Mitosis-specific hyperphosphorylation of Epstein-Barr virus nuclear antigen 2 suppresses its function

Abstract

The Epstein-Barr virus (EBV) nuclear antigen 2 (EBNA-2) is a key gene expressed in EBV type III latent infection that can transactivate numerous promoters, including those for all the other type III viral latency genes as well as cellular genes responsible for cell proliferation. EBNA-2 is essential for EBV-mediated immortalization of primary B lymphocytes. We now report that EBNA-2, a phosphoprotein, is hyperphosphorylated specifically in mitosis. Evidence that the cyclin-dependent kinase p34(cdc2) may be involved in this hyperphosphorylation includes (i) coimmunoprecipitation of EBNA-2 and p34(cdc2), suggesting physical association; (ii) temporal correlation between hyperphosphorylation of EBNA-2 and an increase in p34(cdc2) kinase activity; and (iii) ability of purified p34(cdc2)/cyclin B1 kinase to phosphorylate EBNA-2 in vitro. Hyperphosphorylation of EBNA-2 appears to suppress its ability to transactivate the latent membrane protein 1 (LMP-1) promoter by about 50%. The association between EBNA-2 and PU.1 is also decreased by about 50% in M-phase-arrested cells, as shown by coimmunoprecipitation from cell lysates, suggesting that hyperphosphorylation of EBNA-2 impairs its affinity for PU.1. Finally, endogenous LMP-1 mRNA levels in M phase are around 55% of those in asynchronously growing cells. These results suggest that regulation of gene expression during type III latency may be regulated in a cell-cycle-related manner.

FIG. 1.

EBNA-2 is hyperphosphorylated specifically in mitosis. (A) Left top panel, X50-7 cells arrested by nocodazole (ND); left bottom panel, G₀/G₁, S and G₂/M phases sorted from asynchronously growing cells by FACS (brackets indicate collected fractions). Percentages in each phase are indicated. Right panel, EBNA-2 protein analyzed by immunoblotting with PE2 antibody; tubulin was the loading control. The mitotic index was determined by DAPI staining. (B) Immunoblots of EBNA-2 stably expressed in EBV-negative BJAB cells in nocodazole-arrested M phase (M) or in asynchronous cells (Asy) in whole-cell lysates. (C) Hyperphosphorylation of EBNA-2 in nocodazole-arrested M-phase cells. Immunoprecipitated EBNA-2 from M-phase and asynchronous cells was subjected to immunoblotting directly (−) or after treatment with 2,000 U of λ phosphatase (λPPase) (+). (D) EBNA-2 is hyperphosphorylated in both nocodazole-arrested and untreated mitotic cells. HeLa cells were transfected with EBNA-2 and treated with nocodazole (250 ng/ml) for 18 h, at 20 h after transfection, or were left untreated. M-phase cells were separated from other phases by vigorous shaking. Immunoblots for EBNA-2 of extracts from mitotic cells collected by shake-off from nocodazole-arrested (lane 1) and untreated (lane 3) cell monolayers and from cells remaining adherent (lanes 2 and 4) were analyzed.

FIG. 2.

(A) Physical association of EBNA-2 with p34^cdc2 kinase. Whole cell X50-7 lysates from asynchronous or nocodazole-arrested M-phase cells (A and M, respectively) were immunoprecipitated with p34^cdc2 antibody or normal mouse IgG; immune complexes were separated by electrophoresis. EBNA-2 and p34^cdc2 were detected by immunoblotting. (B) Temporal correlation between EBNA-2 hyperphosphorylation and p34^cdc2 kinase activity. X50-7 cells were continuously treated with nocodazole for the indicated times, and whole-cell lysates were separated by electrophoresis and immunoblotted with EBNA-2 antibody (upper panel). For the kinase assay, p34^cdc2/cyclin B1-kinase complex was immunoprecipitated from the same cell lysates, and kinase activity was determined by [γ-³²p]ATP incorporation into histone H1 (middle panel) and measured with a PhosphorImager (lower panel).

FIG. 3.

Involvement of p34^cdc2 kinase in EBNA-2 phosphorylation. In vitro-transcribed-translated EBNA-2 from rabbit reticulocyte lysates (A) or endogenous EBNA-2 from total lysates of asynchronously growing cells (B) were immunoprecipitated from cells and used as substrates for kinase assays in the presence (+) or absence (−) of purified p34^cdc2/cyclin B1 kinase. Immunoprecipitation with normal mouse IgG served as the negative control. After resolving by electrophoresis and transfer onto membrane, phosphorylation of EBNA-2 was detected by autoradiography (top panel). Equal levels of EBNA-2 were confirmed by immunoblotting (bottom panel).

FIG. 4.

Decreased transactivation of the LMP-1 promoter by EBNA-2 in M-phase-arrested cells. DG75 cells were transfected with reporter plasmid p(−512/+72)LMP1p-luc in addition to 5 μg of pSG5-EBNA-2 expression vector. After transfection, half the cells were arrested at M phase by nocodazole (M); the other half served as the asynchronous control (Asy). Luciferase assays were performed to evaluate promoter activity. In each phase, transactivation of the LMP-1 promoter by EBNA-2 was normalized to the vector control. (A) Comparison of transactivation of the LMP-1 promoter by EBNA-2 in M-phase-arrested and asynchronous control cells. (B) Comparison of SV40 early promoter activity in G₂/M-arrested and asynchronous control cells. Each data point in panels A and B represents the average of three independent experiments done in triplicate. Error bars represent the means ± standard errors of the means. *, P < 0.01. (C) Immunoblot of EBNA-2 with PE2 antibody in asynchronously growing (Asy) and M-phase-arrested (M) cells in a representative experiment; tubulin was used as a loading control.

FIG. 5.

Decreased association of PU.1 and EBNA-2 in M-phase-arrested cells when EBNA-2 is hyperphosphorylated. Immunoprecipitation of EBNA-2 was performed with whole-cell lysates from asynchronously growing (A) and nocodazole-arrested M-phase cells (M). A portion of the immunoprecipitated lysates was separated on a 10% gel, and immunoblotting of EBNA-2 and PU.1 was carried out on the same membrane (top and middle panels). The same samples were separated on an 8% gel to show more clearly the EBNA-2 hyperphosphorylation (bottom panel).

FIG. 6.

Decreased endogenous LMP-1 mRNA and protein levels in M-phase-arrested cells. (A) RPA was performed on total RNA from asynchronous (Asy) and M-phase-arrested X50-7 cells (M) with GAPDH and LMP-1 probes (left panel) or GAPDH and EBNA-2 probe (right panel). Yeast RNA and total RNA from DG75 cells were used as negative controls. A representative result from three independent experiments is shown. (B) Relative mRNA levels of LMP-1 and EBNA-2 were analyzed by normalizing LMP-1 (left panel) or EBNA-2 mRNA (right panel) levels to the GAPDH mRNA level with a PhosphorImager. Each data point represents the average of three independent experiments. Error bars represent the means ± standard errors of the means. *, P < 0.01. (C) Western blot of LMP-1 from total lysates of asynchronous (Asy) and M-phase-arrested X50-7 cells (M). β-actin levels were used as a loading control.

FIG. 7.

Mitotic hyperphosphorylation of EBNA-2 and its possible functional consequences. In the interphase, p34^cdc2 kinase is inactive due to the low level of cyclin B1, its regulatory subunit, and phosphorylation of p34^cdc2. EBNA-2 is hypophosphorylated in the interphase and transcriptionally active. Upon entering the M phase of the cell cycle, p34^cdc2 kinase becomes activated by accumulation of cyclin B1 and dephosphorylation of p34^cdc2. EBNA-2 is hyperphosphorylated during mitosis by p34^cdc2 kinase directly (1) and probably by other kinases as well (2). Association of EBNA-2 with PU.1 decreases, which may be one of the mechanisms whereby transactivation of the LMP-1 promoter by EBNA-2 is impaired. In general, the suppression of EBNA-2 transcriptional activity by hyperphosphorylation may repress all the other EBNA-2-responsive genes during the M phase of the cell cycle.