (Sigma, St. Louis, MO, USA), supplemented with 10% foetal bovine serum (Hyclone, Rockville, MD, USA), 1% penicillin/streptomycin. After three days, non-adherent cells and debris were removed, and the adherent cells were cultured continuously. At near confluence, the cells were re-plated at 100 cells/cm2. Osteogenic and adipogenic differentiation were examined for functional identification, and surface expression of CD11b, CD44, CD45, CD73, CD90, CD105, SCA-1 and STRO-1 were analysed.

Induction of hepatic differentiation in vitro

To determine whether VPA has an effect on in vitro hepatic differentiation of BMSSCs, cells of passage 3 were divided into two groups and exposed to VPA and well-defined cytokines in a sequential way as presented in Fig. 1. Briefly, in group A, cells were pre-treated with 2.5 mM VPA for 3 days, followed by a combined treatment of 10 ng/ml fibroblast growth factor-4 (FGF-4), 20 ng/ml hepatocyte growth factor (HGF), 10 ng/ml oncostatin M (OSM), 1 × ITS (insulin–transferrin–sodium selenite) and 20 μg/l Dex (R&D Systems, Abingdon, UK) at defined time-points that closely resembles the secretion pattern during liver ontogeny. In group B, cells were treated with the cytokines only and devised as control.

image

Figure 1. Schematic presentation of the differentiation protocols. BMSSCs were initially plated at 1 × 103 cells/cm2 on collagen type I coated dishes in expansion medium. When 90% confluency, cells were exposed to valproic acid and different cytokines and growth factors in a sequential way. Group A: experimental cells, Group B: control cells. FGF-4: fibroblast growth factor-4, HGF: hepatocyte growth factor, OSM: oncostatin M, ITS: insulin–transferrin–sodium selenite, Dex: dexamethasone.

Download figure to PowerPoint

Immunofluorescence staining

Cells were fixed with 4% paraformaldehyde, permeabilized with cold methanol, and incubated with primary antibodies, including rabbit anti-alpha-fetoprotein (anti-AFP), rat anti-CK18 and goat anti-albumin (anti-ALB) (Biodesign, Saco, ME, USA). The secondary antibodies including fluorescein isothiocyannate (FITC)-conjugated rabbit anti-goat IgG, goat anti-rabbit IgG, tetrametrylrhodarnine isothiocyante (TRITC) conjugated goat anti-rat immunoglobulin G (IgG) and rabbit anti-goat IgG (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used according to the manufacturer’s instructions. After further washings or counterstained with DAPI, samples were photomicrographed under a confocal laser-scanning microscope (LSM 510; Carl Zeiss, Inc., Jena, Germany).

Periodic acid-Schiff (PAS) stain for glycogen

PAS stain for glycogen was performed as described previously [6].

Detection of cytochrome P450 (CYP) enzyme activities

To enhance CYP expression, basal medium containing 2 mM Phenobarbital (Wako Pure Chemical, Osaka, Japan) was renewed every day for 3 days.

The cells were washed with phosphate buffered saline (PBS) and then treated with 5 mM 7-ethoxyresorufin as a substrate for 3 hrs at 37°C. 7-Ethoxyresorufin-O-deethylase-activity catalysed by CYP2B was detected with a fluorescence detector.

Gene expression analysis by RT-PCR

Total RNA was isolated and complimentary DNA was synthesized by reverse transcription. The PCR primers and conditions to detect gene expression are shown in Table 1. The PCR products were fractionated by 1.2% agarose gel electrophoresis and visualized after staining with ethidium bromide under UV illumination.

Table 1.  Primers, product sizes, and positive controls for PCR analysis
Gene Product size, bp Positive control Primer sequences
Forward Reverse
β-actin200  5′-TTCCTTCTTGGGTATGGAAT-3′ 5′-GAGCAATGATCTTGATCTTC-3′
AFP300Liver 5′-CACTGCTGCAACTCTTCGTA-3′ 5′- CTTTGGACCCTCTTCTGTGA-3′
HNF3-β551Liver 5′-GACCTCTTCCCTTTCTACCG-3′ 5′- TTGAAGGCGTAATGGTGC-3′
ALB475Liver 5′-TCTTCGTCTCCGGCTCTG-3′ 5′- CTGGCAACTTCATGCAAA-3′
TAT619Liver 5′-CTTCAGTCCTGGATGTTCGC-3′ 5′- CAGGGATTGGACGGGTTGTT-3′
FGFR-1IIIc344Brain 5′-CTTGACGTCGTGGAACGATCT-3′ 5′-AGAACGGTCAACCATGCAGAG-3′
FGFR-2IIIb216Lung 5′-CCCATCCTCCAAGCTGGACTG-3′ 5′-CAGAGCCAGCACTTCTGCATTG-3′
FGFR-2IIIc330Lung 5′-CCCATCCTCCAAGCTGGACTG-3′ 5′-TCTCACAGGCGCTGGCAGAAC-3′
FGFR-4179Liver 5′-TTGGGCAAGTGGTTCGT-3′ 5′-CAGCAGGTTGATGATGTTCT-3′
c-Met198Liver 5′-TCGGACAGAGTTTACCACG-3′ 5′-TCCAGGAGGAAGTTCACAT-3′

Real time-PCR analysis

For detailed analysis on FGF-4 and HGF receptors, RNA expression normalized to the housekeeping gene β-actin was measured by real-time PCR using the Mastercycler ep realplex (Eppendorf, Germany) and Real-Time Detection System software. Real time-PCR monitoring was performed by adding the double-stranded DNA dye, SYBR Grean I.

Evaluation of cell proliferation and cell cycle analysis

To determine the cellular proliferation rate after inoculation with VPA, BMSSCs were plated at 1000/cm2, and various concentrations of VPA were added the next day. After 72 hrs, the cells were detached with 0.25% trypsin and total cell numbers were measured with a haemocytometer. For cell cycle assay, dissociated cells were collected and analysis by DNA content (propidium iodide).

Analysis of VPA-mediated epigenetic modification and chromatin structure

Histone acetylation analysis

Cells were exposed to 2.5 mM VPA for 72 hrs and acetylated histones were visualized using anti-acetylated histone H3 (mouse monoclonal to histone H3 [acetyl K9], Abcam, UK) and H4 (Rabbit monoclonal to Histone H4 [acetyl K8], Abcam, UK) antibodies by Western blot and immunofluorescence staining analysis according to the manufacturer’s instructions.

Chromatin structural analysis by electron microscopy

Cells were fixed in 2.5% glutaraldehyde at 4°C, and were rinsed with PBS, dehydrated in increasing ethanol concentrations and embedded in water-permeable London Resin (LR) White plastic. Samples were sectioned at 90-nm thickness and stained for 2 min, in 1% uranyl acetate and for 5 min, with lead citrate, and finally examined on a Philips 100 M electron microscope (Philip Electron Optics, Eindhoven, The Netherlands).

DNase I digestion and Micrococcal nuclease digestion assay

The DNase I digestion and Micrococcal nuclease digestion assay was performed as described previously [19, 20].

Animal treatment

Four- to eight-week-old CD-1® (ICR) mice, wild-type and enhanced green fluorescent proteins (EGFP) transgenic C57BL/6 mice were used in all experiments. All experimental procedures were performed according to institutional guidelines. To induce liver damage, mice were treated with CCl4 (1.0 ml/kg body weight of a 10% solution in mineral oil injected intraperitoneum) twice a week for 4 weeks.

Functional determination of the differentiated hepatocyte-like cells by transplantation assay

After 20 days in vitro differentiation, 5 × 106 hepatocyte-like cells were harvested and transplanted into the liver injury ICR mice by caudal vein. After transplantation, same dose of CCl4- was injected twice a week to maintain persistent liver damage. The mice were killed 3 weeks after transplantation and serum was collected to analyse ALB level, total bilirubin (Tbil) concentration and alanine aminotransferase (ALT) by standard kit and automated instrumentation.

Evaluation of BMSSCs that directly transplanted into liver-injured mice

An EGFP-expressing BMSSCs line was established from C57BL/6 transgenic mice according to the method described above. These cells were used for in vivo tracking analysis of the fate after transplantation and evaluation of the therapeutic effect of the direct BMSSCs transplantation in liver-injured mice.

Wild-type C57BL/6 mice with liver injury were divided into four groups: (1) Model group; (2) PBS group; (3) BMSSCs group 1 (untreated BMSSCs) and (4) BMSSCs group 2 (BMSSCs pre-treated with VPA for 3 days). EGFP-expressing BMSSCs untreated or treated with VPA or control PBS was transplanted by caudal vein (5 × 106 cells/0.1 ml/mouse). The mice were killed 3 weeks after transplantation. In vivo bio-imaging was conducted under a nightowl in vivo imaging system, using WinLight software (Berthold Technologies, Bad Wildbad, Germany). The liver tissues were then fixed by perfusion with 4% paraformaldehyde. After additional fixation in 4% paraformaldehyde for 2 hrs at 4°C, tissues were either cryoprotected in 30% sucrose overnight and embedded in optimal-cutting-temperature (OCT, SAKURA Finetechnical Co., Japan) compound or embedded in paraffin. Sections were screened for the presence of EGFP to evaluate the integrated level of the transplanted BMSSCs into the injured liver, and their subsequent differentiation into hepatocytes by immunofluorescence staining of ALB. To evaluate the therapeutic effect of the direct BMSSCs transplantation in liver-injured mice, the serum collected at 3 weeks after transplantation was analysed. Picro-Sirius red staining was performed to analyse the extent of fibrosis.

Statistical analysis

Statistical analysis was performed using the SPSS version 16.0, and data were expressed as mean ± S.D. Differences between the values were determined by independent samples test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Characterization of culture-expanded BMSSCs

Fibroblast-like cells were cultivated from the bone marrow obtained from 6-week healthy ICR mice (Supporting Fig. 1A). To remove the possible contaminating haematopoietic cells, cells were selected based on plastic adherence and passaged at least three times prior to further use. Flow cytometry revealed that the propagated cell population exhibited CD44+CD73+CD90+ CD105+SCA-1+STRO-1+CD11bCD45 phenotype (Fig. 2). When plated on adipogenic media, cells containing vacuoles that produced lipid could be detected (Supporting Fig. 1B). Staining of alizarin red S and alkaline phosphatase showed that the cultures have the potential to differentiate into osteogenic cells (Supporting Fig. 1C and D). The results obtained above showed that a cell population devoid of haematopoietic cells and macrophages that is phenotypically and functionally equivalent to genuine BMSSCs was obtained.

image

Figure 2. Characterization of mice BMSSCs. Flow cytometry analysis of culture-expanded mice BMSSCs was performed for CD11b, CD44, CD45, CD73, CD90, CD105, SCA-1 and STRO-1. The results shown are representative for five independent experiments. Scale bars represent 50 μm.

Download figure to PowerPoint

VPA treatment significantly improved the in vitro hepatic differentiation of BMSSCs

Morphologically, hepatocyte-like epithelioid cells appeared in group A at day 7, although the cells were still surrounded by spindle-shaped cells, whereas, most cells still display fibroblastic profile in group B (Fig. 3A-i and ii), indicating that the hepatic differentiation in VPA-treated group occurred earlier than in group B. After 21 days, a more homogeneous cell population was developed in VPA-treated group, in which more than 90% cells exhibited a polygonal shape with a clear round nucleus (Fig. 3A-iv).

image

Figure 3. In vitro differentiation of BMSSCs into hepatocyte-like cells. (A) Morphological observation of BMSSCs-derived cells under a phase contrast microscope at day 7 (i, ii) and 21 (iii, iv) respectively. (B) Differentiated cells at day 12 were stained with specific antibodies recognizing AFP (i, ii), CK18 (iii, iv) and albumin (ALB) (v, vi). All sections were counterstained with DAPI (C) The results of RT-PCR analyses showed that BMSSCs cultured in differentiation medium could express a number of liver-specific genes in a time-dependent manner, while no expression of mRNA for such genes was observed in freshly isolated BMSSCs. Mouse β-actin mRNA was used as control for production of cDNA, and liver mRNA was used as positive control. (D) ALB secretion into the medium was measured by ELISA. **P < 0.01, ALB-secretion significantly differs among VPA treatment and VPA un-treatment groups. (E, F) After 20 days’ inducing in differentiation medium, the BMSSC-derived polygonal cells had magenta staining in the region indicating storage of glycogen. 7-Ethoxyresorufin-O-deethylase activity catalysed by CYP2B was detected with a fluorescence detector (i: Group B, ii: Group A). The results shown are representative for at least 8 separate experiments. Scale bars represent 50 μm.

Download figure to PowerPoint

Molecularly, the expressions of liver-specific markers were analysed at both mRNA and protein levels. Although the expression manners of AFP, HNF3-β, ALB and TAT in both groups exhibited time dependency during the differentiation, the expression levels of these genes differed considerably between the two groups. As shown in Fig. 3C, after 7 days, much higher AFP, HNF3-β and ALB mRNA expression levels were observed in group A. By day 14 of differentiation, down-regulation of AFP mRNA expression in VPA-treated group was more pronounced. In addition, the mRNAs of the late liver-specific marker ALB and TAT in group B are much lower compared with group A, suggesting a relatively immature hepatic differentiation status. In parallel, the immunofluorescence staining analyses were performed (Fig. 3B). In accordance with the results obtained at the mRNA level, VPA-treated cells expressed CK18 and ALB more abundantly than cells in control group.

In accordance with the results obtained at the morphological and molecular levels, VPA-exposed cells were more sensitive to PAS staining and expressed functionally active CYP2B with more abundant pattern (Fig. 3E and F). ALB concentrations, secreted into the culture media, were analysed by ELISA. VPA-treated cells significantly up-regulated the ALB secretion rate from day 14 onwards (Fig. 3D). On the contrary, a limited up-regulation of the ALB secretion was detected in group B. Moreover, the hepatocyte-like cells were further evaluated by transplantation analysis. The data from the therapeutic experiment were listed in Table 2. The transplantation of BMSSCs-derived hepatocyte-like cells from both groups reduced serum ALT and Tbil levels, and restored the serum total protein and ALB contents near to the levels in control group.

Table 2.  Therapeutic experiment of transplantation on abnormal liver function in ICR mice with CCl4-induced hepatic fibrosis (n= 6)
Index Control Model PBS BMSSCs-derived hepatocytes (5 × 10 6 ) BMSSCs-derived hepatocytes (5 × 10 6 ) (VPA pre-treated)
  1. ALT, alanine aminotransferase; ALB, albumin; TP, total protein and TBIL, total bilirubin.

  2. Values are means ± S.D. *P < 0.05, **P < 0.01 vs. Model group.

ALT (U/l)27.3 ± 5.3**167.4 ± 58.5189.7 ± 46.548.4 ± 17.1**54.4 ± 23.4**
ALB (g/ l)32.3 ± 1.0*27.9 ± 1.828.1 ± 2.631.1 ± 2.631.3 ± 2.2
TP (g/ l)68.1 ± 5.2*55.6 ± 3.958.5 ± 5.665.3 ± 4.4*64.5 ± 3.5*
TBIL (mg/ l)1.73 ± 0.33**7.93 ± 3.518.02 ± 1.935.82 ± 1.09*5.21 ± 1.16*

Cellular proliferation and cycle analysis of VPA-treated BMSSCs

To determine the effect of VPA on the proliferation of BMSSCs, various concentrations of VPA were employed. The result showed that treatment of VPA inhibited proliferation of BMSSCs in a dose-dependent manner (Fig. 4A). When the concentration of VPA reached 5 mM, most cells detach from the culture dishes. Moreover, cell cycle activity of BMSSCs exposed to 2.5 mM VPA was analysed. The result demonstrated that 2.5 mM VPA treatment did not lead to apoptosis obviously, as very little DNA fragmentation was observed in the apoptotic sub-G0/1 range (Fig. 4B). The proportion of cells in G0/G1-phase increased significantly (by 1.4–1.6-fold) after exposure to 2.5 mM VPA, relative to controls (Fig. 4B-iii). Accordingly, the ratio of S/G2/M, which reflects the proportion of cells completing DNA synthesis and cellular division, decreased (Fig. 4B-iv). This finding suggests that the cell cycle arrest occurs at G1[RIGHTWARDS ARROW]S transition, which facilitates cellular differentiation.

image

Figure 4. (A) Effect of VPA on the proliferation of BMSSCs. Proliferation of BMSSCs was determined by direct cell counting using a haemocytometer at the third day after the addition of VPA. Data for cell proliferation were expressed as a percentage of the initial cell number (n= 6). *P < 0.05 compared with the data in the absence of VPA. (B) Cell-cycle analysis of BMSSCs treated with 2.5 mM VPA for 72 hrs. DNA synthesis and cell-cycle progression of BMSSCs was altered in the presence of 2.5 mM VPA (B-ii), as indicated by the increase in area under the G0/G1 peaks, relative to controls (B-i). (C) Electron microscopy analysis of VPA-induced changes in chromatin structure. Untreated cells displayed dense nucleoli and condensed pattern of heterochromatin (C-i). In contrast, cells treated with 2.5 mM VPA for 72 hrs showed an even dispersion of the chromatin (C-ii). (D) DNase I digestion and micrococcal nuclease digestion of BMSSCs DNA after treatment with 2.5 mM VPA for 72 hrs. (i) DNase I digestion and (ii) micrococcal nuclease digestion. (E) Acetylation of histones H3 and H4 determined by Western blot analysis of BMSSCs treated with 2.5 mM VPA for 72 hrs. (F) Immunofluorescent staining of BMSSCs using histones H3 (i, ii) and H4 (iii, iv) antibodies. VPA clearly increased the amount of acetylated H3 and H4 (ii, iv).

Download figure to PowerPoint

The effects of VPA on histone acetylation and chromatin structure

Analyses for histone acetylation and chromatin structure were performed to ascertain whether the VPA-mediated facilitation of differentiation was accompanied by the alteration of histone acetylation and chromatin decondensation. By Western blot, immunocytochemical and electron microscopic analysis (Fig. 4C, E and F), it was demonstrated that VPA treatment dramatically increased the amount of acetylated H3 and H4, and altered the chromatin structure. As shown in Fig. 4C-i, in untreated BMSSCs, a dense nucleoli and condensed pattern of heterochromatin evenly dispersed in the nuclei. However, in cells treated with VPA (Fig. 4C-ii), a prominent change in the distribution of heterochromatin occurred, causing the chromatin to disperse evenly. For a more quantitative assessment of the chromatin decondensation, we evaluated the susceptibility of DNA to nucleases. As shown in Fig. 4D-i, the intensity increased significantly after VPA treatment, which suggesting that the genomic region of VPA-treated BMSSCs has more labile nucleotide sequence sensitive to mechanical stress like DNase I. When BMSSCs were exposed to micrococcal nuclease, the resulting DNA fragments in VPA-treated samples displayed the characteristic nucleosomal ladder (Fig. 4D-ii) suggestive of chromatin decondensation.

Evaluation of VPA-mediated epigenetic gene expression patterns during hepatic differentiation

Since it was reported that the FGFRs/FGF and c-Met/HGF signalling pathways are involved in different stages of liver development, and epigenetic modification of cell development was mediated by regulation of specific gene expression patterns, we asked whether the VPA-mediated facilitation of differentiation was regulated by the expression patterns of FGF and HGF receptors. BMSSCs of passage 3 were treated with 2.5 mM VPA (72 hrs) or 10 ng/ml FGF-4 (72 hrs) or first 2.5 mM VPA (72 hrs) and second FGF-4 (72 hrs), respectively. Four FGF receptor members, FGFR-1IIIc, FGFR-2IIIb, FGFR-2IIIc and FGFR-4, which bind and respond to FGF-4 [21–24], were selectively chosen for this study. Using specific primers shown in Table 1, we detected that FGFR-1IIIc and FGFR-2IIIc was strongly expressed in BMSSCs, FGFR-2IIIb and FGFR-4 were minimally expressed (Fig. 5A). Following the VPA, FGF-4 or VPA + FGF-4 treatment, the mRNA levels of FGFR-1IIIc, FGFR-2IIIc and c-Met in the BMSSCs were examined by real-time RT-PCR analysis. As shown in Fig. 5B–D, both VPA and FGF-4 treatment slightly increased the expression of FGFR-1IIIc and c-Met mRNA, and VPA + FGF-4 treatment considerably increased the level of FGFR-1IIIc by twofold and c-Met by over fivefold in comparison to untreated control. In contrast, the FGF-4 and VPA + FGF-4 treatments reduced the level of FGFR-2IIIc mRNA. These results suggested that the VPA treatment could alter the gene expression patterns of FGFR-1IIIc, FGFR-2IIIc and c-Met in BMSSCs, which resulted in a more sensitive mode of BMSSCs to FGF-4 and HGF; and the activation of FGFRs/FGF-4 and c-Met/HGF signalling pathways may responsible for the VPA-mediated facilitation of hepatic differentiation.

image

Figure 5. Expression analyses of FGF receptors and HGF receptor (c-Met) in BMSSCs. (A) Expression of FGFRs and c-Met in BMSSCs was detected by RT-PCR. Total RNA extracted from BMSSCs was subjected to RT-PCR using specific primer pairs. RNA quality was evaluated in all samples by parallel RT-PCR for β-actin. Absence of contaminating genomic DNA was ensured by RNA-PCR. (B, C and D) Following the VPA, FGF-4 or VPA + FGF-4 treatment, the mRNA levels of FGFR-1III c, FGFR-2III c and c-Met in the BMSSCs were examined by quantitative real-time RT-PCR analysis. β-actin was used as an internal control. Results show fold variations of treated cells compared to untreated controls. Error bars represent standard deviations of five independent experiments. Control: normal cultured BMSSCs; VPA: BMSSCs were treated with 2.5 mM VPA for 72 hrs; FGF-4: BMSSCs were treated with 10 ng/ml FGF-4 for 72 hrs; VPA + FGF-4: following 72 hrs 2.5 mM VPA treatment, cells were treated with 10 ng/ml FGF-4 for another 72 hrs. *P < 0.05, **P < 0.01, n= 5.

Download figure to PowerPoint

VPA treatment significantly improved the in vivo hepatic differentiation of BMSSCs

The in vivo hepatic differentiation assay was performed in parallel with the studies conducted in vitro to determine whether predictions regarding a promoting effect can be made based on an understanding of the molecular mechanisms of VPA-mediated epigenetic modification. Three weeks after transplantation of EGFP-expressing BMSSCs, mice were killed and the whole livers were imaged under a bio-imaging system and sectioned for morphological and pathological analysis. The EGFP-expressed cells with hepatocyte morphology were detected in all recipient animal livers. As shown in Fig. 6A-i, Bi, there was a much higher frequency event in the VPA pre-treated transplantation group, in which BMSSCs used for transplantation were pre-treated with 2.5 mM VPA for 3 days. These data strongly support the view that VPA pre-treatment is potentially an efficient method for facilitate homing of BMSSCs to the site of liver injury. We further examined the fate of transplanted BMSSCs in the recipients’ livers. Strikingly, all the EGFP-positive cells were found near or not far from the blood vessels (Fig. 6A-ii and B-ii). These EGFP-positive cells had not only acquired proper hepatocyte morphology but also stained positive for hepatocyte-specific antigen such as ALB (Fig. 6A-iii and B-iii).

image

Figure 6. Distribution of EGFP-expressing mouse BMSSCs transplanted into CCl4-treated mice. (A) EGFP-expressing BMSSCs that were pre-treated with 2.5 mM VPA for 3 days were injected by caudal vain into homogeneous liver-injured mice. In vivo imaging analysis of EGFP-expressing cells in fibrotic liver was carried out by charge-coupled device camera 3 weeks after the transplantation. (i) Image shown is charge-coupled device camera image of recipient’s liver 21 days after transplantation. A pseudo-colour luminescent image from blue (least intense) to red (most intense), representing the special distribution of the detected photons emitted from EGFP-expressing cell within the liver. (ii–v) Individual and merged images of liver frozen sections 21 days after transplantation are shown. (B) A control mouse liver transplanted with EGFP-expressing BMSSCs that without VPA treatment. The results shown are representative for at least 10 separate experiments. Scale bars represent 50 μm. P: portal vein.

Download figure to PowerPoint

The EGFP signal-positive tissue sections were prepared for fibrosis analysis by Sirius red staining as previously described [25]. The improvement of BMSSCs transplantation on fibrosis was evaluated by digitalization of the stained area. After 7 weeks of CCl4 treatment, a considerable fibrotic area, stained red, was seen in sections (Fig. 7A). The level of liver fibrosis was reduced to some degree in the BMSSCs group, but there was no significant difference when compared with the control group. In contrast, a significant result was observed in liver receiving VPA-treated BMSSCs (P < 0.05, Fig. 7B-i). As shown in Fig. 7B-ii–iv, transplantation of VPA-treated BMSSCs to some degree restored the ALB production, and significantly suppressed the serum Tbil and ALT levels in the CCl4-injured mice, suggesting that priming the BMSSCs culture with VPA before the transplantation effectively suppressed liver inflammation.

image

Figure 7. Transplantation of the VPA pre-treated BMSSCs causes a significant therapeutic effect in chemically induced liver injury. (A) Liver specimens obtained from animals 14 days after the transplantation were subjected to sirius red staining for fibrosis. (A-i), model group; no transplantation was performed. (A-ii), PBS group; liver-injured mice were transplanted with PBS. (A-iii), liver-injured mice were transplanted with BMSSCs. (A-iv), BMSSCs transplanted into the liver-injured mice were pre-treated with VPA. (Scale bars represent 50 μm) (B) Therapeutic experiment of transplantation on abnormal liver function and score of liver fibrosis in mice with CCl4-induced hepatic fibrosis (n= 6). The degree of liver fibrosis was semi-quantified by measuring the relative areas of fibrosis using computer software (B-i). B-ii, -iii, -iv, recovery of serum albumin, total bilirubin concentration and suppression of liver inflammation in CCl4-injured mice. Blood samples were collected 4 weeks after the transplantation from the experimental and control mice. (*P < 0.05, **P < 0.01).

Download figure to PowerPoint

These results indicate that the pre-treatment of VPA are efficient for the homing of BMSSCs to the site of liver injury and, additionally, for supporting hepatic differentiation in vivo as well as in vitro.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

VPA, together with trichostatin A and sodium butyrate, were direct HDAC inhibitors that participate in the epigenetic modification, have recently emerged as a new class of chemotherapeutic drug that induce cytotoxic or cytostatic effects on tumour cells with remarkable specificity [25]. HDAC inhibitors influence cellular histone acetylation, relax the DNA wrapped around the core histones, and allow transcription of genes involved in important cellular processes. Therefore, HDAC inhibitors have been shown to arrest cell growth by inhibiting DNA synthesis, and induce cellular differentiation and apoptosis by arresting proliferating cells in the G1 phase of the cell cycle in various transformed cells both in vitro and in vivo[26]. Because of the biological similarities between cancer cells and stem cells, much attention has also been paid to the effects of HDAC inhibitors on differentiation of stem cells, and evoking interests to study the relationship between epigenetic modification of chromatin and stem cell differentiation. Indeed, HDAC inhibitors have been employed in the study of stem cells, such as cardiomyocyte differentiation from embryonic stem cells [27], neuronal differentiation, and osteogenic differentiation from adult stem cells [16, 17], and even the expansion of human haematopoietic stem cells [28]. In our present study, we employed a well-established adult stem cell model, BMSSCs, to investigate whether VPA was acting on hepatic cell fate specification from adult stem cells. By in vitro and in vivo experiments, it was clearly demonstrated that VPA significantly increased the hepatic differentiation accompanied with considerable epigenetic modifications, including the increased amount of acetylated histones H3 and H4, the susceptibility of DNA to nucleases, the structural de-condensation of chromatin, the reduced cellular proliferation and cell cycle arrest at G1 to S transition which facilitate the cellular differentiation, as well as the homing of BMSSCs to the site of liver injury and largely improved the liver recovery.

The chromatin modification plays an important role in stem cell decisions based on their ability to regulate the expression of specific genes involved in proliferation, differentiation and apoptosis [17]. The profound effect of VPA shown in our study suggested that the acetylated state have dominant effect on hepatic lineage direction of BMSSCs. Up to now, it has been clear that FGF, HGF and their receptor family members, are activators of hepatic differentiation. The FGF/FGFRs and HGF/c-Met signalling pathways could be involved in different stages of liver development. Thus, the FGF and HGF receptor family members were selectively studied to evaluate whether the VPA-mediated facilitation of hepatic differentiation was regulated by the selective expression of FGF and HGF receptors and their subsequent downstream signalling pathways. The results clearly showed that the FGFR-1IIIc was dramatically up-regulated in response to exogenous treatment of VPA combined with FGF-4, and the FGFR-2IIIc expression was significantly down-regulated. Changes in these two gene-expression patterns were consistent with the previous reports as a normal response to activation of the FGF signal transduction pathway [23]. Interestingly, individual treatment of VPA up-regulated both FGFR-1IIIc and FGFR-2IIIc expressions to a certain degree. Similar results were also acquired in c-Met analysis. However, questions that how VPA induced the expression of c-Met and FGFR-1IIIc through epigenetic modification and whether the induction is specific remain interesting to elucidate. From our results that VPA treatment dramatically increased the amounts of acetylated H3 and H4 as well as the structural decondensation of chromatin, it may be reasonable to hypothesize that VPA relaxes the DNA wrapped around the core histones by H3 and H4 acetylation, allowing genes more facilitated to accept transcriptional factors, leading to an increased transcription of a considerable genes, including c-Met and FGFR-1IIIc, that involved in important cellular processes. The epigenetic modification mediated by VPA is probably universal, since studies have demonstrated that after VPA treatment, hundred of genes belonging to multiple pathways including ribosomal proteins, oxidative phosphorylation, MAPK signalling, focal adhesion, cell cycle, antigen processing and presentation, proteasome, apoptosis, PI3K, Wnt signalling, calcium signalling, TGF-β signalling, and ubiquitin-mediated proteolysis are up-regulated [29]. The dramatic up-regulated gene expression of FGFR-1IIIc and c-Met induced by VPA may largely contribute to trigger and amplify the signalling of FGF-4 and HGF, which may subsequently benefit to the initiation and maturation of hepatic specification. Future studies are needed to determine the precise down stream pathways involved in this process.

In conclusion, this is the first report showing VPA has considerable effect on the hepatic differentiation from BMSSCs. Regulation of FGFRs and c-Met gene expression through post-translational modification of core histones might be the primary initiating event for these effects. This model may be applicable to study endoderm differentiation and offers an unlimited source of functional hepatocyte-like cells applicable for pharmaco-toxicological research and testing. Also this system has numerous potential advantages in clinical application: first, VPA is a well-established drug that is usually well tolerated; second, there is not a problematic limitation of donors since BMSSCs are readily accessible from bone marrow and can be expanded tremendously in vitro and third, the use of BMSSCs is favorable over embryonic stem cells regarding ethical issues. We are encouraged that administration of VPA has the potential of being an excellent choice for liver engineering strategies and cell based therapies using BMSSCs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was supported by the grant of medicine and health key project of Zhejiang Province, Science and Technology Foundation of Ministry of Health of the People’s Republic of China (WKJ2007-2-037), and Shaoxing key project for Science and technology (2007A23008, 2005141).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References