Promoter sequences required for reactivation of Epstein-Barr virus from latency

Abstract

A luciferase reporter system with stably transfected oriP plasmids in Akata Burkitt's lymphoma cells provides a quantitative assay for the BZLF1 Zp promoter in response to B-cell receptor (BCR) activation by cross-linking with anti-immunoglobulin. In this system, detailed kinetic studies of promoter activity are possible. Previously reported promoter elements upstream of -221 from the transcription start and the ZIIR sequence had little effect on the Zp promoter, but the ZI and ZIIIA elements were essential for early activation. The ZIIIB element mediates autoactivation. Mutation of the ZV repressor sequence greatly increased the induction of the promoter but did not make it constitutively active. Zp transcription in response to BCR cross-linking declined after a few hours; this decline was reduced and delayed by acyclovir or phosphonoacetic acid, indicating that viral DNA replication or a late viral gene can play a role in the switch off of the Zp promoter. Late expression of the LMP1 protein may account for this.

FIG. 1.

(A) Linearized representation of the pHEBo:Zp-wt-luc plasmid, which contains a luciferase reporter gene and the Zp promoter region from +12 to −552 relative to the transcription start site. Other elements shown are oriP and the genes for hygromycin (Hyg) and ampicillin (Amp) resistance. (B) Luciferase assay of AK2000/pHEBo:Zp-wt-luc cells treated with anti-Ig. A total of 5 × 10⁶ cells per sample were treated with anti-IgG and harvested at various times postinduction (p.i.) as indicated. Cell extracts were assayed for luciferase activity, shown in relative light units (RLU). (C) Western blot of cells analyzed in panel B for BZLF1 with PCNA (proliferating nuclear antigen) as a loading control. Co, control pHEBo-luc.

FIG. 2.

(A) Relative positions of published elements in the Zp promoter. (B) Example of Southern blot of Zp mutants to determine the copy number of plasmids present in stably transfected AK2000 cell lines. Zp mutations which incorporated an additional EcoRI site at the mutation (MIIR, MIIIA, and MIIIA/B) gave short fragments of characteristic lengths. pHEBo-luc lacks Zp and gives a shorter fragment than the wild type (wt). Two lines are shown for each plasmid. (C) Luciferase assay for pHEBo:Zp-wt-luc, MIA/B, and MIC. Cells were induced with anti-IgG and harvested at the times indicated. The proportion of the different cell lines that was induced in response to anti-Ig was determined by flow cytometry of cells stained with BZ1 antibody to be 15% ± 1% (data not shown). The luciferase activity in relative light units (RLU) has been normalized for the plasmid copy number, protein concentration, and proportion of cells induced. Averages for values from two clones and independent experiments are shown. (D) Luciferase assay for pHEBo:Zp-wt-luc, MIIIA, MIIIB, and MIIIA/B. Cells were induced with anti-IgG and harvested at the times indicated. The luciferase activity was normalized and averaged as for panel C (induction frequency was 14% ± 2%). (E) Luciferase assay for MII mutant data, performed as in panel D. Co, control pHEBo-luc.

FIG. 2.

(A) Relative positions of published elements in the Zp promoter. (B) Example of Southern blot of Zp mutants to determine the copy number of plasmids present in stably transfected AK2000 cell lines. Zp mutations which incorporated an additional EcoRI site at the mutation (MIIR, MIIIA, and MIIIA/B) gave short fragments of characteristic lengths. pHEBo-luc lacks Zp and gives a shorter fragment than the wild type (wt). Two lines are shown for each plasmid. (C) Luciferase assay for pHEBo:Zp-wt-luc, MIA/B, and MIC. Cells were induced with anti-IgG and harvested at the times indicated. The proportion of the different cell lines that was induced in response to anti-Ig was determined by flow cytometry of cells stained with BZ1 antibody to be 15% ± 1% (data not shown). The luciferase activity in relative light units (RLU) has been normalized for the plasmid copy number, protein concentration, and proportion of cells induced. Averages for values from two clones and independent experiments are shown. (D) Luciferase assay for pHEBo:Zp-wt-luc, MIIIA, MIIIB, and MIIIA/B. Cells were induced with anti-IgG and harvested at the times indicated. The luciferase activity was normalized and averaged as for panel C (induction frequency was 14% ± 2%). (E) Luciferase assay for MII mutant data, performed as in panel D. Co, control pHEBo-luc.

FIG. 3.

(A) Reported negative elements in the distal part of Zp (−554 to −221). The positions of each element in Zp relative to the transcription start site are shown (not to scale). (B) Luciferase assay of Zp deletion mutant plasmids in AK2000 cell lines. Luciferase activity in relative light units (RLU) was determined as for Fig. 2C; the proportion of cells that were induced was 10% ± 2%. (C) Comparison of basal levels of Zp-luc activity in AK2000. Untreated cells from AK2000 cell lines containing the indicated plasmids were assayed for luciferase activity in relative light units (RLU), and data were normalized and averaged as for Fig. 2C. Co, control pHEBo-luc.

FIG. 4.

(A) Luciferase assay of Zp mutant plasmids in AK2000 cell lines. The proportion of cells that were induced was 14% ± 1% (data treated as for Fig. 2C). (B) Luciferase assay for pHEBo:MV-Zp-luc in AK2000 cells. The proportion of cells that were induced was 13% ± 2% (data treated as for Fig. 2C). (C) Transient-transfection assay for p294:Zp-MV-BZLF1 in AK31 cells. Cells were transfected by electroporation with 10 μg of p294:Zp-MV-BZLF1 or p294:Zp-552Xwt plasmid DNA and then incubated for 16 h to recover. Cells were then treated with anti-Ig (+Ig) or left untreated (−Ig) and harvested 24 h later. Cell extracts were analyzed by Western blotting probed with the BZ1 antibody. Transfection efficiency was measured by cotransfection of pCMV-βgal. Amounts of extract loaded on the gel were corrected for β-galactosidase activity and protein concentration. Lanes: wt, transfected with p294:Zp-wt-BZLF1; MV, transfected with p294:Zp-MV-BZLF1. Co, control pHEBo-luc.

FIG. 5.

Effect of phosphonoacetic acid (PAA) and acyclovir on Zp-luciferase induction. (A) pHEBo:Zp-wt-luc in AK2000 cells was induced with anti-Ig with or without 0.85 mM phosphonoacetic acid or 100 μg of acyclovir per ml, and samples were assayed at the indicated times for luciferase (data treated as for Fig. 2C). (B to D) Protein samples taken from the experiment shown in panel A were analyzed by Western blotting for BZLF1 (B), BdRF1 (C), or LMP1 (D). Lane BM in panel D contains an extract of the EBV-immortalized B-cell line BM+Akata, a positive control for Akata LMP1.