EBV-encoded miRNAs target ATM-mediated response in nasopharyngeal carcinoma

Abstract

Nasopharyngeal carcinoma (NPC) is a highly invasive epithelial malignancy that is prevalent in southern China and Southeast Asia. It is consistently associated with latent Epstein-Barr virus (EBV) infection. In NPC, miR-BARTs, the EBV-encoded miRNAs derived from BamH1-A rightward transcripts, are abundantly expressed and contribute to cancer development by targeting various cellular and viral genes. In this study, we establish a comprehensive transcriptional profile of EBV-encoded miRNAs in a panel of NPC patient-derived xenografts and an EBV-positive NPC cell line by small RNA sequencing. Among the 40 miR-BARTs, predominant expression of 22 miRNAs was consistently detected in these tumors. Among the abundantly expressed EBV-miRNAs, BART5-5p, BART7-3p, BART9-3p, and BART14-3p could negatively regulate the expression of a key DNA double-strand break (DSB) repair gene, ataxia telangiectasia mutated (ATM), by binding to multiple sites on its 3'-UTR. Notably, the expression of these four miR-BARTs represented more than 10% of all EBV-encoded miRNAs in tumor cells, while downregulation of ATM expression was commonly detected in all of our tested sequenced samples. In addition, downregulation of ATM was also observed in primary NPC tissues in both qRT-PCR (16 NP and 45 NPC cases) and immunohistochemical staining (35 NP and 46 NPC cases) analysis. Modulation of ATM expression by BART5-5p, BART7-3p, BART9-3p, and BART14-3p was demonstrated in the transient transfection assays. These findings suggest that EBV uses miRNA machinery as a key mechanism to control the ATM signaling pathway in NPC cells. By suppressing these endogenous miR-BARTs in EBV-positive NPC cells, we further demonstrated the novel function of miR-BARTs in inhibiting Zta-induced lytic reactivation. These findings imply that the four viral miRNAs work co-operatively to modulate ATM activity in response to DNA damage and to maintain viral latency, contributing to the tumorigenesis of NPC. © 2017 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.

Keywords: ATM serine/threonine kinase (ATM); EBV-miRNAs; Epstein-Barr virus; nasopharyngeal carcinoma; transcriptome sequencing.

Figure 1

Expression of viral miRNAs in EBV‐positive NPCs. (A) The number of EBV‐miRNA reads is indicated per 10 million of the total mapped mature miRNAs to normalize the sequencing depth in each library. The libraries of C666‐1, C17, and the average of four primary NPC‐derived xenografts (xeno‐C666, xeno‐2117, xeno‐1915, and C15) are shown as mean ± SD. (B) The distribution of the individual miR‐BARTs to the total viral miRNAs in the libraries is shown in the pie charts.

Figure 2

ATM is a potential target of several EBV‐encoded miRNAs. (A) Suggested putative miR‐BART recognition sites on the ATM 3'‐UTRs are shown. The seed‐binding regions of the miR‐BARTs are underlined and the bases mutated for the luciferase reporter analysis are marked in red. (B) The expression of miR‐BARTs of interest in primary NPC samples was examined using RT‐qPCR (n = 45). The expression of miR‐BARTs was normalized to EBNA1 for analysis. The low‐expression miR‐BARTs (BART21‐5p and BART20‐3p) and other high‐expression miR‐BARTs were included for comparison. The data shown are the mean ± SEM from the tested samples. (C) Immunoblotting of ATM protein in the NPC samples. Three immortalized normal NP cell lines (NP361, NP550, and NP69), C666‐1, and five NPC xenografts (xeno‐C666, xeno‐2117, xeno‐1915, C15, and C17) were analyzed. (D) ATM protein expression in FFPE specimens was analyzed by immunohistochemistry. NP‐1 and NP‐2 are examples of normal NP epithelia with strong ATM positive stain. NPC‐1 and NPC‐2 are examples of NPC cells that are negative for ATM expression. The infiltrated lymphocytes that served as internal controls were strongly positive. NPC‐3 is an example of NPC cells that are positive for ATM (original magnification × 400). (E) Expression of ATM in the primary NPC samples was demonstrated by RT‐qPCR (number of NPs = 16; number of NPCs = 45). (F) The scatter plot demonstrates the inverse correlation between the miR‐BART of interest and ATM mRNA expression levels in 25 NPC samples. (G) The direct interaction between ATM expression and miR‐BARTs was demonstrated in dual luciferase reporter assays. The pMIR‐REPORT vectors containing the wild‐type ATM binding sites (UTR‐wt) or the mutant binding sites (UTR‐mut) were tested against the corresponding miR‐BART mimics and inhibitors. The relative firefly luciferase activity was normalized to the Renilla luciferase control, and results were taken from at least three independent experiments. The data shown are the mean + SD. miR‐Ctl = miR‐BART mimic control; miR‐BART = miR‐BART mimic; Inh‐NEG = miRNA inhibitor negative control; Inh‐BART = miR‐BART inhibitor. **p < 0.01; ***p < 0.001.

Figure 3

Modulation of ATM expression by EBV‐encoded miRNAs. (A) Downregulation of ATM in NP69 and HeLa cells by ebv‐miRNAs from both clusters. Cells transfected with expression vectors containing BART‐Cluster 1 (miR‐BART‐C1) and BART‐Cluster 2 (miR‐BART‐C2) derived miRNAs or individual miR‐BARTs were analyzed by western blotting. The empty vector and irrelevant miRNA mimic were used as negative controls. (B, C) The ATM expression of BART‐Cluster 1 (C1) and BART‐Cluster 2 (C2) miRNA expressing cells was restored by the indicated miR‐BART inhibitors (Inh‐BARTs). Endogenous ATM expression was measured by western blot and the relative ATM expression is shown under the blots. (D) Endogenous ATM protein expression in C666‐1 cells was restored by the miR‐BART inhibitors. The inhibitors of BART5‐5p, BART7‐3p, BART9‐3p, and BART14‐3p were introduced, individually or together (All 4 Inh‐BARTs), into the C666‐1 cells and protein lysates were collected 48 h post‐transfection for western blot analysis.

Figure 4

EBV‐miRNAs enhance the ionizing radiation (IR) sensitivity of epithelial cells. The miR‐BART transfected cells were treated with different doses of IR and the cells were harvested at 30 min post‐irradiation for immunoblotting analysis. The expression of the basal ATM proteins, phospho‐ATM (p‐ATM), phospho‐CHK2 (p‐CHK2), and γ‐H2AX, was analyzed. Actin was probed as the loading control. ATM knockdown cells (siATM) were included as positive controls.

Figure 5

EBV‐miRNAs suppress the DNA damage response to ionizing radiation. (A) Representative images of the H2AX nuclear foci staining assay. Cells transfected with either miRNA mimics (miR‐NEG) or a combination of four miR‐BART mimics (All 4 miR‐BARTs) were treated with a single dose of 3 Gy irradiation, which was followed by immunostaining with the γ‐H2AX^ser139 antibody 1 h post‐irradiation. Cells containing more than five γ‐H2AX foci in the nucleus were considered positive and the percentage of γ‐H2AX‐positive cells was calculated (n = 100). The mean ± SD for three independent experiments are shown. (B) Comet assays of the DNA repair capacity were performed on NP69 and HeLa cells, which were treated with a single irradiation dose of 10 and 20 Gy, respectively. The tail moment of the irradiated cells at 6 h is shown in the bar charts; mean ± SEM. Student's t‐test was conducted compared with the miR‐NEG control. *p < 0.05; **p < 0.01; ***p < 0.001. (C, D) Clonogenic survival assays. Approximately 200–1600 transfected HeLa cells were seeded into a six‐well plate and treated with a single dose of 0.5, 1 or 2 Gy irradiation. The cells were stained and colonies containing more than 30 cells were counted. The survival fraction was calculated by dividing the plating efficiency of the irradiated cells by the plating efficiency of the untreated cultures. Statistical analyses using Student's t‐test were conducted and compared with the miR‐NEG control. *p < 0.05; **p < 0.01.

Figure 6

BZLF1‐induced virus reactivation is suppressed by miR‐BARTs. (A) The BZLF1‐expressing plasmid (1.25 μg) was co‐transfected with the indicated miR‐BART inhibitors (10 μm) into C666‐1 cells in a six‐well plate. Cells were harvested for immunoblotting analysis after 48 h. The expression of ATM, ATM downstream effectors (p‐ATM and γ‐H2AX), and the early viral lytic proteins (BRLF1 and BMRF1) was analyzed. (B) The synergistic effects of BZLF1 and ATM on p‐ATM, γ‐H2AX, BRLF1, and BMRF1 expression in C666‐1 cells. The cells were co‐transfected with ATM and BZLF1‐expressing plasmids and protein expression was examined by immunoblotting. The two isoforms of BMRF1 are indicated by arrows. The relative expression of the protein level was calculated for comparison.