Regulation of the EBNA1 Epstein-Barr virus protein by serine phosphorylation and arginine methylation

Abstract

The Epstein-Barr virus (EBV) EBNA1 protein is important for the replication and mitotic segregation of EBV genomes in latently infected cells and also activates the transcription of some of the viral latency genes. A Gly-Arg-rich region between amino acids 325 and 376 is required for both the segregation and transcriptional activation functions of EBNA1. Here we show that this region is modified by both arginine methylation and serine phosphorylation. Mutagenesis of the four potentially phosphorylated serines in this region indicated that phosphorylation of multiple serines contributes to the efficient segregation of EBV-based plasmids by EBNA1, at least in part by increasing EBNA1 binding to hEBP2. EBNA1 was also found to bind the arginine methyltransferases PRMT1 and PRMT5. Multiple arginines in the 325-376 region were methylated in vitro by PRMT1 and PRMT5, as was an N-terminal Gly-Arg-rich region between amino acids 41 and 50. EBNA1 was also shown to be methylated in vivo, predominantly in the 325-376 region. Treatment of cells with a methylation inhibitor or down-regulation of PRMT1 altered EBNA1 localization, resulting in the formation of EBNA1 rings around the nucleoli. The results indicate that EBNA1 function is influenced by both serine phosphorylation and arginine methylation.

FIG. 1.

Organization of the EBNA1 protein. A. Schematic representation of EBNA1, showing Gly-Arg-rich (GR) and Gly-Ala repeat (GA) regions. Amino acid numbers are indicated. B. Amino acid sequence of the 325-376 region, showing potentially phosphorylated serine residues in bold.

FIG. 2.

Effects of serine mutations on EBNA1 segregation function and hEBP2 binding assayed in Saccharomyces cerevisiae. A. Assay of the loss of FR-containing plasmids after 11 generations in the absence of selection for the FR plasmid. Assays were performed in the presence of EBNA1 (positive control) or an EBNA1 mutant with the indicated serine mutations or in the absence of EBNA1 (negative control). Serial dilutions of each culture were then grown on nonselective or selective plates with respect to the FR-containing plasmid to determine the number of total yeast cells present (no selection) and the fraction of the cells that contained the plasmid (selection). B. Two-hybrid assay showing the interaction of EBNA1 and EBNA1 serine mutants with hEBP2, as determined by activation of a HIS3 reporter gene. EBNA1 with empty pACT2 (-EBP2) is shown as a negative control (second row). Ten-fold serial dilutions of the cultures were grown on plates containing His or lacking His and containing 0, 2, or 5 mM AT as indicated.

FIG. 3.

Plasmid maintenance ability of EBNA1 serine mutants in human cells. A. Western blot showing the expression of EBNA1 and EBNA1 serine mutants in HeLa cells 3 days posttransfection. B. Southern blot showing the linearized oriP plasmids that were recovered from equal numbers of HeLa cells 2 weeks posttransfection with oriP plasmids expressing EBNA1, no EBNA1 (none), or the EBNA1 serine mutants indicated.

FIG. 4.

Transcriptional effects of EBNA1 serine mutants. HeLa cells were transfected with an FR-CAT reporter plasmid and pc3oriP expressing EBNA1, no EBNA1, or the EBNA1 serine mutants as indicated. Two days later, equal amounts of cell lysates were then assayed for CAT activity by following the acetylation of chloramphenicol in 5-, 15-, and 30-min reactions.

FIG. 5.

Retention of PRMT1 and PRMT5 on EBNA1 affinity columns. Equal amounts of HeLa cell lysates were applied to a 40-μl Affi-gel column resin alone (control) or coupled to purified EBNA1 or EBNA1 deletion mutants (Δ325-376 and Δ61-83). Columns were then washed, and bound proteins were eluted with high salt followed by SDS. A. SDS elutions were analyzed by SDS-PAGE and silver staining, and the indicated bands were identified by MALDI-TOF mass spectrometry. Note that some of the EBNA1 proteins themselves elute from the column with SDS. B. The column eluates in panel A were analyzed by Western blotting using antibodies specific for PRMT5 and PRMT1.

FIG. 6.

In vitro methylation of EBNA1 and EBNA1 mutants. Equal amounts of the purified EBNA1 proteins indicated or the GST-GAR positive control were incubated with S-[methyl-¹⁴C]adenosyl methionine and either PRMT1 (A and B), PRMT5 (C), PRMT3 (D), or PRMT6 (E) as described in Materials and Methods. Reaction mixtures were then analyzed by SDS-PAGE and autoradiography. In each case the major labeled band corresponds to the position of the full-length protein. Note that EBNA1 362-641 has a small fusion at the C terminus, making it run larger than 351-641 (17). In panel B, methylation reactions were conducted for various times ranging from 5 to 40 min and the amount of labeled EBNA1 was quantified. A direct comparison of methylation of GST-GAR and EBNA1 substrates by PRMT1, -3, and -6 is shown in panel F and used amounts of enzyme that gave equal methylation of GST-GAR.

FIG. 7.

In vivo methylation of EBNA1. U2OS cells expressing EBNA1, Δ34-52, Δ325-376, or Δ41-376 were incubated with cycloheximide and then with [³H]methyl-methionine. EBNA1 proteins were then immunoprecipitated with anti-EBNA1 rabbit serum and separated by SDS-PAGE. A. Western blot of immunoprecipitated EBNA1 probed with mouse anti-EBNA1 monoclonal antibody. B. Autoradiograph of the blot in panel A, showing incorporation of [³H]methyl-methionine.

FIG. 8.

Effect of MTA treatment on EBNA1 localization. (A to C) Log-phase HeLa cells expressing EBNA1 were either treated with MTA (+) or mock treated (-) and then were stained with anti-EBNA1 rabbit antiserum and anti-B23 goat antibodies (A), with mouse anti-EBNA1 monoclonal antibody and anti-hEBP2 rabbit serum (B), or with mouse anti-EBNA1 monoclonal antibody and anti-USP7 rabbit serum (C). Secondary antibodies conjugated to Texas Red, FITC, and Cy3 were used to detect mouse, rabbit, and goat primary antibodies, respectively. All cells were counterstained with DAPI and visualized at 400× magnification using similar exposure times. (D) U2OS cells were transfected with a PRMT1 silencing plasmid and then with an EBNA1 expression plasmid. Cells were stained with anti-EBNA1 and anti-PRMT1 antibodies, followed by Texas Red- and FITC-conjugated secondary antibodies, respectively. All cells were counterstained with DAPI and visualized at 400× magnification using similar exposure times. EBNA1 localization was compared in cells with little or no PRMT1 staining (-) and with more obvious PRMT1 staining (+).