Epstein-Barr Virus Viral Processivity Factor EA-D Facilitates Virus Lytic Replication by Inducing Poly(ADP-Ribose) Polymerase 1 Degradation

ABSTRACT

Gammaherpesviruses, including Epstein-Barr virus (EBV), are important human pathogens because they are associated with various tumors. Poly(ADP-ribose) polymerase 1 (PARP1) is a multifunctional host nuclear protein responsible for poly(ADP-ribosyl)ation (PARylation) of target proteins. While PARP1 acts as a negative regulator that suppresses the lytic replication of gammaherpesviruses, viruses are often equipped with various strategies to overcome PARP1 inhibition. However, the mechanisms of how EBV may modulate a repressive host protein, PARP1, are still elusive. In this study, we found that EBV reactivation induced PARP1 downregulation in EBV-infected cells. EBV DNA polymerase processivity factor EA-D, encoded by the BMRF1 gene, directly interacted with the central automodification domain (AD) of PARP1 and was necessary and sufficient to downregulate PARP1 via K29-linked polyubiquitination. Moreover, knockdown of EA-D in B95.8 cells restored PARP1 levels and abrogated the expression of ZTA (also known as ZEBRA), a switch molecule of the EBV life cycle during reactivation. Interestingly, PARP1 PARylated RTA, another key switch molecule, and decreased RTA transactivation on the promoters of the ZTA, BMRF1, and BMLF1 genes. EA-D alleviated the PARylation of RTA and further enhanced RTA-mediated transactivation of these lytic promoters in reporter assays. Taken together, our results suggest that EBV viral processivity factor plays a key role in facilitating lytic replication by inducing PARP1 degradation via its interaction with the PARP1 AD, which is a highly conserved mechanism among gammaherpesviruses to counteract host repressive activity of PARP1 against viral lytic replication.

IMPORTANCE PARP1 acts as a negative regulator of lytic replication in EBV. To successfully enter the reactivation cycle, EBV has developed multiple strategies to counteract the host’s repressive mechanisms. In this study, we investigated how EBV manipulated the host repressive factor PARP1 to facilitate lytic replication. The EBV processivity factor EA-D downregulated PARP1 in a proteasome-dependent manner via its direct binding with PARP1 AD. The knockdown of EA-D restored the PARP1 level and inhibited ZTA expression during reactivation. Interestingly, PARP1 PARylated RTA and EA-D reduced the PARylation of RTA, thereby promoting the ZTA promoter activity. These results suggest that EA-D plays a key role in EBV lytic replication by inducing PARP1 degradation in addition to supporting DNA replication as a viral processivity factor. Given that the KSHV processivity factor also induces PARP1 degradation and enhances RTA function, gammaherpesviruses share a conserved molecular mechanism to overcome the inhibitory effects of PARP1, promoting lytic replication.

INTRODUCTION

Gammaherpesviruses, such as Epstein-Barr virus (EBV), Kaposi’s sarcoma-associated herpesvirus (KSHV), and murine gammaherpesvirus 68 (MHV-68), cause lifelong persistent infections due to their latency and often share molecular mechanisms to regulate their life cycles (1–3). EBV, first isolated from Burkitt’s lymphoma, is associated with various tumors, including Hodgkin’s lymphoma, gastric carcinoma, and nasopharyngeal carcinoma (4–7). Although most tumors are infected with latent EBV, reactivation is also considered an important mechanism by which the viral reservoir is maintained in the infected host. There are two switch molecules, ZTA (also known as Z EBV replication activator, ZEBRA, encoded by BZLF1) and RTA (also known as R, encoded by BRLF1), that are indispensable for EBV reactivation from latency to lytic replication and serve as important transactivators (8–11). When ZTA recognizes the EBV lytic origin (oriLyt), it starts to recruit the viral replication complex, which permits the virus to enter the lytic cycle (12).

PARP1 is an abundant NAD⁺-dependent multifunctional nuclear enzyme responsible for the posttranslational modification of target proteins by consuming NAD⁺ moieties, called poly(ADP-ribosyl)ation (PARylation) (13–16). PARP1 modulates the DNA damage response, cell death, chromatin remodeling, transcription regulation, inflammation, and even tumorigenesis (13–16). PARP1 has also been shown to regulate virus infections, and especially in gammaherpesviruses, it acts as a negative regulator of lytic replication. PARP1 restricts EBV lytic replication by both catalyzing the PARylation of the CCCTC-binding factor (CTCF) insulator protein and binding to the BZLF1 promoter (16–18). Moreover, in KSHV and MHV-68, PARP1 induces the PARylation of KSHV RTA (kRTA) and MHV-68 RTA (mRTA), respectively, thereby inhibiting the kRTA and mRTA transactivation of downstream genes (19–21). Gammaherpesviruses have developed diverse mechanisms to overcome the repressive activities of PARP1 and successfully enter the lytic cycle (22). EBV ZTA overexpression was found to downregulate PARP1 activity (17). The viral PARP1-interacting proteins (vPIPs) of KSHV and MHV-68, a tegument protein encoded by ORF49, sequester PARP1 via direct binding (23). In addition, the DNA processivity factor of KSHV and MHV-68, encoded by ORF59, induces the downregulation of PARP1 in a proteasome-dependent manner via CHFR, a cellular ubiquitin E3 ligase (19, 24). These viral factors contribute to the promotion of viral replication at different stages of the life cycle.

EBV processivity factor EA-D, which is encoded by BMRF1, plays a significant role in virus genome replication as a basic component of the viral DNA replication complex (25). In addition, EA-D was shown to activate the BHLF1 promoter, thereby facilitating lytic replication (25–30). In this study, we found that there was PARP1 downregulation upon reactivation in EBV-positive cells for the first time and studied the role of EBV EA-D in PARP1 degradation and EBV lytic replication. Our results suggest a highly conserved molecular mechanism among gammaherpesviruses where the inhibitory effects of the host factor PARP1 are ablated via interaction with a viral processivity factor to facilitate virus replication.

RESULTS

EBV EA-D induced the downregulation of PARP1 in a proteasome-dependent manner.

To better understand how EBV modulates PARP1 during reactivation, EBV latently infected cells were treated with 12-O-tetradecanoylphorbol-13-acetate (TPA) and sodium butyrate (NaB) for 24 and 48 h and analyzed for PARP1 expression. In the B95.8 cells, EBV reactivation induced PARP1 downregulation over time (Fig. 1A). Consistent with this, viral reactivation also downregulated PARP1 in SNU-719, an EBV-positive gastric adenocarcinoma cell line (Fig. 1B). Unlike the EBV-infected cells, the level of PARP1 was not affected in the EBV-negative BJAB cells treated with TPA (Fig. 1C). These results demonstrate that EBV reactivation induces PARP1 downregulation in EBV-positive cells.

FIG 1.

EBV reactivation induced PARP1 downregulation in a proteasome-dependent manner. (A to C) EBV-positive B95.8 cells (A) and SNU-719 cells (B) were treated with 20 ng/mL TPA and 3 mM NaB and harvested at the indicated time points (A) or after 48 h (B), and EBV-negative BJAB cells (C) were treated with 20 ng/mL TPA and harvested at the indicated time points. The cell lysates were analyzed with Western blots probed with anti-PARP1, anti-EA-D, and anti-α-tubulin. (D) HEK293T cells were transfected with FLAG-EBV EA-D and harvested 48 h posttransfection (hpt). The cell lysates were analyzed with Western blots probed with anti-PARP1, anti-FLAG, and anti-α-tubulin. (E) HeLa cells were transfected with vector alone, FLAG-EBV EA-D, or FLAG-KSHV PF-8 and immunostained with primary antibodies using anti-PARP1 and anti-FLAG at 48 hpt. The proteins were stained with a secondary antibody using Cy3 (PARP1 [red]) and FITC (FLAG [green]), and the nuclei were stained with DAPI (blue). FLAG-KSHV PF-8, with the known ability to induce PARP1 degradation, was included as a positive control. The scale bar indicates 5 μm. (F) HEK293T cells were transfected with vector alone, FLAG-EBV EA-D, or FLAG-KSHV PF-8. At 18 hpt, the medium was changed, and the cells were treated with 1 μM MG132 for an additional 6 h. The cells were harvested and subjected to Western blotting, and the blots were probed with anti-PARP1, anti-FLAG, and anti-α-tubulin. The numbers below the PARP1 blot indicate the relative levels of PARP1.

We previously showed that the viral processivity factors of KSHV and MHV-68 (PF-8 and mPF-8, respectively) induced PARP1 degradation in a proteosome-dependent manner (19, 24). Thus, we set out to examine whether the viral processivity factor of EBV (EA-D) may also downregulate PARP1 levels. The overexpression of EA-D alone was sufficient to decrease PARP1 levels in HEK293T cells (Fig. 1D). When HeLa cells were transfected with the FLAG-EA-D plasmid, the immunofluorescence assay (IFA) results showed that the subcellular localization of EA-D and PARP1 overlapped within the nucleus and that PARP1 downregulation occurred in cells expressing EA-D (Fig. 1E). KSHV PF-8 was included as a positive control, and its results were consistent with previous findings (19, 24). To test whether the degradation of PARP1 was proteasome dependent, MG132, a 26S proteasome complex inhibitor, was incubated with EA-D- or PF-8-transfected cells (Fig. 1F). MG132 treatment restored the downregulation of PARP1 induced by EA-D and PF-8, suggesting that EBV EA-D degrades the PARP1 protein in a proteasome-dependent manner, which is similar to KSHV PF-8 (Fig. 1F).

EA-D induced PARP1 degradation via K29-linked polyubiquitination.

To examine the ubiquitin patterns of PARP1, HEK293T cells were cotransfected with hemagglutinin (HA)-tagged ubiquitin (Ub) and FLAG-EA-D plasmids and subsequently subjected to coimmunoprecipitation (co-IP) assays using the anti-PARP1 antibody. The results showed that EA-D induced the polyubiquitination of PARP1 (Fig. 2A). We further explored which lysine residues of ubiquitin may be related to the polyubiquitin pattern of PARP1 (Fig. 2B and E). As the ubiquitin linkage to the Lys 48 residue (K48) is the linkage type most associated with inducing the degradation of target proteins through the proteasome pathway (31, 32), the association of K48-Ub with PARP1 polyubiquitination was examined. The co-IP results showed that EA-D failed to induce the ubiquitination of PARP1 in the presence of the K48-Ub construct (Fig. 2B). Furthermore, K63-linked ubiquitination of PARP1 was not detected either (Fig. 2C), implying that Ub residues other than K48 and K63 might be related to PARP1 polyubiquitination. When atypical Ub linkages such as K11 or K29 were tested, PARP1 was polyubiquitinated by the K29-Ub construct, but not the K11-Ub or K29R-Ub construct, suggesting that PARP1 polyubiquitination was K29 linkage dependent (Fig. 2D and E). To avoid any effects of EA-D-induced PARP1 degradation on the ubiquitination levels of PARP1, the cells were treated with MG132 following transfection with wild-type (WT)-, K29-, or K29R-Ub constructs in the absence and presence of FLAG-EA-D. The co-IP results were consistent in that PARP1 was polyubiquitinated via K29-Ub (Fig. 2F). Taken together, our results demonstrate that EA-D induced K29-linked polyubiquitination of PARP1, targeting it for proteasome-dependent degradation.

FIG 2.

EA-D induced K29-dependent polyubiquitination of PARP1. (A to E) HEK293T cells were cotransfected with vector alone or FLAG EA-D construct as well as the indicated pRK5-HA-ubiquitin constructs as follows: WT-HA-Ub (A), K48-HA-Ub (B), K63-HA-Ub (C), K11- or K11R-HA-Ub (D), and K29- or K29R HA-Ub (E). The transfected cells were then harvested using the co-IP buffer at 48 hpt. The cell lysates were then pulled down using anti-PARP1 antibody, followed by Western blotting with the blots probed with anti-HA, anti-PARP1, anti-FLAG, and anti-α-tubulin. (F) HEK293T cells were cotransfected with vector alone or the FLAG EA-D construct and the indicated HA-tagged ubiquitin construct (K29 or K29R HA-Ub). At 24 hpt, the cells were treated with 1 μM MG132 for an additional 24 h. The cells were harvested and analyzed with Western blots probed with anti-HA, anti-PARP1, anti-FLAG, and anti-α-tubulin. The numbers below the PARP1 blot indicate the relative levels of PARP1.

EA-D interaction with the automodification domain of PARP1 is critical for PARP1 degradation and efficient lytic replication.

Next, we examined whether EA-D interacts with PARP1 using co-IP assays in HEK293T cells transfected with FLAG-EA-D. KSHV PF-8, which is known to bind and degrade PARP1, was included as the positive control. The reciprocal co-IP results using anti-FLAG and anti-PARP1 antibodies showed that EBV EA-D directly interacted with the endogenous PARP1 (Fig. 3A and B). Interaction of viral EA-D with endogenous PARP1 was also detected in EBV-positive B95.8 cells during reactivation (Fig. 3C). PARP1 consists of three major domains: the DNA binding domain (DBD), the automodification domain (AD) containing the BRCT motif, and the catalytic domain (CAT) (33, 34) (Fig. 3D). To identify the domains of PARP1 responsible for the interaction with EA-D, HEK293T cells were cotransfected with MYC-EA-D and FLAG-tagged domain-only mutants of PARP1 and subjected to co-IP assays using anti-FLAG. The results showed that the AD of PARP1 was sufficient to interact with EA-D (Fig. 3E). Reciprocal co-IP results using the anti-MYC antibody also showed that EA-D interacted only with PARP1 AD (Fig. 3F). In addition, the PARP1 AD deletion (ΔAD) mutant failed to interact with EA-D, as shown in reciprocal co-IP assays (Fig. 3G and H). These results suggest that PARP1 AD is necessary and sufficient to interact with EA-D. To further investigate the functional significance of EA-D interaction with PARP1 AD, B95.8 cells were transfected with the PARP1 WT or ΔAD mutant, followed by reactivation (Fig. 3I). While the PARP1 WT was able to compensate for PARP1 degradation and inhibited viral lytic replication, the PARP1 ΔAD mutant did not restore the PARP1 levels and showed no effects on reactivation, suggesting the importance of EA-D interaction with PARP1 AD in PARP1 degradation to facilitate lytic replication. Taken together, these results demonstrate that EA-D directly interacts with PARP1 via the AD of PARP1, which is critical for PARP1 degradation and efficient lytic replication of EBV.

FIG 3.

EA-D direct interaction with the automodification domain of PARP1 is critical for PARP1 degradation and EBV lytic replication. (A and B) HEK293T cells were transfected with FLAG-EBV EA-D or FLAG-KSHV PF-8 for 48 h and harvested using the co-IP buffer. The cell lysates were then pulled down using anti-PARP1 (A) or anti-FLAG (B) antibody and then analyzed with Western blots with anti-PARP1, anti-FLAG, or anti-α-tubulin. (C) EBV-positive B95.8 cells were treated with 20 ng/mL TPA and 3 mM NaB for 48 h and then harvested using a co-IP buffer. The cell lysates were then pulled down using anti-PARP1 antibody and then analyzed with Western blots probed with anti-PARP1, anti-EA-D, and anti-α-tubulin. The numbers below the PARP1 blot indicate the relative levels of PARP1. (D) Schematic map of PARP1 domains. The starting points of each domain are represented by the amino acid number below the map. (E to H) HEK293T cells were transfected with the indicated PARP1 constructs (WT, CAT, AD, DBD, or ΔAD mutant) and harvested using the co-IP buffer at 48 hpt. The cell lysates then were pulled down using anti-FLAG antibody (E and G) or anti-MYC antibody (F and H), conjugated with protein A/G agarose beads, and then analyzed with Western blots probed with anti-FLAG, anti-MYC, and anti-α-tubulin. (I) B95.8 cells were transfected with vector, FLAG-PARP1 WT, or the ΔAD mutant (1 μg), and then treated with 20 ng/mL TPA and 3 mM NaB at 24 hpt for an additional 24 h. The cell lysates were analyzed with Western blots probed with anti-PARP, anti-FLAG, anti-EA-D, anti-ZTA, and anti-α-tubulin.

EA-D enhances ZTA expression by inhibiting RTA PARylation to facilitate lytic replication.

To examine the role of EA-D-mediated PARP1 degradation during lytic replication, EA-D knockdown B95.8 cells were generated by transducing lentiviruses expressing short-hairpin RNAs for EA-D (shEA-D) or scrambled short-hairpin RNAs (shControl) (Fig. 4). When four independent shEA-D constructs (shEA-D #1 to #4) were tested for EA-D knockdown efficiency in B95.8 cells, shEA-D #1 and shEA-D #3 successfully abolished EA-D expression, suggesting off-target effects of shRNAs are unlikely (Fig. 4A). In these EA-D knockdown B95.8 cells, PARP1 was not degraded, but stabilized during reactivation, confirming that EA-D is critical for PARP1 degradation (Fig. 4B). Moreover, viral ZTA expression in shEA-D cells was abrogated, unlike shControl cells, suggesting that the gene coding for EA-D, an early gene, may act as a positive regulator for expression of ZTA, an immediate early gene (Fig. 4A and B).

FIG 4.

EA-D knockdown restored PARP1 and abrogated ZTA expression during lytic replication. (A) Four independent shRNA constructs targeting EA-D (shEA-D #1 to #4), together with a control shRNA of scrambled sequences (shControl), were tested for the EA-D knockdown efficiency in transduced B95.8 cells following treatment with 20 ng/mL TPA and 3 mM NaB for 24 h. Western blots probed with anti-EA-D, anti-ZTA, and anti-α-tubulin are shown. (B) B95.8 cells transduced with lentiviruses expressing shEA-D #1 and #3 or shControl were treated with TPA and NaB as described above. The cells were harvested at 24 h and analyzed with Western blots probed with anti-PARP1, anti-EA-D, anti-ZTA, and anti-α-tubulin. The numbers below the PARP1 blot indicate the relative levels of PARP1.

Since PARP1 is reported to restrict EBV lytic replication by binding to the BZLF1 promoter (17, 18), we hypothesized that EA-D-induced PARP1 degradation may derepress the BZLF1 promoter and enhance ZTA expression, thereby facilitating viral lytic replication. In addition to ZTA, the RTA of EBV is also a critical immediate early gene product that regulates lytic replication. While PARP1 is shown to PARylate KSHV RTA (kRTA), thereby inhibiting kRTA transactivation (19, 20), interaction between PARP1 and EBV RTA has not previously been examined. HEK293T cells were cotransfected with FLAG-RTA and MYC-EA-D and subsequently subjected to co-IP assays using anti-FLAG antibody to investigate the interactions between RTA and PARP1 (Fig. 5A and B). The results showed that RTA directly interacted with PARP1 as well as EA-D, and that EA-D overexpression attenuated RTA-PARP1 interactions due to reduced PARP1 levels. To delineate whether it may be due to EA-D-mediated PARP1 degradation or its competition for PARP1 binding, MG132 was treated under the same conditions as in Fig. 5A. As MG132 treatment increased the PARP1 levels, the interaction between RTA and PARP1 was also restored, suggesting that EA-D-induced PARP1 degradation attributes to reduced interaction between RTA and PARP1 (Fig. 5B). To further examine whether RTA is targeted by PARP1 for PARylation, HEK293T cells were transfected with FLAG-RTA and co-IP assays using anti-FLAG were conducted (Fig. 5C). Western blot assays using the anti-PAR antibody showed that RTA was indeed PARylated and that EA-D overexpression alleviated the level of RTA PARylation. To explore the functional role of RTA PARylation, the RTA transactivation of the ZTA promoter (Zp) was analyzed using reporter assays when PARP1 was overexpressed (Fig. 5D). RTA transactivation of Zp was reduced by PARP1 overexpression in a dose-dependent manner, suggesting that RTA PARylation may negatively regulate RTA activity. Consistent with this result, EA-D alone enhanced the RTA transactivation of Zp (Fig. 5E). Moreover, EA-D overexpression was able to fully recover PARP1-induced suppression of Zp in a dose-dependent manner (Fig. 5F). Collectively, these results showed that EA-D enhances the ability of RTA to transactivate Zp by reducing the level of PARylated RTA via PARP1 degradation, thereby facilitating virus lytic replication.

FIG 5.

EA-D facilitated viral lytic replication by modulating the PARylation of RTA. (A to C) HEK293T cells were transfected with FLAG-RTA and/or MYC-EA-D in the absence (A and C) or presence (B) of MG132 and harvested at 48 hpt. The cell lysates were subjected to co-IP assays using anti-FLAG antibody and then analyzed with Western blots probed with anti-FLAG, anti-PARP1, anti-MYC, and anti-α-tubulin (A and B) or anti-PAR, anti-FLAG, anti-MYC, and anti-α-tubulin (C). (D to F) HEK293T cells were transfected with EBV RTA (20 ng) and a reporter construct of the ZTA promoter (Zp) (pGL3B-221Zp; 20 ng), together with FLAG-PARP1 (150 ng and 300 ng) (D), FLAG-EA-D (150 ng) (E), or FLAG-PARP1 (100 ng) and FLAG-EA-D (200 ng and 400 ng) (F). At 24 hpt, the cells were harvested with passive lysis buffer and analyzed using a luciferase assay system. Cell lysates were also analyzed for the expression of transfected genes in Western blots. “RTA^‡” images were obtained from a prolonged exposure. Three independent experiments were done, and each transfection was performed in triplicate. Data represents the mean ± SEM. Statistical analysis was performed using the Student's t test (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

EA-D promotes RTA-mediated transactivation on viral promoters, facilitating lytic replication.

Since EA-D affects RTA transactivation via PARP1 degradation, we further tested the effects of PARP1 and EA-D on other RTA-responsive promoters, such as BMRF1p and BMLF1p (35). Both reporter constructs are known to contain an RTA-responsive element (RRE) (Fig. 6A and B). In reporter assays, PARP1 overexpression suppressed RTA-induced transactivation of both BMRF1p and BMLF1p in a dose-dependent manner (Fig. 6C and D). EA-D overexpression not only restored PARP1 inhibition of RTA-mediated activation on BMRF1p and BMLF1p but also further enhanced the activities of both promoters, suggesting that EA-D derepresses PARP1 inhibition and promotes RTA-mediated transactivation of BMRF1p and BMLF1p (Fig. 6E and F). Taken together, our results suggest a highly conserved mechanism among gammaherpesviruses, in which viral processivity factors may act as important regulators of virus lytic replication via PARP1 degradation.

FIG 6.

EA-D derepressed PARP1 inhibition and enhanced RTA-mediated transactivation. (A and B) Schematic diagram of reporter constructs of the BMRF1 promoter (BMRF1p [A]) and the BMLF1 promoter (BMLF1p [B]). The reporter constructs contain the RTA-responsive element (RRE) at the indicated position. (C to F) HEK293T cells were transfected with EBV RTA (20 ng) and a reporter construct of the BMRF1 promoter (pGL2-BMRF1p; 20 ng) (C and E) or the BMLF1 promoter (pGL2-BMLF1p; 20 ng) (D and F), together with FLAG-PARP1 (150 ng and 300 ng) (C and D) or FLAG-PARP1 (100 ng) and FLAG-EA-D (200 ng and 400 ng) (E and F). At 24 hpt, the cells were harvested with passive lysis buffer and analyzed using a luciferase assay system. Three independent experiments were done, and each transfection was in triplicate. Data represents the mean ± SEM. Statistical analysis was performed using the Student's t test (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

DISCUSSION

PARP1 is a multifunctional nuclear factor that is associated with important cellular pathways, such as DNA damage repair, gene expression, inflammation, and tumorigenesis (13–16). PARP1 is also critical for the regulation of viral infections, especially for gammaherpesviruses. PARP1 was reported to restrict EBV lytic replication by catalyzing the PARylation of CTCF, an insulator protein, and by binding to the BZLF1 promoter, which contributes to the maintenance of EBV latency (17, 18). PARP1 also negatively regulates KSHV lytic replication via PARylation of kRTA. However, KSHV vPIP encoded by ORF49 sequesters PARP1 and PF-8 encoded by ORF59 induces PARP1 degradation, which contributes to derepression of PARP1 and enhances the RTA activity. Given that viruses often employ common strategies to surmount inhibitory cellular factors for replication, it is conceivable that gammaherpesviruses may also share a conserved mechanism to overcome PARP1 inhibition.

In this study, we found that EBV reactivation downregulated PARP1 in a proteasome-dependent manner via K29-linked polyubiquitination (Fig. 1 and 2). The viral processivity factor EA-D, encoded by the BMRF1 gene, was necessary and sufficient to induce PARP1 degradation in a proteasome-dependent manner (Fig. 3 and 4). As shown in transfection of the PARP1 AD deletion (ΔAD) mutant into B95.8 cells, direct interaction of EA-D with the central AD of PARP1 was critical for PARP1 degradation and efficient lytic replication of EBV (Fig. 3I). The EA-D knockdown restored PARP1 levels and ablated ZTA expression during reactivation, suggesting that EA-D could promote lytic replication via a mechanism that is distinct from the one as an accessory subunit of the virus DNA replication complex (25, 28–30). Since the ZTA promoter (Zp) can be activated by RTA, we aimed to determine if RTA-induced transactivation could be regulated by EA-D and PARP1. RTA was PARylated and PARP1 decreased the RTA-mediated transactivation of Zp in a dose-dependent manner (Fig. 5). The co-IP assay results showed that EA-D interacted with RTA and alleviated its PARylation level (Fig. 5C). Furthermore, EA-D expression disinhibited PARP1 and enhanced RTA-mediated Zp activation and other RTA-responsive promoters, BMRF1p and BMLF1p (Fig. 5 and 6). To the best of our knowledge, this is the first report identifying the PARylation of EBV RTA and its suppression by PARP1, as well as the alleviation of PARP1 inhibitory effects on RTA-responsive promoters by EA-D.

These results are consistent with our previous findings in KSHV and MHV-68, in which viral processivity factors were found to downregulate PARP1, which in turn enhanced RTA transactivation by reducing their own RTA PARylation (19). Collectively, our results suggest that there is a highly conserved mechanism among oncogenic gammaherpesviruses that functions to curtail the host’s repression of lytic replication by PARP1 with a viral processivity factor. However, the ubiquitin modification of PARP1 in the presence of EBV EA-D differed from that of KSHV PF-8 in that EA-D induces PARP1 degradation via a K29-linked ubiquitin modification, but not a K48-linked one. As KSHV RTA was shown to recruit CHFR, a cellular Ub-E3 ligase for PARP1 degradation, it will be of interest to identify the detailed molecular mechanism to show how EA-D may induce the K29-linked ubiquitination of PARP1 upon reactivation.

Taken together, our results suggest that EBV EA-D plays a pivotal role in facilitating lytic gene expression by inducing PARP1 degradation in addition to promoting viral DNA genome replication. Oncogenic gammaherpesviruses share a conserved molecular mechanism to counteract the host restriction factor PARP1, thereby facilitating viral lytic replication.

MATERIALS AND METHODS

Cells and virus reactivation.

HEK293T and HeLa cells were cultured in complete Dulbecco’s modified Eagle’s medium (DMEM) (PAN-Biotech) containing 10% fetal bovine serum (FBS) (Welgene) and supplemented with 1% penicillin and streptomycin (HyClone) at 37°C and 5% CO₂, while B95.8, SNU-719, and BJAB cells were cultured in complete RPMI (PAN-Biotech) containing 10% FBS (PAN-Biotech) and supplemented with 1% penicillin and streptomycin (HyClone) at 37°C and 5% CO₂. To induce the reactivation of EBV, the B95.8 and SNU-719 cells were treated with 20 ng/mL of 12-O-tetradecanoylphorbol-13-acetate (TPA) (Sigma) and 3 mM sodium butyrate (NaB) (Sigma).

Plasmids and transfection.

To construct the EA-D expression plasmids, EBV EA-D was amplified from EBV genomic DNAs using EA-D-specific primers (forward, 5′-CCGGGATCCATGGAAACCACTCAGACTCTCCG-3′; reverse, 5′-GGGGGAATTCTTAAATGAGGGGGTTAAAGGCCTGC-3′). The PCR product was cloned into the pENTR3C entry vector, and pTAG-EBV EA-D was constructed using the Gateway system (Invitrogen) according to the manufacturer’s instructions. The domain mutant constructs of PARP1 were obtained or constructed as previously described (24, 36). HA-tagged ubiquitin constructs were obtained from Addgene: pRK5-HA-Ubiquitin-K11 and pRK5-HA-Ubiquitin-K29 were provided by Sandra Weller (Addgene no. 22901 and 22903, respectively), pRK5-HA-Ubiquitin-K11R was provided by Josef Kittler (Addgene no. 121154), and pRK5-HA-Ubiquitin-K29R was provided by Ted Dawson (Addgene no. 17602). HEK293T cells in a 100-cm dish or in a 6-well plate were transfected using polyethyleneimine (PEI) (5 mg/mL) (Sigma), while B95.8 cells were transfected using the Neon transfection system (Invitrogen). The transfected cells were incubated at 37°C and 5% CO₂ and subjected to further assays.

Western blot and reagents.

For the Western blot analysis, HEK293T cells were harvested with a lysis buffer (0.5 M Tris-HCl [pH 6.8], glycerol, 10% SDS, β-mercaptoethanol, 0.5% bromophenol blue) and subjected to SDS-PAGE. Resolved proteins were transferred to a polyvinylidene fluoride (PVDF) membrane (Merck Millipore) and probed with a primary antibody against PARP1 (1:2,000) (BD Pharmingen), EA-D (1:500) (Merck), ZTA (1:500) (Santa Cruz Biotechnology), FLAG (1:2,000) (Sigma), MYC (1:1,000) (Homemade), HA (1:500) (homemade), PAR (1:500) (Trevigen), or α-tubulin (1:2,000) (Sigma). After washing, the membrane was incubated with a secondary antibody against anti-mouse IgG (1:5,000) (Cell Signaling) conjugated with horseradish peroxidase (HRP) and exposed to enhanced chemiluminescence (ECL) solution (ELPIS). The immunoblot signals were detected using an LAS-4000 mini (Fujifilm), and the relative protein levels were quantitated using ImageJ.

Coimmunoprecipitation assays.

HEK293T or B95.8 cells were harvested using an IP lysis buffer (20 mM HEPES [pH 7.4], 100 mM NaCl, 0.5% Nonidet P-40, and 1% Triton X-100) with a 1% concentration of protease inhibitor cocktail (Sigma) on ice and lysed in a 4°C rotator for 1 h. The supernatant was then harvested and incubated with anti-PARP1 or anti-FLAG antibodies for 1 h using a rotator at 4°C. Then, protein A/G agarose beads (Santa Cruz Biotechnology; sc-2003) were added and incubated for 15 h at 4°C. After being washed with the IP buffer, the beads were run on a 10% gel for the Western blot.

Immunofluorescence assay.

HeLa cells were seeded onto a cover glass (Paul-Marienfeld) in a 24-well plate. Cells were transfected with plasmids using the Lipofectamine 3000 transfection reagent (Invitrogen) according to the manufacturer’s instructions. Transfected cells were fixed with a fixation solution (4% formaldehyde and 0.15% picric acid in phosphate-buffered saline [PBS]) for 20 min at room temperature and incubated for 45 min at room temperature with a blocking solution (10% normal goat serum [Gibco] and 0.3% Triton X-100 in 0.1% bovine serum albumin [BSA]–PBS). After washing, the cells were probed with anti-FLAG (1:200) (Sigma) and anti-PARP1 (1:1,000) (Cell Signaling) antibodies for 15 h at 4°C. Anti-mouse fluorescein isothiocyanate (FITC) (1:200)-conjugated and anti-rabbit Cy3 (1:200)-conjugated antibodies were incubated for 1 h. Nuclear staining was carried out using a 4′,6-diamidino-2-phenylindole (DAPI) stain (1:1,000). The fluorescence images were obtained at a magnification of ×1,000 using a confocal laser scanning microscope (LSM 5 Exciter; Carl-Zeiss).

Construction of EA-D knockdown B95.8 cells.

To make the EA-D knockdown constructs, four EA-D-targeting oligonucleotides (shEA-D #1-forward [5′-CCGGAACCACTCAGACTCTCCGCTTCTCGAGAAGCGGAGAGTCTGAGTGGTTTTTTTG-3′] and shEA-D #1-reverse [5′-AATTCAAAAAAACCACTCAGACTCTCCGCTTCTCGAGAAGCGGAGAGTCTGAGTGGTT-3′], shEA-D #2-forward [5′-CCGGAAGGGAGGAGTGCTGCAGGTACTCGAGTACCTGCAGCACTCCTCCCTTTTTTTG-3′] and shEA-D #2-reverse [5′-AATTCAAAAAAAGGGAGGAGTGCTGCAGGTACTCGAGTACCTGCAGCACTCCTCCCTT-3′], shEA-D #3-forward [5′-CCGGCAAGTGCTATGACCATGCCCACTCGAGTGGGCATGGTCATAGCACTTGTTTTTG-3′] and shEA-D #3-reverse [5′-AATTCAAAAACAAGTGCTATGACCATGCCCACTCGAGTGGGCATGGTCATAGCACTTG-3′], and shEA-D #4-forward [5′-CCGGTAATCAGCTTCGAGGTCTCCCCTCGAGGGGAGACCTCGAAGCTGATTATTTTTG-3′] and shEA-D #4-reverse [5′-AATTCAAAAATAATCAGCTTCGAGGTCTCCCCTCGAGGGGAGACCTCGAAGCTGATTA-3′]) were annealed and cloned with pLKO.1 TRC cloning vector (Addgene no. 10878) using the AgeI and EcoRI sites according to the TRC cloning protocol. The cloned shEA-D constructs were verified using sequencing. To produce the shEA-D lentiviruses, HEK293T cells were cotransfected with pLKO.1-shEA-D #1~#4, psPAX2, and pMD2.G plasmids as previously described (37). Lentiviruses were harvested and replenished with fresh medium containing 1% penicillin–streptomycin for 3 days and then transferred to B95.8 cells. The transduced cells were treated with puromycin (1 μg/mL) (Sigma) and then selected for 3 days.

Luciferase reporter assays.

HEK293T cells were seeded 24 h prior to transfection in a 24-well plate at a density of 2 × 10⁵ cells/well and transfected with pGL3B-221Zp (Zp), FLAG-EBV RTA, and pTAG-EBV EA-D plasmids. The pGL3B-221Zp construct containing the BZLF1 promoter was provided by Jae Myun Lee (Yonsei University, Republic of Korea). Additional reporter plasmids containing the promoters for BMRF1 and BMLF1 (pGL2-BMRF1 and pGL2-BMLF1) were kindly provided by Shih-Tung Liu (Chang-Gung University, Taiwan) and Li-Kwang Chang (National Taiwan University, Taiwan) (35). The cells were harvested with 1× passive lysis buffer and analyzed using a luciferase assay system following the manufacturer’s instructions (Promega).

Statistical analysis.

Data represent the mean ± standard error of the mean (SEM). Statistical analysis was performed using the Student's t test. A P value of ≤0.05 was considered statistically significant.