Study of the regulatory functions of BamHI-A rightward transcripts (BARTs) in nasopharyngeal carcinoma and EBV- infected epithelial cells

Abstract

Epstein-Barr virus (EBV) is recognized as an essential etiological factor for nasopharyngeal carcinoma (NPC), but mechanistic details of EBV pathogenesis in NPC remain elusive. EBV expresses few viral proteins in NPC cells, but rather a family of abundant non-coding RNAs, including microRNAs (miR-BARTs) and long non-coding RNAs (lnc-BARTs). I investigated the suitability of miR-BARTs for use as biomarkers in diagnosing and determining prognosis of NPC and studied the functions of lnc-BARTs in EBV-infected epithelial cells. Initially screening was performed to identify EBV miRNAs which are secreted outside cells. Three highly expressed extracellular miR-BARTs, miR- BART3, miR-BART7 and miR-BART13, have been identified previously; my study further confirmed that high levels of EBV miR-BART7 are strongly associated with NPC disease. I subsequently analyzed levels of EBV miR-BART7, EBV DNA and anti-EBV EBNA1 antibodies in NPC patients, non-NPC cancer patients and healthy individuals. This revealed that levels of both EBV DNA and miR-BART7 are higher among NPC patients than non-NPC individuals and dramatically decreased in post-treatment plasma. However, detection of EBV miR-BART7 demonstrated significantly higher sensitivity and specificity in recognizing NPC among post-radiotherapy patients with residue or metastatic tumor cells or those with low EBV DNA copy numbers in primary disease. My study further showed that EBV miR-BART7 levels can provide a prognostic evaluation of NPC therapy and that testing for the combination of EBV miR- BART7 and EBV DNA would maximize the clinical application of both biomarkers. To study the function of lnc-BARTs, I used CRISPRi and CRISPRa systems to establish altered levels of lnc-BART expression in the EBV-infected epithelial cell lines, C666-1, AGS-Bx1 and NP460-hTERT-EBV. Expression of lnc-BARTs in CRISPR-edited AGS-Bx1-BART4, NP460-hTERT-EBV-BART and C666-1- BART knockdown cell lines revealed that CRISPR technology was most effective in AGS-Bx1 cells; one of the modified cell lines, AGS-Bx1-BART4, demonstrated an approximately 30-fold induction of lnc-RPMS1. Using CRISPRa-modified AGS-Bx1-BART cell lines, I analyzed the effect of lnc-BARTs on host gene expression by the RNA-seq approach. A total of 570

differentially expressed genes (DEGs) were identified, and the expression of some of these genes further verified by RT-qPCR and Western blot analyses. Gene ontology annotation classified the 570 DEGs into three categories: biological process, cellular component and molecular function. Most genes were further classified into these categories: response to stimulus (199 genes), localization (141 genes), immune system process (99 genes) and biological adhesion (37 genes). KEGG pathway analysis revealed that the DEGs were mostly enriched in cell adhesion molecule clusters, the PI3K-Akt pathway, gene cascades that respond to pathogenic infection and antigen processing categories. Several DEGs selected from the KEGG sub-pathway analysis were found to be involved in epigenetic regulation, cancer metastasis and maintaining EBV latency. The effect of histone deacetylase (HDAC) inhibitors on regulation of EBV lytic gene expression by lnc-RMPS1 was also examined in this study. In summary, this study evaluated the applicability of EBV miR-BART7 in NPC diagnosis and prognostic prediction, established CRISPRa modified AGS- Bx1-BART cell lines and performed RNA-seq analysis to examine the effect of lnc-BARTs on global gene transcription in host cells.