Fibrinogen promotes gallbladder cancer cell metastasis and extravasation by inducing ICAM1 expression

Fibrinogen plays an important role in tumor progression. Here, we explored the role of fibrinogen in gallbladder cancer (GBC) metastasis. The plasma fibrinogen level in M1 GBC patients was higher than in M0 GBC patients, indicating that fibrinogen may participate in GBC metastasis. Treatment of GBC cell lines with fibrinogen promoted metastasis and induced the expression of intercellular adhesion molecule 1 (ICAM1). ICAM1 overexpression promoted metastasis and knockdown inhibited it. The cell adhesion and transendothelial migration of GBC cells were enhanced by fibrinogen treatment and ICAM1 overexpression. In addition, the medium of fibrinogen-treated and overexpression-ICAM1 NOZ cells exhibited enhanced macrophages recruitment. This may work in concert to promote angiogenesis. Immunohistochemistry results on clinical specimens showed that higher fibrinogen levels, higher ICAM1 expression, higher blood vessel density, and higher macrophage levels were present simultaneously. Collectively, this study indicates fibrinogen promotes metastasis and extravasation by inducing ICAM1 expression to enhance tumor cell migration, cell adhesion, transendothelial migration and promote angiogenesis and increase vascular endothelial permeability.


Introduction
Gallbladder cancer, the most common biliary tract malignancy, is highly aggressive [1,2]. The prognosis of GBC is very poor with a 5-year survival rate only ranking to 5-15%, and there is no improvement in last 15 years [3][4][5]. Radical cholecystectomy is the most effective treatment for GBC [6]. However, the majority of patients are diagnosed at advanced stage losing the opportunity of radical surgery, so that they have to be treated with conservative interventions including radiotherapy, chemotherapy, and targeted therapy [7][8][9]. A series of genes associated with GBC have been identified, and drugs have been developed accordingly, but the high rate of liver metastasis at first diagnosis, and the poor prognosis indicate a lack of sufficient knowledge about the metastatic mechanism [2,[10][11][12]. Therefore, it is critical to identify the novel biomarker and the therapeutic target for the treatment of GBC.
Fibrinogen is biosynthesized by hepatocytes and secreted into the blood [13]. During hemostasis, it is converted to fibrin by thrombin, and recruits various cells [14,15]. In addition to its role in blood coagulation, fibrinogen is involved in a variety of diseases such as cardiovascular diseases, cancers, neurological disorders, and amyloidosis [16]. Higher plasma fibrinogen levels are associated with a worse prognosis in esophageal, lung, renal, and gallbladder cancers [17][18][19]. Accumulating evidences have demonstrated that fibrinogen plays an important role in tumor progression [20][21][22]. In hematogenous pulmonary metastasis, fibrinogen facilitates the stable adhesion and survival of tumor cells after intravenous injection [23]. The platelet-fibrinogen axis impedes the nature killer cell elimination of tumor cells to increase metastatic potential [24]. Fibrinogen has no effect on tumor growth generally but supports both tumor growth and metastasis in colon cancer [25,26]. In colon cancer model, fibrinogen plays an important role in tumor growth and dissemination. Tumors harvested from fibrinogen-deficient mice show slow cell proliferation, increased tumor necrosis, and decreased tumor vascular density [25]. Fibrinogen deficiency results in a significant decrease in inflammation-driven adenoma formation following azoxymethane/ dextran sodium sulfate challenge [26]. Fibrinogen also activates focal adhesion kinase, inhibits p53 and its downstream targets, thereby promoting colon tumor cellular proliferation and preventing senescence [27]. Hence, the mechanism of fibrinogen involved in cancer development and progression is unclear. In our study, we revealed the facilitative effects of fibrinogen in GBC metastasis in vitro and in vivo.

Patient's data
The cohort was derived from the Chinese Research Group of Gallbladder Cancer (CRGGC) study, a national multicenter retrospective cohort of patients with GBC. The protocol of the CRGGC was approved by the Committee for Ethics of Xinhua Hospital, Shanghai Jiao Tong University School of Medicine and registered on ClinicalTrials.gov. We included patients with M0 stage GBC according to following criteria: (1) with primary malignancy of gallbladder confirmed by histological studies; (2) who underwent radical cholecystectomy; (3) with no distant metastasis; (4) with histologically confirmed T stages of T1b-T4 (excluding Tis and T1a); (5) with available follow-up data; and (6) with available preoperative plasma fibrinogen data. The flowchart of the patients with GBC from the CRGGC study from 15 hospitals in 10 provinces operated between 2010 and 2017 was included in this study (Fig. S1).

Statistical analysis
All data are shown as mean ± standard deviation (SD). The χ 2 test was used to compare the distributions of categorical variables, t test or the Wilcoxon rank sum test was used for the distributions of continuous variables, and the log-rank test was used for survival data.
All statistical analyses were performed using R version 3.6.2 (R Foundation for Statistical Computing, Vienna, Austria). Differences with a two-sided P < 0.05 were considered statistically significant.

Cell proliferation assay
The CCK-8 assay and colony formation assay were used to detect cell proliferation. For CCK-8 assay, cells (1 × 10 3 ) were seeded into 96-well plates. The CCK-8 reagent (Yeasen, China) was added to each well. After 2 h incubation, the absorbance at 450 nm was measured by a microplate reader (Bio-TEK, Saxony, USA). For colony formation assay, 500 cells were seeded into six-well plates and cultured for 2 weeks. The colonies were fixed by 4% paraformaldehyde and stained by crystal violet.

Cell migration assay
The migration ability of GBC cells was evaluated by using wound-healing assay and Transwell system (Corning, USA). For wound-healing assay, we scratched the well of six-well plates with 70-80% confluence by using a 200-μL pipette tip. After scratching, it was monitored by microscope at 0 h, 24 h, and 48 h. For Transwell system, NOZ (2 × 10 4 ) and OCUG-1 (4 × 10 4 ) cells were seeded in the upper chamber containing 200 μL serum-free medium, respectively. The low chambers were added 600 μL medium with 10% FBS. For macrophage migration assay, THP-1(10 × 10 4 ) cells were first seeded in the upper chamber and the medium with 10% FBS, and 100 ng/mL PMA was added in upper chamber and low chamber. After 24-48 h culture, the THP-1 were induced polarization into M0 phase. Subsequently, the medium in upper chambers was replaced with 200 μL serum-free medium, and the medium in low chambers was replaced with 600 μL conditional medium. After 24 h incubation, the inserts were fixed in paraformaldehyde and then stained with crystal violet. The cells on surface of the inserts were removed by cotton swab. Finally, we randomly selected five fields to photograph.

Quantitative real-time PCR
Total cellular RNA was extracted by TRIzol reagent (Invitrogen, USA). A reverse transcription reagent kit (TaKaRa, Dalian, China) was used to synthesize cDNA. The relative gene expression was calculated by the 2 −ΔΔT method. GAPDH was used as internal control. The gene primers are listed in supplement Table S1.

Western blot analysis
Total proteins were obtained from cells by using RIPA lysis buffer (Beyotime Biotechnology, Shanghai, China) containing protease inhibitor cocktail. Bicinchoninic acid (BCA) protein assay kit (Beyotime Biotechnology, Shanghai, China) was used to quantify the concentration of proteins. Subsequently, proteins were separated by SDS-PAGE and then transferred onto 0.45 μm PVDF membranes (Millipore). The membranes were blocked by 5% skim milk in TBST buffer. After that, they were incubated with a series of primary antibodies. ICAM1: A19300, ABclonal; β-actin: AC038, ABclonal. After TBST buffer washing, the membranes were incubated with secondary antibody. The enhanced chemiluminescent detection reagent (Rockford, IL, USA) was used to determine target protein. β-actin was used as internal loading control.

Angiogenesis assay
The growth factor-reduced Matrigel (Corning, USA) was added into 96-welll plate and polymerized at room temperature for 1 h. The HUVECs (1 × 10 4 ) with serum starvation for 6 h were resuspended on different media and seeded into the labeled wells. After incubated at incubator for 6 h, the images of the capillary network were taken by a microscope.

Transendothelial migration assay
The HUVECs were seeded in the upper insert and formed a confluent monolayer. After that, tumor cells were added to the upper insert with serum-free medium. The low chambers were added 600 μL medium with 10%FBS. After 24 h incubation, we randomly selected five fields to photograph by a fluorescence microscope. Tumor cells were labeled with GFP. Different tumor cell groups include fibrinogen treatment, overexpression, and knockdown of ICAM1. HUVECs were treated with fibrinogen when performing vascular permeability assays.

RNA-seq analysis
The total RNA was exacted with TRIzol reagent, and then sequencing libraries were generated using NEBNext® UltraTM RNA Library Prep Kit for Illumina® (NEB, USA). The libraries were subjected to sequence with Illumina Novaseq platform. Gene Ontology (GO) enrichment analysis of differentially expressed genes was implemented by the clus-terProfiler R package, in which gene length bias was corrected.

Immunofluorescence assay
The 12-well plates were previously laid with sterile cover glasses. NOZ cells were seeded into the plates and incubated for 24 h. The cells were fixed by 4% paraformaldehyde and then permeabilized in 0.2% Triton X-100 at room temperature. After blocked by 0.1% bovine serum albumin (BSA), they were incubated with FGA (ab92572, abcam), FGB (ab189490, abcam), and FGG (ab119948, abcam) at 4 °C overnight. The cells were followed by incubation with secondary antibody Alexa Fluor 488 or Alexa Fluor 594 conjugated secondary antibody (Yeasen Biotech, Shanghai, China). Finally, DAPI was used to counterstain the cells, and Leica microscope was used to image the cells.

Single-cell RNA-sequencing data analysis
The single-cell RNA-sequencing data and analysis method were based on our previous study [28].

In vivo experiments
For xenografted model, cells (1 × 10 6 ) were resuspended with 100 μL PBS and then subcutaneously into BALB/C nude mice. Five days later, the tumor volumes were measured every 5 days and calculated by the formula: Volume (mm 3 ) = 0.5 × length (mm) × width (mm 2 ). After 25 days, the mice were sacrificed and tumors were harvested and weighed. For intrasplenic injection model, cells (2 × 10 6 ) which expressed luciferase were injected into spleen. Splenectomy was performed 5 min later after cell injection. Four weeks later, the metastases were monitored by using the IVIS@ lumine II system after intraperitoneal injection of luciferin. All animal studies were approved by the Xinhua Hospital Ethics Committee Affiliated to Shanghai Jiao Tong University School of Medicine.

Clinal specimens
The human specimens were obtained from biobank of Shanghai Key Laboratory of Biliary Tract Disease Research. All participants of this study were approved by the Ethics Committees of Xinhua Hospital.

Patients with M1 stage GBC had higher fibrinogen levels
Patients with GBC from the CRGGC study who were from 15 hospitals in 10 provinces diagnosed between 2010 and 2017 were involved in this study. A total of 871 patients with GBC who underwent radical cholecystectomy and 460 patients with M1 stage GBC at diagnosis were identified in the same registry.  (Table 1). Metastatic GBC patients had higher average fibrinogen levels than patients with GBC who underwent radical cholecystectomy (Fig. 1).

Exogenous fibrinogen promoted GBC metastasis
To explore the reasons for higher average fibrinogen levels observed in metastatic GBC patients, we investigated the role of fibrinogen in GBC metastasis. To evaluate the effect of fibrinogen on the proliferation and migration of GBC cells, we added exogenous fibrinogen to medium to culture NOZ and OCUG-1 cell lines. The results of the CCK-8 assay and clone formation assay showed that fibrinogen had no effect on GBC cell proliferation ( Fig. 2A). However, the results of the wound-heal assay and transwell assay showed that the migration ability of fibrinogen-treated GBC cells was enhanced (Fig. 2B). Furthermore, we chose two mouse models, a subcutaneous xenograft model and an intrasplenic injection model, to investigate the role of fibrinogen in vivo. In the subcutaneous xenograft model, we injected individual NOZ cells with different fibrinogen treatments. After 25 days, there were no significant differences in the tumor growth curves, volume, or weight between the two groups (Fig. 2C). In the intrasplenic injection model, liver metastasis was more extensive in the experimental group, which was injected with fibrinogen-treated NOZ cells (Fig. 2D). The in vitro and in vivo results consistently showed that fibrinogen promoted GBC cells metastasis, but not proliferation.

Fibrinogen-induced ICAM1 expression promoted metastasis
To investigate the mechanisms underlying the fibrinogenmediated promotion of GBC cell metastasis, we verified the expression levels of fibrinogen receptors in the NOZ cell line. ICAM1 was highly expressed in NOZ cells treated with fibrinogen, whereas the expression of other receptors did not change (Fig. 3A). The role of ICAM1 was also investigated. Wild-type NOZ cells exhibited a low expression level of ICAM1, and OCUG-1 cell exhibited a high level (Fig. 3B). Thus, we established a stable overexpression-ICAM1 (OE-ICAM1) NOZ cell line and a knockdown-ICAM1 (sh-ICAM1) OCUG-1 cell line (Fig. 3C). Overexpression and knockdown of ICAM1 did not affect cell proliferation in vitro or in vivo (Fig. 3D−E). Conversely, the migration ability of GBC cells affected; overexpression of ICAM1 promoted migration, and knockdown inhibited it (Fig. 3F). On the other hand, overexpression of ICAM1 also promoted liver metastasis in vivo (Fig. 3G).

Fibrinogen and ICAM1 facilitated GBC extravasation and angiogenesis
Based on previous studies indicating an intricate relationship among fibrinogen, cell adhesion, and angiogenesis, we hypothesized that fibrinogen enhanced the adhesion capacity of GBC cells and promoted angiogenesis [16]. Moreover, the RNA-seq data showed that cell adhesion and angiogenesis were enriched in OE-ICAM1 NOZ cells compared with OE-Control NOZ cells (Fig. 4A). Both fibrinogen treatment and ICAM1 overexpression enhanced cell adhesion, whereas ICAM1 knockdown suppressed it (Fig. 4B). In addition, we used a transendothelial migration model to simulate extravasation. Fibrinogen-treated NOZ cells and OE-ICAM1 NOZ cells were more likely to cross HUVEC monolayer, while sh-ICAM1 OCUG-1 cells were less, showing a higher potential for extravasation (Fig. 4C). In addition, the number of migrated wild-type NOZ cells increased when the HUVEC monolayer was replaced with a fibrinogen-treated monolayer (Fig. S2). We hypothesized that fibrinogen might facilitate the extravasation of tumor cells by enhancing the transendothelial migration ability and vascular permeability. However, the results of the angiogenesis experiments differed from our expectation. We used a co-culture system for these experiments. The conditional medium that cultured cells with fibrinogen added in medium or cultured OE-ICAM1 NOZ cells promoted angiogenesis slightly (Fig. S3). Nevertheless, the medium containing fibrinogen or ICAM1 dramatically promoted angiogenesis (Fig. S3), which was in accordance with previous studies [29,30]. To elucidate the enrichment of angiogenesis, we performed a macrophage recruitment assay. As expected, more macrophages were recruited by conditional medium which were cultured cells with fibrinogen or cultured OE-ICAM1 NOZ cells (Fig. 4D). Consistently, in specimens with higher plasma fibrinogen levels, we observed higher ICAM1 expression, concomitant with higher blood vessel density, and higher macrophage levels (Fig. 4E).

Discussion
Gallbladder cancer is a highly malignant tumor commonly diagnosed at an advanced stage, with a high incidence of hepatic metastasis [31]. Although numerous studies have investigated the causes of GBC metastasis, the underlying mechanisms remain poorly understood [10,11,[32][33][34]. Therefore, a comprehensive understanding of the mechanism of GBC metastasis is valuable for prognosis prediction and the development of therapy. This study focused on the relationship between fibrinogen and GBC metastasis. Previous studies have shown that higher fibrinogen levels were associated with a worse prognosis in digestive tumors as well as GBC [18,19]. In our study, the patients with M1 stage GBC had higher fibrinogen levels than patients with GBC who underwent radical cholecystectomy. Therefore, fibrinogen may promote GBC metastasis. Owing to the low incidence of GBC, studies with large sample sizes are scarce. Additionally, the enrollment of metastatic GBC patients might be less representative of the whole patient population, because most patients with M1stage GBC lack fibrinogen data.
We also investigated the mechanism by which fibrinogen promotes GBC metastasis. Fibrinogen is closely associated with tumor metastasis [16]. For example, in a fibrinogen-deficient mouse model, metastasis was diminished after the injection of tumor cells, but growth and angiogenesis were unaffected [23]. Similarly, GBC cells cultured with fibrinogen exhibited enhanced metastasis ability, with no variation in their proliferation ability. On the other hand, GBC cell-to-HUVEC adhesion was enhanced by fibrinogen, contributing to metastasis. Transendothelial migration and angiogenesis during metastasis were also simulated, and the transendothelial migration ability of GBC cells treated with fibrinogen was enhanced. However, there were no significant differences in angiogenesis in the co-culture system. These phenomena may have resulted from the fibrinogen depletion or ICAM1 deficiency in the in vitro model. In contrast, both fibrinogen and ICAM1 can promote angiogenesis independently. Tumor angiogenesis is facilitated via macrophage recruitment [35,36]. The coculture medium effectively recruited macrophages, demonstrating a potential mechanism of angiogenesis. Furthermore, vascular endothelial permeability was increased by fibrinogen in vitro, consistent with previous study [37][38][39]. We found that ICAM1 expression was induced by exogenous fibrinogen. Higher ICAM1 expression is associated with great aggression and a worse prognosis Fig. 4 Fibrinogen and ICAM1 facilitated GBC extravasation and angiogenesis. A The GO enrichment showed that adheres junction and blood vessel morphogenesis were significant. B The images of NOZ cells with different fibrinogen treatments, OE-Control NOZ cells, OE-ICAM1 NOZ cells, sh-Control OCUG-1 cells, and sh-ICAM1 OCUG-1 cells adhered HUVEC monolayer. C The images of NOZ cells with different fibrinogen treatments, OE-Control NOZ cells, OE-ICAM1 NOZ cells, sh-Control OCUG-1 cells, and sh-ICAM1 OCUG-1 cells crossed HUVEC monolayer. D The images of macrophage recruitment after co-culture with conditional medium (CM). CM 1: the medium that cultured NOZ after 2 days; CM 2: the medium that cultured NOZ cells with 50 μg/mL FIB after 2 days; CM 3: the medium that cultured NOZ cells with 100 μg/mL FIB after 2 days; CM 4: the medium that cultured OE-Control NOZ cells after 2 days; CM 5: the medium that cultured OE-ICAM1 NOZ cells after 2 days. Fibrinogen was added separately at a concentration of 100 μg/ mL. ICAM1 was added separately at a concentration of 1 μg/mL. E Representative images of HE, ICAM, CD31, and CD68 IHC staining in GBC tissues with different preoperative plasma fibrinogen levels ◂ in many tumors [30]. The metastatic ability, including migration, cell adhesion, and transendothelial migration, was enhanced in OE-ICAM1 NOZ cell line. Moreover, the conditional medium which cultured OE-ICAM1 NOZ cells exhibited enhanced macrophage recruitment ability. These results indicated that fibrinogen and ICAM1 play important roles in GBC metastasis. Therefore, they may serve as new targets for GBC treatment. Considering that the treatment researches about fibrinogen and ICAM1 is being carried out, this is very helpful for the future treatment of GBC [40,41].
Our study systematically investigated the role of fibrinogen in GBC metastasis. Fibrinogen promotes GBC cell metastasis and extravasation by inducing ICAM1 expression to enhance tumor cell migration, cell adhesion, transendothelial migration and promote angiogenesis and increase vascular endothelial permeability. Previously, several hypotheses have been proposed for the molecular mechanism of fibrinogen in tumorigenesis and progression. First, the fibrinogen influences cell proliferation, metastasis, and angiogenesis through interaction with growth factors [20,42]. Second, the fibrinogen binds to several cell types to promote adhesion, motility, and invasion [20,43]. Finally, fibrinogen interaction with platelets facilitates the protection of tumor cells from nature killer cytotoxicity [24].Therefore, the effect of fibrinogen in GBC metastasis has been partly studied in other cancers or diseases.
This study had the following limitations. (1) Due to the extensive application of in vitro cell models in this study, relationships among fibrinogen, tumor cells, vascular endothelial cells, and macrophages in tumor microenvironments were unclear. (2) The mechanism by which ICAM1 participates in metastasis remains unclear. It has been reported that ICAM1 would be induced by IL35 and promoted the extravasation and metastasis of pancreatic ductal adenocarcinoma cell [44]. However, in this study, ICAM1 was not blocked to verify whether the phenomena occur in GBC. (3) During metastasis, fibrinogen protects circulating tumor cells (CTCs) from natural killer cells, whereas ICAM1 contributes to the formation of CTCs [24,45]. The role of fibrinogen and ICAM1 in blood circulation was not studied. (4) Many different types of cells can synthesize fibrinogen [42,46]. We investigated the expression levels of related genes and found that GBC cells could synthesize fibrinogen slightly (Fig. S4). However, endogenous fibrinogen, which may participate in GBC biological behavior, was not evaluated.