Phosphorylation of Epstein-Barr virus ZEBRA protein at its casein kinase 2 sites mediates its ability to repress activation of a viral lytic cycle late gene by Rta

Abstract

ZEBRA, a member of the bZIP family, serves as a master switch between latent and lytic cycle Epstein-Barr virus (EBV) gene expression. ZEBRA influences the activity of another viral transactivator, Rta, in a gene-specific manner. Some early lytic cycle genes, such as BMRF1, are activated in synergy by ZEBRA and Rta. However, ZEBRA suppresses Rta's ability to activate a late gene, BLRF2. Here we show that this repressive activity is dependent on the phosphorylation state of ZEBRA. We find that two residues of ZEBRA, S167 and S173, that are phosphorylated by casein kinase 2 (CK2) in vitro are also phosphorylated in vivo. Inhibition of ZEBRA phosphorylation at the CK2 substrate motif, either by serine-to-alanine substitutions or by use of a specific inhibitor of CK2, abolished ZEBRA's capacity to repress Rta activation of the BLRF2 gene, but did not alter its ability to initiate the lytic cycle or to synergize with Rta in activation of the BMRF1 early-lytic-cycle gene. These studies illustrate how the phosphorylation state of a transcriptional activator can modulate its behavior as an activator or repressor of gene expression. Phosphorylation of ZEBRA at its CK2 sites is likely to play an essential role in proper temporal control of the EBV lytic life cycle.

FIG. 1.

In vitro phosphorylation of ZEBRA by CK2. Purified recombinant wild-type ZEBRA (Z), Z(S167A), Z(S173A), and Z(S167A/S173A) proteins were incubated with purified recombinant CK2 in the presence of [γ-³²P]ATP. The reaction mixtures were resolved by SDS-PAGE (10% polyacrylamide), and the gel was stained with silver nitrate (B). The extent of ³²P labeling of each protein was determined by autoradiography after drying the gel (A).

FIG. 2.

Mutations at S167 and S173 alter the phosphorylation state of ZEBRA in vivo. (A) Immunoprecipitation of metabolically labeled ZEBRA. HH514-16 cells were transfected with empty vector (CMV) or expression vectors encoding wild-type ZEBRA (Z), Z(S186A), and Z(S167A/S173A/S186A). Fifteen hours after transfection, the cells were labeled with [³²P]orthophosphate or a mixture of [³⁵S]methionine and [³⁵S]cysteine. The in vivo-labeled ZEBRA proteins were immunoprecipitated and separated by SDS-PAGE (8% polyacrylamide), and the ZEBRA bands were identified by autoradiography. (B) Tryptic phosphopeptide maps of wild-type ZEBRA and the CK2 site mutants. Gel slices containing ZEBRA polypeptides were excised, and the protein was extracted and subjected to trypsin digestion. The phosphorylated peptides were detected by autoradiography after separation on TLC plates in two dimensions. The major phosphopeptides are circled and designated with the letters A to D. (C) Phosphoaminoacid analysis of in vivo-labeled ZEBRA proteins in the absence and presence of mutations at S167 and S173. A fraction of each tryptic digest was hydrolyzed with 6 N HCl and separated by electrophoresis on TLC plates in two dimensions. The radioactive phosphoamino acids were identified by virtue of comigration with nonradioactive phosphoserine, phosphothreonine, and phosphotyrosine, which were stained with ninhydrin. Pi, free phosphate; pSer, phosphoserine; pThr, phosphothreonine.

FIG. 3.

The ³⁵S-labeled tryptic peptide that contains S167 and S173 migrates at the same position as phosphopeptide D. (A) Immunoprecipitation of ZEBRA from HH514-16 cells transfected with empty vector (CMV) or expression vectors encoding Z(S186A) or Z(C171A/S186A) and labeled with [³⁵S]methionine and [³⁵S]cysteine. The immunoprecipitated products were separated on an 8% polyacrylamide-SDS gel, and the ZEBRA bands were visualized by autoradiography. (B) Tryptic peptide maps of Z(S186A) and Z(C171A/S186A) labeled with ³⁵S. The dashed circle indicates the position of the tryptic peptide that contains C171, S167, and S173.

FIG. 4.

Mutations of the CK2 phosphorylation motif in ZEBRA specifically abolish its ability to suppress activation of BLRF2 (LR2) by Rta. (A) Mutations S167A and S173A do not affect the capacity of ZEBRA to synergize with Rta in activation of EA-D in Raji cells. (B) The Z(S167A/S173A) mutant is deficient in repressing Rta activation of LR2. Raji cells were transfected with Rta and either wild-type ZEBRA (Z) or the Z(S167A/S173A) mutant. Extracts prepared at different times after transfection were assessed for expression of the LR2 product by immunoblotting. (C) Effect of single-point mutations in the CK2 site on the capacity of ZEBRA to act as a repressor of Rta. Extracts of transfected Raji cells were examined for LR2 protein 30 h after transfection. (D) Mutations at several other serine and threonine residues in its activation domain did not alter the competence of ZEBRA to repress LR2 expression. In all experiments, the membranes were immunoblotted with antibodies against EA-D, ZEBRA, Rta, and BLRF2 protein (LR2).

FIG. 5.

DRB inhibits CK2 phosphorylation of ZEBRA in vitro. Purified recombinant CK2 was incubated with 250 ng of purified bacterially expressed ZEBRA and [γ-³²P]ATP. The phosphorylation level of ZEBRA was assessed in the absence or presence of increasing concentrations of the specific CK2 inhibitor DRB. The reaction mixtures were incubated for 15 min at 37°C, and the phosphorylated proteins were separated by SDS-PAGE (10% polyacrylamide). The gel was stained with Coomassie blue (B) and dried and autoradiographed (A).

FIG. 6.

The CK2 inhibitor DRB restores the ability of Rta to activate LR2 expression in the presence of wild-type ZEBRA. Ten micrograms of plasmid DNA was transfected into 10⁷ Raji cells. The DNA contained equal amounts of Rta plus ZEBRA expression vector or Rta expression vector plus empty vector. Eight hours after transfection, the cells were treated with the CK2 inhibitor. Cells were harvested after a total period of 30 h, and extracts were prepared and analyzed by Western blotting for the indicated proteins. (A) Concentrations of DRB from 20 to 40 μM inhibit ZEBRA's ability to repress activation of LR2 by Rta. (B) Cells were harvested either 40 h (lanes 2 and 3) or 20 h (lanes 4 and 5) after transfection, and the level of LR2 expression was compared with that of cells treated with the inhibitor between h 21 and 40 (lanes 6 and 7). The numbers at the bottom of each figure represent the relative level of LR2 expression following different treatments. (The level of LR2 expression in the absence of any treatment was equal to 1.)

FIG. 7.

Inhibition of CK2-mediated phosphorylation of ZEBRA allows expression of LR2 in cells in which lytic viral DNA replication is blocked by PAA. HH514-16 cells were transfected with 5 μg of DNA containing equal amounts of Rta and ZEBRA expression vector or Rta and empty vector. The transfected cells were untreated (lanes 1, 2, and 5), treated with 0.4 mM PAA (lanes 3 and 6), or treated with 50 μM DRB together with 0.4 mM PAA (lanes 4 and 7). Cell extracts prepared 40 h after transfection were analyzed by Western blotting with antibodies against Rta, ZEBRA, and LR2 (A) or EBNA1 and FR3 (B).