The Epstein-Barr virus (EBV) noncoding RNAs, EBV-encoded RNA 1 (EBER1) and EBER2, are the most abundant viral transcripts in all types of latently infected human B cells, but their function remains unknown. We carried out heterokaryon assays using cells that endogenously produce EBERs to address their trafficking, as well as that of the La protein, because EBERs are quantitatively bound by La in vivo. Both in this assay and in oocyte microinjection assays, EBERs are confined to the nucleus, suggesting that their contribution to viral latency is purely nuclear. EBER1 does not bind exportin 5; therefore, it is unlikely to act by interfering with microRNA biogenesis. In contrast, La, which is a nuclear phosphoprotein, undergoes nucleocytoplasmic shuttling independent of the nuclear export protein Crm1. To ensure that small RNA shuttling can be detected in cells that are negative for EBER shuttling, we demonstrate the shuttling of U1 small nuclear RNA.

EBER 1, asmall noncoding viral RNA abundantly expressed in all cells transformed by Epstein-Barr virus (EBV), has been shown to associate with the human ribosomal protein L22. Here we present in vitro binding studies using purified RNAs and recombinant proteins. Electrophoretic mobility-shift assays (EMSAs) show that recombinant L22 (rL22) and maltose-binding protein (MBP)-tagged L22 protein bind EBER 1i nv itro, both forming three specific protein-dependent mobility shifts. Use of am ixture of rL22 and MBP-L22 indicates that these three shifts contain one, two, or three L22 proteins per EBER 1m olecule. EMSAs performed with EBER 1d eletion constructs and EBER 1s tem-loops inserted into an onbinding RNA, HSUR 3, identify stem-loops I, III, and IV as L22 binding sites. The existence of multiple L22 binding sites on EBER 1inside cells is demonstrated by in vivo UV cross-linking. Our results are discussed with respect to the function of EBER 1i nE BV-infected human Bc ells.

Oligodeoxyribonucleotides containing thymidine and 8-oxo-2'-deoxyadenosine can form pyr.pur.pyr type triplexes with double-stranded DNA. Unlike triplexes whose third strands contain thymidine and deoxycytidine, the stability of these triplexes is independent of pH. We have prepared d-ps-TAAATAAATTTTTAT-L [I(A)], where A is 8-oxo-2'-deoxyadenosine, ps is 4'-hydroxymethyl-4,5',8- trimethylpsoralen and L is a 6-amino-2-(hydroxymethyl)hexyl linker. The oligomer is designed to interact with a homopurine sequence in the promoter region of the human gene coding for the 92 kDa form of collagenase type IV. Oligomer I(A) and oligomer I(C), which contains 2'-deoxycytidine in place of 8-oxo-2'-deoxycytidine, both form stable triplexes at pH 6.2, but only I(A) forms a stable triplex with a model duplex DNA target at pH 7.5, as determined by UV melting experiments. Triplex formation is stabilized by the presence of the psoralen group. Upon irradiation both I(A) and I(C) form photoadducts with the DNA target at pH 6.2, but only I(A) forms a photoadduct at pH 7.5. In these photoreactions oligomer I(A) appears to selectively form a photoadduct with a C in the purine-rich strand of the duplex target. Although a T residue is present in the pyrimidine-rich strand of the target at the duplex/triplex junction, essentially no adduct formation takes place with this strand, nor is interstrand cross-linking observed. The extent of photoadduct formation decreases with increasing temperature, behavior which is consistent with the UV melting curve of the triplex. A tetramethylrhodamine derivative of I(A) was prepared and found to cross-link less extensively than I(A) itself. Oligomer I(A) is completely resistant to hydrolysis when incubated for 24h in the presence of 10% fetal bovine serum at 37 degree C, although it is hydrolyzed by S1 nuclease. The properties of oligomer I(A) suggest that 8-oxo- containing oligomers may find utility as antigene oligonucleotide reagents.

In Epstein-Barr virus (EBV) latent infection, the EBV-encoded RNAs EBER1 and EBER2 accumulate in the host cell nucleus to ~106 copies. While the expression of EBERs in cell lines is associated with transformation, a mechanistic explanation of their roles in EBV latency remains elusive. To identify EBER-specific gene expression features, we compared the proteome and mRNA transcriptome from BJAB cells (an EBV-negative B lymphoma cell line) stably transfected with an empty plasmid or with one carrying both EBER genes. We identified ~1800 proteins with at least 2 SILAC pair measurements, of which only 8 and 12 were up- and downregulated ≥ 2-fold, respectively. One upregulated protein was PIK3AP1, a B-cell specific protein adapter known to activate the PI3K-AKT signaling pathway, which regulates alternative splicing and translation in addition to its pro-survival effects. In the mRNA-seq data, the mRNA levels for some of the proteins changing in the SILAC data did not change. We instead observed isoform switch events. We validated the most relevant findings with biochemical assays. These corroborated the upregulation of PIK3AP1 and AKT activation in BJAB cells expressing high levels of both EBERs and EBNA1 (a surrogate of Burkitt’s lymphoma EBV latency I) relative to those expressing only EBNA1. The mRNA-seq data in these cells showed multiple upregulated oncogenes whose mRNAs are enriched for 3´-UTR AU-rich elements (AREs), such as ccl3, ccr7, il10, vegfa and zeb1. The CCL3, CCR7, IL10 and VEGFA proteins promote cell proliferation and are associated with EBV-mediated lymphomas. In EBV latency, ZEB1 represses the transcription of ZEBRA, an EBV lytic phase activation factor. We previously found that EBER1 interacts with AUF1 in vivo and proposed stabilization of ARE-containing mRNAs. Thus, the ~106 copies of EBER1 may promote not only cell proliferation due to an increase in the levels of ARE-containing genes like ccl3, ccr7, il10, and vegfa, but also the maintenance of latency, through higher levels of zeb1.