Epstein-Barr Virus-Induced miR-155 Attenuates NF-κB Signaling and Stabilizes Latent Virus Persistence^{▿ †}

Cells and plasmids.

Raji is a type III latency B-cell line derived from Burkitt's lymphoma. LCL3473 (LCL) is a type III latency B-cell line derived from primary lymphoblasts transformed with EBV strain B95-8. DG-75 is an EBV-negative, B-cell line derived from Burkitt's lymphoma. These cell lines were maintained in RPMI supplemented with 10% fetal bovine serum, glutamine, penicillin, and streptomycin sulfate. 293T, EBV-positive 293-ZKO, and D98 cells were maintained in Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum, glutamine, penicillin, and streptomycin sulfate. 5x κB-L reporter plasmid that contained five copies of an NF-κB-responsive element upstream of luciferase was kindly provided by R. Kaufman, The Wistar Institute, Philadelphia, PA. miR-155 sensor plasmid was kindly provided by R. Renne (University of Florida, Gainesville). pGL4.74[hRluc/TK] was from Promega. pCMV-Flag-LMP1 was constructed by PCR amplification of LMP1 with primers introducing a HinDIII-XbaI site cloned in frame to pFLAG-CMV2 (Sigma). pCMV-FLAG-EBNA2 was constructed by PCR amplification of EBNA2 with primers introducing a Asp718-HinDIII site cloned in frame to pFLAG-CMV2. pFLAG-CMV2 was used as vector control. miRIDIAN miR-155 mimic (Dharmacon old number C-300138-01/new number C-300647-05), inhibitor (Dharmacon old number I-300138-01/new number IH-300647-06), and negative control of mimic or inhibitor were purchased from Dharmacon. The miR-155 inhibitor is a single-stranded chemically enhanced oligonucleotide designed antisense to the mature miR-155 (UUAAUGCUAAUCGUGAUAGGGG), and the miR-155 mimic is a double-stranded RNA oligonucleotide miRNA with an effective mimic of endogenous mature miR-155 function (18).

Primary B-cell infection.

Infectious particles of EBV were isolated from the supernatants of B95-8 cell cultures. Supernatants were passed through a 0.45-μm-pore-size filter and then used to infect primary human B cells. Human B cells were derived from Hypaque fractionation of peripheral blood from anonymous donors.

miRNA array analysis.

RNA was isolated from human B-cell cultures by using RNazol isolation reagent. Total RNA was isolated from B-cell cultures prior to EBV infection, and at 10 days postinfection or at 10 days in culture after mock infection. RNA was then labeled by reverse transcription (RT) and assayed for miRNA by hybridization on glass slide arrays using an oligonucleotide array for 245 miRNAs as described previously (19). Microarray images were analyzed by using Genepix Pro 6.0 (Axon Instruments). Average values of the replicate spots of each miRNA were background subtracted, normalized, and further analyzed. Normalization was performed by using global median method calculated using R (www.r-project.org) with Bioconductor (www.bioconductor.org). Finally, we selected the miRNAs measured as present in at least as many samples as the smallest class in the data set (50%). Absent calls were thresholded to 4.5 (log₂ scale) before statistical analysis, representing the average minimum intensity level detectable in the system. miRNAs that are differentially expressed in different tissues were identified by using the Significance Analysis of Microarrays (SAM; version 3.0) application with a threshold difference in expression set to 2, the s0 percentile set to 0.05 (default), and the number of permutations set to 100 (default). The SAM application calculates a score for each gene on the basis of the change of expression relative to the standard deviation of all measurements. Only mature miRNAs that are differentially expressed are reported. A tree cluster was generated by the hierarchical cluster analysis showing the separation of each tissue. For hierarchical clustering, we used average linkage metrics and centered Pearson correlation (Cluster 3.0). Java Treeview 1.1.1 was used for tree visualization.

Luciferase assays.

293T cells were transfected with Lipofectamine 2000 (Invitrogen) according to the manufacturer's specifications. A total of 2 × 10⁵ 293T cells were transfected with 20 μM miRNA mimic and 200 ng of reporter DNA with Renilla internal control vector pGL4.74 (Promega) and then harvested at 48 h posttransfection. For luciferase assay of DG75 and LCL cells, 3 × 10⁶ cells were centrifuged and resuspended in 100 μl of T solution (Amaxa) containing 60 μM miRNA mimic or inhibitor and 2 μg of reporter DNA with Renilla internal control vector pGL4.74 and then electroporated on an Amaxa Nucleofector using program A-23. After nucleofection, the cells were transferred to a flask containing 10 ml of fresh, complete RPMI 1640 medium and incubated at 37°C for 48 h. The relative luciferase activity was assayed by using Promega Dual-Luciferase reporter assay system. All data points were the average of at least three independent transfections.

Transfection of miR-155 mimic and inhibitor.

Nucleofection was used for transfection of all B lymphocytes including, DG75, Raji, and LCL cells. Briefly, 5 × 10⁶ cells were centrifuged and resuspended in 100 μl of T solution (Amaxa) containing 150 μM miRNA mimic or inhibitor and 20 μM siGLO (Dharmacon) and electroporated on an Amaxa Nucleofector using the program A-23. Mimic or inhibitor negative controls with siGLO were transfected in the same way. After nucleofection, the cells were transferred to a flask containing 20 ml of fresh, complete RPMI 1640 medium and incubated at 37°C for 48 h. The fluorescein isothiocyanate-positive cells were sorted and used in RT-PCR and Western blotting. For genome copy number assays, cells were transfected twice, with the second transfection at 48 h later. The cells were then selected by fluorescence-activated cell sorting for siGlo at 96 h after original transfection and subjected to genomic DNA purification (Promega genome DNA isolation kit).

Quantitative RT-PCR and Northern blotting.

Total RNA was prepared in TRIzol reagents (Invitrogen) in accordance with the manufacturer's instructions. RT-PCR was done as previously described. Real-time PCR was performed with Sybr green probe in an ABI Prism 7000 according to the manufacturer's specified parameters. Primer sequences for RT-PCRs were as follows: BIC, 5′-TCAAGAACAACCTACCAGAGACCTT-3′ and 5′-TCCTGGTTTGTGCCACCAT-3′; LMP-1, 5′-CCCTCAACAAGCTACCGATGAT-3′ and 5′-TCTGCCCTCGTTGGAGTTAGA-3′; EBNA-2, 5′-GGGATGCCTGGACACAAGA and 5′-GTCCATGCCCGACGTCATA-3′; β-actin, 5′-GCCATGGTTGTGCCATTACA-3′ and 5′-GGCCAGGTTCTCTTTTTATTTCTG-3′; IKKɛ, 5′-TGCGTGCAGAAGTATCAAGC-3′ and TACAGGCAGCCACAGAACAG-3′; BCL2, 5′-GAGGATTGTGGCCTTCTTTG-3′ and 5′-ACAGTTCCACAAAGGCATCC-3′; CCL5, 5′-CCATATTCCTCGGACACCAC-3′ and 5′-TGTACTCCCGAACCCATTTC-3′; and EBNA1, 5′-GGTCGTGGACGTGGAGAAAA-3′ and 5′-GGTGGAGACCCGGATGATG-3′.

For Northern blot analysis, 20 μg of total RNA was resolved on a 15% denaturing polyacrylamide gel and electrotransferred to Hybond N⁺ nylon membrane (Amersham). The membranes were cross-linked under UV. Prehybridization and hybridization were performed at 42°C in 10× Denhardt solution, 6× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate), and 0.1% sodium dodecyl sulfate (SDS). Oligonucleotides complementary to miRNA-155 or U6 snRNA were end labeled with [γ-³²P]ATP and used as probes for Northern analysis. The sequences of the oligonucleotides were 5′-CCCCTATCACGATTAGCATTAA-3′ (miR-155) and 5′-GCAGGGGCCATGCTAATCTTCTCTGTATCG-3′ (U6 snRNA). All of the probes were washed twice for 5 min at 25°C in 6× SSC-0.1% SDS.

Western blotting and antibodies.

The Western blot was preblocked with 5% dried milk in Tris-saline (20 mM Tris [pH 8.0], 200 mM NaCl), incubated with antibody in the blocking agent (diluted from 1:1,000 to 1:2,000), followed by four washes with Tris-saline containing 0.05% IPEGAL. Secondary antibody was diluted 1:2,000 in blocking agent, washed as described above, and then detected by enhanced chemiluminescence (Amersham). The following mouse monoclonal antibodies were used anti-IKKɛ (Imgenex) and anti-PCNA (Santa Cruz).

Genome copy number assay.

Cells (ca. 10⁶ cells per sample) were collected and resuspended in 100 μl of SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris [pH 8.0]). After brief sonication, immunoprecipitation dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris [pH 8.0], 167 mM NaCl) was added to 1 ml, followed by incubation with proteinase K for 2 to 3 h at 50°C. Then, 300 μl was removed and subjected to phenol-chloroform extraction and ethanol precipitation. Precipitated DNA was then assayed by real-time PCR using primers for the DS region of EBV and normalized by the cellular DNA signal at the actin gene locus. The primers for DS were 5′-ATGTAAATAAAACCGTGACAGCTCAT-3′ and 5′-TTACCCAACGGGAAGCATATG-3′ and for actin were 5′-GCCATGGTTGTGCCATTACA-3′ and 5′-GGCCAGGTTCTCTTTTTATTTCTG-3′.

Southern blotting.

Purified genomic DNA was digested with BamHI and analyzed by 0.7% agarose gel electrophoresis and Southern blotting. EBV DNA was identified by using the OriLyt sequence amplified by the primers 5′-GCCCGTTGGGTTTCATTAAG-3′ and 5′-CCAGGAAGTGGCGAGCAT-3′. PCR products were labeled with digoxigenin using a PCR DIG probe synthesis kit (Roche) according to the manufacturer's protocol. Oligonucleotide probes for telomere repeats (6x TTAGGG) were labeled with a DIG oligonucleotide 3′ end-labeling kit (Roche).

EBV primary infection induces miR-155 and BIC RNA.

To determine whether primary infection of EBV induces cellular miRNAs, we assayed human peripheral blood cells before and after EBV infection for changes in miRNA abundance. RNA was isolated from peripheral blood cells before infection and at 10 days postinfection. As a secondary control for days in cell culture, we also assayed miRNA levels in peripheral blood cells after 10 days in culture after mock infection. A comparison of seven different donor samples revealed a clustering of at least two cellular miRNAs, miR-155 and miR-210, that were consistently induced by EBV infection relative to preinfection and mock infection controls (Fig. 1A). miR-338, miR-346, and miR-369 were induced by EBV infection relative to preinfection but not compared to mock infections (data not shown). No miRNAs were found to be consistently repressed by EBV infection in this analysis.

BIC RNA is elevated in EBV positive B-cell lines.

To begin to validate the findings from the array analysis, we focused on miR-155, which has been previously implicated in B-cell malignancies. miR-155 is known to be derived from the noncoding BIC RNA, which can be readily detected by RT-PCR analysis. RNA from the seven peripheral blood cell samples were assayed for BIC RNA levels relative to actin after EBV infection or in preinfected and mock-infected controls (Fig. 1B). EBV infection induced BIC RNA relative to actin in all seven samples, with no detectable activation in the preinfected or mock-infected controls. BIC activation levels varied for each donor sample, with samples 1, 5, and 6 showing greater than 10- to 200-fold induction by EBV. The detection of mature miR-155 was also verified by Northern blotting of total cellular RNA (Fig. 1C, upper panel). Mature miR-155 was readily detected in EBV-infected samples from donors 6 and 5, and to a lesser extent donor 4, reflecting the pattern observed for the BIC precursor RNA. EBV infection had no effect on control U6 snRNA (Fig. 1C, lower panel). The time course of BIC induction after EBV infection was assayed with a new donor sample by using RT-PCR for BIC relative to actin (Fig. 1D). We found that BIC was strongly activated by 19 days postinfection but could be detected by 8 days postinfection relative to mock-infected controls. The expression of EBV growth-transforming genes LMP1 and EBNA2 was also assayed in donor samples 4, 5, and 6 (Fig. 1E). BIC RNA levels correlated with expression levels of EBNA2 and LMP1, with donor sample 6 showing the highest levels of all three RNAs relative to actin controls. These findings suggest that EBV infection and gene products contribute to the elevated expression of BIC RNA.

BIC RNA is elevated in EBV positive B-cell lines.

EBV infects both B lymphocytes and epithelial cells. To determine whether all EBV-positive cells have elevated BIC, we assayed EBV-positive and -negative cell lines for BIC levels, relative to cellular actin (Fig. 2A). We found that EBV-positive lymphoblastoid cell lines had the highest levels of BIC expression. The EBV-positive Burkitt's lymphoma cell line Raji had a modest increase in BIC levels relative to an EBV-negative Burkitt's lymphoma cell line DG75. In contrast, BIC levels were not significantly elevated relative to actin in EBV-positive or -negative epithelial cell lines derived from 293 HEK cells (293T or ZKO-293) or a HeLa-derived cell line (D98/HR1). The levels of EBV gene products LMP1 and EBNA2 were also assayed (Fig. 2B and C). LMP1 and EBNA2 mRNA were expressed in LCL, Raji, and ZKO-293 cells, while D98/HR1 cells express LMP1 only, and 293T and DG75 cells do not express either EBV gene product. These findings suggest that BIC activation by EBV depends on cooperating B-cell-specific factors.

LMP1 induces BIC RNA in B lymphocytes.

To further explore the mechanism of BIC activation by EBV, we assayed the ability of the EBV gene products EBNA2 or LMP1 to activate BIC mRNA in EBV-negative cell lines derived from B lymphocytes or epithelial cells. We found that BIC levels were induced ∼7-fold by transfection of LMP1 in B-cell line DG75. These levels were similar to that observed for BIC RNA in EBV-positive Raji cells. EBNA2 also induced BIC, but to a lesser extent (∼3-fold). Combining EBNA2 and LMP1 did not lead to a greater induction of BIC than when LMP1 was transfected alone. Mock transfection had no effect on BIC levels. LMP1 and EBNA2 expression levels were also monitored in transfected cells (Fig. 3B and C). Although LMP1 and EBNA2 levels were readily detected in transfected cells, the levels of expression were not as high as those observed in EBV-positive Raji cells. We also assayed the ability of LMP1 and EBNA2 to activate BIC RNA in 293T cells (Fig. 3D). In contrast to the B-cell line DG75, BIC levels were not activated above basal levels in 293T cell lines. EBNA2 and LMP1 mRNA was expressed to higher levels in 293T cells than were observed in Raji cells (Fig. 3E and F). These data indicate that LMP1, and to a lesser extent EBNA2, can activate BIC RNA in DG75 B cells but not in 293T cells.

miR-155 downmodulates NF-κB signaling.

To explore the functional significance of BIC and miR-155 activation in EBV-infected B cells, we assayed the effects of an miR-155 inhibitor in EBV-positive LCL cell lines. The miR-155 inhibitor was demonstrated to efficiently activate the expression of a luciferase construct containing the miR-155 target sequence at its 3′ UTR (Fig. 4A). The miR-155 inhibitor, but not a scrambled control inhibitor, efficiently activated luciferase in LCL cell lines that are known to express high endogenous levels of miR-155 (see Fig. S1 in the supplemental material). We then tested the ability of the miR-155 inhibitor to affect several cell functions, including cell cycle profile, cellular proliferation, and apoptosis induction (data not shown). We did not detect any significant effect of miR-155 inhibitor with these functional assays. We therefore reasoned that miR-155 may play a modulatory effect on the LMP1 signaling pathways. The major LMP1 signaling target is through NF-κB. We therefore tested whether miR-155 inhibitor affected NF-κB activity in LCL cells by using a luciferase reporter construct that contains five copies of NF-κB response elements (5x κB-L). We found that miR-155 inhibitor activated 5x κB-L ∼2-fold in LCL cells (Fig. 4B). To further explore a role of miR-155 in NF-κB signaling, we assayed a miR-155 mimic in EBV-negative DG75 B cells (Fig. 4C). The miR-155 mimic reduced 5x κB-L activity by ∼5-fold, while a control miRNA mimic had no detectable effect. The miR-155 mimic had no significant effect on 5x κB-L activity relative to the control in EBV-negative 293T cell lines (Fig. 4D). This suggests that miR-155 downmodulates the relatively high levels of NF-κB activity in EBV-positive B cells but has no significant effect on low or constitutive levels of NF-κB in 293T cells.

Computational studies and luciferase indicator assays have implicated IKKɛ mRNA as a potential target for translational modulation by miR-155 (8, 25, 33). To test whether miR-155 inhibitor affected IKKɛ protein levels, we assayed LCL cells by Western blotting for expression of IKKɛ relative to control protein PCNA (Fig. 5A). We found that miR-155 inhibitor elevated IKKɛ protein levels ∼2-fold in LCL cells, with little detectable effect on the levels of PCNA. We also found that the miR-155 mimic in EBV-negative DG75 cells caused an ∼2-fold reduction in IKKɛ, with no detectable effect on PCNA. Given the fact of IKKɛ being involved in both NF-κB signaling and IFN antiviral response, we further investigated the expression of NF-κB-regulated BCL2 and IFN-regulated gene CCL5 (Fig. 5B and C). We examined the mRNA levels for IKKɛ, BCL2, and CCL5 in EBV-positive LCLs (Fig. 5B) or Raji BL cells (Fig. 5C) transfected with control or miR-155 inhibitor. RT-PCR was used to measure mRNA levels relative to cellular actin. We found that miR-155 inhibitor activated BCL2 and CCL5 mRNA levels relative to actin in both cell types, while IKKɛ mRNA levels were highly activated in Raji cells but not LCL cells. These findings are consistent with a role of miR-155 in the translational inhibition of IKKɛ and suggest that miR-155 modulates both NF-κB and IFN-responsive genes through downmodulation of IKKɛ.

The IFN-associated antiviral response may also play a role in the stable maintenance of the viral genome during latency. To explore this possibility, we first measured the effect of miR-155 inhibitor on the EBV genome copy number (Fig. 6A and B). We found that miR-155 inhibitor caused an ∼4-fold reduction in the EBV genome copy number in both Raji and LCL cells as measured by quantitative PCR of viral relative to cellular DNA. EBV genome maintenance during latency requires sufficient expression of the virus-encoded EBNA1 protein. We therefore examined whether miR-155 inhibitor reduced EBNA1 mRNA expression (Fig. 6C and D). We found that miR-155 inhibitor modestly reduced EBNA1 mRNA ∼2-fold in Raji (Fig. 6C) but only ∼1.3-fold in LCL cells (Fig. 6D) relative to control inhibitor. The loss of EBV DNA was further measured by Southern blotting after transfection of miR-155 inhibitor in Raji cells (Fig. 6E). We found an ∼2-fold loss of EBV genomes relative to telomere repeat DNA. These findings indicate that miR-155 contributes to the stable maintenance of EBV genomes during latent infection and may exert this effect, partly, through regulation of EBNA1 mRNA levels.