Arctigenin inhibits cholangiocarcinoma progression by regulating cell migration and cell viability via the N-cadherin and apoptosis pathway

Northeast Thailand has the highest incidence of cholangiocarcinoma (CCA) in the world. The lack of promising diagnostic markers and appropriate therapeutic drugs is the main problem for metastatic stage CCA patients who have a poor prognosis. N-cadherin, a cell adhesion molecule, is usually upregulated in cancers and has been proposed as an important mediator in epithelial-mesenchymal transition (EMT), one of the metastasis processes. Additionally, it has been shown that arctigenin, a seed isolated compound from Arctium lappa, can inhibit cancer cell progression via suppression of N-cadherin pathway. In this study, we investigated the protein expression of N-cadherin and its correlation with clinicopathological data of CCA patients, as well as the impact of arctigenin on KKU-213A and KKU-100 CCA cell lines and its underlying mechanisms. Immunohistochemistry results demonstrated that high expression of N-cadherin was significantly associated with severe CCA stage (p = 0.027), and shorter survival time (p = 0.002) of CCA patients. The mean overall survival times between low and high expression of N-cadherin were 31.6 and 14.8 months, respectively. Wound healing assays showed that arctigenin significantly inhibited CCA cell migration by downregulating N-cadherin whereas upregulating E-cadherin expression. Immunocytochemical staining revealed that arctigenin suppressed the expression of N-cadherin in both CCA cell lines. Furthermore, flow cytometry and western blot analysis revealed that arctigenin significantly reduced CCA cell viability and induced apoptosis via the Bax/Bcl-2/caspase-3 pathway. This research supports the use of N-cadherin as a prognostic marker for CCA and arctigenin as a potential alternative therapy for improving CCA treatment outcomes.


Introduction
Cholangiocarcinoma (CCA) originates from epithelial cells of intrahepatic and extrahepatic bile ducts and is a common form of liver cancer (Khan et al. 2019). CCA is the main public health problem in northeast Thailand, with the highest prevalence worldwide (Kamsa-Ard et al. 2018). The absence of clear symptoms in CCA patients remains a major diagnosis and treatment problem. Most patients are diagnosed with progressive or metastatic stage, with an extremely poor prognosis and not suitable for surgery (Aljiffry et al. 2009;Banales et al. 2020). When the disease reaches an advance stage, palliative treatment is the only option possible. Many chemotherapy drugs, such as cisplatin and gemcitabine, have been commonly used to treat CCA patients, but response rates are poor and survival times of patients are short (Okusaka et al. 2010). The exploration of appropriate therapeutic agents for CCA is therefore important.

3
Epithelial-mesenchymal transformation (EMT) is a major point of the tumor metastasis mechanism acting by enhancing cell invasion and migration (Kalluri and Weinberg 2009). There are a group of proteins that act as EMT biomarkers, such as epithelial marker, E-cadherin and mesenchymal marker, and N-cadherin. An alteration of the role of E-cadherin and N-cadherin contributing to tumor metastasis has been investigated in many cancers (Loh et al. 2019). Techasen et al. reported that low E-cadherin protein expression in CCA tissues was strongly correlated with positive metastasis status, and they proposed that E-cadherin could be used as a CCA metastasis marker (Techasen et al. 2014). Previous studies have reported that high expression of N-cadherin predicts poor prognosis in patients with colorectal, prostate, and gallbladder cancers (Gravdal et al. 2007;Yi et al. 2014;Yan et al. 2015). It appears that acquiring N-cadherin is a crucial step in epithelial cancer metastasis and disease progression (Kalluri and Weinberg 2009;Gheldof and Berx 2013). Therefore, N-cadherin expression is a possible potential biomarker for cancer prognosis.
Natural plant therapeutic properties are currently being used to suppress cancer cell metastasis by the EMT process (Wang and Jiang 2012;Avila-Carrasco et al. 2019;Jiang et al. 2020;Wang et al. 2020). Curcumin, the main phytochemical in turmeric, can reverse TGF-β-1-induced EMT in gastric cancer cells by upregulating E-cadherin and downregulating N-cadherin expression (Cheng et al. 2018). Resveratrol, a grape extract, has been shown to completely inhibit cancer cell motility, and reverted the N-to E-cadherin switch in colon cancer and CCA cells (Thongchot et al. 2018;Yuan et al. 2019). Gedunin, a compound extracted from Azadirachta indica, blocked pancreatic cancer cell metastasis by downregulating the expression of mesenchymal markers, N-cadherin and vimentin, while upregulating the expression of epithelial marker, E-cadherin (Subramani et al. 2017). Moreover, Paeoniflorin, a compound derived from plants in the Paeonia lactiflora Pallas, inhibit glioblastoma cell by suppression of TGF-β signaling pathway via decreasing the expression of snail, N-cadherin and vimentin .
Arctigenin is a member of the Asteraceae family and has pharmacological properties such as anti-oxidant, antiinflammatory, anti-cancer, and anti-metastasis (He et al. 2018). Arctigenin can inhibit cell invasion and migration in hepatocellular carcinoma, colorectal, and lung cancers via suppression of EMT process (Han et al. 2016;Xu et al. 2017;Lu et al. 2019). In addition, Li et al. discovered that arctigenin inhibits cell proliferation and causes apoptosis in colon cancer cells (Li et al. 2016). However, there is no information on the effect of arctigenin on CCA cell progression.
The aims of this study were to investigate the expression of N-cadherin protein in CCA tissues, to determine whether there are any associations with a patient's clinicopathological data, as well as to investigate whether arctigenin could be used as a new therapeutic for inhibiting CCA progression focusing on the switching of N-to E-cadherin pathway. Additionally, the mechanism of arctigenin's action was also examined.

Human CCA tissues
This study was approved by the Human Ethics Committee of Khon Kaen University, based on the ethics of human specimen experimentation of the National Research Council of Thailand (HE631304). The paraffin-embedded CCA tissues (n = 81) were collected from Srinagarind Hospital and were kept in the biospecimen bank of Cholangiocarcinoma Research Institute (CARI), Khon Kaen University, Thailand.

Cell lines and cell culture
The CCA cell lines, KKU-213A (JCRB1557) (Sripa et al. 2020) and KKU-100 (JCRB1568) (Sripa et al. 2005), were developed by Prof. Banchob Sripa at Cholangiocarcinoma Research Institute, Khon Kaen University, Thailand. Both cell lines were deposited to the Japanese Collection of Research Bioresources (JCRB), Osaka, Japan, for all complete identification of characteristics. In our study, cell lines were purchased from JCRB cultured in Ham's F-12 containing 10% fetal bovine serum and penicillin/streptomycin (50 U/ml and 50 µg/ml) and incubated in a humidified incubator of 5% CO 2 at 37 °C.

Chemicals
Arctigenin (Sigma-Aldrich, USA) (purity ≥ 95%) stock solution of 50 mM was prepared by dissolving arctigenin in dimethyl sulfoxide (DMSO) and stored at − 80 °C. Stock solution was diluted with culture medium to prepare the working solutions which were at the desired concentrations. The final concentration of DMSO was 0.4% (v/v) in all treatment conditions.

Immunohistochemical staining for N-cadherin expression
Formalin-fixed paraffin-embedded tissue of 81 CCA patients were deparaffinized in xylene and dehydrated in graded ethanol series. For antigen retrieval, section slides were heated in sodium citrate in microwave oven for 10 min. The endogenous peroxidase activity in the tissues were blocked by incubating the slide in 0.3% hydrogen peroxide (H 2 O 2 ) 1 3 solution in PBS for 30 min. Subsequently, 10% skim milk in PBS was added for 30 min to block non-specific substances and all slides were incubated with mouse anti-Ncadherin monoclonal antibody (mAb, dilution 1:100; BD Transduction Laboratories™, USA) in humid chamber at 4 °C overnight. After washing, slides were incubated with horseradish peroxidase-conjugated Envision™ secondary antibody (DAKO, Denmark) for 1 h and 30 min. DAB peroxidase (HRP) substrate kit (Vector Laboratories, CA), which stained the antigen brown, and hematoxylin counterstaining were used to visualize the reaction. The slides were then dehydrated and mounted, and stained sections were investigated under a light microscope.

Immunohistochemical scoring for N-cadherin expression
The Histoscore (H-score), a semi-quantitative method, was applied for immunohistochemical staining as described by Detre et al. (1995). The evaluations were recorded as percentages of positively stained cancer cells in each of the four intensity categories, which were denoted as 0 (no staining), 1 (weak staining), 2 (moderate staining), and 3 (strong staining). The total number of cells in each field and the number of cells stained at each intensity were counted. The average percentage positive was calculated, and the following formula was applied.
The range of scores were from 0 to 300. The expression level of protein was classified as low or high according to the mean value of the H-score.

Cell migration assay
Cell migration was determined by using a wound healing assay. The KKU-213A and KKU-100 cells were seeded in 24-well plates (10 5 cells/well) and incubated to complete 100% cell confluence. After treatment with arctigenin at concentrations of 100 and 200 μM, wounds were made by pipette tip scratching and monitored by microscopy at 18 h for KKU-213A and 24 h for KKU-100.

Cell proliferation assay
Cell proliferation was determined by using an MTT colorimetric assay. The cells were seeded in 96-well plates (2,000 cells/well) and were incubated for 24 h. Cells were treated with a series concentration of arctigenin and were incubated at 37 °C in 5% CO 2 incubator. After 48 h treatment, the cells were wash with PBS and incubated with MTT reagent at 37 °C for 2 h. Subsequently, the MTT reagent was removed and DMSO was added to dissolve formazan crystals. The optical density was measured at an absorbance of 540 nm using a microplate reader. Data was presented as mean ± SD from two independent experiments. Arctigenin concentrations that produce 50% cell viability (the half maximal inhibitory concentration; IC 50 ) were calculated from curves constructed by plotting cell survival (%) versus arctigenin concentration (µM). The values were represented as the percentage of cell viability comparing with the control which is 0 μM of arctigenin treatment.

Flow cytometric assay for apoptosis evaluation
The apoptotic cell distribution was determined using the Alexa Fluor ® 488 annexin V/Dead Cell Apoptosis Kit (Inv-itrogen™, USA) according to the manufacturer's protocol. Briefly, after 48 h of arctigenin treatment at concentrations of 50, 100, and 200 μM, the cells were trypsinized with trypsin and washed with cold PBS. After PBS was removed, and cells were resuspended with 100 μl of 1 × annexinbinding buffer, 2.5 μl of Alexa Fluor ® 488 annexin V, and 1 μl of 100 μg/ml PI. The stained cells were analyzed by flow cytometry and measured the fluorescence emission at 530 nm and 575 nm using 488-nm excitation.

Western blot analysis
Protein was extracted from cell pellets by using a RIPA cell lysis buffer and the protein concentration was determined by using a Pierce™ BCA Protein assay kit (Pierce Biotechnology, Rockford, IL, USA). The protein solutions were boiled at 95 °C for 5 min on hot plate and cooled on ice. An appropriate amount of protein was loaded onto an SDS-PAGE (4% stacking gel and 8% separating gel) and was transferred onto a PVDF membrane (Merck, Billerica, MA, USA). The transferred membrane was blocked in 5% skim milk in Tris-buffered saline containing 0.1% Tween 20 (TTBS) at room temperature for 1 h before being incubated with primary antibody overnight at 4 °C. The antibodies used in this study included mouse anti-N-cadherin monoclonal antibody (mAb, dilution 1:250; BD Transduction Laboratories™, USA), mouse anti-E-cadherin monoclonal antibody (mAb, dilution 1:1000; BD Transduction Labora-tories™, USA), rabbit anti-caspase-3 monoclonal antibody (mAb, dilution 1:1000; Abcam, UK), rabbit anti-BCL-2 polyclonal antibody (dilution 1:1000; Proteintech, USA), rabbit anti-BAX polyclonal antibody (dilution 1:1000; Proteintech, USA), and mouse anti-β actin monoclonal antibody (mAb, dilution 1:20,000; Sigma-Aldrich, USA). After incubation, the membranes were rinsed with TTBS, incubated with peroxidase-labelled secondary antibodies for 1 h, and rinsed with TTBS. Finally, the membranes were exposed to an Amersham™ ECL™ Prime Western Blot Detection Reagent (GE Healthcare, Buckinghamshire, UK) for chemiluminescence detection using an Amersham Imager™ 600.

Immunocytochemistry analysis
CCA cells were seeded on slide chambers and treated with or without arctigenin for 48 h. After incubation, cells were fixed for 30 min in 4% paraformaldehyde in PBS. Fixed cells were incubated in 3% (w/v) bovine serum albumin in PBS for non-specific binding blocking. Cells were incubated overnight at 4 °C with the primary antibodies specific for N-cadherin (mAb, dilution 1:100; BD Transduction Labo-ratories™, USA). After washing, cells were incubated with secondary antibodies conjugated with Alexa flour 555 (Inv-itrogen™, USA) for 1 h. The preparations were mounted with mounting solution containing a 4,6-diamidino-2-phenylindole (DAPI) for nuclear staining. Slides were visualized under a fluorescence microscope on an Inverted Microscope EVOS™ Imaging System.

Statistical analysis
IBM SPSS V.23.0 statistical package (SPSS Inc., Chicago, IL) was used for statistical analysis. Data were represented as mean ± SD. The associations between N-cadherin expression and clinicopathological parameters of CCA patients was calculated by chi-square test. The log-rank analysis was used to compare survival distributions and the Kaplan-Meier method was plotted for survival analysis. The quantitative bar chart for functional analysis and the dose-response curve for IC 50 was calculated and created by GraphPad Prism version 8.0 for windows (GraphPad software, San Diego, CA). The statistical significance between different groups was detected by one-way ANOVA. p < 0.05 was considered statistically significant.

N-cadherin expression in CCA and its correlation with clinicopathological data
The characteristics of all CCA patients are summarized in Table 1. The mean age of patients was 60 years (range 38-83 years), with 52 males and 29 females. Most patients had non-papillary histology type, and stage III and stage IV CCA at the time of diagnosis. The representative IHC staining of N-cadherin in 81 CCA tissue samples is shown in Fig. 1a. N-cadherin was predominantly localized in cytoplasmic of all CCA tissues. Overall, the expression of N-cadherin was lower in normal bile ducts compared to tumor cells. Using a mean value of H-score of the IHC result as the cutoff point, N-cadherin expression was found to be elevated in 43.2% (35/81) of CCA tissues. As shown in Table 2, high expression of N-cadherin was significantly associated with severe CCA stage (p = 0.027). In addition, high expression of N-cadherin was significantly associated with shorter survival time of CCA patients as shown in Fig. 1b (p = 0.002). The mean overall survival times between low and high expression of N-cadherin were 31.6 and 14.8 months, respectively.

Arctigenin inhibits CCA cell metastasis via reversing of N-to E-cadherin pathway
To further assess the anti-metastatic effect of arctigenin, a wound healing assay was performed to determine whether arctigenin affects the migration ability of KKU-213A and KKU-100 CCA cells. Cell movements were observed 18 and 24 h after arctigenin treatment. In the presence of arctigenin, CCA cell migration was significantly suppressed in a dose-dependent manner, whereas control cells migrated toward the scratched site ( Fig. 2a and b). As compared to the control condition, the percentage of wound area in 100 and 200 μM of arctigenin was significantly different in both CCA cell types ( Fig. 2c and d).
As N-cadherin has been reported to be associated with the metastasis process, we further determined whether arctigenin effect on N-cadherin expression by western blot analysis. As shown in Fig. 2e, arctigenin treatment significantly decreased the expression of N-cadherin and increased the expression of E-cadherin, an epithelial marker in KKU-213A. Furthermore, immunocytochemistry analysis revealed that arctigenin suppressed N-cadherin expression in KKU-213A and KKU-100 cells compared to controls (Fig. 2f).

Arctigenin inhibits CCA cell proliferation and induces apoptosis via Bax/Bcl-2 and caspase-3 pathway
The effects of arctigenin with a concentration range of 5-250 μM on KKU-213A and KKU-100 cell lines were determined using MTT assay. As shown in Fig. 3, cell proliferation rates of KKU-213A and KKU-100 CCA cell lines were significantly reduced after treatment with arctigenin when compared with the control. After 48 h of treatment, the IC 50 in KKU-213A cells was 98.16 μM, while in KKU-100 cells was 120.60 μM.
To identify whether arctigenin inhibited CCA cell proliferation by inducing apoptosis. KKU-213A and KKU-100 cells were treated with a series of arctigenin concentrations. Cells were stained with Annexin V-FITC/PI for flow cytometry analysis. As shown in Fig. 4a, KKU-213A cells treated with 50, 100, and 200 μM of arctigenin for 48 h showed increased apoptotic rate by approximately 60.85%, 73.8%, and 89.3%, respectively. For KKU-100 cells, the apoptotic rates were found to increase approximately 4.35%, 14.4%, and 18.00% after treatment with series of arctigenin concentrations, respectively. The  results indicated that arctigenin significantly induced CCA apoptosis, as shown in the right quadrants of the flow cytometry graphs. All concentrations showed the capacity to strongly induce CCA apoptosis except for dose 50 μM of arctigenin in KKU-100 (Fig. 4b). In addition, the expression of apoptosis-related proteins in KKU-213A cells was investigated using western blot analysis. As shown in Fig. 4c, the expression of pro-caspase-3 and Bcl-2 was decreased, whereas the expression of Bax was increased after arctigenin treatment.

Discussion
Epithelial-mesenchymal transition (EMT) is an important step in the process of tumor metastasis through transforming epithelial cells to mesenchymal phenotypes (Steinestel et al. 2014). The loss of epithelial markers such as E-cadherin while increasing mesenchymal markers such as N-cadherin is the most remarkable occurrence in cancer cells that develop metastatic properties (Suresh and Nathawat 2014). Techasen et al. reported that low E-cadherin protein expression in CCA tissues was significantly associated with positive metastasis status. They suggested that E-cadherin might act as a central modulator of tumor cell phenotype and can also serve as a potential metastasis marker in CCA (Techasen et al. 2014). Aside from epithelial markers, mesenchymal markers should also be identified. N-cadherin is a member of the calciumdependent adhesion molecule family of classical cadherins, which specifically mediate cell-cell adhesion. It is typically absent in normal epithelial cells, and the aberrant expression of N-cadherin in epithelial cancer cells is a well-documented characteristic of epithelial malignancies (Tomita et al. 2000;Nakajima et al. 2004;Lascombe et al. 2006;Choi et al. 2013). Several studies have found a correlation between high N-cadherin protein expression and cancer progression in a variety of cancers (Ning et al. 2013;Okubo et al. 2017;Abufaraj et al. 2018). Increasing N-cadherin expression was associated with late stage and lymph node metastasis in patients with breast and prostate cancers (Ning et al. 2013;Drivalos et al. 2016). According to previous reports, patients with elevated N-cadherin expression in breast and lung cancers have a lower survival rate than patients with low N-cadherin expression (Hui et al. 2013;Bock et al. 2014). Similarly, increased N-cadherin immunoreactivity was related to a severe CCA stage and a shorter patient survival time in our study. These findings suggest the N-cadherin could be used as the prognostic and targeted marker for CCA treatment. Accumulating evidence suggests that without curative operation, the effectiveness of standard chemotherapy for CCA treatment is not satisfactory (Okusaka et al. 2010). The development of effective therapeutic strategies in CCA is of considerable therapeutic benefit. Natural products are major sources for development of novel anti-cancer drugs due to their potentially low toxicity and potential effectiveness. Arctigenin is a bioactive compound which is isolated from Arctium lappa L., a popular medicinal herb usually used for anti-influenza treatment in Asia. Arctigenin has attracted the attention of researchers due to promising therapeutic effects including anti-inflammation, anti-infection, and anti-cancer (Gao et al. 2018). Therapeutic effects of arctigenin have been demonstrated in various cancers in vitro and in vivo, especially the role in EMT modulation (Yao et al. 2011, Wang et al. 2014Lu et al. 2019). In CT26 colorectal cancer, arctigenin significantly inhibited lung metastasis through increasing the expression of E-cadherin and decreasing the expressions of N-cadherin. In addition, arctigenin induced apoptosis in CT26 cells through the intrinsic apoptotic pathway (Han et al. 2016).
Results from our study demonstrated that arctigenin significantly suppressed CCA cell migration in a dose-dependent manner. Arctigenin repressed the expression of N-cadherin and induced a dramatic increase in the expression of E-cadherin in CCA cell lines. Similar results have also been observed in hepatocellular carcinoma, colorectal, and lung cancers (Han et al. 2016;Xu et al. 2017;Lu et al. 2019). This data supported that arctigenin exhibited an anti-EMT effect on CCA related to the regulation of N-cadherin/Ecadherin switching. Therefore, arctigenin could be used as anti-metastasis agent in CCA.
In addition, our study found that arctigenin plays a role in anti-proliferation and induced apoptosis in CCA cell lines. We performed cell proliferation assays after treatment with arctigenin. Our results show that arctigenin significantly inhibited the CCA cell viability up to 50%. This result was consistent with previous study which reported that arctigenin had an anti-proliferation role in breast cancer cells. MCF-7 human breast cancer cells were treated with 200 μM arctigenin and their viability was found to decrease by up to 50% (Maxwell et al. 2018). In addition, we performed flow cytometry to analyze whether the anti-proliferation effect of arctigenin was mediated via the apoptosis pathway. After being treated with arctigenin, CCA cells showed a significant increase in cellular apoptosis. Apoptosis is controlled by two major pathways including intrinsic and extrinsic pathways. The intrinsic pathway is controlled by the Bcl-2 family proteins (Cavalcante et al. 2019). The extrinsic pathway is induced by activation of the death cell receptor. The activation of the death cell receptor triggers the development of a death-inducing signaling complex, which includes the initiator caspase. There is cross-talking between two apoptotic pathways that have common apoptotic markers including Bid and caspase-3 (Wali et al. 2013). Our results showed that arctigenin treatment reduced Bcl-2 and pro-caspase-3 expression, whereas it induced Bax expression. Similar results were found by Li et al. who showed that arctigenin induces apoptosis in HT-29 colon cancer cells by caspase-3 activation (Li et al. 2016). Moreover, a significant increase in apoptosis marker as represented by the ratio of Bax to Bcl-2 protein was observed in prostate cancer by arctigenin treatments compared to control (Wang et al. 2017). Consequently, arctigenin is thought to cause CCA apoptosis through the Bax, Bcl-2, and caspase-3 pathways.
Considering that the toxicity of chemotherapy to normal cells is one of the major challenges of cancer chemotherapy. In our study, cytotoxicity of arctigenin on MMNK-1 which is the immortalized cholangiocyte cell line was performed. Our results demonstrated that whereas arctigenin is effective against CCA cells, its toxicity effect is not cancer cell specific. According to the literature, using of nanoparticulate drug carriers can solve many issues in drug delivery to the cancer cells such as improving drug solubility and stability, reducing adverse effects in non-target organs, and concentrating drugs at the cancer site (Khan and Gurav 2018). Our challenge is to conduct more research into the transport of arctigenin into specific bile duct cancers to resolve these issues. As a result, arctigenin could be used to improve CCA treatment outcomes as a natural product selection.

Conclusion
These studies showed for the first time that high N-cadherin expression was correlated with severe CCA stage and shorter patient survival times of CCA patient, indicating that N-cadherin could be used as a prognostic or targeted marker for CCA. Results also highlight the use of arctigenin, a seed isolated compound from Arctium lappa, to inhibit CCA progression. Our findings revealed that arctigenin inhibited metastasis in CCA cells via reversing of N-to E-cadherin switching. In addition, arctigenin suppressed cell viability and induced apoptosis through Bax/Bcl-2/caspase-3 pathway. This study suggests that arctigenin could be used as an alternative therapeutic agent to improve CCA treatment outcomes. Taken together, our findings suggest that arctigenin could be used as a new natural compound, either alone or in combination with chemotherapy drugs, to improve CCA treatment. However, a combined analysis of arctigenin and chemotherapy is needed to confirm the efficacy in both the in vitro and preclinical settings.