Different distributions of Epstein-Barr virus early and late gene transcripts within viral replication compartments

Abstract

Productive replication of the Epstein-Barr virus (EBV) occurs in discrete sites in nuclei, called replication compartments, where viral genome DNA synthesis and transcription take place. The replication compartments include subnuclear domains, designated BMRF1 cores, which are highly enriched in the BMRF1 protein. During viral lytic replication, newly synthesized viral DNA genomes are organized around and then stored inside BMRF1 cores. Here, we examined spatial distribution of viral early and late gene mRNAs within replication compartments using confocal laser scanning microscopy and three-dimensional surface reconstruction imaging. EBV early mRNAs were mainly located outside the BMRF1 cores, while viral late mRNAs were identified inside, corresponding well with the fact that late gene transcription is dependent on viral DNA replication. From these results, we speculate that sites for viral early and late gene transcription are separated with reference to BMRF1 cores.

Fig 1

RNAPII becomes localized inside BMRF1 cores at late stages of productive replication. Tet-BZLF1/B95-8 cells were treated with doxycycline to induce lytic replication and harvested at the indicated postinduction times. After treatment with mCSK buffer, they were fixed. (A) Harvested cells were stained with anti-BMRF1 (green), anti-BALF2 (red), and anti-RNAPII (yellow) antibodies and observed by laser scanning confocal microscopy. The 2D images show brightest-point projections. Lower panels are merged images of the indicated combinations of the proteins. Pol II, RNA polymerase II. (B) Lytic replication-induced cells were stained with anti-BMRF1 (green) and anti-RNAPII (red). Projections of 60 images collected at 0.33-μm steps in the z axis are displayed as 3D topographical reconstruction images of BMRF1 and RNAPII (left and middle panels, respectively). Representative 3D surface reconstruction images are presented. Right panels are merged 3D surface reconstruction images. (C) The ratio of cells in which RNAPII was located outside the BMRF1 core to total BMRF1-positive cells was determined (black bar). White bars show the ratios of cells in which RNAPII was located inside the BMRF core to total BMRF1-positive cells. We regard cells as inside when 50% or more of RNAPII signal was detected inside the core. More than 100 cells were counted at each of the indicated time points.

Fig 2

BALF4, BLLF1, and BcLF1 are late (L) genes. (A) Lytic replication-induced tet-BZLF1/B95-8 cells were cultured in the absence or presence of PAA (phosphonoacetic acid) and harvested at 48 h postinduction. Total RNAs were purified using TriPure isolation reagent and applied for real-time RT-PCR as described in Materials and Methods. Levels of BMRF1, BALF4, BcLF1, and BLLF1 mRNAs were normalized to GAPDH mRNA levels. The value obtained in the absence of PAA was set to 100%, and data are expressed as means ± standard deviations (SD) from 3 biological replicates. dox, doxycycline. (B) Expression profiles for mRNA levels of EBV early (E) and L genes. Lytic replication-induced tet-BZLF1/B95-8 cells were harvested at 0, 24, and 48 h postinduction. Total RNAs were purified using TriPure isolation reagent and applied for real-time RT-PCR as described in Materials and Methods. Levels of BMRF1, BALF5, BALF4, BcLF1, and BLLF1 mRNAs were normalized to GAPDH mRNA levels. The value obtained at 0 h postinduction was set to 1, and data are expressed as means ± SD from 3 biological replicates.

Fig 3

Alteration of viral gene transcription foci at early and late stages of EBV lytic replication. Lytic replication-induced tet-BZLF1/B95-8 cells were hybridized with DIG-labeled RNA FISH probes and then exposed to anti-BMRF1 antibodies (magenta), anti-DIG antibodies (green), and DAPI (blue). The data are presented as 3D reconstruction images (lower panels; merged images are at the right) and corresponding 2D images (upper panels). The 2D images show brightest-point projections of 60 images collected at 0.33-μm steps on the z axis. (A) Lytic replication-induced tet-BZLF1/B95-8 cells were harvested at 0 h postinduction and hybridized with antisense mRNA probes for BMRF1 (upper panels) and BALF4 (lower panels), followed by exposure to anti-BMRF1 monoclonal antibodies (magenta). (B and C) Lytic replication-induced tet-BZLF1/B95-8 cells were harvested at 24 h postinduction and hybridized with antisense mRNA probes for BMRF1 (B) and BALF5 (C), followed by exposure to anti-BMRF1 monoclonal antibodies (magenta). More than 30 cells were counted, and the ratio of the cells in which FISH signals were located outside the BMRF1 core to total BMRF1-positive cells was determined (circle chart, black). (D to F) Lytic replication-induced tet-BZLF1/B95-8 cells were harvested at 48 h postinduction and hybridized with antisense mRNA probes for BALF4 (D), BcLF1 (E), and BLLF1 (F), followed by exposure to anti-BMRF1 monoclonal antibodies (magenta). x-z 2D merged images, which show brightest-point sections in the y axis, are also shown. More than 30 cells were counted, and the ratio of the cells in which FISH signals were located inside the BMRF1 core to total BMRF1-positive cells was determined (circle chart, gray). (B′ to F′) As negative controls, cells were hybridized with a sense RNA probe for BMRF1 (B′), BALF5 (C′), BALF4 (D′), BcLF1 (E′), and BLLF1 (F′), followed by exposure to anti-BMRF1 monoclonal antibodies (magenta).

Fig 4

BcRF1, the EBV late gene transactivator, is located inside the BMRF1 core. Tet-BZLF1/B95-8 cells were transfected with HA-BcRF1 expression vector and then treated with doxycycline to induce lytic replication. Cells were harvested at 48 h postinduction, treated with mCSK buffer, fixed, and stained with anti-BMRF1 (green) and anti-HA (red) antibodies. (A) The data are presented as 3D surface reconstruction images of BMRF1 and HA-BcRF1 (left and middle panels). A merged 3D surface reconstruction image is also presented (right panel). (B) The data are presented as serial x-y, x-z, and y-z 3D surface reconstruction images (3D) and corresponding 2D images (2D). The 2D images show brightest-point sections in the x, y, or z axis.