Regulation of Epstein-Barr virus OriP replication by poly(ADP-ribose) polymerase 1

Abstract

Poly(ADP-ribose) polymerase (PARP) is an abundant, chromatin-associated, NAD-dependent enzyme that functions in multiple chromosomal processes, including DNA replication and chromatin remodeling. The Epstein-Barr virus (EBV) origin of plasmid replication (OriP) is a dynamic genetic element that confers stable episome maintenance, DNA replication initiation, and chromatin organization functions. OriP function depends on the EBV-encoded origin binding protein EBNA1. We have previously shown that EBNA1 is subject to negative regulation by poly(ADP-ribosyl)ation (PARylation). We now show that PARP1 physically associates with OriP in latently EBV-infected B cells. Short hairpin RNA depletion of PARP1 enhances OriP replication activity and increases EBNA1, origin recognition complex 2 (ORC2), and minichromosome maintenance complex (MCM) association with OriP. Pharmacological inhibitors of PARP1 enhance OriP plasmid maintenance and increase EBNA1, ORC2, and MCM3 occupancy at OriP. PARylation in vitro inhibits ORC2 recruitment and remodels telomere repeat factor (TRF) binding at the dyad symmetry (DS) element of OriP. Purified PARP1 can ribosylate EBNA1 at multiple sites throughout its amino terminus but not in the carboxy-terminal DNA binding domain. We also show that EBNA1 linking regions (LR1 and LR2) can bind directly to oligomers of PAR. We propose that PARP1-dependent PARylation of EBNA1 and adjacently bound TRF2 induces structural changes at the DS element that reduce EBNA1 DNA binding affinity and functional recruitment of ORC.

FIG. 1.

PARP1 interacts with the DS region of OriP in vivo. ChIP assays were performed with anti-PARP1 antibody or an IgG control using EBV-positive LCL or Mutu I cells. ChIP DNA was analyzed by quantitative PCR with primers specific for EBV regions DS or OriLyt.

FIG. 2.

PARP1 depletion enhances OriP-dependent plasmid DNA replication. (A) 293 cells were cotransfected with either shRNA targeting PARP1 or a nontargeting control and with OriP-EBNA1 containing plasmid and subjected to Western blotting. (B) Cells were assayed for transient DNA replication at 72 h posttransfection using DpnI resistance assays. DpnI/BamHI-resistant DNA (top panel) was compared to the total recovered DNA linearized by BamHI (lower panel). The ratio of the DpnI/BamHI to BamHI DNA is indicated below as the percent wild-type control replication (% wt). Experiments are shown in duplicate, and the average of four independent replicates (data not shown) was quantified.

FIG. 3.

PARP inhibitors increase EBNA1 occupancy at the DS site. (A) 293-ZKO cells were cultured in the presence or absence of 3 mM 3-AB without hygromycin selection. EBV episome maintenance was assayed by FACS analysis of the percentage of GFP-positive cells on the indicated days. The curve reflects the rate of episome loss derived from at least three independent experiments. (B to D) ChIP assays were performed with antibodies specific for EBNA1 or control IgG and analyzed using primers specific for the DS or OriLyt region in Raji cells treated with or without 3-AB (3 mM for 72 h) (B), Mutu I cells treated with or without 3-AB (3 mM for 72 h) (C), or Raji cells treated with or without PJ34 (5 μM for 72 h) (D). (E and F) ChIP assays were performed with antibodies specific for MCM3, ORC2, or control IgG and analyzed using primers specific for the DS or OriLyt region in Mutu I or Raji cells.

FIG. 4.

Increased EBNA1 binding and prereplication complex formation at OriP in PARP-depleted cells. (A) 293-ZKO cells were transfected with control or PARP1-targeting shRNA and assayed by Western blotting for depletion of PARP1 or EBNA1. (B) 293-ZKO cells transfected with control or shPARP1 expression vectors, as shown in panel A, were subjected to immunoprecipitation with anti-EBNA1 or control IgG. Immunoprecipitates were then analyzed by Western blotting with anti-PAR or anti-EBNA1. (C) ZKO-293 cells transfected with control or shPARP1 were assayed by ChIP with antibodies to EBNA1 or control IgG and analyzed by quantitative PCR at the DS and control OriLyt DNA regions. (D) The same experiment as in panel C, except that ChIP was performed with antibodies to MCM3, ORC2, or control IgG. α, anti.

FIG. 5.

Poly(ADP-ribosyl)ation alters protein binding to the DS element. (A) Raji nuclear extracts were subjected to affinity chromatography with biotinylated DS element or control BKS DNA. Purified complexes coupled with beads were then incubated with buffer control, NAD (1 mM), or ATP (1 mM). Proteins were then eluted and assayed by SDS-PAGE and Western blotting with antibodies for TRF1, hRap1, TRF2, tankyrase 1, PARP1, EBNA1, ORC2, or MCM3. The asterisk indicates a cleaved form of PARP1 that binds specifically to the DS site. (B) Purified recombinant PARP1 and EBNA1 proteins were bound to biotinylated DS element DNA immobilized on magnetic beads and then incubated with or without NAD⁺ (1 mM). Proteins were then eluted and analyzed by SDS-PAGE and Western blotting with anti-EBNA1 or anti-PAR antibody. (C) DS element DNA affinity-bound PARP1 and EBNA1 were incubated without or with NAD and then separated by centrifugation for bound (B) and unbound (U) proteins. Samples were then analyzed by SDS-PAGE and Western blotting with antibodies to EBNA1 (top panels) or anti-PAR (lower panel).

FIG. 6.

PARylation alters protein-DNA interactions at telomere repeat binding sites in the DS element. DNase I footprinting of DS element DNA using purified EBNA1, TRF1, and PARP1 with (+) or without (−) NAD, as indicated above each lane. EBNA1 binding sites (EBS) and TRF binding sites are indicated to the right. (B) Reaction mixtures were also incubated with or without 3-AB, as indicated. The asterisks indicate DNase I-hypersensitive sites induced by EBNA1, TRF1, PARP1, and NAD.

FIG. 7.

EBNA1 domains involved in PARylation and PAR binding. (A) Purified GST-EBNA1 fusion proteins were assayed for in vitro poly(ADP-ribosyl)ation in the presence of NAD (1 mM) and then analyzed by SDS-PAGE and Western blotting with anti-PAR, anti-GST or anti-PARP1 antibody. (B) Purified GST-EBNA1 fusion proteins were assayed as above in the absence of NAD. (C) Purified GST, GST-LR1, or GST-LR2 protein was incubated with ³²P-labeled PAR and assayed by EMSA. (D) Schematic of EBNA1 protein domains. Linking regions 1 and 2 (LR1 and LR2), glycine-alanine repeats (GAr), nuclear localization sequence (NLS), and the DNA binding domain (DBD) are indicated. PAR acceptor sites are indicated. (D) Schematic summary of EBNA1 domains involved in PARylation and PAR binding as determined in panels A to C.

FIG. 8.

Model of PAR-dependent nucleoprotein remodeling at the DS site. In the absence of PAR, EBNA1 can recruit ORC through interactions with LR1 and LR2, and telomere repeats are bound primarily by TRF2. PARylation induces changes in EBNA1 DNA binding and in LR1 and LR2 recruitment of ORC. PARylation also alters protein interactions with telomere repeats at the DS element. The precise binding sites of PARP1 with proteins and DNA are speculative.