EBNA3C coactivation with EBNA2 requires a SUMO homology domain

Abstract

Epstein-Barr virus (EBV) nuclear antigen 3C (EBNA3C) is critical for EBV immortalization of infected B lymphocytes and can coactivate the EBV LMP1 promoter with EBNA2. EBNA3C amino acids 365 to 545 are necessary and sufficient for coactivation and are required for SUMO-1 and SUMO-3 interaction. We found that EBNA3C but not EBNA3CDelta343-545 colocalized with SUMO-1 in nuclear bodies and was modified by SUMO-2, SUMO-3, and SUMO-1. EBNA3C amino acids 545 to 628 and amino acids 30 to 365 were also required for EBNA3C sumolation and nuclear body localization but were dispensable for coactivation, indicating that EBNA3C sumolation is not required for coactivation. Furthermore, EBNA3C amino acids 476 to 992 potently coactivated with EBNA2 but EBNA3C amino acids 516 to 922 lacked activity, indicating that amino acids 476 to 515 are critical for coactivation. EBNA3C amino acids 476 to 515 include DDDVIEV(507-513), which are similar to SUMO-1 EEDVIEV(84-90). EBNA3C m1 and m2 point mutations, DDD(507-509) mutated to AAA and DVIEVID(509-513) mutated to AVIAVIA, respectively, diminished SUMO-1 and SUMO-3 interaction in directed yeast two-hybrid and glutathione S-transferase pulldown assays. Furthermore, EBNA3C m1 and m2 did not coactivate the LMP1 promoter with EBNA2. Overexpression of wild-type SUMO-1, SUMO-3, and the SUMO-conjugating enzyme UBC9 coactivated the LMP1 promoter with EBNA2. Since EBNA2 activation is dependent on p300/CBP, the possible effect of EBNA3C on p300-mediated transcription was assayed. EBNA3C potentiated transcription of p300 fused to a heterologous DNA binding domain, whereas EBNA3C m1 and m2 did not. All of these data are consistent with a model in which EBNA3C upregulates EBNA2-mediated gene activation by binding to a sumolated repressor and inhibiting repressive effects on p300/CBP and other transcription factor(s) at EBNA2-regulated promoters.

FIG. 1.

EBNA3C colocalizes with GFP-SUMO-1 in nuclear bodies and disperses endogenous PML. HeLa cells were transfected with 2 μg of EBNA3C (A to D, H to J) and/or GFP-SUMO-1 (B to G) expression vectors. Endogenous PML and transfected EBNA3C were visualized by staining with anti-PML rabbit polyclonal antibody and A10 EBNA3C monoclonal antibody, followed by fluorophore-conjugated secondary antibody.

FIG. 2.

EBNA3C and EBNA3B are modified by SUMO-3, but EBNA3A is not. 293-T cells were transfected with 4 μg of HA-tagged SUMO-3 expression plasmid and 4 μg of EBNA3A, EBNA3B, or EBNA3C expression plasmid. After 48 h HA-tagged proteins were immunoprecipitated (HA ip) from clarified cell lysates with HA antibody. Immunoprecipitates were separated on 8% polyacrylamide gels, transferred to nitrocellulose, and blotted with EBV-immune human serum, which recognizes all the EBNA proteins. The square bracket indicates the position of SUMO-3-modified EBV nuclear proteins, while the asterisk indicates the position of unmodified protein.

FIG. 3.

Modification by SUMO-2 and SUMO-3 polymers requires EBNA3C amino acids 343 to 545. Cell extracts were prepared from 293-T cells transfected with 2 μg of empty pSG5 vector (lane 1) or 2 μg of plasmid expressing either wild-type (WT) EBNA3C (lanes 2 to 5) or EBNA3C Δ343-545 (lanes 6 to 9). In addition, 2 μg of expression vectors encoding HA-tagged SUMO-1 (S1, lanes 3 and 7), SUMO-2 (S2, lanes 4 and 8), and SUMO-3 (S3, lanes 5 and 9) and pSG5 vector (V, lanes 1, 2, and 6) were cotransfected where indicated. Proteins were immunoprecipitated with A10 EBNA3C antibody, resolved on a sodium dodecyl sulfate-6% polyacrylamide gel, transferred to nitrocellulose, and immunoblotted with HA and EBNA3C antibodies.

FIG. 4.

EBNA3C amino acids 185 to 365 and 545 to 628 are required for SUMO-3 modification. 293-T cells were transfected with wild-type or mutant EBNA3C expression vectors and with HA-SUMO-3 expression vector and empty pSG5 vector (V). HA-SUMO-3 modification of EBNA3C was assessed by HA antibody-specific immunoprecipitation and immunoblot for the indicated EBNA3C deletion mutants with the EBNA3C-specific A10 monoclonal antibody. A 3% input (3% in) was loaded adjacent to the HA immunoprecipitation (ip).

FIG. 5.

EBNA3C sumolation is required for assembly into nuclear dots but not for LMP1 promoter coactivation with EBNA2. (A) HeLa cells were transfected with 4 μg of plasmid DNA encoding wild-type EBNA3C (1 to 992) or EBNA3C deletion mutants. EBNA3C was detected with the EBNA3C A10 monoclonal antibody and Texas Red-conjugated anti-mouse immunoglobulin secondary antibody. (B) BJAB cells were transfected with10 μg of the (−512/+72) LMPp-Luc reporter plasmid, 1 μg of vector (V), or 1 μg of EBNA2 (E2) expression vector and with 2 μg of the indicated EBNA3C deletion mutant plasmid. Transfection efficiency was normalized by cotransfected β-galactosidase expression, and results are reported as activation above that with EBNA2 alone. A representative experiment is shown in which the mean of three replicates was calculated ± standard error.

FIG. 6.

EBNA3C amino acids 476 to 516 are critical for EBNA3C coactivation with EBNA2. BJAB cells were transfected with 10 μg of the p(−512/+72)LMP-Luc reporter, 1 μg of EBNA2 (E2) expression vector, or 2 μg of wild-type EBNA3C, the indicated EBNA3C deletion mutant, or vector control plasmid (V). Activity above that with EBNA2 alone after β-galactosidase normalization is indicated. The average and standard error of two replicates from two independent experiments are reported. The panel below shows an immunoblot with A10 EBNA3C-specific monoclonal antibody.

FIG. 7.

EBNA3C has colinear homology with SUMO-1. B95-8 (type I) EBNA3C and human SUMO-1 and SUMO-3 amino acid sequences are shown. Residues conserved between human and herpesvirus papio EBNA3C are underlined.

FIG. 8.

EBNA3C amino acids 507 to 515 are important for SUMO-3 binding. (A) Wild-type EBNA3C and deletion mutants were expressed in 293-T cells. After 24 h, the cells were lysed in GST binding buffer. Lysates were precleared with GST beads and then incubated with either GST or GST-SUMO-3 beads for 1 h at 4°C. The beads were washed three times with binding buffer. Bound proteins were eluted, resolved on an 8% polyacrylamide gel, transferred to nitrocellulose, and immune blotted with rabbit polyclonal anti-EBNA3C antibody (ExAlpha Corporation). (B) The EBNA3C and EBNA3C m1 and m2 mutants were tested in the same assay with A10 EBNA3C-specific monoclonal antibody.

FIG. 9.

EBNA3C amino acids 507 to 515 are critical for coactivation of the LMP1 promoter with EBNA2. (A) BJAB cells were transfected with 10 μg of (−512/+72) LMPp-Luc reporter construct, 1 μg of EBNA2 (E2) expression vector, or 1 μg of wild-type and mutant EBNA3C expression vector. Activity above that with EBNA2 is indicated after β-galactosidase normalization. The average and standard error of duplicates in a representative experiment are reported. (B) EBNA3C and EBNA3C mutants were tested for repression of EBNA2 activity with a promoter with eight upstream copies of the Cp RBP-Jκ binding site. Quantities of transfected plasmids are the same as in A except that 10 μg of EBNA3C expression plasmids and 2 μg of EBNA2 expression plasmid were used in the repression assay. Results are the average of two replicates from a representative experiment. Expression of EBNA3C and EBNA3C mutants was detected with the A10 EBNA3C monoclonal antibody.

FIG. 10.

SUMO-1, SUMO-3, and UBC9 can coactivate with EBNA2 at the LMP1 promoter. Coactivation does not depend on SUMO-3 modification. (A) LMP1 promoter (−512/+72) LMPp-Luc reporter was transfected with β-galactosidase into BJAB cells with 10 μg of empty vector (V) or plasmid expressing SUMO-1 (S1), SUMO-3 (S3), or UBC9 (U9) alone and with 1 μg of EBNA2 (E2) expression vector. Luciferase values were normalized for cotransfected β-galactosidase reporter. Activity above that with the vector alone is shown. Results are representative of three independent experiments. (B) Luciferase reporter assays were done as in A with 10 μg of cotransfected SUMO-3 (S3), SUMO-3 K₁₁R (S3 KR), or SUMO-3 with terminal diglycines mutated to alanines (S3 GGAA). Luciferase above that with EBNA2 alone is shown. Values represent the average and standard error of duplicates from a representative experiment.

FIG. 11.

EBNA3C SUMO homology domain is required for EBNA3C coactivation with p300, but p300 sumolation in the cell cycle regulatory domain is not required. (A) BJAB cells were transfected with a plasmid containing multimerized papillomavirus E2 binding sites upstream of a promoter and the luciferase reporter (5 μg), with pGK-βgal expression vector control, with 5 μg of vector (V) and E2-p300, and with the indicated amounts of EBNA3C expression plasmid. The mean activity above E2-p300 activity after β-galactosidase normalization of duplicates from a representative experiment is reported. Expression levels from one of the two replicates for wild-type EBNA3C and mutant EBNA3C transfection are shown in the immune blot with A10 antibody. (B) Same as A except that 2 μg of E2-p300 and E2-p300 K₁₀₂₀R, K₁₀₂₄R and 5 μg of EBNA3C expression plasmid and vector were cotransfected where indicated. Values are normalized relative to that with the reporter alone (V).