ARRB1 Drives Gallbladder Cancer Progression by Facilitating TAK1/MAPK Signaling Activation

Gallbladder carcinoma (GBC) is the most common malignancy of the biliary tract, with a dismal 5-year survival of 5%. Recently, ARRB1, as a molecular scaffold, has been proposed to participate in the progression of multiple malignancies. However, the effect and regulatory mechanisms of ARRB1 in GBC have not been investigated. Our study aimed to explore the biological functional status and the possible molecular mechanisms of ARRB1 with respect to GBC progression. The survey showed that human GBC tissues exhibited increased levels of ARRB1 compared with normal tissues, and the high expression of ARRB1 was associated with poor prognosis of GBC patients. A series of in vitro and in vivo functional experiments based on knockdown of ARRB1 uncovered that ARRB1 enhanced GBC cell proliferation, migration, and invasion. Furthermore, we reported that TAK1, a component of the TNF /MAPK pathway, is a vital downstream effector of ARRB1. In addition, siTAK1 could abolish the functional changes between ARRB1 overexpression GBC cells and control ones. Our data revealed that ARRB1 facilitated the carcinogenesis and development of GBC through TNF/TAK1/MAPK axis, suggesting that ARRB1 may be a promising biomarker and treatment target for GBC patients.


Introduction
Gallbladder carcinoma (GBC) is the fifth most common gastrointestinal malignancy worldwide, with extremely poor prognosis [1]. Radical cholecystectomy is still the main therapeutic schedule for GBC. However, only fewer than 10% of GBC patients are suitable for surgical excision due to the delayed diagnosis [2]. For those unresectable GBC patients, the approved chemotherapy is only a palliative therapy method with low efficacy [3,4]. The etiology of GBC is intricate and multifactorial. Various risk factors associated with the development of GBC include geographical distribution, genetic susceptibility, race, older age, gender, gallstone (size > 3 cm) and chronic inflammation [5,6]. In view of the few therapeutic options, high recurrence and inadequate understanding of disease mechanisms, it is necessary to further study the mechanisms of GBC pathogenesis, explore new sensitive molecular biomarkers and screen more effective treatment targets for prolonging the survival.
β-arrestin 1(ARRB1), as a well-known primary effector of the GPCR pathway, has been shown to promote several stages in the progression of different cancers, including leukaemia, breast cancer, lung cancer, colon cancer and laryngeal carcinoma [7]. As a molecular scaffold, ARRB1 could regulate cellular Ivyspring International Publisher function by interacting with other partner proteins. For ovarian cancer, ARRB1 is activated by ET-1R and cooperates with p300 to participate in the interaction between HIF-1α, which enhances the transcription of genes required for tumor cell invasion and angiogenesis [8]. Furthermore, ARRB1 is a potential prognostic biomarker for lung cancer and could predict the response to the chemotherapy of EGFR inhibitor [9]. Our previous study has shown that ARRB1 could participate in the aggressiveness of hepatocellular carcinoma through regulating CD97 [10]. However, the function of ARRB1 on the development and prognosis of GBC has not been verified so far.
In this study, we utilized clinical samples and GBC cell lines to reveal the effects of ARRB1 on the GBC biological behavior. Further mechanism study revealed that ARRB1 could mediated GBC progression in a TAK1-dependent manner. Our findings may aid in the development of novel diagnostic and therapeutic strategies targeting GBC.

Cell Lines and Cell Culture
The GBC cell lines NOZ and EH-GB1 were obtained from the Chinese Academy of Life Sciences (Shanghai, China). The cell lines SGC-996 and GBC-SD were kindly supplied by Xinhua Hospital (Shanghai, China). The SGC-996 cells were cultured in medium 1640 (Gibco, California), and NOZ, GBC-SD and EH-GB1 cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS, Gibco, USA) at 37 ℃ in humidified air containing 5% carbon dioxide (CO2).

Patient Samples
Surgical GBC samples and adjacent normal specimens were obtained from 62 patients at the Affiliated Changzhou NO.2 People's Hospital of Nanjing Medical University (Jiangsu, China) from January 2014 to December 2018. All the specimens were paraffin-embedded for immunohistochemical staining or extracted total protein for western blot. This study was approved by the Institutional Ethics Committee of Nanjing Medical University and informed consent was signed before surgery. All specimens were confirmed by pathologists and the pathological diagnosis was established according to the American Joint Committee on Cancer criteria (2002).

Immunohistochemistry
Immunohistochemistry of ARRB1 and Ki67 was performed using 5 µm paraffin sections. The sections were deparaffined in xylene and rehydrated. Antigen retrieval was performed by heating sections with citrate buffer, then removed endogenous peroxidase with 3% H2O2. Slides were washed with TBS-Triton 3 times. Non-specific binding was blocked using 5% goat serum for 30 minutes. Then sections were incubated with specific antibody (1:100). Complete the experiment and following DAB staining was done according to the instructions (KIT9730 IHC Kit, MXB, China). The scores of the staining were indexed on the intensity and percentage of the immunoreactive cells, and classified into three groups (Strong, Medium, Weak / Negative) [11]. For statistical analysis, the strong group was defined as high group (n=27), and the others were defined as low group (n=35).

Knockdown or overexpression of ARRB1 expression in GBC cell lines
Lentiviral vectors (labeled with luciferase) containing human ARRB1 short-hairpin RNA (sh-ARRB1) and negative control (sh-NC) were a gift from Professor Beicheng Sun (Nanjing University Medical School, China). Sh-ARRB1 and sh-NC sequences are presented in Supplementary Table S1. Transfected NOZ cells were verified with puromycin and RT-PCR. For overexpression of ARRB1, sequence of ARRB1 were subcloned into the lentiviral vector lentivirus vector GV248 (Corues Biotechnology, Nanjing). Recombinant lentiviruses co-transfected into 293T cells with packaging plasmids were harvested, and utilized to infect GBC-SD cells after mixing with polybrene (10 μg/ml, Sigma). Stable cells were verified by Western blotting.
Quantification was performed by comparing the Ct values of each sample with the 2 -ΔΔCt method and normalization to β-actin. Values were expressed as fold induction in comparison with controls.

Colony formation assay
The GBC cells were separately seeded and incubated in 6-well plastic plates with 500 cells for 14 days, then stained in paraformaldehyde and crystal violet successively. This assay was repeated three times. The number of clone colonies was counted and compared.

5-Ethynyl-2′-deoxyuridine (EdU) incorporation assay
The EdU assay was performed with the EdU detection kit (Ribo Biotechnology, Guangzhou, China) according to the manufacturer's instructions. Briefly, the cells were seeded into 96-well plates (1~5 × 10 4 per well) and cultured overnight at 37℃ with 5% CO 2 . After 4 h of incubation in EdU medium, the cells were fixed and permeabilized, then incubated with reaction mixture, the percentage of EdU staining cells were imaged and counted under the fluorescence microscopy.

Cell Apoptosis and Cell Cycle
Cell apoptosis was measured with the Annexin V-APC/7-AAD apoptosis kit according to the manufacturer's recommendation (Vazyme, Nanjing, China). For cell cycle analysis, the transfected NOZ and GBC-SD cells were stained with propidium iodide for 30 min, then analyzed through flow cytometry.

Wound healing experiment
The stable cells in the logarithmic growth phase were seeded in a 6-well plate. When the cells reached almost 90% fusion, vertical wound was scratched with the head of pipette, and lightly washed with PBS, then added medium containing with 2.5% FBS. The scratch width was recorded. After cultured for 24 hours, the condition of scratch healing was measured again.

Transwell assay
Matrigel (BD Biosciences, USA) were precoated on the upper chamber of 24-well transwells (8-μm, Millipore, USA). GBC cells (8 × 10 4 ) were suspended in serum-free medium and infused into the upper chamber. Medium supplemented with 20% FBS was applied to the lower chamber. After 24 h at 37℃ with 5% CO 2 , cells maintained in the lower surface were fixed with 4% methanol and dyed in 0.5% crystal violet. Transwell assays were independently repeated three times. The visual fields were photographed and counted under microscope.

Nude mouse xenograft experiments
Specific pathogen-free 6-week-old female BALB/c nude mice were obtained from Model Animal Research Center of Yangzhou University and housed in the standard conditions. The NOZ cells (1 × 10 6 / per) were subcutaneously injected into the groin of the mice (n = 6). The tumor size was measured every 5 days using a vernier caliper (tumor volume = 0.5 × width 2 × length). Four weeks later, all nude mice were euthanized and the tumor specimens were exteriorized and measured. Furthermore, for the tail vein metastasis model, NOZ cells (1 ×10 6 cells/100 μl) were injected into the tail vein of mouse under anesthesia. Five weeks later, after humanly killed, the lung and liver metastatic samples were sectioned and stained with H&E to assess the extent of metastasis.

Statistical Analysis
GraphPad Prism 6 and SPSS software were used for plotting and statistical analysis. All data was expressed as mean ± standard error of the mean. Student t-test was applied to calculate statistical significance for data sets following normal distribution. Kaplan-Meier curve was used for survival analysis. P-values less than 0.05 were considered statistically significant.  The data in this table is analyzed using chi-square test. * indicates P value < 0.05

Upregulated Immunohistochemical Staining of ARRB1 Proteins in GBC Tissues
Assessed with immunohistochemical staining, the expression of ARRB1 tended to be higher in GBC tissues than that in cholelithiasis specimens, and ARRB1 immunoreactivity was mainly observed in the cytoplasm of tumor cells (Figure 1 A). Among the 62 GBC tissue samples, 43.55% (27/62) of cases were strong stained, 22.58% (14/62) of cases were medium stained, 33.87% (21/62) of cases were weak or negatively stained, the opposite of that only 6.67% (2/30) of the cholecystitis tissues showed strong staining of ARRB1 protein (Figure 1 B). The western blot data showed that the protein level of ARRB1 was obviously upregulated in GBC tissues compared with the nontumor counterparts and cholecystitis tissues (Figure 1 C).

Increased ARRB1 Expression Correlates with Aggressive Clinicopathologic Characteristics and Poor Prognosis in GBC Patients
To investigate the clinical consequences of ARRB1 expression, we divided 62 GBC patients into two categories (high ARRB1 expression and low ARRB1 expression), and analyzed the correlation between ARRB1 overexpression and the pathological features as well as the disease progression of GBC. As summarized in Table 1, high levels of ARRB1 expression were significantly correlated with tumor size (P = 0.015) and lymphatic metastasis (P = 0.009). Importantly, with regard to overall survival (OS), high ARRB1 expression was correlated with worse OS rate (Figure 1 D). These findings suggested that the upregulated ARRB1 was significantly related with poor prognosis of GBC.

Knockdown of ARRB1 regulates proliferation, migration, and invasion of GBC cell in vitro
To further elucidate the biological role of ARRB1 in GBC, we first detected the expression of ARRB1 in GBC cell lines and found that ARRB1 was higher expressed in NOZ and lower expressed in GBC-SD cell lines compared with others respectively (Figure 1  E, F). As NOZ cells showed higher ARRB1 expression, we transfected them with lentiviral shRNAs (shRNA-1, shRNA-2 and shRNA-3) specially against ARRB1 and the negative control (sh-NC). The interference efficiency of shRNAs were tested by qRT-PCR and western blot (Figure 2 A, B). Because of more effective suppression, shRNA-2 was selected for further experiments.    Figure S1 B). Secondly, to investigate the role of ARRB1 on GBC cell motility, we performed wound healing and transwell migration assays, and found that loss of ARRB1 expression significantly suppressed random motility, migration, and invasion of NOZ cells (Figure 2 G, H). It also significantly altered the level of the related apoptosis protein (Bcl-2), as well as the epithelial-mesenchymal transition (EMT) proteins (Vimentin, E-cadherin and N-cadherin) (Figure 2 I). Collectively, these results strongly implicated that ARRB1 was an oncogenic gene in GBC cells.

ARRB1 promotes tumor progression in vivo
The results above suggested that ARRB1 plays important roles in GBC tumor progression in vitro. To extend our investigations in vivo, NOZ cells were injected into BALB/c nude mice to construct the xenograft model. Tumors expressing normal level of ARRB1 grew rapider than NOZ-shARRB1 ones. After 28 days, we observed the tumor size distinctly diminished on account of ARRB1 downregulation (Figure 3 A, B). Confirmed with in vitro experiments, lower Ki-67 expression was tested in shARRB1 group than that in sh-NC group by IHC (Figure 3  The metastatic model of the nude mice was established by the tail intravenous injection of NOZ cells. The number and size of liver metastatic nodules per nude mice was significantly declined in ARRB1 knockdown group compared with control group (Figure 3 E, F). Furthermore, the pulmonary metastasis lesions in group injected with NOZ-shARRB1 cells were smaller and fewer assessed both by visual inspection and microscope (Figure 3 F,  G). All observations of in vivo experiment distinctly suggested that ARRB1 could promote GBC tumorigenesis and metastasis.

ARRB1 silencing restrains the expression of TAK1 and inhibits MAPK pathway in GBC
Multiple past studies have shown that ARRB1 regulates specific cellular functions by interacting with specific partner proteins, including PI3K and MAPK pathway [7]. To explore the downstream regulatory mechanisms by which ARRB1 acts in GBC, we exploited RNA sequencing to compare the transcriptome of sh-NC and sh-ARRB1 groups. Many genes were differentially expressed when ARRB1 was depleted ( Figure S1 E). Further KEGG analysis identified that the most significantly regulated pathway in the ARRB1 knockdown GBC cells was TNF signaling (Figure 4 A). Recent work has demonstrated that TNF-α could directly induce hepatic ARRB1 expression and enhance hepatocellular carcinogenesis via ARRB1-AKT interaction by binding to boost Akt phosphorylation [13]. Additionally, in intestinal epithelial cells of colitis, TNF-α/ARRB1-dependent signaling in hematopoietic and non-hematopoietic cells differentially regulates colitis pathogenesis in modulating MAPK pathways [14]. Preliminary literatures combined with our results indicated that ARRB1 may be a critical TNF signaling regulator in GBC. To verify the hypothesis, we performed RT-qPCR to confirm the expression of these candidate genes with a change of greater than 2.5-fold involved in TNF signaling. The results showed that shRNA-mediated ARRB1 could regulated the expression levels of TAK1 (MAP3K7), TRAF3 and NIK ( Figure S1 F, G). Notably, TAK1 was the most strongly downregulated gene when ARRB1 was knocked down in GBC cells. Subsequently, western blotting and immunofluorescence (IF) staining both revealed that TAK1 protein level was dramatically repressed after ARRB1 depletion (Figure  4 B, C). Synchronously, same tendency was found in the phosphorylation of Erk, p38 and JNK (Figure 4 B). After treated with siRNA-TAK1(sc-36606), the expression of TAK1 in NOZ and GBC-SD cells reduced, while the ARRB1 level was not affected by this treatment (Figure 4 D). These findings suggested that ARRB1 positively coordinates the level of TAK1 to activate the TNF/MAPK signaling pathway.

ARRB1 controls GBC biological functions partly through TAK1
To investigate whether TAK1 is indeed required for the biological effects of ARRB1 depletion in GBC, we designed the following rescue experiments. First, Lv-ARRB1 lentivirus was used to upregulate ARRB1 expression in GBC-SD cell line, which exhibited lowexpression level of ARRB1. The raised expression of ARRB1 was verified by western blot detection, and the expression of TAK1 increased accordingly ( Figure  5 A).
Functionally, ectopic expression of ARRB1 in GBC-SD cells evidently promoted cell viability and proliferation, which could be partially rescued by TAK1 suppression (Figure 5 B, C). As for apoptosis, the percentage of apoptosis showed significant differences in ARRB1 overexpressed GBC cells in comparison with the Lv-NC or adding si-TAK1 synchronously (

Discussion
Gallbladder carcinoma is a highly lethal and aggressive disease, and the therapeutic outcome of non-operative treatment for GBC is not satisfactory. The poor prognosis of this disease is also due to early metastasis and delayed diagnosis, as the 5-year survival rate is less than 6% [15]. Therefore, it is urgent to increase our understanding of the pathogenesis and molecular mechanism of GBC for screening out more effective therapeutic target.
ARRB1 is a member of β-arrestin family, mostly recruited associated with signaling desensitization of G-protein coupled receptors (GPCR), to achieve spatiotemporal specificity of different signaling complexes [16,17]. ARRB1 could act as cytosolic, nuclear scaffold or signal transducer, controlling in multifaceted signaling processes, such as cell proliferation, metastasis and drug resistance [18]. For ovarian cancer, the recruitment of ARRB1 showed the remarkable ability as a checkpoint converging pathway on β-catenin signaling to promote the invasion and metastasis of ovarian cancer cells [19]. In glioblastoma, knockdown of ARRB1 decreases cell viability, metastasis and glycolysis by suppressing Src signaling [20]. Our data firstly showed that ARRB1 expression was upregulated in human GBC tissues compared with normal gallbladder epithelium tissues. In addition, the related poor clinical characteristics suggested that ARRB1 assuredly played an important role in progression of GBC. Through a series of in vitro and in vivo assays, we confirmed that inhibition of ARRB1 restrained the GBC cell proliferation, metastasis and tumor growth. Moreover, RNA sequencing and our following research showed that ARRB1 deletion decreased the activation of TNF/MAPK signaling pathway in GBC. TNF-α is a key inflammatory cytokine responding to chronic inflammation which is a major carcinogenic mechanism of gallbladder cancer [21]. The promote relationships between TNF-α and Vascular Endothelial Growth Factor-C (VEGF-C), ARRB1 and VEGF-C has been confirmed in many diseases including pulmonary hypertension, colitis and gallbladder cancer [22][23][24]. To further clarify the downstream signaling mechanism involved by ARRB1 in TNF-α/MAPK, we used ARRB1overexpressed GBC-SD cells and siTAK1 to further demonstrate that ARRB1 positively regulated the TAK1/MAPK signaling pathway. TAK1, thought to mediate much of the intracellular actions of TNF-α, is closely involved in inflammation-related diseases [25]. Our study complemented the inflammationcarcinogenic mechanism of ARRB1/TAK1, and firstly verified the important role of ARRB1/TAK1 in the regulation of proliferation, apoptosis and metastasis in GBC.
In this study, we conclude that ARRB1 regulates GBC progression by modulating the TAK1/MAPK pathway. However, emerging data shows that ARRB1 may be driven by many non-GPCR pathways, such as endothelin-1 TGF-β and VEGFR [26,27]. The carcinogenic mechanism of inflammation linked to TGF-β or TNF-α and its regulation of ARRB1 in GBC or other cancers still needs to be fully elucidated. According to the results of RNA sequencing, we focused on and verified the ARRB1/TAK1/MAPK signal path. Interestingly, in portal hypertensive gastropathy, ARRB1 was confirmed to regulated ER stress-induced mucosal epithelial apoptosis through suppression of the TNF-α/p65 signaling pathway activation [28]. Synchronously, ARRB1 also mediated the development of nonalcoholic fatty liver disease through TRAFs, which is another important member of TNF-α pathway [29]. The specific mechanism of ARRB1 participation in the downstream pathway of TNF-α/MAPK and whether it forms a network system with other pathways, such as PI3K and NF-κB, are worthy of further study. Focus on clinical translational applications; we have noticed that the TAK1 inhibition-NG25 could significantly inhibit tumor cell growth in different cancers [30,31]. Combined with our results, it is of great research value to exploit ARRB1 or its downstream TAK1 as a therapeutic target for GBC, we speculate that inhibition of ARRB1 or TAK1 with specific molecule compounds may be effective for GBC treatment.
Altogether, our study firstly investigated that ARRB1 was related to the poor prognosis in GBC, and functioned as an oncogenic gene. Knockdown of ARRB1 restrained the cell proliferation by promoting apoptosis, and inhibit the migrative and invasive ability of GBC cells in vitro and in vivo. Moreover, the tumor-promoting effect of overexpression ARRB1 may be partially related the activations of TAK1/MAPK axis. Therefore, ARRB1 may serve as a potential target for cancer diagnosis of GBC and related therapies.