Tissue microarrays (TMAs) were constructed from archival paraffin embedded tissue samples of 68 intrahepatic CCA patients, 46 HCC patients and tissues from 6 normal healthy livers. These patients underwent liver resection at Srinagarind Hospital, Khon Kaen University, Thailand during 1999-2010. Diagnosis of both CCA and HCC patients were evaluated by clinical data, imaging analysis, tumor markers, and pathology. Immunohistochemical studies for pathological diagnosis included antibodies to CK7, cancer antigen or carbohydrate antigen 19-9 (CA19-9), HepPar1 and alpha-fetoprotein (AFP). The tumor tissues were verified based on the following criteria: as CCA when either CK7⁺ or HepPar1^-, with or without CA19-9⁺, were found; or as HCC when either HepPar1⁺ or CK7^-, with or without AFP⁺, were found. The study protocol was approved by The Human Research Ethics Committee, Khon Kaen University, Thailand (HE551407).

TMAs of 68 O. viverrini-associated CCA cases and 46 HCC cases were generated manually from the paraffin-embedded tissues. In brief, four randomly selected regions from each paraffin block were identified on a hematoxylin-eosin (HE)-stained slide, after which the slide was aligned with the surface of the original paraffin block to locate the sampling areas. The designated areas in the paraffin block were punched with a 1-mm-diameter needle before each punched tissue was then manually transferred to a new recipient paraffin block to generate a TMA block. Five-micrometer-thick sections were cut from the TMA block, and applied on silane-coated slides (Sigma, St. Louis, MO, United States). After TMA construction, to confirm the presence of intact tumor tissue, a HE stained section of the TMA block was prepared and reviewed by two independent pathologists.

An immunohistochemical reaction was performed on 5 μm-thick sections of TMA on silane-coated slides (Sigma) by an immunoperoxidase method. Tissue samples were deparaffinized and dehydrated before the endogenous peroxidase activity was blocked by adding 3% H₂O₂ in phosphate buffer saline (PBS) for 30 min. After washing with PBS, pH 7.4, and blocking with 2% skim milk in PBS, pH 7.4, for 30 min, the samples were incubated with rabbit polyclonal anti-ANXA1 antibody (1:100, Santa Cruz, Heidelberg, Germany) diluted in 2% nonfat dried milk, followed by a horseradish peroxidase (HRP)-conjugated secondary antibody (1:400, GE healthcare, Piscataway, NJ, United States). The appearance of the brown color corresponding to the peroxidase activity was developed using 3,3-diaminobenzidine tetrahydrochloride as a chromogen and counterstained with Mayer’s hematoxylin. Negative controls were performed in a similar manner but omitting the primary antibody. The ANXA1 staining was scored based on signal intensity and positive area as follows: negative, < 10%; weak (+), 10%-25%; moderate (++), 26%-75%; and strong (+++), > 75%. Consensus evaluation from at least two of the three investigators was considered acceptable.

The Animal Ethics Committee of Khon Kaen University, Thailand approved the study protocols for the animal experiments (AEKKU 17/2552). Thirty male Syrian golden hamsters (Mesocricetus auratus) aged between 4 and 6 wk were obtained from the Animal Unit, Faculty of Medicine, Khon Kaen University, Khon Kaen, Thailand. Animals were divided into two groups; normal control group and infection with O. viverrini plus N-nitrosodimethylamine (NDMA) (O. viverrini + NDMA). In the treated group, hamsters induced by infection with O. viverrini metacercariae isolated from naturally infected cyprinid fish by pepsin (Wako, Japan) digestion and subsequently treated with 12.5 ppm NDMA at the same time point as described previously[23]. NDMA was given in drinking water for 2 mo, and withdrawn thereafter until the animals were sacrificed at 21 d, and then at 3 and 6 mo post-treatment (n = 5 for each sub-group). Animals were anaesthetized and killed with an overdose of diethyl ether. The liver was dissected and was placed in 10% buffered formalin and used to evaluate histopathological changes and immunohistochemical studies[23].

Four human CCA cell lines, namely M156, M055, M213 and M214 were isolated from intrahepatic CCA patients from northeastern Thailand. CCA tissues were characterized as M156, M055 and M214 moderately differentiated CCA and M213 adenosquamous cell carcinoma. Those CCA cell lines were used to assess ANXA1 gene expression and compared to an immortalized human cholangiocyte cell line (MMNK1). In addition, the M214 CCA cell line was used to verify ANXA1-related molecule regulation. These CCA cell lines were kindly provided by Associate Professor Banchop Sripa. All cell culture materials (media, serum and antibiotics) were purchased from Gibco, Invitrogen (Auckland, New Zealand). The CCA cells were cultured in HAM’s F-12 medium supplemented with 10% heat-inactivated fetal bovine serum, 1% Penicillin-streptomycin, 1.176 g/L sodium bicarbonate and adjusted to pH 7.1 with 1 mol/L HCl. The cells were incubated at 37 °C under a humidified 5% CO₂ atmosphere.

Protein was extracted from the CCA cell lines and the concentration was measured by the Bradford assay (Bio-Rad, Hercules, United States) according to the manufacturer’s instructions. Twenty micrograms of protein were separated on a 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis gel and transferred to a polyvinylidene difluoride membrane (Amersham Bioscience, Piscataway, NJ, United States) for 2 h at 60 V. The membrane was incubated overnight at 4 °C with rabbit polyclonal anti-ANXA1 antibody (1:1000, Santa Cruz, Heidelberg, Germany) diluted in 2% nonfat dried milk/phosphate buffered saline with Tween 20 (PBS-T). Subsequently, the membrane was incubated with an appropriate HRP-conjugated secondary antibody (1:3000, GE healthcare) diluted in 2% nonfat dried milk/PBS-T. The immunoreactive materials were developed by enhanced chemiluminescence using the ECL Western blotting Detection Reagent (GE Healthcare).

Total RNA extraction was performed from various CCA cell lines (3 × 10⁵ cells) of each experiment using the TRIzol reagent (Invitrogen, Carlsbad, CA, United States). An aliquot of total RNA was reverse transcribed into cDNA using reverse transcriptase (Invitrogen) following the manufacturer’s protocol. Polymerase chain reaction (PCR) was carried out in duplicate in a 20-μL volume using Faststart Universal SYBR Green Master (ROX, Roche Applied Science, Penzberg, Germany) using the following sets of primers: matrix metalloproteinase 2 (MMP2) (5′-TTGATGGCATCGCTCAGATC-3′ and 5′-CTGCGAAGAACACAGCCTTC-3′), MMP9 (5′-CATTGTCATCCAGTTTGGT G-3′ and 5′-ACCACAACTCGT CGTCGTC-3′), transforming growth factor (TGF)-β (5′-ACATCGACTTTCGCAAGGAC-3′ and 5′-TGGTTGTAG AGGGCAAGGAC-3′), nuclear factor-κB (NF-κB) (5′-GCTTTGCAAACCTGGGAATA-3′ and 5′-CAAGG TCAGAATGCACCAGA-3′), and GAPDH (5’-AGAAGACTGTGGATGGCCCC-3’ and 5’-TGACCTTGCCCACAGCCTT-3’). Reactions were performed in the ABI 7500 thermal cycler (Applied Biosystems, Foster City, CA, United States). All data were analyzed using 7500 system software with a cycle threshold (Ct) in the linear range of amplification and then processed by the 2_‾^ΔΔCt method.

The M214 CCA cells were transiently transfected with siRNA against the ANXA1 gene (Silencer^® Select siRNA; siRNA ID s1380, Ambion, TX, United States) using the lipofectamine™ 2000 transfection reagent (Invitrogen) following the manufacturer’s protocol with some modifications. In brief, 5 μL of 5 μmol/L stock concentration siRNA or scrambled siRNA and 5 μL lipofectamine™ 2000 transfection reagent were separately diluted in 250 μL OPTI-MEM I medium (Invitrogen). The diluted siRNA solution was mixed with the diluted transfection reagent and incubated for 20 min at room temperature before being added to a six-well plate seeded with 1 × 10⁵ CCA cells in 2 mL transfection medium. The level of ANXA1 mRNA was accessed at 48 and 72 h after transfection. The negative controls were performed by using Silencer^® Negative Control siRNA No. 1 (Ambion), a non-targeted sequence.

The data were expressed as mean ± SD. To compare the data between groups, statistical significance was determined by Student’s t-test. The χ² test was used to analyze the correlation between ANXA1 expression and the categorical variables regarding clinicopathological parameters. Survival analysis was done by Kaplan-Meier and log-rank tests. Statistical analyses were performed using SPSS version 15 (SPSS, Inc, Chicago, IL). A P value of less than 0.05 was considered statistically significant.

ANXA1 expression in TMA obtained from 68 CCA and 46 HCC patients was examined using immunohistochemistry. Expression of ANXA1 was observed in the cytoplasm of epithelial bile duct tumor cells and some of inflammatory cells but was seen faintly or with no staining in tumor stroma and in hepatocytes (Figure 1). Sixty-four cases were positive (94.1%, 64/68) for ANXA1 expression. A high expression was coded in 21 cases for ++ and 24 cases for +++ or 66.2% (45/68 cases) and low expression where 4 cases were negative and 19 cases for + or 33.8% (23/68 cases) (Table l; Figure 1). The histological feature in the tubular type showed that high expression (78.8%, 26/33) was significantly higher than low expression (21.2%, 7/33) (P < 0.05) (Table 1).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Immunostaining of annexin A1 protein expression in cholangiocarcinoma tissue microarrays. A-D: Tissue microarray of cholangiocarcinoma samples was stained with anti-annexin A1 antibody and counter stained with Mayer's hematoxylin and represented by A for negative, B for +, C for ++, and D for +++ when expression was < 10%, 10%-25%, 26%-75% and > 75%, respectively (magnification, × 200).

To determine ANXA1 expression distinguishing CCA from HCC, ANXA1 expression was determined in TMA of 46 HCC patients. Slight expression of ANXA1 was observed in the cytoplasm of some inflammatory cells, but was seen faintly or with no staining in tumor cells and hepatocytes. All histologic patterns of HCC samples 100% (46/46) were negative or of low expression (44 for negative and 2 for +) as shown in Figure 2. In addition, no staining or faintly stained immunoreactivity in normal liver tissues of healthy individuals (6/6) was also found. ANXA1 protein as an immunohistological marker could detect CCA and differential CCA from HCC in the primary liver cancer similar to CK7 and CA19-9 (Figure 3). ANXA1 immunohistochemistry showed high sensitivity (94%) and specificity (100%) and the positive prediction value was 100%.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Immunostaining of annexin A1 in hepatocellular carcinoma tissue microarrays. Tissue microarray of hepatocellular carcinoma (HCC) samples was stained with anti-annexin A1 (ANXA1) antibody and counter stained with Mayer's Hematoxylin. ANXA1 shows low expression in all histologic patterns of HCC. A: Trabecular pattern; B: Broad trabecular pattern; C: Trabecular with pseudoacinar pattern; D: Solid growth pattern (magnification, × 40).

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Illustration of the comparative immunohistochemical stains in 3 representative cases. The 3 representative cases include tubular type cholangiocarcinoma (CCA) (Tub CCA), papillary type CCA (Pap CCA) and hepatocellular carcinoma (HCC). Immunopositivity of cytokeratin 7 (CK7), annexin A1 (ANXA1), hepatocyte paraffin 1 (HepPar1) and alpha-fetoprotein (AFP) appeared as brown cytoplasmic staining of tumor cells, while positivity to carbohydrate antigen 19-9 (CA19-9) appeared as cytoplasmic and luminal staining of tumor (magnification, × 40). Noted that CCA are CK7⁺/CA19-9⁺/ANXA1⁺ and HCC is HepPar1⁺/AFP⁺/ANXA1^-. -: Negative result; +: Positive result.

The correlation between ANXA1 expression and clinicopathological parameters was analyzed as shown in Table 1. There were no correlations between ANXA1 expression level and age, sex, tumor location, tumor size, gross type, lymph node metastasis (Table 1) and patients’ survival outcome (data not shown). Notably, high ANXA1 expression levels were positively correlated with the histological features of the tubular type (P = 0.03) but not the papillary type. All 26 cases of high expression in the tubular type had positive lymph nodes.

To evaluate the expression of ANAX1 in the different stages of CCA tissues in O. viverrini + NDMA-induced CCA in the hamster model, immunohistochemistry was used to evaluate the inflamed tissues at 21 d, when the tumors began at 3 mo, and tumor progression at 6 mo post-treatment. The percentage of CCA 60% (3/5) at 3 mo, and 100% (5/5) at 6 mo post-treatment were described previously[23]. Here, it was shown that the expression of ANXA1 was observed mainly in the cytoplasm of the epithelial bile ducts, some inflammatory cells (large cell, macrophage-like cell), and tumor cells which increased with time as bile duct proliferation and CCA development progressed (Figure 4). The level of its expression at 21 d was +, 3 mo was ++, and 6 mo was +++ for all five animals per group. In addition, low or a little expression was found in the normal livers with no changes in each time-sacrificed group.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Immunostaining of annexin A1 protein expression in hamster tissues. A-D: Hamster tissues were stained with anti-annexin A1 antibody and counterstained with Mayer's hematoxylin represented by negative for A (normal at 6 mo post-treatment), grade + for B (21 d post-treatment), grade ++ for C (3 mo post-treatment) and grade +++ for D (6 mo post-treatment) (magnification, × 200). Ov: Opisthorchis viverrini; BD: Bile duct lumen.

Since ANXA1 was positive in CCA tissue but not in HCC tissue, the expression was further confirmed in various intrahepatic human CCA cell lines compared to MMNK1. Real-time reverse transcription (RT)-PCR revealed that a significantly increased expression of the ANXA1 gene was observed in all four CCA cell lines including M156, M055, M213 and M214 compared to in MMNK1 (P < 0.01) (Figure 5).

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Expression of the annexin A1 gene in various cholangiocarcinoma cell lines. The mRNA expression level of annexin A1 (ANAX1) was evaluated by real-time reverse transcription-polymerase chain reaction in various cholangiocarcinoma (CCA) cell lines including M156, M055 and M214 moderately differentiated CCA, M213 adenosquamous cell carcinoma and compared to an immortalized human cholangiocyte cell line (MMNK1). Data are derived from duplicate independent experiments and presented as mean ± SD, P < 0.01, P < 0.001 vs MMNK1 is significantly different by the Student t-test.

In addition, ANXA1 was also investigated whether it was involved in MMPs in CCA metastasis by silencing ANXA1 using siRNA to target of ANXA1 gene expression in M214 CCA cells. Real-time RT-PCR and Western blotting analysis showed that the mRNA level of cells transfected with siRNA at 48 h was suppressed by 87.3% and the protein level was decreased by 80% compared to that of controls (Figure 6). In vitro proliferation of the ANXA1 silenced cells was also reduced (data not shown). Suppression of the ANXA1 gene significantly inhibited the mRNA expression of MMP2 by 50% and MMP9 by 45% (P < 0.05), and TGF-β by 15%, but induced NF-κB by 70% (Figure 7).

Open in New Tab   Full Size Figure   Download Figure

Figure 6 Effect of annexin A1 knockdown on the M214 cholangiocarcinoma cell line. A, B: Suppression of annexin A1 (ANXA1) expression at the translational level (at 48 h) and transcriptional level (at 48 and 72 h) in knockdown M214 cells was evaluated by Western blotting (A) and real-time reverse transcription-polymerase chain reaction (B).

Open in New Tab   Full Size Figure   Download Figure

Figure 7 Effect of gene expression in the knockdown of the annexin A1 gene in the M214 cholangiocarcinoma cell line. A: Relative transforming growth factor-β (TGF-β) mRNA expression; B: Relative matrix metalloproteinase (MMP) 2 mRNA expression; C: Relative MMP9 mRNA expression; D: Relative nuclear factor (NF)-κB mRNA expression. Real-time reverse transcription-polymerase chain reaction was used to confirm the expression of TGF-β, MMP2, MMP9 and NF-κB in the knockdown of the annexin A1 (ANXA1) gene in the M214 cholangiocarcinoma cell line relative to glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Data are derived from duplicate independent experiments and presented as mean ± SD, and P < 0.05 vs scramble is significantly different by the Student t-test.

Cholangiocarcinoma (CCA) is associated with late presentation, has high mortality rate, and poses challenges for diagnostic. The histological CCA is difficult distinguish from hepatocellular carcinoma (HCC). New diagnostic markers with better diagnostic differentiation are required. Annexin A1 (ANXA1) is a multipotent protein involved in several functions including regulation of cell differentiation, apoptosis, and carcinogenesis. ANXA1 involved in this process is frequently overexpressed in many types of cancers. Moreover, its expression in the same cancer is controversial.

Although the expression of ANXA1 has been demonstrated in many types of cancers, its expression in CCA and HCC is controversial. Moreover, its expression at different stages of CCA development and its regulation in CCA has not been reported.

In this study, strong expression of ANXA1 was observed in CCA, but was low in HCC, which was similar to the available markers for CCA (cytokeratin 7 and carbohydrate antigen 19-9), implying its expression could present as a new marker for differential diagnosis of primary liver cancer. ANXA1 increased expression along with cholangiocarcinogenesis in hamsters induced by Opisthorchis viverrini infection and N-nitrosodimethylamine administration. Its expression was positively correlated with transforming growth factor-β (TGF-β) and matrix metalloproteinase (MMP) but negatively with the nuclear factor κB (NF-κB) signaling pathway. ANXA1 is involved in carcinogenesis of chronic inflammation related-CCA. ANXA1 is a promising biomarker and a potential for a therapeutic target in this aggressive cancer.

This study demonstrated that the ANXA1 protein was expressed in CCA, but had low expression in HCC and therefore it may provide a new diagnostic marker in CCA by the immunohistochemistry technique. By silencing of ANXA1 gene expression in the in vitro study the results demonstrate that ANXA1 represents a potential therapeutic target for the future intervention in CCA patients.

Annexin A1 (also named macrocortin, renocortin, lipomodulin and lipocortin 1) is a 37 kDa protein member of an annexin superfamily which has a binding or annexing property to acidic phospholipid in a calcium dependent manner. It also has been shown to play a critical role in the regulation of inflammatory cell activation, cellular processes and involved in carcinogenesis.

This is a good paper with sensible use of different experimental techniques. The authors examined the expression of ANXA1 in CCA and HCC tissues microarrays; ANXA1 protein had high sensitivity and specificity for immunohistochemistry diagnosis in CCA. Suppression of ANXA1 gene expression showed it inhibiting of TGF-β and MMPs but induction NF-κB expression. The results are interesting, represent a new immunohistochemistry analysis of CCA and ANXA1 is a promising therapeutic potential of drugs targeting in CCA.