SNHG8 Promotes the Progression of Epstein-Barr Virus-Associated Gastric Cancer via Sponging miR-512-5p and Targeting TRIM28

Abstract

SNHG8, a family member of small nucleolar RNA host genes (SNHG), has been reported to act as an oncogene in gastric carcinoma (GC). However, its biological function in Epstein-Barr virus (EBV)-associated gastric cancer (EBVaGC) remains unclear. This study investigated the role of SNHG8 in EBVaGC. Sixty-one cases of EBVaGC, 20 cases of non-EBV-infected gastric cancer (EBVnGC), and relative cell lines were studied for the expression of SNHG8 and BHRF1 (BCL2 homolog reading frame 1) encoded by EBV with Western blot and qRT-PCR assays. The relationship between the expression levels of SNHG8 and the clinical outcome in 61 EBVaGC cases was analyzed. Effects of overexpression or knockdown of BHRF1, SNHG8, or TRIM28 on cell proliferation, migration, invasion, and cell cycle and the related molecules were determined by several assays, including cell proliferation, colony assay, wound healing assay, transwell invasion assay, cell circle with flow cytometry, qRT-PCR, and Western blot for expression levels. The interactions among SNHG8, miR-512-5p, and TRIM28 were determined with Luciferase reporter assay, RNA immunoprecipitation (RIP), pull-down assays, and Western blot assay. The in vivo activity of SNHG8 was assessed with SNHG8 knockdown tumor xenografts in zebrafish. Results demonstrated that the following. (1) BHRF1 and SNHG8 were overexpressed in EBV-encoded RNA 1-positive EBVaGC tissues and cell lines. BHRF1 upregulated the expressions of SNHG8 and TRIM28 in AGS. (2) SNHG8 overexpression had a significant correlation with tumor size and vascular tumor thrombus. Patients with high SNHG8 expression had poorer overall survival (OS) compared to those with low SNHG8 expression. (3) SNHG8 overexpression promoted EBVaGC cell proliferation, migration, and invasion in vitro and in vivo, cell cycle arrested at the G2/M phase via the activation of BCL-2, CCND1, PCNA, PARP1, CDH1, CDH2 VIM, and Snail. (4) Results of dual-luciferase reporter assay, RNA immunoprecipitation, and pull-down assays indicated that SNHG8 sponged miR-512-5p, which targeted on TRIM28 and promoted cancer malignant behaviors of EBVaGC cells. Our data suggest that BHRF1 triggered the expression of SNHG8, which sponged miR-512-5p and upregulated TRIM28 and a set of effectors (such as BCL-2, CCND1, CDH1, CDH2 Snail, and VIM) to promote EBVaGC tumorigenesis and invasion. SNHG8 could be an independent prognostic factor for EBVaGC and sever as target for EBVaGC therapy.

Keywords: EBVaGC; SNHG8; TRIM28; cell proliferation; invasion; miR-512-5p; migration.

Figure 1

Expression of BFRF1 in GC and GC cell lines and its regulation with SNHG8 and TRIM28. (A, B) BHRF1 was expressed in EBVaGC tissues at different levels, while it was almost not expressed on EBVnGC. (C) The SNHG8 expression was much higher in EBVaGC tissues than that in EBVnGC tissues (p = 0.0159). RQ = 2^−ΔΔCt, where Ct values were generated from qPCR. (D) The expression of BHRF1 and SNHG8 was much higher in EBV⁺ AGS-BX cells than in EBV^- GC cell lines (AGS, MKN28). (E) Overexpression of EBV protein BFRF1 resulted in an increased TRIM28 demonstrated by Western blot analysis. (F) Overexpression of EBV protein BFRF1 enhanced the transcription of SNHG8 and TRIM28 demonstrated by qRT-PCR. ****p < 0.0001.

Figure 2

Poor prognosis of EBVaGC was associated with the upregulation of SNHG8. (A) EBER in gastric cancer detected by in situ hybridization. Left: EBER is positive in EBV-associated gastric cancer (EBVaGC). Right: EBER is negative in non-EBV-associated gastric cancer (EBVnGC). (B) The expression of SNHG8 was higher in 61 cases of EBVaGC than that in their paracancerous tissues (p < 0.001). (C) Kaplan–Meier curves showed that patients with a high level of SNHG8 had a poor OS (p = 0.018).

Figure 3

Overexpression of SNHG8 promoted the proliferation, migration, and invasion of EBVaGC. (A) AGS-BX1 cells with SNHG8 gene overexpression (AGS-BX1/SNHG8-OE)had a high level of the expression of SNHG8 demonstrated by qPCR as compared to vector alone control (AGS-BX1/NC) and wild-type cells (AGS-BX1). The overexpression of SNHG8 promoted the malignant behaviors of AGS-BX1/SNHG8-OE cells in the following aspects: (B) increasing cell proliferation as measured with MTS assay; (C) enhancing colony formation; (D) stopping cells at G2/M; (E, F) increasing the expression of proliferation-related genes of BCL-2, CCND1, C-PARP, and PCNA, confirmed by qRT-PCR and Western blot; (G, H) promoting the migration measured with the wound healing assay and the invasion measured with the transwell invasive assay; (I) and enhancing the expression of invasion/metastasis-related-genes, Snai1, VIM, CDH1, and CDH2 as demonstrated by qRT-PCR and Western blot (J, K). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. ns mean not statistical significance.

Figure 4

Silencing SNHG8 inhibited the proliferation, migration, and invasion of EBVaGC. (A) SNHG8 gene knockdown in AGS-BX1 cells (AGS-BX1/SNHG8-SH) had a low expression of SNHG8 quantified by qPCR as compared to vector alone control (AGS-BX1/NC) and wild-type cells (AGS-BX1). SNHG8 silencing reduced the malignant behaviors of AGS-BX1/SNHG8-SH cells in following aspects: the cell proliferation as measured with MTT assay (B); colony formation (C); cells in G0/1 (D); the expressions of proliferation-related-genes of BCL-2, CCND1, C=PARP, and PCNA demonstrated by qRT-PCR and Western blot (E, F); the migration measured with the wound healing assay (G, H); the invasion measured with the transwell invasive assay (I); and the expressions of invasion/metastasis-related-genes of Snail, VIM, CDH1, and CDH2 demonstrated by qRT-PCR and Western blotting (J, K). **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 5

SNHG8 exerted its function through sponging hsa-miR-512-5p and upregulating of TRIM28. (A, B) Sequence alignment of hsa-miR-512-5p with the putative binding sites in the wild-type and mutant regions of SNHG8 and TRIM28. (C, D) Hsa-miR-512-5p expression was decreased in AGS-BX1/SNHG8-OE cells and increased in AGS-BX1/SNHG8-SH cells tested with qRT-PCR. (E, F) The expression of SNHG8 and TRIM28 was reduced in AGS-BX1/miR-512-5p mimic cells tested with qRT-PCR. (G–J) Dual-luciferase reporter assay showed that hsa-miR-512-5p mimics reduced the intensity of fluorescence in HEK293T or AGS-BX1 cells transfected with SNHG8-Wt or TRIM28-Wt, but not in the controls of SNHG8-Mut or TRIM28-Mut vector. (K–M) TRIM28 expression was increased in AGS-BX1/SNHG8-OE cells and decreased in AGS-BX1/SNHG8-SH cells tested with qRT-PCR and Western blot analysis. (N) SNHG8 expression was measured with qRT-PCR after RNA pull-down assay using Bio-miR-NC and Bio-miR-512-5p. (O) SNHG8 and miR-512-5p levels were determined by qRT-PCR after Ago 2 or IgG RIP assay. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 6

Enhancing of TRIM28 was critical for SNHG8-mediated malignant behaviors. (A, B) TRIM28 expression was increased in AGS-BX1/SNHG8-OE cells and reduced by silencing of TRIM28 tested by RT- PCR and Western blot. (C) SNHG8 enhanced cell proliferation could be reduced by silencing of TRIM28 tested with MTS assay. (D) SNHG8 enhanced colony formation could be reduced by silencing of TRIM28. (E) SNHG8-enhanced cell cycle stopping at G2/M could be reduced by silencing of TRIM28 as detected by flow cytometry. (F, G) SNHG8-enhanced expression of BCL-2, CCND1, C-PARP, and PCNA could be reduced by silencing of TRIM28 tested with qRT-PCR and Western Blot. (H, I) SNHG8-enhanced cell migration could be reduced by silencing of TRIM28 tested with wound healing assay. (J) SNHG8-enhanced cell invasion could be reduced by silencing of TRIM28 tested with transwell assay. (K, L) SNHG8-enhanced expressions of invasion/metastasis-related-genes of VIM and CDH2 could be reduced by silencing of TRIM28 tested with qRT-PCR and Western blot. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 7

SNHG8 promoted the growth of EBVaGC tumor in zebrafish. (A) The growth of zebrafish was not affected by the injection of AGS-BX1 cells in 3 days. (B) Representative images of zebrafish from the control (AGS-BX1-NC) and AGS-BX1-SNHG8-SH groups with zoom-in of cancer cells. (C, D) Analysis of cancer cell numbers of each group in zebrafish assay. AGS-BX1 cells transfected with red florescence gene (AGS-BX1-NC, AGS-BX1-SNHG8-SH) could form tumor in zebrafish, which could be reduced by silencing of SNHG8 as determined by red fluorescence integral density and cancer cells in zebrafish abdominal cavity. ***p < 0.001; ****p < 0.0001.