Down‐regulation of interferon regulatory factor 2 binding protein 2 suppresses gastric cancer progression by negatively regulating connective tissue growth factor

Abstract Interferon regulatory factor 2 binding protein 2 (IRF2BP2) is a transcriptional repressor involved in regulating gene expression and other biological processes, including tumorigenesis. However, the clinical significance and roles of IRF2BP2 in human gastric cancer (GC) remain uncertain. Clinical GC tissues were obtained from GC patients at the First Affiliated Hospital of Nanchang University. Immunohistochemistry (IHC) was conducted to detect the IRF2BP2 protein in clinical paraffin specimens. Cell proliferation, migration and invasion were evaluated by MTT, colony formation assays and transwell assays. Co‐immunoprecipitation was conducted to detect the interaction between TEA domain family members 4 (TEAD4) and vestigial‐like family member 4 (VGLL4) or Yes‐associated protein 1 (YAP1). Dual‐luciferase reporter assay was used to confirm the binding of miR‐101‐3p to the 3′‐UTR. The expression of IRF2BP2 was significantly higher in GC tissues than in normal tissues. Patients with higher IRF2BP2 protein expression had lower survival. IRF2BP2 knockdown inhibited proliferation, migration, invasion and epithelial‐mesenchymal transition in GC cells. IRF2BP2 knockdown decreased the mRNA and protein levels of connective tissue growth factor (CTGF). The interaction between IRF2BP2 and VGLL4 increased the binding of TEAD4 to YAP1, resulting in the transcriptional coactivation of CTGF. In addition, miR‐101‐3p suppressed the expression of CTGF by directly targeting the 3′‐UTR of IRF2BP2. Taken together, these findings provide a model for the role of miR‐101‐3p‐IRF2BP2‐CTGF signalling axis in GC and a novel insight into the mechanism of GC progression and metastasis.


| INTRODUC TI ON
Gastric cancer (GC) is one of the most prevalent gastrointestinal malignancies, and mortality from GC is the third leading cause of cancer-related deaths worldwide. 1 The incidence of GC is highest in eastern Asia, and approximately 42% of cases occurred in China. 2 Early diagnosis of GC is challenging, and most patients are diagnosed at an advanced stage. Despite tremendous advances in surgery, chemotherapy, radiotherapy and targeted molecular therapy, the overall effectiveness of treatment is low, with the 5-year survival rate being <35%. 3 Furthermore, the median overall survival (OS) of GC is currently <12 months. 4 Therefore, exploring new diagnostic and prognostic markers is essential for developing targeted therapies for GC.
The interferon regulatory factor 2 binding protein 2 (IRF2BP2) gene encodes a nuclear protein that contains an N-terminal zinc finger and a C-terminal RING finger domain of the C3HC4 subclass that interacts with the C-terminal transcriptional repression domain of interferon regulatory factor 2 (IRF2), which is a class of transcription factors that can regulate interferon expression. [5][6][7] IRF2BP2 is an IRF2-dependent transcriptional corepressor that can inhibit both enhancer-activated transcription and baseline transcription, and inhibition is not mediated by histone deacetylase activity. 8 IRF2BP2 also performs IRF-2-independent functions, including the negative regulation of the nuclear factor of activated T cell 1 (NFAT1)-mediated transactivation of NFAT-responsive promoters, consequently affecting the cell cycle, differentiation and apoptosis. 9 IRF2BP2 is regulated by cancer-related molecules. Koeppel et al 10  an IRF2BP2-retinoic acid receptor alpha fusion protein in acute promyelocytic leukaemia. 11,12 These studies suggest that IRF2BP2 may play a role in tumorigenesis and cancer progression.
Connective tissue growth factor (CTGF) is a matricellular protein of the cysteine-rich angiogenic inducer 61 (Cyr61)/CTGF/Nov family involved in many physiological and pathological processes, including carcinogenesis and regulation of the tumour microenvironment. In GC patients, elevated CTGF expression is strongly correlated with lymph node metastases, peritoneal dissemination and poor prognosis. [13][14][15][16] Moreover, CTGF expression levels are positively associated with the levels of vascular endothelial growth factors C and D (VEGF-C and VEGF-D). 13 A study suggests that CTGF promotes GC cell proliferation by inducing the expression of cyclin D1, 16 and the down-regulation of CTGF inhibits GC cell metastasis and decreases the expression and proteolytic activity of both matrix metalloproteinase (MMP)-2 and MMP-9. 15 CTGF binds to multiple cell surface receptors in a context-dependent manner and functions as an oncogene in different types of tumours. [17][18][19] CTGF is also one of the target genes downstream of the transcriptional coactivator Yes-associated protein 1 (YAP1), which is an essential effector of the Hippo pathway. YAP1 accumulates in the nucleus, where it binds primarily to DNA-binding transcription factors TEA domain family members 1-4 (TEAD1-4) and transcriptionally coactivates CTGF, contributing to tumorigenesis and cancer progression. 20,21 VGLL4 serves as a transcriptional corepressor in the nucleus by binding to TEAD4 and blocking transcriptional coactivation; moreover, VGLL4 competes with YAP1 for binding to TEAD4. 21,22 A previous study indicates that VGLL4 binds to IRF2BP2 to promote PD-L1 expression and induces immune evasion through IRF2 inhibition in lung cancer. 23 The binding of VGLL4 to IRF2BP2 also activates the expression of VEGF-A in muscle cells. 24 This evidence implies that IRF2BP2 may cross talk with Hippo pathway and CTGF via binding to VGLL4.
The Cancer Genome Atlas (TCGA) database shows the amplification of the IRF2BP2 gene in the majority of malignant tumours, including GC, which prompted us to focus on the clinical significance and roles of IRF2BP2 in human GC. This study evaluated the effect of the overexpression of IRF2BP2 in GC, indicating that IRF2BP2 regulated the expression of CTGF in a YAP1-dependent manner and that the miR-101-3p-IRF2BP2-CTGF axis could be a potential prognostic marker and therapeutic target in GC.

| Patients and clinical specimens
Paraffin-embedded GC tissue samples (n = 65) and adjacent noncancerous gastric tissues (n = 20) were from patients who underwent resection at the First Affiliated Hospital of Nanchang University between January 2009 and December 2011. Complete clinicopathological data of all the patients were available (Table 1). Fresh GC tissues (n = 40) and paired adjacent noncancerous tissues were stored in liquid nitrogen before use (Table S2). All patients agreed to participate in the study and provided written informed consent.

| Immunohistochemistry
Immunohistochemistry was performed as previously described. 25 Slides were incubated with IRF2BP2 rabbit polyclonal antibody (1:100; Abcam, ab180891) in a humidified chamber overnight at 4°C, washed thrice with PBS and incubated with secondary antibody for 50 min at 37°C. Sections were washed, developed with 3,3′-diaminobenzidine tetrahydrochloride and counterstained with haematoxylin before mounting. All scores were evaluated by two pathologists who were blinded to the pathological information.
Immunohistochemical grading standards were performed according to the previously described methods. 25

| Immunoblotting
The proteins were extracted from GC tissues or cells by lysis buffer

| RNA extraction and real-time quantitative PCR
Total RNA from GC cell lines and tissues was extracted using relative to the internal control (GAPDH) was analysed using the 2 −ΔΔC T method.

| Cell proliferation and colony formation assays
Forty-eight hours after transfection, SGC-7901 or BGC-823 cells were plated in 96-well plates at a density of 2000 cells/well for MTT assays or seeded in six-well plates at a density of 300 cells/well for colony formation assays following the methods described previously. 25

| Migration and invasion assays
Cell invasion and migration were conducted using 8-µm transwell inserts (Costar) coated with or without 60 µL of Matrigel (BD Biosciences), which were placed into each well of a 24-well plate. Cells were stained with crystal violet and counted by direct microscopic visualization.

| Co-immunoprecipitation
The cell lysates were centrifuged, and the supernatants were transferred to test tubes. The IRF2BP2 antibody (Proteintech, 18847-1-AP), TEAD4 antibody (Proteintech 12418-1-AP) or IgG (served as a control) was incubated with the corresponding supernatant at 4°C for 4 hours. After that, protein A/G beads (Santa Cruz Biotechnology) were added to the tubes, and the mixtures were incubated at 4°C for 2 hours. Beads were washed thrice with lysis buffer, and Western blotting was performed to detect the bound proteins.

| Vector construction and dual-luciferase reporter assay
The present study predicted the binding of miR-101-3p to the 3′-

| Xenografted tumour model and staining
BALB/c-nu mice (5-6 weeks old) were purchased from the SLACCAS Experimental Animal Company (Shanghai). The mice were randomly divided into two groups (six animals per group).
For the subcutaneous assay, 1 × 10 7 SGC-7901 cells that transfected with lentivirus encoding IRF2BP2 shRNAs or scramble shRNA were subcutaneously injected. The tumour volume was measured every 3 days with a calliper and calculated using the formula (L × W 2 )/2, where L is the length diameter and W is the width diameter of the tumour. After 21 days, the tumours were excised. All mice were handled according to the institutional guidelines.

| Statistical analysis
Data were analysed using SPSS software version 20.0 (SPSS). The difference between the two groups was determined by Student's t test. Survival curves were plotted using the Kaplan-Meier method and compared by the log-rank test. Correlations between IRF2BP2 mRNA and CTGF mRNA expression were estimated using the Spearman correlation coefficient. P values <.05 were considered statistically significant.

| IRF2BP2 expression is up-regulated in GC cell lines and tissues
The bioinformatic analysis of the TCGA database indicated that IRF2BP2 was amplified in most types of tumours, including GC (http://www.cbiop ortal.org/) ( Figure 1A), and the level of IRF2BP2 mRNA in primary GC tissues was significantly higher than that in normal tissues (http://ualcan.path.uab.edu/cgi-bin/TCGAE xResu ltNew2.pl?genen am=IRF2B P2&ctype =STAD) ( Figure 1B). Western blotting was used to determine the expression of IRF2BP2 in GC tissues and cell lines. The IRF2BP2 protein levels were significantly higher in all GC cell lines than in the human immortalized gastric epithelial cell line GES-1 ( Figure 1C). Moreover, IRF2BP2 protein expression was elevated in GC tissues relative to adjacent noncancerous gastric tissues (n = 8) ( Figure 1D). To further confirm the prognostic value of IRF2BP2 proteins, IHC was used to detect the expression of these proteins in GC tissues (n = 65). The results showed that IRF2BP2 was predominantly localized to the nucleus, IRF2BP2 protein expression was higher in GC tissues, and the TNM stage was positively correlated with the expression level of IRF2BP2; in contrast, IRF2BP2 expression was lower in adjacent noncancerous gastric tissues ( Figure 2B). Next, the relationship between IRF2BP2 protein expression and the clinicopathological characteristics of GC was evaluated. IRF2BP2 expression was closely associated with age (P = .036), tumour size (P = .0006), TNM stage (P = .0002), depth of invasion (P = .0002), lymph node metastasis (P = .0008) and vascular invasion (P = .03) (Table 1). Furthermore, the clinical follow-up data for patients with GC indicated that high IRF2BP2 expression contributed to poor OS (P = .0057) ( Figure 2C), and the 5-year survival rate was comparatively higher in patients with lower IRF2BP2 expression (Table S1).

| Predicted prognostic value of IRF2BP2 in GC patients
F I G U R E 1 IRF2BP2 expression is up-regulated in GC. A, An analysis of the database in cBioPortal showed that IRF2BP2 was amplified in most tumours. B, The TCGA database indicated that the expression of IRF2BP2 mRNA was significantly higher in primary GC tissues (n = 415) than in normal tissues (n = 34) (P < .001). C, The IRF2BP2 protein levels were higher in GC cell lines than in the GES-1 cell line. D, IRF2BP2 protein expression was higher in GC tissues than in adjacent noncancerous gastric tissues. GC, gastric cancer; IRF2BP2, interferon regulatory factor 2 binding protein 2; TCGA, The Cancer Genome Atlas

| Knockdown of IRF2BP2 inhibits GC cell proliferation and invasion
To

| The binding of IRF2BP2 to VGLL4 increases the binding of TEAD4 to YAP1, leading to the transcriptional coactivation of CTGF expression
This study further explores the specific mechanism of function of IRF2BP2. Data from the proteomic database STRING (https :// string-db.org/cgi/netwo rk.pl?taskI d=gcaog yyOxVi0) predicted that IRF2BP2 could interact with VGLL4 ( Figure 4A). Moreover, studies proved that the binding of IRF2BP2 to VGLL4 played an important role in immune evasion and angiogenesis. 23,24 It has been established that VGLL4 is a transcriptional corepressor that

| The expression of CTGF and IRF2BP2 is positively correlated in clinical GC samples
In clinical samples, 40 pairs of fresh GC tissue and adjacent noncancerous tissues were used (Table S2). The mRNA levels of IRF2BP2 and CTGF were higher in GC tissues than in paired noncancerous tissues ( Figure 5A). Moreover, the up-regulation of IRF2BP2 mRNA was positively associated with the up-regulation of CTGF mRNA (R = 0.581) ( Figure 5B). We further analysed the relationship between the mRNA expression of CTGF or IRF2BP2 and clinicopathological features, including TNM stage, lymphatic metastasis, depth of invasion and differentiation. High CTGF mRNA expression was correlated with the depth of invasion and lymphatic metastasis, and CTGF mRNA expression in patients with TNM stage III was significantly higher than that in patients with TNM stage I or II ( Figure 5C). In addition, IRF2BP2 mRNA expression was positively associated with the depth of invasion, and the level of IRF2BP2 mRNA expression in patients with TNM stage III was significantly higher than that in patients with lower TNM stage ( Figure 5D).

| IRF2BP2 depletion impedes xenograft tumour growth
To evaluate the biological significance of IRF2BP2, SGC-7901 cells expressing scramble shRNA or IRF2BP2 shRNA were inoculated to build a xenograft mouse model. Tumour size was monitored for 21 days. The results suggest that the knockdown of IRF2BP2 significantly inhibits tumour growth ( Figure 7A). To confirm the mechanisms identified in cell lines, the tumours were removed from mice and subjected to RT-qPCR and Western blotting. Both the mRNA and protein levels of CTGF were significantly decreased F I G U R E 5 Expression of IRF2BP2 and CTGF mRNA in fresh GC samples. A, The results of RT-qPCR showed that the mRNA levels of IRF2BP2 and CTGF were higher in 40 fresh GC samples than in the paired adjacent normal tissues. B, IRF2BP2 mRNA levels were positively correlated with CTGF mRNA levels in these 40 samples. C, Relationship between CTGF mRNA expression and tumour stage or differentiation. D, Relationship between IRF2BP2 mRNA expression and tumour stage or differentiation. *P < .05, **P < .01. CTGF, connective tissue growth factor; GC, gastric cancer; IRF2BP2, interferon regulatory factor 2 binding protein 2 upon IRF2BP2 knockdown, which is consistent with in vitro findings ( Figure 7B,C).

| D ISCUSS I ON
IRF2BP2 has been reported to be involved in the malignancy of breast cancer, leukaemia and chondrosarcoma. 11,12,37 The TCGA database revealed that the IRF2BP2 gene was amplified in most tumours, including GC. In addition, the level of IRF2BP2 mRNA was higher in primary GC tissues than in normal tissues. Therefore, our study focused on the roles and mechanisms of IRF2BP2 in GC.
In the present study, the IRF2BP2 protein expression levels were significantly increased in GC cell lines and tissues, indicating Hippo pathway is composed of a kinase cascade, and the transcriptional coactivator YAP1 is up-regulated in most human tumours. [45][46][47] Therefore, the Hippo-YAP1 signalling pathway and TEADs are potential targets for cancer therapy. [48][49][50] In this study, we assessed whether the interaction between was also found to inhibit the p53-mediated transactivation of the p21 and the BAX genes. 10 MicroRNAs play important roles in transcriptional regulation and are frequently involved in human tumours. [30][31][32] MicroRNA-101-3p acts as a tumour suppressor in GC by targeting the serum response factor directly and suppressing the proliferation and invasion of GC cells induced by HOX Transcript Antisense RNA (HOTAIR). 36 Yan et al 54 reported that lncRNA SNHG6 promoted cell proliferation and EMT by sponging miR-101-3p, leading to poor prognosis in GC.
Luciferase activity assays were performed in present study to con- Taken together, our findings indicate that the expression of IRF2BP2 is increased in GC, which is closely related to proliferation, migration and invasion, and contributes to poor prognosis in GC. The binding of IRF2BP2 to VGLL4 weakens the interaction between TEAD4 and VGLL4 and increases the binding of TEAD4 to YAP1, resulting in the transcriptional coactivation of CTGF expression. In addition, miR-101-3p inhibits IRF2BP2 expression by targeting the 3′-UTR of IRF2BP2 mRNA, consequently suppressing CTGF expression in a YAP1-dependent manner. We propose that the miR-101-3p-IRF2BP2-CTGF axis has a role in GC, and IRF2BP2 is a potential prognostic marker and therapeutic target in GC.

ACK N OWLED G EM ENTS
The authors thank the Department of General Surgery and Department of Pathology in the First Affiliated Hospital of Nanchang University for sharing the clinicopathological information of patients.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
All the patients agreed to participate in our study and provided written informed consent. All mice were cared in accordance with the National Guides for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee of the First Affiliated Hospital of Nanchang University.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets supporting the conclusions of this article are included within the article.