Table of Contents

    Topic review

    EBV Lytic Induction therapy

    View times: 151
    Submitted by: Alan Chiang

    Definition

    Epstein-Barr virus (EBV) lytic induction therapy is an emerging virus-targeted therapeutic approach that exploits the presence of EBV in tumor cells to confer specific killing effects against EBV-associated malignancies. Efforts have been made in the past years to uncover the mechanisms of EBV latent-lytic switch and discover different classes of chemical compounds that can reactivate the EBV lytic cycle. Despite the growing list of compounds showing potential to be used in the lytic induction therapy, only a few are being tested in clinical trials with varying degrees of success. This review will summarize the current knowledge on EBV lytic reactivation, the major hurdles of translating the lytic induction therapy into clinical settings and highlight some potential strategies in the future development of this therapy for EBV-related lymphoid and epithelial malignancies.

    1. Introduction

    Epstein–Barr virus (EBV) infects more than 90% of adults worldwide. While its primary infection is often asymptomatic, it can manifest as infectious mononucleosis (IM) in adolescents and young adults [1]. EBV is also associated with lymphomas such as endemic Burkitt lymphoma (BL), Hodgkin lymphoma (HL), T-/NK-, and B-cell non-Hodgkin lymphoma as well as epithelial carcinomas, which include undifferentiated nasopharyngeal carcinoma (NPC) and a subset of gastric carcinoma (EBVaGC) [2][3][4]. The biphasic lifecycle of EBV allows it to establish latency subsequent to primary infection in which viral gene expression is limited to those that are responsible for tumorigenesis, apoptosis inhibition, immune evasion, and so on [5]. Owing to the limited choice and the low expression of these viral proteins, it is difficult to target EBV-positive tumor cells specifically. In most cases, treatment against EBV-positive lymphomas is similar to those of EBV-negative lymphomas of the same histology, for example, chemotherapy, radiation, and tumor resection [6]. Therapeutic strategies that target EBV in the associated malignancies can result in highly specific killing effects to the tumor cells, but spare the normal cells from toxic effects.

    Occasionally, the latent virus within the infected cells enters into lytic cycle, in which >70 viral proteins are produced [5]. The switch occurs upon the expression of immediate early (IE) proteins, BZLF1 (Zta), and BRLF1 (Rta), which transactivate Zta and Rta promoters (Zp and Rp) and activate the expression of viral genes for viral replication, such as BMRF1, BALF1, and BGLF4, as well as that for production of virions, such as BLLF1 and BFRF3 [7]. The activation of IE proteins and promoters can be achieved through post-translational modification of activators or repressors, modulation of cellular signaling pathways, epigenetic regulation, such as DNA methylation; histone modification; cellular stresses, for example, oxidative stress, hypoxia, autophagy, and inflammation, as well as through modulation of host and viral micro RNAs [8][9][10][11]. Owing to the massive number of viral proteins expressed during the lytic cycle, they may be potentially utilized for EBV-specific therapies. One such therapy is the lytic induction therapy in which EBV is reactivated into the lytic cycle that confers cytotoxicity of antiviral drugs to achieve specific killing effects against EBV-positive cells. Although there were many studies in the past decades studying the lytic induction therapy, only a few were conducted in the setting of clinical trials.

    2. Overview of the Lytic Induction Therapy

    Lytic induction therapy is an emerging virus-targeted therapeutic approach that exploits the presence of EBV in tumor cells to confer specific killing effects against EBV-associated malignancies. This strategy involves two classes of compounds, that is, chemical lytic inducers and nucleoside analogue antiviral pro-drugs. The EBV lytic cycle is first being reactivated by the chemical lytic inducers producing an array of lytic proteins, one of which is the viral protein kinase encoded by BGLF4 [7]. This kinase phosphorylates and converts nucleoside analog anti-viral pro-drugs, such as ganciclovir, to their cytotoxic forms, consequently killing their host cells. More importantly, the phosphorylated drugs can be transferred to adjacent cells, which leads to a “bystander killing” effect [12] (refer to Figure 1) [13]. As a result, the success of this method relies heavily on the effectiveness of lytic inducers in reactivating EBV lytic cycle, emphasizing the importance of investigating a broad variety of compounds in order to enable and consolidate this form of therapy for EBV-associated malignancies.

    Figure 1. Overview of Epstein–Barr virus (EBV) lytic induction therapy. EBV lytic cycle is first reactivated by chemical inducers in which the viral protein kinase encoded by BGLF4 is produced. BGLF4 then activates the nucleoside analogue antiviral pro-drug into its cytotoxic form, and consequently results in a specific killing effect on EBV-positive cells. Moreover, the activated drugs can be transferred to adjacent cells, resulting in a “bystander killing” effect. GCV, valganciclovir.

    3. Lytic Inducers

    The lytic induction potential and the modes of lytic reactivation of different compounds, such as histone deacetylase (HDAC) inhibitors, chemotherapeutics agents, phorbol esters, butyrates, and novel compounds, in various cell lines harboring EBV have been summarized in detail in a recent review [14]. Despite having a continuously growing list of lytic inducers that can potentially be incorporated into the lytic induction therapy, very few drugs have been tested in clinical settings [15]. The only clinical trial study to date that has shown a promising outcome tested the effect of combining lytic inducers, gemcitabine (GCb) and valproic acid (VPA), with valganciclovir (GCV) on patients with end-stage NPC [16]. As different classes of lytic inducers have been addressed in detail in other reviews [17][18], we will briefly summarize the mechanisms of EBV lytic reactivation and outline the lytic inducers that possess the corresponding reactivation mechanism.

    EBV lytic cycle can be reactivated by modulating different signaling pathways of the host, for example, by activating protein kinase C (PKC) directly or together with mitogen-activated protein kinase (MAPK) family consisting of extracellular-signal-regulated kinase (ERK), c-Jun N-terminal kinases (JNK), and p38 signaling pathways . HDAC inhibitors such as suberanilohydroxamic acid (SAHA), romidepsin, valproic acid (VPA), trichostatin A (TSA), and sodium butyrate (NaB) [19][20]; phorbol esters such as tetradecanoylphorbol acetate (TPA) [21]; and microtubule depolymerization compounds such as colchicine and vinblastine [22] have been shown to activate the PKC and/or JNK and p38 signaling to reactivate EBV lytic cycle. TPA activates nuclear factor-κB (NF-kB) and activator protein 1 (AP-1) that mediate the activation of JNK, which may interact with Zp through the binding of c-Jun to the ZI and ZII elements . Another study revealed that Zp activation via PKC-δ activation requires the ZID element, which allows binding of the transcription factor Sp1 [23]. Proteasome inhibitor such as bortezomib and endoplasmic reticulum (ER) stress inducers such as thapsigargin and tunicamycin, on the other hand, can induce EBV lytic cycle by activating ER stress/unfolded protein response (UPR), which induces JNK and/or C/EBP-β and activates Zp through C/EBP-binding sites in ZII and ZIIIB elements [24][25]. UPR-induced lytic reactivation was also observed in clofoctol treatment, which mediates the activation of the PERK-XBP1 axis [26].

    Activation of PI3K/Akt signaling pathway can also reactivate EBV lytic cycle. Compounds that possess this property include chemotherapeutic drugs such as gemcitabine, doxorubicin, cis-platinum, and 5-FU [27] and immunosuppressive drug such as methotrexate [28]. Phosphoinositide 3-kinases (PI3K) activation was shown to be required for Rta activation of Zp and BMRF1 promoters, albeit the exact mechanism has not been completely elucidated [29]. Immunomodulatory agents such as lenalidomide and thalidomide suppress Ikaros, which can regulate EBV latency as well as activate PI3K signaling [30].

    Cellular stress-related signaling pathways involving ATM and p53 can also be associated with reactivation of EBV lytic cycle. Reactive oxygen species (ROS) inducers such as H2O2, methylnitronitrosoguanidine (MNNG), and the chemotherapeutic drug gemcitabine activate p53, which subsequently binds to the Sp1-binding element in Zp and Rp and activates the lytic cycle [31][32]. Additionally, chloroquine can reactive EBV lytic cycle by chromatin remodeling through the activation of the ATM pathway and the downstream phosphorylation of KAP1/TRIM28 [33], allowing the access of cellular transcription factors to activate the viral promoters [34].

    Induction of hypoxia has been shown to reactivate EBV lytic cycle through the binding of hypoxia-inducible factor 1 (HIF-1) to the hypoxia response element motif on Zp and/or by the activation of the ERK1/2 signaling pathway [35][36]. Iron chelators such as deferoxamine, Dp44mT, and a novel compound known as C7 were found to stabilize HIF-1a, which subsequently leads to the reactivation of EBV lytic cycle [36]. Apart from stabilizing HIF-1a, C7 was also found to reactivate EBV lytic cycle through the activation of the ERK1/2-autophagy (ATG5) axis [36]. In addition to C7, autophagy induction through the PKCδ-p38 MAPK axis by combination of TPA and NaB has also been shown to promote EBV lytic cycle [37].

    In addition to the above lytic reactivation pathways, other mechanisms such as induction of psychological stress by glucocorticoids such as hydrocortisone and dexamethasone [38], as well as inhibition of NF-kB signaling by antiretroviral medication such as azidothymidine [39],anti-inflammatory drugs or natural compounds such as aspirin [40] and curcuminoids [41], have been found to reactivate EBV lytic cycle. The detailed mechanisms for the reactivation have not been completely delineated. Large-scale screenings of chemical compounds have also identified several novel organic compounds, named E11 and A10 [42], and tetrahydrocarboline derivatives, named C09, C50, C51, C60, and C67, which can induce EBV lytic induction through as yet undetermined mechanisms [43].

    4. Weaknesses and Concerns Related to the Lytic Inducing Compounds

    As mentioned in the previous sections, many efforts have been made in the past years to uncover the mechanisms of chemical compounds in reactivating EBV lytic cycle in both EBV-positive lymphomas and epithelial carcinomas. Despite having the potential of being incorporated into EBV lytic induction therapy regimens, as shown in in vitro testing and Phase I/II clinical trial [44], these compounds have major weaknesses in their action. For instance, they have relatively low efficiencies in the reactivation of EBV lytic cycle. Table 1 summarizes the efficiencies of EBV lytic induction by the different compounds from multiple studies. In general, HDAC inhibitors such as NaB could reactivate 2–60% of EBV-positive B cells into lytic cycle, while SAHA could reactivate 30–65% of EBV-associated epithelial cells (AGS-BX1, HA, and HK1-EBV) into lytic cycle [45][46][47]. VPA could induce around 10% of AGS-EBV cells, while the percentage was low in LCL and C666-1 cells [48]. Novel compounds identified by our group such as C7, E11, C8, E7, and A10 could induce 30–60% of AGS-BX1 cells into lytic cycle [42]. Follow-up studies on C7, the best-performing compound identified, showed its ability to induce 6–12% of HA, C666-1, and NPC43 cells into lytic cycle [42]. Another new class of compounds, curcuminoids, were shown to induce 20–50% of AGS-BX1, C666-1, and HONE1-EBV cells into lytic cycle [41]. Combination of lytic compounds such as VPA and cisplatin was able to induce 50% of AGS-EBV cells into lytic cycle [48], while 40–70% could be achieved in AGS-BX1, HONE1-EBV, and C666-1 cells treated with VPA together with gemcitabine [41]. The above studies showed that a considerable proportion of cells are refractory to lytic cycle induction by most compounds studied. This refractory population greatly hinders the implementation of these lytic inducers into the lytic induction therapy and the translation to clinical settings.

    Second, these lytic compounds rely heavily on the cellular background for inducing EBV lytic cycle. For example, HDAC inhibitor, VPA, could induce EBV lytic cycle in EBV-associated epithelial carcinomas such as C666-1 and AGS-EBV cells [48], but not in EBV-positive lymphomas such as HH514-16, Raji, and Akata cells [49][50]. NaB was shown to induce lytic cycle in EBV-positive lymphoma cell lines including P3HR-1, B95.8, Raji, Daudi [51], and AK2003 , as well as in EBV-associated epithelial carcinoma cell line, AGS-BX1 , but does so very weakly in NPC cells . SAHA could induce lytic cycle in AGS-BX1, HA , AK2003, and C666-1  cells, but not in NPC43  and LCLs [47]. Similar results were found in chemotherapy agents such as 5-FU and cis-platinum, which could induce lytic cycle in Akata and AGS-EBV cells, but not in LCLs. For other classes of compounds such as tetrahydrocarboline derivatives , curcuminoids, iron chelators, and the novel compounds, lytic induction studies were only examined in either EBV-positive lymphoma cells or a subset of EBV-associated epithelial carcinoma cell lines, thus limiting general conclusions on their abilities to reactivate EBV lytic cycle in both cell types (Table 1). None of the compounds studied to date could induce EBV lytic cycle in all EBV-positive cell lines and their action dependent on cellular background and EBV latency states greatly hinder the incorporation of the available inducers in clinically relevant lytic induction therapeutic regimens.

    Lastly, the concern of promoting viral dissemination through chemical induction of EBV lytic cycle has to be addressed with caution. Most of the chemical compounds studied reactivate a complete EBV lytic cycle with production of virions. For instance, supernatant from HONE1-EBV cells induced with SAHA could transduce 71% of Daudi cells in an EBV transduction assay . This raises the concern of promoting viral dissemination in the midst of the therapy. Indeed, a pilot study on the efficacy and safety of romidepsin in treating extranodal natural killer/T-cell lymphoma found a substantial increase in viremia in these patients [52]. The novel compound C7  and anti-bacterial antibiotic, clofoctol, were found to induce the expression of immediately early and early lytic proteins, but not late lytic proteins. Moreover, EBV virions were not produced after lytic induction by these two compounds. The reactivation of EBV lytic cyle without production of virions puts them as potentially suitable candidates for incorporation in lytic induction therapy with minimal risk of viral dissemination.

    All of the previously studied compounds have at least one of the three major weaknesses mentioned above. For instance, HDAC inhibitors appear to be efficient in reactivating 30–50% of the cell population into EBV lytic cycle in both EBV-positive lymphoma and epithelial carcinoma cells, but their induction of full viral lytic cycle raises concerns in promoting EBV dissemination during the therapy. On the other hand, C7 and clofoctol are able to induce EBV lytic cycle without production of virions, but a relatively low percentage (10–20%) of cells can be induced into lytic cycle. Therefore, efforts such as structural refinements, as demonstrated in studies by Tikhmyanova et al. [53] and our group , will be important in promoting the utility of these compounds in lytic induction therapy. Apart from refining the currently available lytic inducers, combination of these different classes of compounds, repurposing of other classes of clinically available compounds, or designing novel chemical molecules or peptides can be employed to facilitate the translation of lytic induction therapy for EBV-associated malignancies into the clinics. These strategies will be discussed in detail in section 5.

    Table 1. Summary of the efficiency of lytic induction of Epstein–Barr virus (EBV) of the lytic inducers and the cell types in which lytic cycle can be induced *.

    Class

    Compound

    Cell Type that Can Be Induced

    (% of Cell Population)

    Cell Type that Cannot Be Induced

    Ref.

    HDAC inhibitors

    NaB

    HH514-16, B95.8

    Raji

    [49][50]

    AHS-BX1, BL-AK2003

    LCLs

    [47]

    TSA

    HH514-16, P3J-HR1

    Raji, B95.8, Akata

    [49][50]

    AHS-BX1, BL-AK2003

    LCLs

    [47]

    VPA

    LCL (low), C666-1 (low), AGS-EBV (10%)

    /

    [48]

    AGS-BX1

    LCLs, BL-AK2003

    [47]

    TPA

    B95.8, Raji

    HH514-16

    [49][50]

    SAHA

    AGS-BX1, BL-AK2003

    LCLs, NPC43

    [47]

    HK1-EBV, HONE-1-EBV, HA (30–65%), C666-1

    /

    [46]

    Romidepsin

    HA (75%), C666-1 (6%)

    NPC43

    [46]

    DNA methyltransferase

    inhibitor

    AZC (5 ara2'-deoxycytidine)

    HH514-16

    /

    [49][50]

    RaeI (80%)

    /

    [54]

    Iron chelators

    Deferoxamine,  Dp44mT

    AGS-BX1, SNU719, HA

    /

    [36]

    Deferasirox, Deferiprone

    AGS-BX1, SNU719

    /

    [36]

    Novel compounds

    C7

    AGS-BX1, SNU719, HONE1-EBV, YCCEL-1, HA (10%), C666-1 (6%), NPC43 (12%)

    /

    [36]

    E11

    AGS-BX1 (60%), HONE1-EBV, YCCEL-1

    SNU719, C666-1

    [42]

    C8

    AGS-BX1 (30%), HONE1-EBV, C666-1

    SNU719, YCCEL-1

    [42]

    E7

    AGS-BX1 (30%), HONE1-EBV, C666-1, SNU719

    YCCEL-1

    [42]

    A10

    AGS-BX1 (30%), HONE1-EBV, C666-1, YCCEL-1

    SNU719

    [42]

    Chemotherapeutic agents

    5-FU

    Akata, AGS-EBV (24–28%)

    LCL

    [17]

    Gemcitabine

    AGS-EBV (30%), Akata, AGS-EBV (24–28%)

    LCL

    [17]

    Doxorubicin

    LCL [17]

    LCL [48]

    [17]

    Taxol

    LCL [17]

    LCL [48]

    [17]

    5 aza-CR

    Akata, AGS-EBV (24–28%)

    /

    [27]

    Immunomodulatory agents

    Lenalidomide, thalidomide, pomalidomide

    B95.8, D4 LCL, DAUDI, KEM-I, MUTU-I

    /

    [17]

    Anti-bacterial antibiotic

    Clofoctol

    Akata (40%), SNU719 (2%), C666-1 (0.5%), LCLs (0.5%)

    /

    [26]

    Curcuminoids

    41

    AGS-BX1 (40–60%), C666-1 (10–30%), HONE1-EBV (20-40%)

    SNU719

    [41]

    EF24

    AGS-BX1 (50–70%), C666-1 (10–30%), HONE1-EBV (40–60%)

    SNU719

    [41]

    Tetrahydrocarboline derivatives

    C09, C50, C53, C60, C67

    MutuI, LCL, Akata, C666-1

    /

    [43]

    ER stress inducers

    Thapsigargin

    LCL

    /

    [25]

    ROS inducer

    N-Methyl-N’-Nitro-N-Nitrosoguanidine (MNNG)

    HA, C666-1, NA (70%)

    /

    [32]

    * HDAC, histone deacetylase; NaB, sodium butyrate; TSA, trichostatin A; VPA, valproic acid; TPA, tetradecanoylphorbol acetate; SAHA, suberanilohydroxamic acid; 5 aza-CR, 5-azacytidine; ER, endoplasmic reticulum; ROS, reactive oxygen species.

    5. Potential drugs and strategies in the future development of lytic induction therapy

    5.1. Combining currently available lytic inducers for induction of EBV

    Different lytic inducers have been combined in previous studies for reactivating lytic cycle of EBV. The combination between TPA and NaB was found to enhance EA-D expression by 1.5-15 fold more than that by either compound alone in Raji cells [55]. Combination of VPA with cisplatin could induce 50% of AGS-EBV cells into lytic cycle with 1.5-5 fold increase relative to treatment with either compound alone. Additionally, when lenalidomide was combined with doxorubicin or melphalan, lytic induction was enhanced in Daudi and Mutu-I cells. These studies showed that combining different classes of lytic inducers with divergent modes of action in lytic reactivation of EBV could complement one another and achieve a higher efficiency in the induction lytic cycle of EBV. Iron chelators and SAHA could reactivate EBV lytic cycle by stabilizing HIF-1a [56] and activating the PKC-d pathway, respectively . Our group showed that iron chelators could reactivate the lytic cycle through autophagy-dependent pathways while SAHA’s action was independent of autophagy. These two compounds might synergize with one another in inducing lytic cycle of EBV (Figure 2a). Combining iron chelators with immunomodulatory agent such as lenalidomide will also be of interest. Lenalidomide could reactivate lytic cycle of EBV by suppressing Ikaros , which is a transcription factor that was found to upregulate the expression of cellular factors responsible for maintaining EBV latency [57][58]. Direct activation of Zta promoter by iron chelator through HIF-1a binding together with the suppression of inhibitory factors that prevent Zta transactivation of other lytic genes by lenalidomide may provide a feed-forward loop for lytic reactivation of EBV (Figure 2b). Some criteria may need to be considered in the design of combination therapy. First, matching the kinetics of lytic induction of different compounds will be important. Our group found that combination of C7 and SAHA could only enhance lytic reactivation when the treatment duration of C7 matched with its reactivation kinetics . Second, compounds of the same class may not utilise the same mode of action in inducing lytic cycle of EBV. VPA antagonized the reactivation of lytic cycle of EBV by other compounds of the same class such as NaB, TSA, AzaCdR, MS-275, apicidin and SAHA and uniquely enhanced expression of some cellular genes . Similar antagonism was also observed when romidepsin, another HDAC inhibitor thought to have similar action as that of SAHA, was combined with C7. Therefore, in-depth study should be performed to delineate the modes of action of different lytic inducers before deciding on the combination therapy.