Epstein-Barr virus reactivation is regulated by various cellular signaling pathways and can be activated under various stimuli, such as hypoxia (Jiang et al., 2006), 12-O-tetradecanoylphorbol-13-acetate (TPA) (Baumann et al., 1998), B-cell receptor (BCR) activation (Murata and Tsurumi, 2014), and reactive oxygen species (ROS) inducers (Yiu et al., 2020). Although how these stimuli initially induce EBV reactivation remains incompletely understood, the cascades following reactivation are triggered uniformly by viral immediate-early proteins Zta and Rta. They are the first two viral proteins actively expressed upon EBV reactivation, while both of them are silenced in EBV latently infected cells (Israel and Kenney, 2003; Kenney and Mertz, 2014). The expression of Zta and Rta marks the onset of the EBV lytic cycle, initiating a multi-stage biological process in which EBV-encoded lytic proteins and various host factors play complex and pivotal roles. Zta and Rta can also activate their promoters (Israel and Kenney, 2003), further boosting the expression of subsequent EBV early and late genes through positive feedback. They serve as transcriptional activators for lytic gene expression by binding to host transcription factors and specific response elements (Urier et al., 1989; Gruffat et al., 1990; Heilmann et al., 2012). Additionally, they are involved in viral genome replication after the onset of reactivation (Rennekamp et al., 2010; El-Guindy et al., 2013). Therefore, the expression of Zta and Rta is essential for the initiation of EBV reactivation, triggering a series of intricate cascading events, including EBV lytic gene expression, viral genome replication, virus particle assembly, and the release of progeny virions.

Given the crucial roles of Zta and Rta in EBV reactivation, transcription factors orchestrate the regulation of BZLF1 or BRLF1 transcription by specifically targeting their promoters (Figure 1A and Table 1). Various transcriptional activators and repressors are demonstrated to bind to BZLF1 and BRLF1, whose expression then determines the state of EBV infection (Spitz and Furlong, 2012; Lambert et al., 2018; Pope and Medzhitov, 2018). ZEBs silence the promoter of BZLF1 (Zp) in a cell-dependent manner, thus promoting the maintenance of viral latency. In particular, ZEB1 directly binds to the Zp, and its overexpression reduces the activity of Zp, suggesting its role in suppressing EBV lytic replication (Feng et al., 2007; Yu et al., 2007). Similar to ZEB1, ZEB2/SIP1 can bind to the same sites on Zp and inhibit BZLF1 expression to maintain EBV latency (Ellis et al., 2010). PARP1 has also been shown to bind Zp and suppress EBV reactivation, further contributing to cell proliferation and LMP1-mediated oncogenesis (Martin et al., 2016; Lupey-Green et al., 2017; Hulse et al., 2018). Additionally, IRF4 negatively regulates the activity of Zp, while EBF1 negatively regulates the activity of both Zp and Rp (Bristol et al., 2022). In contrast, the MEF2 transcription factor family binds and activates Zp (Murata et al., 2013). XBP-1 can activate Rp independently, whereas the activation of Zp by XBP-1 requires cooperation with PKD (Bhende et al., 2007). KLF4 and BLIMP1 are two differentiation-dependent transcription factors that synergistically induce EBV lytic replication. Among them, KLF4 can bind to and activate both Zp and Rp (Nawandar et al., 2015). BLIMP1 may bind to Rp and indirectly induce transcription of Zp and Rp (Reusch et al., 2015). KLF4 and BLIMP1 also synergistically activate LMP1, which promotes lytic replication in EBV-infected epithelial cells by enhancing the expression of Zta and Rta (Nawandar et al., 2017). HIF-1α directly binds to Zp, not Rp, to induce EBV reactivation, which requires phosphorylated p53 as a co-activator for the induction of the BZLF1 gene (Kraus et al., 2017; Kraus et al., 2020). The Hippo signaling effectors YAP and TAZ interact with Zp in epithelial cells to activate BZLF1 expression, and TEADs enhance the activation of YAP/TAZ on Zp (Van Sciver et al., 2021a). In epithelial cells, ΔNp63α maintains EBV latency, while in Burkitt lymphoma, TAp63α restrains the activation of EBV. ΔNp63α can block the activation of Zp which is mediated by KLF4 and Rta protein (Van Sciver et al., 2021b). The transcriptional activator C-Jun binds to and activates Zp (Flemington and Speck, 1990; Liang et al., 2002), whereas ARKL1 interacts with C-Jun, inhibiting the activity of C-Jun and negatively regulating EBV reactivation. The interaction between ARKL1 and C-Jun is mediated by the CK2 kinase-regulating subunit CK2β (Siddiqi et al., 2019). KAP1 is also an important factor in the maintenance of viral latency, whose sumoylation binds to the lytic promoters and suppresses their transcriptional activity (Bentz et al., 2015). Additionally, IFI16 and KAP1 function together to silence BZLF1 expression (Xu et al., 2022). In summary, these cellular factors can directly bind to Zp and Rp or recruit other factors for binding, thereby modulating their transcription activity and initiating subsequent processes in EBV reactivation.

The transcription of Zp and Rp is also regulated by epigenetic modifications. Zp and Rp are heavily methylated and silenced in latent infection, whereas histone acetylation of Zp and Rp enhances gene accessibility and promotes the binding of transcription factors during lytic replication (Speck et al., 1997; Miller et al., 2007; Woellmer and Hammerschmidt, 2013; Daskalogianni et al., 2015; Guo and Gewurz, 2022). In epithelial cells, TAF-Iβ recruits host methyltransferase MLL1 to promote H3K4 dimethylation and H4K8 acetylation, leading to the chromatin modifications of Zp region and the following EBV reactivation (Mansouri et al., 2013). CAF1, HIRA, and ASF1 are host factors that cooperatively load histone H3 and H4 onto the EBV genome, inhibiting the activity of IE promoter mediated by H3K27me3 and thus hindering lytic reactivation. Depletion of CAF1 protein induces EBV reactivation in Burkitt lymphoma cells (Zhang et al., 2020). JDP2 has been shown to bind to the Z II element of Zp and recruit HDAC3, reducing histone acetylation on the promoter and inhibiting EBV reactivation (Murata et al., 2011). The host factor C-Jun facilitates the binding of TET1 to Zp, enhancing the demethylation of Zp and activating BZLF1 expression (Zhang et al., 2016). When these inhibitory histone modifications are relieved, the transcription factors can bind to the promoters of lytic genes and activate EBV lytic replication (Jung et al., 2007; Murata et al., 2012; Guo and Gewurz, 2022). DNA methylation is not always detrimental to EBV lytic replication, for instance, Zta can specifically bind to the methyl-Zta responsive element (meZRE) and activate lytic gene promoters in a methylation-dependent manner, thereby disrupting epigenetic silencing (Bergbauer et al., 2010; Kalla et al., 2012; Almohammed et al., 2018). Furthermore, Zp and Rp can be targeted by various miRNAs. Following BCR stimulation in plasma cells, the increasing miR-429 disrupts EBV latency by targeting Zp (Lin et al., 2010). EBV miR-BHRF1-3 targets the 3′UTR region of BZLF1 gene and attenuates Zta expression (Fachko et al., 2022). Similarly, EBV miR-BART20-5p binds to the 3′ UTR region of these two IE genes, resulting in transcription inhibition (Jung et al., 2014). EBV miR-BART18 targets the 3′ UTR of MAP3K2 mRNA, affecting the expression of both BZLF1 and BRLF1 (Qiu and Thorley-Lawson, 2014).

Therefore, the activity of IE promoter, Zp and Rp, is modulated by transcriptional regulatory factors and epigenetic modifications, which collectively control EBV reactivation through regulating the expression of IE genes.

In addition to the transcription activity, the host proteins interact with Zta or Rta and modulate their capabilities in EBV reactivation (Figure 1B and Table 1). PLSCR1 can directly interact with Zta, antagonize the interaction between Zta and CBP, and further block the transcriptional activity of the Zta-CBP complex (Kusano and Ikeda, 2019). In Zta-mediated transcriptional activation, TORC2 acts as a co-activator and interacts with the Zta-CREB complex to activate Zp transcription, which involves the calcineurin-TORC pathway (Murata et al., 2009). PAX5 interacts directly with Zta and inhibits Zta-mediated lytic gene expression (Raver et al., 2013). PAX5 can also bind to viral TRs (tandem terminal repeats) and regulate the expression of viral genes (Arvey et al., 2012; Lee et al., 2015; Lee et al., 2016). IQGAP2 co-localizes with Rta in the nucleus and participates in the activation of Rp (Lin et al., 2023). Oct-1 and Oct-2 induce EBV reactivation through interactions with Rta and Zta, respectively. In particular, Oct-1 synergistically induces viral reactivation through direct interaction with Rta, which is independent of Oct-1 binding with DNA (Robinson et al., 2011). In contrast, Oct-2 negatively regulates virus reactivation by inhibiting the functions of Zta protein (Robinson et al., 2012). MCAF1 serves as an interaction partner for both Zta and Rta, facilitating their synergistic activation. The interaction between MCAF1 and Rta promotes the formation of Sp1-MCAF1-Rta complex at the Sp1 binding site, which ultimately results in the activation of Rp (Chang et al., 2005). In addition, the Zta-MCAF1-Rta complex also binds to the Zta response element (ZRE) and synergistically promotes the expression of EBV lytic genes (Chang et al., 2010). RanBPM simultaneously interacts with SUMO-E2 enzyme Ubc9 and Rta, which promotes SUMO-1-mediated Rta sumoylation and the transactivation activity of Rta (Chang et al., 2008).

Post-translational modifications (PTMs), such as phosphorylation, ubiquitination, sumoylation, and other covalent modifications are capable of regulating the associated activity of Zta or Rta (Walsh et al., 2005). The phosphorylation of Zta and Rta has been shown to affect the binding of Zta to DNA, and the transactivation capabilities of both Zta and Rta (Francis et al., 1999; Bhende et al., 2005; Asai et al., 2006). Furthermore, four lysine residues on Zta are targets of ubiquitination, and the ubiquitination of Zta is essential for progeny production in the viral lytic replication (Zhao et al., 2020). All SUMO isoforms 1, 2, and 3 can induce Zta sumoylation at the lysine residue 12 (Murata et al., 2010). Although sumoylation does not alter Zta localization, it inhibits the transactivation ability of Zta and the Zta-mediated lytic cycle (Hagemeier et al., 2010). In contrast, Rta transactivation activity is enhanced during SUMO-1-mediated sumoylation (Chang et al., 2004). Two E3 ubiquitin ligases, RNF4 and TRIM5α, interact with Rta and promote its ubiquitination, restricting viral lytic replication (Yang et al., 2013; Huang et al., 2017). RNF4 can target SUMO-2-conjugated Rta and knockouts of both RNF4 and TRIM5α increase Rta expression. PIAS1 enhances Rta-mediated activation of EBV lytic genes via sumoylation, but PIAS1 inhibits Zta and C/EBP β-mediated Zp activation, and this inhibition does not rely on its SUMO ligase activity (Zhang et al., 2017). Therefore, the host proteins interact with viral IE proteins and modulate their activities, potentially depending on the post-translational modifications, which play crucial roles in EBV reactivation.

The expression of EBV lytic genes determines the progression of a lytic cycle (Figure 2A and Table 2). During viral DNA replication, EBV-encoded proteins not only regulate the replication of viral genome but also hijack or antagonize host factors to control host DNA replication. BGLF4, an EBV-encoded protein kinase, not only activates the expression of BGLF3, but together they control post-replication processes and regulate the expression of late genes (El-Guindy et al., 2014). BGLF4 also phosphorylates UXT at the threonine-3 site, weakening the interaction between UTX and p65, which represses NF-κB-mediated BZLF1 and BRLF1 expression and promotes the viral lytic cycle (Chang et al., 2012). EBV-encoded nuclear protein SM is essential for lytic gene expression. SM binds to and stabilizes mRNA, modulating the splicing of cellular and viral pre-mRNAs, which depends on the interaction between SM and the SR splicing factor, specifically SRp20 (Verma et al., 2010). Additionally, as a transcription activator, SM binds to and recruits cellular XPB (xeroderma pigmentosum group B-complementing protein, a subunit of TFII H), subsequently targeting the promoters of late genes (Verma et al., 2020). SM also binds to CBC, eIF4F, and PABP, thereby enhancing the translation initiation of mRNPs, which are involved in protein synthesis (Mure et al., 2018). BRRF1 interacts with TRAF2 and induces lytic gene expression in a TRAF2-dependent manner. BRRF1 also co-activates lytic replication with p53. In the absence of p53, BRRF1-mediated activation of lytic genes is significantly reduced (Hagemeier et al., 2011).

The cofactor of viral DNA polymerase, BMRF1, is a key protein for lytic DNA replication. BMRF1 promotes the activation of BcLF1 and BLLF1 through interacting with the transcription regulator BRG1 (Su et al., 2017). The cellular pRb facilitates lytic reactivation, whose deletion inhibits lytic replication and BMRF1 expression in differentiated epithelial cells (Myers et al., 2023). The host factor TRAF6, once activated by polyubiquitin chain conjugation, binds to LMP1 and mediates NF-κB signaling, which inhibits the viral lytic cycle. However, EBV-encoded BPLF1 interacts with and deubiquitinates TRAF6, thereby inhibiting the LMP1-mediated NF-κB pathway and promoting viral DNA replication (Saito et al., 2013). Meanwhile, BPLF1 promotes the translation of EBNA1 mRNA (Liu et al., 2023). The host SMC5/6 complex recognizes and occupies R-loops, which are essential for DNA double-strand separation and the loading of core replication proteins during the lytic cycle, inhibiting the synthesis of EBV genome (Yiu et al., 2022). EBV-encoded BNRF1 membrane protein degrades the SMC5/6 complex through calpain proteolysis and the E3 ubiquitin ligase Cullin-7 dependent proteasome degradation pathway, allowing EBV to escape host restriction on viral genome replication and the production of infectious progeny virions (Yiu et al., 2022). Therefore, EBV-encoded proteins are key regulators in controlling the processes of viral lytic replication.

After the EBV genome completes replication, the virus synthesizes capsid proteins and assembles progeny virus particles within the host cell, ultimately releasing infectious virus particles to infect other cells (Figure 2A and Table 2). EBV-encoded BALF3 is involved not only in viral DNA synthesis but also in the packaging and release of mature virus particles (Chiu et al., 2014). BBLF1 is a myristoylated and palmitoylated protein, whose myristoylation enhances its stability and facilitates its membrane anchoring in the trans-Golgi network (TGN). Through interaction with PACS-1, a cytoplasmic protein, BBLF1 mediates retrograde transport from endosomes to the TGN and supports the production of progeny viruses (Chiu et al., 2012). The NED region of BBLF1 is key to the release of mature virus particles (Uddin et al., 2023). BDLF4, a viral protein expressed early in viral lysis, is a component of the lytic gene regulatory complex. BDLF4 knockout severely restricts the expression of late genes and reduces the production of progeny virus (Watanabe et al., 2015). By hijacking the host CDK2 complex, BDLF4 is phosphorylated at multiple amino acid residues, preventing its degradation through ubiquitin-mediated pathways. Among these residues, the phosphorylation of Thr-91 is critical for maintaining BDLF4 stability (Sato et al., 2019). BGLF2 enhances Zta expression in both B cells and gastric cancer cells, promoting the release of progeny viruses through the p38 signaling pathway. In contrast, its absence impairs reactivation triggered by inducers (Liu and Cohen, 2016). BMLF1 activates the UPR (unfolded protein response) sensor ATF6 by initiating the proteolytic cleavage, which transcriptionally activates the GRP78 promoter, a central regulator in the UPR, via ER stress response elements. Although the deletion of GRP78 does not affect the expression of EBV cleavage genes, it reduces the production of viral particles (Chen L. -W. et al., 2021). BKRF4, a late lytic protein, has a limited impact on viral gene replication and protein expression, yet it is crucial for the production of infectious progeny viruses (Masud et al., 2017). BPLF1 modulates the DNA damage response induced by EBV lytic replication, collaborating with RAD18 to enhance the production of infectious viruses (Kumar et al., 2014). BPLF1 also deubiquitinates TOP2, inhibiting the DNA damage response induced by TOP2 and promoting viral production (Li et al., 2021). These EBV-encoded proteins are responsible for regulating virus assembly and progeny production by modulating various cellular factors.

To successfully replicate and propagate within the infected cells, EBV has evolved various strategies to evade the host’s innate immune surveillance (Figure 2B and Table 2). EBV-encoded BGLF2 is an essential factor to counteract the interferon (IFN)-mediated antiviral response. BGLF2 can recruit the SHP1 phosphatase to STAT1 and inhibit the phosphorylation of STAT1. Additionally, BGLF2 recruits STAT2 to the E3 ubiquitin ligase Cullin 1, leading to its ubiquitination and degradation (Jangra et al., 2021). EBV G-protein-coupled receptor BILF1 recruits the E3 ligase UFL1 (UFM1 specific ligase 1) to facilitate the interaction between MAVS and UFM1 (ubiquitin fold modifier 1), leading to the MAVS UFMylation. This process results in MAVS degradation via the mitochondrial-lysosomal pathway, thus blocking the activation of NLRP3 inflammasomes (Yiu et al., 2023). EBV-encoded SM protein inhibits the transcriptional activation of DHX9, counteracting its antiviral function and reducing the effects of host immunity during reactivation (Fu et al., 2019). Cellular RNA Helicase DHX9 enhances the antiviral innate immunity in the host, limiting the production of EBV viral particles. TRIM24 and TRIM33 are cellular antiviral factors that inhibit BZLF1 expression, however, Zta interacts with these two proteins in EBV reactivation, leading to the destruction of the TRIM24-TRIM28-TRIM33 complex and the degradation of TRIM24 and TRIM33 (De La Cruz-Herrera et al., 2023). Autophagy impairs virus replication and transmission, but BPLF1 targets and inhibits the recruitment of selective autophagy receptor SQSTM1/p62 to LC3, reducing the adverse effects of autophagy on the lytic cycle (Ylä-Anttila et al., 2021).

In conclusion, the replication of the EBV genome, the production of progeny viruses, and the host’s immune response induced by various factors constitute an interconnected regulatory network. The balance among these components ultimately determines the progression following reactivation.

In addition to the EBV-encoded proteins described above, the process of EBV reactivation is also regulated by distinct miRNAs (Figure 2C and Table 2). EBV miR-BART20-5p blocks BAD-induced Caspase-3-dependent cellular apoptosis, thus contributing to the maintenance of EBV latency (Kim et al., 2016). Four EBV miR-BARTs (BART5-5p, BART7-3p, BART9-3p, and BART14-3p) cooperatively regulate and suppress ATM, resulting in the disruption of EBV reactivation (Lung et al., 2018). Another EBV miR-BART5-3p maintains latency by suppressing p53 expression (Zheng et al., 2018). EBV miR-BART9-3p, along with its homologous miR-141-3p, targets and inhibits the expression of the host transcription factor FOXO3 during BCR cross-linking, thereby reducing its inhibitory effect on BCR-induced viral lytic activation (Chen Y. et al., 2021). In contrast, EBV miR-BHRF1-2-5p, miR-BART2-5p, and the host miR-17-5p have been shown to weaken the BCR-mediated reactivation of EBV in vivo (Chen et al., 2019).