Trps1 Regulates Biliary Epithelial-Mesenchymal Transition and Has Roles during Biliary Fibrosis in Liver Grafts: A Preliminary Study

Objective To investigate the role(s) of Trps1 in non-anastomotic biliary stricture (NABS) following liver transplantation. Methods Immunohistochemical and histological techniques were used to detect Trps1, E-cadherin, CK19, vimentin, α-SMA, and collagen deposition. Human intrahepatic biliary epithelial cells (HIBECs) were infected with a Trps1 adenovirus, or transfected with Trps1 short-interfering RNAs (siRNAs). Reverse transcription polymerase chain reaction (RT-PCR) assays and western blotting were used to determine expression levels of epithelial and mesenchymal markers, and Trps1 in HIBECs. Results Expression of Trps1 and epithelial markers was down-regulated or absent in NABS liver samples. Mesenchymal markers were seen in biliary epithelial cells (BECs), with collagen deposited around the bile duct. Trps1 expression positively correlated with epithelial markers. Expression of epithelial marker mRNAs and proteins in HIBECs decreased with prolonged cold preservation (CP), while mesenchymal marker expression increased. A 12-h CP period led to increased Trps1 mRNA and protein levels. Expression of E-cadherin was increased in HIBECs following Trps1 adenovirus infection and CP/reperfusion injury (CPRI), with vimentin expression levels reduced and CPRI-mediated epithelial-mesenchymal transition (EMT) inhibited. Transfection of HIBECs with Trps1 siRNAs in conjunction with CPRI revealed that E-cadherin expression was decreased, vimentin expression was increased, and CPRI-mediated EMT was promoted. Conclusion Trps1 is involved in NABS pathogenesis following liver transplantation and negatively correlates with BEC EMT and biliary fibrosis in liver grafts. Trps1 demonstrates antagonistic effects that could reverse EMT.


Introduction
Non-anastomotic biliary stricture (NABS) is a major biliary complication after liver transplantation, and can result in biliary cirrhosis, graft failure, re-transplantation, or death. NABS is a significant bottleneck that limits the improvement of liver transplantation efficacy. Pathologically, NABS progresses with biliary fibrosis of the graft, resulting in ineffective or excessive repair of the injured bile duct epithelium. Cold preservation/reperfusion injury (CPRI) is an independent factor of biliary fibrosis following liver transplantation, however the mechanisms involved require clarification [1].
Previous studies into the mechanisms of biliary fibrosis in liver grafts have primarily focused upon the pathway of TGF-β-mediated hepatic stellate cell activation. Cells that synthesize matrix are derived from the inherent source (activated hepatic stellate cell) and other heterologous cell populations [2]. Hepatocyte or biliary epithelial cells (BECs) are the major matrix-synthesizing cells in liver and biliary tract fibrosis during epithelial-mesenchymal transition (EMT) [3].
EMT is closely related to embryogenesis, injury repair, organ fibrosis, and tumor invasion/metastasis. EMT has been shown to promote organ fibrosis in multiple studies [4,5]. During the development of primary biliary cirrhosis and primary sclerosing cholangitis, EMT of bile duct epithelial cells is a major source of matrix-producing cells, and is a critical factor in liver and biliary tract fibrosis [6,7]. In vitro studies have confirmed that mature BECs can undergo EMT in the presence of particular pro-inflammatory cytokines [8]. Liu X et al. [9] observed EMT of BECs during biliary fibrosis following orthotopic liver transplantation in Sprague-Dawley rats.
The development and progression of EMT requires specific micro-environmental conditions and precise regulation of intracellular signals. Certain signaling molecules, including transforming growth factor (TGF)-β, Notch, Wnt, matrix metalloproteinases, Snail, Twist, and Zeb have been shown to stimulate EMT [10,11]. Several other signaling molecules, such as tricho-rhino-phalangeal syndrome type 1 (Trps1) and bone morphogenetic protein 7 (BMP-7) have been shown to inhibit EMT [12,13]. Trps1 is an atypical GATA transcription factor that specifically binds to the GATA motif of the Trps1 gene. It competitively inhibits Trps1 expression that is activated by other GATA transcription factors (GATA-1, -2, and -3) [14]. Recent studies have shown that Trps1 is involved in antagonizing EMT in renal tubular epithelial cells [15], alveolar epithelial cells [16], and breast cancer cells [17]. Whether Trps1 is involved in the negative regulation of BEC EMT, and inhibits biliary fibrosis in grafts following liver transplantation remains unknown. We sought investigate the role(s) of Trps1 in NABS following liver transplantation.

Patients and tissue samples
We used six paraffin-embedded secondary liver transplant samples, taken from January 2010 to June 2012, that were histopathologically diagnosed with biliary fibrosis of the graft at the Southwest Hospital of the Third Military Medical University. A summary of patient characteristics are shown in Table 1, and all diagnoses were made by a full-time pathology doctor. Normal bile duct surrounding hepatic cavernous hemangioma tumor tissue was used as a control (7 cases). We obtained patient consent and approval from the Research Ethics Committee of the Southwest Hospital of the Third Military Medical University to use these clinical materials in our study.

Masson's trichrome staining
Specimens were fixed in 10% buffered formalin, embedded in paraffin, sectioned (4-μm thickness), and stained with hematoxylin and eosin. Paraffin sections were also stained with 1% (v/ v) HCl and Masson's composite staining solution (Sigma) was added drop-wise for 5-10 min. Sections were subsequently washed with distilled water, treated with 1% phosphotungstic acid, and incubated with aniline blue staining solution for 5-10 min. After treatment with 1% glacial acetic acid for 1 min and dehydration with 95% (v/v) alcohol, sections were mounted with neutral gum. Collagen fibers around the bile duct stained blue. For each sample, we randomly selected five fields of view at a high power (×200 magnification), and calculated the proportion of blue staining around the bile duct using Image-pro software.

Cell culture and in vitro CPRI
HIBECs (P5100) were incubated at 37°C/5% CO 2 and passaged while in log phase using 0.25% (w/v) trypsin. Replication of CPRI was achieved by placing cells at 4°C in UW solution for 12, 24, and 48 h, followed by transfer to regular culture medium containing serum, and then incubating at 37°C for 1 h [18].

Cell infection and transfection
HIBECs were seeded in 6-well plates (2 × 10 5 cells/well) and cultured for 24 h or until 60-80% confluent. Cells were infected with either recombinant adenovirus containing Trps1 using a multiplicity of infection (MOI) of 80, or an empty virus control. Virus was allowed to adsorb to cells for 2 h at 37°C/5% CO 2 before the medium was replaced with complete RPMI 1640 containing FBS. Cells were maintained at 37°C/5% CO 2 for another 48 h before cells were harvested. For siRNA transfection, HIBECs were plated in 6-well plates (2 × 10 5 cells/well) and cultured for 24 h or until 60-80% confluent. We added Trps1 or control siRNAs at an MOI of 100 in the presence of 250 μl of serum-free, antibiotic-free RPMI 1640. Cultures were mixed gently and incubated at 37°C/5% CO 2 for 6 h. Medium was replaced with RPMI 1640 containing serum and incubated for another 24 h before cells were harvested.

Trps1 immunohistochemistry
Dry and sterile coverslips were placed in 6-well plates and HIBECs seeded (1 × 10 5 cells/well) in 3 ml of medium containing 10% (v/v) FBS. Cultures were incubated at 37°C/5% CO 2 until cells had attached to coverslips. The culture medium was aspirated and cells rinsed with D-Hank's solution before fixing with acetone (4°C for 10 min). Coverslips were removed, dried, subjected to Trps1 immunohistochemistry and subsequent staining with DAB substrate according to a two-step protocol. Coverslips were thoroughly rinsed with tap water, re-stained, and dehydrated. Coverslips were sealed with mounting medium.

Quantitative PCR (qPCR) assays
Total RNA was prepared from HIBECs using Trizol reagent. Reverse transcription of total RNA into cDNA was conducted using a ReverTra Ace RT-PCR kit according to the manufacturer's instructions. Specific primers targeting Trps1, E-cadherin, CK19, Vimentin, and α-SMA were designed, using Gene Tool, against gene sequences in GenBank (Table 2). We used β-actin as an internal control reference gene. Primers were synthesized by Shanghai Jierui Bioengineering Co., Ltd. Amplification reactions were performed in a 20-μl volume containing 2 μl of cDNA template, 10 μl of 2× PCR Mix Taq, 10 pmol of forward primer, and 10 pmol of reverse primer. Amplicons were separated by electrophoresis at 120 V for 20 min on a 2% (W/ V) agarose gel containing ethidium bromide and visualized under UV light. Images were acquired using Quantity One 4.5 analysis software (Bio-Rad) and band intensities determined. The ratio of the integral optical density (IOD) for each PCR product to the IOD of the β-actin amplicon was used to express the relative gene content in HIBECs.

Western blotting
Total protein was prepared from HIBECs using cell lysis buffer, and quantified using Coomassie blue staining. An equal amount of protein per sample was loaded onto polyacrylamide gels,   overnight. Membranes were then incubated with a horseradish peroxidase (HrP)-conjugated secondary antibody at 37°C for 2 h. Membranes were washed, incubated with chemiluminescent substrate for 5 min, and exposed to X-ray film. Bands were quantitatively analyzed using Quantity One 4.5 software. All experiments were repeated at least three times. Results are expressed as the intensity ratios for the proteins of interest against the internal control.

Scratch assays
HIBECs were seeded into 6-well plates and mitomycin C (1 μg/ml) added to confluent cultured cells for 2 h to inhibit proliferation. Monolayers were then uniformly 'scratched' with the tip of a 1-ml pipet. Culture medium was replaced and cells left room temperature; after 12 h, the migratory distances of cells along the scratches were observed and velocities compared.

Statistical analysis
We used SPSS 13.0 (SPSS Inc.) to perform two-sample comparisons using independent t-tests.
Multiple-sample comparisons were performed using χ 2 tests. The correlation between two variables was analyzed using linear regression.

HIBECs morphology and migration
Following a 12-h CP period and 1 h of RI, normally ovoid HIBECs were transformed into fibroblast-like cells with a spindle morphology. Additionally, their cell migratory speed was significantly increased from 8.3 ± 2.2 μm/h to 32.5 ± 4.7 μm/h (p = 0.02; Fig 1).

Immunohistochemistry and collagen deposition
Positive staining for E-cadherin and CK19 was seen on the cellular membrane and in the cytoplasm. In control groups, epithelial markers were normally expressed in all BECs In the NABS group, expression of some of the epithelial markers was absent Vimentin and α-SMA were expressed at variable levels in the NABS group and not detected in the control group Trps1 was mainly expressed in the nuclei of BECs, with a small fraction expressed in the cytoplasm. In the control group, Trps1s was normally expressed in almost all BECs however in the NABS group, biliary epithelial Trps1 expression was less extensive (Fig 2). In the NABS group, bile duct hyperplasia was distinct, and a large amount of collagen was deposited around the bile ducts. There appeared to be severe damage to BECs as distinguished by altered cell morphologies. The BECs in the control group had normal morphologies and very little collagen deposited around the bile duct. Samples from NABS patients exhibited significantly reduced expression levels of E-cadherin, CK-19, Vimentin, α-SMA, and Trps1 (P < 0.01; Table 3) in comparison with those for the control group. Significant differences in biliary fibrosis were also seen between the two groups (P < 0.01), with NABS samples having more severe bile duct fibrosis (Table 4). In NABS samples, Trps1 expression levels positively correlated with epithelial marker expression, and negatively correlated with mesenchymal marker expression, and degree of biliary fibrosis in the liver graft (Table 5).

Trps1 expression in HIBECs
The Trps1 protein was primarily seen in the nucleus, and to a lesser extent in the cytoplasm, of cultured HIBECs (Fig 3). Following 12 h of CP, expression levels of the Trps1 gene and protein were significantly increased (p = 0.018 and 0.031, respectively) in HIBECs compared with those in untreated controls. In contrast, these were significantly decreased in HIBECs subjected to 24-or 48-h CP (p = 0.026 and 0.009, respectively), with a continued decrease observed after 48 h of CP (Fig 4).

Epithelial/mesenchymal marker expression in HIBECs
In HIBECs subjected to 12, 24, or 48 h of CP, mRNA transcription and protein expression levels of E-cadherin were significantly decreased (p = 0.018 and 0.011, respectively), as were CK19  levels (p = 0.039 and 0.044, respectively). These decreases were more pronounced as the CP period was prolonged. In contrast, mRNA and protein expression levels in HIBECs subjected to CPRI were significantly increased for Vimentin (p = 0.002 and 0.026, respectively) and α-SMA  (p = 0.035 and 0.016, respectively) compared with those in untreated controls. These differences became more distinct as the CP period was prolonged (Fig 4). Trps1 mRNA expression levels positively correlated with those for epithelial markers (E-cadherin and CK19), and negatively correlated with those for mesenchymal markers (Vimentin and α-SMA; Table 6).

Effects of Trps1 expression levels on EMT
Cells infected with the Trps1 adenovirus showed increased Trps1 mRNA levels compared with those in uninfected cells or cells infected with the empty control virus. When the infected HIBECs were subjected to CPRI, we observed alterations in the expression levels of genes associated with EMT. There was a significant increase in E-cadherin mRNA and protein levels (p = 0.002 and p < 0.001, respectively), and a significant decrease in Vimentin mRNA and protein levels (p < 0.001, p = 0.017), with CPRI-mediated EMT inhibited.Trps1 mRNA levels were decreased in cells transfected with the Trps1-specific siRNA compared with those in untransfected cells or those transfected with the control siRNA (Fig 5). In contrast, HIBECs transfected with the Trps1-specific siRNA and subjected to CPRI showed significantly decreased mRNA and protein expression levels for E-cadherin (p = 0.032 and 0.044, respectively). Vimentin mRNA and protein levels were significantly increased compared with those in untreated control cells (p = 0.047 and 0.031, respectively), with CPRI-mediated EMT enhanced (Figs 6 and 7). Recent studies investigating the relationship between Trps1 and tumor malignancy have revealed reduced cellular expression levels of Trps1 at the primary sites of endometrial carcinoma and breast, prostate, pancreatic, and colon cancers [20][21][22][23][24]. EMT was induced and cancer cells acquired the ability to migrate, thereby invading and disseminating to distant sites. However, the functions of Trps1 in wound healing and organ fibrosis have only been confirmed in the kidneys of experimental animals. Huang et al. [25] used a rat model of acute renal injury involving bilateral renal artery occlusion. They found that Trps1 expression was up-regulated during renal injury repair and negatively correlated with renal injury markers, indicating that it played roles in renal repair and regeneration. Gai et al. [26] used a model of unilateral ureteral obstruction in mice to generate kidney fibrosis. They confirmed that Trps1 +/mice experienced stronger EMT than wild-type Trps1 +/+ mice, supporting the hypothesis that Trps1 has a role in inhibiting EMT.
Using samples from NABS patients that underwent secondary liver transplantation, we found that Trps1 expression was either significantly reduced (P < 0.01) or absent when compared with that in normal bile duct tissues. Correlation between Trps1 changes in BECs and liver grafts, and expression of epithelial and mesenchymal markers confirm the involvement of Trps1 in the regulation of biliary EMT and fibrosis after liver transplantation, and therefore might play critical antagonist roles.
CPRI independently affects biliary fibrosis in liver grafts. Prolonged CP of liver grafts aggravates biliary endothelial damage during liver transplantation, and is a major reason for postsurgery biliary stricture and subsequent transplant complications [27][28][29]. A possible mechanism by which this occurs is CPRI-mediated EMT in BECs. Our findings confirm that BECs adopt mesenchymal characteristics following CPRI, an indicator that CPRI induces EMT in HIBECs. Prolonged CP (12-48 h) led to a more distinct form of EMT in BECs.
Our results regarding Trps1 expression patterns implies a role for Trps1 as a critical antagonist of EMT, and also suggests an important role for Trps1 in inhibiting and reversing the pathogenesis of organ fibrosis. Our results are consistent with the negative regulatory effects of Trps1 seen in other epithelial tissues and cells, where Trps1 expression initially increases and then gradually decreases. We postulate that increased Trps1 expression in BECs subjected to CP for less than 12 h might represent a compensatory process that protects cells from EMT progression. However, as the period of CP was prolonged, these compensatory effects declined and Trps1 expression levels gradually decreased, allowing for EMT to progress rapidly.
To the best of our knowledge, we are the first to confirm that Trps1 is involved in CPRI-mediated EMT in BECs, highlighting the antagonistic role of Trps1. CPRI inhibits Trps1 expression in BECs, subsequently stimulating EMT. This serves as an important factor in the development of biliary fibrosis in liver grafts. We propose that Trps1 represents a potential new target for the treatment of NABS that accompanies liver transplantation, however further research in alternative animal models is required.