A Novel Persistence Associated EBV miRNA Expression Profile Is Disrupted in Neoplasia

We have performed the first extensive profiling of Epstein-Barr virus (EBV) miRNAs on in vivo derived normal and neoplastic infected tissues. We describe a unique pattern of viral miRNA expression by normal infected cells in vivo expressing restricted viral latency programs (germinal center: Latency II and memory B: Latency I/0). This includes the complete absence of 15 of the 34 miRNAs profiled. These consist of 12 BART miRNAs (including approximately half of Cluster 2) and 3 of the 4 BHRF1 miRNAs. All but 2 of these absent miRNAs become expressed during EBV driven growth (Latency III). Furthermore, EBV driven growth is accompanied by a 5–10 fold down regulation in the level of the BART miRNAs expressed in germinal center and memory B cells. Therefore, Latency III also expresses a unique pattern of viral miRNAs. We refer to the miRNAs that are specifically expressed in EBV driven growth as the Latency III associated miRNAs. In EBV associated tumors that employ Latency I or II (Burkitt's lymphoma, Hodgkin's disease, nasopharyngeal carcinoma and gastric carcinoma), the Latency III associated BART but not BHRF1 miRNAs are up regulated. Thus BART miRNA expression is deregulated in the EBV associated tumors. This is the first demonstration that Latency III specific genes (the Latency III associated BARTs) can be expressed in these tumors. The EBV associated tumors demonstrate very similar patterns of miRNA expression yet were readily distinguished when the expression data were analyzed either by heat-map/clustering or principal component analysis. Systematic analysis revealed that the information distinguishing the tumor types was redundant and distributed across all the miRNAs. This resembles “secret sharing” algorithms where information can be distributed among a large number of recipients in such a way that any combination of a small number of recipients is able to understand the message. Biologically, this may be a consequence of functional redundancy between the miRNAs.本研究首次对体内来源的正常感染组织与肿瘤性感染组织中的EB病毒（Epstein-Barr virus, EBV）微RNA（microRNA, miRNA）开展了全面的表达谱分析。本研究揭示了体内正常感染细胞在表达限制性病毒潜伏程序时的独特病毒miRNA表达模式：生发中心细胞对应潜伏II型（Latency II），记忆性B细胞对应潜伏I/0型（Latency I/0）。本次检测的34种miRNA中，有15种完全未被检出，其中包含12种BART miRNAs（约占Cluster 2的一半）与4种BHRF1 miRNAs中的3种。除其中2种外，其余缺失的miRNA均在EBV介导的细胞增殖（潜伏III型，Latency III）过程中得以表达。此外，EBV介导的细胞增殖过程中，生发中心与记忆性B细胞内表达的BART miRNAs水平会出现5~10倍的下调。因此，潜伏III型同样具有独特的病毒miRNA表达模式。本研究将EBV介导增殖过程中特异性表达的miRNA命名为潜伏III型相关miRNA。对于采用潜伏I型或II型的EBV相关肿瘤（包括伯基特淋巴瘤、霍奇金病、鼻咽癌与胃癌），其体内潜伏III型相关的BART miRNAs（而非BHRF1 miRNAs）会出现上调。由此可见，EBV相关肿瘤中BART miRNA的表达存在失调。本研究首次证实，潜伏III型特异性基因（即潜伏III型相关BART基因簇）可在这类肿瘤中表达。尽管EBV相关肿瘤的miRNA表达模式极为相似，但通过热图聚类（heat-map/clustering）或主成分分析（principal component analysis）对表达数据进行分析时，可轻易区分不同肿瘤类型。系统分析结果显示，区分不同肿瘤类型的表达信息具有冗余性，且分布于所有检测的miRNA之中。这一现象类似于“秘密共享”（secret sharing）算法：信息可分发至大量接收方，仅需少量接收方的任意组合即可还原完整信息。从生物学角度而言，这一现象可能源于不同miRNA之间存在功能冗余性。