SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Functional pluripotent characteristics have been observed in specific subpopulations of hepatic cells that express some of the known cholangiocyte markers. Although evidence indicates that specific cytokines, granulocyte macrophage colony-stimulating factors (GM-CSFs), and stem cell factors (SCFs) may be candidate treatments for liver injury, the role of these cytokines in intrahepatic biliary epithelium remodeling is unknown. Thus, our aim was to characterize the specific cytokines that regulate the remodeling potentials of cholangiocytes after 70% partial hepatectomy (PH). The expression of the cytokines and their downstream signaling molecules was studied in rats after 70% PH by immunoblotting and in small and large murine cholangiocyte cultures (SMCCs and LMCCs) by immunocytochemistry and real-time polymerase chain reaction (PCR). There was a significant, stable increase in SCF and GM-CSF levels until 7 days after PH. Real-time PCR analysis revealed significant increases of key remodeling molecules, such as S100 calcium-binding protein A4 (S100A4) and miR-181b, after SCF plus GM-CSF administration in SMCCs. SMCCs produced significant amounts of soluble and bound SCFs and GM-CSFs in response to transforming growth factor-beta (TGF-β). When SMCCs were incubated with TGF-β plus anti-SCF+GM-CSF antibodies, there was a significant decrease in S100A4 expression. Furthermore, treatment of SMCCs with SCF+GM-CSF significantly increased matrix metalloproteinases (MMP-2 and MMP-9) messenger RNA as well as miR-181b expression, along with a reduction of metalloproteinase inhibitor 3. Levels of MMP-2, MMP-9, and miR-181b were also up-regulated in rat liver and isolated cholangiocytes after PH. Conclusion: Our data suggest that altered expression of SCF+GM-CSF after PH can contribute to biliary remodeling (e.g., post-transplantation) by functional deregulation of the activity of key signaling intermediates involved in cell expansion and multipotent differentiation. (HEPATOLOGY 2012;;55:209–221)

In addition to playing important roles in the regulation of ductal secretion, cholangiocytes are the target cells of apoptotic, proliferative, and regenerative events leading to changes in biliary damage (e.g., after carbon tetrachloride [CCl4] acute administration), hyperplasia (e.g., after bile duct ligation), and regeneration (e.g., after 70% hepatic resection).1, 2 During liver development, both hepatocytes and cholangiocytes arise from common bipotential progenitors called hepatoblasts3 or hepatic stem cells (HSCs).4 The HSCs or hepatoblasts in the liver parenchyma differentiate into hepatocytes, whereas those adjacent to the portal mesenchyme differentiate into cholangiocytes.4, 5 During liver regeneration, the process of hepatic wound healing, a complex cascade of inflammatory signaling, recruits inflammatory cells, stimulates hepatobiliary cell proliferation, directs cell migration, and induces vascularization to restore tissue integrity.6 Previous studies related to biliary wound healing focused mainly on the inducers of compensatory biliary proliferation after bile duct insult.7

Cytokines that are candidates in liver-remodeling processes include stem cell factor (SCF), erythropoietin (EPO), granulocyte colony-stimulating factor (G-CSF), and granulocyte macrophage colony-stimulating factor (GM-CSF); these regulate bone marrow production of circulating red cells, white cells, and platelets. These cytokines act on stem cells, leading to lineage specific differentiation.8 SCF regulates the differentiation of CD34-positive stem cells, whereas other factors, such as EPO, G-CSF, and GM-CSF, modulate the synthesis of more specific cell types.9 Colony-stimulating factors (CSFs) are involved in hepatic inflammation (via direct effects on the vascular endothelium, and/or neutrophil recruitment and activation) as well as in hepatic repair and regeneration.10

The potential mechanisms involved in direct actions and/or induction of additional factors that promote liver regeneration and repair are unclear. Multiple studies have suggested that there is an intricate regulatory system involved in hepatic regeneration after injury. Although many cytokines have proliferative effects on hepatobiliary cells both in vitro and in vivo, no single molecule has proven to be a key factor responsible for controlling hepatocyte or proliferation of subpopulations sharing cholangiocyte markers and in vivo remodeling during liver injury. Although evidence suggests that these factors may be candidate treatments for liver injury, either as potential hepatoprotectants or as hepatoreparative agents, the role of one or more of these cytokines in hepatobiliary remodeling after partial hepatectomy (PH) is undefined.

Hepatocyte proliferation may be blocked if the tissue injury is too severe.6 During this process, cholangiocytes of the portal ductules and canals of Hering (small tubules lined by epithelium with biliary morphology, which connect the hepatocyte bile canalicular network to the portal biliary ductules) begin expressing hepatocyte-associated transcription factors.11 It has been suggested that cholangiocytes can acquire stem cell phenotypes and, subsequently, become hepatocytes, restoring liver regeneration when hepatocytes cannot proliferate,12 but an alternative interpretation is that liver regeneration is derived from HSCs.4, 13, 14 Furthermore, HSCs are located within the canals of Hering and have markers shared with biliary epithelia,4, 13-15 and they expand in disease conditions14 before the formation of oval cells, progenitor populations occurring in livers of hosts exposed to oncogenic insults.16 Therefore, HSCs are one of the most important regenerative alternatives during conditions where hepatocytes fail to proliferate. The current study elucidated the possible role of cytokine-mediated remodeling during liver regeneration, especially their synergistic effects on the proliferation of cholangiocytes and their mesenchymal partners, stellate cells,17 from human, rat, and mouse bile ducts.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Cell Lines and Cultures.

Our small (SMCC) and large (LMCC) murine cholangiocytes were isolated from healthy mice (BALB/c) and immortalized by the introduction of the SV40 large-T antigen gene.18 Healthy human intrahepatic biliary epithelial cells (HiBECs), human hepatocytes, and mediums were obtained from ScienCell Research Laboratories (San Diego, CA). All other cell-culture media and supplements were obtained from Invitrogen (Carlsbad, CA).

Animal Protocols and 70% Hepatectomy Model.

Male Fisher 344 rats (75-100 g) were purchased from Charles River (Wilmington, MA). The 70% partial hepatectomy (PH) surgery was performed according to the classical model of Higgins and Anderson. Tissues were collected, and intrahepatic cholangiocytes were isolated from the removed liver tissues, as previously described.1, 2, 19, 20

Purified Cholangiocytes and Normal Rat Intrahepatic Cholangiocyte Cultures.

Virtually pure cholangiocytes were isolated by immunoaffinity separation1, 2, 19, 20 with a monoclonal antibody (a gift of Dr. R. Faris) against an unidentified antigen expressed by all intrahepatic rat cholangiocytes.21 The in vitro experiments were performed in freshly isolated rat cholangiocytes (IRCs) and our polarized normal rat intrahepatic cholangiocyte cultures (NRICs).22

In Vitro Proliferation and Migration Assay.

Commercially available kits were used for the proliferation and migration assays in hepatobiliary cells (details in Supporting Information).

Western Blotting.

Protein was extracted from cultured cells or homogenized tissues, and western blotting was performed as previously described.23 Please see the Supporting Information for more details.

Real-Time Polymerase Chain Reaction Assays for Mature microRNAs.

The microRNA (miRNA) was isolated as previously described,23 and the expression of specific mature miR-181b was verified by real-time polymerase chain reaction (PCR) analysis, using a TaqMan Human MicroRNA Assay kit (Applied Biosystems, Foster City, CA).

Enzyme-Linked Immunosorbent Assay.

SCF (SCF enzyme-linked immunosorbent assay [ELISA] kit; BioSource International, Camarillo CA, USA) and GM-CSF (GM-CSF ELISA kit; BioSource) ELISAs were performed according to the manufacturer's instructions.

Statistical Analysis.

All data are expressed as mean ± standard error (SE). The differences between groups were analyzed by the Student t test when two groups were analyzed or analysis of variance if more than two groups were analyzed. P < 0.05 was used to indicate statistically significant differences.

Please see Supporting Information for more detailed information of this section.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Altered Expression of CSF and Their Receptors After PH.

Animals underwent 70% PH or sham and were sacrificed at 0, 1, 3, 5, and 7 days after surgery. Liver samples were obtained and analyzed for SCF, GM-CSF, G-CSF, and EPO messenger RNA (mRNA) expression by real-time PCR. Sham animals demonstrated high mRNA levels of SCF, but not GM-CSF, G-CSF, and EPO, in the liver, suggesting a large baseline hepatic reservoir of SCF (Fig. 1A). There was a significant decline in SCF expression during the first 24 hours after PH, followed by a significant, stable increase in SCF levels up to 7 days after PH (Fig. 1B). Concurrent with this increase in hepatic SCF mRNA levels, there was a significant increase in SCF receptor mRNA levels (Fig. 1C). Meanwhile, a moderate increase of GM-CSF, but not receptor mRNA, levels was observed after PH. Interestingly, the significant increases of SCF and GM-CSF were also observed in IRCs from total liver after PH (Fig. 1D). No changes in the expression of other CSFs and their receptors, such as G-CSF and EPO, were observed after PH.

thumbnail image

Figure 1. The expressions of CSFs and their receptors in rat liver after 70% PH. Animals underwent 70% PH or sham control, and liver tissues as well as IRCs were obtained at various times postoperatively. RNA was isolated, and complementary DNA was generated by reverse transcription using MMLV reverse transcriptase. Basal mRNA expressions of SCF, EPO, G-CSF, and GM-CSF (A), as well as the alterations of CSFs (B) and their receptors (C) in total liver tissues, and the expressions of CSF in IRCs (D) after PH, were quantified by real-time PCR using SYBR Green as the fluorophore, and expressed relative to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA expression assessed concurrently in the same samples. (A) *P < 0.05, when compared with the EPO group. (B andC) *P < 0.05, when compared with the sham control group.

Download figure to PowerPoint

CSFs Promote the Proliferation of Primary Hepatocytes but Not Cholangiocytes.

We evaluated in vitro the effects of CSF on the proliferation of human and mouse hepatocytes and cholangiocytes. Cells were exposed to media alone or 0.1, 1, 10, 20, and 50 ng/mL of CSFs, and proliferation was measured at 72 hours by (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) (MTS) assays.24 Exposure to 10 and 20 ng/mL of SCF and GM-CSF resulted in a significant increase in the proliferation of cells of the hepatocytic lineage, but not in subpopulations sharing cholangiocyte markers (i.e., HSCs and biliary epithelia), compared with incubation with media alone (Supporting Fig. 1), suggesting different proliferative roles between cells with hepatocytic markers and those with cholangiocyte traits during regeneration after PH.

Synergistic Role of Cytokines in Proliferation of Cells With Biliary Markers.

The synergistic effects of SCF in combination with other cytokines have been demonstrated by many studies during the regeneration of different organ systems.25, 26 To determine whether the association between liver regeneration and the biliary epithelium is cytokine dependent, we examined the proliferation of cell subpopulations with cholangiocyte markers (e.g., HSCs and biliary epithelia) before or after stimulation with SCF+GM-CSF, G-CSF, and EPO. We observed an enhancement of cell proliferation in small murine epithelial subpopulations with biliary markers after stimulation with SCF+GM-CSF for 72 hours (Fig. 2). Addition of this cytokine combination did not change the proliferation index in other human and mouse cholangiocyte subpopulations. These results indicate a regenerative potential in either HSCs and/or committed biliary epithelial progenitors (i.e., small cholangiocytes) after stimulation with SCF+GM-CSF.

thumbnail image

Figure 2. SCF promotes synergistic cellular proliferation in combination with GM-CSF in small murine cells with biliary markers. Epithelial cells in 96-well plates with 10% fetal bovine serum containing culture medium and were treated with different cytokines for various concentrations; as indicated below the panels, in combination with SCF (10 ng/mL). After 72 hours, cell proliferation was assessed using the MTS assay, and the proliferation index was derived. Mean ± SE from four independent experiments, each in triplicate, at each point is illustrated. GM-CSF in combination with SCF significantly increased cellular proliferation at higher concentrations (≥10 ng/mL) in normal human hepatocytes and SMCCs. *P < 0.05, when compared with control SCF-only group.

Download figure to PowerPoint

SCF+GM-CSF Induce Subpopulations of Biliary Epithelia to Become Competent to Mitogen Signal.

Isolated cholangiocyte subpopulations include cells that are small (7-9 μm) or large (14-15 μm).2, 20 The small ones comprise the HSCs and the committed biliary progenitors, whereas the large ones are cholangiocytes.4 These usually acquire long survival potency, maintain a certain level of differentiation, and do not respond to single-cytokine stimulation. To test the capacity of cholangiocyte subpopulations to proliferate, isolated small (i.e., HSCs and committed biliary epithelia) and large cholangiocytes were exposed for 3 days to different combinations of SCF, EPO, G-CSF, and GM-CSF in a medium without fetal calf serum, and DNA synthesis was estimated in these subpopulations (Fig. 3A). In the absence of stimulation and in single CSF-treated cultures, low levels of DNA synthesis were detected in both small and large biliary subpopulations. In contrast, a combination of SCF+GM-CSF induced DNA synthesis in at least 30% of small cells with biliary markers at day 3, whereas less than 12% of large cholangiocytes incorporated bromodeoxyuridine (BrdU) when stimulated by SCF in the presence of GM-CSF. Measurement of the mitotic index (MI) in cultures stimulated with SCF+GM-CSF for 5 days showed a peak of division at day 4, reaching 25% of small cells with biliary markers (Fig. 3B). The MI matched the BrdU incorporation level.

thumbnail image

Figure 3. SCF+GM-CSF promote cell-cycle progression in SMCC. (A) SMCCs and LMCCs were stimulated with SCF plus other CSFs (all at 10 ng/mL) for 72 hours, and percentages of BrdU-labeled cells were determined by immunocytochemistry after 24-hour BrdU incorporation at day 3. Results represent means ± SE (n = 8). (B) SMCCs were stimulated with the combination of SCF/GM-CSF (10 ng/mL) for 5 days. MI was determined every day after a 24-hour colcemid treatment. Percentages of dividing cells were determined. Values represent means ± SE (n = 6). (C) SMCCs were stimulated with (1) combination of SCF+GM-CSF for 48 hours, (2) SCF then GM-CSF for 24 hours each, (3) GM-CSF then SCF for 24 hours each, and (4) SCF for 48 hours, followed by 2 days without stimulation and then GM-CSF for 48 hours. BrdU was incorporated during the last 24 hours of treatment, and percentages of BrdU-labeled cells were determined. Results are means ± SE (n = 6). (D) Alterations of cell-cycle proteins in SCF- and/or GM-CSF-treated SMCCs. SMCCs were maintained without stimulation or treated with SCF, GM-CSF alone, or SCF+GM-CSF (10 ng/mL) for 3 days. Protein extracts from fractions enriched for cells with biliary markers were analyzed by immunoblotting using antibodies against cyclin D1, Cdk1, Cdk2, and β-actin.

Download figure to PowerPoint

We next tested whether the addition of GM-CSF before or after SCF would influence the replication of small cells with biliary markers. Cultures were costimulated with GM-CSF and SCF for 48 hours or stimulated for 24-hour periods with GM-CSF before or after SCF treatment (Fig. 3C). In all cases, cells replicated their DNA, and BrdU incorporation levels were similar. In addition, a pause of 2 days after SCF treatment did not change the responsiveness of cells to GM-CSF.

SCF Promotes Cell-Cycle Progression to Late G1, Whereas GM-CSF Is Necessary for S Entry.

To further analyze the respective role of the combination of the cytokines in cell-cycle progression, we studied the expression levels of cyclin D1, Cdk2, and Cdk1 in small and large cholangiocyte subpopulations. Short-term cholangiocyte cultures, known to undergo one cell cycle under SCF stimulation, were used as controls.

In untreated and GM-CSF-stimulated cultures, as well as in untreated short-term cultures, cyclin D1 and Cdk1 were not detected (Fig. 3D), suggesting that cells were blocked in G1 upstream from the mitogen-associated restriction point in mid-late G1. In contrast, in GM-CSF+SCF-stimulated cultures and in SCF-treated short-term cultures, cyclin D1, Cdk2, and Cdk1 expression levels markedly increased, thus confirming the cell progression into the S and M phases. These results suggested that SCF alone promoted progression up to late G1, and that GM-CSF stimulation was necessary for the G1/S transition.

Contribution to Regenerative Potentials by Cytokines in Small Cells With Biliary Markers.

Our previous studies have demonstrated that small cholangiocytes (originated from small bile duct branches and resistant to CCl4-induced apoptosis)2, 20de novo proliferate to compensate for the loss of large biliary mass.2, 4 It is possible that the combination of SCF and GM-CSF may alter the induction of remodeling potentials in cholangiocytes during liver regeneration. Alkaline phosphatase (AP) activity has been used as a biomarker for tissue repair and remodeling. We next compared StemTAG AP activity (Cell Biolabs, Inc., San Diego, CA) in mouse cholangiocytes after stimulation with SCF, GM-CSF, or a combination of the two. AP activity was scarcely detected 72 hours after stimulation with SCF or GM-CSF alone in small and large murine cells with biliary markers (Supporting Fig. 2A). In contrast, a more intense AP staining was observed when small cells with biliary markers were stimulated with the SCF+GM-CSF combination. mRNA levels of the SCF receptor were equally decreased after stimulation with either SCF alone or SCF combined with GM-CSF. In addition, GM-CSF receptor mRNA levels were similar between stimulation with SCF alone and SCF+GM-CSF (Supporting Fig. 2B). These results indicate that enhanced remodeling potential of small cells with biliary markers after stimulation with SCF+GM-CSF is the result of the activation of downstream signaling pathways, rather than increased levels of CSF receptors. Together, the data suggest that there are altered tissue-repair capabilities with SCF+GM-CSF in small mouse cells with biliary markers, but not in the presence of either SCF or GM-CSF alone.

Characterization of Tissue Remodeling Marker S100A4 in Cytokine-Treated Cholangiocytes.

S100 calcium-binding protein A4 (S100A4), a member of the S100 family of proteins, is of particular interest as a marker of chronic tissue remodeling. It plays an important role in matrix remodeling by up-regulating the expression of matrix metalloproteinases (MMPs). We assessed the effect of cytokine combinations on S100A4 mRNA expression. Quantitative real-time PCR analysis revealed a 3.8 ± 0.2-fold increased expression of S100A4 after SCF+GM-CSF administration in small cells with biliary markers, compared to SCF alone (P < 0.01), suggesting a significant remodeling event after treatment (Fig. 4A, B). Interestingly, the same treatment combination also induced a moderate increase of S100A4 protein expression in human intrahepatic cholangiocytes (Fig. 4C, bottom panel). In contrast, no significant changes were observed in S100A4 expression in small and large cells with biliary markers by SCF or GM-CSF alone or other SCF combinations.

thumbnail image

Figure 4. SCF regulates mesenchymal marker S100A4 in combination with GM-CSF in small murine cholangiocytes. Total RNA was isolated from SMCCs (A) and LMCCs (B) treated with CSF+SCF, and quantitative real-time PCR for mesenchymal marker S100A4 was performed using a superarray quantitative PCR assay kit. Quantitative data representing the mean and SE from three experiments performed in triplicate are presented in the bar graph. The expression of S100A4 was normalized to that of the GAPDH gene control. SCF promotes synergistic mesenchymal transition or transformation in combination with GM-CSF only in small murine cells with biliary markers. Differences between any of the other groups were not significant. (C) Immunocytochemistry for S100A4 and CK-19 was performed in SMCCs and HiBECs. An increase in S100A4 expression is observed in both cell lines after SCF+GM-CSF treatment (10 ng/mL, 72 hours). **P < 0.01, compared with expression in the SCF-only group.

Download figure to PowerPoint

Involvement of SCF and GM-CSF in Transforming Growth Factor Beta–Dependent Signaling in Cholangiocytes.

Transforming growth factor-beta (TGF-β) has been associated with intracellular matrix deposition and hepatic/biliary tissue repair and damage.27 It is a cytokine that stimulates mesenchymal proliferation, inhibits epithelial growth, and is important in organogenesis. Because previous studies have suggested that TGF-β is produced in response to PH,28 we next determined whether TGF-β would stimulate SCF and GM-CSF production and release. SMCC and LMCC in vitro were stimulated with 10 ng/mL of TGF-β or media alone and were harvested after 72 hours of incubation. Supernatants or supernatants plus cells were collected for SCF and GM-CSF measurement by ELISA assay. Supernatant levels of SCF/GM-CSF were used to estimate levels of soluble SCF/GM-CSF; for quantitation of soluble plus bound SCF/GM-CSF, supernatants plus cells were sonicated and SCF/GM-CSF levels in this solution were used as an estimate of soluble plus bound SCF/GM-CSF. Small mouse cholangiocytes produced significant amounts of both soluble and soluble plus bound SCF in response to TGF-β, when compared to controls, whereas large cholangiocytes only showed a slight increase at the time point studied (Fig. 5A-D).

thumbnail image

Figure 5. TGF-β stimulates SCF and GM-CSF production and release in SMCCs. SMCC and LMCC mouse cells were stimulated with 10 ng/mL of TGF-ß or media alone. Supernatants and supernatants plus cells were harvested at 72 hours, and SCF and GM-CSF levels were measured by ELISA. Supernatant SCF and GM-CSF levels were used to determine levels of soluble SCF and GM-CSF (A and C). For quantitation of soluble plus bound SCF and GM-CSF (B and D), supernatants plus cells were sonicated and SCF and GM-CSF levels in this solution were measured as an estimate of soluble plus bound SCF. There were no significant differences noted in levels of soluble, compared with soluble plus bound, SCF in cells incubated in media alone. There were significant increases in soluble SCF and GM-CSF levels at the 72 hours time point from cells treated with TGF-ß in SMCCs (*P < 0.05 versus media alone). In addition, levels of soluble plus bound SCF and GM-CSF were significantly increased in SMCCs after TGF-ß treatment, compared with treatment with media alone (*P < 0.05 versus media alone). Furthermore, levels of soluble plus bound SCF and GM-CSF were significantly increased, compared with levels of soluble SCF alone, in SMCCs after TGF-ß treatment at 72 hours. Data are expressed as the mean ± SE. (E) Anti-SCF+GM-CSF partially blocked TGF-β-induced changes in small cells with biliary markers. SMCCs and LMCCs were incubated with 10 ng/mL of TGF-β with anti-SCF+GM-CSF antibodies (10 μg/L) or control antiserum for 72 hours, total RNA was isolated, and quantitative real-time PCR for mesenchymal marker S100A4 was performed using a superarray quantitative PCR assay kit. Quantitative data represent the mean and SE from three experiments. The expression of S100A4 was normalized to that of the GAPDH gene control. A significant decrease in S100A4 mRNA expression was noted in SMCCs treated with TGF-β with anti-SCF+GM-CSF antibodies, compared to control TGF-β group, suggesting that SCF+GM-CSF are involved in TGF-β-induced changes in this system. *P < 0.05; **P < 0.01, relative to control; #P < 0.05, compared with expression in control TGF-β group.

Download figure to PowerPoint

Because TGF-β is a well-known S100A4 inducer and because the above experiments suggest that TGF-β can induce SCF/GM-CSF production and release, the next experiments were designed to evaluate whether SCF+GM-CSF and TGF-β-induced S100A4 expression in small cholangiocytes would occur via a related pathway. At a dose of 10 ng/mL, both TGF-β and SCF+GM-CSF induced significant cell remodeling after 72 hours of incubation, as measured by the incorporation of S100A4 mRNA expression (Fig. 5E). Next, the effects of SCF+GM-CSF blockade in the presence of TGF-β were measured. When small cholangiocytes were incubated with 10 ng/mL of TGF-β plus anti-SCF and GM-CSF antibodies (10 μg/L), a significant decrease in S100A4 mRNA expression was noted, suggesting that SCF+GM-CSF plays an important role in TGF-β-induced remodeling in this system. Meanwhile, when large cholangiocytes were incubated with the same combinations, only a slight decrease in S100A4 expression level was observed, which did not reach statistical significance (Fig. 5E).

SCF+GM-CSF Enhance miR-181 Targeting of Tissue Inhibitor of Metalloproteinases-3.

Our recent data suggested that miRNA-181 is critical for hepatic cell remodeling and differentiation and target 3′-UTR of tissue inhibitor of metalloproteinases-3 (TIMP-3), leading to TIMP-3 mRNA degradation.29 TIMP-3 has unique domains that interact with extracellular matrix (ECM) components and, unlike the other TIMPs, is mainly bound to tissue matrix.30 Because miR-181 has been shown to be regulated by TGF-β, we performed experiments to determine whether SCF+GM-CSF would affect the miR-181 targeting of TIMP-3 in SMCCs. To this end, small cholangiocytes were transfected with pre-miR-181b or control pre-miRNA in combinations with the treatments of SCF+GM-CSF; cell lysates were obtained to determine the levels of TIMP-3 protein. Levels of TIMP-3 protein and mRNA were reduced in small cells with biliary markers and cotransfected with the miR-181b expression plasmid (pre-miR-181b); SCF+GM-CSF further reduced TIMP-3 protein (Fig. 6A). The level of TIMP-3 expression in small cells with biliary markers was lower than in LMCCs (Fig. 6B). Importantly, TIMP-3 was almost fully degraded in cells simultaneously treated with SCF+GM-CSF and transfected with miR-181b expression vectors; the ability of miR-181b targeting of TIMP-3 degradation was offset when SCF and GM-CSF were knocked down by specific antibodies (Fig. 6C). Levels of TIMP-3 protein in anti-SCF+GM-CSF-treated cells were also higher than in control small cells with biliary markers (Fig. 6C). Therefore, SCF+GM-CSF accelerates miR-181b-induced TIMP-3 protein degradation, whereas anti-SCF+GM-CSF counteracts this process. It is of note that TIMP-3 protein is almost absent when SCF+GM-CSF and miR-181b were treated/transfected simultaneously. These findings suggest that the level of TIMP-3 in SMCCs is regulated by SCF+GM-CSF and miR-181b.

thumbnail image

Figure 6. SCF+GM-CSF facilitates miR-181b-induced TIMP-3 degradation. (A and C) SMCCs were transfected with control or miR-181b precursors with or without cotreatment with SCF+GM-CSF (10 ng/mL, 72 hours). (A) Western blotting with TIMP-3 antibody and β-actin was used as internal control, whereas (C) is real-time PCR analysis for TIMP-3 mRNA level, and GAPDH was used as internal control. (B) Total RNA from SMCCs and LMCCs treated with a different combination of SCF+GM-CSF indicated in the bottom were subjected to real-time PCR analyses for TIMP-3 mRNA level. GAPDH was used as internal control. mRNA expression values relative to control SMCCs group are shown (n = 6). (D) Western blotting using TIMP-3 antibodies in SMCCs treated with a different combination of SCF+GM-CSF for 72 hours, as indicated in the bottom. β-actin was used as the loading control. *P < 0.05, relative to the no treatment/control group.

Download figure to PowerPoint

Alteration of Matrix-Remodeling Enzymes: MMPs in Cholangiocytes After Treatments of SCF+GM-CSF.

PH triggers the hepatocyte/proliferation of subpopulations sharing biliary epithelial markers/apoptosis and hepatic/biliary matrix remodeling, all of which are important events in the regenerating liver. MMPs are a family of zinc-containing neutral proteinases involved in matrix remodeling in both normal and pathophysiological processes. The onset of ECM production after PH correlates with a peak in the expression of TGF-β mRNA, which stimulates collagen synthesis. To confirm the functional relevance of SCF+GM-CSF-dependent modulation of regeneration, we assessed the expression of MMPs involved in cell remodeling. Treatment of small mouse cholangiocytes with SCF+GM-CSF significantly increased MMP-2 and MMP-9 mRNA expressions, when compared to SCF or GM-CSF alone. Compared with the controls, the expression of both MMP-2 and MMP-9 was increased in TGF-β-treated small cholangiocytes by 2.1- ± 0.4-fold and 3.6- ± 0.7-fold, respectively. Furthermore, the expression of both MMP-2 and MMP-9 were decreased after the addition of antibodies of SCF+GM-CSF in TGF-β treated small cells with biliary markers (Fig. 7). The up-regulations of MMP-2, MMP-9, as well as S100A4 were also observed in rat liver tissues as well as in IRCs 3 days after PH using the specific PCR Array kit from SABiosciences (#PAHS-033A; Frederick, MD) (Fig. 8A-C), along with the significant increase of miR-181b level and the reduced TIMP-3 expression by zymogen gel assay and real-time PCR analysis (Fig. 8D). Our findings provide evidence of a link between SCF+GM-CSF and the expression of mediators of cell remodeling in biliary cell subpopulations and lines. These data suggest that altered expression of SCF+GM-CSF after PH can contribute to biliary remodeling by functional deregulation of activity of key signaling intermediates.

thumbnail image

Figure 7. SCF+GM-CSF modulate mRNA expression of MMPs. (A and B) LMCCs and SMCCs were treated with SCF and SCF plus EPO, G-CSF, GM-CSF, or controls. Quantitative real-time PCR was performed for MMP-2, MMP-9, and GAPDH mRNA expression. SCF+GM-CSF treatment increased MMP-9 mRNA expression in SMCCs. *P < 0.05, relative to SCF-only group. (C and D) SMCCs and LMCCs were incubated with 10 ng/mL of TGF-β with anti-SCF+GM-CSF antibodies (10 μg/L) or control antiserum for 72 hours, total RNA was isolated and quantitative real-time PCR for MMP-2 and MMP-9 was performed using a superarray quatitative PCR assay kit. Expression of MMPs was normalized to that of the GAPDH gene control. Inhibition of SCF+GM-CSF reduced both MMP-2 and MMP-9 mRNA expression in SMCCs, compared with controls. (E and F) Immunocytochemistry for MMP-2 (E) and MMP-9 (F) was performed in SMCCs. An increase in MMP-2 expression, along with the enhanced expression of MMP-9, was observed in SCF+GM-CSF-treated cholangiocytes. *P < 0.05; **P < 0.01, relative to control.

Download figure to PowerPoint

thumbnail image

Figure 8. Alterations of MMP-2, MMP-9, TIMP-3, miR-181b, and S100A4 in rat liver and isolated intrahepatic rat cholangiocytes after PH. Rats underwent 70% hepatectomy or sham control, and liver tissue was obtained for specific PCR array assay (PAHS-033A; SABiosciences, Frederick, MD) at 72 hours postoperatively. We choose this PCR array assay because this is the only 84-gene group PCR array, which includes S100A4 and other remodeling molecules. (A and B( Total RNAs from liver tissues/isolated IRCs of control and PH were characterized in technical triplicates, and the relative expression levels and P values for each gene in the related samples are plotted against each other in the scatter and volcano plots, depicting the relative expression levels (Log10) for selected genes in treated versus control panels (A) or in P values versus fold changes (B). Genes for which the difference in expression levels was greater than a factor of 3 and/or P < 0.05 are shown out of the cut-off lines on the graph. MMP-2, MMP-9 and S100A4 were among the top up-regulated genes in the PH group, when compared to sham control rats with P < 0.05. (C and D) Total RNAs from liver tissues/isolated IRCs of control and PH (3 days) were subjected to MMP zymogen gel assay (C) and real-time PCR assay (D), respectively. MMP-2, MMP-9, and miR-181b were significantly up-regulated, whereas TIMP-3 was reduced in total liver tissues as well as IRCs after PH (72 hours). *P < 0.05, relative to controls.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Although a role for cholangiocytes in liver regeneration has been proposed, the molecular mechanisms by which cytokines modulate hepatobiliary remodeling and subsequently contribute to hepatic repair are unknown. We show that SCF+GM-CSF expressions are significantly increased after PH, compared to normal controls. Moreover, we demonstrate that the combination of SCF+GM-CSF stimulates specific cholangiocyte subpopulations in proliferation, remodeling, migration, and associated mesenchymal cell activation via TGF-β-dependent signaling and downstream effects involving the expression of miR-181b as well as matrix-remodeling enzymes TIMP-3, MMP-2, and MMP-9. The identification of SCF+GM-CSF as the important regulators of remodeling events in vitro emphasizes an essential synergistic role of these cytokines in mediating hepatic regeneration and repair and provides insight into the contribution of altered remodeling in recovering from severe liver injury.

Cholangiocytes and hepatocytes share embryologic origins. This common heritage contributes traits carried into adulthood. If specific cells exist within the adult liver with multipotential capacity, they should be able to differentiate into either hepatocytes or cholangiocytes in particular circumstances, such as when late-lineage-stage hepatocytes are lost or their regeneration is blocked. Plasticity between intrahepatic cholangiocytes and hepatocytes has been assumed,17, 31 suggesting that terminally differentiated cells of one lineage are able to directly differentiate into another lineage or undergo transdifferentiation.32 However, the data are interpretable also as expansion and differentiation of a stem cell population.4 The extent to which liver stem cells mediate liver regeneration has been hotly debated. One of the primary reasons for this controversy is the use of multiple definitions for the HSC.33 Recent studies have demonstrated that meticulously isolated, rigorously characterized gallbladder epithelial cells cultured under defined in vitro conditions acquire hepatocyte-like properties, such as the ability to synthesize bile acids and take up low-density lipoprotein, without expression of oval cell or hematopoietic stem cell markers.12 The recent discovery of biliary tree stem cell populations34 again provides an alternative interpretation of the expansion and differentiation of stem cells, as opposed to transdifferentiation. Therefore, specific subpopulations of cells that express some of the known cholangiocyte markers can be hypothesized to contain a multipotent stem cell population when exposed to certain environmental conditions. Our studies have suggested that such cells could attain functional pluripotent characteristics after synergistic treatment of SCF+GM-CSFs, and subsequently, they could be used to repopulate damaged livers.

Inadequate liver regeneration is still an unsolved problem in major liver resection and living donor liver transplantation. The studies have implicated the usage of cytokines as the exogenous stimulators of liver regeneration in various animal models of liver resection and liver transplantation.35 SCF and GM-CSF have been demonstrated to affect cellular differentiation and proliferation in various types of cells besides hepatobiliary epithelial cells. SCF enhances growth and differentiation when combined with other cytokines.36 SCF and GM-CSF induce synergistic proliferation and differentiation in myeloid progenitor cells, including the megakaryoblastic cell line, MO7e.37 Synergistic effects are of critical importance biologically, because hematopoietic stem cells and early progenitor cells require growth factors in combination for self-renewal and differentiation. Moreover, in the bone marrow microenvironment, the physiological actions of SCF occur in combination with other growth factors and ECM proteins. The combination of SCF+GM-CSF, promoting remodeling within the liver, supports the possibility of TGF-β-dependent mechanisms contributing to synergistic hepatic repair in response to SCF and GM-CSF.

Our current study indicates that there are relatively high constitutive SCF and GM-CSF levels within the liver during regeneration. SCF and GM-CSF are normally found as the transmembrane molecules under homeostatic conditions and are solubilized after inflammatory stimuli to induce the proper enzyme release to cleave it from the cell surface. This procedure may allow a substantial SCF and GM-CSF reservoir to be accumulated on the cell surface, ready for release upon the environmental condition changes. We hypothesize that solubilized SCF and GM-CSF released during liver regeneration could interact via their receptors with the surrounding specific subpopulations of cells with biliary marker populations, functioning as a remodeling agent within the damaged tissue. Although hepatobiliary homeostasis and regeneration likely involve multiple complex mechanisms and pathways, our current data strongly suggest that SCF and GM-CSF play significant synergistic roles in reestablishing the homeostasis of the biliary system.

The expression of downstream mediators of biliary remodeling could be modulated by SCF+GM-CSF. Therapeutic strategies to increase SCF+GM-CSF may be potentially useful to rebuild the hepatobiliary system after liver injury. Knowledge of specific processes, such as biliary proliferation, migration, remodeling, and mesenchymal transition, that are regulated by CSFs, and the identification of critical targets for SCF and GM-CSF, provides novel insights into mechanisms in the development and remodeling of the intrahepatic biliary epithelium.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_24673_sm_SuppFig1.tif616KSupporting Informatin Figure 1
HEP_24673_sm_SuppFig2.tif518KSupporting Informatin Figure 2

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.