Hide-and-seek by Epstein-Barr virus: evasion of innate immunity

Transcription

1 Hide-and-seek by Epstein-Barr virus: evasion of innate immunity Michiel van Gent

2 Hide-and-seek by Epstein-Barr virus: evasion of innate immunity PhD thesis, Utrecht University, the Netherlands ISBN: Cover: Michiel van Gent Lay-out: Michiel van Gent Printed by Uitgeverij BOXPress, proefschriftmaken.nl Michiel van Gent, Utrecht, the Netherlands. All rights reserved. No parts of this thesis may be reproduced, stored in a retrieval system or transmitted in any form or by any means without permission of the author. The copyright of articles that have been published has been transferred to the respective journals. Printing of this thesis was financially supported by the Netherlands Society of Medical Microbiology (NVMM) and the Royal Netherlands Society for Microbiology (KNVM), Infection & Immunity Utrecht, and University Medical Center Utrecht.

3 Hide-and-seek by Epstein-Barr virus: evasion of innate immunity Ontsnapping aan het afweersysteem door het Epstein-Barr virus (met een samenvatting in het Nederlands) Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof.dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op dinsdag 19 mei 215 des middags te uur door Michiel van Gent geboren op 29 november 1984 te Nijmegen

4 Promotor: Prof. dr. E.J.H.J. Wiertz Copromotor: Dr. M.E. Ressing

5

6 Commissie: Prof. dr. F. J. M. van Kuppeveld Prof. dr. J. M. Middeldorp Prof. dr. L. Meyaard Prof. dr. J.A.G. van Strijp Dr. D. van Baarle Paranimfen: Rutger D. Luteijn Marc A. J. van den Heuvel

7 Contents Chapter 1 General introduction Partly published in Current topics in Microbiology and Immunology, 215; in press 9 Chapter 2 EBV Lytic-Phase Protein BGLF5 Contributes to TLR9 Downregulation during Productive Infection Published in The Journal of Immunology, 211; 186, Chapter 3 Silencing the shutoff protein of Epstein-Barr virus in productively infected B cells points to (innate) targets for immune evasion Published in Journal of General Virology, 215; 96, Chapter 4 Epstein-Barr Virus Large Tegument Protein BPLF1 Contributes to Innate Immune Evasion through Interference with Toll-Like Receptor Signaling Published in PLoS Pathogens, 214; 1(2), e Chapter 5 Epstein-Barr virus mirna BART16 suppresses type I interferon signaling Manuscript in preparation 121 Chapter 6 Summarizing discussion 147 Addendum List of abbreviations Nederlandse samenvatting Dankwoord Curriculum vitae

8

9 CHAPTER 1 General introduction Maaike E. Ressing Michiel van Gent Anna M. Gram Marjolein Hooykaas Sytse Piersma Emmanuel J. Wiertz Partly published in Current topics in Microbiology and Immunology (215), in press

10 Chapter 1 Part I: Epstein-Barr virus Herpesviruses Herpesviruses are enveloped, double-stranded DNA viruses that are highly disseminated in nature. Many animal species are infected with one or more herpesviruses and over 2 herpesvirus species have been identified to date [1]. In contrast to most RNA viruses that cause brief infections and are rapidly cleared, herpesviruses persist for life in the infected host, usually without overt symptoms. Herpesvirus genomes are relatively stable over long periods of time, in sharp contrast to quickly evolving RNA viruses such as influenzavirus [1,2]. The narrow host range of herpesviruses, together with millions of years of coevolution with their host, has led to intimate interactions between herpesviruses and their hosts [3], and many herpesviruses encode a set of unique genes without homologs in other family members. This is for example reflected by the delicate balance between immune activation and evasive strategies. Initially, herpesvirus classification was based on virion architecture, which led to the inclusion of an extended set of divergent viruses into the herpesvirus order [1]. Following more recent availability of sequencing data, the more specific Herpesviridae family was established that encompasses herpesviruses of mammals, birds, and reptiles. The nine human-specific herpesviruses are Herpes Simplex virus (HSV)-1 and 2 (HHV1 and 2), Varicella Zoster virus (VZV, HHV3), Epstein-Barr virus (EBV, HHV4), Human Cytomegalovirus (HCMV, HHV5), HHV-6A and B, HHV-7, and Kaposi s Sarcoma-associated Herpesvirus (KSHV, HHV8) [1,4]. These are further subdivided in the alpha-, beta-, and gamma subfamilies (Table I). Each herpesvirus generally establishes latency in long-lived cells, for instance neurons or memory B cells, and induces lytic infection in permissive cells closer to the surface of the infected individual (e.g. epithelial cells) to facilitate spread to other individuals [1]. In general, and especially for the gammaherpesviruses EBV and KSHV, pathogenesis is not an essential part of the normal viral life cycle. Nevertheless, their ubiquitous presence renders herpesviruses a significant clinical burden. Epstein-Barr virus and clinical implications The human gammaherpesvirus EBV is the first human tumor virus identified and was first observed by electron microscopy in Burkitt s lymphoma cells in 1964 [5]. EBV was implicated as the causative agent of infectious mononucleosis (IM) in 1968, a few years later its transforming capacity was established in in vitro-infected B cells [1], and the entire genome was sequenced in 1984 [6]. EBV has a wide geographical distribution, infecting over 9% of the world adult population [1]. Two EBV subtypes (1 and 2) are distinguished that differ in the sequence of several latency-associated proteins (described below), with subtype 1 being most prevalent in the majority of human populations [7,8]. Although type 1 strains have increased transforming capacity, there is no difference in disease correlation between the two 1

11 General introduction Table 1. Human herpesviruses HHV* Name Subfamily Cell tropism Pathophysiology Latent infection Lytic infection HHV1 HSV-1 α Neurons Epithelial cells Orofacial infections, encephalitis HHV2 HSV-2 α Neurons Epithelial cells Genital and neonatal infections HHV3 VZV α Neurons Epithelial cells Chickenpox, shingles HHV4 EBV γ B cells Epithelial cells, B cells HHV5 HCMV β Macrophages, lymphocytes, epithelial cells HHV6A HHV6A β Monocytes, macrophages HHV6B HHV6B β Monocytes, macrophages Macrophages, lymphocytes, epithelial cells CD4 + T cells CD4 + T cells Infectious mononucleosis, lymphoma (BL, HL), carcinoma (NPC, gastric) Congenital infections, retinitis, hepatitis Exanthem subitum Exanthem subitum HHV7 HHV7 β T cells T cells Exanthem subitum HHV8 KSHV γ Lymphocytes Lymphocytes Kaposi s sarcoma *HHV, human herpesvirus; adapted from Paludan et al., 211 [121] 1 subtypes [1]. Historically, EBV was acquired in the early years of life, a situation that is still seen in developing countries [7]. Primary infection during childhood is generally believed to proceed asymptomatically, although some studies suggest that symptoms may also arise in younger patients [9]. Nowadays, primary infection in developed countries is postponed until adolescence and leads to IM in over 7% of cases [1,11]. IM is a usually self-limiting disease characterized by large numbers of atypical CD8 + T cells in the blood that only leads to mortality in rare cases [12]. Besides IM, EBV s transforming capacity is associated with malignancies that originate from persistently infected cells [11]. For example, EBV infection of epithelial cells is involved in a subset of nasopharyngeal carcinomas (NPCs) and approximately 1% of gastric carcinomas [1,13,14]. NPCs are relatively rare in the Western world and are predominantly observed in certain populations in South-East Asia, North Africa and indigenous people in the Arctic region [15]. Most prominent malignancies originating from B cells, the other primary target cell of EBV, are Burkitt s lymphoma and Hodgkin s lymphoma [16 18]. Less frequently observed are T- and NK-cell lymphomas originating from EBV-infected cells [1,19,2]. EBVassociated malignancies arise in particular in immunocompromised individuals [1]. These include patients undergoing transplantations and receiving immunosuppressive drugs (post-transplant lymphoproliferative disease or PTLD) as well as HIV-patients, although this latter group is declining since the introduction of effective treatment methods that limit HIV infection. These observations illustrate the importance of the immune system in controlling EBV infection. Besides inducing tumor growth, EBV infection is associated with several other clinical diseases. X-linked Lymphoproliferative disease (XLP-1) patients are particularly vulnerable to EBV infection-related morbidity and mortality. XLP-1 is caused by a mutation in SAP (SLAMassociated protein), an adaptor protein that is expressed on T, NK, NKT, and B cells and interacts with the SLAM receptor on target cells to inhibit IFN-γ production and increase NK 11

12 Chapter 1 cell-mediated killing [21 24]. Defects in SAP lead to impaired NK and CD4 + T cell responses, resulting in reduced numbers of EBV-specific memory B cells and plasma cells, and impaired long-term antibody responses in those patients that survive primary infection [1]. Chronic Active EBV (CAEBV) is another rare disorder mainly occurring in Asia and Central and South America. CAEBV presents after primary EBV infection with recurrent IM-like symptoms such as fever and lymphadenopathy, and lymphoproliferative disease affecting many organs [1]. Although the exact cause is unknown, a deficient EBV-specific immune response is believed to play a role in CAEBV. Finally, in contrast to these EBV-associated disorders originating from latently infected cells, oral hairy leukoplakia is a non-malignant disease observed in late-stage AIDS patients caused by lytic EBV replication in the outer layer of epithelial cells of the tongue in the absence of proper immune surveillance [1]. Viral life cycle EBV infection starts by transmission through saliva (although some observations also suggest transmission by sexual intercourse), after which the virus enters epithelial cells in the nasopharyngeal area to replicate and spread to naïve B cells in the underlying lymphoid tissues [1,18]. In these B cells, EBV mimics the endogenous cellular differentiation pathways leading to memory B cells by going through several latency stages until true latency is established in memory B cells. This sequence commences with latency III, also termed growth transcription program, characterized by expression of latent genes EBNA-1, EBNA-2, EBNA- 3a, EBNA-3b, EBNA-3c, EBNA-LP, LMP1, and LMP2, together with the non-coding RNAs EBER1 and EBER2 and EBV-encoded mirnas (Figure 1) [25]. LMP1 and LMP2 mimic the essential differentiation signals normally provided by costimulatory molecule CD4 and the B-cell receptor, respectively [1]. Lymphoblastoid cell lines (LCLs), established upon in vitro infection of primary B cells in the absence of a functioning immune system, display a latency III expression pattern. This is representative of post-transplant lymphoproliferative disorders that arise in immunosuppressed individuals. During natural infection of healthy individuals, latency III is replaced by latency II (or the default transcription program ) characterized by expression of EBNA-1, LMP1, LMP2, the EBERs, and the EBV mirnas. These gene products provide a survival advantage to the infected B cells as they enter germinal centers, which enables these cells to compete with endogenously activated, maturing B cells. The latency II expression pattern is observed in cells and cell lines derived from EBV-associated Hodgkin s lymphomas [16] and nasopharyngeal carcinomas [15]. Upon exiting the germinal center, the B cell has differentiated into a long-lived memory cell and all viral protein expression has ceased (latency ). Only during cell division, EBNA1 is expressed to ensure replication and maintenance of the viral genome (latency I), a stage that is represented in Burkitt s lymphoma cells [17,18]. In healthy, asymptomatic carriers 1-5 per million B cells in the peripheral blood carries the EBV genome [26]. 12

13 General introduction To be transmitted to new cells and hosts, new viral particles are produced during the lytic or productive phase of infection. Both B cells and epithelial cells are permissive to productive infection, which involves sequential expression of the full repertoire (>8) of viral lytic genes (Figure 1). The transition from latent to lytic infection is induced by the immediate-early (IE) transactivators BZLF1 (Zta, ZEBRA, Z) or BRLF1 (Rta, R) that trigger expression of over 3 early (E) genes involved in viral DNA replication. During the late (L) stage of productive infection, >3 structural proteins are expressed and assembled into new viral particles. Productive EBV infection is associated with a general inhibition of (host) protein synthesis, termed host shutoff, instigated by mrna degradation by the viral alkaline exonuclease BGLF5 [27]. 1 EBV genome The EBV genome is ~18 kilobase pairs in length and encodes approximately 9 identified or predicted ORFs (Figure 1). The exact transcriptome may however be more elaborate, as studies suggest that the actual transcriptome of herpesviruses can be significantly more complex than predicted from ORF annotations alone [28]. Forty-one genes with essential roles during the herpesvirus life cycle are conserved throughout the entire herpesvirus family ( core genes). Gammaherpesviruses additionally share homologous genes necessary to maintain the viral genome during cell division [1]. The remaining EBV gene products are likely tailored to the host and cell types infected, being (more) specific to EBV. While many functions of the EBV gene products have been elucidated, several proteins remain without any identified function, and novel purposes for proteins with known functions keep being identified [1]. Identification of EBV lytic gene functions is compromised by the fact that this virus, in contrast to related KSHV, establishes latency upon in vitro infection of B and epithelial cells. Productive infection can be induced chemically or by cellular stimulation such as B-cell receptor crosslinking using antibodies. Unfortunately, in all cases induction efficiencies hardly exceed a few percent, severely limiting biochemical characterization of productively-infected cells. Therefore, a system was set up consisting of EBV + Akata BL B cells that more efficiently enter productive infection upon B-cell receptor crosslinking (on average 4-8%) and express a ratcd2-gfp fusion protein on the cell surface that allows identification and sorting of productively infected B cells to high purity [29]. This system greatly facilitated studies towards the function of lytic gene products, and has been used to successfully identify various adaptive immune evasion strategies adopted by EBV [3]. The discovery of viral mirnas in EBV-infected B cells approximately ten years ago added another layer of complexity to the regulation of viral and host protein expression in infected cells [31]. Over 9% of the more than 24 viral mirnas identified to date are encoded by herpesvirus family members, and EBV expresses 44 mature mirnas during all latency stages and during productive infection [31 35]. Three EBV mirna precursors are encoded 13

14 Chapter 1 Daudi deletion P3HR1 deletion BHRF1 mirnas LMP2a-b BNRF1 EBER1-2 BCRF1 LP LP LP LP LP LP BCRF2 BWRF1 BWRF1 BWRF1 BWRF1 BWRF1 BWRF1 BWRF1 BYRF1 BHLF1 BFLF BPLF1 BOLF1 BFRF BRLF1 BRRF1 BRRF2 BKRF1 BKRF2 BKRF3 BKRF LMP2a LMP1 LMP2b BVRF1.5 BVRF2 BdRF1 BILF2 LF3 LF2 LF1 BILF1 BALF5 ECRF4 BALF4 BARF BALF3 BALF2 BALF1 BARF1 BNLF2a,b BZLF1 BBLF4 BBRF1 BBRF2 BBLF3 BBLF2 BBRF3 BBLF1 BGLF5 BGLF4 BGLF3.5 BGLF3 BGRF1 BGLF2 BGLF1 BDLF4 BDRF1 BDLF3 BDLF2 BDLF1 BcLF1 BcRF1 BTRF1 BXLF2 BXLF1 BXRF1 BVRF1 BFLF1 BFRF1a BFRF1 BFRF2 BORF1 BORF2 BaRF1 BMRF1 BMRF2 BMLF1 BSLF2 BSLF1 BSRF1 BLLF3 BLRF1 BLRF2 BLLF2 BLLF1 BLRF3 BERF1 BERF2a BERF2b BERF3 BERF3 BZLF2 BHRF1 EBNA2 EBNA3a EBNA3b EBNA3c Raji deletion EBNA1 BART cluster 1 mirnas BART cluster 2 mirnas BART mirnas B95-8 deletion Raji deletion 14

15 General introduction Figure 1. Schematic representation of the linear Epstein-Barr virus genome with open reading frames (ORFs). EBV ORFs are named based on the BamHI fragment they are located in (alphabetically ordered, based on size and starting with the largest). LF (leftward fragment) and RF (rightward fragment) indicate the direction of the ORF. Latency-associated ORFs are depicted in red, the immediate-early transactivators in yellow, and the other productive-phase associated ORFs in blue. Numbers refer to kb pairs. The reference type 1 EBV genome is a composite of the B95-8 and Raji genomes, as the former contains a deletion around the BART region (indicated by a red line) encompassing Raji ORFs LF1, LF2, and LF3 including many of the BART mirnas. The Raji strain lacks the EBNA3c and BZLF2 coding sequence, as well as a 2.9 kb sequence in the BamHI A fragment. The Daudi and P3HR1 EBV strains contain the indicated deletions around the EBNA2 region. Purple text indicates gene products with a known role in innate or adaptive immune evasion. BPLF1, BGLF5, and the BART cluster1 mirnas (indicated in green) are discussed in this thesis. Adapted from Longnecker et al., 213 [1]. 1 in the BHRF1 cluster and another 4 are contained within introns of the non-coding BART transcript (BART cluster 1 and 2, Figure 1) [36]. Because of their relatively recent identification the role of many of these EBV mirnas remains elusive, although some have been shown to regulate viral protein expression and cellular processes such as apoptosis and immune responses [36]. Potential mirna targets have been obtained through bio-informatics and immunoprecipitation approaches, yet experimental validation is often lacking and is complicated by the short (and therefore abundant) mirna target sites and the potential targeting of hundreds of genes by single mirnas [37]. However, the small size and lack of immunogenicity marks mirnas as appealing mediators of viral immune evasion that may especially be active during latent infection, when virtually all protein expression is abolished. Taken together, EBV is a clinically relevant agent due to its causal links with IM, various malignancies of mainly epithelial and lymphoid origin, and several other clinical disorders. Nonetheless, the exact contribution of EBV to pathogenesis remains obscure in many cases. Our insights into the elaborate and fine-tuned interactions of EBV with its human host gradually increases, yet remains far from complete. An important component of these interactions is the interplay between the virus and the human immune system, which forms an important determinant in disease outcome as illustrated by several observations for herpesviruses in general and EBV specifically (discussed in part II of this chapter). A significant part of the EBV proteins and mirnas may be involved in regulating virus-host interactions. Especially for the innate immune system, many of these functions have not been identified yet. This thesis focuses on the identification of novel innate immune evasion strategies employed by EBV. Increasing our understanding of the interactions of EBV with the immune responses keeping viral replication and cellular outgrowth in check, will provide a better insight into the pathogenesis of EBV and its associated malignancies and will aid in improving current and novel therapeutic strategies [38]. The following parts of this chapter sequentially present our current state of knowledge on several components of the innate immune system involved in sensing EBV, and the strategies EBV has adopted to prevent elimination by the host immune system. 15

16 Chapter 1 Part II: Epstein-Barr virus and the innate immune system The primary target cells of EBV, epithelial cells and in particular B cells, are equipped with various receptors and components of the innate and adaptive immune systems. Activation of the innate immune system induces direct antiviral responses [39] and for instance type I IFN responses and natural killer (NK) cells have proven crucial in the containment of many herpesvirus infections [4 43]. In addition, innate immune responses play an important role in activating and orchestrating adaptive antiviral responses [44]. In this way, potent CD8 + and CD4 + T cells responses are initiated upon primary EBV infection that can quickly detect and kill reactivated cells [45]. This potentially limits the lifespan and virus yield from productively infected cells, which are viable for several days in vitro [29]. Priming of optimized T cell responses by the innate immune system has proven crucial for the control of latent and lytic EBV infection [45 47]. The importance of the host immune system in controlling herpesvirus infections is further illustrated by the emergence of EBV-associated malignancies in immunosuppressed individuals [48 5]. This second part of this chapter describes the intracellular components of the innate immune system dealing with virus infection (part 2.1) and the components important for sensing EBV infection (part 2.2), schematically presented in Figure Innate immune responses to viral infection The innate immune system forms a first line of defense against invading pathogens. Firstly, mechanical barriers, for example the skin and mucosal surfaces, interfere with invasive infection. Cell-intrinsic antiviral immune mechanisms directly restrict viral replication [51,52]. In addition, the innate immune system comprises a broad range of cells with innate effector functions, including epithelial cells, dendritic cells, phagocytic macrophages, NK cells, and neutrophils. These and other (immune) cells are equipped with a diverse repertoire of pattern-recognition receptors (PRRs) that function as sentinels for conserved pathogenassociated molecular patterns (PAMPs) or danger-associated molecular patterns (DAMPs) in their surroundings [51,53 55]. Activation of these receptors by their respective ligands induces direct antiviral effector mechanisms and orchestrates ensuing adaptive immune responses to limit establishment and spread of (viral) infection [44,56,57]. Pattern-recognition receptors The first and best characterized PRRs are members of the Toll-like receptor (TLR) family, identified approximately fifteen years ago and comprising ten members in humans [58 6]. These transmembrane proteins are expressed at the cell surface and in endosomal compartments of immune and non-immune cells. TLRs expressed at the plasma membranes 16

17 General introduction recognize mainly lipid and protein components of invading pathogens, whereas ligands for the endosomal TLRs consist of nucleic-acids such as single stranded (ss)rna, double stranded (ds)rna, and unmethylated CpG DNA [61]. The C-type lectin receptors (CLRs) comprise another family of transmembrane PRRs and mainly detect glycosylated PAMPs from viruses, fungi, and bacteria [62,63]. Cytoplasmic families of PRRs include the RNA-sensors retinoic acid-inducible gene (RIG) and melanoma differentiation-associated gene 5 (MDA5) that belong to the RIG-I-like receptor family (RLR) [64 67]. In addition, over twenty nucleotidebinding, oligomerization domain (NOD)-like receptors (NLRs) are encoded in the human genome. They are classified in different subfamilies based on the protein-protein interaction domain present in their N-terminal effector binding region [64,68]. The best characterized NLRs are the caspase-recruitment domain (CARD)-containing NLRCs and the pyrin domain (PYD) NLRPs [64,68]. The NLRCs NOD1 and NOD2 sense cytoplasmic bacterial components (e.g. peptidoglycan), whereas the NLRPs contribute to inflammasome function [69 72]. More recently, over ten cytoplasmic DNA sensors such as DNA-dependent activator of IFNregulatory factors (DAI) and cgas, and the AIM2-like receptors (ALRs) AIM2 and IFI16 have been identified [73 8]. These ALRs contain a pyrin protein interaction domain similar to the NLRPs [81]. Finally, the double-stranded RNA-dependent protein kinase R (PKR) and the 2,5 -oligoadenylate synthetase (OAS) are generally classified as cytoplasmic PRRs [82]. Activation of these PRRs induces intracellular signaling cascades that for example culminate in the activation of transcription factors NF-κB and/or interferon-regulatory factors (IRFs), inducing expression of proinflammatory effectors and type I interferons (IFNs) [54,83,84]. Alternatively, ALRs can induce generation and secretion of IL-1β and IL-18 through activation of inflammasomes, multimeric protein complexes dependent on caspase 1 [77,78]. There is ample crosstalk between these immune pathways and many PRRs synergistically activate multiple signaling pathways to tailor the response to individual pathogens [58,83,84]. 1 Proinflammatory signaling pathways leading to NF-κB activation Upon ligand binding, TLRs and NLRCs induce signaling pathways that activate transcription factor NF-κB and induce expression of proinflammatory mediators [54]. Activated, dimerized TLRs interact with Toll-IL-1 receptor (TIR)-domain containing adaptors such as MyD88 and TRIF [57,83,85]. MyD88, recruited by all TLRs except TLR3, forms a complex (the Myddosome ) with IL-1 receptor-associated kinase (IRAK) family members [86]. Subsequent IRAK4-mediated phosphorylation of IRAK1 leads to activation of the central signaling component tumor necrosis factor receptor-associated factor (TRAF) 6, a RING-domain E3 ubiquitin ligase [87 89]. Autoubiquitination of TRAF6 recruits regulatory subunits TAB1, TAB2, and TAB3 to form a complex with TAK1, driving TAK1 activation [9 92]. TAK1 subsequently catalyzes the formation of lysine(k)63-linked polyubiquitin chains on NFκB essential modulator (NEMO, or IκB kinase (IKK)γ), the regulatory subunit of the IKK 17

18 Chapter 1 complex additionally consisting of IKKα and IKKβ. Activation of the IKK complex leads to phosphorylation and ubiquitination of the inhibitor of NF-κB, IκBα. K48-linked ubiquitination targets IκBα for proteasomal degradation, releasing NF-κB to translocate into the nucleus and initiate transcription of genes encoding proinflammatory cytokines. Similarly, NLRs NOD1 and NOD2 recruit the kinases RIP2 and TRAF6 and converge at the downstream signaling pathway leading to NF-κB activation at TAK1 activation [7,93 96]. These signaling cascades are extensively regulated by post-translational modifications such as phosphorylation and ubiquitination [92,97 1]. Many viruses interfere with or mimic these regulatory processes to steer the innate response in a direction that favors viral replication [11,12]. IRF activation and production of type I interferons TLR3 and TLR4 activate TRIF, an adaptor molecule whose activation leads to production of type I IFNs [54,13]. TRIF recruits the E3 ubiquitin ligase TRAF3 [14,15]. TRAF3 activation is followed by induction of IKK-related kinases TANK-binding kinase 1 (TBK1) and IKKi, together with NEMO. TBK1 and IKKi phosphorylate the transcription factors interferon regulator factor (IRF) 3 and IRF7, causing their homodimerization and nuclear translocation to induce type I IFN production. Alternatively, TRIF also interacts with TRAF6 to induce NF-κB activation and delayed production of proinflammatory cytokines. In plasmacytoid DCs, known for their ability to secrete large amounts of type I IFNs in response to viral infection, TLR7 and TLR9 induce IRF7 transcriptional activity in an alternative, MyD88- dependent manner [16]. In this case, phosphorylation and nuclear translocation of IRF7 is induced by a complex containing MyD88, IRAK-1, TRAF6, TRAF3, IKKα. The RLRs induce activation of TRAF3 [54] through a CARD-domain interaction between the RLR and IFNβ-promoter stimulator 1 (IPS-1) present on the mitochondrial membrane [54,17]. The downstream signaling intermediates are shared with the TLR signaling pathway. Finally, signaling induced by the cytoplasmic DNA sensors converges at the central adaptor molecule stimulator of interferon genes (STING) that induces type I IFN production through TBK1 and IRF3 [18 111]. Type I interferons Type I interferons are important in limiting virus spread by induction of cell-intrinsic antimicrobial states in autocrine and paracrine fashion [112,113]. The most well-defined type I IFNs, IFNα and IFNβ, bind to the heterotrimeric type I IFN receptor (IFNAR) on the surface of target cells [114]. In the canonical signaling pathway, IFNAR ligation activates the receptorassociated protein tyrosine kinases Janus kinase 1 (JAK1) and tyrosine kinase 2 (TYK2), followed by phosphorylation and nuclear translocation of signal transducer and activator of transcription 1 (STAT1) and STAT2 [115]. The STAT proteins associate with interferonregulatory factor 9 (IRF9) to form the IFN-stimulated gene factor 3 (ISGF3) complex. This 18

19 General introduction complex binds IFN-stimulated response elements (ISREs) and recruits various chromatin remodeling and transcriptional activator complexes to induce transcription of several hundreds of interferon-stimulated genes (ISGs) [112, ]. The exact outcome of IFNAR ligation largely depends on the cell type, state of the immune response, and crosstalk with other signaling pathways [119]. The ISG-encoded proteins include transcription factors and many other effector proteins that restrict viral replication by interfering with the viral life cycle at various stages and by altering cellular processes such as protein translation and lipid metabolism [12] Detection of EBV by PRRs Several of the PRRs described earlier contribute to the innate detection of EBV infection [39,121, ]. These include the membrane-bound TLR2, TLR3, TLR7, and TLR9, as well as some cytoplasmic RLRs. In addition, a role for the cytoplasmic DNA sensors is now also emerging. TLR2 The lipopeptide sensor TLR2 is expressed on the surface of monocytes and, although mainly known for detecting bacterial infections, can also respond to certain viral PAMPs [54,83]. TLR2 is activated by components of HSV-1 [ ], VZV [127], and HCMV [128,129]. For HSV-1, however, the ability to activate TLR2 varies between different strains of HSV-1 [13]. Although the exact viral ligands remain unknown, several observations suggest the glycoproteins in the viral membranes are involved. In the case of EBV, TLR2 is activated by the viral nonstructural protein dutpase (encoded by BLLF3) in a MyD88-dependent fashion [131,132]. Exosomal secretion of dutpase even induces NF-κB activation and cytokine secretion in uninfected DCs and PBMCs [133]. We also observed TLR2-mediated NF-κB activation by EBV particles [134]. Paradoxically, the exact role of TLR2 in control of herpesvirus infections is osbcure, as both deleterious and protective effects of TLR2 stimulation on viral replication have been reported. For example, TLR2-deficiency protects neonatal mice against HSV-1-associated encephalitis [135], while TLR2 stimulation is beneficial for the host during HSV-2 and MCMV infection in mouse models [125,136]. Nonetheless, the association between disease severity during HSV- 2 and HCMV infection with certain TLR2 mutations in human genetic studies supports a protective function for TLR2 during human herpesvirus infections [135,137]. TLR3 The endosomal dsrna sensor TLR3 is constitutively expressed in various cell types and relies on the adaptor molecule TRIF to induce type I IFN production [138]. TLR3 is important for activation of conventional DCs (cdcs) and subsequent priming of T cell responses, as well 19

20 Chapter 1 IFNAR1 IFNAR2 TLR2 TLR7 TLR3 Endosome Nucleus Cytosol Endoplasmic reticulum MyD88 IRAK1 IRAK4 ub ub ub TRAF6 TAB2/3 TAK1 TAB1 ub ub ub IKKα NEMO IKKβ ub ub ub IκBα NF-κB BGLF5 NOD1 NOD2 RIP2 BPLF1 EBNA1 BPLF1 BZLF1 BRLF1 LF2 BGLF4 LMP1 MyD88 TRAF3 TBK1 NEMO IRF3/7 IKKi LMP1 BGLF5 TLR9 MyD88 TRIF LMP2a RIG-I MDA5 IPS-1 Mitochondrion TBK1 IRF3/7 type I IFNs BZLF1 NF-κB BGLF4 Proinflammatory cytokines type I IFNs DAI cgas AIM2 IFI16 STING mir-bart16-5p p CBP p STAT1 STAT2 IRF9 ISRE LMP2a/b JAK1 TYK2 p STAT1 STAT2 p EBNA1 LMP1 IRF9 EBNA2 Interferon stimulated genes (ISGs) e.g. ISG15 and IFIT1 2

21 General introduction Figure 2. Schematic overview of the innate signaling pathways that play a role in EBV infection. EBV components are sensed by various pattern recognition receptors (PRRs). These include the MyD88-dependent (green) and TRIF-dependent (red) TLRs at the cell surface or in the endosomal compartment, as well as the cytosolic NLRs (orange), RLRs (pink), and DNA sensors including ALRs (blue). The PRRs in purple are known to be directly activated by EBV components. Ligation of PRRs induces signaling cascades resulting in production of proinflammatory cytokines and type I IFNs by the transcription factors NF-κB and the IRFs. Type I IFNs are secreted and activate the IFNAR receptor in an autocrine and paracrine manner, inducing JAK/STAT signaling pathways. This induces the transcriptional activity of ISRE promoter elements and leads to production of interferon stimulated genes that perform various antiviral activities. EBV encodes several proteins and mirnas implicated in modulation of these innate responses (grey, gene products described in this thesis are depicted in bold). 1 as cross-presentation of viral antigens to antigen-presenting cells (APCs) [139]. Expression of this TLR is strongly induced by the actions of IFNs during viral infection, establishing a positive feedback loop [14]. Potential dsrna ligands for TLR3 are produced by many DNA viruses (and positive-stranded RNA viruses) at some point during their replication [141]. Exactly how cytoplasmic RNA species are detected by endosomal TLRs remains elusive, although it is speculated that authophagy is involved [142]. TLR3 plays a pivotal role in T-cell priming during HSV-1 infection [143]. The finding that two cases of HSV-1-associated encephalitis were homozygous for a rare TLR3 mutation and that a less active TLR3 variant was present in a patient with recurrent HSV-2-related meningitis illustrates the involvement of TLR3 in herpesvirus infection [ ]. In mice, TLR3-deficiency leads to increased viral loads and reduced serum cytokine levels during MCMV infection [147]. In the case of EBV, the EBV-encoded small RNAs EBER1 and EBER2 form dsrna structures that are recognized by TLR3. Exosome-mediated release of EBERs from EBV-infected cells leads to activation and maturation of DCs [148]. Notably, sera of patients with IM or CAEBV contain elevated amounts of EBERs, and the concomitant release of large amounts of proinflammatory cytokines may add to the pathogenesis of these disorders [148]. TLR7 Together with TLR9, the ssrna sensor TLR7 is involved in the IFN-α response to MCMV infection [149]. Exactly where ssrna ligands for TLR7 during herpesvirus infection are derived from is unknown. They may be generated during viral replication, but the viral mrna found in virions of all herpesvirus subfamilies provide an alternative source of ligands that potentially activate TLR7 at early times of infection [15 152]. Although there is no evidence of TLR7 stimulation by EBV components, primary EBV infection increases TLR7 expression. Furthermore, TLR7 activation by exogenous ligands induces EBV-encoded LMP-1 expression, priming cells for IFN responses following TLR3 or TLR9 stimulation [153,154]. 21

22 Chapter 1 TLR9 Expression of the endosomal dsdna sensor TLR9 is restricted to plasmacytoid DCs (pdcs) and B cells [54,83]. pdcs are well-known for their ability to secrete copious amounts of type I IFNs during viral infection [155]. As opposed to other immune cells, in which type I IFN production is mostly induced by cytoplasmic PRRs, pdcs express membrane-bound TLR7 and TLR9 to induce type I IFNs in response to pathogen-derived nucleic acids. TLR9 is activated by murine herpesviruses MCMV [147,149,156,157] and MHV68 [158]. Studies with these murine models demonstrate the importance of TLR9-mediated pdc activation and IFN production in the control of herpesvirus infections [147,156,158]. Unmethylated, CpG-containing motifs that serve as TLR ligands are abundantly present within herpesvirus genomes and all three families of human herpesviruses are recognized by TLR9 [126, ]. We observed TLR9-mediated induction of NF-κB activity by EBV virions [167], and pdcs control EBV infection by inducing TLR9-mediated IFN-α secretion [159,16]. The DNA serving as TLR9 ligand during EBV infection is expected to come from virionassociated viral genomes early after release in infected cells, as the EBV genome circularizes and gets extensively methylated following establishment of latency [168,169]. Paradoxically, exogenous stimulation of TLR9 on B cells benefits transformation and proliferation of infected cells, but on the other hand decreases cell survival and reduces the efficiency of productive infection [17 173]. RIG-I-Like receptors As viral replication occurs intracellularly, it seems plausible that infection is sensed by cytoplasmic PRRs in addition to the TLRs that scan the extracellular milieu [174]. In contrast to restricted expression of TLRs, RLRs are expressed by nearly all cells in the human body, and are strongly induced by type I IFNs. Until now, however, our knowledge on RLR mediated recognition of herpesviruses remains far from complete [121]. HSV-1 is known to induce IFN responses in macrophages mediated by MDA5 and MAVS [175]; and RIG-I and MAVS are involved in KSHV sensing [176]. Concerning EBV, the EBV-encoded non-coding RNA molecules (EBERs) are detected by RIG-I, leading to secretion of IL-1 and type I IFNs [ ]. Although it may also be caused by a lack of attention this area of research received, the limited amount of reports on RLR involvement in herpesvirus infections suggests that these receptors are not a major player in detections of herpesvirus infections. DNA sensors Cytoplasmic DNA sensors were identified more recently and their involvement in viral infection is only beginning to be elucidated. The first indications for induction of antiviral responses by herpesviral DNA were found a long time ago [164,18], and the more recent identification of DAI as a cytoplasmic sensor for HSV-1- and HCMV-derived DNA has finally 22

23 General introduction delivered a potential receptor mediating these effects [8,181,182]. In addition, the cytosolic sensors DHX9 and DHX36 induce cytokine secretion in pdcs following HSV-1 infection [183], and AIM2 induces inflammasome activation during MCMV infection [184]. Finally, the inflammasome-inducing ALR IFI16 senses HSV-1 [73,185], HCMV [186], VZV [187], and KSHV [188,189]. Although herpesviruses replicate in the nucleus, proteasomal degradation of virions upon entrance of a cell leads to release of viral genomic DNA that may serve as cytoplasmic ligand [169]. Latent EBV infection is associated with constitutive IFI16 activity [19], and consequent EBV-induced inflammasome activity may explain the elevated levels of the inflammasome-dependent cytokine IL-18 observed during acute EBV infection [191]. Analogous to observations for other PRRs, some herpesviruses seem to have evolved to use DNA sensor activation to their benefit: HCMV and MCMV rely on IFI16 activation for successful establishment of infection [192,193]. 1 Adaptive immunity The innate immune system activated during primary EBV infection and viral reactivation directly affects viral replication, yet is also important in orchestrating the ensuing adaptive immune response [44]. Adaptive antiviral immunity relies on a powerful, memory-based response of virus-specific B and T lymphocytes. Once EBV infection has been established, the virus resides intracellularly most of the time. Detection and elimination of EBV-infected cells depends to a large extent on T cells, which recognize virus-derived peptides in the context of surface HLA molecules. Primary EBV infection induces strong, virus-specific T cell responses targeting both lytic and latent EBV-derived epitopes. In addition, HLA class II-restricted CD4 + T cells are an essential component of the adaptive immune response against EBV, especially since EBV infects B lymphocytes cells that express high levels of HLA II. EBV-specific memory T helper responses can eliminate virus-producing cells either directly through their cytotoxic capacity, or indirectly through activation of CD8 + cytotoxic T cells and B cells. Indeed, following lytic EBV infection, CD4 + T cells against lytic antigens are readily detectable in the peripheral blood of infected individuals [194]. The importance of CD4 + T cell responses in controlling EBV infection is illustrated by the increased frequencies of EBVassociated malignancies that occur in patients with a defective CD4 + T cell compartment, such as HIV-infected individuals [195]. Taken together, PRR-mediated detection of herpesviruses including EBV is important to limit viral replication by inducing direct antiviral responses and through regulation of the adaptive response. Especially the membrane-bound DNA sensor TLR9 and the more recently identified cytosolic DNA sensors (e.g. ALRs) have been implicated extensively in detection of EBV and herpesviruses in general. It thus seems that DNA serves as the main inducer of innate antiviral responses during herpesvirus infection, especially for the early induction 23

24 Chapter 1 prior to viral replication [121]. This may not be surprising given the DNA genomes these viruses possess, analogous to the importance of RNA sensors in detection of RNA-virus infection [196]. Besides the question of which receptors are activated during EBV infection, another issue that remains unclear is the purpose of receptor stimulation. For example, despite the ample in vitro and in vivo evidence of TLR7, TLR8, and TLR9 activation during herpesvirus infection, IRAK- 4 deficient young children with defective type I IFN-production following activation of these TLRs do not display an increased vulnerability to herpesviruses, but only to bacterial infections [197]. Although this might suggest a minor role for IRAK4-mediated signaling pathways in controlling EBV infection at this age, the IRAK4 deficiency may also be compensated for by other innate systems, illustrating the redundancy in type I IFN-inducing systems. However, also the previously mentioned TLR2 and IFI16 have both beneficial and detrimental effects on viral replication. These combined observations show that the effect of PRR-associated signaling pathways on infection is likely multifactorial, depending on the local circumstances at the site of infection or for example the age of the host. Moreover, viruses may have evolved to use the antiviral defenses for their benefit, as exemplified by the dependence of HCMV and MCMV on IFI16 activation for successful infection. 24

25 Part III: Immune evasion by Epstein-Barr virus General introduction Even in the face of functional immunity in healthy individuals, EBV persists for life in over 9% of the world population. The long coevolution between the virus and its host has led to a fine-tuned interaction of EBV with the human immune system. During the latent phase, the number of viral gene products is limited, thereby reducing the amount of viral targets available for immune detection. With the development of model systems to study EBV infection of B and epithelial cells, it has become apparent that EBV encodes a wide range of gene products that are non-essential for replication in vitro, but contribute to creating an in vivo environment that is beneficial to maintenance of the viral episome and occasional reactivation of EBV. Following the identification of multiple EBV immune evasion molecules and elucidation of their mechanisms of action, it is estimated that over half of the viral gene products is dedicated to functions that modulate anti-viral responses of the host. Furthermore, the incorporation of EBV mrnas and non-coding mirnas into viral particles provides a means to immediately express immune modulatory gene products in newly infected cells [198]. The last part of this chapter describes the strategies adopted by EBV to interfere with host immunity, focussing on the innate immune system during both the latent and productive phases of infection (Figure 2) Immune evasion during productive infection Reduction of Toll-like receptor expression Like other γ-herpesviruses, EBV inhibits cellular protein synthesis in productively infected cells through global mrna destabilization. This process, termed shutoff, is mediated by the EBV DNase (alkaline exonuclease) BGLF5, which is expressed with early kinetics during the productive phase of infection [199]. BGLF5 s additional RNase function utilizes the same catalytic site as its DNase activity, yet the substrate-binding site appears only partly shared by DNA and RNA substrates [2]. The promiscuous RNA degradation induced by EBV BGLF5 can affect immunologically relevant proteins, including TLR2 and TLR9 that are capable of sensing EBV infection [131,167,21]. The observation that BGLF5 did not downregulate TLR4, a TLR not reported to contribute to EBV detection, suggests some selectivity in shutoff [131,21]. By analogy to herpes simplex virus (HSV)-2 infection of epithelial cells, where shutoff induced by the virion host shutoff (vhs) protein contributes to downregulation of the PRRs TLR2, TLR3, RIG-I, and MDA-5 [22], the effects of BGLF5 may well extend to additional immune components. Indeed, many TLRs are differentially expressed in lytically EBV-infected B cells by mechanisms that remain elusive and may involve shutoff effects [167]. In vivo studies on HSV and the murine γ-herpesvirus MHV68 indicate that shutoffinduced reductions in protein levels mainly prevent production of newly synthesized effector 25

26 Chapter 1 molecules rather than reducing the levels of existing PRRs [23 25]. It remains to be assessed to what extent EBV BGLF5 affects PRR expression levels, and subsequent signal transduction, in vivo. Modulation of IRF signaling and type I interferon production A number of lytic EBV proteins interferes with the actions of host IRFs, the transcription factors that induce type I IFN production. The immediate-early EBV transactivator BZLF1 interacts with IRF7 and inhibits its transcriptional activity on the IFN-α4 and IFN-β promoters to prevent induction of an antiviral environment [26]. The other immediate-early EBV transactivator, BRLF1, reduces expression of both IRF3 and IRF7, thereby inhibiting production of IFN-β [27]. The EBV tegument protein LF2, which is present in virions and gains access to the cell immediately upon virus entry, targets IRF7 and suppresses IFN-α production [28]. Finally, the EBV protein kinase BGLF4 phosphorylates and inhibits IRF3 transcriptional activity, thereby reducing IFN-β expression in response to treatment with the TLR3 agonist poly(i:c) [29]. Recently, Dunmire et al. reported that a clear systemic IFN response is observed during acute EBV infection, but this response lacks some key components compared to observations for other viruses [21]. This may reflect the successful actions of the immune evasion mechanisms employed by EBV to repress secretion of interferon responsive genes. Interference with NF-κB and inflammatory pathways In general, productive EBV infection is associated with a reduction in NF-κB-dependent gene expression [211]. Viral BZLF1 and cellular NF-κB reciprocally inhibit each other s expression and, as a consequence, higher levels of NF-κB in the absence of BZLF1 favor EBV latency, whereas increased expression of BZLF1 upon lytic cycle induction overwhelms the limiting amount of NF-κB [ ]. While NF-κB is still translocated to the nucleus, its transcriptional activity is suppressed by BZLF1, preventing induction of antiviral immune effector mechanisms [214]. TLR signaling pathways leading to NF-κB activation are tightly controlled by posttranslational modifications, such as phosphorylation and ubiquitination [97,1]. EBV encodes lytic proteins that interfere with these modifications. For example, phosphorylation of the NF-κB coactivator UXT is targeted by BGLF4 to suppress NF-κB activity [215]. BPLF1, the EBV homolog of the conserved herpesvirus deubiquitinase, reverses ubiquitination of several TLR signaling intermediates [134]. This inhibits NF-κB activation and cytokine production following TLR stimulation [134,216], and promotes viral genome replication [216]. Being a component of the EBV tegument, BPLF1 could act both in productively as well as in newly infected cells [134,217]. 26

27 General introduction Interference with innate effector molecules The abovementioned transcription factors induced upon PRR engagement greatly alter cellular gene expression to effectuate diverse effector mechanisms. Among these is secretion of effector molecules, such as proinflammatory cytokines and interferons, which act in autocrine and paracrine ways. A number of EBV gene products interferes with the functions of these innate effector molecules. EBV counteracts the pleiotropic host cytokine colony-stimulating factor 1 (CSF-1), which stimulates macrophage differentiation and IFN-α secretion. To this end, EBV encodes a soluble form of the CSF-1 receptor, BARF1, that neutralizes the effects of host CSF-1 in vitro, leading to reduced IFN-α secretion by EBV-infected mononuclear cells [218,219]. Mutating the BARF1 homologue in a related rhesus macaque lymphocryptovirus decreases viral load during primary infection and leads to a lower persistence setpoint in vivo [22]. EBV BZLF1 counteracts innate effector molecules in several ways. First, BZLF1 downregulates the receptors for TNF-α and IFN-γ to reduce cellular responsiveness to these cytokines [ ]. Second, BZLF1 induces the suppressor of cytokine signaling SOCS3, which inhibits JAK/ STAT signaling and thereby favors a state of type I IFN-irresponsiveness [224]. Additionally, SOCS3 reduces IFN-α production by monocytes. Third, BZLF1 causes expression of the immunosuppressive cytokine TGF-β [225] and disrupts the formation of PML-bodies [226], which can have antiviral activity [227]. The early-expressed EBV-encoded dutpase (encoded by the BLLF3 gene) also modulates cytokine-induced responses. In a mouse model, dutpase compromises lymphocyte responses to stimulation, e.g. secretion of IFN-γ [228]. In human cells, EBV dutpase has seemingly opposing effects: it induces NF-κB activation in a TLR2/MyD88-dependent way [132,133,229], it inhibits lymphocyte proliferation, and it induces production of both proinflammatory cytokines as well as IL-1 [23,231]. Following this strategy, EBV appears to exploit the advantageous effects of NF-κB activation, while limiting ensuing antiviral T-cell responses. 1 Adaptive immune evasion EBV compromises activation of both CD8 + and CD4 + T cells by interfering at various stages of the HLA class I and class II antigen presentation pathways, in particular during the productive phase of infection. As all nucleated human cells express HLA class I molecules (HLA I), also EBV-infected cells will be targeted by antiviral CD8 + T cells. To prevent recognition by EBVspecific (memory) T cells, EBV encodes at least three proteins that independently interfere with antigen presentation through downregulating surface expression of HLA I in distinct ways. These three viral proteins are expressed early during the replicative cycle of EBV and act in concert to prevent recognition by CD8 + T cells [232,233]. Firstly, degradation of HLA I-encoding mrnas by EBV BGLF5 reduces peptide presentation at the cell surface and, in turn, inhibits T-cell recognition [27,199]. During productive infection, silencing of BGLF5 27

28 Chapter 1 expression by 75% using shrnas reduces, but not completely blocks HLA I expression [21], indicating that other EBV gene products contribute to HLA I downregulation. One of these is BNLF2a, which expressed in isolation or in the context of EBV infection causes reduced CD8 + T cell recognition. BNLF2a depletes peptides from the ER (the HLA I loading compartment) through inhibition of peptide import by the transporter associated with antigen presentation (TAP) [234,235]. Furthermore, BILF1, encoding a constitutively active G-proteincoupled receptor (GPCR) downregulates HLA I from the cell surface [236]. The underlying mechanism involves reduced transport of HLA I from the trans-golgi network, as well as an increased turnover from the cell surface and, subsequently, enhanced degradation via lysosomal proteases [236]. Most HLA I haplotypes are downregulated by BILF1, yet HLA-C alleles appear resistant [237]; the latter could deviate NK cells. EBV also interferes with HLA II-restricted antigen presentation in various ways. The surface glycoprotein gp42 has initially been described as an entry receptor for EBV, binding to HLA class II molecules present on B cells. Additionally, gp42 acts as an immune evasion molecule. Its association with HLA class I/peptide complexes blocks T cell receptor (TCR)-class II interactions and interferes with activation of CD4 + T cells [238]. The viral interaction partners of gp42, gh and gl, cooperate to increase HLA II evasion (unpublished observation). In addition, the early host shut-off protein BGLF5 decreases cell-surface HLA II by degradation of HLA II mrnas [27]. Besides these ORFs that directly impair HLA II recognition, other EBV gene products indirectly interfere with CD4 + T cell immunity. The immediate-early protein BZLF1 impairs IFN-γ-signaling, thereby inhibiting CIITA promoter activity and, as a result, decreasing HLA II surface levels [222]. More recently, BZLF1 has been shown to impair HLA II presentation post-transcriptionally by interfering with the function of the invariant chain [239]. EBV also encodes a viral IL-1 homologue (BCRF1) that serves as an anti-inflammatory cytokine that is able to inhibit and modulate CD4 + T cell priming and effector functions [23]. Moreover, BCRF1 has been shown to inhibit co-stimulatory molecules on human monocytes, which potentially results in inefficient priming and expansion of CD4 + T cells [24]. 3.2 Immune evasion during latency EBV severely limits viral protein expression during latent infection to avoid recognition by the host immune system. Different forms of latency reflect the various stages leading from primary infection of the naive B cell to growth transformation. Immunomodulatory functions have been ascribed to several individual EBV proteins expressed during the various latency stages. EBNA1 EBNA1, expressed during all latency stages, contains a long glycine-alanine repeat that 28

29 General introduction inhibits translation as well as proteasomal degradation of EBNA1 through interference with processing by the 19S proteasomal subunit [ ]. This strategy ensures sufficient EBNA1 levels to maintain the viral genome [245], while decreasing protein turnover to minimize viral antigen presentation to CD8 + T cells. Initially, EBNA1-specific CD8 + T-cell responses were indeed not observed in vitro [242]. However, later studies did report EBNA1-specific T cell responses initiated by endogenously presented EBNA1-derived antigens [ ]. Potential source of these antigens include defective ribosomal products that lack the glycine-alanine repeat, or cross-presentation of exogenous antigens released by EBV-infected cells. Other immune evasive actions of EBNA1 include inhibition of the canonical NF-κB pathway by interfering with phosphorylation of the IKK complex signaling intermediate [249] and modulation of the STAT1 and TGF-β signaling pathways [25]. 1 EBNA2 EBNA2 applies a double-edged strategy by inducing low-level IFN-β production that leads to interferon-stimulated gene (ISG) production in BL cell lines [251], whereas anti-proliferative effects are neutralized by EBNA2-mediated inhibition of selected ISGs [252,253] and enhanced transcriptional activity of STAT3 [254] following IFN-α production. STAT3 modulates IFNinduced immune responses through STAT1 and suppresses production of inflammatory mediators [255]. In addition, EBNA2 upregulates the IL-18 receptor on BL cells [256]. IL-18 plays a role in regulating innate and adaptive immune responses and is elevated in certain EBV-associated malignancies [191]. LMP1 LMP1 promotes B cell growth and survival by mimicking constitutive CD4 signaling to activate NF-κB, JNK, MAPK, JAK/STAT and PI3K signaling pathways [257]. These pathways affect many immunological processes and allow LMP1 to steer the host immune response [258]. LMP1-mediated NF-κB activation in EBV-immortalized B cells results in type I IFN production that stimulates STAT1 expression in autocrine and paracrine fashion [ ]. STAT2 activity is inhibited by LMP1 [262]. LMP1-mediated upregulation of IRF7 benefits EBV by promoting cell growth, while at the same time an inhibitory IRF7 splice variant is induced to repress the adverse effects of type I IFN-production [ ]. Furthermore, LMP1-mediated induction of JAK/STAT signaling pathways may be advantageous to EBV as the antiviral activities of ISGs prevent superinfection and facilitate establishment of latency [268,269]. Finally, LMP-1 mediated NF-κB activation reduces TLR9 surface expression [27] and supplies growth benefits to infected cells [271,272]. LMP2a and 2b LMP2a inhibits NF-κB activity, IL-6 production, and subsequent JAK/STAT signaling 29

30 Chapter 1 pathways in carcinoma cell lines [273]. In contrast LMP2a induces NF-κB activation in B cells and uses the subsequently increased levels of anti-apoptotic Bcl-2 to protect infected cells from apoptosis in a transgenic mouse model [274]. Furthermore, LMP2a and LMP2b accelerate turnover of IFN receptors, resulting in decreased responsiveness of epithelial cells to IFN-α and IFN-γ [275]. EBERs Two approximately 17 nucleotide long RNA molecules called EBER-1 and EBER-2 are expressed during each of the latency stages [276]. The EBERs bind dsrna-dependent protein kinase R (PKR) in vitro [277] and inhibit apoptosis, but it is still a subject of debate whether suppression of IFN-α-induced apoptosis by the EBERs is due to inhibition of PKR [ ]. EBV mirnas EBV also expresses a number of mirnas, named the BART and BHRF1 mirnas [31,36, ]. These mirnas are organized in clusters and are expressed during all latency stages (and thus in EBV-associated tumors), although their expression levels differ between these stages [33 35]. mirnas have been detected in EBV virions, allowing their release in newly infected cells [152]. Moreover, EBV mirnas can be transferred via exosomes from EBV-infected cells to uninfected recipient cells in vitro, and may in this way regulate uninfected cells and cell types not typically infected by EBV [285,286]. Although our knowledge on the function of the EBV mirnas is limited, knock-out EBV strains have indicated a role for the BHRF1 mirnas and, to a lesser extent, for the BART mirnas in the early phase of B cell transformation [ ]. Experiments in mice showed that the BHRF1 mirnas were not crucial for infection, but knock-out virus did show delayed kinetics compared to the wild type virus [29]. Apart from these growth transforming and anti-apoptotic functions, EBV mirnas target several host genes involved in anti-viral immunity. Among the first EBV mirna targets identified was CXCL-11, a T cell attracting chemokine downregulated by EBV BHRF1-3 [291]. Stress-induced NK cell ligands have been specifically investigated as potential viral mirna targets. Initially, MICB was identified as a target of the HCMV encoded mir-ul112, and later it became apparent that also mirnas encoded by KSHV and EBV downregulate MICB expression. Inhibiting EBV mirna BART2-5p results in increased NK cell killing in vitro [292]. Inflammasomes are induced by various cytoplasmic and nuclear sensors (e.g. NLRP3 and IFI16) and lead to production of the inflammatory cytokines IL-1β and IL-18 [293]. Although EBV has so far only been observed to activate inflammasomes through IFI16 [19], EBV mirna BART15 downregulates the alternative inflammasome-activating sensor NLRP3 [294]. Co-culture of monocytic recipient cells with EBV + B cells secreting BART15-containing exosomes results in decreased IL-1β production. Additionally, EBV mirnas regulate the IFN-γ-STAT1 pathway in EBV + NK cells by 3

31 General introduction downregulating IFNγ transcriptional regulator T-bet (BART2-5p), IFN-γ (BART2-5p), and STAT1 (BART8) [295,296]. Inhibition of BART6-3p in a BL cell line caused upregulation of the IL-6 receptor chains (p8 and gp13) at the mrna and protein level, indicating that BART6-3p may affect IL-6 signaling [297]. Technological developments have greatly aided the search for mirna targets. A number of RISC immunoprecipitation screens on EBV infected B cells have generated a long list of putative mirna targets that await functional validation, including C-type lectin receptors and PML body components [289,298 32]. 1 Scope of this thesis In order to successfully establish infection and replicate, EBV has to withstand host antiviral responses. The innate immune system forms an important barrier against EBV infection by inducing direct antiviral responses and by orchestrating the adaptive immune response. During the long coevolution with the human host, EBV has acquired many immune evasion strategies that have led to a delicate balance between immune activation and evasion during infection. Over the past decades, many adaptive evasion strategies have been identified, but knowledge on the innate evasion strategies acquired by EBV is lagging behind. The aim of this thesis is to elucidate novel strategies that are employed by EBV to interfere with PRRmediated innate immunity. Chapter 2 describes the downregulation of several TLR family members from the surface of B cells during productive EBV infection. We further present the contribution of the EBV-encoded shutoff protein BGLF5 to downregulation of TLR9, a PRR capable of sensing EBV. In chapter 3, an shrna-approach is employed to silence BGLF5 expression and thereby inhibit the associated shutoff of protein synthesis in the infected B cell. We identify TLR2 and CD1d as additional targets of BGLF5. Chapter 4 describes the immune evasive properties of the conserved herpesvirus large tegument protein, BPLF1, that acts as a deubiquitinase in EBV-infected cells. We provide evidence for the deubiquitination of several signaling intermediates in the TLR pathway by BPLF1 to inhibit NF-κB activation and reduce production of inflammatory cytokines. Chapter 5 reports the contribution of an EBV mirna, mir-bart16, to immune evasion by downregulating expression of the cellular CREB-binding protein (CBP) to inhibit type I IFN-induced responses. Chapter 6 summarizes the findings presented in this thesis and discusses future perspectives of EBV immune evasion research. 31

32 Chapter 1 References 1. Longnecker RM, Kieff E, Cohen JI (213) Epstein- Barr Virus. In: Knipe DM, Howley PM, editors. Field s Virology. Philadelphia: Lippincott Williams & Wilkins. pp Hayward GS (1999) KSHV strains: the origins and global spread of the virus. Seminars in Cancer Biology 9: McGeoch DJ, Cook S, Dolan A, Jamieson FE, Telford EAR (1995) Molecular Phylogeny and Evolutionary Timescale for the Family of Mammalian Herpesviruses. Journal of Molecular Biology 247: Pellet PE, Roizman B (213) Herpesviridae. In: Knipe DM, Howley PM, editors. Fields Virology. Philadelphia: Lippincott Williams & Wilkins. pp Epstein MA, Achong BG, Barr YM (1964) Virus Particles in Cultured Lymphoblasts from Burkitt s Lymphoma. Lancet 283: Baer R, Bankier AT, Biggin MD, Deininger PL, Farrell PJ, et al. (1984) DNA sequence and expression of the B95-8 Epstein-Barr virus genome. Nature 31: Hjalgrim H, Friborg J, Melbye M (27) The epidemiology of EBV and its association with malignant disease. In: Arvin A, Campadelli-Fium G, Mocarski E, Moore P, Roizman B et al., editors. Human Herpesviruses: Biology, Therapy, and Immunoprophylaxis. Cambridge: Cambridge University Press. 8. Sample J, Young L, Martin B, Chatman T, Kieff E, et al. (199) Epstein-Barr virus types 1 and 2 differ in their EBNA-3A, EBNA-3B, and EBNA-3C genes. Journal of Virology 64: Chan CW, CHIANG AKS, Chan KH, au ASY (23) Epstein-Barr virus-associated infectious mononucleosis in Chinese children. The Pediatric Infectious Disease Journal Balfour HH, Odumade OA, Schmeling DO, Mullan BD, Ed JA, et al. (213) Behavioral, Virologic, and Immunologic Factors Associated With Acquisition and Severity of Primary Epstein-Barr Virus Infection in University Students. Journal of Infectious Diseases 27: Kutok JL, Wang F (26) Spectrum of Epstein-Barr Virus-Associated Diseases. Annu Rev Pathol Mech Dis 1: Okano M, Gross TG (1996) Epstein-Barr virusassociated hemophagocytic syndrome and fatal infectious mononucleosis. Am J Hematol 53: Shibata D, Weiss LM (1992) Epstein-Barr virusassociated gastric adenocarcinoma. Am J Pathol 14: Imai S, Koizumi S, Sugiura M, Tokunaga M, Uemura Y, et al. (1994) Gastric carcinoma: monoclonal epithelial malignant cells expressing Epstein-Barr virus latent infection protein. Proc Natl Acad Sci U S A 91: Yu MC, Yuan JM (22) Epidemiology of nasopharyngeal carcinoma. Seminars in Cancer Biology 12: Kuppers R (212) New insights in the biology of Hodgkin lymphoma. ASH Education Program Book 212: Rowe M, Kelly GL, Bell AI, Rickinson AB (29) Burkitt s lymphoma: The Rosetta Stone deciphering Epstein-Barr virus biology. Semin Cancer Biol 19: Thorley-Lawson DA, Allday MJ (28) The curious case of the tumour virus: 5 years of Burkitt s lymphoma. Nat Rev Micro 6: Jones JF, Shurin S, Abramowsky C, Tubbs RR, Sciotto CG, et al. (1988) T-Cell Lymphomas Containing Epstein- Barr Viral DNA in Patients with Chronic Epstein-Barr Virus Infections. N Engl J Med 318: Kawa-Ha K, Ishihara S, Ninomiya T, Yumura-Yagi K, Hara J, et al. (1989) CD3-negative lymphoproliferative disease of granular lymphocytes containing Epstein-Barr viral DNA. J Clin Invest 84: Parolini S, Bottino C, Falco M, Augugliaro R, Giliani S, et al. (2) X-Linked Lymphoproliferative Disease: 2b4 Molecules Displaying Inhibitory Rather than Activating Function Are Responsible for the Inability of Natural Killer Cells to Kill Epstein-Barr Virus-Infected Cells. The Journal of Experimental Medicine 192: Bottino C, Falco M, Parolini S, Marcenaro E, Augugliaro R, et al. (21) Gntb-A, a Novel Sh2d1a-Associated Surface Molecule Contributing to the Inability of Natural Killer Cells to Kill Epstein-Barr Virus-Infected B Cells in X-Linked Lymphoproliferative Disease. The Journal of Experimental Medicine 194: Hislop AD, Palendira U, Leese AM, Arkwright PD, Rohrlich PS, et al. (21) Impaired Epstein- Barr virus-specific CD8+ T-cell function in X-linked lymphoproliferative disease is restricted to SLAM family- 32

33 positive B-cell targets. Blood 116: Palendira U, Low C, Chan A, Hislop AD, Ho E, et al. (211) Molecular Pathogenesis of EBV Susceptibility in XLP as Revealed by Analysis of Female Carriers with Heterozygous Expression of SAP. PLoS Biol 9: e General introduction expression profile is disrupted in neoplasia. PLoS Pathog 7: e Barth S, Meister G, Grasser FA (211) EBV-encoded mirnas. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms 189: Kang MS, Kieff E (215) Epstein-Barr virus latent genes. Exp Mol Med 47: e Bartel DP (29) MicroRNAs: Target Recognition and Regulatory Functions. Cell 136: Babcock GJ, Decker LL, Volk M, Thorley-Lawson DA (1998) EBV Persistence in Memory B Cells In Vivo. Immunity 9: Rowe M, Glaunsinger B, van LD, Zuo J, Sweetman D, et al. (27) Host shutoff during productive Epstein-Barr virus infection is mediated by BGLF5 and may contribute to immune evasion. Proc Natl Acad Sci U S A 14: Johnson LS, Willert EK, Virgin HW (21) Redefining the genetics of Murine Gammaherpesvirus 68 via transcriptome-based annotation. Cell Host & Microbe 7: Ressing ME, Keating SE, van LD, Koppers-Lalic D, Pappworth IY, et al. (25) Impaired transporter associated with antigen processing-dependent peptide transport during productive EBV infection. J Immunol 174: Ressing ME, Horst Dl, Griffin BD, Tellam J, Zuo J, et al. (28) Epstein-Barr virus evasion of CD8+ and CD4+ T cell immunity via concerted actions of multiple gene products. Seminars in Cancer Biology 18: Pfeffer S, Zavolan M, Grasser FA, Chien M, Russo JJ, et al. (24) Identification of Virus-Encoded MicroRNAs. Science 34: Griffiths-Jones S, Grocock RJ, van Dongen S, Bateman A, Enright AJ (26) mirbase: microrna sequences, targets and gene nomenclature. Nucleic Acids Research 34: D14-D Amoroso R, Fitzsimmons L, Thomas WA, Kelly GL, Rowe M, et al. (211) Quantitative studies of Epstein-Barr virus-encoded micrornas provide novel insights into their regulation. J Virol 85: Martorelli D, Muraro E, Merlo A, Turrini R, Fae DA, et al. (212) Exploiting the Interplay between Innate and Adaptive Immunity to Improve Immunotherapeutic Strategies for Epstein-Barr-Virus-Driven Disorders. Journal of Immunology Research 212: Chijioke O, Azzi T, Nadal D, Munz C (213) Innate immune responses against Epstein Barr virus infection. Journal of Leukocyte Biology 94: Dupuis S, Jouanguy E, Al-Hajjar S, Fieschi C, Al- Mohsen IZ, et al. (23) Impaired response to interferona/b and lethal viral disease in human STAT1 deficiency. Nat Genet 33: Salazar-Mather TP, Lewis CA, Biron CA (22) Type I interferons regulate inflammatory cell trafficking and macrophage inflammatory protein 1a delivery to the liver. J Clin Invest 11: Biron CA, Byron KS, Sullivan JL (1989) Severe Herpesvirus Infections in an Adolescent without Natural Killer Cells. N Engl J Med 32: Rager-Zisman B, Quan PC, Rosner M, Moller JR, Bloom BR (1987) Role of NK cells in protection of mice against herpes simplex virus-1 infection. The Journal of Immunology 138: Iwasaki A, Medzhitov R (21) Regulation of Adaptive Immunity by the Innate Immune System. Science 327: Rickinson AB, Moss DJ (1997) Human Cytotoxic T Lymphocyte Responses to Epstein-Barr Virus Infection. Annu Rev Immunol 15: Hislop AD, Taylor GS, Sauce D, Rickinson AB (27) Cellular Responses to Viral Infection in Humans: Lessons from Epstein-Barr Virus. Annu Rev Immunol 25: Motsch N, Alles J, Imig J, Zhu J, Barth S, et al. (212) MicroRNA profiling of Epstein-Barr virus-associated NK/T-cell lymphomas by deep sequencing. PLoS One 7: e Braaten DC, Sparks-Thissen RL, Kreher S, Speck SH, Virgin HW (25) An Optimized CD8+ T-Cell Response Controls Productive and Latent Gammaherpesvirus Infection. Journal of Virology 79: Qiu J, Cosmopoulos K, Pegtel M, Hopmans E, Murray P, et al. (211) A novel persistence associated EBV mirna 48. Diebold J, Raphael M, Prevot S, Audouin J (1997) Lymphomas associated with HIV infection. Cancer 33

34 Chapter 1 surveys 3: Knowles DM (1999) Immunodeficiency-associated lymphoproliferative disorders. Modern pathology : an official journal of the United States and Canadian Academy of Pathology, Inc 12: McClain KL, Leach CT, Jenson HB, Joshi VV, Pollock BH, et al. (1995) Association of Epstein-Barr Virus with Leiomyosarcomas in Young People with AIDS. N Engl J Med 332: Basset C, Holton J, O Mahony R, Roitt I (23) Innate immunity and pathogen-host interaction. Vaccine 21, Supplement 2: S12-S Yan N, Chen ZJ (212) Intrinsic antiviral immunity. Nat Immunol 13: Akira S, Uematsu S, Takeuchi O (26) Pathogen Recognition and Innate Immunity. Cell 124: Takeuchi O, Akira S (21) Pattern Recognition Receptors and Inflammation. Cell 14: Chen GY, Nunez G (21) Sterile inflammation: sensing and reacting to damage. Nat Rev Immunol 1: Koyama S, Ishii KJ, Coban C, Akira S (28) Innate immune response to viral infection. Cytokine 43: Takeuchi O, Akira S (29) Innate immunity to virus infection. Immunological Reviews 227: Kawai T, Akira S (211) Toll-like Receptors and Their Crosstalk with Other Innate Receptors in Infection and Immunity. Immunity 34: O Neill LAJ, Golenbock D, Bowie AG (213) The history of Toll-like receptors [mdash] redefining innate immunity. Nat Rev Immunol 13: Kawai T, Akira S (21) The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nat Immunol 11: Kumar H, Kawai T, Akira S (29) Toll-like receptors and innate immunity. Biochemical and Biophysical Research Communications 388: Osorio F, Reis-áe-áSousa C (211) Myeloid C-type Lectin Receptors in Pathogen Recognition and Host Defense. Immunity 34: Geijtenbeek TBH, Gringhuis SI (29) Signalling through C-type lectin receptors: shaping immune responses. Nat Rev Immunol 9: Loo YM, Gale J (211) Immune Signaling by RIG-I-like Receptors. Immunity 34: Schlee M, Roth A, Hornung V, Hagmann CA, Wimmenauer V, et al. (29) Recognition of 5GÇ Triphosphate by RIG-I Helicase Requires Short Blunt Double-Stranded RNA as Contained in Panhandle of Negative-Strand Virus. Immunity 31: Schmidt A, Schwerd T, Hamm W, Hellmuth JC, Cui S, et al. (29) 5 -triphosphate RNA requires basepaired structures to activate antiviral signaling via RIG-I. Proceedings of the National Academy of Sciences 16: Pichlmair A, Schulz O, Tan CP, Rehwinkel J, Kato H, et al. (29) Activation of MDA5 Requires Higher-Order RNA Structures Generated during Virus Infection. Journal of Virology 83: Ting JPY, Lovering RC, Alnemri ES, Bertin J, Boss JM, et al. (28) The NLR Gene Family: A Standard Nomenclature. Immunity 28: Wen H, Miao E, Ting J (213) Mechanisms of NODlike Receptor-Associated Inflammasome Activation. Immunity 39: Moreira LO, Zamboni DS (212) NOD1 and NOD2 Signaling in Infection and Inflammation. Frontiers in Immunology 3: Shaw PJ, Lamkanfi M, Kanneganti TD (21) NOD-like receptor (NLR) signaling beyond the inflammasome. Eur J Immunol 4: Kanneganti TD, Lamkanfi M, Nunez G (27) Intracellular NOD-like Receptors in Host Defense and Disease. Immunity 27: Unterholzner L, Keating SE, Baran M, Horan KA, Jensen SB, et al. (21) IFI16 is an innate immune sensor for intracellular DNA. Nat Immunol 11: Sun L, Wu J, Du F, Chen X, Chen ZJ (213) Cyclic GMP- AMP Synthase Is a Cytosolic DNA Sensor That Activates the Type I Interferon Pathway. Science 339: Wu J, Chen ZJ (214) Innate Immune Sensing and Signaling of Cytosolic Nucleic Acids. Annu Rev Immunol 32: Paludan S, Bowie AG (213) Immune Sensing of DNA. Immunity 38: Hornung V, Ablasser A, Charrel-Dennis M, 34

35 Bauernfeind F, Horvath G, et al. (29) AIM2 recognizes cytosolic dsdna and forms a caspase-1 activating inflammasome with ASC. Nature 458: Roberts TL, Idris A, Dunn JA, Kelly GM, Burnton CM, et al. (29) HIN-2 Proteins Regulate Caspase Activation in Response to Foreign Cytoplasmic DNA. Science 323: Wang Z, Choi MK, Ban T, Yanai H, Negishi H, et al. (28) Regulation of innate immune responses by DAI (DLM-1/ZBP1) and other DNA-sensing molecules. Proceedings of the National Academy of Sciences 15: General introduction 89. Gohda J, Matsumura T, Inoue Ji (24) Cutting Edge: TNFR-Associated Factor (TRAF) 6 Is Essential for MyD88- Dependent Pathway but Not Toll/IL-1 Receptor Domain- Containing Adaptor-Inducing IFN-b (TRIF)-Dependent Pathway in TLR Signaling. The Journal of Immunology 173: Chen ZJ (212) Ubiquitination in signaling to and activation of IKK. Immunological Reviews 246: Ajibade AA, Wang HY, Wang RF (213) Cell typespecific function of TAK1 in innate immune signaling. Trends in Immunology 34: Takaoka A, Wang Z, Choi MK, Yanai H, Negishi H, et al. (27) DAI (DLM-1/ZBP1) is a cytosolic DNA sensor and an activator of innate immune response. Nature 448: Ratsimandresy RA, Dorfleutner A, Stehlik C (213) An update on PYRIN domain-containing pattern recognition receptors: from immunity to pathology. Frontiers in Immunology Dauber B, Wolff T (29) Activation of the Antiviral Kinase PKR and Viral Countermeasures. Viruses 1: Jiang X, Chen ZJ (211) The role of ubiquitylation in immune defence and pathogen evasion. Nat Rev Immunol 12: Hasegawa M, Fujimoto Y, Lucas PC, Nakano H, Fukase K, et al. (28) A critical role of RICK/RIP2 polyubiquitination in Nod-induced NF-kB activation. EMBO J 27: Park JH, Kim YG, McDonald C, Kanneganti TD, Hasegawa M, et al. (27) RICK/RIP2 Mediates Innate Immune Responses Induced through Nod1 and Nod2 but Not TLRs. The Journal of Immunology 178: Mogensen TH (29) Pathogen Recognition and Inflammatory Signaling in Innate Immune Defenses. Clin Microbiol Rev 22: Honda K, Taniguchi T (26) IRFs: master regulators of signalling by Toll-like receptors and cytosolic patternrecognition receptors. Nat Rev Immunol 6: Chandran B (21) Early Events in Kaposi s Sarcoma- Associated Herpesvirus Infection of Target Cells. Journal of Virology 84: Lin SC, Lo YC, Wu H (21) Helical assembly in the MyD88:IRAK4:IRAK2 complex in TLR/IL-1R signaling. Nature 465: Kollewe C, Mackensen AC, Neumann D, Knop J, Cao P, et al. (24) Sequential Autophosphorylation Steps in the Interleukin-1 Receptor-associated Kinase-1 Regulate its Availability as an Adapter in Interleukin-1 Signaling. Journal of Biological Chemistry 279: Jiang Z, Ninomiya-Tsuji J, Qian Y, Matsumoto K, Li X (22) Interleukin-1 (IL-1) Receptor-Associated Kinase-Dependent IL-1-Induced Signaling Complexes Phosphorylate TAK1 and TAB2 at the Plasma Membrane and Activate TAK1 in the Cytosol. Molecular and Cellular Biology 22: Nembrini C, Kisielow J, Shamshiev AT, Tortola L, Coyle AJ, et al. (29) The Kinase Activity of Rip2 Determines Its Stability and Consequently Nod1- and Nod2-mediated Immune Responses. Journal of Biological Chemistry 284: Kobayashi K, Inohara N, Hernandez LD, Galan JE, Nunez G, et al. (22) RICK/Rip2/CARDIAK mediates signalling for receptors of the innate and adaptive immune systems. Nature 416: Wertz IE, Dixit VM (21) Signaling to NF-kB: Regulation by Ubiquitination. Cold Spring Harbor Perspectives in Biology Boone DL, Turer EE, Lee EG, Ahmad RC, Wheeler MT, et al. (24) The ubiquitin-modifying enzyme A2 is required for termination of Toll-like receptor responses. Nat Immunol 5: Brummelkamp TR, Nijman SMB, Dirac AMG, Bernards R (23) Loss of the cylindromatosis tumour suppressor inhibits apoptosis by activating NF-kB. Nature 424: Deribe YL, Pawson T, Dikic I (21) Post-translational modifications in signal integration. Nat Struct Mol Biol 17:

36 Chapter Isaacson MK, Ploegh HL (29) Ubiquitination, Ubiquitin-like Modifiers, and Deubiquitination in Viral Infection. Cell Host & Microbe 5: Randow F, Lehner PJ (29) Viral avoidance and exploitation of the ubiquitin system. Nat Cell Biol 11: Kawasaki T, Kawai T (214) Toll-Like Receptor Signaling Pathways. Frontiers in Immunology 5: Hacker H, Redecke V, Blagoev B, Kratchmarova I, Hsu LC, et al. (26) Specificity in Toll-like receptor signalling through distinct effector functions of TRAF3 and TRAF6. Nature 439: Oganesyan G, Saha SK, Guo B, He JQ, Shahangian A, et al. (26) Critical role of TRAF3 in the Toll-like receptordependent and -independent antiviral response. Nature 439: Colonna M, Trinchieri G, Liu YJ (24) Plasmacytoid dendritic cells in immunity. Nature Immunology 5: Kawai T, Takahashum K, Sato S, Coban C, Kumar H, et al. (25) IPS-1, an adaptor triggering RIG-I- and Mda5-mediated type I interferon induction. Nature Immunology 6: Ishikawa H, Barber GN (28) STING is an endoplasmic reticulum adaptor that facilitates innate immune signalling. Nature 455: Sun W, Li Y, Chen L, Chen H, You F, et al. (29) ERIS, an endoplasmic reticulum IFN stimulator, activates innate immune signaling through dimerization. Proceedings of the National Academy of Sciences 16: Zhong B, Yang Y, Li S, Wang YY, Li Y, et al. (28) The Adaptor Protein MITA Links Virus-Sensing Receptors to IRF3 Transcription Factor Activation. Immunity 29: Burdette DL, Vance RE (213) STING and the innate immune response to nucleic acids in the cytosol. Nat Immunol 14: Ivashkiv LB, Donlin LT (214) Regulation of type I interferon responses. Nat Rev Immunol 14: McNab F, Mayer-Barber K, Sher A, Wack A, O Garra A (215) Type I interferons in infectious disease. Nat Rev Immunol 15: Platanias LC (25) Mechanisms of type-i- and type- II-interferon-mediated signalling. Nat Rev Immunol 5: Levy DE, Darnell JE (22) STATs: transcriptional control and biological impact. Nat Rev Mol Cell Biol 3: Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, et al. (211) A diverse range of gene products are effectors of the type I interferon antiviral response. Nature 472: MacMicking JD (212) Interferon-inducible effector mechanisms in cell-autonomous immunity. Nat Rev Immunol 12: Rusinova I, Forster S, Yu S, Kannan A, Masse M, et al. (213) INTERFEROME v2.: an updated database of annotated interferon-regulated genes. Nucleic Acids Research 41: D14-D van Boxel-Dezaire AHH, Rani MRS, Stark GR (26) Complex Modulation of Cell Type-Specific Signaling in Response to Type I Interferons. Immunity 25: Schoggins JW, Rice CM (211) Interferon-stimulated genes and their antiviral effector functions. Current Opinion in Virology 1: Paludan SR, Bowie AG, Horan KA, Fitzgerald KA (211) Recognition of herpesviruses by the innate immune system. Nat Rev Immunol 11: Ning S (211) Innate immune modulation in EBV infection. Herpesviridae 2: Di Paolo NC (214) Recognition of Human Oncogenic Viruses by Host Pattern-Recognition Receptors. Frontiers in Immunology 5: Kurt-Jones EA, Chan M, Zhou S, Wang J, Reed G, et al. (24) Herpes simplex virus 1 interaction with Toll-like receptor 2 contributes to lethal encephalitis. Proc Natl Acad Sci U S A 11: Sorensen LN, Reinert LS, Malmgaard L, Bartholdy C, Thomsen AR, et al. (28) TLR2 and TLR9 Synergistically Control Herpes Simplex Virus Infection in the Brain. The Journal of Immunology 181: Sarangi PP, Kim B, Kurt-Jones E, Rouse BT (27) Innate Recognition Network Driving Herpes Simplex Virus-Induced Corneal Immunopathology: Role of the Toll Pathway in Early Inflammatory Events in Stromal Keratitis. Journal of Virology 81: Wang JP, Kurt-Jones EA, Shin OS, Manchak MD, 36

37 Levin MJ, et al. (25) Varicella-Zoster Virus Activates Inflammatory Cytokines in Human Monocytes and Macrophages via Toll-Like Receptor 2. Journal of Virology 79: Boehme KW, Guerrero M, Compton T (26) Human Cytomegalovirus Envelope Glycoproteins B and H Are Necessary for TLR2 Activation in Permissive Cells. The Journal of Immunology 177: General introduction 138. Alexopoulou L, Holt AC, Medzhitov R, Flavell RA (21) Recognition of double-stranded RNA and activation of NF-kB by Toll-like receptor 3. Nature 413: Schulz O, Diebold SS, Chen M, Naslund TI, Nolte MA, et al. (25) Toll-like receptor 3 promotes crosspriming to virus-infected cells. Nature 433: Compton T, Kurt-Jones EA, Boehme KW, Belko J, Latz E, et al. (23) Human Cytomegalovirus Activates Inflammatory Cytokine Responses via CD14 and Toll- Like Receptor 2. J Virol 77: Sato A, Linehan MM, Iwasaki A (26) Dual recognition of herpes simplex viruses by TLR2 and TLR9 in dendritic cells. Proc Natl Acad Sci U S A 13: Miettinen M, Sareneva T, Julkunen I, Matikainen S (21) IFNs activate toll-like receptor gene expression in viral infections. Genes & Immunity 2: Weber F, Wagner V, Rasmussen SB, Hartmann R, Paludan SR (26) Double-Stranded RNA Is Produced by Positive-Strand RNA Viruses and DNA Viruses but Not in Detectable Amounts by Negative-Strand RNA Viruses. J Virol 8: Gaudreault E, Fiola S, Olivier M, Gosselin J (27) Epstein-Barr Virus Induces MCP-1 Secretion by Human Monocytes via TLR2. J Virol 81: Deretic V, Saitoh T, Akira S (213) Autophagy in infection, inflammation and immunity. Nat Rev Immunol 13: Ariza ME, Glaser R, Kaumaya PTP, Jones C, Williams MV (29) The EBV-Encoded dutpase Activates NFkB through the TLR2 and MyD88-Dependent Signaling Pathway. The Journal of Immunology 182: Ariza ME, Rivailler P, Glaser R, Chen M, Williams MV (213) Epstein-Barr Virus Encoded dutpase Containing Exosomes Modulate Innate and Adaptive Immune Responses in Human Dendritic Cells and Peripheral Blood Mononuclear Cells. PLoS ONE 8: e van Gent M, Braem SGE, de Jong A, Delagic N, Peeters JGC, et al. (214) Epstein-Barr Virus Large Tegument Protein BPLF1 Contributes to Innate Immune Evasion through Interference with Toll-Like Receptor Signaling. PLoS Pathog 1: e Kijpittayarit S, Eid AJ, Brown RA, Paya CV, Razonable RR (27) Relationship between Toll-Like Receptor 2 Polymorphism and Cytomegalovirus Disease after Liver Transplantation. Clinical Infectious Diseases 44: Szomolanyi-Tsuda E, Liang X, Welsh RM, Kurt- Jones EA, Finberg RW (26) Role for TLR2 in NK Cell- Mediated Control of Murine Cytomegalovirus In Vivo. Journal of Virology 8: Bochud PY, Magaret AS, Koelle DM, Aderem A, Wald A (27) Polymorphisms in TLR2 Are Associated with Increased Viral Shedding and Lesional Rate in Patients with Genital Herpes Simplex Virus Type 2 Infection. Journal of Infectious Diseases 196: Davey GM, Wojtasiak M, Proietto AI, Carbone FR, Heath WR, et al. (21) Cutting Edge: Priming of CD8 T Cell Immunity to Herpes Simplex Virus Type 1 Requires Cognate TLR3 Expression In Vivo. The Journal of Immunology 184: Zhang SY, Jouanguy E, Ugolini S, Smahi A, Elain Gl, et al. (27) TLR3 Deficiency in Patients with Herpes Simplex Encephalitis. Science 317: Ranjith-Kumar CT, Miller W, Sun J, Xiong J, Santos J, et al. (27) Effects of Single Nucleotide Polymorphisms on Toll-like Receptor 3 Activity and Expression in Cultured Cells. Journal of Biological Chemistry 282: Willmann O, Ahmad-Nejad P, Neumaier M, Hennerici MG, Fatar M (21) Toll-Like Receptor 3 Immune Deficiency May Be Causative for HSV-2- Associated Mollaret Meningitis. European Neurology 63: Tabeta K, Georgel P, Janssen E, Du X, Hoebe K, et al. (24) Toll-like receptors 9 and 3 as essential components of innate immune defense against mouse cytomegalovirus infection. Proceedings of the National Academy of Sciences of the United States of America 11: Iwakiri D, Zhou L, Samanta M, Matsumoto M, Ebihara T, et al. (29) Epstein-Barr virus (EBV)-encoded small RNA is released from EBV-infected cells and activates signaling from Toll-like receptor 3. The Journal of Experimental Medicine 26:

38 Chapter Zucchini N, Bessou G, Traub S, Robbins SH, Uematsu S, et al. (28) Cutting Edge: Overlapping Functions of TLR7 and TLR9 for Innate Defense against a Herpesvirus Infection. The Journal of Immunology 18: Bresnahan WA, Shenk T (2) A Subset of Viral Transcripts Packaged Within Human Cytomegalovirus Particles. Science 288: Sciortino MT, Suzuki M, Taddeo B, Roizman B (21) RNAs Extracted from Herpes Simplex Virus 1 Virions: Apparent Selectivity of Viral but Not Cellular RNAs Packaged in Virions. Journal of Virology 75: Jochum S, Ruiss R, Moosmann A, Hammerschmidt W, Zeidler R (212) RNAs in Epstein-Barr virions control early steps of infection. Proceedings of the National Academy of Sciences 19: E1396-E Martin HJ, Lee JM, Walls D, Hayward SD (27) Manipulation of the Toll-Like Receptor 7 Signaling Pathway by Epstein-Barr Virus. Journal of Virology 81: Valente RM, Ehlers E, Xu D, Ahmad H, Steadman A, et al. (212) Toll-Like Receptor 7 Stimulates the Expression of Epstein-Barr Virus Latent Membrane Protein 1. PLoS ONE 7: e Reizis B, Bunin A, Ghosh HS, Lewis KL, Sisirak V (211) Plasmacytoid Dendritic Cells: Recent Progress and Open Questions. Annu Rev Immunol 29: Krug A, French AR, Barchet W, Fischer JAA, Dzionek A, et al. (24) TLR9-Dependent Recognition of MCMV by IPC and DC Generates Coordinated Cytokine Responses that Activate Antiviral NK Cell Function. Immunity 21: Hokeness-Antonelli KL, Crane MJ, Dragoi AM, Chu WM, Salazar-Mather TP (27) IFN-ab-Mediated Inflammatory Responses and Antiviral Defense in Liver Is TLR9-Independent but MyD88-Dependent during Murine Cytomegalovirus Infection. The Journal of Immunology 179: Guggemoos S, Hangel D, Hamm S, Heit A, Bauer S, et al. (28) TLR9 Contributes to Antiviral Immunity during Gammaherpesvirus Infection. The Journal of Immunology 18: Fiola S, Gosselin D, Takada K, Gosselin J (21) TLR9 Contributes to the Recognition of EBV by Primary Monocytes and Plasmacytoid Dendritic Cells. The Journal of Immunology 185: Lim WH, Kireta S, Russ GR, Coates PTH (27) Human plasmacytoid dendritic cells regulate immune responses to Epstein-Barr virus (EBV) infection and delay EBV-related mortality in humanized NOD-SCID mice. Blood 19: Krug A, Luker GD, Barchet W, Leib DA, Akira S, et al. (23) Herpes simplex virus type 1 activates murine natural interferon-producing cells through toll-like receptor 9. Blood 13: Lund J, Sato A, Akira S, Medzhitov R, Iwasaki A (23) Toll-like Receptor 9-mediated Recognition of Herpes Simplex Virus-2 by Plasmacytoid Dendritic Cells. The Journal of Experimental Medicine 198: Varani S, Cederarv M, Feld S, Tammik C, Frascaroli G, et al. (27) Human Cytomegalovirus Differentially Controls B Cell and T Cell Responses through Effects on Plasmacytoid Dendritic Cells. The Journal of Immunology 179: Rasmussen SB, Sorensen LN, Malmgaard L, Ank N, Baines JD, et al. (27) Type I Interferon Production during Herpes Simplex Virus Infection Is Controlled by Cell-Type-Specific Viral Recognition through Toll-Like Receptor 9, the Mitochondrial Antiviral Signaling Protein Pathway, and Novel Recognition Systems. Journal of Virology 81: Lund JM, Linehan MM, Iijima N, Iwasaki A (26) Cutting Edge: Plasmacytoid Dendritic Cells Provide Innate Immune Protection against Mucosal Viral Infection In Situ. The Journal of Immunology 177: Wuest T, Austin BA, Uematsu S, Thapa M, Akira S, et al. (26) Intact TRL 9 and type I interferon signaling pathways are required to augment HSV-1 induced corneal CXCL9 and CXCL1. Journal of Neuroimmunology 179: van Gent M, Griffin BD, Berkhoff EG, van Leeuwen D, Boer IGJ, et al. (211) EBV Lytic-Phase Protein BGLF5 Contributes to TLR9 Downregulation during Productive Infection. The Journal of Immunology 186: Woellmer A, Arteaga-Salas JM, Hammerschmidt W (212) BZLF1 Governs CpG-Methylated Chromatin of Epstein-Barr Virus Reversing Epigenetic Repression. PLoS Pathog 8: e Horan KA, Hansen K, Jakobsen MR, Holm CK, Soby S, et al. (213) Proteasomal Degradation of Herpes Simplex Virus Capsids in Macrophages Releases DNA to the Cytosol for Recognition by DNA Sensors. The Journal of Immunology 19: Younesi V, Shirazi F, Memarian A, Amanzadeh A, Jeddi-Tehrani M, et al. (214) Assessment of the 38

39 effect of TLR7/8, TLR9 agonists and CD4 ligand on the transformation efficiency of Epstein-Barr virus in human B lymphocytes by limiting dilution assay. Cytotechnology 66: Iskra S, Kalla M, Delecluse HJ, Hammerschmidt W, Moosmann A (21) Toll-Like Receptor Agonists Synergistically Increase Proliferation and Activation of B Cells by Epstein-Barr Virus. Journal of Virology 84: Noack J, Jordi M, Zauner L, Alessi D, Burch A, et al. (212) TLR9 agonists induced cell death in Burkitt s lymphoma cells is variable and influenced by TLR9 polymorphism. Cell Death Dis 3: e Zauner L, Melroe GT, Sigrist JA, Rechsteiner MP, Dorner M, et al. (21) TLR9 triggering in Burkitt s lymphoma cell lines suppresses the EBV BZLF1 transcription via histone modification. Oncogene 29: Malmgaard L, Melchjorsen J, Bowie AG, Mogensen SrC, Paludan SrR (24) Viral Activation of Macrophages through TLR-Dependent and -Independent Pathways. The Journal of Immunology 173: Melchjorsen J, Rintahaka J, Soby S, Horan KA, Poltajainen A, et al. (21) Early Innate Recognition of Herpes Simplex Virus in Human Primary Macrophages Is Mediated via the MDA5/MAVS-Dependent and MDA5/ MAVS/RNA Polymerase III-Independent Pathways. Journal of Virology 84: : General introduction 181. DeFilippis VR, Sali T, Alvarado D, White L, Bresnahan W, et al. (21) Activation of the Interferon Response by Human Cytomegalovirus Occurs via Cytoplasmic Double-Stranded DNA but Not Glycoprotein B. Journal of Virology 84: DeFilippis VR, Alvarado D, Sali T, Rothenburg S, Fruh K (21) Human Cytomegalovirus Induces the Interferon Response via the DNA Sensor ZBP1. Journal of Virology 84: Kim T, Pazhoor S, Bao M, Zhang Z, Hanabuchi S, et al. (21) Aspartate-glutamate-alanine-histidine box motif (DEAH)/RNA helicase A helicases sense microbial DNA in human plasmacytoid dendritic cells. Proceedings of the National Academy of Sciences 17: Rathinam VAK, Jiang Z, Waggoner SN, Sharma S, Cole LE, et al. (21) The AIM2 inflammasome is essential for host defense against cytosolic bacteria and DNA viruses. Nat Immunol 11: Johnson KE, Chikoti L, Chandran B (213) Herpes Simplex Virus 1 Infection Induces Activation and Subsequent Inhibition of the IFI16 and NLRP3 Inflammasomes. Journal of Virology 87: Li T, Chen J, Cristea IM (213) Human cytomegalovirus tegument protein pul83 inhibits IFI16- mediated DNA sensing for immune evasion. Cell Host & Microbe 14: West JA, Wicks M, Gregory SM, Chugh P, Jacobs SR, et al. (214) An Important Role for Mitochondrial Antiviral Signaling Protein in the Kaposi s Sarcoma-Associated Herpesvirus Life Cycle. Journal of Virology 88: Samanta M, Iwakiri D, kanda T, Imaizumi T, Takada K (26) EB virus-encoded RNAs are recognized by RIG-I and activate signaling to induce type I IFN. The EMBO Journal 25: Samanta M, Iwakiri D, Takada K (28) Epstein-Barr virus-encoded small RNA induces IL-1 through RIG-Imediated IRF-3 signaling. Oncogene 27: Ablasser A, Bauernfeind F, Hartmann G, Latz E, Fitzgerald KA, et al. (29) RIG-I-dependent sensing of poly(da:dt) through the induction of an RNA polymerase III-transcribed RNA intermediate. Nat Immunol 1: Ishii KJ, Coban C, Kato H, Takahashi K, Torii Y, et al. (26) A Toll-like receptor-independent antiviral response induced by double-stranded B-form DNA. Nat Immunol 187. Nour AM, Reichelt M, Ku CC, Ho MY, Heineman TC, et al. (211) Varicella-Zoster Virus Infection Triggers Formation of an Interleukin-1b (IL-1b)-processing Inflammasome Complex. Journal of Biological Chemistry 286: Kerur N, Veettil M, Sharma-Walia N, Bottero V, Sadagopan S, et al. (211) IFI16 Acts as a Nuclear Pathogen Sensor to Induce the Inflammasome in Response to Kaposi Sarcoma-Associated Herpesvirus Infection. Cell Host & Microbe 9: Singh VV, Kerur N, Bottero V, Dutta S, Chakraborty S, et al. (213) Kaposi s Sarcoma-Associated Herpesvirus Latency in Endothelial and B Cells Activates Gamma Interferon-Inducible Protein 16-Mediated Inflammasomes. Journal of Virology 87: Ansari MA, Singh VV, Dutta S, Veettil MV, Dutta D, et al. (213) Constitutive Interferon-Inducible Protein 16-Inflammasome Activation during Epstein-Barr Virus Latency I, II, and III in B and Epithelial Cells. Journal of Virology 87:

40 Chapter van de Veerdonk FL, Wever PC, Hermans MHA, Fijnheer R, Joosten LAB, et al. (212) IL-18 Serum Concentration Is Markedly Elevated in Acute EBV Infection and Can Serve as a Marker for Disease Severity. Journal of Infectious Diseases 26: Cristea IM, Moorman NJ, Terhune SS, Cuevas CD, O Keefe ES, et al. (21) Human Cytomegalovirus pul83 Stimulates Activity of the Viral Immediate-Early Promoter through Its Interaction with the Cellular IFI16 Protein. Journal of Virology 84: Hertel L, De Andrea M, Azzimonti B, Rolle A, Gariglio M, et al. (1999) The Interferon-Inducible 24 Gene, a Member of the Ifi 2 Family, Is Not Involved in the Antiviral State Induction by IFN-a, but Is Required by the Mouse Cytomegalovirus for Its Replication. Virology 262: Mackay LK, Long HM, Brooks JM, Taylor GS, Leung CS, et al. (29) T Cell Detection of a B-Cell Tropic Virus Infection: Newly-Synthesised versus Mature Viral Proteins as Antigen Sources for CD4 and CD8 Epitope Display. PLoS Pathog 5: e Petrara MR, Freguja R, Gianesin K, Zanchetta M, De Rossi A (213) Epstein-Barr Virus-Driven Lymphomagenesis in the Context of Human Immunodeficiency Virus Type 1 Infection. Frontiers in Microbiology Rehwinkel J, Tan CP, Goubau D, Schulz O, Pichlmair A, et al. (21) RIG-I Detects Viral Genomic RNA during Negative-Strand RNA Virus Infection. Cell 14: Ku CL, von Bernuth H, Picard C, Zhang SY, Chang HH, et al. (27) Selective predisposition to bacterial infections in IRAK-4-deficient children: IRAK-4- dependent TLRs are otherwise redundant in protective immunity. The Journal of Experimental Medicine 24: Jochum S, Moosmann A, Lang S, Hammerschmidt W, Zeidler R (212) The EBV Immunoevasins vil-1 and BNLF2a Protect Newly Infected B Cells from Immune Recognition and Elimination. PLoS Pathog 8: e Zuo J, Thomas W, van Leeuwen D, Middeldorp JM, Wiertz EJHJ, et al. (28) The DNase of Gammaherpesviruses Impairs Recognition by Virus- Specific CD8+ T Cells through an Additional Host Shutoff Function. Journal of Virology 82: Horst D, Burrows SR, Gatherer D, van WB, Bell MJ, et al. (212) Epstein-Barr virus isolates retain their capacity to evade T cell immunity through BNLF2a despite extensive sequence variation. J Virol 86: van Gent M, Gram AM, Boer IGJ, Geerdink RJ, Lindenbergh MFS, et al. (214) Silencing the shutoff protein of Epstein-Barr virus in productively infected B cells points to (innate) targets for immune evasion. Journal of General Virology. 22. Yao XD, Rosenthal KL (211) Herpes simplex virus type 2 virion host shutoff protein suppresses innate dsrna antiviral pathways in human vaginal epithelial cells. Journal of General Virology 92: Murphy E, Vanicek J, Robins H, Shenk T, Levine AJ (28) Suppression of immediate-early viral gene expression by herpesvirus-coded micrornas: Implications for latency. Proc Natl Acad Sci U S A 15: Pasieka TJ, Lu B, Crosby SD, Wylie KM, Morrison LA, et al. (28) Herpes Simplex Virus Virion Host Shutoff Attenuates Establishment of the Antiviral State. Journal of Virology 82: Sheridan V, Polychronopoulos L, Dutia BM, Ebrahimi B (214) A Shutoff and Exonuclease Mutant of Murine Gammaherpesvirus-68 Yields Infectious Virus and Causes RNA loss in Type I Interferon Receptor Knock-Out cells. Journal of General Virology. 26. Hahn AM, Huye LE, Ning S, Webster-Cyriaque J, Pagano JS (25) Interferon Regulatory Factor 7 Is Negatively Regulated by the Epstein-Barr Virus Immediate-Early Gene, BZLF-1. Journal of Virology 79: Bentz GL, Liu R, Hahn AM, Shackelford J, Pagano JS (21) Epstein-Barr virus BRLF1 inhibits transcription of IRF3 and IRF7 and suppresses induction of interferon-ß. Virology 42: Wu L, Fossum E, Joo CH, Inn KS, Shin YC, et al. (29) Epstein-Barr Virus LF2: an Antagonist to Type I Interferon. Journal of Virology 83: Wang JT, Doong SL, Teng SC, Lee CP, Tsai CH, et al. (29) Epstein-Barr Virus BGLF4 Kinase Suppresses the Interferon Regulatory Factor 3 Signaling Pathway. Journal of Virology 83: Dunmire SK, Odumade OA, Porter JL, Reyes-Genere J, Schmeling DO, et al. (214) Primary EBV Infection Induces an Expression Profile Distinct from Other Viruses but Similar to Hemophagocytic Syndromes. PLoS ONE 9: e Keating Sa, Prince S, Jones M, Rowe M (22) The Lytic 4

41 Cycle of Epstein-Barr Virus Is Associated with Decreased Expression of Cell Surface Major Histocompatibility Complex Class I and Class II Molecules. Journal of Virology 76: Dreyfus DH, Nagasawa M, Pratt JC, Kelleher CA, Gelfand EW (1999) Inactivation of NF-kB by EBV BZLF- 1-Encoded ZEBRA Protein in Human T Cells. The Journal of Immunology 163: Gutsch DE, Holley-Guthrie EA, Zhang Q, Stein B, Blanar MA, et al. (1994) The bzip transactivator of Epstein-Barr virus, BZLF1, functionally and physically interacts with the p65 subunit of NF-kappa B. Molecular and Cellular Biology 14: Morrison TE, Kenney SC (24) BZLF1, an Epstein- Barr virus immediate-early protein, induces p65 nuclear translocation while inhibiting p65 transcriptional function. Virology 328: Chang LS, Wang JT, Doong SL, Lee CP, Chang CW, et al. (212) Epstein-Barr Virus BGLF4 Kinase Downregulates NF-kB Transactivation through Phosphorylation of Coactivator UXT. Journal of Virology 86: Saito S, Murata T, Kanda T, Isomura H, Narita Y, et al. (213) Epstein-Barr Virus Deubiquitinase Downregulates TRAF6-Mediated NF-kB Signaling during Productive Replication. Journal of Virology 87: Johannsen E, Luftig M, Chase MR, Weicksel S, Cahir- McFarland E, et al. (24) Proteins of purified Epstein-Barr virus. Proceedings of the National Academy of Sciences of the United States of America 11: Cohen JI, Lekstrom K (1999) Epstein-Barr Virus BARF1 Protein Is Dispensable for B-Cell Transformation and Inhibits Alpha Interferon Secretion from Mononuclear Cells. Journal of Virology 73: Strockbine LD, Cohen JI, Farrah T, Lyman SD, Wagener F, et al. (1998) The Epstein-Barr Virus BARF1 Gene Encodes a Novel, Soluble Colony-Stimulating Factor-1 Receptor. Journal of Virology 72: Ohashi M, Fogg MH, Orlova N, Quink C, Wang F (212) An Epstein-Barr Virus Encoded Inhibitor of Colony Stimulating Factor-1 Signaling Is an Important Determinant for Acute and Persistent EBV Infection. PLoS Pathog 8: e Bristol JA, Robinson AR, Barlow EA, Kenney SC (21) The Epstein-Barr Virus BZLF1 Protein Inhibits Tumor Necrosis Factor Receptor 1 Expression through Effects on Cellular C/EBP Proteins. Journal of Virology 84: General introduction 222. Morrison TE, Mauser A, Wong A, Ting JPY, Kenney SC (21) Inhibition of IFN-y Signaling by an Epstein-Barr Virus Immediate-Early Protein. Immunity 15: Morrison TE, Mauser A, Klingelhutz A, Kenney SC (24) Epstein-Barr Virus Immediate-Early Protein BZLF1 Inhibits Tumor Necrosis Factor Alpha-Induced Signaling and Apoptosis by Downregulating Tumor Necrosis Factor Receptor 1. Journal of Virology 78: Michaud F, Coulombe F, Gaudreault E, Paquet- Bouchard C, Rola-Pleszczynski M, et al. (21) Epstein- Barr Virus Interferes with the Amplification of IFNa Secretion by Activating Suppressor of Cytokine Signaling 3 in Primary Human Monocytes. PLoS ONE 5: e Cayrol C, Flemington EK (1995) Identification of cellular target genes of the Epstein-Barr virus transactivator Zta: activation of transforming growth factor beta igh3 (TGF-beta igh3) and TGF-beta 1. Journal of Virology 69: Adamson AL, Kenney S (21) Epstein-Barr Virus Immediate-Early Protein BZLF1 Is SUMO-1 Modified and Disrupts Promyelocytic Leukemia Bodies. Journal of Virology 75: Saffert RT, Kalejta RF (28) Promyelocytic leukemianuclear body proteins: herpesvirus enemies, accomplices, or both? Future Virology 3: Padgett DA, Hotchkiss AK, Pyter LM, Nelson RJ, Yang E, et al. (24) Epstein-Barr virus-encoded dutpase modulates immune function and induces sickness behavior in mice. J Med Virol 74: Waldman WJ, Williams J, Lemeshow S, Binkley P, Guttridge D, et al. (28) Epstein-Barr virus-encoded dutpase enhances proinflammatory cytokine production by macrophages in contact with endothelial cells: Evidence for depression-induced atherosclerotic risk. Brain, Behavior, and Immunity 22: Brooks DG, Trifilo MJ, Edelmann KH, Teyton L, McGavern DB, et al. (26) Interleukin-1 determines viral clearance or persistence in vivo. Nat Med 12: Glaser R, Litsky ML, Padgett DA, Baiocchi RA, Yang EV, et al. (26) EBV-encoded dutpase induces immune dysregulation: Implications for the pathophysiology of EBV-associated disease. Virology 346: Quinn LL, Zuo J, Abbott RJM, Shannon-Lowe C, Tierney RJ, et al. (214) Cooperation between Epstein-Barr Virus Immune Evasion Proteins Spreads Protection from CD8+ T Cell Recognition across All Three Phases of the 41 1

42 Chapter 1 Lytic Cycle. PLoS Pathog 1: e Ressing ME, Horst D, Griffin BD, Tellam J, Zuo J, et al. (28) Epstein-Barr virus evasion of CD8(+) and CD4(+) T cell immunity via concerted actions of multiple gene products. Semin Cancer Biol 18: Hislop AD, Ressing ME, van Leeuwen D, Pudney VA, Horst Dl, et al. (27) A CD8+ T cell immune evasion protein specific to Epstein-Barr virus and its close relatives in Old World primates. The Journal of Experimental Medicine 24: Horst D, van LD, Croft NP, Garstka MA, Hislop AD, et al. (29) Specific targeting of the EBV lytic phase protein BNLF2a to the transporter associated with antigen processing results in impairment of HLA class I-restricted antigen presentation. J Immunol 182: Zuo J, Currin A, Griffin BD, Shannon-Lowe C, Thomas WA, et al. (29) The Epstein-Barr Virus G-Protein- Coupled Receptor Contributes to Immune Evasion by Targeting MHC Class I Molecules for Degradation. PLoS Pathog 5: e Griffin BD, Gram AM, Mulder A, van Leeuwen D, Claas FHJ, et al. (213) EBV BILF1 Evolved To Downregulate Cell Surface Display of a Wide Range of HLA Class I Molecules through Their Cytoplasmic Tail. The Journal of Immunology 19: Ressing ME, van Leeuwen D., Verreck FA, Gomez R, Heemskerk B, et al. (23) Interference with T cell receptor-hla-dr interactions by Epstein-Barr virus gp42 results in reduced T helper cell recognition. Proc Natl Acad Sci U S A 1: Zuo J, Thomas WA, Haigh TA, Fitzsimmons L, Long HM, et al. (211) Epstein-Barr Virus Evades CD4+ T Cell Responses in Lytic Cycle through BZLF1-mediated Downregulation of CD74 and the Cooperation of vbcl-2. PLoS Pathog 7: e Salek-Ardakani S, Arrand JR, Mackett M (22) Epstein-Barr Virus Encoded Interleukin-1 Inhibits HLA-Class I, ICAM-1, and B7 Expression on Human Monocytes: Implications for Immune Evasion by EBV. Virology 34: Apcher S, Komarova A, Daskalogianni C, Yin Y, Malbert-Colas L, et al. (29) mrna Translation Regulation by the Gly-Ala Repeat of Epstein-Barr Virus Nuclear Antigen 1. Journal of Virology 83: Blake N, Lee S, Redchenko I, Thomas W, Steven N, et al. (1997) Human CD8+ T Cell Responses to EBV EBNA1: HLA Class I Presentation of the (Gly-Ala)-Containing Protein Requires Exogenous Processing. Immunity 7: Levitskaya J, Coram M, Levitsky V, Imreh S, Steigerwald-Mullen PM, et al. (1995) Inhibition of antigen processing by the internal repeat region of the Epstein- Barr virus nuclear antigen-1. Nature 375: Yin Y, Manoury B, Fahraeus R (23) Self-Inhibition of Synthesis and Antigen Presentation by Epstein-Barr Virus-Encoded EBNA1. Science 5: Hochberg D, Middeldorp JM, Catalina M, Sullivan JL, Luzuriaga K, et al. (24) Demonstration of the Burkitt s lymphoma Epstein-Barr virus phenotype in dividing latently infected memory cells in vivo. Proceedings of the National Academy of Sciences 11: Lee SP, Brooks JM, Al-Jarrah H, Thomas WA, Haigh TA, et al. (24) CD8 T Cell Recognition of Endogenously Expressed Epstein-Barr Virus Nuclear Antigen 1. The Journal of Experimental Medicine 199: Tellam J, Connolly G, Green KJ, Miles JJ, Moss DJ, et al. (24) Endogenous Presentation of CD8+ T Cell Epitopes from Epstein-Barr Virus-encoded Nuclear Antigen 1. The Journal of Experimental Medicine 199: Voo KS, Fu T, Wang HY, Tellam J, Heslop HE, et al. (24) Evidence for the Presentation of Major Histocompatibility Complex Class IGÇôrestricted Epstein-Barr Virus Nuclear Antigen 1 Peptides to CD8+ T Lymphocytes. The Journal of Experimental Medicine 199: Valentine R, Dawson C, Hu C, Shah K, Owen T, et al. (21) Epstein-Barr virus-encoded EBNA1 inhibits the canonical NF-kappaB pathway in carcinoma cells by inhibiting IKK phosphorylation. Molecular Cancer 9: Wood VHJ, O Neil JD, Wei W, Stewart SE, Dawson CW, et al. (27) Epstein-Barr virus-encoded EBNA1 regulates cellular gene transcription and modulates the STAT1 and TGF[beta] signaling pathways. Oncogene 26: Kanda K, Kempkes B, Bornkamm GW, Gabain Av, Decker T (1999) The Epstein-Barr Virus Nuclear Antigen 2 (EBNA2), a Protein Required for B Lymphocyte Immortalization, Induces the Synthesis of Type I Interferon in Burkitts Lymphoma Cell Lines. bchm 38: Aman P, von Gabain A (199) An Epstein-Barr virus immortalization associated gene segment interferes specifically with the IFN-induced anti-proliferative response in human B-lymphoid cell lines. The EMBO 42

43 Journal 9: Kanda K, Decker T, Aman P, Wahlstrom M, von Gabain A, et al. (1992) The EBNA2-related resistance towards alpha interferon (IFN-alpha) in Burkitt s lymphoma cells effects induction of IFN-induced genes but not the activation of transcription factor ISGF-3. Molecular and Cellular Biology 12: Muromoto R, Ikeda O, Okabe K, Togi S, Kamitani S, et al. (29) Epstein-Barr virus-derived EBNA2 regulates STAT3 activation. Biochemical and Biophysical Research Communications 378: Ho HH, Ivashkiv LB (26) Role of STAT3 in Type I Interferon Responses: Negative Regulation of STAT1- dependent Inflammatory Gene Activation. Journal of Biological Chemistry 281: Pagès F, Galon J, Karaschuk G, Dudziak D, Camus M, et al. (24) Epstein-Barr virus nuclear antigen 2 induces interleukin-18 receptor expression in B cells. Blood 15: Kieser A (27) Signal transduction by the Epstein- Barr virus oncogene latent membrane protein 1 (LMP1). Signal Transduction 7: Middeldorp JM, Pegtel DM (28) Multiple roles of LMP1 in Epstein-Barr virus induced immune escape. Seminars in Cancer Biology 18: General introduction 264. Ning S, Campos AD, Darnay BG, Bentz GL, Pagano JS (28) TRAF6 and the Three C-Terminal Lysine Sites on IRF7 Are Required for Its Ubiquitination-Mediated Activation by the Tumor Necrosis Factor Receptor Family Member Latent Membrane Protein 1. Molecular and Cellular Biology 28: Zhang L, Pagano JS (2) Interferon Regulatory Factor 7 Is Induced by Epstein-Barr Virus Latent Membrane Protein 1. Journal of Virology 74: Zhang L, Wu L, Hong K, Pagano JS (21) Intracellular Signaling Molecules Activated by Epstein-Barr Virus for Induction of Interferon Regulatory Factor 7. Journal of Virology 75: Zhao Y, Xu D, Jiang Y, Zhang L (21) Dual Functions of Interferon Regulatory Factors 7C in Epstein-Barr Virus- Mediated Transformation of Human B Lymphocytes. PLoS ONE 5: e Gires O, Kohlhuber F, Kilger E, Baumann M, Kieser A, et al. (1999) Latent membrane protein 1 of Epstein-Barr virus interacts with JAK3 and activates STAT proteins. The EMBO Journal 18: Zhang J, Das SC, Kotalik C, Pattnaik AK, Zhang L (24) The Latent Membrane Protein 1 of Epstein- Barr Virus Establishes an Antiviral State via Induction of Interferon-stimulated Genes. Journal of Biological Chemistry 279: Najjar I, Baran-Marszak F, Le Clorennec C, Laguillier C, Schischmanoff O, et al. (25) Latent Membrane Protein 1 Regulates STAT1 through NF-kB-Dependent Interferon Secretion in Epstein-Barr Virus-Immortalized B Cells. Journal of Virology 79: Richardson C, Fielding C, Rowe M, Brennan P (23) Epstein-Barr Virus Regulates STAT1 through Latent Membrane Protein 1. Journal of Virology 77: Xu Y, Rodriguez-Huete A, Pari GS (26) Evaluation of the Lytic Origins of Replication of Kaposi s Sarcoma- Associated Virus/Human Herpesvirus 8 in the Context of the Viral Genome. J Virol 8: Geiger TR, Martin JM (26) The Epstein-Barr Virus- Encoded LMP-1 Oncoprotein Negatively Affects Tyk2 Phosphorylation and Interferon Signaling in Human B Cells. Journal of Virology 8: Leslie E, Ning S, Kelliher M, Pagano JS (27) Interferon Regulatory Factor 7 Is Activated by a Viral Oncoprotein through RIP-Dependent Ubiquitination. Molecular and Cellular Biology 27: Fathallah I, Parroche P, Gruffat H, Zannetti C, Johansson H, et al. (21) EBV Latent Membrane Protein 1 Is a Negative Regulator of TLR9. The Journal of Immunology 185: Cahir-McFarland ED, Davidson DM, Schauer SL, Duong J, Kieff E (2) NF-kB inhibition causes spontaneous apoptosis in Epstein-Barr virus-transformed lymphoblastoid cells. Proceedings of the National Academy of Sciences 97: Izumi KM, Kieff ED (1997) The Epstein-Barr virus oncogene product latent membrane protein 1 engages the tumor necrosis factor receptor-associated death domain protein to mediate B lymphocyte growth transformation and activate NF-kB. Proceedings of the National Academy of Sciences 94: Stewart S, Dawson CW, Takada K, Curnow J, Moody CA, et al. (24) Epstein-Barr virus-encoded LMP2A regulates viral and cellular gene expression by modulation of the NF-kB transcription factor pathway. Proceedings of the National Academy of Sciences of the United States of America 11:

44 Chapter Swanson-Mungerson M, Bultema R, Longnecker R (21) Epstein-Barr virus LMP2A imposes sensitivity to apoptosis. Journal of General Virology 91: Shah KM, Stewart SE, Wei W, Woodman CBJ, O Neil JD, et al. (29) The EBV-encoded latent membrane proteins, LMP2A and LMP2B, limit the actions of interferon by targeting interferon receptors for degradation. Oncogene 28: Lerner MR, Andrews NC, Miller G, Steitz JA (1981) Two small RNAs encoded by Epstein-Barr virus and complexed with protein are precipitated by antibodies from patients with systemic lupus erythematosus. Proceedings of the National Academy of Sciences 78: Clarke PA, Schwemmle M, Schickinger J, Hilse K, Clemens MJ (1991) Binding of Epstein-Barr virus small RNA EBER-1 to the double-stranded RNA-activated protein kinase DAI. Nucleic Acids Research 19: Komano J, Maruo S, Kurozumi K, Oda T, Takada K (1999) Oncogenic role of Epstein-Barr virus-encoded RNAs in Burkitt s lymphoma cell line Akata. J Virol 73: Nanbo A, Inoue K, Adachi-Takasawa K, Takada K (22) Epstein-Barr virus RNA confers resistance to interferon-alpha-induced apoptosis in Burkitt s lymphoma. EMBO J 21: Nanbo A, Yoshiyama H, Takada K (25) Epstein- Barr virus-encoded poly(a)- RNA confers resistance to apoptosis mediated through Fas by blocking the PKR pathway in human epithelial intestine 47 cells. J Virol 79: Ruf IK, Lackey KA, Warudkar S, Sample JT (25) Protection from interferon-induced apoptosis by Epstein- Barr virus small RNAs is not mediated by inhibition of PKR. J Virol 79: Cai X, Schafer A, Lu S, Bilello JP, Desrosiers RC, et al. (26) Epstein-Barr virus micrornas are evolutionarily conserved and differentially expressed. PLoS Pathog 2: e Grundhoff A, Sullivan CS, Ganem D (26) A combined computational and microarray-based approach identifies novel micrornas encoded by human gammaherpesviruses. RNA 12: Zhu JY, Pfuhl T, Motsch N, Barth S, Nicholls J, et al. (29) Identification of novel Epstein-Barr virus microrna genes from nasopharyngeal carcinomas. J Virol 83: Meckes DG, Jr., Shair KH, Marquitz AR, Kung CP, Edwards RH, et al. (21) Human tumor virus utilizes exosomes for intercellular communication. Proc Natl Acad Sci U S A 17: Pegtel DM, Cosmopoulos K, Thorley-Lawson DA, van Eijndhoven MAJ, Hopmans ES, et al. (21) Functional delivery of viral mirnas via exosomes. Proc Natl Acad Sci U S A 17: Feederle R, Linnstaedt SD, Bannert H, Lips H, Bencun M, et al. (211) A viral microrna cluster strongly potentiates the transforming properties of a human herpesvirus. PLoS pathogens 7: e Seto E, Moosmann A, Gromminger S, Walz N, Grundhoff A, et al. (21) Micro RNAs of Epstein- Barr virus promote cell cycle progression and prevent apoptosis of primary human B cells. PLoS Pathog 6: e Vereide DT, Seto E, Chiu YF, Hayes M, Tagawa T, et al. (214) Epstein-Barr virus maintains lymphomas via its mirnas. Oncogene 33: Wahl A, Linnstaedt SD, Esoda C, Krisko JF, Martinez- Torres F, et al. (213) A Cluster of Virus-Encoded MicroRNAs Accelerates Acute Systemic Epstein-Barr Virus Infection but Does Not Significantly Enhance Virus- Induced Oncogenesis In Vivo. Journal of Virology 87: Xia T, O Hara A, Araujo I, Barreto J, Carvalho E, et al. (28) EBV MicroRNAs in Primary Lymphomas and Targeting of CXCL-11 by ebv-mir-bhrf1-3. Cancer Res 68: Nachmani D, Stern-Ginossar N, Sarid R, Mandelboim O (29) Diverse herpesvirus micrornas target the stressinduced immune ligand MICB to escape recognition by natural killer cells. Cell Host Microbe 5: Rathinam V, Fitzgerald K (21) Inflammasomes and Anti-Viral Immunity. J Clin Immunol 3: Haneklaus M, Gerlic M, Kurowska-Stolarska M, Rainey AA, Pich D, et al. (212) Cutting edge: mir-223 and EBV mir-bart15 regulate the NLRP3 inflammasome and IL-1beta production. J Immunol 189: Huang WT, Lin CW (214) EBV-encoded mir- BART2-5p and mir-bart8 inhibit the IFN-gamma- STAT1 pathway associated with disease progression in nasal NK-cell lymphoma. Am J Pathol 184: Lin TC, Liu TY, Hsu SM, Lin CW (213) Epstein-Barr 44

45 virus-encoded mir-bart2-5p inhibits T-bet translation with secondary suppression of p53 in invasive nasal NK/ T-cell lymphoma. Am J Pathol 182: Ambrosio MR, Navari M, Di LL, Leon EA, Onnis A, et al. (214) The Epstein Barr-encoded BART-6-3p microrna affects regulation of cell growth and immuno response in Burkitt lymphoma. Infect Agent Cancer 9: 12. General introduction Dolken L, Malterer G, Erhard F, Kothe S, Friedel CC, et al. (21) Systematic analysis of viral and cellular microrna targets in cells latently infected with human gamma-herpesviruses by RISC immunoprecipitation assay. Cell Host Microbe 7: Gottwein E, Corcoran DL, Mukherjee N, Skalsky RL, Hafner M, et al. (211) Viral microrna targetome of KSHV-infected primary effusion lymphoma cell lines. Cell Host Microbe 1: Kuzembayeva M, Chiu YF, Sugden B (212) Comparing proteomics and RISC immunoprecipitations to identify targets of Epstein-Barr viral mirnas. PLoS One 7: e Riley KJ, Rabinowitz GS, Yario TA, Luna JM, Darnell RB, et al. (212) EBV and human micrornas co-target oncogenic and apoptotic viral and human genes during latency. EMBO J 31: Skalsky RL, Corcoran DL, Gottwein E, Frank CL, Kang D, et al. (212) The viral and cellular microrna targetome in lymphoblastoid cell lines. PLoS Pathog 8: e

46

47 CHAPTER 2 Epstein-Barr virus lytic phase protein BGLF5 contributes to Toll-like receptor 9 downregulation during productive infection *Michiel van Gent *Bryan D. Griffin *Eufemia G. Berkhoff Daphne van Leeuwen Ingrid G. J. Boer Marlyse Buisson Franca C. Hartgers Wim P. Burmeister Emmanuel J. Wiertz Maaike E. Ressing * equal contributions Published in The Journal of Immunology (211), 186:

48 Chapter 2 Abstract Viruses employ a wide range of strategies to modulate the host immune response. The human γ-herpesvirus Epstein-Barr virus (EBV), causative agent of infectious mononucleosis and several malignant tumours, encodes proteins that subvert immune responses, notably those mediated by T cells. Less is known about EBV interference with innate immunity, more specifically at the level of Toll-like receptor (TLR)-mediated pathogen recognition. The viral double-stranded DNA sensor TLR9 is expressed on B cells, a natural target of EBV infection. Here, we show that EBV particles trigger innate immune signaling pathways through TLR9. Furthermore, using an in vitro system for productive EBV infection, it has now been possible to compare the expression of TLRs by EBV - and EBV + human B cells during the latent and lytic phases of infection. Several TLRs were found to be differentially expressed either in latently EBV-infected cells or after induction of the lytic cycle. In particular, TLR9 expression was profoundly decreased at both the RNA and protein levels during productive EBV infection. We identified the EBV lytic phase protein BGLF5 as a protein that contributes to down-regulating TLR9 levels through RNA degradation. Reducing the levels of a patternrecognition receptor capable of sensing the presence of EBV provides a mechanism by which the virus could obstruct host innate antiviral responses. 48

49 EBV lytic-phase protein BGLF5 reduces TLR9 expression Introduction EBV is a ubiquitous human herpesvirus that targets B lymphocytes and persistently infects more than 9% of adults worldwide (1). In the majority of cases, persistent infection remains subclinical. However, EBV can induce lymphoproliferative disorders in immunocompromised patients and is also strongly associated with several malignant tumours of lymphoid and epithelial origin, such as Burkitt s lymphoma (BL) and nasopharyngeal carcinoma (2). During childhood, primary infection is typically asymptomatic, but if delayed until adolescence it manifests as infectious mononucleosis in approximately 5% of cases (3). This syndrome arises due to vigorous, non-specific T cell activation producing a large number of atypical lymphocytes (2). A subsequent virus-specific T-cell response reduces the number of infected B cells, though fails to completely eradicate the virus. Instead, EBV persists for life in memory B cells, where up to 9 viral proteins are expressed in the latent stage of the viral life cycle. This reduction in the number of viral antigens presented to the immune system aids latently infected B cells in escaping destruction. Transmission to another host requires the production of new virions, which is facilitated by occasional viral reactivation and is characterized by the expression of up to 8 viral proteins during the lytic phase. During this replicative phase of infection, dedicated immune evasion molecules encoded within the viral genome are thought to extend the life span of virus-producing cells. Indeed, the past years have witnessed the identification of a number of EBV immunoevasins targeting adaptive immune responses, particularly the T cell-mediated response (4). A compelling paradox of primary EBV infection is the failure of the apparently robust T-cell response to result in viral clearance. High T cell numbers may suggest that prior innate immune responses were triggered by viral infection. However, were EBV to subvert the innate immune response upon primary infection, this could result in a poorly tailored adaptive immune response, possessing insufficient antiviral specificity in its early stages, and thereby allowing a number of infected B cells to escape elimination and establish a persistent infection. To examine this hypothesis, we set out to investigate whether EBV can modulate innate immune responses. Successful elimination of most pathogens requires crosstalk between the innate and adaptive arms of immunity. TLRs occupy a pivotal position in this regard, detecting the presence of microbial pathogens through recognition of pathogen-associated molecular patterns (5). TLRs are responsible for host detection of several herpesviruses (6). For instance, TLR2 senses human cytomegalovirus (7), HSV-1 (8), as well as EBV (9,1). TLR3 detects viral dsrna, including the non-coding EBV EBER molecules (11), and also dsrna intermediates produced during viral replication, as seen with Kaposi s sarcoma-associated herpesvirus (12). Additionally, TLR7 and 9, expressed to high levels by plasmacytoid dendritic cells (pdc), appear to play a role in the recognition of EBV (13-15). TLR9 is activated by DNA sequences containing 2 49

50 Chapter 2 unmethylated deoxy-cytidylate-phosphate-deoxyguanylate (CpG) dinucleotides flanked by two 5 purines and two 3 pyrimidines, which are abundant in certain dsdna viral genomes, including those of herpesviruses (16,17). Indeed, TLR9 also recognizes HSV-1 (18) and -2 (19), and mouse cytomegalovirus (2). TLR triggering upon ligand recognition activates intracellular signaling networks that ultimately lead to the production of a wide range of immunoregulatory molecules, some of which possess direct antiviral activity, while others serve to orchestrate the adaptive immune response (21-23). This is exemplified by TLR9 signaling in B cells: in response to CpG, purified human B cells display up-regulation of the T cell co-stimulatory molecules CD8 and CD86, as well as HLA class II molecules (16,24). TLR9 signaling in B cells also directly induces CXCR3 chemokines capable of attracting and activating Th1 cells (25). Furthermore, while pdcs are considered more important in this regard, some studies provide evidence of a role for TLR9 activation in B cells in production of IL-12, tilting the T-cell response towards Th1 polarity (26). In the present study, we examined whether TLRs are differentially expressed by EBV - and EBV + B cells during the latent and lytic phases of infection. The fact that reactivation occurs in only a small percentage of EBV-infected cells (1-5%) has limited studies on productive EBV infection so far. To overcome this, we used a strategy to select populations of productively EBV-infected BL cells based on the inducible expression of a lytic phase reporter following viral reactivation (27). To our knowledge, this is the only system that allows isolation of lytically EBV-infected B cells (AKBMs) to high purity for subsequent studies. Previously, this system proved essential in the identification of EBV lytic phase genes interfering with the machinery of adaptive immune responses (27,28). Importantly, phenotypes of e.g. HLA class I down-regulation observed at the surface of lytically induced AKBM cells, recapitulated those reported for the small fraction of EBV + B-LCL that spontaneously reactivate (29). This demonstrates that the AKBM system, while unique in its ability to allow isolation of a pure population of lytically infected cells, remains representative of productive EBV infection of B cells. Here, we use this system to explore the effects of EBV infection on innate immunity, and specifically on TLRs. Results Differential expression of TLRs by B cells during the latent and lytic phase of EBV infection To examine whether EBV affects TLR expression in B cells, we initially compared the TLR expression profile of primary B cells to that of two EBV - BL cell lines, BJAB and AK31. AK31 is the EBV - counterpart of the AKBM cell line used for EBV reactivation studies. RT-PCR was used to monitor expression of TLR 1-1 mrnas (Figure 1A), with 293 cells transfected to express individual TLRs serving as positive controls for TLR detection. Primary B cells, BJAB, 5

51 EBV lytic-phase protein BGLF5 reduces TLR9 expression and AK31 displayed common expression of TLRs 1, 6, 9, and 1. Furthermore, low levels of TLR7 were observed in both BL cell lines, while this signal was very weak in primary B cells. Expression of TLRs 2, 5, and 8, was not detectable in any of the B cell types tested, although low levels of TLR2 have been reported to be expressed on naïve and memory B cells (34). Primary B cells and the two EBV - BL lines differed only in levels of TLR3 (only present in BJAB) and TLR4 (hardly detectable in AK31). Thus, the TLR expression profile of primary B cells is largely retained in the EBV - BL cells, setting the stage for studies on the effect of EBV on expression of TLR1, 6, 7, 9, and 1. We next assessed whether TLR expression in B cells is altered by EBV infection. To this end, we compared the Akata-derived EBV - AK31 to EBV + AKBM cells that display a latency I phenotype and to AKBM cells undergoing lytic EBV replication. AKBM cells express a rat CD2-GFP reporter only after induction of the lytic phase by cross-linking of surface IgG (27). Cells receiving anti-igg treatment for 16 h were immunomagnetically sorted for surface rat CD2 expression to high purity. As a control, AK31 cells constitutively expressing rat CD2-GFP were treated in an identical manner to exclude any secondary effects of anti-igg antibody treatment or immunomagnetic sorting. RT-PCR was then conducted to investigate the effects of latent and lytic EBV infection on a range of TLRs expressed by Akata-derived cells, reflecting those detected in primary B cells. We found that latent infection leads to reduced expression of TLRs 1 and 1 (Figure 1B, compare lanes 1 and 3), while levels of TLRs 6, 7, and 9 were not down-regulated by 2 A B Cells AK31 AKBM TLR ND Positive control (293-TLR) αigg TLR1 Primary B cells TLR6 BJAB TLR7 AK31 TLR9 TLR1 18S BGLF5 Lane Figure 1. TLR mrnas are differentially expressed during latent and lytic phases of EBV infection. (a) cdna was prepared from human primary B cells and EBV - BJAB and AK31 BL cells. Expression levels of TLR1-1 were determined by PCR analysis. cdna from 293 cells stably expressing individual TLRs was used as positive control for specific TLR detection. ND: not determined. (b) EBV - AK31 and EBV + AKBM cells were incubated with (+) or without (-) anti-igg for 16 h. Cells treated with anti-igg were immunomagnetically sorted for surface rat CD2 expression to over 95% purity. cdna was generated and expression levels of the indicated gene products were determined by PCR analysis. 51

52 Chapter 2 latent EBV gene expression. This phenotype was reproduced in another pair of EBV - BL cells (Mutu I, see Supplementary Figure 1). When replicating EBV, AKBM cells displayed reduced mrna expression of TLRs 1, 6, 7, 9, and 1, with abrogation of TLR9 expression being most pronounced (Figure 1B, compare lanes 3 and 4). The detection of the EBV lytic cycle transcript BGLF5 confirmed induction of the productive phase of the viral life cycle in AKBM cells treated with anti-human IgG. As a control indicating that the differences in signal were not the result of differences in RNA extraction efficiencies, the level of 18S RNA was similar for all cells tested. Thus, these data show that EBV modulates TLR mrna levels during both the latent and lytic stages of infection, with TLR9 expression during productive infection undergoing a particularly strong inhibition. TLR9 expression during EBV lytic phase is strongly reduced at both the RNA and protein level To determine the extent of the observed striking reduction in TLR9 expression in the lytic phase, we performed real-time PCR analysis on cdna generated from EBV - AK31 as well as latent and lytic EBV + AKBM cells (Figure 2A). TLR9 mrna levels, normalized to 18S RNA levels, were reduced by 86% in AKBM cells during EBV lytic phase compared to latently infected or uninfected cells; anti-igg treatment of AK31 cells had no such effect. These realtime PCR results therefore confirm and extend our RT-PCR data (Figure 1). To examine whether TLR9 mrna levels correlated with TLR9 protein expression, Western Blot analyses on cell lysates of EBV - AK31 and EBV + AKBM cells were performed. TfR was used as a control and its expression was not affected by cross-linking of surface IgG in either AK31 or AKBM cells (Figure 2B). However, while TLR9 was detected in cell lysates of EBV - AK31 cells and EBV + AKBM cells during EBV latent phase, it was virtually absent from lysates of AKBM cells during EBV lytic phase (Figure 2B, compare lanes 1-3 to 4). Since the EBV - AK31 cells were examined under the same conditions as the EBV + AKBM cells, the reduction of TLR9 expression during EBV lytic cycle is a specific feature of the virus, exhibited during the productive phase of infection. Such a dramatic decrease in the steady-state levels of a cellular protein have not previously been observed after induction of the lytic cycle for 16-2 h. Notably, despite displaying reduced mrna expression, steady-state levels of HLA class I heavy chain, HLA class II α and β chains, TAP1 and TAP2 all remain unaffected at this time (27,28). To probe whether the dramatic reduction in TLR9 protein relative to other cellular proteins could be explained by a shorter half-life, we performed pulse-chase analyses to compare the relative stabilities of TLR9 and TfR protein. 293 cells stably expressing TLR9 were metabolically labelled for 3 min before immunoprecipitating TLR9 and TfR at the indicated chase time points. Both TLR9 and TfR were detected to similar levels up to at least 48 min after pulselabeling (Figure 2C). Quantification of protein levels revealed almost identical stabilities for 52

53 EBV lytic-phase protein BGLF5 reduces TLR9 expression TLR9 and TfR proteins (Figure 2D). Thus, despite displaying equivalent stability, steady-state levels of TLR9 protein are dramatically reduced after one day of productive EBV infection, while TfR levels remain largely unaltered. A Fold change in TLR9 mrna B Cells TLR AK31 AKBM AK31 AKBM α-igg α-igg Cells C D Chase (min) TfR Lane Chase (min) Figure 2. TLR9 expression is reduced during EBV lytic phase both at the mrna and protein level. (a) EBV + AKBM cells were cultured for 16 h with or without anti-igg and subjected to immunomagnetic sorting to isolate productively infected cells. To exclude any secondary effects of antibody treatment and/or immunomagnetic sorting, EBV- AK31 cells were treated in the same manner. Levels of TLR9 mrna were determined by quantitative real-time PCR analysis and normalized to 18S RNA expression. Data are expressed relative to TLR9 mrna levels in non-igg-treated cells. Standard deviation of Ct values in 18S and TLR9 amplification replicates was less than 2% in each case. (b) Protein expression of TLR9 and TfR in post-nuclear lysates of the cells described in A was determined by Western Blot analysis. (c) 293-TLR9 cells were labeled for 3 min with [ 35 S]Met/Cys and chased for the times indicated. TLR9 and TfR were immunoprecipitated from cell lysates and all samples were analyzed by SDS-PAGE. (d) The relative stabilities of TLR9 and TfR proteins were evaluated by quantification of the experiment depicted in C. The small shift in mobility observed for TfR is explained by maturation of its N-linked glycans over time. % Intensity (relative to t=) TLR9 TfR TLR9 1 TLR9 TfR 8 2 TLR9-mediated intracellular signaling is activated by EBV In order to explore the functional relevance of the strong reduction in TLR9 protein seen in lytically infected cells, we examined whether TLR9 can recognize EBV. BJAB B cells were monitored by flow cytometry for surface expression of the activation marker CD54 (ICAM- 1) upon overnight stimulation with increasing concentrations of EBV particles or the known TLR9 activating ligand CpG. Virus particles caused a dose-dependent increase in CD54 expression, with the highest concentration used providing a similar induction to that observed with CpG (Figure 3A), suggestive of TLR activation by EBV. As BJAB cells express a range of TLRs other than TLR9 (Figure 1A), we determined whether TLR9 in particular can detect EBV. 293 cells stably expressing TLR9 (293-TLR9) and control 293T cells were transfected with an NF-κB-inducible firefly luciferase reporter gene. Stimulation 53

54 Chapter 2 with TNFα verified the responsiveness of the reporter system to NF-κB activation in both cell lines (not shown). Addition of CpG to 293-TLR9 cells confirmed the functionality of the TLR9 receptor (Figure 3B). Interestingly, addition of EBV particles to the TLR9-expressing cells activated NF-κB in a dose-dependent manner. This effect required TLR9, since the highest concentration of EBV particles used failed to trigger signaling through NF-κB in control 293T cells. The absence of detectable viral latent or lytic gene expression in 293T cells following culture in the presence of EBV (not shown), renders involvement of infection unlikely. These combined data demonstrate that TLR9 can sense the presence of EBV, highlighting the significance of TLR9 down-regulation during the EBV lytic cycle. The EBV lytic phase protein BGLF5 reduces TLR9 expression through mrna degradation To identify the EBV gene responsible for TLR9 down-regulation, we first determined the stage of the lytic phase at which TLR9 protein expression is decreased. PAA inhibits viral DNA replication and late lytic phase gene expression and was used here to discriminate early from late EBV protein expression (28). Irrespective of the PAA-mediated block of late lytic phase gene expression, TLR9 protein levels were reduced in EBV + AKBM cells during the lytic phase when compared to anti-igg-treated EBV - AK31 cells (Supplementary Figure 2). As a control, no reduction in β-actin levels was observed under the same conditions. Thus, an EBV gene product expressed (immediate) early during EBV lytic cycle causes the reduction in TLR9 expression. A Counts Surface CD54 (ICAM-1) on BJAB cells B Fold induction of NF-kB activity unstimulated CpG 5. vp/ml 25. vp/ml vp/ml vp/ml vp/ml - EBV - EBV CpG 293T 293-TLR9 Stimulus Cells Figure 3. EBV particles activate TLR9 signaling. (a) BJAB B cells were stimulated for 18 h with or without CpG-ODN26 (1 μg/ml) or increasing amounts of purified EBV particles, as indicated. Surface expression of CD54 (ICAM-1) was determined by flow cytometry. The dashed line represents unstained control cells. (b) 293T and 293-TLR9-CFP cells, transiently co-transfected with NF-κB-luciferase and phrl-tk (constitutively expressing Renilla luciferase) were stimulated for 6 h with or without CpG-ODN26 (1 μg/ml) or increasing amounts of purified EBV particles (see Materials and Methods). Cell lysates were assayed for firefly and Renilla (for normalizing transfection efficiency) luciferase. Results (mean±sd) are presented relative to cells treated with medium alone (-). 54

55 EBV lytic-phase protein BGLF5 reduces TLR9 expression Previously, we described a block in de novo protein synthesis occurring during productive EBV infection, referred to as host shutoff (28). The EBV protein responsible for virusinduced host shutoff was identified as the early lytic phase protein BGLF5, the viral alkaline exonuclease that also causes enhanced mrna turnover. Recently we have shown that recombinant BGLF5 acts not only as a DNase in vitro, but also as an Mn 2+ -dependent RNase, mediating the degradation of GFP mrna (33). Here, we addressed the question of whether BGLF5 could degrade TLR9 mrna. We observed that in vitro transcribed TLR9 mrna was A - + BGLF5 B 1 2 DNA RNA Mn 2+ EDTA Mn 2+ EDTA % TLR9 mrna remaining upon addition of BGLF Lane Mn 2+ EDTA Buffer C vector only BGLF5 (2x) D TLR9 transfection BGLF5 BNLF2a Counts Counts TLR9 BGLF5 BGLF5 (5x) BGLF5 (3x) TFR Counts Counts E Counts BGLF5 (1x) Counts BNLF2a GFP - BGLF5 BNLF2a Figure 4. EBV lytic phase protein BGLF5 degrades TLR9 mrna and reduces TLR9 protein expression. (a) Following in vitro transcription, TLR9 mrna was incubated with or without recombinant BGLF5 protein for 3 min at 37 C in the presence of either Mn 2+ or EDTA. TLR9 mrna and linearized plasmid DNA were visualized with ethidium bromide in a 1.5% agarose gel. (b) RNA bands were quantified using Quantity One software; the bars depict the percentage TLR9 mrna that remains after incubation with BGLF5, with TLR9 levels in buffer alone set at 1%. (c) 293T cells were transiently co-transfected with puno-tlr9-ha and pcdna3-ires-gfp, pcdna3- BNLF2a-IRES-GFP, or increasing amounts of pcdna3-bglf5-ires-gfp. 4 h post transfection, GFP expression was measured by flow cytometry. (d) Expression of TLR9-HA, BGLF5-HA, and TfR was examined by Western Blot analysis using post-nuclear lysates of the cells described in C. (e) The results of the TLR9 immunoblot depicted in D were quantitated using Quantity One software and normalized to TfR levels. The bars represent the percentage of TLR9 expression upon cotransfection with BGLF5 or control BNLF2a, with the level of TLR9 alone set at 1%. Intensity of TLR9 staining (% of control)

56 Chapter 2 strongly reduced upon addition of recombinant BGLF5, provided Mn 2+ ions were present (Figure 4A, compare lanes 1 and 3 ). No degradation was seen in the presence of EDTA (Figure 4A, compare lanes 2 and 4). As BGLF5 DNase activity is also dependent on the presence of bivalent ions, remaining plasmid DNA was also degraded by recombinant BGLF5 in Mn 2+ - containing buffer. Quantification of RNA levels demonstrated that EBV BGLF5 degraded TLR9 mrna in vitro (Figure 4B). To examine whether BGLF5 expressed in cells can reduce levels of TLR9, 293T cells were transiently transfected with TLR9 in combination with increasing amounts of BGLF5, or another EBV lytic phase protein BNLF2a (35), used as a control. Both BGLF5 and BNLF2a were expressed from a bicistronic vector that also encodes GFP. At 4 h post-transfection, cellular GFP expression was measured by flow cytometry as an indication of transfection efficiency. On average, cells co-transfected with the controls, vector alone or BNLF2a, were over 95% GFP + (Figure 4C). In the case of BGLF5 co-transfections, we have previously seen that GFP is susceptible to host shutoff. Thus, the decreasing GFP levels observed with increasing doses of BGLF5 are indicative of BGLF5-mediated enhanced turnover of GFP mrna. TLR9 protein was detected by Western blot analysis of post-nuclear cell lysates (Figure 4D). Co-expression of BGLF5 led to decreasing levels of TLR9 protein in a dose-dependent manner. TfR, used as a control protein, was found to be expressed to similar extents in all samples. Quantification of the Western blot data indicated a reduction in TLR9 protein levels of up to 5% with the highest concentration of BGLF5 co-expressed, when normalized for TfR protein levels (Figure 4E). The specificity of BGLF5-mediated down-regulation of TLR9 was demonstrated by the failure of BNLF2a to cause a reduction in TLR9 expression. Thus, BGLF5 can degrade TLR9 mrna and, thereby, decrease levels of TLR9 protein, providing compelling evidence that this EBV lytic cycle product contributes to down-regulation of TLR9 expression observed during productive viral infection. Discussion Although our understanding of evasion from adaptive immune responses by EBV has increased greatly over recent years (4), less is known about the interactions of the virus with innate immunity. Here, we examined the interplay between EBV and TLRs. We show that expression of TLRs 1, 6, 7, 9, and 1 is down-regulated in human B cells undergoing productive EBV infection. In particular, TLR9 protein levels were dramatically decreased. We demonstrate that TLR9 signaling is activated in the presence of EBV particles, revealing the reduction of TLR9 levels to be a potentially useful immune evasion strategy employed by the virus. We established that the EBV early protein BGLF5 degrades TLR9 mrna in vitro, providing a mechanism for its contribution to TLR9 down-regulation. To our knowledge, this is the first demonstration of an EBV lytic cycle gene product directly affecting expression levels 56

57 EBV lytic-phase protein BGLF5 reduces TLR9 expression of a pattern-recognition receptor capable of alerting the immune system to viral infection. While the expression of several TLRs was altered during productive EBV infection, the striking reduction in TLR9 at both RNA and protein levels was of particular interest since the motif recognized by this receptor is widely present in various herpesvirus genomes, including EBV. This would supply an abundance of potential TLR9-activating ligands both during primary infection and virus production. In fact, a number of recent publications indicate that TLR9 is involved in detecting EBV, leading to production of IL-8 in primary monocytes and IFN-α in pdcs (13,14). As for B cells, we here show that stimulation of BJAB cells with EBV increases surface CD54 expression. Likewise, Iskra et al. found upregulation of CD8 on primary B cells following co-culture with EBV (36). While EBV recognition by monocytes and pdcs appears to depend on TLR9 acting in tandem with TLRs 2 and 7, this may be different for B cells that do not or only weakly express TLRs 2 and 7. In line with this, we now demonstrate that stimulation with EBV particles leads to NF-κB activation in 293 cells only when TLR9 is expressed, but not in its absence, thereby establishing that TLR9 in isolation can specifically recognize EBV. Most likely, this TLR9-mediated recognition of EBV does not require prior cellular infection as we did not detect signs of infection of 293T cells (not shown). Indeed, this is in agreement with recent findings showing that pdcs can recognize UV-irradiated EBV particles in a TLR9-dependent fashion (13). As opposed to infection not being required for TLR9 recognition of EBV, the presence of viral DNA as well as endosomal maturation are necessary (13). In latently infected EBV + BL cells, the expression levels of several TLRs were also found to be altered. Observing decreased expression of TLRs 1 and 1 in both the Akata and Mutu I model systems of latency suggests that EBV latency-associated gene products may modulate TLR expression. A recent report indicated decreased TLR9 and increased TLR7 expression within hours after primary EBV infection of naïve B cells that had most likely entered latency III (15). We did not detect decreased TLR9 mrna expression in either latently infected Akata or Mutu I cells, and observed only a modest increase in TLR7 in latently infected Akata cells. One explanation for this discrepancy could be the different stages of latency examined, with both AKBM and Mutu I cells existing in latency I. Another possible reason for the decrease in TLR9 mrna seen by Martin et al. could be due to the actions of EBV lytic cycle gene products that undergo transient expression upon primary infection. An example of this latter phenomenon is provided by the immediate but transient expression of the viral bcl-2 homologue genes, BALF1 and BHRF1, upon infection of primary B cells by EBV (37). Thus, it is tempting to speculate that transient expression of BGLF5 upon primary infection could mediate downregulation of TLR9. In our search for the EBV modulator of TLR9 expression during productive infection, we focused on (immediate) early lytic phase genes, as inhibition of late viral gene expression failed to block the down-regulation of TLR9 observed during productive EBV infection. An 2 57

58 Chapter 2 appropriate candidate was BGLF5, which appears 3-6 h after lytic cycle induction. Here, we find that expression of BGLF5 in isolation is sufficient to cause an approximately 5% decrease in TLR9 steady-state protein levels in transfected cells. The extent of TLR9 down-regulation achieved in 293T cells upon expression of BGLF5 (Figure 4 C and D) is less than that observed upon induction of the viral lytic cycle in AKBM cells (Figure 2B). However, it should be noted that upon co-transfection of 293T cells, levels of BGLF5 protein approximated those observed during lytic infection of B cells, while expression of TLR9 far exceeded that present in latently-infected AKBM cells (not shown). Thus, higher relative levels of TLR9 expressed in transfected 293T cells may have resulted in a less robust BGLF5-mediated down-regulation. Additionally, other EBV lytic cycle gene products may to contribute to the stronger decrease of TLR9 seen in productively infected cells (further discussed below). Finally, in addition to identifying BGLF5 as an EBV protein contributing to decreased levels of TLR9 during productive infection, we offer a mechanistic explanation for this effect by demonstrating the ability of BGLF5 to degrade TLR9 mrna in vitro. During productive EBV infection, the decrease in TLR9 mrna levels is more pronounced than the down-regulation of other cellular components tested, including β 2 -microglobulin transcripts, suggesting that a further mechanism operates to diminish TLR9 expression at this stage in the viral life cycle. One possibility involves down-regulation of TLR9 transcription, potentially through activation of NF-κB (38) induced by viral gene products. More remarkable than the decrease in TLR9 mrna levels is the strong down-regulation of steady-state TLR9 protein levels seen at 16 h post-induction of the lytic cycle in AKBM cells. At this time, while mrna expression and surface protein levels of, for instance, HLA molecules are both diminished, no reduction in total protein levels is detected by Western blot analysis. Therefore, the dramatic reduction in total TLR9 protein levels in lytically infected cells, combined with metabolic labeling experiments showing TLR9 protein to have stability similar to that of the TfR, raise the prospect that EBV also targets TLR9 post-translationally. The existence of viral strategies other than host shutoff targeting TLR9 would provide an intriguing parallel to EBV-mediated subversion of adaptive antigen presentation pathways: BNLF2a and BILF1 combine with BGLF5 to sabotage the HLA class I antigen presentation (35,39), while gp42 complements the effect of BGLF5 on HLA class II-restricted T cell recognition (28,4). A fundamental role for TLRs in controlling viral disease is supported by the correlation between severe disease upon viral infection and loss-of-function mutations in TLRs themselves (41). Additionally, HSV encephalitis can arise from a deficiency in UNC-93B, a protein required for translocation of the nucleotide-sensing TLRs (TLRs 3, 7/8 and 9) to endolysosomes, which is critical for signaling (42-45). Interestingly, our preliminary data indicate that expression of UNC-93B transcripts is also suppressed during productive EBV infection. The recent identification of the first virus-encoded proteins that subvert TLR signaling further emphasizes the importance of TLRs in the antiviral response. Among these, vaccinia virus 58

59 EBV lytic-phase protein BGLF5 reduces TLR9 expression A46R can interact with the Toll-like IL-1 receptor (TIR)-domain-containing adaptors that associate with TLRs, thus reducing downstream Interferon-regulatory factor (IRF) and NFκB activation (46). The hepatitis C virus protease NS3-4A cleaves the TIR-domain containing adaptor inducing IFN-β (TRIF), thereby inhibiting the antiviral effects of TLR3 signaling (47). Direct targeting of TLRs has also been reported, with E6 and E7 of human papillomavirus causing diminished TLR9 levels through negatively regulating promoter activity, thereby functionally impairing TLR9 signaling (48). For EBV, some mechanisms by which viral gene products can thwart components of the innate immune response have emerged in recent years. For instance, the non-polyadenylated, untranslated RNAs, EBER-1 and -2, expressed during latency, inhibit the activation of the dsrna-stimulated protein kinase R, thereby providing resistance to IFN-induced apoptosis (49). Various immunoevasive capabilities have also been assigned to the immediate-early transactivator BZLF1, including association with the NF-κB subunit p65, resulting in its nuclear location while impairing its transcriptional ability (5), inhibition of IRF7, thereby hampering type I IFN production during viral reactivation (51), and down-regulation of the TNF receptor gene promoter activity compromising the effects of TNFα on an infected cell (52). More recently, the EBV tegument protein LF2 was shown to block dimerization of IRF7, while, the virion-associated kinase BGLF4 was found to curtail IRF3 transactivation ability (53,54). These latter examples indicate that EBV virus particles can begin countering the host innate immune response upon cellular entry without the need for prior gene expression. Here, we demonstrate the ability of the EBV early lytic cycle gene product BGLF5, expressed following viral reactivation in B cells, to reduce expression of TLR9, capable of recognizing pathogen-associated molecular patterns present in the EBV genome. The biological relevance of this effect is underscored by the demonstration that recognition of EBV particles by TLR9 leads to activation of the transcription factor NF-κB. This constitutes a novel manner in which EBV can thwart the host immune response, thus extending the time window for the virus to establish persistent infection and to produce new virions. 2 Acknowledgements We are grateful to Drs Paola Massari (Boston University School of Medicine, USA), Kees Melief (Dept. Immunohematology and Blood Transfusion, Leiden University Medical Center, The Netherlands), Sjoerd van der Burg (Dept. Clinical Oncology, Leiden University Medical Center, The Netherlands), and Debbie van Baarle (Dept. Immunology, University Medical Center Utrecht, The Netherlands) for helpful discussions and for generously sharing reagents. Furthermore, we thank Femke Stalpers (Dept. Medical Microbiology, University Medical Center Utrecht, The Netherlands) and Adriaan de Wilde (Dept. Medical Microbiology, Leiden University Medical Center, The Netherlands) for technical assistance and Dr Carla de Haas 59

60 Chapter 2 and Daniëlle Horst (Dept. Medical Microbiology, University Medical Center Utrecht, The Netherlands) for critically reading the manuscript. Materials and Methods Cells Human embryonic kidney 293 cells constitutively expressing functional human TLR genes were purchased from Invivogen (San Diego, CA), with the exception of 293 cells stably transfected with human TLR9 (3) (a gift from G.B. Lipford, Coley Pharmaceutical Group, Wellesley, USA) and 293 cells expressing a TLR9-CFP fusion protein (31) (kindly provided by P. Massari, Boston University School of Medicine, Dept. of Infectious Diseases, Boston MA, USA). 293T cells were maintained in DMEM (Invitrogen, Carlsbad, CA) supplemented with 1% FBS (EU approved, Invitrogen), 2 mm L-glutamine, 1 U/mL penicillin, and 1 µg/ml streptomycin. Cell lines stably expressing TLRs were cultured in selective antibiotics as recommended. The B cell lines BJAB, AKBM, AK31, 2A8, and Mutu I clone 3 and clone 9 (kindly provided by J. Sixbey, St. Jude Children s Research Hospital, Depts. of Infectious Diseases and Virology & Molecular Biology, Memphis, TN) (32), were maintained in RPMI 164 medium (Invitrogen) supplemented with 1% FBS (Standard Quality, EU approved, PAA Laboratories, GmbH, Pasching, Austria), 2 mm L-glutamine, 1 U/mL penicillin, and 1 µg/ml streptomycin. In addition, AK31 and AKBM cells were cultured in the presence of 1 mg/ml geneticin or.5 mg/ ml hygromycin B, respectively. Primary human B cells were prepared from peripheral blood donated by adult volunteers. Mononuclear cells were enriched by centrifugation on Ficoll-Paque (GE Healthcare, Waukesha, WI). B cells were positively selected from PBMCs by immunomagnetic labeling using CD19 Microbeads and LS columns (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer s protocols. The purity of the population was assessed by monitoring CD19 expression. Antibodies The following specific mouse monoclonal antibodies were used: anti-tlr9 (Imgenex IMG-35A, clone 26C593.2), antihuman transferrin receptor (TfR) H68.4 (Roche Diagnostics, Indianapolis, IN) for Western Blot analysis and anti-tfr CD71 (BD Pharmingen, San Diego, CA) for flow cytometry; 72A1 (ATCC, Middlesex, United Kingdom), recognizing the BLLF1-encoded gp35/22 late membrane antigen; OX34 specific for the extracellular domain of rat CD2 (kindly provided by M. Rowe, University of Birmingham Medical School, Birmingham, United Kingdom); and a phycoerythrinconjugated antibody directed against human CD54/ICAM-1 (#555511, BD Pharmingen, San Diego, CA). The rat monoclonal antibody 3F1 is directed against the influenza virus-derived HA-tag (Roche Diagnostics). Horseradish peroxidase (HRP)-conjugated secondary anti-species antibodies (Dako, Glostrup, Denmark), and secondary goat antimouse allophycocyanin-conjugated antibody (Leinco Technologies, St. Louis, MO) were used for Western Blot analysis and flow cytometry, respectively. Induction of the EBV lytic cycle To study the latent and lytic phase of EBV infection, we used EBV + B cells that express a rat CD2-GFP reporter protein during lytic phase, designated AKBM cells. The derivation of the AKBM cells and the isolation of populations in EBV lytic phase have been described previously (27). In brief, the human EBV + BL cell line Akata was stably transfected with a reporter plasmid (phebo-prbmrf1-ratcd2-gfp). The EBV lytic phase can be induced by cross-linking the surface IgG of the AKBM cells with 5 μg/ml goat F(ab )2 fragments to human IgG (Cappel, MP Biomedicals, Solon, OH). To isolate EBV + B cells in lytic phase, cells were stained with rat CD2-specific antibody (OX34) and were positively selected by magnetic cell sorting with anti-mouse IgG2a/b microbeads and MS columns (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the instructions of the manufacturer. Sorted populations of greater than 9% purity were obtained in this manner. Discrimination between (immediate) early and late stages of lytic phase was achieved by inhibition of viral DNA replication and late lytic phase gene expression, using phosphonoacetic acid (PAA). PAA (ph 7.4 in 1 mm Hepes) was added 1 h prior to induction of the EBV lytic phase at a final concentration of 3 μg/ml. To exclude any secondary effects of PAA, antibody treatment used for induction of EBV lytic phase or magnetic cell sorting, we used a subclone of Akata that has lost its EBV genome (AK31 cells) in parallel. AK31 cells constitutively express pegfpn1-rat CD2 in part of the population. 6

61 EBV lytic-phase protein BGLF5 reduces TLR9 expression RNA isolation and PCR Total RNA was extracted from 1-3 x 1 6 cells using Trizol reagent (Invitrogen Life Science, Carlsbad, CA) or RNeasy Plus Mini kit (Qiagen Benelux B.V., The Netherlands) and treated with DNase (TURBO DNAse-free kit, Applied Biosystems) according to manufacturer s protocols. 1 μg RNA was used for cdna transcription using random hexamers and the Moloney murine leukemia virus reverse transcriptase (RT; Finnzymes, Espoo, Finland). To confirm the absence of genomic DNA, the reactions were also performed without RT. The resulting cdna was used diluted 1:1 with water, and 1-2 μl was used for amplification with Taq or Pfu DNA Polymerase as indicated in Table I. TLR-specific primer sequences and the corresponding number of amplification cycles for semi-quantitative PCR analysis are listed in Table I. DNA fragments of expected length were visualized by 1% agarose gel electrophoresis and ethidium bromide staining. Quantitative PCR reactions were performed in duplicate using a final concentration of 2 nm probe and 3 nm primers (18S rrna), or a TaqMan Gene Expression Assay (TLR9) according to the instructions of the manufacturer (Applied Biosystems, Foster City, CA). The amplifications were performed on an ABI/PRISM 75 sequence detector system (Applied Biosystems). 18S rrna was amplified using the forward primer 5 -AGTCCCTGCCCTTTGTACACA-3, the reverse primer 5 -GATCCGAGGGCCTCACTAAAC-3, and the probe 5 -CGCCCGTCGCTACTACCGATTGG-3, labeled at the 5 end with the reporter dye JOE and at the 3 end with the quencher dye carboxytetramethylrhodamine (TAMRA). Gene expression was normalized to the house-keeping gene 18S rrna and calculations were performed using the comparative Ct method (User Bulletin #2, ABI Prism 77 Sequence Detection System, P/N ) to assess the difference in TLR9 mrna levels in AK31/AKBM cells treated with or without anti-igg. 2 Transient transfections For assessment of the effects of BGLF5 on TLR9 protein levels, 293T cells were seeded at a density of 2.5 x 1 5 cells/ml, 2 ml/well, in a 6-well plate. After 24 h, cells were transfected with or without.125 μg of puno-htlr9-ha (Invivogen), and.625 μg (5X, relative to the quantity of puno-tlr9-ha transfected), 1.25 μg (1X), 2.5 μg (2X), or 3.75 μg (3X) of pcdna3-bglf5-ha-ires-nlsgfp (BGLF5) or 3.75 μg (3X) of pcdna3-bnlf2a-ires-nlsgfp (BNLF2a) using Lipofectamine 2 according to the instructions of the manufacturer (Invitrogen). The total amount of DNA was maintained at 4 μg in each case using pcdna3-ires-nlsgfp (kindly provided by E. Reits, Academic Medical Center, Amsterdam, The Netherlands). Lysates were generated 4 h post-transfection, as described below. Table I. Primer sequences for detection of TLR expression by RT PCR analysis TLR Forward primer Reverse primer # cycles 1 AAAAGAAGACCCTGAGGGCC TCTGAAGTCCAGCTGACCCT 4 a 2 AACCCTAGGGGAAACATCTCT GGAATATGCAGCCTCCGGAT 4 a 3 AAATTGGGCAAGAACTCACAGG GTGTTTCCAGAGCCGTGCTAA 45 a 4 TACAAAATCCCCGACAACCTC AGCCACCAGCTTCTGTAAACT 4 b 5 TGCATTAAGGGGACTAAGCCTC AAAAGGGAGAACTTTAGGGACT 4 a 6 TTGACAGTTTTGAGACTTTCCC TGAACCTCTGGTGAGTTCTG 45 a 7 TCCAGTGTCTAAAGAACCTGG TGGTATATATACCACACATCCC 4 b 8 TAATAGGCTCAAGCACATCCC TCCCAGTAAAACAAATGGTGAG 45 a 9 GTGCCCCACTTCTCCATG GGCACAGTCATGATGTTGTTG 38 a 1 TGACCACAATTCATTTGACTACTC TTGAATACTTTTGGGCAAGCACC 4 a Controls Forward primer Reverse primer # cycles 18S GTAACCCGTTGAACCCCATT GATCCGAGGGCCTCACTAAAC 25 a GAPDH GGATGATGTTCTGGAGAGCC CATCACCATCTTCCAGGAGC 32 a BGLF5 GGGACATATTGCGAAATGGC CACTGTGGCGGACGTAGTC 3 b a amplified with Taq DNA polymerase b amplified with Pfu DNA polymerase 61

62 Chapter 2 Western blotting Western Blot analysis was performed as previously described (27). In brief, for the detection of TLR9, HA-tagged BGLF5 and TfR, post-nuclear lysates were generated using.5% NP-4 buffer (.5% NP-4, 5 mm Tris HCl [ph 7.5], 5 mm MgCl2, 1 µm leupeptin, and 1 mm 4-(2-aminoethyl)benzenesulfonyl fluoride). Lysates were denatured in reducing sample buffer (final concentration: 2% SDS, 5 mm Tris-HCl [ph 8.], 1% glycerol, 5% 2-mercaptoethanol,.5% bromophenol blue), and solubilized proteins equivalent to 2 x 15 cells (for 293T cells) or 1 x 16 cells (for AK31 and AKBM cells) were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and transferred to polyvinylidene difluoride membranes (GE Healthcare, Waukesha, WI). Proteins of interest were detected by incubating the membranes with specific antibodies followed by HRP-conjugated specific secondary antibodies (Dako). Bound HRP-labeled antibodies were visualized using ECL Plus detection kit (GE Healthcare). For quantification of the protein expression levels, scanned films were analyzed with Quantity One software (Bio-Rad). Pulse-chase analysis For pulse-chase experiments, cells were cultured in methionine (Met)- and cysteine (Cys)-free RPMI medium (Lonza BioWhittaker, Fisher Scientific, Pittsburgh, PA) for 6 min at 37 C (starvation) prior to metabolic labeling for 3 min with [35S]Met/Cys (25 µci/ml; 35S Redivue Promix, a mixture of >7% [35S]Met and <3% [35S]Cys; Amersham, Buckinghamshire, United Kingdom) (pulse). Incorporation of the label was terminated by replacing the medium with culture medium supplemented with 1 mm Met and.1 mm Cys (chase). Precleared post-nuclear NP-4 cell lysates were subjected to immunoprecipitation with TLR9- or TfR-specific antibodies for 2-4 h at 4 C. Immune complexes were washed with NET buffer (.5% NP-4, 5 mm EDTA, 5 mm Tris-Hcl [ph 7.4], 15 mm NaCl), boiled for 5 min in reducing sample buffer and subjected to SDS-PAGE. Dried gels were exposed to a phosphorimaging screen, which was scanned with Personal Molecular Imager FX (Bio-Rad) and analyzed with Quantity One software. EBV stimulation of BJAB BJAB were seeded in a 48-well plate at a density of 5 x 15 cells/ml and stimulated for 18 h with CpG oligodeoxyribonucleotide (ODN) 26 type B (Invivogen, 1 μg/ml) or various amounts of EBV B95.8 purified virus particles (1 x 14, 5 x 14, 1 x 14, 25 x 14 or 5 x 14 virus particles/ml; Lot number 16-99, Advanced Biotechnologies, Columbia, MD). Surface expression of CD54 was determined by flow cytometry. Flow cytometry Cell surface expression of specific molecules was determined by using fluorochrome-conjugated antibody or indirectly by using unlabeled primary antibody together with goat anti-mouse allophycocyanin-conjugated antibody. GFP expression in AK31 cells, AKBM cells upon EBV lytic phase induction, or 293T cells upon transient transfection was measured without further staining. Cells were fixed with 2% paraformaldehyde and analyzed on a FACS Calibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ), using CellQuest Pro (BD Biosciences, San Jose, CA) and Flowjo software (TreeStar Inc., Ashland, OR). TLR reporter assays To examine activation of NF-κB by EBV particles, 293T and 293-TLR9-CFP cells were seeded in 96-well plates at a density of 2 x 15 cells/ml, 2 μl/well, 24 h before transfection with lipofectamine 2 following the manufacturer s instructions. Cells were transfected with 15 ng NF-κB-luciferase (a gift from P. Moynagh, NUI Maynooth, Ireland), 8 ng of phrl-tk (constitutively expressing Renilla luciferase) and 135 ng of pcdna3.1 to give a total amount of 23 ng DNA per well. 4 h post-transfection, cells were stimulated with TNF-α (1 ng/ml), CpG oligo-deoxyribonucleotide (ODN) 26 type B (Invivogen, 1 μg/ml) or various amounts of EBV B95.8 purified virus particles (5 x 14, 1 x 14, 2 x 14 or 4 x 14 virus particles per ml) for a further 6 h. NF-κB-induced firefly luciferase and Renilla luciferase activity were assayed using the Luciferase Assay Reagent (Promega, Madison, WI) and Renilla Luciferase Assay System (both Promega, Madison, WI, USA), respectively, according to the instructions of the manufacturer. Luminescence was measured with the LB94 Mithras Research II microplate reader (Berthold Technologies, Bad Wildbad, Germany). Degradation of in vitro transcribed mrna The pcr2.1-topo-tlr9-ha plasmid was linearized with SpeI restriction enzyme and used for in vitro transcription with T7 polymerase (Invitrogen) according to the manufacturer s protocol. Assays were performed in reaction buffer 62

63 EBV lytic-phase protein BGLF5 reduces TLR9 expression (25 mm NaCl and 2 mm Tris, ph7.5, supplemented with 1 mm MnCl2 or 5 mm EDTA in the presence or absence of 1 μm BGLF5 as previously described (33). The samples were incubated at 37 C for 3 min and stopped by the addition of EDTA to a final concentration of 5 mm before analysis on a 1.5% agarose gel stained with ethidium bromide. Gels were visualized with Gel Doc XR (Bio-Rad) and quantified using Quantity One software (Bio-Rad). Supporting information Akata EBV - 2 Mutu I TLR1 TLR9 TLR1 EBNA1 Lane Supplementary Figure 1. TLRs 1 and 1 are down-regulated in EBV latency. Expression levels of the indicated mrnas were determined by RT-PCR analysis. cdna was prepared from Akataderived EBV- (2A8) and EBV+ (AKBM) cells, as well as from Mutu I-derived EBV- (clone 9) and EBV+ (clone 3) cells. EBNA1 was used as a control for EBV latency I in the AKBM and the Mutu I clone 3 BL cells. counts A EBV gp35 B CD71 α-igg treated cells PAA AKBM AK TLR9 aspecific band β-actin Supplementary Figure 2. TLR9 expression is reduced at an early stage during the EBV lytic phase. (a) EBV+ AKBM cells treated with anti-human IgG antibody and with or without PAA were stained for gp35 and measured by flow cytometry. For comparison, expression of TfR was determined. Light grey lines: EBV+ AKBM cells in lytic phase without PAA treatment (gated on GFP+ cells). Dark grey lines: EBV+ AKBM cells arrested early in lytic phase by addition of PAA (gated on GFP+ cells). Black lines: FACS-staining control (no primary antibody). (b) Total protein levels of TLR9 in lysates from sorted populations of EBV- AK31 and EBV+ AKBM cells, treated with antiigg antibody together with or without PAA, were determined by Western Blot analysis. The specific TLR9 band is indicated; ns: non-specific band. For comparison, protein expression of β-actin was determined. 63

64 Chapter 2 References 1. Rickinson, A. B., and E. Kieff. 27. Epstein-Barr Virus. In Field s Virology, Vol. 2. D. M. Knipe and P. M. Howley, eds. Lippincott Williams & Wilkins, Philadelphia, pp Kutok, J. L., and F. Wang. 26. Spectrum of Epstein- Barr virus-associated diseases. Annu. Rev. Pathol. 1: Tattevin, P., T. Y. Le, S. Minjolle, A. Person, J. M. Chapplain, C. Arvieux, R. Thomas, and C. Michelet. 26. Increasing incidence of severe Epstein-Barr virus-related infectious mononucleosis: surveillance study. J. Clin. Microbiol. 44: Ressing, M. E., D. Horst, B. D. Griffin, J. Tellam, J. Zuo, R. Khanna, M. Rowe, and E. J. Wiertz. 28. Epstein-Barr virus evasion of CD8(+) and CD4(+) T cell immunity via concerted actions of multiple gene products. Semin. Cancer Biol. 18: Akira, S., S. Uematsu, and O. Takeuchi. 26. Pathogen recognition and innate immunity. Cell 124: Bowie, A. G., and I. R. Haga. 25. The role of Toll-like receptors in the host response to viruses. Mol. Immunol. 42: Boehme, K. W., M. Guerrero, and T. Compton. 26. Human cytomegalovirus envelope glycoproteins B and H are necessary for TLR2 activation in permissive cells. J. Immunol. 177: Kurt-Jones, E. A., M. Chan, S. Zhou, J. Wang, G. Reed, R. Bronson, M. M. Arnold, D. M. Knipe, and R. W. Finberg. 24. Herpes simplex virus 1 interaction with Toll-like receptor 2 contributes to lethal encephalitis. Proc. Natl. Acad. Sci. U. S. A 11: Gaudreault, E., S. Fiola, M. Olivier, and J. Gosselin. 27. Epstein-Barr virus induces MCP-1 secretion by human monocytes via TLR2. J. Virol. 81: Ariza, M. E., R. Glaser, P. T. Kaumaya, C. Jones, and M. V. Williams. 29. The EBV-encoded dutpase activates NF-kappa B through the TLR2 and MyD88-dependent signaling pathway. J. Immunol. 182: Iwakiri, D., L. Zhou, M. Samanta, M. Matsumoto, T. Ebihara, T. Seya, S. Imai, M. Fujieda, K. Kawa, and K. Takada. 29. Epstein-Barr virus (EBV)-encoded small RNA is released from EBV-infected cells and activates signaling from Toll-like receptor 3. J. Exp. Med. 26: West, J., and B. Damania. 28. Upregulation of the TLR3 pathway by Kaposi s sarcoma-associated herpesvirus during primary infection. J. Virol. 82: Fiola, S., D. Gosselin, K. Takada, and J. Gosselin. 21. TLR9 contributes to the recognition of EBV by primary monocytes and plasmacytoid dendritic cells. J. Immunol. 185: Quan, T. E., R. M. Roman, B. J. Rudenga, V. M. Holers, and J. E. Craft. 21. Epstein-Barr virus promotes interferon-alpha production by plasmacytoid dendritic cells. Arthritis Rheum. 62: Martin, H. J., J. M. Lee, D. Walls, and S. D. Hayward. 27. Manipulation of the toll-like receptor 7 signaling pathway by Epstein-Barr virus. J. Virol. 81: Hemmi, H., O. Takeuchi, T. Kawai, T. Kaisho, S. Sato, H. Sanjo, M. Matsumoto, K. Hoshino, H. Wagner, K. Takeda, and S. Akira. 2. A Toll-like receptor recognizes bacterial DNA. Nature 48: Takeshita, F., C. A. Leifer, I. Gursel, K. J. Ishii, S. Takeshita, M. Gursel, and D. M. Klinman. 21. Cutting edge: Role of Toll-like receptor 9 in CpG DNA-induced activation of human cells. J. Immunol. 167: Krug, A., G. D. Luker, W. Barchet, D. A. Leib, S. Akira, and M. Colonna. 24. Herpes simplex virus type 1 activates murine natural interferon-producing cells through toll-like receptor 9. Blood 13: Lund, J., A. Sato, S. Akira, R. Medzhitov, and A. Iwasaki. 23. Toll-like receptor 9-mediated recognition of Herpes simplex virus-2 by plasmacytoid dendritic cells. J. Exp. Med. 198: Tabeta, K., P. Georgel, E. Janssen, X. Du, K. Hoebe, K. Crozat, S. Mudd, L. Shamel, S. Sovath, J. Goode, L. Alexopoulou, R. A. Flavell, and B. Beutler. 24. Tolllike receptors 9 and 3 as essential components of innate immune defense against mouse cytomegalovirus infection. Proc. Natl. Acad. Sci. U. S. A 11: Takeda, K., T. Kaisho, and S. Akira. 23. Toll-like receptors. Annu. Rev. Immunol. 21: Moynagh, P. N. 25. TLR signalling and activation of IRFs: revisiting old friends from the NF-kappaB pathway. Trends Immunol. 26: Samuel, C. E. 21. Antiviral actions of interferons. Clin. Microbiol. Rev. 14:778, table. 24. Marshall, J. D., K. Fearon, C. Abbate, S. Subramanian, P. Yee, J. Gregorio, R. L. Coffman, and N. G. Van. 23. Identification of a novel CpG DNA class and motif that optimally stimulate B cell and plasmacytoid dendritic cell functions. J. Leukoc. Biol. 73: Kato, A., T. Ogasawara, T. Homma, J. Batchelor, S. Imai, H. Wakiguchi, H. Saito, and K. Matsumoto. 24. CpG oligodeoxynucleotides directly induce CXCR3 chemokines in human B cells. Biochem. Biophys. Res. Commun. 32: Ladell, K., M. Dorner, L. Zauner, C. Berger, F. Zucol, M. Bernasconi, F. K. Niggli, R. F. Speck, and D. Nadal. 27. Immune activation suppresses initiation of lytic Epstein-Barr virus infection. Cell Microbiol. 9: Ressing, M. E., S. E. Keating, L. D. van, D. Koppers- Lalic, I. Y. Pappworth, E. J. Wiertz, and M. Rowe. 25. Impaired transporter associated with antigen processingdependent peptide transport during productive EBV infection. J. Immunol. 174: Rowe, M., B. Glaunsinger, L. D. van, J. Zuo, D. Sweetman, D. Ganem, J. Middeldorp, E. J. Wiertz, and M. 64

65 EBV lytic-phase protein BGLF5 reduces TLR9 expression E. Ressing. 27. Host shutoff during productive Epstein- Barr virus infection is mediated by BGLF5 and may contribute to immune evasion. Proc. Natl. Acad. Sci. U. S. A 14: Keating, S., S. Prince, M. Jones, and M. Rowe. 22. The lytic cycle of Epstein-Barr virus is associated with decreased expression of cell surface major histocompatibility complex class I and class II molecules. J. Virol. 76: Bauer, S., C. J. Kirschning, H. Hacker, V. Redecke, S. Hausmann, S. Akira, H. Wagner, and G. B. Lipford. 21. Human TLR9 confers responsiveness to bacterial DNA via species-specific CpG motif recognition. Proc. Natl. Acad. Sci. U. S. A 98: Latz, E., A. Schoenemeyer, A. Visintin, K. A. Fitzgerald, B. G. Monks, C. F. Knetter, E. Lien, N. J. Nilsen, T. Espevik, and D. T. Golenbock. 24. TLR9 signals after translocating from the ER to CpG DNA in the lysosome. Nat. Immunol. 5: Chodosh, J., V. P. Holder, Y. J. Gan, A. Belgaumi, J. Sample, and J. W. Sixbey Eradication of latent Epstein-Barr virus by hydroxyurea alters the growthtransformed cell phenotype. J. Infect. Dis. 177: Buisson, M., T. Geoui, D. Flot, N. Tarbouriech, M. E. Ressing, E. J. Wiertz, and W. P. Burmeister. 29. A bridge crosses the active-site canyon of the Epstein-Barr virus nuclease with DNase and RNase activities. J. Mol. Biol. 391: Bernasconi, N. L., N. Onai, and A. Lanzavecchia. 23. A role for Toll-like receptors in acquired immunity: upregulation of TLR9 by BCR triggering in naive B cells and constitutive expression in memory B cells. Blood 11: Hislop, A. D., M. E. Ressing, L. D. van, V. A. Pudney, D. Horst, D. Koppers-Lalic, N. P. Croft, J. J. Neefjes, A. B. Rickinson, and E. J. Wiertz. 27. A CD8+ T cell immune evasion protein specific to Epstein-Barr virus and its close relatives in Old World primates. J. Exp. Med. 24: Iskra, S., M. Kalla, H. J. Delecluse, W. Hammerschmidt, and A. Moosmann. 21. Toll-like receptor agonists synergistically increase proliferation and activation of B cells by epstein-barr virus. J. Virol. 84: Altmann, M., and W. Hammerschmidt. 25. Epstein- Barr virus provides a new paradigm: a requirement for the immediate inhibition of apoptosis. PLoS. Biol. 3:e Takeshita, F., K. Suzuki, S. Sasaki, N. Ishii, D. M. Klinman, and K. J. Ishii. 24. Transcriptional regulation of the human TLR9 gene. J. Immunol. 173: Zuo, J., A. Currin, B. D. Griffin, C. Shannon-Lowe, W. A. Thomas, M. E. Ressing, E. J. Wiertz, and M. Rowe. 29. The Epstein-Barr virus G-protein-coupled receptor contributes to immune evasion by targeting MHC class I molecules for degradation. PLoS. Pathog. 5:e Ressing, M. E., L. D. van, F. A. Verreck, R. Gomez, B. Heemskerk, M. Toebes, M. M. Mullen, T. S. Jardetzky, R. Longnecker, M. W. Schilham, T. H. Ottenhoff, J. Neefjes, T. N. Schumacher, L. M. Hutt-Fletcher, and E. J. Wiertz. 23. Interference with T cell receptor-hla-dr interactions by Epstein-Barr virus gp42 results in reduced T helper cell recognition. Proc. Natl. Acad. Sci. U. S. A 1: Zhang, S. Y., E. Jouanguy, S. Ugolini, A. Smahi, G. Elain, P. Romero, D. Segal, V. Sancho-Shimizu, L. Lorenzo, A. Puel, C. Picard, A. Chapgier, S. Plancoulaine, M. Titeux, C. Cognet, B. H. von, C. L. Ku, A. Casrouge, X. X. Zhang, L. Barreiro, J. Leonard, C. Hamilton, P. Lebon, B. Heron, L. Vallee, L. Quintana-Murci, A. Hovnanian, F. Rozenberg, E. Vivier, F. Geissmann, M. Tardieu, L. Abel, and J. L. Casanova. 27. TLR3 deficiency in patients with herpes simplex encephalitis. Science 317: Casrouge, A., S. Y. Zhang, C. Eidenschenk, E. Jouanguy, A. Puel, K. Yang, A. Alcais, C. Picard, N. Mahfoufi, N. Nicolas, L. Lorenzo, S. Plancoulaine, B. Senechal, F. Geissmann, K. Tabeta, K. Hoebe, X. Du, R. L. Miller, B. Heron, C. Mignot, T. B. de Villemeur, P. Lebon, O. Dulac, F. Rozenberg, B. Beutler, M. Tardieu, L. Abel, and J. L. Casanova. 26. Herpes simplex virus encephalitis in human UNC-93B deficiency. Science 314: Kim, Y. M., M. M. Brinkmann, M. E. Paquet, and H. L. Ploegh. 28. UNC93B1 delivers nucleotide-sensing tolllike receptors to endolysosomes. Nature 452: Park, B., M. M. Brinkmann, E. Spooner, C. C. Lee, Y. M. Kim, and H. L. Ploegh. 28. Proteolytic cleavage in an endolysosomal compartment is required for activation of Toll-like receptor 9. Nat. Immunol. 9: Ewald, S. E., B. L. Lee, L. Lau, K. E. Wickliffe, G. P. Shi, H. A. Chapman, and G. M. Barton. 28. The ectodomain of Toll-like receptor 9 is cleaved to generate a functional receptor. Nature 456: Stack, J., I. R. Haga, M. Schroder, N. W. Bartlett, G. Maloney, P. C. Reading, K. A. Fitzgerald, G. L. Smith, and A. G. Bowie. 25. Vaccinia virus protein A46R targets multiple Toll-like-interleukin-1 receptor adaptors and contributes to virulence. J. Exp. Med. 21: Li, K., E. Foy, J. C. Ferreon, M. Nakamura, A. C. Ferreon, M. Ikeda, S. C. Ray, M. Gale, Jr., and S. M. Lemon. 25. Immune evasion by hepatitis C virus NS3/4A proteasemediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc. Natl. Acad. Sci. U. S. A 12: Hasan, U. A., E. Bates, F. Takeshita, A. Biliato, R. Accardi, V. Bouvard, M. Mansour, I. Vincent, L. Gissmann, T. Iftner, M. Sideri, F. Stubenrauch, and M. Tommasino. 27. TLR9 expression and function is abolished by the cervical cancer-associated human papillomavirus type 16. J. Immunol. 178: Nanbo, A., K. Inoue, K. chi-takasawa, and K. Takada. 22. Epstein-Barr virus RNA confers resistance to interferon-alpha-induced apoptosis in Burkitt s lymphoma. EMBO J. 21: Morrison, T. E., and S. C. Kenney. 24. BZLF1, an Epstein-Barr virus immediate-early protein, induces p65 nuclear translocation while inhibiting p65 transcriptional function. Virology 328:

66 Chapter Hahn, A. M., L. E. Huye, S. Ning, J. Webster-Cyriaque, and J. S. Pagano. 25. Interferon regulatory factor 7 is negatively regulated by the Epstein-Barr virus immediateearly gene, BZLF-1. J. Virol. 79: Morrison, T. E., A. Mauser, A. Klingelhutz, and S. C. Kenney. 24. Epstein-Barr virus immediate-early protein BZLF1 inhibits tumor necrosis factor alpha-induced signaling and apoptosis by downregulating tumor necrosis factor receptor 1. J. Virol. 78: Wu, L., E. Fossum, C. H. Joo, K. S. Inn, Y. C. Shin, E. Johannsen, L. M. Hutt-Fletcher, J. Hass, and J. U. Jung. 29. Epstein-Barr virus LF2: an antagonist to type I interferon. J. Virol. 83: Wang, J. T., S. L. Doong, S. C. Teng, C. P. Lee, C. H. Tsai, and M. R. Chen. 29. Epstein-Barr virus BGLF4 kinase suppresses the interferon regulatory factor 3 signaling pathway. J. Virol. 83:

67 EBV lytic-phase protein BGLF5 reduces TLR9 expression 2 67

68

69 CHAPTER 3 Silencing the shutoff protein of Epstein-Barr virus in productively infected B cells points to (innate) targets for immune evasion Michiel van Gent Anna M. Gram Ingrid G. J. Boer Ruben J. Geerdink Marthe F. S. Lindenbergh Robert Jan Lebbink Emmanuel J. Wiertz Maaike E. Ressing Published in Journal of General Virology (215), 96:

70 Chapter 3 Abstract During productive infection with Epstein-Barr virus (EBV), a dramatic suppression of cellular protein expression is caused by the viral alkaline exonuclease BGLF5. Among the proteins downregulated by BGLF5 are multiple immune components. Here, we show that shutoff reduces expression of the innate EBV-sensing Toll-like receptor-2 and the lipid antigenpresenting CD1d molecule, thereby identifying these proteins as novel targets of BGLF5. To silence BGLF5 expression in B cells undergoing productive EBV infection, we employed an shrna approach. Viral replication still occurred in these cells, albeit with reduced late gene expression. Surface levels of a group of proteins, including immunologically relevant molecules such as CD1d and HLA class I and class II, were only partly rescued by depletion of BGLF5, suggesting that additional viral gene products interfere with their expression. Our combined approach thus provides a means to unmask novel EBV (innate) immune evasion strategies that may operate in productively infected B cells. 7

71 Silencing BGLF5 expression in EBV-producing B cells Main text Herpesviruses are large enveloped DNA viruses that establish lifelong persistence in infected hosts. To achieve persistence, many herpesvirus gene products are dedicated to preventing elimination of virus-producing cells. For instance, members of all three herpesvirus subfamilies encode proteins that specifically interfere with antigen presentation to T cells (Griffin et al., 21). Prior to adaptive immunity, innate responses are elicited upon sensing of infection through pattern-recognition receptors, such as the Toll-like receptors (TLRs) (Iwasaki & Medzhitov, 21;Paludan et al., 213). These innate antiviral responses are also subject to herpesvirus immune evasion (Feng et al., 213;Ning, 211;Paludan et al., 213). Productive infection by α- and γ-herpesviruses induces a global inhibition of protein synthesis resulting from enhanced mrna degradation (Gaglia et al., 212). For the γ-herpesviruses, this shutoff is mediated by the viral alkaline exonuclease (AE) (Covarrubias et al., 29;Glaunsinger & Ganem, 24;Rowe et al., 27). AE proteins are conserved throughout the herpesvirus family, reflecting their critical DNase function in processing of newly synthesized viral genomes; their additional RNase-based shutoff function is unique to γ-herpesviruses (Glaunsinger & Ganem, 24;Rowe et al., 27). Shutoff appears broadly active and affects expression of most cellular proteins (Clyde & Glaunsinger, 211). As such, herpesvirusinduced shutoff provides a general strategy to dampen anti-viral immune activation. 3 Additional, more specific, immune evasive mechanisms operating in herpesvirus-infected cells could be masked by the general effects of shutoff. Indeed, the first examples of T cell escape by dedicated herpesvirus immunoevasins were identified in the absence of shutoff: α-herpesvirus-encoded inhibitors of antigen presentation by HLA class I (HLA I) molecules were identified using shutoff-defective mutant viruses (Koppers-Lalic et al., 23;York et al., 1994), and multiple HLA I evasion strategies were identified for the β-herpesvirus human cytomegalovirus that lacks a virus-encoded shutoff function (Barnes & Grundy, 1992). Cooperative targeting of a single immune pathway by multiple viral gene products has emerged as a common theme (Jones et al., 1995;Ressing et al., 28). Compared to α-herpesviruses, shutoff by γ-herpesviruses has been discovered more recently. Consequently, less is known about the role of AE-mediated shutoff in immune evasion during productive γ-herpesvirus infection. The prototypic human γ-herpesvirus, Epstein-Barr virus (EBV), naturally infects B cells, which form the latent virus reservoir in vivo (Rickinson & Kieff, 27). For production of new viral progeny, EBV reactivates from a small percentage of latently infected B cells. The AE protein of EBV, BGLF5, is expressed during this lytic phase of infection. Earlier, we have reported 71

72 Chapter 3 that cellular expression of BGLF5 downregulates immunologically relevant proteins, such as HLA molecules and TLR9 (Rowe et al., 27;van Gent et al., 211;Zuo et al., 28), providing a means of general immune evasion. In this study, we aimed to evaluate BGLF5 s effects in the context of productive EBV infection in B cells. Several approaches have been used to eliminate expression of individual herpesvirus genes from infected cells, one of which is based on the use of bacterial artificial chromosomes (BACs) (Delecluse et al., 28). Using this approach, deletion of BGLF5 was shown to perturb EBV replication in transfected 293T cells, resulting in reduced viral yields (Feederle et al., 29a). Studying EBV mutants during productive infection of B cells has been more difficult. Here, we have employed the EBV + Akata B cell line AKBM, in which cross-linking of the B cell receptor with anti-human IgG reactivates EBV in 1-4% of cells. Productively infected B cells can be identified and sorted on the basis of induced expression of a reporter protein, ratcd2-gfp (rcd2-gfp) (Ressing et al., 25). Using this system, we have elucidated several immune evasion mechanisms acting during the productive phase of EBV infection (Horst et al., 29;Ressing et al., 25;van Gent et al., 211;van Gent et al., 214). As our approach to suppress BGLF5-mediated shutoff during productive infection in B cells, lentivirus-delivered shrnas were introduced into these EBV + AKBM cells. Ten candidate shrnas that target sites within the BGLF5 coding sequence were cloned into a lentiviral vector (Fig. S1) (Lebbink et al., 211). Two of these considerably reduced BGLF5 levels in EBV-producing B cells, whereas the expression of a control protein, transferrin receptor (CD71), was not substantially affected (Fig. S2). Combining the two shrnas (referred to as shbglf5) reduced BGLF5 protein levels by 6-75% compared to lytically induced control AKBM cells (Fig. 1a, S2). Also mrna levels of BGLF5 were markedly reduced in productively infected AKBM-shBGLF5 cells, whereas levels of another EBV transcript, BNLF2a, remained unchanged (Fig. 1b). Thus, stable expression of specific shrnas through lentiviral transduction substantially reduced BGLF5 levels during productive EBV infection in B cells. AE proteins, through their conserved DNase function, are required for processing of replicated herpesvirus genomes in infected cells. We examined whether silencing of BGLF5 in B cells interfered with progression through the EBV replication cycle. Upon reactivation, Figure 1. Silencing BGLF5 in productively EBV-infected B cells. AKBM B cells stably expressed no shrna, a combination of two BGLF5-targeting shrnas (shbglf5), or an shrna targeting Fas (shcontrol). By 2 hours of anti-human IgG treatment (+ anti-igg), a population of cells had entered the EBV lytic cycle, with concomitant expression of the rcd2-gfp reporter. (a) Intracellular BGLF5 levels were determined by flow cytometry, with percentages indicating BGLF5 levels compared to those in shcontrol cells. In over 1 experiments, around 15-4% BGLF5 protein expression remained in lytic AKBM-shBGLF5 cells. See Fig. 72

73 Silencing BGLF5 expression in EBV-producing B cells (a) AKBM B cells (b) AKBM B cells - shrna shcontrol shbglf5 (c) + anti-igg % % BGLF5 AKBM B cells No primary Ab Latent cells (GFP-) Lytic cells (GFP+) shcontrol shbglf anti-igg BGLF5 BNLF2a 18S rrna Lane 3 - shrna - shrna shcontrol shbglf5 IE E L BZLF1 BGLF5 rcd2-gfp gp35 ghgl 1 % % 1 % 1 % % 1 % % % 1 % % % 1% 1 15% 1 85% % 1 1% 24% 1 76% 23% 77% % 99% 1 7% 1 93% % 1 1% 2% 1 8% 3% 7% % 99% 1 1 % 1 1% 46% 54% 1 41% 59% Ab - anti-igg (latent) + anti-igg (lytic) EBV antigen S2 for further details. (b) Semi-quantitative RT-PCR analysis was performed to determine mrna levels of BGLF5, an EBV control transcript (BNLF2a), or a cellular control RNA (18S ribosomal RNA). Almost pure populations of productively EBV-infected cells were obtained through magnetic sorting of lytically induced cells labeled with mouse-anti-rcd2 Ab and anti-mouse-magnetic beads (Miltenyi) (Ressing et al., 25). (c) Effects of shrnas on progression through the replicative cycle were monitored by flow cytometric analysis of several EBV antigens (Table S1, Fig. S3). Percentages indicate viral antigen-negative (left) and -positive cells (right) within the productively infected population (lytic cells); solid lines, lytic AKBM cells gated on expression of the rcd2-gfp reporter protein (GFP + ); dashed lines, latently EBV-infected cells (GFP - ); grey lines, no primary Ab. 73

74 Chapter 3 immediate-early (IE), early, and late herpesvirus proteins are sequentially expressed (Table S1). Anti-IgG treatment of AKBM-shBGLF5 and control cells caused similar amounts of B cells to become positive for the IE transactivator BZLF1 and the early-expressed rcd2-gfp reporter (Figs 1c, S3), both of which precede expression of the BGLF5 protein. In contrast, the proportion of AKBM-shBGLF5 cells expressing late proteins gp35, gh, and gl was substantially reduced. Thus, silencing BGLF5 appears to hamper entry into the late phase of productive EBV infection in B cells. To evaluate the effects of shutoff, relying on BGLF5 s RNase activity, in EBV-producing B cells, we examined the influence of BGLF5 silencing on the downregulation of various surface proteins. In control cells, EBV reactivation caused a minor reduction in CD71 surface levels, while HLA I and II were strongly downregulated (Figs 2a and S4a, upper panels), which is in line with earlier observations (Ressing et al., 25). In lytic AKBM-shBGLF5 cells, surface display of HLA I and II was partly rescued (Fig. 2a and S4a, lower panels), supporting a contribution of shutoff to evasion from T cell detection during EBV replication in B cells. Still, levels of these antigen presenting molecules remained markedly reduced on BGLF5-silenced cells, which could reflect the specific effects on HLA I expression mediated by two dedicated EBV lytic cycle proteins, BNLF2a and BILF1 (Ressing et al., 28;Zuo et al., 29). The analysis was extended to a panel of additional cellular proteins detectable at the surface of latently EBV-infected B cells (Table S2). Following viral reactivation, cellular display of the markers tested was reduced to varying degrees (Fig. 2b,d, S4b). CD58, CD119, CD1, and CD45 were marginally affected and, therefore, the effect of BGLF5 silencing was difficult to evaluate (group I; Fig. 2b,d, S4b), as was the case for CD71 (Fig. 2a). Surface levels of another group of proteins, comprising CD38, CD47, CD19, and CD2, were strongly reduced during productive EBV infection of control cells and they remained downregulated in induced AKBM-shBGLF5 cells (group II, Fig. 2b,d, S4b). The phenotype for this latter group of proteins resembles that of the peptide-presenting HLA I and II complexes. We also included the non-classical HLA molecule CD1d in this analysis. CD1d molecules present lipid antigens to invariant natural killer T (inkt) cells that express a semi-invariant T cell receptor as well as NK cell markers. inkt cells act at the interface of innate and adaptive immunity: they rapidly produce polarizing cytokines when activated, for instance in response to viral infection (Horst et al., 212a;Kinjo et al., 213). Induction of the EBV lytic cycle in AKBM-CD1d cells caused a dramatic decrease in surface appearance of human CD1d molecules (Fig. 2c,d, S4c). Although CD1d expression was partly restored when BGLF5 was silenced, it remained far below the levels observed on latently infected cells. Thus, CD1drestricted antigen presentation appears a novel target of EBV immune evasion, in part mediated by the shutoff protein BGLF5. 74

75 Silencing BGLF5 expression in EBV-producing B cells (a) AKBM B cells Ab shcontrol shbglf CD71 HLA I HLA II antigen expression BGLF No primary Ab Latent cells (GFP-) Lytic cells (GFP+) (b) AKBM B cells Ab shcontrol shbglf CD CD Downregulation in lytically-infected AKBM cells CD CD CD CD CD19 CD antigen expression (c) AKBM-CD1d cells Ab shcontrol shbglf CD1d BGLF antigen expression Figure 2. BGLF5 silencing rescues expression of surface proteins on EBV-producing B cells. EBV reactivation was induced in AKBM-shBGLF5 or -shcontrol cells. (a,b) Surface levels of the cellular antigens in Table S2, in (b) ordered by increasing (d) % expression in lytically-infected AKBM (-CD1d) cells (compared to latently infected cells) 1% shbglf HLA I CD2 CD1d HLA II 2 CD38 CD47 CD19 Group II CD58 CD71 CD shcontrol 8 CD1 CD119 Group I downregulation from EBV-producing cells. (c) Surface levels of CD1d on AKBM cells that stably expressed human CD1d molecules after lentiviral transduction (AKBM- CD1d cells). The extent of BGLF5 silencing was visualized by intracellular staining. Solid lines, productively EBVinfected (GFP + ) cells; dashed lines, latently infected (GFP - ) cells; grey, no primary Ab (dot plots in Fig. S4). (d) Relative protein levels on productively versus latently EBV - infected shcontrol cells (horizontal axis) were plotted against those in shbglf5 cells (vertical axis). Values were obtained by dividing geometric mean fluorescence intensities by background signals (isotype control or without primary antibody) and denoted as percentages expression in lytically compared to latently infected cells, as determined in at least four independent experiments (average ± SD; values for CD58, CD45, and CD119 based on two replicates). 1% 75

76 Chapter 3 The combined data imply that B-cell proteins whose surface display remains markedly reduced on BGLF5-silenced cells (Fig. 2d, group II) are likely to be downregulated by additional EBV lytic phase proteins, for instance to effectuate reduced recognition of virus-producing B cells by the immune system. To complement the studies performed in naturally infected AKBM-CD1d-shBGLF5 cells, we investigated BGLF5 s effects on the non-classical antigen-presenting molecule CD1d in cells expressing BGLF5 in isolation. MJS-CD1d cells were transiently transfected with BGLF5 and reduction of GFP and surface HLA I levels cells confirmed induction of shutoff. The BGLF5-transfected cells displayed reduced surface expression of CD1d (Fig. 3a). This CD1d downregulation was, however, less pronounced than that on B cells expressing all EBV gene products (Fig. 2c), reminiscent of the phenotype for HLA I (Fig. 2a) (Rowe et al., 27). These results show that BGLF5 reduces CD1d levels and that other viral factors, absent from the transfected MJS cells, are likely to add to the robust CD1d downregulation observed during productive EBV infection of B cells. Earlier, we have found that expression of innate sensors, namely several TLRs, is reduced upon EBV reactivation in AKBM cells and that BGLF5 contributes to the downregulation of TLR9 (van Gent et al., 211). TLR2, 3, and 9 sense EBV particles (Gaudreault et al., 27;Iwakiri et al., 29;van Gent et al., 211), yet no evidence for TLR4-mediated recognition of EBV has been reported (Gaudreault et al., 27). Here, we monitored the influence of BGLF5 on TLR2 and TLR TLR2 and 293-TLR4 cells were transiently transfected with the empty IRES- GFP vector, wild-type BGLF5, or a catalytically inactive mutant, BGLF5 D23S. TLR2 levels were reduced on wild-type BGLF5-expressing cells, but not on control cells (Fig. 3b). In contrast, TLR4 levels were not affected by any of the transfected gene products. Thus, BGLF5-mediated shutoff appears to target TLR2, a pattern-recognition receptor sensing EBV. To conclude, this study shows that lentivirus-delivered shrnas can successfully be applied to our system for productive EBV infection of B cells to achieve stable silencing of BGLF5. A similar approach in EBV-transformed B-LCLs yielded around 75% knockdown of viral gene expression, which was sufficient to reveal a hierarchy in immune evasive properties of BNLF2a, BILF1, and BGLF5 (Quinn et al., 214). Figure 3. BGLF5 contributes to downregulation of CD1d and TLR2. (a,b) MelJuSo cells were lentivirally transduced to stably express human CD1d molecules (MJS-CD1d). MJS-CD1d cells (a) and 293-TLR2/CD14 cells (Kurt-Jones et al., 22) or 293-TLR4/CD14/MD2 cells (Invitrogen) (b) were transiently transfected with a pcdna3-ires-nlsgfp vector without insert or encoding BGLF5 or the catalytic mutant BGLF5 D23S (Horst et al., 212b). At 48 hours post-transfection, surface levels of HLA I, CD1d, TLR2, and TLR4 were determined by flow cytometry. Downregulation of GFP served as a measure for BGLF5 s shutoff function. Dashed lines, untransfected GFP - cells; solid lines, transfected GFP + cells; grey lines, secondary Ab only. Data shown are from one experiment representative of three independent experiments. 76

77 Silencing BGLF5 expression in EBV-producing B cells (a) MJS-CD1d cells No primary Ab Untransfected cells (GFP-) Transfected cells (GFP+) Ab antigen expression Control HLA I CD1d HLA I CD1d Control BGLF5 transfection (b) 293-TLR2 cells Ab Control antigen GFP HLA I TLR antigen expression HLA I TLR Control BGLF5 BGLF5 D23S transfection 3 GFP antigen expression 293-TLR4 cells Ab antigen Control HLA I TLR HLA I TLR Control BGLF5 BGLF5 D23S transfection GFP antigen expression 77

78 Chapter 3 In our system of EBV-producing AKBM cells, a comparable reduction of BGLF5 protein levels interfered with viral replication (Fig. 1). While confirming that knockdown in EBVproducing B cells was sufficiently robust to observe a phenotype, this observation extends earlier studies in 293T cells transfected with a BGLF5 deletion mutant EBV BAC (Feederle et al., 29a). Within the EBV genome, BGLF5 occurs in tandem with BGLF4, which codes for the EBV protein kinase that can regulate EBV late gene expression (El-Guindy et al., 214). Since BGLF4 is translated from a transcript encoding both BGLF4 and BGLF5, expression of both proteins was lost from BGLF5-deleted virus-producing 293T cells (Feederle et al., 29b). Likewise, the use of RNA interference to silence BGLF5 expression in B cells will target both BGLF5 and BGLF5+BGLF4 transcripts. When applying this approach to AKBM B cells, we have focused on the shutoff effects that are selectively induced by BGLF5. During productive EBV infection, a broad range of B cell surface proteins is downregulated in the presence of BGLF5, and this effect is partly reversed upon silencing of BGLF5 (Fig. 2). These findings are in agreement with promiscuous shutoff by γ-herpesvirus AE proteins, deduced from mrna target analysis (Clyde & Glaunsinger, 211) and metabolic labeling experiments (Rowe et al., 27). Still, some gene products escape shutoff and TLR4 appears to be one of them (Fig. 3). Based on our current data, two groups of host surface proteins can be discriminated. The first group comprises proteins that are downregulated to a limited extent during EBV replication; the costimulatory molecules CD8 and CD86 can be included in this group (Ressing et al., 25). The second group is more strongly downregulated, likely by multiple EBV lytic proteins, and their surface levels remain substantially reduced when BGLF5 is silenced. For some proteins belonging to this latter group, a causative role for BGLF5 in their downregulation has been confirmed through transient transfection experiments, i.e. for HLA I and II (Rowe et al., 27), for TLR9 (van Gent et al., 211), and for CD1d and TLR2 (this study, Fig. 3). These combined findings support the notion that BGLF5 contributes to EBV-induced immune evasion during productive infection of B cells. A recent study in BGLF5-silenced B-LCLs revealed a minor role for BGLF5 in CD8 + T cell evasion when compared to BNLF2a and BILF1 (Quinn et al., 214). Along the same line, we observed only partially rescued surface display of B cell proteins upon BGLF5 knockdown during productive EBV infection (Fig. 2). Residual downregulation could result from the ~25% BGLF5 protein expression that remained in induced AKBM-shBGLF5 cells and/or from additional EBV-encoded shutoff function(s), such as that recently reported to be exerted by BZLF1 (Park et al., 214). In vivo studies on the α-herpesvirus HSV-1 and the murine γ-herpesvirus MHV68 suggest that the immune evasive functions of shutoff mainly affect newly synthesized proteins induced by type I interferons (Murphy et al., 23;Pasieka et al., 28;Sheridan et al., 214). The above observations, together with the absence of shutoff from β-herpesviruses, point to a relatively small contribution of shutoff to immune evasion. This would provide a rationale as to why herpesviruses have acquired additional, specific immune 78

79 Silencing BGLF5 expression in EBV-producing B cells evasion mechanisms to synergistically achieve the proper timing and extent of immune interference. Analogous to all other herpesviruses studied, EBV encodes multiple gene products that act in concert to prevent T cell activation (Ressing et al., 28). Additional EBV strategies interfering with innate immunity continue to be identified (Ning, 211). Here, we have added CD1d and TLR2 to the target list of EBV BGLF5. Furthermore, we have identified a group of B cell surface proteins including CD1d whose expression is likely downregulated by EBV lytic phase proteins besides BGLF5. Acknowledgements This work was financially supported by the Netherlands Scientific Organization (NWO Vidi grant to M.E.R.) and the Dutch Cancer Foundation (grant UU to E.J.W, R.J.L., and M.E.R). 3 Supporting information Table S1. Viral antigens expressed during productive EBV infection ORF Name Timing Function Ab clone BZLF1 Z, Zta, ZEBRA IE Lytic cycle transactivator BZ.1 BNLF2a E HLA I downregulation N/A BGLF5 EBV AE, DNase E Shutoff and exonuclease 311H BLLF1 gp35 L Major envelope glycoprotein, attachment to CR2 72A1 BXLF1 gp85/gh L Viral membrane fusion/entry E1D1 BKRF2 gp25/gl L Chaperone of gh E1D1 rcd2-gfp rat CD2-GFP fusion protein E Lytic reporter, magnetic sorting OX-34 (α-ratcd2) Continues on next page 79

80 Chapter 3 Table S2. Expression of cellular surface proteins by AKBM cells CD# Name Function Ab Not expressed by latently EBV-infected AKBM cells CD1d Non-classical HLA molecule, lipid presentation to inkt cells CD11b Integrin alpha M Complement receptor 3 subunit BD5519 CD11c Integrin alpha X Complement receptor 4 subunit BD CD21 CR2 Complement receptor 2, B cell coreceptor component (with CD19), EBV B cell receptor BC IM473U CD23 FcεRII IgE receptor BD CD25 IL-2 receptor alpha-chain BD34111 CD32 B cell coreceptor RDI-CD32abm- 7PE CD35 CR1 Complement receptor 1 BD CD4 Costimulation SC65263 CD44 Variety of lymphocyte functions BD55989 CD54 ICAM-1 Intercellular adhesion molecule BD CD132 Common gamma chain (γc) Common subunit of at least six interleukin receptors BD CD162 P-selectin glycoprotein ligand-1 Neutrophil recruitment BD55655 CD181 CXCR1 IL-8 receptor R&D-FAB33P CD191 CCR1 Chemokine receptor, inflammatory responses R&D-FAB145P CD192 CCR2 Chemokine receptor, monocyte chemotaxis R&D-FAB151P CD282 Toll-Like Receptor 2 Pattern-recognition receptor ebioscience CD284 Toll-Like Receptor 4 Pattern-recognition receptor Biolegend Marginally downregulated during productive EBV infection (group I) CD1 Neprilysin Zinc-dependent metalloprotease BD34921 CD45 Leukocyte common antigen Protein tyrosine phosphatase BD CD58 LFA-3 Adhesion molecule, antigen-presenting cell - T cell interaction BD CD71 Transferrin receptor Iron uptake BD CD119 IFNγRα Ligand binding α chain of IFNγ receptor BD Substantially downregulated during productive EBV infection (group II) CD19 B cell coreceptor component, reduces activation threshold BD CD2 Enables optimal B cell immune response BD CD38 Cyclic ADP ribose hydrolase Multifunctional ectoenzyme, regulation of intracellular Ca 2+ levels BD CD47 Integrin associated protein Range of cellular processes BD55647 HLA I Classical HLA molecule, peptide presentation to CD8 + T cells W6/32 HLA II Classical HLA molecule, peptide presentation to CD4 + T cells L243 8

81 Silencing BGLF5 expression in EBV-producing B cells M A D V D E L E D P M E E M T S Y T F A R F L R S P E T E A F V R N L D R P P Q M P ATGGCCGACGTGGATGAGCTCGAGGATCCCATGGAGGAGATGACCTCCTACACGTTTGCCCGCTTCCTCCGCAGTCCGGAGACTGAGGCCTTTGTCCGTAACCTTGACCGTCCACCTCAGATGCC B95.8 ATGGCCGACGTGGATGAGCTCGAGGATCCCATGGAGGAGATGACCTCCTACACGTTTGCCCGCTTCCTCCGCAGTCCGGAGACTGAGGCCTTTGTCCGTAACCTTGACCGTCCGCCTCAGATGCC Akata M A D V D E L E D P M E E M T S Y T F A R F L R S P E T E A F V R N L D R P P Q M P shrna1 A M R F V Y L Y C L C K Q I Q E F S G E T G F C D F V S S L V Q E N D S K D G P S L GGCCATGCGCTTTGTCTATCTCTATTGCCTCTGTAAACAAATACAAGAGTTCTCTGGTGAAACTGGCTTCTGTGACTTTGTCTCCTCGTTAGTCCAAGAGAATGACAGCAAGGACGGTCCCTCCC B GGCCATGCGCTATGTCTATCTCTATTGCCTCTGTAAACAAATACAAGAGTTCTCTGGTGAAACTGGCTTCTGTGACTTTGTCTCCTCGTTAGTCCAAGAGAATGACAGCCAGGACGGTCCCTCCC Akata A M R Y V Y L Y C L C K Q I Q E F S G E T G F C D F V S S L V Q E N D S Q D G P S L 251 shrna2 shrna3 K S I Y W G L Q E A T D E Q R T V L C S Y V E S M T R G Q S E N L M W D I L R N G TGAAATCCATTTACTGGGGGCTACAGGAGGCCACCGACGAGCAGAGGACTGTTCTCTGCTCGTACGTGGAGTCCATGACCAGGGGGCAGTCTGAGAACCTGATGTGGGACATATTGCGAAATGGC B95.8 TGAAATCCATTTACTGGGGGCTACAGGCGGCCACCGACGAGCAGAGGACTGTTCTCTGCTCGTACGTGGAGTCCATGACCAGGGGGCAGTCTGAGAACCTGATGTGGGACATATTGCGAAATGGC Akata K S I Y W G L Q A A T D E Q R T V L C S Y V E S M T R G Q S E N L M W D I L R N G 376 I I S S S K L L S T I K N G P T K V F E P A P I S T N H Y F G G P V A F G L R C E D ATAATTTCCTCTTCCAAGCTGCTCTCCACCATTAAGAATGGACCCACCAAGGTGTTTGAGCCAGCTCCCATCTCCACAAATCACTACTTTGGGGGACCTGTGGCCTTTGGCCTGCGGTGTGAGGA ATAATTTCCTCTTCCAAGCTGCTCTCCACCATTAAGAATGGACCCACCAAGGTGTTTGAGCCAGCTCCCATCTCCACAAATCACTACTTTGGGGGACCTGTGGCCTTTGGCCTGCGGTGTGAGGA I I S S S K L L S T I K N G P T K V F E P A P I S T N H Y F G G P V A F G L R C E D B95.8 Akata shrna4 shrna5 T V K D I V C K L I C G D A S A N R Q F G F M I S P T D G I F G V S L D L C V N V E CACGGTCAAGGACATTGTCTGTAAGCTCATCTGCGGGGACGCATCCGCCAACCGTCAATTTGGCTTTATGATTAGTCCCACGGATGGCATTTTTGGGGTGTCTCTGGATCTTTGCGTCAATGTGG B CACGGTCAAGGACATTGTCTGTAAGCTCATCTGCGGGGACGCATCTGCCAACCGTCAATTTGGCTTTATGATTAGTCCCACGGATGGCATTTTTGGGGTGTCTCTGGATCTTTGCGTCAATGTGG Akata T V K D I V C K L I C G D A S A N R Q F G F M I S P T D G I F G V S L D L C V N V E S Q G D F I L F T D R S C I Y E I K C R F K Y L F S K S E F D P I Y P S Y T A L Y AGTCACAGGGAGACTTTATACTGTTCACCGACCGGAGCTGCATTTATGAGATTAAGTGCCGCTTCAAGTACCTCTTTTCCAAGTCAGAATTTGACCCCATCTACCCATCCTACACTGCGCTTTAC B AGTCACAGGGAGACTTTATACTGTTCACCGACCGGAGCTGCATTTATGAGATTAAGTGCCGCTTCAAGTACCTCTTTTCCAAGTCAGAGTTTGACCCCATCTACCCATCCTACACTGCGCTTTAC Akata S Q G D F I L F T D R S C I Y E I K C R F K Y L F S K S E F D P I Y P S Y T A L Y 3 shrna6 K R P C K R S F I R F I N S I A R P T V E Y V P D G R L P S E G D Y L L T Q D E A W AAGAGGCCATGCAAGAGGTCATTTATCAGATTTATCAATTCTATAGCTCGTCCTACCGTGGAGTACGTCCCGGATGGGCGGTTGCCCTCGGAGGGTGACTATCTGTTGACGCAGGATGAGGCCTG B AAGAGGCCATGCAAGAGGTCATTTATCAGATTTATCAATTCTATAGCTCGTCCTACCGTGGAGTACGTACCGGATGGGCGGTTGCCCTCGGAGGGTGACTATCTGTTGACGCAGGATGAGGCCTG Akata K R P C K R S F I R F I N S I A R P T V E Y V P D G R L P S E G D Y L L T Q D E A W shrna7 shrna8 N L K D V R K R K L G P G H D L V A D S L A A N R G V E S M L Y V M T D P S E N A G GAATCTTAAAGATGTCCGTAAGCGCAAACTGGGCCCCGGTCATGACCTGGTGGCAGACAGCCTAGCTGCCAACAGGGGGGTGGAGTCTATGCTCTACGTAATGACGGACCCAAGCGAAAATGCGG B GAATCTTAAAGATGTCCGTAAGCGCAAACTGGGCCCCGGTCATGACCTGGTGGCAGACAGCCTAGCTGCCAACAGGGGGGTGGAGTCTATGCTCTACGTAATGACGGACCCAAGCGAAAATGCGG Akata N L K D V R K R K L G P G H D L V A D S L A A N R G V E S M L Y V M T D P S E N A G 11 shrna9 R I G I K D R V P V N I F I N P R H N Y F Y Q V L L Q Y K I V G D Y V R H S G G G GGCGCATTGGTATTAAAGACCGGGTCCCAGTCAACATCTTCATCAATCCACGGCACAACTACTTCTACCAGGTGCTCCTCCAATACAAAATTGTTGGAGACTACGTCCGCCACAGTGGGGGTGGC B95.8 GGCGCATTGGTATTAAAGACCGGGTCCCAGTCAACATCTTCATCAATCCACGGCACAACTACTTCTACCAGGTGCTCCTCCAATACAAAATTGTTGGAGACTACGTCCGCCACAGTGGGGGTGGT Akata R I G I K D R V P V N I F I N P R H N Y F Y Q V L L Q Y K I V G D Y V R H S G G G 1126 K P G R D C S P R V N I V T A F F R K R S P L D P A T C T L G S D L L L D A S V E I AAGCCCGGGAGAGACTGCTCACCCCGGGTGAACATTGTGACGGCCTTCTTTCGAAAACGGTCGCCTCTAGACCCGGCGACCTGCACGCTCGGCTCAGACCTGCTTCTGGACGCCTCGGTGGAGAT B95.8 AAGCCCGGGAGAGACTGCTCACCCCGGGTGAACATTGTGACGGCCTTCTTTCGAAAACGGTCGCCTCTAGACCCGGCGACCTGCACGCTCGGCTCAGACCTGCTTCTGGACGCCTCGGTGGAGAT Akata K P G R D C S P R V N I V T A F F R K R S P L D P A T C T L G S D L L L D A S V E I shrna1 P V A V L V T P V V L P D S V I R K T L S T A A G S W K A Y A D N T F D T A P W V P TCCCGTGGCGGTGCTGGTGACACCCGTGGTCCTGCCGGACTCTGTCATCCGTAAGACCTTGAGCACCGCGGCTGGCTCCTGGAAAGCGTACGCAGACAATACTTTTGACACCGCGCCATGGGTGC B TCCCGTGGCGGTGCTGGTGACACCCGTGGTCCTGCCGGACTCTGTCATCCGTAAGACCTTGAGCACCGCGGCTGGCTCCTGGAAAGCGTACGCAGACAATACTTTTGACACCGCGCCATGGGTGC Akata P V A V L V T P V V L P D S V I R K T L S T A A G S W K A Y A D N T F D T A P W V P 1376 S G L F A D D E S T P. CCTCTGGTCTCTTTGCCGACGACGAGTCAACTCCATAG B95.8 CCTCTGGTCTCTTTGCCGACGACGAGTCAACTCCATAG Akata S G L F A D D E S T P. Figure S1. shrna target sequences within EBV BGLF5. Nucleotide and amino acid alignments are depicted for the BGLF5 open reading frames from EBV strains B95.8 and Akata. Numbers indicate nucleotide positions and boxes show inter-strain differences. Ten candidate shrna target sites within the BGLF5 coding sequence (in grey) were selected using a prediction algorithm (the Hannon lab website: (Paddison et al., 24). As a control shrna, a target sequence (tatgcagaggatgaaagattaa) within the cellular gene encoding the Fas receptor was used. shrnas were cloned into a lentiviral vector derived from psicor (Jacks Lab, MIT), in which the U6 promoter was altered to allow for sticky cloning of shrnas in between a BstXI and XhoI site, and in which an EF1α promoter was used to drive expression of a cassette encoding a puromycin resistance marker, the ribosome skipping peptide T2A, and mcherry. The resulting vectors were used for the generation of replication-deficient self-inactivating lentivirus stocks, as described (Lebbink et al., 211). 81

82 Chapter 3 (a) AKBM B cells + anti-igg - shrna - shrna shrna1 shrna2 shrna3 shrna5 shrna6 shrna8 shrna9 shrna rcd2-gfp AKBM B cells - shrna - shrna shrna1 shrna2 shrna3 shrna5 shrna6 shrna8 shrna9 shrna1+9 22% CD71 BGLF5 (b) No primary Ab Latent cells (GFP-) + anti-igg Lytic cells (GFP+) 1 % 8 1 1% % % % % % % % BGLF5 1 % 8 1 1% % % % % % % 8 1 1% % CD71 82

83 Silencing BGLF5 expression in EBV-producing B cells Figure S2. Experimental selection of shrna target sequences that silence BGLF5 protein expression. AKBM B cells were lentivirally transduced and puromycin selected (>95% mcherry-positive) to stably express single BGLF5-targeting shrnas or a combination of two BGLF5 shrnas (shrnas #1 and #9, referred to as shbglf5 in the remainder of the manuscript). Productive EBV infection was induced in these cells by treatment with anti-human IgG Abs (+ anti-igg: 5 μg/ml; Cappel, MP Biochemicals) for 2 hours (Ressing et al., 25). Subsequently, cells were fixed, permeabilized, stained for intracellular expression of BGLF5 (Ab 311H; intracellular staining; (Horst et al., 212b)), and analyzed by flow cytometry. The transferrin receptor (CD71; Ab BD555534) was taken along as a control protein and was visualized at the surface of unfixed and non-permeabilized cells ( surface staining). Results of a representative experiment with seven out of the ten shrnas tested are depicted as dot plots (a) and as overlay histograms (b). Solid lines, lytically induced AKBM cells gated on expression of the rcd2-gfp reporter protein (GFP + ); dashed lines, latently EBV-infected cells (GFP - ); grey lines, no primary Ab. Percentages denote BGLF5 protein expression levels in productively infected AKBM cells with BGLF5-targeting shrnas compared to those of control cells without shrna. Geometric mean fluorescence intensities of specifically stained cells (+ primary Ab + secondary APC-conjugated Ab) were divided by the background signal in the absence of primary Ab (second step only). The resulting fold inductions in lytic cells compared to latent cells were expressed as percentages with the expression in -shrna control cells set at 1%. Over 1 such experiments have been performed with shbglf5 (#1 + #9) yielding 6-85% silencing of BGLF5 protein expression. Of note, screening of the shrnas specific for B95.8 BGLF5 by transient transfections in 293T cells (data not shown) had little predictive value for their efficacy during productive EBV infection in B cells, in our hands. 3 Continues on next page 83

84 Chapter 3 AKBM B cells 1 4 Control shrna shrna shcontrol shbglf intracellular surface 1 4 IE E L BZLF1 1 4 BGLF5 Control gp EBV antigen rcd2-gfp rcd2-gfp ghgl staining Ab - anti-igg (latent) + anti-igg (lytic) 84

85 Silencing BGLF5 expression in EBV-producing B cells Figure S3. Silencing BGLF5 expression in AKBM B cells affects progression through the EBV replicative cycle. AKBM B cells stably expressing no shrnas (-shrna), a control Fas-targeting shrna (shcontrol), or a combination of two BGLF5-specific shrnas (shbglf5) were treated with anti-human IgG for 2 hours. Entry into the EBV replicative cycle was monitored by flow cytometry. Abs used for intracellular stainings were directed against the IE transactivator BZLF1 and early (E)-expressed BGLF5. Abs used for surface stainings vizualised the late (L) EBV glycoproteins gp35 and ghgl. Dot plots depict the, early-expressed, rcd2-gfp lytic cycle reporter (horizontal axis) and the indicated viral antigens (vertical axis). Histograms of this experiment are presented in Fig. 1c. (a) AKBM B cells Ab shcontrol Control Surface CD71 HLA I HLA II Intracellular Control BGLF antigen shbglf rcd2-gfp (b) AKBM B cells Surface Ab Control CD58 CD119 CD1 CD45 CD38 CD47 CD19 CD2 shcontrol shbglf antigen (c) AKBM-CD1d cells Surface Ab shcontrol shbglf5 Control CD1d rcd2-gfp Intracellular Control BGLF antigen rcd2-gfp Figure S4. Silencing BGLF5 expression in AKBM B cells affects surface display of cellular antigens. AKBM B cells (a,b) and AKBM-CD1d cells (c) stably expressing a control shrna targeting Fas (shcontrol) or a combination of two shrnas targeting BGLF5 (shbglf5) were treated with anti-human IgG for 2 hours to induce productive EBV infection. Intracellular BGLF5 levels and downregulation of various B cell surface markers were analyzed by flow cytometry and depicted as dot plots. Histograms of this experiment are presented in Fig

86 Chapter 3 References Barnes, P. D. & Grundy, J. E. (1992).Down-regulation of the class I HLA heterodimer and B2-microglobulin on the surface of cells infected with cytomegalovirus. Journal of General Virology 73, Clyde, K. & Glaunsinger, B. A. (211).Deep sequencing reveals direct targets of gammaherpesvirus-induced mrna decay and suggests that multiple mechanisms govern cellular transcript escape. PLoS One 6, e Covarrubias, S., Richner, J. M., Clyde, K., Lee, Y. J. & Glaunsinger, B. A. (29).Host shutoff is a conserved phenotype of gammaherpesvirus infection and is orchestrated exclusively from the cytoplasm. J Virol 83, Delecluse, H. J., Feederle, R., Behrends, U. & Mautner, J. (28).Contribution of viral recombinants to the study of the immune response against the Epstein-Barr virus. Seminars in Cancer Biology 18, El-Guindy, A., Lopez-Giraldez, F., Delecluse, H. J., McKenzie, J. & Miller, G. (214).A Locus Encompassing the Epstein-Barr Virus bglf4 Kinase Regulates Expression of Genes Encoding Viral Structural Proteins. PLoS Pathog 1, e1437. Feederle, R., Bannert, H., Lips, H., Muller-Lantzsch, N. & Delecluse, H. J. (29a).The Epstein-Barr Virus Alkaline Exonuclease BGLF5 Serves Pleiotropic Functions in Virus Replication. Journal of Virology 83, Feederle, R., Mehl-Lautscham, A. M., Bannert, H. & Delecluse, H. J. (29b).The Epstein-Barr virus protein kinase BGLF4 and the exonuclease BGLF5 have opposite effects on the regulation of viral protein production. J Virol 83, Feng, P., Moses, A. & Fruh, K. (213).Evasion of adaptive and innate immune response mechanisms by y-herpesviruses. Current Opinion in Virology 3, Gaglia, M. M., Covarrubias, S., Wong, W. & Glaunsinger, B. A. (212).A common strategy for host RNA degradation by divergent viruses. Journal of Virology 86, Gaudreault, E., Fiola, S., Olivier, M. & Gosselin, J. (27). Epstein-Barr Virus Induces MCP-1 Secretion by Human Monocytes via TLR2. Journal of Virology 81, Glaunsinger, B. & Ganem, D. (24).Lytic KSHV Infection Inhibits Host Gene Expression by Accelerating Global mrna Turnover. Molecular Cell 13, Griffin, B. D., Verweij, M. C. & Wiertz, E. J. H. J. (21). Herpesviruses and immunity: The art of evasion. Veterinary Microbiology 143, Horst, D., Geerdink, R. J., Gram, A. M., Stoppelenburg, A. J. & Ressing, M. E. (212a).Hiding Lipid Presentation: Viral Interference with CD1d-Restricted Invariant Natural Killer T (inkt) Cell Activation. Viruses 4, Horst, D., Burmeister, W. P., Boer, I. G. J., van Leeuwen, D., Buisson, M., Gorbalenya, A. E., Wiertz, E. J. H. J. & Ressing, M. E. (212b).The Bridge in the Epstein-Barr Virus Alkaline Exonuclease Protein BGLF5 Contributes to Shutoff Activity during Productive Infection. Journal of Virology 86, Horst, D., van Leeuwen, D., Croft, N. P., Garstka, M. A., Hislop, A. D., Kremmer, E., Rickinson, A. B., Wiertz, E. J. H. J. & Ressing, M. E. (29).Specific Targeting of the EBV Lytic Phase Protein BNLF2a to the Transporter Associated with Antigen Processing Results in Impairment of HLA Class I-Restricted Antigen Presentation. The Journal of Immunology 182, Iwakiri, D., Zhou, L., Samanta, M., Matsumoto, M., Ebihara, T., Seya, T., Imai, S., Fujieda, M., Kawa, K. & other authors (29).Epstein-Barr virus (EBV)-encoded small RNA is released from EBV-infected cells and activates signaling from toll-like receptor 3. The Journal of Experimental Medicine 26, Iwasaki, A. & Medzhitov, R. (21).Regulation of Adaptive Immunity by the Innate Immune System. Science 327, Jones, T. R., Hanson, L. K., Sun, L., Slater, J. S., Stenberg, R. M. & Campbell, A. E. (1995).Multiple independent loci within the human cytomegalovirus unique short region down-regulate expression of major histocompatibility complex class I heavy chains. Journal of Virology 69, Kinjo, Y., Kitano, N. & Kronenberg, M. (213).The role of invariant natural killer T cells in microbial immunity. J Infect Chemother 19, Koppers-Lalic, D., Rychlowski, M., van Leeuwen, D., Rijsewijk, F. A. M., Ressing, M. E., Neefjes, J. J., Bienkowska-Szewczyk, K. & Wiertz, E. J. H. J. (23). Bovine herpesvirus 1 interferes with TAP-dependent peptide transport and intracellular trafficking of MHC class I molecules in human cells. Arch Virol 148, Kurt-Jones, E. A., Mandell, L., Whitney, C., Padgett, A., Gosselin, K., Newburger, P. E. & Finberg, R. W. (22). Role of Toll-like receptor 2 (TLR2) in neutrophil activation: GM-CSF enhances TLR2 expression and TLR2-mediated interleukin 8 responses in neutrophils. Blood 1, Lebbink, R. J., Lowe, M., Chan, T., Khine, H., Wang, X. & McManus, M. T. (211).Polymerase II Promoter Strength Determines Efficacy of microrna Adapted shrnas. PLoS One 6, e Murphy, J. A., Duerst, R. J., Smith, T. J. & Morrison, L. A. (23).Herpes Simplex Virus Type 2 Virion Host Shutoff Protein Regulates Alpha/Beta Interferon but Not Adaptive Immune Responses during Primary Infection In Vivo. Journal of Virology 77, Ning, S. (211).Innate immune modulation in EBV infection. Herpesviridae 2, 1. Paddison, P. J., Cleary, M., Silva, J. M., Chang, K., Sheth, N., Sachidanandam, R. & Hannon, G. J. (24).Cloning of short hairpin RNAs for gene knockdown in mammalian 86

87 Silencing BGLF5 expression in EBV-producing B cells cells. Nat Meth 1, Paludan, S. R., Bowie, A. G., Horan, K. A. & Fitzgerald, K. A. (213).Recognition of herpesviruses by the innate immune system. Nat Rev Immunol 13, 614. Park, R., El-Guindy, A., Heston, L., Lin, S. F., Yu, K. P., Nagy, M., Borah, S., Delecluse, H. J., Steitz, J. & other authors (214).Nuclear Translocation and Regulation of Intranuclear Distribution of Cytoplasmic Poly(A)-Binding Protein Are Distinct Processes Mediated by Two Epstein Barr Virus Proteins. PLoS One 9, e Pasieka, T. J., Lu, B., Crosby, S. D., Wylie, K. M., Morrison, L. A., Alexander, D. E., Menachery, V. D. & Leib, D. A. (28).Herpes Simplex Virus Virion Host Shutoff Attenuates Establishment of the Antiviral State. Journal of Virology 82, Quinn, L. L., Zuo, J., Abbott, R. J. M., Shannon-Lowe, C., Tierney, R. J., Hislop, A. D. & Rowe, M. (214). Cooperation between Epstein-Barr Virus Immune Evasion Proteins Spreads Protection from CD8+ T Cell Recognition across All Three Phases of the Lytic Cycle. PLoS Pathog 1, e Ressing, M. E., Horst, D. l., Griffin, B. D., Tellam, J., Zuo, J., Khanna, R., Rowe, M. & Wiertz, E. J. H. J. (28).Epstein- Barr virus evasion of CD8+ and CD4+ T cell immunity via concerted actions of multiple gene products. Seminars in Cancer Biology 18, Ressing, M. E., Keating, S. E., van Leeuwen, D., Koppers- Lalic, D., Pappworth, I. Y., Wiertz, E. J. H. J. & Rowe, M. (25).Impaired Transporter Associated with Antigen Processing-Dependent Peptide Transport during Productive EBV Infection. The Journal of Immunology 174, Rickinson, A. B. & Kieff, E. (27). Epstein-Barr Virus. In Field s Virology, 5 edn, pp Edited by D. M. Knipe & P. M. Howley. Philadelphia: Lippincott Williams & Wilkins. Rowe, M., Glaunsinger, B., van, L. D., Zuo, J., Sweetman, D., Ganem, D., Middeldorp, J., Wiertz, E. J. & Ressing, M. E. (27).Host shutoff during productive Epstein-Barr virus infection is mediated by BGLF5 and may contribute to immune evasion. Proc Natl Acad Sci U S A 14, Sheridan, V., Polychronopoulos, L., Dutia, B. M. & Ebrahimi, B. (214).A Shutoff and Exonuclease Mutant of Murine Gammaherpesvirus-68 Yields Infectious Virus and Causes RNA loss in Type I Interferon Receptor Knock-Out cells. Journal of General Virology 95, van Gent, M., Griffin, B. D., Berkhoff, E. G., van, L. D., Boer, I. G., Buisson, M., Hartgers, F. C., Burmeister, W. P., Wiertz, E. J. & other authors (211).EBV lytic-phase protein BGLF5 contributes to TLR9 downregulation during productive infection. J Immunol 186, van Gent, M., Braem, S. G. E., de Jong, A., Delagic, N., Peeters, J. G. C., Boer, I. G. J., Moynagh, P. N., Kremmer, E., Wiertz, E. J. & other authors (214).Epstein-Barr Virus Large Tegument Protein BPLF1 Contributes to Innate Immune Evasion through Interference with Toll-Like Receptor Signaling. PLoS Pathog 1, e1396. York, I. A., Roop, C., Andrews, D. W., Riddell, S. R., Graham, F. L. & Johnson, D. C. (1994).A cytosolic herpes simplex virus protein inhibits antigen presentation to CD8+ T lymphocytes. Cell 77, Zuo, J., Currin, A., Griffin, B. D., Shannon-Lowe, C., Thomas, W. A., Ressing, M. E., Wiertz, E. J. H. J. & Rowe, M. (29).The Epstein-Barr Virus G-Protein-Coupled Receptor Contributes to Immune Evasion by Targeting MHC Class I Molecules for Degradation. PLoS Pathog 5, e1255. Zuo, J., Thomas, W., van Leeuwen, D., Middeldorp, J. M., Wiertz, E. J. H. J., Ressing, M. E. & Rowe, M. (28).The DNase of Gammaherpesviruses Impairs Recognition by Virus-Specific CD8+ T Cells through an Additional Host Shutoff Function. Journal of Virology 82,

88

89 CHAPTER 4 Epstein-Barr virus large tegument protein BPLF1 contributes to innate immune evasion through interference with Toll-like receptor signaling Michiel van Gent Steven G. E. Braem Annemieke de Jong Nezira Delagic Janneke G. C. Peeters Ingrid G. J. Boer Paul N. Moynagh Elisabeth Kremmer Emmanuel J. Wiertz Huib Ovaa Bryan D. Griffin Maaike E. Ressing Published in PLoS Pathogens (214), 1(2):e1396

90 Chapter 4 Abstract Viral infection triggers an early host response through activation of pattern recognition receptors, including Toll-like receptors (TLR). TLR signaling cascades induce production of type I interferons and proinflammatory cytokines involved in establishing an anti-viral state as well as in orchestrating ensuing adaptive immunity. To allow infection, replication, and persistence, (herpes)viruses employ ingenious strategies to evade host immunity. The human gamma-herpesvirus Epstein-Barr virus (EBV) is a large, enveloped DNA virus persistently carried by more than 9% of adults worldwide. It is the causative agent of infectious mononucleosis and is associated with several malignant tumors. EBV activates TLRs, including TLR2, TLR3, and TLR9. Interestingly, both the expression of and signaling by TLRs is attenuated during productive EBV infection. Ubiquitination plays an important role in regulating TLR signaling and is controlled by ubiquitin ligases and deubiquitinases (DUBs). The EBV genome encodes three proteins reported to exert in vitro deubiquitinase activity. Using active site-directed probes, we show that one of these putative DUBs, the conserved herpesvirus large tegument protein BPLF1, acts as a functional DUB in EBV-producing B cells. The BPLF1 enzyme is expressed during the late phase of lytic EBV infection and is incorporated into viral particles. The N-terminal part of the large BPLF1 protein contains the catalytic site for DUB activity and suppresses TLR-mediated activation of NF-κB at, or downstream of, the TRAF6 signaling intermediate. A catalytically inactive mutant of this EBV protein did not reduce NF-κB activation, indicating that DUB activity is essential for attenuating TLR signal transduction. Our combined results show that EBV employs deubiquitination of signaling intermediates in the TLR cascade as a mechanism to counteract innate anti-viral immunity of infected hosts. 9

91 EBV deubiquitinase inhibits TLR signaling Author Summary Epstein-Barr virus (EBV) is a human herpesvirus that persistently infects >9% of adults worldwide. One factor underlying the ability of EBV to establish such widespread and lifelong infections is its capacity to escape elimination by the human immune system. Among the first lines of defense against viral infection is the human Toll-like receptor (TLR) system. These receptors can detect the presence of viruses and initiate an intracellular protein signaling cascade that leads to the expression of immune response genes. The activation status of many proteins in this signaling cascade is regulated by the addition of ubiquitin tags. EBV has previously been reported to encode enzymes, called deubiquitinases (DUBs), which are capable of removing such ubiquitin tags from substrate proteins. In our study, we found that one of these enzymes, BPLF1, functions as an active DUB during EBV production in infected cells before being packaged into newly produced viral particles. Furthermore, our study provides insight into the way in which EBV can subvert the human immune response, as we show that BPLF1 can remove ubiquitin tags from proteins in the TLR signaling cascade. This inhibits TLR signaling and decreases the expression of immune response genes. 4 91

92 Chapter 4 Introduction Herpesviruses are large enveloped DNA viruses that establish widespread persistent infections. The long coevolution has led to a delicate balance between virus and host. For instance, the human gamma-herpesvirus Epstein-Barr virus (EBV) is carried by over 9% of the adult world population, mostly without overt symptoms [1], even though the virus is also causally involved in infectious mononucleosis and a number of malignancies of lymphoid and epithelial origin [2]. Upon primary infection, EBV establishes a lifelong latent infection in memory B cells, characterized by expression of a limited set of viral gene products. For transmission, viral particles are generated during the productive phase of EBV infection, during which the full repertoire of viral lytic genes is expressed. To successfully establish infection and replicate, herpesviruses including EBV must withstand elimination by host defense mechanisms. A first line of host defense is posed by the innate immune system. Innate responses are initiated upon recognition of conserved pathogenassociated molecular patterns (PAMPs) by host pattern-recognition receptors (PRRs). Resulting signaling cascades culminate in the production of type I interferons and proinflammatory cytokines, whose actions limit viral replication by direct anti-viral effects and through tailoring ensuing adaptive immunity [3]. Among the PRRs contributing to anti-viral immunity are membrane-bound Toll-like receptors (TLRs) and cytosolic RIG-I-like receptors (RLRs). The importance of TLRs for controlling herpesvirus infection in vivo is exemplified by an increased susceptibility to MCMV [4 6] or HSV [7,8] in TLR2, TLR3, TLR7, and/or TLR9 knockout mice as well as in mice lacking the TLR-signaling adaptor MyD88. In humans, genetic studies found an increased incidence of herpesvirus encephalitis in individuals with a defect in the TLR3 pathway, whereas susceptibility to pathogens outside the herpesvirus family was not altered [9 13]. TLRs sense PAMPs from a wide variety of pathogens and a number of herpesvirus-derived TLR ligands has now been identified [14]. For EBV, they include virion components that trigger cell-surface displayed TLR2 [15,16] and virus-derived nucleic acids, such as dsrna intermediates and genomic DNA, that are sensed by intracellular TLR3, TLR7, and TLR9 [17 2]. Upon ligand binding, dimerized TLRs interact with Toll-IL-1 receptor (TIR)-domain containing adaptors [21,22]. All TLRs except TLR3 recruit adaptor protein MyD88, which leads to phosphorylation of IL-1 receptor-associated kinase (IRAK)-1 and subsequent activation of tumor necrosis factor-associated factor (TRAF)6. To regulate signal transduction and recruit kinase complexes, TRAF6 catalyzes the formation of lysine(k)63-linked polyubiquitin chains on itself and on NF-κB essential modulator (NEMO, or IκB kinase (IKK)γ). Activation of the IKK complex (comprising IKKα, IKKβ, NEMO) then leads to phosphorylation and 92

93 EBV deubiquitinase inhibits TLR signaling ubiquitination of the inhibitor of NF-κB, IκBα. This K48-linked ubiquitination targets IκBα for proteasomal degradation, thereby allowing NF-κB to translocate into the nucleus and initiate transcription of genes encoding proinflammatory cytokines. TLR signaling can also lead to the production of type I interferons (IFN). Stimulation of TLR7 and TLR9, expressed by plasmacytoid dendritic cells (pdc) for instance, leads to production of type I IFNs through activation of transcription factor IRF7 (signaled through IRAK1/ TRAF6/ IKKα). TLR3 or TLR4 ligation results in type I IFNs production following signaling through the adaptor molecule TIR domain-containing adapter inducing IFN-β (TRIF). In addition to this, TRIF-mediated signaling can induce (delayed) activation of NF-κB by triggering TRAF6. These pathways offer several possibilities for interference by (herpes)viruses. Indeed, we have previously reported that mrna expression levels of several TLRs were reduced in latently and productively EBVinfected B cells [23]. Through post-translational ubiquitin modification, TLR signaling pathways are tightly regulated by cellular ubiquitin ligases and deubiquitinases (DUBs) [24]. The cellular DUBs A2 [25] and CYLD [26] for instance, downmodulate TLR signaling through deubiquitination of TRAF6. Several viruses interfere with the expression of cellular DUBs or encode DUBs themselves [27,28]. Using probes that identify DUBs by interacting with their catalytic sites, conserved DUB activity was found in the large tegument protein of several herpesviruses [29 31]. The EBV homolog of this large tegument protein, BPLF1, consists of 3149 amino acids and its DUB activity is contained in the first N-terminal 24 residues [31]. In the absence of the cloned full-length BPLF1 gene, or BPLF1-specific monoclonal antibodies allowing immunoprecipitation of virally expressed protein, in vitro studies so far have used epitopetagged BPLF1 fragments encoding up to the first 325 most N-terminal amino acids. Besides BPLF1, a bioinformatics screen identified two other EBV proteins with putative DUB activity, BSLF1 and BXLF1 [32]. Since EBV can activate TLR2, TLR3, TLR7, and TLR9, it would be of benefit if EBV were to evade innate immune activation. Here, we examined whether EBV DUB activity was present during productive infection in B cells and whether it could inhibit innate immune signaling. 4 Results The large tegument protein BPLF1 acts as an active deubiquitinase during productive EBV infection To determine if EBV expresses active viral DUBs, we profiled lytically and latently EBVinfected B cells. Productive infection was induced in the latent EBV + B cell line AKBM by treatment with anti-human IgG antibodies (Abs) to crosslink the B cell receptor and thereby reactivate EBV [33]. At 26 hours of induction, up to 52% of cells had entered the lytic cycle, as visualized by rat CD2-GFP reporter expression (Fig. 1a). Post-nuclear cell lysates were 93

94 Chapter 4 incubated with a fluorescently labeled ubiquitin (Ub)-based, active-site directed probe (Ub- VME) [36]. This probe interacts with the active site cysteines of DUBs to form a covalent adduct. The fluorescently-labeled active DUB can then be visualized following SDS-PAGE by in-gel imaging. Using this approach, numerous Ub-VME-reactive species were visible in latently infected AKBM cells, likely representing cellular DUBs (Fig. 1b, lane 1). After 9 and 26 hours of anti-igg treatment, an additional high molecular weight band of >2 kda appeared, suggestive of induced expression of a DUB at late times of productive EBV infection (arrow, lanes 4 and 5, and Fig. S1 for enlarged view). To investigate whether this DUB was virus-encoded, we compared the DUB profiles of EBV + B cells (AKBM) to those of control EBV - B cells (AK31). Both cell lines were either mocktreated or treated with anti-igg for 24 hours and active DUBs were visualized in post-nuclear lysates (Fig. 1c, lanes 1-4). The high molecular weight species reacting with the Ub-VME probe in productively infected AKBM cells (Fig. 1c, lane 4) was not observed in the EBV - AK31 cells, even after treatment with anti-igg (Fig. 1c, lanes 2). This suggests that the active DUB appearing in productively infected B cells is encoded by EBV. Earlier, three EBV proteins were reported to display DUB activity in vitro: BPLF1, BSLF1, and BXLF1 [31,32]. To examine whether any of these corresponded to the active DUB observed in productively infected B cells, the DUB profile of EBV-infected B cells was compared to those of 293T cells that transiently expressed the Flag-tagged BPLF1, BSLF1, or BXLF1 protein (Fig. 1c). By intracellular FACS staining, the BXLF1 and BSLF1 proteins were detected in around 5% of cells (Fig. 1d). In contrast, full-length BPLF1 (FL BPLF1) was detectable in only up to 27% of cells, likely due to the large size of the protein (~315 amino acids). So far, others have used N-terminal fragments of the BPLF1 protein comprising the catalytic site for DUB activity in their studies (varying in length from 25 to 325 aa) [38 41]. Thus, we transfected 293T cells in parallel with a vector encoding the N-terminal 325 amino acids of BPLF1, which yielded higher transfection efficiencies (up to 53% positive cells, Fig. 1d) than full-length BPLF1. Ub-VME-labeled bands corresponding to the BXLF1 or BSLF1 proteins were not observed by DUB profiling of transfected 293T cells (Fig. 1c, lanes 7 and 8), indicating that these display very little or no DUB activity. Both EBV proteins contain nuclear localization signals and (a proportion of) these proteins are present in the nucleus. Although intracellular FACS staining (Fig. 1d) demonstrated expression of Flag-BSLF1 as well as Flag-BXLF1 proteins in transfected cells, they could have been absent from the post-nuclear cell lysates used for DUB profiling. However, Flag-specific immunoblots of the gels used for fluorescence imaging showed that both BSLF1 and BXLF1 were present in the post-nuclear lysates (Fig. 1e, lanes 2 and 3). In our hands neither BSLF1 nor BXLF1 exerted in vitro cleavage of fluorogenic Ub-AMC substrates or reduction in Ub-protein adducts in 293T cells expressing HA-ubiquitin (data not shown). Furthermore, we did not find specific Ub-VME-reactive proteins corresponding to the sizes of these two proteins in productively EBV-infected AKBM cells (Fig. 1c, lane 4). These combined 94

95 EBV deubiquitinase inhibits TLR signaling B C Fluorescence, + Ub-VME EBVAK31 αigg - + EBV+ AKBM cells + αigg (hrs) hrs 13% 3 hrs Fluorescence, + Ub-VME kda kda EBV+ AKBM - + C AKBM cells % 293T cells FL BP LF 1 BX LF 1 BS LF 1 BP LF 1 A Transf. FL BPLF1 + Ub 97 3% Lane Immunoblot anti-bplf1 (1F2) 293T cells EBVEBV+ AK31 AKBM αigg BP C kda BS LF 1 BP LF 1 LF 1 LF C 27% 2 F Immunoblot anti-flag 293T cells Transf. FL BPLF1 Lane C 4 LF E 293T cells % 3 BS D 2 BX rcd2-gfp BPLF1 + Ub 1 52% 26 hrs * 39 LF 9 hrs BX 36% FL 6 hrs Transf. kda FL BPLF1 + Ub % BXLF1 BSLF % BSLF1 BPLF1 53% BXLF Lane 1 * * BPLF1 + Ub Flag-tag (FITC) Lane Figure 1. EBV BPLF1 is a DUB active during productive infection. EBV+ AKBM BL cells were treated with anti-human IgG (αigg) to induce viral replication. (a) At the indicated times post induction, percentages of productively EBV-infected AKBM cells were determined by flow cytometric analysis of induced ratcd2-gfp reporter expression. (b) In post-nuclear AKBM cell lysates, active DUBs were labeled with a fluorescent Ub-VME probe, resolved by SDS-PAGE, and visualized by in-gel fluorescence imaging. The arrow indicates a band appearing in AKBM cells after 9 hours of lytic cycle induction (lanes 4 and 5). (c) DUB profiles of EBV- AK31-rCD2-GFP cells (AK31, lanes 1 and 2), EBV+ AKBM cells (lanes 3 and 4), and transfected 293T cells (lanes 5-9). In parallel with the EBV+ AKBM cells, EBV- control AK31 cells were treated with αigg for 24 hours; this resulted in productive infection in 56% of AKBM cells; AK31 cells included a population of 45% that expressed ratcd2-gfp from a constitutive promoter irrespective of αigg treatment (data not shown). For comparison, 293T cells were transfected with constructs encoding three (putative) EBV DUBs: BPFL1, BXLF1, BSLF1, or the 325 aa N-terminal part of BPLF1. The asterisk marks a smaller fragment observed upon transfection of 293T cells with full-length or the N-terminal part of BPLF1 (lanes 6 and 9). Left and right panels are parts of one gel displayed at different exposures. (d) Sixteen hours post-transfection, percentages of 293T cells expressing BPLF1, BXLF1, BSLF1 or the N-terminal fragment of BPLF1 were determined by intracellular FACS staining for the Flag-tag. (e) Immunoblot of part of the gel in c probed with an anti-flag Ab to detect transfected BXLF1 and BSLF1 (sequential staining following 1F2, see in f). (f) Immunoblot of the gel in c probed with the BPLF1-specific mouse monoclonal Ab 1F2. Both panels are part of one gel presented at different exposures. 95 4

96 Chapter 4 results imply that BSLF1 and BXLF1 do not behave as active DUBs when expressed in cells. In contrast, transfection of full-length BPLF1 in 293T cells resulted in expression of an active DUB visible as a specific Ub-VME reactive band migrating over 2 kda (Fig. 1c, lane 6). The size of this band corresponded to that observed in lytically infected AKBM cells (lane 4), suggesting that the large DUB expressed during productive EBV infection is BPLF1. Upon cellular expression of full-length BPLF1, an additional smaller active DUB fragment was visible around 42 kda (asterisk, lane 6). Interestingly, transfection of the N-terminal 325 aa of BPLF1 also yielded this smaller fragment, in addition to the expected band at 46 kda (corresponding to 36 kda of BPLF1 covalently linked to the Ub-VME probe of 1 kda, lane 9). The active fragment of ~42 kda (corresponding to ~28 aa) could have arisen upon alternative expression or processing of BPLF1. Others have also reported the presence of a smaller fragment upon expression of an N-terminal part of BPLF1 [38,42]; here, we show for the first time that a similar fragment is generated upon expression in 293T cells of full-length BPLF1. To enable the study of BPLF1 protein expression during natural EBV infection (i.e. without (Flag-)epitope tag), we generated monoclonal Abs against two peptides in the N-terminal domain of this protein (based on the B95.8 EBV strain, see Fig. S2a). In immunoblots of transfected 293T cells, mab 1F2 (epitope: BPLF1 residues 78-94) reacted with N-terminal fragments of BPLF1 of both the B95.8 and Akata EBV strains (Fig. S2b, Fig. 1f lane 9, and Fig. S3b lane 1). Importantly, this BPLF1-specific Ab visualized the full-length protein and shorter fragment in transfected 293T cells and also the active DUB expressed in productively infected AKBM cells (Fig. 1f, lanes 6 and 4, respectively, and Fig. S3), indicative of BPLF1 being expressed during the EBV lytic cycle. Thus, for two out of the three putative EBV DUBs, BSLF1 and BXLF1, we did not find indications for DUB activity in cells. The large tegument protein BPLF1, on the other hand, appeared to interact with a functional probe both in transfected 293T cells and in EBV-producing B cells, implying that it acts as an active DUB during late times of productive EBV infection. Since both full-length BPLF1 and the N-terminal part yielded the same active smaller fragment, a construct expressing the N-terminal 325 aa of BPLF1 was used in further assays. BPLF1 interferes with TLR-mediated NF-κB activation In view of the critical role of (de)ubiquitination in regulating TLR signal transduction, we next addressed the question: does EBV exploit virus-encoded DUBs to evade innate immune activation signaled through TLRs? To this end, we monitored NF-κB activity in 293T cells transfected with an NF-κB-responsive firefly luciferase reporter. Co-transfection of MyD88, the adaptor molecule used by all TLRs except TLR3, resulted in robust activation of NF-κB (~5-fold induction of luciferase activity compared to control cells; Fig. 2a). Upon additional introduction of each of the three EBV proteins in these cells, expression of BPLF1 resulted in a dose-dependent inhibition of NF-κB activation. BXLF1 and BSLF1 did not substantially 96

97 EBV deubiquitinase inhibits TLR signaling A NF-κB activation (fold induction) BPLF1 293T cells BXLF1 BSLF1 A2 C NF-κB activation (%) TLR3 cells NF-κB reporter IFNβ reporter C C BPLF1 A2 IFNβ promoter activation (%) C C BPLF1 A2 α-ha α-flag α-β actin MyD88 BPLF1 BXLF1 MyD88 Actin + poly(i:c) + poly(i:c) B NF-κB activation (%) C C BPLF1 BXLF1 BSLF1 A2 C BPLF1 293-TLR2/CD14 cells A2 BXLF1 BSLF1 + MALP-2 + EBV D secreted IL-8 (ng/ml) C C BXLF1 BSLF1 A2 BPLF1 293-TLR2/CD14 cells C BXLF1 BSLF1 A2 BPLF1 + MALP-2 + EBV 4 Figure 2. BPLF1 interferes with TLR-mediated NF-κB activation. 293T cells were transiently transfected with a firefly luciferase reporter construct responsive to either NF-κB activation or IFN-β promoter activity, an HSV TK-promoter driven Renilla luciferase plasmid (for normalization), and plasmids encoding HA-BPLF1, Flag-BXLF1, Flag-BSLF1 or cellular control DUB A2. (a) An NF-κB responsive firefly luciferase reporter and 4 or 8 ng of vectors encoding the EBV or control proteins were cotransfected in 293T cells and the TLR signaling cascade was initiated by expressing adaptor protein MyD88. Sixteen hours posttransfection, cells were lysed and luciferase activity was determined. In addition, protein expression was analyzed in post-nuclear cell lysates by immunoblot using anti-ha and anti-flag Abs; β-actin served as loading control. (b) 293-TLR2/CD14 cells were cotransfected with an NF-κB responsive reporter and 32 ng of plasmids encoding the indicated proteins. Starting at 16 hours post-transfection, cells were stimulated with 1 ng/ml MALP-2 or 6.5 x 1 6 EBV particles/ml for 7 hours, after which they were lysed and luciferase activity was measured. (c) Similarly, 293-TLR3 cells were cotransfected with 16 ng of the indicated genes together with an NF-κB responsive reporter (left panel) or a reporter under control of the IFN-β promoter (right panel). Starting 16 hours post-transfection, cells were stimulated for 7 hours with 1 μg/ml poly(i:c), after which they were lysed and luciferase activity was determined. Data are presented as percentages firefly luciferase activity relative to stimulated control samples and normalized for transfection efficiency using Renilla luciferase values, mean ± SD. (d) 293-TLR2/CD14 cells were transfected with empty control vector or plasmids encoding EBV BPLF1 or cellular A2. Starting at 24 hours posttransfection, cells were stimulated with 1 μg/ml MALP2 or 6.5 x 1 6 EBV particles/ml for 8 hours. IL-8 secretion in the culture supernatants was determined by ELISA. 97

98 Chapter 4 affect NF-κB activation following MyD88 signaling (Fig. 2a), in line with their undetectable DUB activity in cells. Immunoblots revealed comparable amounts of MyD88 protein in each sample. The BPLF1 and BXLF1 proteins were clearly detectable in the post-nuclear cell lysates. BSLF1 protein levels, however, appeared much lower (data not shown), representing low expression and/or a predominantly nuclear localization of this EBV protein. Overexpression of the cellular DUB A2 in these cells, as a control known to inhibit TLR signaling, was also difficult to detect by immunoblot (data not shown), yet strongly counteracted NF-κB activation (Fig. 2a). As a more physiological alternative to initiating TLR signaling by transfection of MyD88, we stimulated 293 cells that stably expressed TLR2 and CD14 (293-TLR2/CD14) or TLR3 (293- TLR3) with their cognate ligands. TLR2-mediated NF-κB activation upon MALP-2 stimulation was inhibited by BPLF1 and A2, but not by BXLF1 or BSLF1 (Fig. 2b). In contrast to the MyD88-dependent TLR2, TLR3 activates NF-κB through a different adaptor molecule, TRIF. This pathway, stimulated by poly(i:c) treatment, was also repressed by EBV BPLF1 as well as by the control DUB A2 (Fig. 2c, left panel). In addition, TLR3 stimulation induces type I IFN production. Activation of an IFN-β-reporter was also strongly reduced upon expression of BPLF1 and control A2 in poly(i:c)-stimulated 293-TLR3 cells (Fig. 2c, right panel). Finally, we tested what the consequences would be of EBV DUB expression for TLR activation of endogenous proinflammatory cytokine production (Fig. 2d). Stimulation of 293-TLR2/ CD14 cells with MALP-2 ligand induced the production of IL-8 and this was counteracted by expression of BPLF1. As EBV has also been reported to activate TLR2 [16], we performed the same experiments using viral particles rather than MALP-2 to stimulate TLR2 signaling. In response to EBV particles, both NF-κB activation and IL-8 production occurred to levels comparable to those observed for MALP-2 stimulation in 293-TLR2/CD14 cells (Fig. 2b and d). No NF-κB activation or IL-8 production was observed upon stimulation of TLR2-negative 293-CD14 cells (not shown) indicating that the EBV-induced response was mediated by TLR2. Moreover, BPLF1 reduced NF-κB-luciferase and IL-8 production also upon TLR stimulation by EBV; BXLF1 and BSLF1 had no influence in this context (Fig. 2b and d). These results indicate that EBV BPLF1 inhibits proinflammatory cytokine production by cells in which the innate immune system has been activated by viral particles. Taken together, we found that one of the (putative) EBV DUBs, the large tegument protein BPLF1, counteracts both MyD88- and TRIF-dependent TLR signal transduction pathways. The ability of BPLF1 to modulate TLR signaling correlates with its deubiquitinase activity To assess whether BPLF1 relies on deubiquitination to interfere with TLR signaling, we used three previously reported BPLF1 mutants (C61A, D86/9R, and D86/9G). Replacing the catalytic cysteine residue 61 by an alanine in BPLF1 C61A has been shown to abolish enzymatic activity [31]. Besides acting as a DUB, BPLF1 is capable of hydrolyzing Nedd8 conjugates and 98

99 EBV deubiquitinase inhibits TLR signaling affects the major cellular neddylated proteins, the cullins [38,39]. Two mutants, BPLF1 D86/9R and BPLF1 D86/9G, contain altered charged residues at the protein surface that almost completely prevent interaction with the cullin proteins and, therefore, no longer deneddylate them. While the catalytic site C61 residue is unaltered, enzymatic activities of these two mutants are differentially affected, probably due to conformational changes in the protein [39]. Transient transfection of these BPLF1 variants in 293T cells resulted in similar protein expression levels (Fig. 3a, FACS; Fig. 3b, immunoblot). In agreement with earlier reports, DUB labeling with the fluorescent Ub-VME probe was abrogated by the C61A catalytic site mutation, whereas BPLF1 D86/9R and BPLF1 D86/9G maintained reactivity (Fig. 3b middle and lower panel, in gel-fluorescence and subsequent immunoblot, respectively). The immunoblot depicted in figure 3b (lower panel) clearly shows the larger and smaller BPLF1 fragments (lane 3) and their 1 kda shift upon covalent attachment of the active site probe (lanes 2, 4, and 5). Differences among the BPLF1 variants became more apparent in a cellular assay for DUB activity, in which ubiquitinated proteins in 293T cells were visualized through transfection with HA-tagged ubiquitin (Fig. 3c). In this assay, potent deubiquitination was mediated by wild-type (wt) BPLF1 (lane 3) and, to a slightly lesser extent, by BPLF1 D86/9G (lane 6); BPLF1 D86/9R had largely lost its capacity to remove HA-ubiquitin from cellular proteins (lane 5), as did the catalytic mutant BPLF1 C61A (lane 4). We next examined the effects of expressing wt BPLF1 or the three mutants on TLR signaling. To this end, 293T cells were transfected to transiently (co-)express the TLR2 protein (Fig. 3d). NF-κB activation following MALP-2 treatment of these cells was inhibited by wt BPLF1 and, 4 to a lesser extent, by BPLF1 D86/9G. In contrast, both the catalytically inactive mutant BPLF1 C61A as well as BPLF1 D86/9R did not reduce NF-κB activation. As TLR interference by these BPLF1 variants (Fig. 3d) correlates with their reactivity towards intracellular Ub-protein adducts (Fig. 3c), we conclude that the ability of BPLF1 to downmodulate TLR signaling depends on its intracellular DUB activity. BPLF1 acts at multiple levels in the TLR signal transduction cascade Regulation of TLR responses involves K48- and K63-linked ubiquitination of various signaling intermediates. To delineate where EBV BPLF1 acts to inhibit TLR signaling, we performed the following experiments. Transcription of type I IFNs and proinflammatory cytokines is induced by NF-κB after its translocation into the nucleus upon release from the cytosolic inhibitor IκBα. Downstream of TLR stimulation, IκBα is targeted for proteasomal degradation by K48-linked ubiquitination. To investigate whether BPLF1 interferes with this process, we assessed IκBα levels in 293- TLR2/CD14 cells (Fig. 4a). In response to the TLR2 ligand MALP-2, IκBα degradation became apparent in TLR2-expressing cells after 3 minutes of stimulation and this was counteracted by the cellular DUB A2. Cellular expression of the EBV DUB BPLF1 likewise prohibited 99

100 Chapter 4 A % C B BPLF1 33% 45% 44% 38% wt C61A D86/9R D86/9G IB: anti-flag +Ub-VME Fluorescence +Ub-VME IB: anti-bplf1 kda Lane C wt C61A D86/9R D86/9G BPLF1 * BPLF1 * + Ub-VME - Ub-VME BPLF1 Flag-tag (FITC) C kda C C wt C61A D86/9R D86/9G BPLF HA-Ub 37 Lane IB: anti-ha Ubiquitinated proteins IB: anti-flag BPLF1 IB: anti-actin β actin Ub-HA staining intensity (%) C C wt C61A D86/9R + HA-Ub D86/9G D NF-κB activation (%) C 293T + TLR2 C wt C61A D86/9R D86/9G + MALP-2 Figure 3. BPLF1-mediated modulation of TLR signaling correlates with deubiquitinase activity. 293T cells were transiently transfected with an empty vector control or plasmids encoding Flag-tagged BPLF1 or mutants thereof (BPLF1 C61A, BPLF1 R86/R9, or BPLF1 G86/G9 ). (a) Sixteen hours post-transfection, the percentage of cells expressing BPLF1 was determined by intracellular flow cytometric analysis using an anti-flag Ab. (b, upper panel) Post-nuclear cell lysates were prepared and immunoblotted using an anti-flag Ab. The asterisk marks a smaller active fragment observed upon expression of BPLF1. (b, middle panel) Active DUBs in these cell lysates were labeled using the fluorescent Ub-VME probe (causing a concomitant size increase of 1 kda) and visualized by in-gel imaging. (b, lower panel) Immunoblot of the gel in c probed with BPLF1-specific Ab 2E5. Unlabeled BPLF1 migrates at 36 kda and 32 kda, while the upper two bands of 46 kda and 42 kda represent probe-bound BPLF1. (c) 293T cells were cotransfected with HA-ubiquitin and BPLF1 variants. After 4 hours, total lysates were prepared and immunoblotted for HA-Ub-protein adducts using an anti-ha Ab (left panel; right panel shows the quantification of average staining intensities from four independent experiments). Expression levels of BPLF1 variants were assessed using an anti-flag Ab; β-actin served as loading control. (d) 293T cells were cotransfected with vectors encoding TLR2, an NF-κB responsive firefly luciferase reporter, HSV TK-driven Renilla luciferase for normalization, and BPLF1 variants (16 ng). Following treatment with 1 ng/ml MALP-2 for 7 hours, cells were lysed and luciferase activity was measured. The data are presented as percentage firefly luciferase activity relative to stimulated control sample and normalized for transfection efficiency using Renilla luciferase values, mean ± SD. 1

101 EBV deubiquitinase inhibits TLR signaling IκBα degradation, whereas the catalytically inactive mutant BPLF1 C61A did not. Thus, BPLF1 interferes with TLR-mediated NF-κB signaling at, or upstream of, the level of IκBα degradation in a manner dependent on its DUB activity. Several signaling intermediates upstream of IκBα are regulated through K63-linked ubiquitination (Fig. 4b). To dissect these steps, we stimulated the TLR signaling pathway at different levels by transfection of 293T cells with the activator proteins MyD88, IRAK1, TRAF6, or IKKα, and monitored ensuing NF-κB activation in luciferase assays (Fig. 4c). Feasibility of this approach is illustrated by two controls. Firstly, IκBα blocks NF-κB activation downstream of IKKα, and IκBα expression therefore inhibited responses induced by MyD88 down to IKKα. Secondly, A2 is known to deubiquitinate TRAF6; A2 expression inhibited NF-κB activation mediated by MyD88, IRAK1, and TRAF6, but not IKKα. EBV BPLF1 reduced signaling induced by expression of all four signaling intermediates. This inhibitory effect occurred through deubiquitination, since the catalytic mutant BPLF1 C61A did not alter NF-κB activation (Fig. 4c). These data point to the EBV DUB targeting ubiquitinated signaling molecules at or upstream of the IKKα/IKKβ/NEMO-complex. Confocal microscopy was employed to study cellular colocalization of the EBV-encoded DUB and three ubiquitinated intermediates in the TLR signaling pathway, a prerequisite for their direct deubiquitination by BPLF1. 293T cells transfected with BPLF1 and either TRAF6, NEMO, or IκBα were fixed and permeabilized and both EBV and signaling proteins were visualized with specific antibodies. Colocalization in the cytoplasm was observed for BPLF1 with TRAF6 and with NEMO (Fig. 5a). Transfection of IκBα results in nuclear localization of this signaling intermediate [43] and, therefore, colocalization with BPLF1 was detected in the nucleus (Fig. 5a). We next determined whether cytoplasmic colocalization with BPLF1 affected K63-linked polyubiquitination of TRAF6 or NEMO, by comparing cells expressing TRAF6 or NEMO in the absence or presence of DUBs. Immunoprecipitation from post-nuclear lysates of 293T cells co-expressing HA-ubiquitin revealed a smear of polyubiquitinated TRAF6 (Fig. 5b, upper panel lane 1) and polyubiquitinated NEMO (Fig. 5b, lower panel lane 1). Co-expression of wt BPLF1 in these cells caused a marked reduction in the amounts of ubiquitinated TRAF6 and 4 NEMO (lanes 2), which was not induced by the catalytically inactive DUB mutant BPLF1 C61A (lanes 3). As a positive control, A2 also reduced poly-ubiquitination of TRAF6 (upper panel lane 4), but not of NEMO (lower panel lane 4). These data strongly indicate that cytoplasmic BPLF1 can remove K63-linked ubiquitin moieties from TRAF6 and NEMO. Finally, to directly assess if BPLF1 can enzymatically act on both K63-linkages as well as K48-linkages present on ubiquitinated TRAF6 and NEMO, versus IκBa, we compared the in vitro deubiquitinase activity of isolated (in)active DUBs towards these targets (Fig. 5c). In line with the results observed upon their cellular coexpression (Fig. 5b), in vitro incubation of ubiquitinated TRAF6 or NEMO with wt BPLF1 resulted in removal of polyubiquitin chains, which did not occur upon incubation with the inactive mutant BPLF1 C61A (Fig. 5c, 11

102 Chapter 4 A MALP-2 (mins) αikba 293-TLR2/CD C wt C6 1A TLR A2 MyD88 anti-flag αtfr MALP-2 (mins) αikba B BPLF1 293-CD TLR2/CD14 cells transfected with: C BPLF1 wt IRAK1 TRAF αtfr C61A MALP-2 (mins) αikba IKKα IKKβ NEMO Ub A IκBα αtfr Lane 4 Ub Ub NF-κB C BP C LF C6 1 1A A2 IκB α MyD IRAK TRAF IKKα C BP C LF C6 1 1A A2 IκB α NF-κB activation (%) 5 1 C BP C LF C6 1 1A A2 IκB α 75 NF-κB activation (%) 1 C BP C LF C6 1 1A A2 IκB α NF-κB activation (%) NF-κB activation (%) C Figure 4. BPLF1 interferes with TLR signal transduction at multiple levels. (a) 293-TLR2/CD14 cells and control 293-CD14 cells were transiently transfected with plasmids encoding Flagtagged BPLF1, catalytically inactive mutant BPLF1C61A, or cellular control DUB A2. At 4 hours post-transfection, cells were treated with 1 ng/ml MALP-2 for indicated time periods and IκBα levels in post-nuclear cell extracts were determined by immunoblotting with an anti-iκbα Ab. Expression of transfected Flag-tagged proteins was assessed by immunoblotting with an anti-flag Ab; TfR served as loading control. (b) Schematic representation of the signaling cascade that induces NF-κB following TLR activation. Ub in green is K63-linked, Ub in purple is K48linked. (c) 293T cells were cotransfected with an NF-κB responsive firefly luciferase reporter, HSV TK promoterdriven Renilla luciferase for normalization, and plasmids expressing BPLF1, BPLFC61A, or controls IκBα and A2 (13.5 ng). TLR signaling was induced by expression of activator proteins MyD88, IRAK1, TRAF6, or IKKα. After 16 hours of transfection, luciferase activity was measured. Results are depicted as percentage firefly luciferase activity normalized using Renilla luciferase values relative to stimulated control sample, mean ± SD. left and middle panels). The cellular DUB A2 only affected ubiquitination of TRAF6, but not of NEMO. Since K48-linked (as opposed to K63-linked) polyubiquitination results in proteasomal degradation, IκBα transfected cells were additionally stimulated with TNFα and treated with a proteasome inhibitor to obtain sufficient amounts of ubiquitinated signaling protein for evaluation in an in vitro DUB assay. Polyubiquitinated IκBα was also targeted by BPLF1, but not by the inactive mutant or A2 (Fig. 5c, right panel), showing that the EBV DUB can interfere with K48-ubiquitin-regulated proteasomal degradation of IκBα. Taken together, our results indicate that BPLF1 inhibits NF-κB activation by targeting multiple K48- and K63-ubiquitin-regulated steps along the TLR signaling pathway. 12

103 EBV deubiquitinase inhibits TLR signaling BPLF1 is expressed during the late phase of productive EBV infection and is incorporated into viral particles Conflicting information has been published on the timing of expression of the EBV DUB during productive infection, with BPLF1 being classified either as a late viral gene based on promoter consensus sequences [44], or as an early gene based on the kinetics of mrna expression [38]. Our new BPLF1-specific monoclonal Ab provides the opportunity to address this issue by determining natural BPLF1 expression kinetics at the protein level. We examined expression of EBV proteins in post-nuclear lysates of AKBM cells at various times after induction of the lytic cycle (Fig. 6a and b). The immediate early viral protein BZLF1 was readily detected at 4 hours post induction and onwards, followed by the early BGLF5 protein, and the late structural glycoprotein gp42 at 8-12 hours post induction. The BPLF1 protein appeared in productively infected AKBM cells after 8-12 hours, coincident with the timing of gp42 expression (Fig. 6b). To confirm the late kinetics of BPLF1 expression, phosphonoacetic acid (PAA) was added during anti-igg treatment of EBV + AKBM cells and the EBV - AK31 control cells (Fig. 6c). PAA is an inhibitor of viral DNA replication and prevents late gene expression in productively infected B cells [45]. While BZLF1 and BGLF5 proteins were expressed in lytic AKBM cells irrespective of PAA treatment (lanes 6 and 7), the inhibitor blocked synthesis of both BPLF1 and the control late EBV protein gp42 (lane 7). Finally, we analyzed purified EBV particles (strain B95.8): in addition to the envelope protein gp42, we also found the full-length BPLF1 protein to be present (Fig. 6c, lane 1). Taken together, these results show that the EBV DUB BPLF1 is expressed during the late phase of productive infection and is incorporated into viral particles, providing various opportunities for interference with TLR signaling. 4 Discussion In this study, we have shown that the large tegument protein BPLF1 acts as an active DUB during productive EBV infection. By removal of K63- and/or K48-linked ubiquitin chains from signaling intermediates, such as TRAF6, NEMO, and/or IκBα, BPLF1 inhibits TLR signaling through both MyD88- and TRIF-dependent pathways. This leads to reduced NF-κB activation and proinflammatory cytokine production in response to EBV. The first herpesvirus-encoded DUB was identified by covalent binding of a ubiquitin-based active site-directed probe to the large tegument protein expressed in HSV-1-infected cells [3]. The catalytic site of this DUB appeared to be conserved in the large tegument protein of all herpesviruses [31]. Based on the structure of the MCMV homologue, M48, the herpesvirus DUBs form a unique class of cysteine proteases that do not share homology with cellular deubiquitinating enzymes, except for the catalytic site triad [3,46]. Both in vitro and in vivo studies support important roles for the conserved tegument DUBs during the viral life cycle. 13

104 Chapter 4 A TRAF6 NEMO TO-PRO-3 Sign. interm. BPLF1 Merge Merge Staining TRAF6 (α-flag) BPLF1 (2E5) TRAF6 + BPLF1 NEMO BPLF1 (α-flag) NEMO + BPLF1 B IP: TRAF6 IB: HA IP: TRAF6 IB: TRAF6 IB: Flag kda Flag-TRAF HA-Ub Lane BPLF1 kda C WT C WT BPLF1 C61A A2 C61A A2 Ub(n)- TRAF6 TRAF6 A2 TRAF6 BPLF His-NEMO HA-Ub C α-flag IκBα kda α-ha 1 75 C C WT BPLF1 C61A A Flag-TRAF HA-Ub kda 25 TRAF6 1 α-ha -Ubn 75 5 C C WT IκBα BPLF1 (α-flag) IκBα + BPLF1 BPLF1 C61A A2 α-his 5 A2 75 TRAF α-flag 5 37 BPLF1 37 Lane Lane IP: His IB: HA IP: His IB: His IB: Flag Lane C C WT BPLF1 Ub(n)- NEMO Ub1-NEMO Ub1-NEMO NEMO A2 C61A A2 BPLF His-NEMO IκBα HA-Ub HA-Ub kda 25 NEMO IκBα 1 -Ubn α-ha 75 -Ubn Ub1-NEMO 5 Ub1-NEMO NEMO α-iκbα 37 IκBα A2 A2 75 α-flag 5 BPLF1 37 BPLF1 Lane Figure 5. EBV DUB BPLF1 can target K63- and K48-ubiquitinated signaling intermediates. (a) Fluorescence micrographs of 293T cells that were cotransfected with plasmids encoding Flag-TRAF6 and HA- BPLF1 (upper panel), NEMO and Flag-BPLF1 (middle panel), or IκBα and Flag-BPLF1 (lower panel). Sixteen hours post-transfection, signaling intermediates (column 2) and BPLF1 (column 3) were labeled using anti-flag (red; for TRAF6 in upper panel; for BPLF1 in middle and lower panels), anti-bplf1 (green), anti-nemo (green), and anti- IκBα (green) Abs as indicated. Nuclei were visualized using TO-PRO-3 (blue, column 1). Column 4 and 5 show merged pictures from red and green channels without and with nuclear stain (blue), respectively. Scale bar 5 μm. (b) 293T cells were cotransfected with plasmids encoding either Flag-TRAF6 (upper panel) or His-NEMO (lower panel) together with HA-ubiquitin and BPLF1, BPLF1 C61A, or A2. Expression of Flag-tagged proteins was assessed by immunoblotting (IB) with an anti-flag Ab. TRAF6 and NEMO were precipitated (IP) from post-nuclear lysates 14

105 EBV deubiquitinase inhibits TLR signaling For example, an HCMV UL48 mutant virus yielded reduced viral titers in cultured cells [47]. A mutant of HSV, from which the homologous tegument gene UL36 had been completely deleted, did not assemble virions (late in infection) and failed to enter the nucleus (at early stages of infection) [48]. MHV-68 expressing a catalytically inactive DUB encoded by ORF64 had reduced in vitro replication capacity and was cleared more efficiently in vivo [49]. For EBV, knockout or knockdown of BPLF1 expression in lytically infected 293T or B cells, respectively, resulted in reduced virus production [38,4,5]. So far, one viral substrate of BPLF1 has been found: deubiquitination of the subunit BaRF1 reduces activity of the EBV ribonucleotide reductase complex [4]. Among the reported cellular targets of BPLF1 is PCNA [41], deubiquitination of which prevents recruitment of DNA damage response elements. The cullins form cellular substrates for deneddylation by BPLF1. As a consequence of active BPLF1 expression, the cullin-ring ligase neddylation cycle is interrupted leading to an S-phase cell cycle arrest that favors EBV replication [38,39]. Although the N-terminal fragment of the homologous large tegument protein of EBV has been reported to exert DUB activity [31,38,4,51], earlier attempts were unsuccessful in finding the active BPLF1 protein in EBV-infected B cells [52]. Combining our EBV lytic cycle system with a sensitive fluorescently labeled Ub-VME probe and new BPLF1-specific monoclonal Abs has now revealed an active DUB protein corresponding to full-length BPLF1 in EBV-producing B cells (Fig. 1). This complements the finding by Whitehurst et al. [41] of endogenous BPLF1 expression in lytically induced 293T-BAC EBV cells. While our manuscript was under revision, Gastaldello et al. [42] reported detection of BPLF1 as an active deneddylase as well as DUB in a related cell line, Akata-BX1, thereby confirming the reproducibility of our findings. BPLF1 is expressed as a late lytic cycle gene product and is incorporated into viral particles (Fig. 6). BPLF1 s presence in virions is also supported by detection of this high molecular weight protein (indicated as large tegument protein, LTP) among EBVs structural proteins [53]. Two other reported putative EBV DUBs, BXLF1 and BSLF1, were not detectable as active DUBs in EBV-producing B cells or in transfected cells (Fig. 1). Cellular expression of full-length as well as a 325 aa N-terminal domain of BPLF1 yielded an additional smaller fragment, as also observed by others [38,42]. Based on its size this fragment corresponds to residues 1- ~28 of BPLF1, including the active site cysteine, and was shown to exert DUB activity (Figs 1, 3, 5). Our preliminary data indicate that full-length BPLF1 disappeared in time (at ~48 hours and detected using an anti-traf6 or anti-his Ab, respectively. HA-tagged ubiquitin adducts were visualized on immunoblot using anti-ha Abs. (c) 293T cells were cotransfected with plasmids encoding HA-ubiquitin together with Flag-TRAF6 (left panel), His-NEMO (middle panel), or IκBα (right panel). IκBα-transfected cells were additionally treated with TNFα and the proteasome inhibitor MG132 for 2 hours before lysis. Twenty-four hours post-transfection, TRAF6, NEMO, and IκBα were precipitated from post-nuclear cell lysates and incubated in vitro with separately purified Flag-BPLF1, Flag-BPLF1C61A, or Flag-A2 proteins for 4 hours at 37 C. Flag-tagged DUBs were detected by immunoblotting with anti-flag Ab and signaling intermediates were visualized using anti-flag (TRAF6), anti-his (NEMO), or anti-iκbα Abs. HA-ubiquitin adducts were detected using an anti-ha Ab. 15 4

106 Chapter 4 in lytic AKBM cells, Fig. S3c; from ~72 hours in 293T cell transfected with FL BPLF1, most clearly visible as loss from the cytosol of the C-terminal HA-tag, Fig. S3b) and a smaller fragment appeared (reactive with the DUB probe and Abs specific for epitopes spanning BPLF1 aa 2-17 (peptide A) and (peptide B); Figs S2 and S3). This likely occurs due to processing of the full-length protein, as it coincided with appearance of N-terminal fragments of BPLF1 migrating around 32 and 25 kda (detectable with anti-bplf1 peptide A Ab, Fig. S3b; and with anti-flag Ab, not shown). Shorter fragments of BPLF1 of 32 kda would lack nuclear localization sequences (aa ) and might localize preferably to the cytoplasm. The ~32 kda fragment was indeed observed in cytoplasmic (and nuclear) fractions, yet the ~25 kda form was restricted to the nucleus (Fig. S3b). This is in accordance with a very recent publication, where a nuclear ~25 kda fragment of BPLF1 was found to result from cleavage of the viral protein by caspase 1; the resulting small molecular size ( 4 kda) could allow diffusion through the nuclear pore, which would thus facilitate deneddylase activity towards the cullin substrates in the nucleus [42]. We here propose that the ~32 kda fragment could exert DUB activity towards substrates that reside in the cytoplasm. Intracellular processing could thus regulate BPLF1 s (subcellular) localization and activity. This is reminiscent of the dual role of UL36 in HSV-1 infection: at late times, full-length UL36 functions as a structural/ tegument protein in virion assembly from virus-producing cells, whereas at early times the protein is cleaved upon entry into newly infected cells in order to allow nuclear release of the viral DNA [54 56]. Since full-length and/or truncated large tegument proteins with DUB activity have additionally been reported for HCMV [57] and MHV-68 [29], this may represent a general feature among these conserved viral DUBs. In the current study, we have found that BPLF1 interferes with innate immune activation by targeting multiple intermediates along the TLR signal transduction pathway, including TRAF6, NEMO, and IκBα (Figs 2, 4, and 5). Interestingly, Saito et al. very recently also reported TRAF6 to be deubiquitinated by BPLF1 [5], albeit in a different context. They reported that BPLF1 contributed to virus production by repressing NF-κB signaling, which is induced in latently infected EBV + cells by the EBV latent membrane protein 1 (LMP1) and signaled through TRAF6 [5]. Although in our system, BPLF1 is not detectable in latently EBV-infected B cells, simultaneous expression of BPLF1 and LMP1 could occur in B cells during productive infection. In agreement with published results [58], low levels of LMP1 were expressed from 8 hours of lytic induction of AKBM B cells and higher levels of the shorter, lytic form of LMP1 (lylmp1) were detected in cell lysates of 24 and 48 hours post-induction (Fig. S4). LyLMP1 does not activate NF-κB signaling itself and in fact appears to counteract (latent FL) LMP1 signaling [59,6]. Whereas during latency LMP1 provides proliferative and survival signals through NF-κB, lylmp1 could cooperate with BPLF1 to reduce NF-κB activation in B cells producing new virions. In our study, we addressed another implication of BPLF1 s DUB function for virus-host interactions. BPLF1 inhibited NF-κB activation and proinflammatory 16

107 EBV deubiquitinase inhibits TLR signaling A B AKBM cells % EBV+ AKBM cells αigg (hrs) kda BPLF1 hrs BZLF1 (IE) BGLF5 (E) 17% 5 gp42 (L) 4 hrs 37 β-actin C 8 hrs αigg 49% PAA 37 1 Virus particles Lane 36% EBVAK31 EBV+ AKBM kda BPLF hrs 4 37 BZLF1 (IE) BGLF5 (E) % gp42 (L) hrs β-actin rcd2-gfp Lane Figure 6. BPLF1 is expressed during the late phase of lytic EBV infection and is incorporated into viral particles. EBV+ AKBM cells were treated with anti-human IgG (αigg) to induce productive infection. (a) At the indicated times post-induction, percentages of AKBM cells undergoing productive infection were determined by flow cytometric analysis of ratcd2-gfp reporter expression. (b) EBV protein expression in post-nuclear cell lysates was determined by immunoblotting with Abs specific for BZLF1 (immediate early, IE), BGLF5 (early, E), and gp42 (late, L). β-actin served as loading control. (c) Immunoblot of post-nuclear lysates of EBV B95.8 particles (lane 1), EBV- AK31 cells (lanes 2-4), and EBV+ AKBM cells (lanes 5-7). EBV+ AKBM cells and EBV- control AK31 cells were treated with αigg Abs for 24 hours, resulting in productive infection in 26% of AKBM cells; AK31 cells expressed ratcd2-gfp constitutively in ~4% of the population (data not shown). Phosphonoacetic acid (PAA, 3 μg/ml) treatment starting 1 hour prior to anti-igg treatment was used to inhibit late protein expression; β-actin served as loading control. cytokine production in response to TLR stimulation, e.g. by EBV particles (Fig. 2). In addition, we have identified likely targets downstream of TRAF6 that are relevant to this TLR inhibition: BPLF1 colocalizes with and can deubiquitinate K63-Ub-TRAF6, K63-Ub-NEMO, and K48-UbIκBα (Figs 4 and 5). The mechanistic experiments were mostly performed in 293T cells with BPLF1 expressed in isolation, yet the expression levels of the signaling components we studied appeared comparable in the B cells used for productive EBV infection (data not shown). Employing mutants of BPLF1 (D86/9) that retained an unaltered catalytic site, but no longer 17

108 Chapter 4 interact with the cullins [39], we observed that inhibition of NF-κB activation correlated with DUB activity of the mutants (Fig. 3). From this, we deduced that the TLR evasive properties of BPLF1 rely on its deubiquitinase, rather than its deneddylase, activity. This places EBV BPLF1 in the arena of viral proteins that target protein ubiquitination. As ubiquitination represents a post-translational modification that plays a regulatory role in NF-κB- and IRF-mediated immunity through protein degradation and signal transduction, it comes as no surprise that viruses have found ways to interfere with the process. On the one hand, ubiquitin conjugation can be altered in virus-infected cells. Examples include poxvirus-induced inhibition of ubiquitin-dependent degradation of IκBα via sequestration of the cellular E3 ligase involved [61] and KSHV Rta-encoded E3-ligase activity that invokes degradation of IRF7 [62]. On the other hand, increased deubiquitination imposed by viruses can also shift the balance. For instance, the cellular DUB UCHL1 appears to be upregulated by HPV16 infection leading to suppressed innate signaling through TRAF3 and NEMO [63]. The HSV tegument protein ICP [64] recruits the cellular DUB USP7 into the cytosol to switch off TLR signaling to NF-κB. More recently, virus-encoded DUBs were found to dampen innate responses towards both RNA viruses [65,66] and DNA viruses, adenovirus [67] and KSHV [68]. Here we demonstrate that EBV BPLF1 also exploits its DUB activity to modulate TLR signaling. Herpesviruses are ancient viruses and have evolved in close association with their hosts, leading to acquisition of extensive strategies to perturb host immunity. A common strategy emerged from studies on viral evasion of host adaptive immune mechanisms. This is exemplified by EBV, where multiple viral proteins target the HLA class I-restricted antigen presentation pathway to ensure the optimal timing and extent of inhibition of host immunity [69]. Viruses have adopted a similar strategy to achieve innate immune evasion. Poxviruses have already been found to encode a number of inhibitors for the NF-κB pathway [7]. During productive EBV infection, the past years have also witnessed identification of a number of TLR evasion strategies. The viral tegument protein LF2 prevents activation of IRF7 [71] and the immediate-early transactivator BZLF1 negatively regulates IRF7-induced responses [72]. Of the early EBV early proteins, the viral kinase BGLF4 inhibits IRF3 activation [73] and the shutoff protein BGLF5 reduces TLR expression levels [23]. Here, we have reported on the large tegument protein BPLF1, expressed at late times of infection, downregulating TLR signaling through its DUB activity. Since tegument proteins such as BPLF1 are packaged into viral particles, they are transferred directly into newly infected cells (Fig. 7). Its close association with the EBV capsid would suggest that BPLF1 is present when PAMPs are activating innate receptors and cleavage of the N-terminal domain of BPLF1 might then allow deubiquitination to take place. It is tempting to speculate that deubiquitination by this EBV protein may exert its immunomodulatory functions in particular during primary infection. If and how this occurs awaits future elucidation. 18

109 EBV deubiquitinase inhibits TLR signaling TLR2 (a) MyD88 IRAK4 IRAK1 (a) TLR3 A2 (d) TRAF6 Ub Ub Ub Ub Ub TRIF (g) TAB1 TAB2- TAB3 TAK1 K63-Ub (e) BPLF1 (b) IKKα IKKβ NEMO Ub Ub Ub Ub Ub (e) (e) P P IκBα NF-κB K63-Ub Ub Ub Ub Ub Ub Proteasomal degradation IκBα K48 Ub P (h) (f) 4 Cytoplasm NF-κB Nucleus Proinflammatory cytokines (c) BPLF1 Secreted cytokines e.g. IL-8 Figure 7. Schematic model of BPLF1-mediated TLR evasion during EBV infection. During EBV infection, EBV components are sensed by host TLRs (a). These receptors activate signaling pathways through adaptor molecules MyD88 and TRIF, culminating in activation of transcription factor NF-κB (b). NF-κB induces production of proinflammatory cytokines (c) and obstructs viral replication. TLR signaling pathways are extensively regulated by ubiquitination, a process governed by cellular ubiquitin ligases and deubiquitinases (e.g. A2, d). The EBV-encoded DUB BPLF1 counteracts TLR-mediated NF-κB activation and can interfere with K63- and K48-linked ubiquitination of signaling intermediates, for example on TRAF6, NEMO, and IκBα (e). BPLF1 is expressed as a full-length protein during the late phase of productive EBV infection, is incorporated into the tegument of viral particles (f), and can subsequently be released into newly infected cells (g). It is as yet unclear where BPLF1 is processed to yield a shorter active fragment of ca. 28 aa (see Fig. 1). Thus, (processed) BPLF1 could exert its immunomodulatory functions towards TLR signaling not only in EBV-producing B cells (h), but also during primary infection (g). 19

110 Chapter 4 Acknowledgments We gratefully acknowledge T. Chiba (University of Tsukuba, Japan) and A. Baldwin (UNC Lineberger Comprehensive Cancer Center, Chapel Hill, USA) for generously sharing reagents and constructs. Materials and Methods Cells AKBM cells are derived from an EBV + Burkitt s lymphoma (BL) cell line (EBV strain Akata) by stable transfection with the plasmid phebo-prbmrf1-ratcd2-gfp yielding inducible rat CD2-GFP expression in productively EBV-infected cells (see below) [33]. AK31 cells are derived from an EBV-negative subclone of the parental Akata BL cell line by stable transfection to achieve constitutive expression of the ratcd2-gfp fusion protein in part of the population. Jijoye cells are EBV + B cells that express LMP1 [34]. AKBM, AK31, and Jijoye cells were maintained in RPMI 164 medium (Invitrogen) supplemented with 1% FBS (Sigma), 2 mm L-glutamine, 1 U/mL penicillin, 1 μg/ml streptomycin, and.3 mg/ml hygromycin B (AKBM cells) or 1 mg/ml geneticin (AK31 cells). Human embryonic kidney 293T cells were maintained in Dulbecco s modified Eagle medium (DMEM; Invitrogen) supplemented with 1% FBS (Sigma) and 2 mm L-glutamine (Invitrogen). 293 cells expressing human TLR2 and CD14 (293-TLR2/CD14) or CD14 alone (293-CD14) were described previously [35] and 293 cells expressing human TLR3 (293- TLR3) were obtained from InvivoGen (San Diego, CA). Productive EBV infection and EBV particles Productive EBV infection was induced in AKBM cells as described [33]. Briefly, B cell surface IgG was cross-linked by incubation with 5 μg/ml goat F(ab )2 fragments to human IgG (Cappel; MP Biomedicals) for the times indicated. Percentages of productively infected cells were determined based on induced expression of the rat CD2-GFP reporter protein. To discriminate (immediate) early and late stages of productive infection, viral DNA replication and late phase gene expression were inhibited by treatment with 3 μg/ml phosphonoacetic acid (PAA, ph7.4 in 1 mm HEPES, Sigma Aldrich) starting 1 hour prior to anti-igg treatment. To exclude secondary effects of antibody (Ab) treatment or the presence of PAA, control EBV - AK31 cells were included in parallel. Directly pelleted EBV (strain B95.8) was purchased from Advanced Biotechnologies Inc. (batch #16-176) and lysates of 5 x 1 1 particles/ml were prepared in reducing sample buffer (2 mm Tris-HCl [ph6.8], 3% SDS, 1% glycerol, 1 mm EDTA, 5 mm DTT,.5% bromophenol blue). Plasmids BPLF1 (aa 1-325), BSLF1, or BXLF1 in frame with an N-terminal HA- (YPYDVPDYA) or Flag- (DYKDDDDK) tag were PCR amplified from EBV B95.8 genomic DNA using the primers listed in Table 1. An additional Flag-tagged BPLF1 (aa 1-325) construct was generated by PCR amplification based on EBV strain Akata. Fragments were cloned into Gateway donor vectors and subcloned into the mammalian expression vector pmaxcloning (Lonza). Full-length BPLF1 (FL BPLF1) with an N-terminal Flag-tag and C-terminal HA-tag was engineered in pmaxcloning by joining three overlapping fragments of B95.8 BPLF1 amplified from genomic DNA (see Table 1), using unique restriction sites present inside the BPLF1 open reading frame (EcoRI and ClaI). The resulting sequence was identical to the B95.8 BPLF1 open reading frame except for a deletion spanning nucleotides (two out of eight PASAA repeats) and a T to A mutation at nucleotide position BPLF1 (aa 1-325) mutants C61A (TGC to GCC), D86/9R (two times GAC to CGC), or D86/9G (two times GAC to GGC) were generated by site-directed mutagenesis using Pfu Turbo polymerase (Agilent Technologies) and the primers listed in Table 1. Correctness of all constructs was verified by sequencing. Other plasmids used were: pcdna3-flag-myd88 (M. Muzio, Milan, Italy), IRAK1, Flag-TRAF6 (Tularik), pcdna3- Flag-IKKα (A. Baldwin, Chapel Hill, USA), IκBα, His-NEMO (IKKγ), Flag-A2, pcdna3.1-ha-ubiquitin (T. Chiba, University of Tsukuba, Japan), and puno-htlr2 (InvivoGen). 11

111 EBV deubiquitinase inhibits TLR signaling Transfections To transiently express individual EBV proteins, 293T cells (1.5 x 1 6 ) were transfected in suspension with a total of 4 μg vector DNA using Lipofectamine 2 according to manufacturer s instructions (Invitrogen) and seeded in 6-well plates. 16 hours post-transfection, cells were lysed for DUB profiling or immunoblotting (see below). To visualize general protein ubiquitination, 293T cells (1.5 x 1 6 ) were transfected with a combination of plasmids encoding HA-Ub (1 μg) and BPLF1 variants (2 μg), supplemented with empty vector (1 μg). Cells were seeded in 6-well plates and lysed 4 hours post-transfection in reducing sample buffer (2 mm Tris-HCl [ph6.8], 3% SDS, 1% glycerol, 1 mm EDTA, 5 mm DTT,.5% bromophenol blue). To study IκBα degradation, 293T cells (1 x 1 6 ) were transfected in suspension using 3 μg DNA and 13 μl PEImax (Brunschwig Chemie). After 4 hours, cells were stimulated with 1 ng/ml MALP-2 and post-nuclear lysates were prepared. DUB profiling Active DUBs were labeled in cell lysates as previously described [36]. In brief, post-nuclear cells were prepared in lysis buffer (5 mm Tris-HCl, 25 mm sucrose, 5 mm MgCl 2, 1 mm DTT,.5% CHAPS, and.1% NP4). DUB labeling was performed for 3 min at 37 C in lysis buffer containing 1 mg/ml protein extract, 1 μm 5-carboxytetramethylrhodamine (TMR)-conjugated ubiquitin-based probe, and two equivalents (v/v relative to probe) 5 mm NaOH to neutralize the ph. Reactions were terminated by the addition of reducing sample buffer prior to heating at 9 C for 1 min, after which proteins were resolved by SDS-PAGE (4%-12% gradient gel) and visualized by in-gel fluorescence imaging. Where indicated, gels were further used for immunoblotting. Table 1. Primer sequences for cloning EBV putative DUBs. Gene product Amino acids F/R* Primer sequence (5-3 ) Flag-FL BPLF1 -HA F GGGGTACCCAATTGCCACCATGGATTACAAGGATGACGACGATAAG AGTAACGGCGACTGGGGGCAAAGC R GTGGGGACCTGCAGATGTCCCG F GTCCGACTCTGAAGAAGCGGAGAGCG R F CGGGGCCGCAAAGAACCCCACG R ATAGTTTAGCGGCCGCTTAGGATGAGCGTTTGGGAGAGCTGATTCTGC ATAGTTTAGCGGCCGCTTAAGCGTAATCTGGAACATCGTATGGGTA CAGATACAAAAACTTGAGTCTCTCGAGG Flag-BSLF F GGGCTCGAGACCATGGATTACAAGGATGACGACGATAAGTCCGCC CCCGTCGTCATCAAGG R GGGTCTAGACTAGTTCGGGAGAGTCTCTGAGAAG Flag-BXLF F GGGGACAAGTTTGTACAAAAAAGCAGGCTTCCTCGAGACCATGGAT TACAAGGATGACGACGATAAGGCTGGATTTCCAGGAAAGGAGGCC GCTGGATTTCCAGGAAAGGAGGCC R GGGGACCACTTTGTACAAGAAAGCTGGGTCTCTAGACTAGTCCCGA TTTCCCCTCTC Flag-BPLF F GGGGACAAGTTTGTACAAAAAAGCAGGCTTCCTCGAGACCATGGAT TACAAGGATGACGACGATAAGAGTAACGGCGACTGGGGGCAAAGC R GGGGACCACTTTGTACAAGAAAGCTGGGTCTCTAGACTAAGGACTA TACCTGGCGGCAGGG HA-BPLF F GGGGTACCCAATTGCTCGAGCCACCATGTACCCATACGATGTTCCA GATTACGCTAGTAACGGCGACTGGGGGCAAAGC *F: forward, R: reverse R GGGGACCACTTTGTACAAGAAAGCTGGGTCTCTAGACTAAGGACTA TACCTGGCGGCAGGG 4 111

112 Chapter 4 Antibodies (Abs) BPLF1-specific rat monoclonal Abs were prepared by immunizing Lou/C rats with OVA-coupled peptides encompassing BPLF1 amino acids 2-17 (peptide A: SNGDWGQSQRTRGTGP) and amino acids (peptide B: PLTSRPELDEVLDEGAR) of B95.8. To detect other EBV proteins, the following Abs were used: mouse anti-bzlf1, mouse anti-bglf5 (311H) [37], rabbit polyclonal anti-gp42 (PB1112), and mouse anti-lmp1 (clone CS1-4). Other Abs included: mouse anti-flag (M2, F3165, Sigma-Aldrich), rat anti-ha (3F1, , Roche Diagnostics), mouse anti-his (5H1, HyTest Ltd), mouse antiβ-actin (Clone C4, MAB151R, Millipore), mouse anti-transferrin receptor (H68.4, 13-68, Life Technologies), mouse anti-p97 (#612183, BD Transduction laboratories), and rabbit anti-histon H3 (Cell Signaling Tech #9715). Specific Abs against signaling intermediates included rabbit anti-traf6 (H274, sc-7221, Santa Cruz, for immunoprecipitation) and mouse anti-traf6 (sc-849, Santa Cruz, for immunoblotting), rabbit anti-nemo (sc-833, Santa Cruz), rabbit anti-iκbα (sc-23, Santa Cruz), and rabbit anti-nf-κb (sc-372, Santa Cruz). For flow cytometry, goat anti-mouse allophycocyanin-conjugated (APC; M11, Leinco Technologies) or fluorescein isothiocyanate-conjugated (FITC; Dako) Abs were used as second step. For immunoblotting, HRP-conjugated secondary Abs used were: anti-rat Ig (light-chain specific, # , Jackson), anti-mouse Ig (light-chain specific, # , Jackson), and anti-rabbit Ig (43-5, Southern Biotech). Secondary Abs used for immunofluorescence were anti-mouse Ig-Cy3 ( , Jackson), anti-rabbit Ig-Dylight 488 ( , Jackson), and anti-rat Ig-Dylight 488 ( , Jackson). Flow cytometry For intracellular staining, cells were fixed with PBS/1% formaldehyde and permeabilized using.5% saponin (Sigma- Aldrich). Samples were measured on a FACSCantoII or FACSCalibur flow cytometer (BD Biosciences) and analyzed using FACSDiva (BD Biosciences) and FlowJo software (Tree Star). Immunoblotting Immunoblot analysis was performed as described [33]. In brief, post-nuclear cell lysates were prepared as follows: cells were lysed using NP4 lysis buffer (.5% Igepal-CA63, 5 mm Tris HCl [ph7.5], 15 mm NaCl, 1 μm Leupeptin, and 1 mm 4-(2-aminoethyl)benzenesulfonyl fluoride), post-nuclear lysates were clarified by centrifugation, and proteins in the lysates were denatured with reducing sample buffer. Where indicated, nuclear fractions were prepared by resuspending the nuclear pellet obtained after centrifugation in NP4 lysis buffer through sonification. Solubilized proteins were resolved by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Bio-Rad). Membranes were incubated with specific primary Abs followed by HRP-conjugated secondary Abs and reactive protein bands were visualized using ECL Plus detection kit (GE Healthcare) and ImageQuant LAS 4 imager (GE Healthcare Life Sciences). Densitometric quantification of staining intensity was performed using Quantity One software (Bio-Rad). Luciferase reporter assays and IL-8 ELISA Transfections were conducted in suspension using Lipofectamine 2 (Invitrogen) according to the manufacturer s protocol in 96-well flat-bottom plates. Briefly, 293(T) cells (5 x 1 4 ) were transfected with a total of 2 ng DNA containing 75 ng pgl3-firefly luciferase controlled by either five NF-κB binding sites (NF-κB-luc), part of the human E-selectin promoter (pelam-luc), or the human IFN-β promoter together with 4 ng of HSV thymidine kinase driven Renilla luciferase for normalization (pgl3-tk vector), and indicated amounts of plasmids encoding EBV genes or controls. Where indicated, 293T cells were cotransfected with 5 ng puno-tlr2. TLR signaling was induced either by cotransfecting activator proteins MyD88 (15 ng), IRAK1 (8 ng), TRAF6 (4 ng), or IKKα (75 ng) or by treatment for 7 hours with 1 ng/ml TLR2 agonist MALP-2 (sc , Santa Cruz), 1 μg/ml TLR3 agonist poly(i:c) (HMW, InvivoGen), or 6.5 x 1 6 EBV particles/ml (purified virus, batch 4H12-PV, Advanced Biotechnologies Inc) starting at 16 hours posttransfection. Cells were lysed in Passive Lysis Buffer and Firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System according to manufacturer s instructions (Promega). Data were normalized for Renilla luciferase activity and presented as percentage firefly luciferase activity relative to the stimulated transfected control sample, mean ± S.D. To study effects on cellular IL-8 production, 293-TLR2/CD14 cells were transfected in 96-well flat-bottom plates with 2 ng EBV or control plasmids and were stimulated with 1 ng/ml MALP-2 or 6.5 x 1 6 EBV particles/ml (batch 4H12-PV) at 24 hours post-transfection in 1 μl DMEM (transfection efficiencies around 6%). Following 8 hours of 112

113 EBV deubiquitinase inhibits TLR signaling stimulation, levels of IL-8 secreted in the culture supernatants were measured by ELISA using PeliKine Compact human IL-8 ELISA kit (Sanquin Reagents) according to manufacturer s instructions. Cellular deubiquitination assay 293T cells were transfected with HA-ubiquitin together with either Flag-TRAF6 or His-NEMO and empty vector, FlagBPLF1, Flag-BPLF1C61A, or Flag-A2. Post-nuclear extracts of transfected cells were prepared in lysis buffer (5 mm Tris-HCl [ph7.5], 15 mm NaCl,.5% Igepal [w/v], 1 mm dihiothreitol [DTT], 1 mm phenylmethylsulfonyl fluoride (PMSF), 25 μg/ml Leupeptin, 25 μg/ml aprotinin, 1 mm benzamidine, 1 μg/ml trypsin inhibitor), in case of TRAF6 supplemented with 5 mm NaF and 1 mm Na3VO4. Samples containing TRAF6 were denatured by addition of SDS to a final concentration of 1% and heating at 95 C for 5 minutes, after which SDS was diluted by adding additional lysis buffer. TRAF6 was precipitated by incubating with anti-traf6 Ab overnight at 4 C with constant agitation and for an additional 5 hours with protein A-sepharose beads. His-NEMO was precipitated by incubation with nickel-nta agarose beads (Qiagen) overnight at 4 C. After washing, precipitated proteins were eluted in reducing sample buffer and analyzed by immunoblotting. In vitro deubiquitination of signaling intermediates 293T cells were cotransfected with HA-ubiquitin and either Flag-TRAF6, His-NEMO, or IκBα. 24 hours post transfection, IκBα-transfected cells were additionally treated with 5 μm MG132 and 1 ng/ml TNFα for 2 hours. Post-nuclear cells lysates were prepared using NP4 lysis buffer with protease inhibitor cocktail (Roche) and 2 mm NEM. To precipitate (ubiquitinated) TRAF6, NEMO, or IκBα, cell lysates were incubated overnight at 4 C with constant agitation with antiflag M2 Abs and protein G sepharose beads, nickel NTA beads (Qiagen), or anti-iκbα Abs and protein A sepharose beads, respectively. Flag-TRAF6(-Ubn) was released from the beads using 1 μg/ml Flag peptide (Sigma) and His-NEMO(-Ubn) was eluted using 25 mm imidazole. The thus obtained ubiquitinated signaling proteins were incubated for 4 hrs at 37 C in DUB assay buffer (5 mm HEPES-NaOH ph8., 1% glycerol, 3 mm DTT) with Flag-tagged BPLF1, BPLF1C61A, or A2 DUB proteins that were purified from separately transfected 293T cells using anti-flag Abs and protein G sepharose beads. Samples were analyzed by immunoblot analysis. Immunofluorescence 293T cells (3.) were transfected with plasmids encoding BPLF1 and signaling intermediates (125 ng each) using Lipofectamine 2 and cultured overnight in Lab-Tek II chamber slide system (Thermo Scientific). Cells were fixed using 3% paraformaldehyde in PBS for 3 minutes at 37 C and permeabilized by incubation with PBS/1% Triton X-1/.1% sodium citrate for 5 minutes at room temperature. After blocking for 6 minutes with PBS/5% BSA, samples were incubated with primary Abs in PBS/1% BSA/.1% Triton X-1 for 6 minutes, washed, and incubated for 6 minutes with fluorescently labeled secondary Abs together with DNA stain TO-PRO-3 (Life Technologies). Slides were embedded in Mowiol (Brunschwig Chemie) and analyzed using a confocal microscope (Leica SP5) and Leica Application Suite Advance Fluorescence software. Supporting information EBV+ AKBM cells + αigg (hrs) kda 191 Figure S1. Enlarged view of the upper part of the gel depicted in Fig. 1b. For details, see legend Fig. 1b

114 Chapter 4 A B95.8 Akata B95.8 Akata Peptide A (2-17) Peptide B (78-94) MSNGDWGQSQRTRGTGPVRGIRTMDVNAPGGGSGGSALRILGTASCNQAHCKFGRFAGIQCVSNCVLYLVKSFLAGRPLTSRPELDEVLDEGARLDALMRQ MSNGDWGQSQRPRGTGPVRGIRTMDVNAPGGGSGGSALRILGTASCNQAHCKFGRFAGIQCVSNCVLYLVKSFLAGRPLTSRPELDEVLDEGARLDALMRQ * SGILKGHEMAQLTDVPSSVVLRGGGRVHIYRSAEIFGLVLFPAQIANAVVQSLAEVLHGSYNGVAQFILYICDIYAGAIIIETDGSFYLFDPHCSQKDAAP SGILKGHEMAQLTDVPSSVVLRGGGRVHIYRSAEIFGLVLFPAQIANSAVVQSLAEVLHGSYNGVAQFILYICDIYAGAIIIETDGSFYLFDPHCQKDAAP B95.8 GTPAHVRVSTYAHDILQYVGAPGAQYTCVHLYFLPEAFETEDPRIFMLEHYGVYDFYEANGSGFDLVGPELVSSDGEAAGTPGADSSPPVMLPFERRIIPY Akata GTPAHVRVSTYAHDILQYVGAPGAQYTCVHLYFLPEAFETEDPRIFMLEHYGVYDFYEANGSGFDLVGPELVSSDGEAAGTPGADSSPPVMLPFERRIIPY B95.8 NLRPLPSRSFTSDSFPAARYSP... Akata NLRPLPSRSFTSDSFPAARYSP... B IB: kda anti-bplf1 mab aa 2-17 (2E5) B95.8 control control Akata B95.8 control Akata aa (1F2) Akata B95.8 anti-flag BPLF1 * C IP: IP: Flag Native 21E9 2E5 Denaturing Lane Lane IB: anti-flag 21E9 1F2 Flag-BPLF1 (B95.8) Flag-BPLF1 (Akata) Flag-BXLF1 Figure S2. BPLF1-specific monoclonal antibodies. Rat monoclonal Abs were generated against peptides encompassing residues 2-17 (peptide A) and residues (peptide B) of EBV strain B95.8-encoded BPLF1. (a) Sequence alignment of the N-terminal parts of BPLF1 (aa 1-325) derived from the EBV strains B95.8 and Akata. An asterisk indicates the amino acid difference between these strains at position 12 (T12P). (b) Reactivity of two BPLF1-specific (IgG2a) mabs 2E5 and 1F2 was tested in immunoblot. 293T cells were transiently transfected with plasmids encoding the N-terminal domains of BPLF1 from EBV strains B95.8 and Akata; transfection efficiencies were comparable (~6% positive cells). Sixteen hours after transfection, post-nuclear cell lysates were prepared, separated by SDS-PAGE, and immunoblots were stained with an anti-flag Ab or anti-bplf1 Abs directed against peptide A (2E5) or peptide B (1F2). Equal loading was demonstrated by comparable band intensity observed upon staining for the Flag-tag (lanes 5 and 6); this was further supported by the staining with Ab 1F2 that reacts with an epitope identical in both EBV strains (lanes 3 and 4). The T12P amino acid difference strongly reduced 2E5-mediated detection of Akata-derived BPLF1 compared to B95.8-derived BPLF1 (lanes 1 and 2). An asterisk indicates the smaller fragment arising upon cellular expression of BPLF1. Since this 32 kda band is recognized by both the anti-flag Ab (tag at N-terminus) as well as the two BPLF1-reactive Abs (aa 2-17 and 78-94), it likely represents a truncated N-terminus of BPLF1 (residues 1- ~28). (c) BPLF1-specific monoclonal Abs were tested in immunoprecipitation experiments. 293T cells were transfected with Flag-tagged N-terminal domains of B95.8- and Akata derived BPLF1, or Flag-BXLF1 as control. Immunoprecipitations were performed in post-nuclear lysates using monoclonal anti-bplf1 Abs 21E9 (peptide B, lanes 2 and 4), 2E5 (peptide A, lane 3), and 1F2 (peptide A, lane 5), or anti-flag Ab (lane 1) under native conditions or after denaturing proteins by incubating with 1% SDS as indicated. Ab 21E9 appeared ineffective in precipitating BPLF1, while 2E5 and, to a lesser extent, 1F2 precipitated B95.8 as well as Akata-derived BPLF1 under native and denaturing conditions, respectively. 114

115 EBV deubiquitinase inhibits TLR signaling A FL BPLF1 Flag-BPLF1 HA-BPLF1 αflag αbplf1 1F2 2E5 Flag * * Flag * * HA * * 1 N-terminal αha fragment 325 NLS * C61 ** D86/9 αha HA 3149 aa numbers B anti-bplf1 (2E5) anti-ha BPLF1 FL BPLF C N C N C N C N Transfection (hrs) Fraction FL BPLF1 Flag-BPLF1/ * FL BPLF1 HA-BPLF1 C kda BPLF1 C 293T FL BPLF1 EBV+ AKBM Lane IB: α-bplf1 1F2 + αigg (hrs) FL BPLF1 * BPLF1 4 anti-p anti-histon H3 Lane Figure S3. BPLF1 expression in time. (a) Schematic representation of the EBV BPLF1 protein. Green boxes indicate peptides used to generate BPLF1- specific rat monoclonal Abs (2E5, anti-peptide A, aa 2-17; 1F2, anti-peptide B, aa 78-94). Asterisks denote aa substitutions of the mutants used in this study. NLS: nuclear localization signal. Numbers refer to aa positions in EBV strain B95.8. (b) 293T cells were transfected with plasmids encoding Flag-tagged BPLF1, HA-tagged BPLF1 or full length BPLF1 containing an N-terminal Flag-tag and C-terminal HA-tag (Flag-FL BPLF1 -HA). At 24, 48, and 72 hours post-transfection, cytosolic (C) and nuclear (N) fractions were prepared and analyzed by immunoblotting with anti-bplf1 (2E5) and anti-ha Abs. Adequate separation of cytosolic and nuclear fractions was evaluated by immunoblot analysis of cytosolic p97 and nuclear histon H3 using specific Abs. Left and right panels are part of the same gels presented at different exposures. The asterisk indicates the ~32 kda cytosolic fragment observed over time upon expression of FL BPLF1. (c) Immunoblot using BPLF1-specific Ab 1F2 shows BPLF1 expression and processing in EBV + AKBM cells (lanes 4-7) and transfected 293T cells (lanes 1-3). AKBM cells were treated with anti- IgG Abs to induce productive infection and post-nuclear lysates were prepared after the indicated time periods. 293T cells were transfected with constructs encoding the N-terminal domain of BPLF1 (aa 1-325), full-length BPLF1, or an empty control plasmid (C). The asterisk indicates the smaller BPLF1 fragment observed at 24 and 48 hours after induction of productive EBV infection in AKBM cells (lanes 6 and 7) and upon expression of full-length BPLF1 in transfected 293T cells (lane 3). 115

116 Chapter 4 cells Jijoye EBV+ AKBM αigg (hrs) LMP kda LMP1 lylmp1 TfR 1 Lane Figure S4. LMP1 expression during productive EBV infection in AKBM cells. EBV + AKBM cells were treated with anti-human IgG (αigg) to induce productive infection. At the indicated times post-induction, expression of LMP1 was determined in post-nuclear cell lysates by immunoblotting using a specific Ab (lanes 2-6). LMP1 expression was observed starting 8 hours after induction of lytic infection. Lytic LMP1 (lylmp1), an inhibitory variant of LMP1 that counteracts LMP-1 mediated activation of signaling pathways, is expressed at 24 hours post infection. Samples were the same as used for Figure 6b. EBV + latency III cell line Jijoye constitutively expressing LMP1 was included as a positive control (lane 1). 116

117 References 1. Rickinson AB, Kieff E (27) Epstein-Barr Virus. In: Knipe DM, Howley PM, editors. Field s Virology. Philadelphia: Lippincott Williams & Wilkins. pp Kutok JL, Wang F (26) Spectrum of Epstein-Barr Virus- Associated Diseases. Annu Rev Pathol Mech Dis 1: Iwasaki A, Medzhitov R (21) Regulation of Adaptive Immunity by the Innate Immune System. Science 327: Szomolanyi-Tsuda E, Liang X, Welsh RM, Kurt-Jones EA, Finberg RW (26) Role for TLR2 in NK Cell-Mediated Control of Murine Cytomegalovirus In Vivo. Journal of Virology 8: Tabeta K, Georgel P, Janssen E, Du X, Hoebe K, et al. (24) Toll-like receptors 9 and 3 as essential components of innate immune defense against mouse cytomegalovirus infection. Proceedings of the National Academy of Sciences of the United States of America 11: Zucchini N, Bessou G, Traub S, Robbins SH, Uematsu S, et al. (28) Cutting Edge: Overlapping Functions of TLR7 and TLR9 for Innate Defense against a Herpesvirus Infection. The Journal of Immunology 18: Davey GM, Wojtasiak M, Proietto AI, Carbone FR, Heath WR, et al. (21) Cutting Edge: Priming of CD8 T Cell Immunity to Herpes Simplex Virus Type 1 Requires Cognate TLR3 Expression In Vivo. The Journal of Immunology 184: Sorensen LN, Reinert LS, Malmgaard L, Bartholdy C, Thomsen AR, et al. (28) TLR2 and TLR9 Synergistically Control Herpes Simplex Virus Infection in the Brain. The Journal of Immunology 181: Casrouge A, Zhang SY, Eidenschenk C, Jouanguy E, Puel A, et al. (26) Herpes Simplex Virus Encephalitis in Human UNC-93B Deficiency. Science 314: Guo Y, Audry M, Ciancanelli M, Alsina L, Azevedo J, et al. (211) Herpes simplex virus encephalitis in a patient with complete TLR3 deficiency: TLR3 is otherwise redundant in protective immunity. The Journal of Experimental Medicine 28: Perez de Diego R, Sancho-Shimizu V, Lorenzo L, Puel A, Plancoulaine S, et al. (21) Human TRAF3 Adaptor Molecule Deficiency Leads to Impaired Toll-like Receptor 3 Response and Susceptibility to Herpes Simplex Encephalitis. Immunity 33: Sancho-Shimizu V, rez de Diego R, Lorenzo L, Halwani R, Alangari A, et al. (211) Herpes simplex encephalitis in children with autosomal recessive and dominant TRIF deficiency. J Clin Invest 121: Zhang SY, Jouanguy E, Ugolini S, Smahi A, Elain G, et al. (27) TLR3 Deficiency in Patients with Herpes Simplex Encephalitis. Science 317: Paludan SR, Bowie AG, Horan KA, Fitzgerald KA (211) Recognition of herpesviruses by the innate immune EBV deubiquitinase inhibits TLR signaling system. Nat Rev Immunol 11: Ariza ME, Glaser R, Kaumaya PTP, Jones C, Williams MV (29) The EBV-Encoded dutpase Activates NFkB through the TLR2 and MyD88-Dependent Signaling Pathway. The Journal of Immunology 182: Gaudreault E, Fiola S, Olivier M, Gosselin J (27) Epstein-Barr Virus Induces MCP-1 Secretion by Human Monocytes via TLR2. J Virol 81: Fiola S, Gosselin D, Takada K, Gosselin J (21) TLR9 Contributes to the Recognition of EBV by Primary Monocytes and Plasmacytoid Dendritic Cells. The Journal of Immunology 185: Iwakiri D, Zhou L, Samanta M, Matsumoto M, Ebihara T, et al. (29) Epstein-Barr virus (EBV)-encoded small RNA is released from EBV-infected cells and activates signaling from Toll-like receptor 3. The Journal of Experimental Medicine 26: Lim WH, Kireta S, Russ GR, Coates PTH (27) Human plasmacytoid dendritic cells regulate immune responses to Epstein-Barr virus (EBV) infection and delay EBV-related mortality in humanized NOD-SCID mice. Blood 19: Quan TE, Roman RM, Rudenga BJ, Holers VM, Craft JE (21) Epstein-Barr virus promotes interferon-alpha production by plasmacytoid dendritic cells. Arthritis & Rheumatism 62: Mogensen TH (29) Pathogen Recognition and Inflammatory Signaling in Innate Immune Defenses. Clinical Microbiology Reviews 22: Takeuchi O, Akira S (29) Innate immunity to virus infection. Immunological Reviews 227: van Gent M, Griffin BD, Berkhoff EG, van Leeuwen D, Boer IGJ, et al. (211) EBV Lytic-Phase Protein BGLF5 Contributes to TLR9 Downregulation during Productive Infection. The Journal of Immunology 186: Wertz IE, Dixit VM (21) Signaling to NF-kB: Regulation by Ubiquitination. Cold Spring Harbor Perspectives in Biology 2: Boone DL, Turer EE, Lee EG, Ahmad RC, Wheeler MT, et al. (24) The ubiquitin-modifying enzyme A2 is required for termination of Toll-like receptor responses. Nat Immunol 5: Brummelkamp TR, Nijman SMB, Dirac AMG, Bernards R (23) Loss of the cylindromatosis tumour suppressor inhibits apoptosis by activating NF-[kappa]B. Nature 424: Isaacson MK, Ploegh HL (29) Ubiquitination, Ubiquitin-like Modifiers, and Deubiquitination in Viral Infection. Cell Host Microbe 5: Randow F, Lehner PJ (29) Viral avoidance and exploitation of the ubiquitin system. Nat Cell Biol 11:

118 Chapter Gredmark S, Schlieker C, Quesada V, Spooner E, Ploegh HL (27) A Functional Ubiquitin-Specific Protease Embedded in the Large Tegument Protein (ORF64) of Murine Gammaherpesvirus 68 Is Active during the Course of Infection. Journal of Virology 81: Kattenhorn LM, Korbel GA, Kessler BM, Spooner E, Ploegh HL (25) A Deubiquitinating Enzyme Encoded by HSV-1 Belongs to a Family of Cysteine Proteases that Is Conserved across the Family Herpesviridae. Mol Cell 19: Schlieker C, Korbel GA, Kattenhorn LM, Ploegh HL (25) A Deubiquitinating Activity Is Conserved in the Large Tegument Protein of the Herpesviridae. Journal of Virology 79: Sompallae R, Gastaldello S, Hildebrand S, Zinin N, Hassink G, et al. (28) Epstein-Barr Virus Encodes Three Bona Fide Ubiquitin-Specific Proteases. Journal of Virology 82: Ressing ME, Keating SE, van Leeuwen D, Koppers- Lalic D, Pappworth IY, et al. (25) Impaired Transporter Associated with Antigen Processing-Dependent Peptide Transport during Productive EBV Infection. The Journal of Immunology 174: Bornkamm GW, Hudewentz J, Freese UK, Zimber U (1982) Deletion of the Nontransforming Epstein-Barr Virus Strain P3HR-1 Causes Fusion of the Large Internal Repeat to the DSL Region. Journal of Virology 43: Kurt-Jones EA, Mandell L, Whitney C, Padgett A, Gosselin K, et al. (22) Role of Toll-like receptor 2 (TLR2) in neutrophil activation: GM-CSF enhances TLR2 expression and TLR2-mediated interleukin 8 responses in neutrophils. Blood 1: de Jong A, Merkx R, Berlin I, Rodenko B, Wijdeven RHM, et al. (212) Ubiquitin-Based Probes Prepared by Total Synthesis To Profile the Activity of Deubiquitinating Enzymes. ChemBioChem 13: Tsai CH, Liu MT, Chen MR, Lu J, Yang HL, et al. (1997) Characterization of Monoclonal Antibodies to the Zta and DNase proteins of Epstein-Barr Virus. J Biomed Sci 4: Gastaldello S, Hildebrand S, Faridani O, Callegari S, Palmkvist M, et al. (21) A deneddylase encoded by Epstein-Barr virus promotes viral DNA replication by regulating the activity of cullin-ring ligases. Nat Cell Biol 12: Gastaldello S, Callegari S, Coppotelli G, Hildebrand S, Song M, et al. (212) Herpes virus deneddylases interrupt the cullin-ring ligase neddylation cycle by inhibiting the binding of CAND1. Journal of Molecular Cell Biology 4: Whitehurst CB, Ning S, Bentz GL, Dufour F, Gershburg E, et al. (29) The Epstein-Barr Virus (EBV) Deubiquitinating Enzyme BPLF1 Reduces EBV Ribonucleotide Reductase Activity. Journal of Virology 83: Whitehurst CB, Vaziri C, Shackelford J, Pagano JS (212) Epstein-Barr Virus BPLF1 Deubiquitinates PCNA and Attenuates Polymerase eta Recruitment to DNA Damage Sites. Journal of Virology 86: Gastaldello S, Chen X, Callegari S, Masucci MG (213) Caspase-1 Promotes Epstein-Barr Virus Replication by Targeting the Large Tegument Protein Deneddylase to the Nucleus of Productively Infected Cells. PLoS Pathog 9: e Sachdev S, Hoffmann A, Hannink M (1998) Nuclear Localization of IkBa Is Mediated by the Second Ankyrin Repeat: the IkBa Ankyrin Repeats Define a Novel Class of cis-acting Nuclear Import Sequences. Molecular and Cellular Biology 18: Schmaus S, Wolf H, Schwarzmann F (24) The Reading Frame BPLF1 of Epstein-Barr Virus: A Homologue of Herpes Simplex Virus Protein VP16. Virus Genes 29: Rowe M, Glaunsinger B, van Leeuwen D, Zuo J, Sweetman D, et al. (27) Host shutoff during productive Epstein-Barr virus infection is mediated by BGLF5 and may contribute to immune evasion. Proceedings of the National Academy of Sciences 14: Schlieker C, Weihofen WA, Frijns E, Kattenhorn LM, Gaudet R, et al. (27) Structure of a Herpesvirus- Encoded Cysteine Protease Reveals a Unique Class of Deubiquitinating Enzymes. Mol Cell 25: Kim ET, Oh SE, Lee YO, Gibson W, Ahn JH (29) Cleavage Specificity of the UL48 Deubiquitinating Protease Activity of Human Cytomegalovirus and the Growth of an Active-Site Mutant Virus in Cultured Cells. Journal of Virology 83: Roberts APE, Abaitua F, O Hare P, McNab D, Rixon FJ, et al. (29) Differing Roles of Inner Tegument Proteins pul36 and pul37 during Entry of Herpes Simplex Virus Type 1. Journal of Virology 83: Gredmark R, Isaacson MK, Kattenhorn L, Cheung EJ, Watson N, et al. (29) A Gammaherpesvirus Ubiquitin- Specific Protease Is Involved in the Establishment of Murine Gammaherpesvirus 68 Infection. Journal of Virology 83: Saito S, Murata T, Kanda T, Isomura H, Narita Y, et al. (213) Epstein-Barr Virus Deubiquitinase Downregulates TRAF6-Mediated NF-kB Signaling during Productive Replication. Journal of Virology 87: Ernst R, Claessen JHL, Mueller B, Sanyal S, Spooner E, et al. (211) Enzymatic Blockade of the Ubiquitin- Proteasome Pathway. PLoS Biol 8: e Ovaa H, Kessler BM, Rolén U, Galardy PJ, Ploegh HL, et al. (24) Activity-based ubiquitin-specific protease (USP) profiling of virus-infected and malignant human cells. Proceedings of the National Academy of Sciences of the United States of America 11: Johannsen E, Luftig M, Chase MR, Weicksel S, Cahir- McFarland E, et al. (24) Proteins of purified Epstein-Barr 118

119 EBV deubiquitinase inhibits TLR signaling virus. Proceedings of the National Academy of Sciences of the United States of America 11: Abaitua F, Hollinshead M, Bolstad M, Crump CM, O Hare P (212) A Nuclear Localization Signal in Herpesvirus Protein VP1-2 Is Essential for Infection via Capsid Routing to the Nuclear Pore. Journal of Virology 86: Jovasevic V, Liang L, Roizman B (28) Proteolytic Cleavage of VP1-2 Is Required for Release of Herpes Simplex Virus 1 DNA into the Nucleus. Journal of Virology 82: Schipke J, Pohlmann A, Diestel R, Binz A, Rudolph K, et al. (212) The C Terminus of the Large Tegument Protein pul36 Contains Multiple Capsid Binding Sites That Function Differently during Assembly and Cell Entry of Herpes Simplex Virus. Journal of Virology 86: Wang J, Loveland AN, Kattenhorn LM, Ploegh HL, Gibson W (26) High-Molecular-Weight Protein (pul48) of Human Cytomegalovirus Is a Competent Deubiquitinating Protease: Mutant Viruses Altered in Its Active-Site Cysteine or Histidine Are Viable. Journal of Virology 8: Vazirabadi G, Geiger TR, Coffin III, Martin JM (23) Epstein-Barr virus latent membrane protein-1 (LMP-1) and lytic LMP-1 localization in plasma membrane-derived extracellular vesicles and intracellular virions. Journal of General Virology 84: Pandya J, Walling DM (26) Oncogenic Activity of Epstein-Barr Virus Latent Membrane Protein 1 (LMP-1) Is Down-Regulated by Lytic LMP-1. Journal of Virology 8: Erickson KD, Martin JM (2) The Late Lytic LMP- 1 Protein of Epstein-Barr Virus Can Negatively Regulate LMP-1 Signaling. Journal of Virology 74: Mansur DS, Maluquer de Motes C, Unterholzner L, Sumner RP, Ferguson BJ, et al. (213) Poxvirus Targeting of E3 Ligase ßTrCP by Molecular Mimicry: A Mechanism to Inhibit NF-kB Activation and Promote Immune Evasion and Virulence. PLoS Pathog 9: e Yu Y, Wang SE, Hayward GS (25) The KSHV Immediate-Early Transcription Factor RTA Encodes Ubiquitin E3 Ligase Activity that Targets IRF7 for Proteosome-Mediated Degradation. Immunity 22: Karim R, Tummers B, Meyers C, Biryukov JL, Alam S, et al. (213) Human Papillomavirus (HPV) Upregulates the Cellular Deubiquitinase UCHL1 to Suppress the Keratinocyte s Innate Immune Response. PLoS Pathog 9: e Daubeuf S, Singh D, Tan Y, Liu H, Federoff HJ, et al. (29) HSV ICP recruits USP7 to modulate TLR-mediated innate response. Blood 113: van Kasteren PB, Beugeling C, Ninaber DK, Frias- Staheli N, van Boheemen S, et al. (212) Arterivirus and Nairovirus Ovarian Tumor Domain-Containing Deubiquitinases Target Activated RIG-I To Control Innate Immune Signaling. Journal of Virology 86: van Kasteren PB, Bailey-Elkin BA, James TW, Ninaber DK, Beugeling C, et al. (213) Deubiquitinase function of arterivirus papain-like protease 2 suppresses the innate immune response in infected host cells. Proceedings of the National Academy of Sciences 11: E838-E Balakirev MY, Jaquinod M, Haas AL, Chroboczek J (22) Deubiquitinating Function of Adenovirus Proteinase. Journal of Virology 76: Inn KS, Lee SH, Rathbun JY, Wong LY, Toth Z, et al. (211) Inhibition of RIG-I mediated signaling by Kaposi s sarcoma-associated herpesvirus-encoded deubiquitinase ORF64. Journal of Virology 85: Ressing ME, Wiertz EJHJ (28) Manipulation of the immune response by Epstein-Barr virus and Kaposi s sarcoma-associated herpesvirus: Consequences for tumor development. Seminars in Cancer Biology 18: Mohamed MR, McFadden G (29) NF-kB inhibitors: Strategies from poxviruses. Cell Cycle 8: Wu L, Fossum E, Joo CH, Inn KS, Shin YC, et al. (29) Epstein-Barr Virus LF2: an Antagonist to Type I Interferon. Journal of Virology 83: Hahn AM, Huye LE, Ning S, Webster-Cyriaque J, Pagano JS (25) Interferon Regulatory Factor 7 Is Negatively Regulated by the Epstein-Barr Virus Immediate-Early Gene, BZLF-1. Journal of Virology 79: Wang JT, Doong SL, Teng SC, Lee CP, Tsai CH, et al. (29) Epstein-Barr Virus BGLF4 Kinase Suppresses the Interferon Regulatory Factor 3 Signaling Pathway. Journal of Virology 83:

120

121 CHAPTER 5 Epstein-Barr virus mirna BART16 suppresses type I interferon signaling *Marjolein Hooykaas *Michiel van Gent Ingrid G. J. Boer Elisabeth Kruse Dik van Leenen Marian Groot Koerkamp Frank Holstege Maaike E. Ressing # Emmanuel J. Wiertz # Robert Jan Lebbink *,# equal contributions Manuscript in preparation

122 Chapter 5 Abstract Herpesviruses are well known for establishing an intricate relationship with their (human) host, in particular with respect to host immune responses. The prototypic human gammaherpesvirus Epstein-Barr virus (EBV), causally involved in infectious mononucleosis and various malignancies, expresses over 4 mature mirnas whose functions remain poorly understood. In this study we identify EBV-encoded mir-bart16 as a novel viral immune evasion factor that interferes with type I IFN responses. mir-bart16-5p directly targets transcripts of the cellular histone acetylase (HAT) CREBB-binding protein (CBP), a component of the type I IFN signaling pathway. CBP protein levels are downregulated in mir-bart16 expressing cells, and inhibition of endogenously expressed mir-bart16 in EBV-transformed B cells results in increased CBP levels. Consistent with the important role of CBP in type I IFN signaling, mir-bart16 expression interferes with ISRE-promoter activity and production of interferon-stimulated genes ISG15 and IFIT1 in response to IFN-α treatment. These combined data show that EBV employs viral mirnas to tamper with IFN-induced gene transcription. This provides a way for EBV to obstruct induction of an antiviral state, thereby favoring viral replication and ultimately the occurrence of EBV-associated malignancies. 122

123 EBV mirna BART16 suppresses type I IFN signaling Introduction Herpesviruses are large, enveloped DNA viruses causing lifelong infections. They are widespread, as exemplified by the prototypic human gammaherpesvirus Epstein-Barr virus (EBV) that infects over 9% of the adult population worldwide. Although EBV infection is mostly asymptomatic [1], the virus causes infectious mononucleosis and is associated with various epithelial and lymphoid malignancies [2]. Following primary infection, EBV establishes latency in memory B cells, characterized by the expression of a limited set of viral gene products. Transmission to other cells and other individuals is facilitated by occasional reactivation from these latently infected B cells through expression of the full repertoire of viral lytic genes. A first line of defense against incoming pathogens is posed by the innate immune system. Several families of pattern-recognition receptors (PRRs) sense conserved pathogen-associated molecular patterns (PAMPs), leading to production of type I interferons and proinflammatory cytokines [3]. Type I interferons (IFNs) are particularly important in limiting the spread of viral pathogens as they induce a cell-intrinsic antimicrobial state in infected and neighboring cells. thereby shaping the innate and adaptive immune response [4]. All type I IFN members, including IFN-α and IFN-β, interact with the heterodimeric type I IFN receptor (IFNAR) present on the surface of target cells. IFNAR ligation activates the receptor-associated protein tyrosine kinases Janus kinase 1 (JAK1) and tyrosine kinase 2 (TYK2), followed by phosphorylation and nuclear translocation of signal transducer and activator of transcription 1 (STAT1) and STAT2. The STAT proteins subsequently associate with interferon-regulatory factor 9 (IRF9) to form the IFN-stimulated gene factor 3 (ISGF3) complex. This complex binds IFN-stimulated response elements (ISREs) present in the promoter of interferon-stimulated genes (ISGs) and subsequently recruits chromatin remodeling and transcriptional activator complexes, thereby inducing transcription of these genes [4]. The induced ISGs perform antiviral functions by interfering with various aspects of the viral replication cycle [5]. Primary EBV infection is associated with a systemic interferon response [6]. EBV-mediated TLR9 activation in plasmacytoid dendritic cells (pdcs) leads to IFN-α secretion [7 9] and conventional DCs secrete IFNs upon recognition of EBV-encoded non-coding RNAs (EBERs) by TLR3 and RIG-I [1,11]. Additionally, IFNs limit primary EBV infection and replication [12,13], and recent studies in humanized mouse models suggest that IFN-α production by pdcs is crucial for controlling EBV infection [14]. Many viruses attenuate the adverse effects of IFN on virus replication by interfering with either IFN production or downstream signaling pathways [15 17]. Also EBV has adopted strategies to attenuate the IFN response to successfully establish and maintain infection. The immediate-early transactivators BZLF1 and BRLF1, BARF1, tegument component LF2, and the viral kinase BGLF4 interfere with transcriptional activity of interferon regulatory factors (IRFs) during productive infection 5 123

124 Chapter 5 [18 22]. In addition, EBV latency-associated proteins EBNA2 and LMP2 interfere with type I IFN-mediated functions [23,24]. MicroRNAs (mirnas) are small noncoding RNA molecules of nucleotides in length that post-transcriptionally regulate gene expression. MiRNAs emerged as key regulators of many cellular processes such as proliferation, differentiation, apoptosis, and (innate) immunity [25]. Virus-encoded mirnas, first identified in EBV-infected cells, regulate both virus and host protein expression [26,27]. Their small size, lack of immunogenicity, and ability to post-transcriptionally regulate (host) gene expression make mirnas appealing candidates for viral immune evasion strategies. Accordingly, several (herpes-)virus-encoded mirnas affect host immunity [28,29] by targeting the expression of chemokines [3,31], inflammasome components [32], NK cell ligands [33 35], antigen presentation [31], and components of innate immune signaling pathways [36 38]. EBV expresses more than 4 mature mirnas [39 41]. Three EBV mirna precursors are encoded in the BHRF1 cluster and another 22 are contained within introns of the non-coding BART transcript (BART cluster 1 and 2) [42]. The functional relevance of most EBV-encoded mirnas remains unknown. Bio-informatics and immunoprecipitation approaches have identified potential targets for (EBV) mirnas, yet experimental validation of these is often lacking. In the present study, we found that EBVencoded mirnas interfere with type I IFN-mediated immune responses and demonstrate that mir-bart16-5p downregulates cellular histone acetylase CBP. Results EBV mirnas attenuate type I IFN signaling To study interference of EBV mirnas with type I IFN-induced signaling pathways, a cluster of 1 BART mirnas (mir-bart1, 3, 4, 5, 6, 15, 16, 17, 18 and 21, designated BART cluster 1 ) was cloned into an intron separating two egfp-encoding exons in a lentiviral expression vector. Following lentiviral transduction of 239T cells, the activity of the type I IFN signaling pathway was probed using a reporter encoding firefly luciferase under the control of five IFNstimulated response element (ISRE) repeats (Fig. 1a). We observed reduced IFN-α-induced firefly luciferase activity in BART cluster 1-expressing cells as compared to control cells, suggesting that one or multiple EBV mirnas interfere with the IFN-α-induced signaling pathway (Fig. 1b). To screen ISRE-promoter activity at the single cell level by flow cytometry, 293T cells were transduced with a vector encoding the fluorescent protein mcherry under the control of ISRE promoter elements (293T-ISRE-mCherry; Fig. 1c). Treatment of these cells with IFN-α induced mcherry reporter expression in a dose-dependent manner; high IFN-α concentrations induced mcherry reporter expression in the entire population, whereas low concentrations only gave rise to reporter expression in a proportion of the cells (Fig. 1d). Abolishing expression of two crucial components of the canonical type I IFN signaling route 124

125 EBV mirna BART16 suppresses type I IFN signaling (IFNAR2, one of the two chains of the heterodimeric receptor, and IRF9) each ablated mcherry upregulation, illustrating integrity of this reporter system (S1 Fig.). Similar as observed in the IFN luciferase experiments, ectopic expression of BART cluster 1 mirnas in 293T-ISREmCherry cells potently abrogated IFN-α stimulated mcherry-reporter expression at both IFN concentrations tested (Fig. 1e). These results indicate that one or multiple mirnas within the BART cluster 1 interfere(s) with the induction of ISRE promoter activity in response to type I interferon. A B 293T cells 5x ISRE firefly luc D Relative cell number 293T-ISRE-mCherry cells mcherry unstimulated 55 IU/mL IFN-α 11 IU/mL IFN-α Relative luciferase activity (%) IFN-α (5 IU/mL) control BART cluster 1 E Relative cell number Relative cell number 293T-ISRE-mCherry cells IFN-α: IU/mL 15 IU/mL control BART cluster 1 5 C 5x ISRE mcherry Relative cell number 11 IU/mL mcherry Fig 1. EBV mirnas attenuate ISRE promoter activity in response to type I interferon signaling. (a) A reporter encoding firefly luciferase under control of five ISRE promoter elements was used to probe ISRE promoter activity in 293T cells. (b) 293T cells transduced to stably express EBV BART cluster 1 or an empty control vector were transiently transfected with the reporter depicted in (a) (75 ng) and an HSV TK-promoter driven Renilla luciferase plasmid for normalization (4 ng). Following stimulation with 5 IU/mL IFN-α2c for 7 hours, cells were lysed and luciferase activities were determined. Values are presented as percentages firefly luciferase activity relative to stimulated control samples and normalized for transfection efficiency using Renilla luciferase values, mean ± SD of three replicates. Data are representative of three independent experiments. (c) To determine ISRE promoter activity at single cell level, 293T cells were transduced with a reporter encoding mcherry under control of five ISRE promoter elements (293T-ISRE-mCherry). (d) Clonal 293T-ISRE-mCherry cells were either left untreated (dashed line) or stimulated with 55 IU/mL (grey line) or 11 IU/mL (black line) IFN-α2c for 29 hours followed by flow cytometric analysis of mcherry expression. (e) 293T-ISRE-mCherry cells were transduced to express BART cluster 1 (solid line) or an empty control vector (dashed line). Following treatment with, 15, or 11 IU/mL IFN-α2c for 29 hours, mcherry expression levels were determined by flow cytometry. 125

126 Chapter 5 Inhibition of mir-bart16-5p rescues type I interferon signaling To pinpoint the mirnas within BART cluster 1 responsible for the observed inhibition of type I IFN signaling, we used tough decoy (TuD) mirna inhibitors to specifically repress given mirnas by competing with endogenous mirna targets for mirna binding [43]. We constructed twenty TuD mirna inhibitors to specifically interfere with the function of the 5 and 3 arms of the ten EBV mirnas present in BART cluster 1. These mirna inhibitors were introduced into 293T-ISRE-mCherry cells by lentiviral transduction, and their ability to individually abrogate inhibition of IFN signaling caused by BART cluster 1 mirnas was monitored by flow cytometry. Inhibiting mir-bart16-5p most potently rescued the response to IFN-α treatment (Figs. 2a and b). Similarly, inhibiting mir-bart16-5p increased ISREfirefly reporter induction in IFN-α-treated 293T cells expressing BART cluster 1 (Fig. 2c). These results suggest that mir-bart16-5p is the dominant inhibitor of IFNα-mediated ISRE induction within BART cluster 1. Expression of mir-bart16-5p attenuates type I interferon signaling Following the observation that inhibition of mir-bart16-5p relieved BART cluster 1-mediated repression of IFN-α-induced signaling, we investigated the impact of expressing mir-bart16 individually. As we previously observed, introducing BART cluster 1 into 293T-ISRE-mCherry cells reduced reporter induction following IFN-α treatment and expression of mir-bart16 individually caused a similar effect (Fig. 3a). The inhibitory effect of mir-bart16 was less potent than that of the entire mirna cluster (27% inhibition for mir-bart16 alone vs 52% for BART cluster 1), suggesting that additional mirnas within the cluster impact this pathway. To verify the contribution of mir-bart16-5p to the inhibition of IFN signaling, we introduced specific inhibitors into these cells. The suppression of the ISRE-mCherry reporter induction in cells expressing BART cluster 1 as well as mir-bart16 alone was (partially) reversed by the mir-bart16-5p-specific inhibitor (Figs 3b and c). These results indicate that mir-bart16-5p is indeed responsible for attenuating IFN-α-mediated Fig 2. EBV mir-bart16-5p interferes with ISRE promoter activity in response to type I interferon signaling. (a) 293T-ISRE-mCherry cells expressing TuD inhibitors targeting the indicated mirnas were left untreated or treated with 15, 15, or 15 IU/mL IFN-α2c followed by flow cytometric analysis of mcherry levels. Depicted are the ratios of the geometric means of cells with inhibitor versus without inhibitor (averages of two technical replicates ± SD). Data are representative of three independent experiments. (b) 293T-ISRE-mCherry cells without (dashed lines) or with BART cluster 1 were transduced with an empty control vector (grey lines) or inhibitors targeting either mir- BART4-5p, mir-bart16-5p, mir-bart17-3p, or mir-bart18-3p (black lines). The cells were treated with 15 IU/ ml IFN-α2c for 29 hours and mcherry levels were determined by flow cytometry. (c) 293T cells expressing BART cluster 1 were transduced with inhibitors for mir-bart3-3p, mir-bart5-5p, mir-bart15-3p, mir-bart16-5p, mir- BART17-3p, or a control (-). Sixteen hours after transient transfection with a firefly luciferase reporter responsive to ISRE promoter activity (75 ng) and an HSV TK-promoter driven Renilla luciferase plasmid for normalization (4 ng), cells were left untreated (-) or stimulated with 5 IU/mL IFN-α2c (+) for 7 hours. Luciferase activity was determined following cell lysis. Values represent percentages firefly luciferase activity relative to stimulated control samples and normalized for transfection efficiency using Renilla luciferase values, mean ± SD. Data is representative of two independent experiments. 126

127 EBV mirna BART16 suppresses type I IFN signaling Control BART cluster 1 mcherry Inhibitor 5 IU/mL IFN-α p 5-5p 15-3p 16-5p 17-3p Relative cell number Relative cell number Inhibited mirna: BART16-5p BART18-3p Relative cell number Relative cell number Relative luciferase activity (%) BART cluster 1 +, inhibitor + BART cluster 1 +, inhibitor BART cluster 1 - Inhibited mirna: BART4-5p BART17-3p B C Inhibited mirna: - BART1 BART3 BART4 BART5 BART6 BART15 BART16 BART17 BART18 BART21 ISRE-mCherry reporter expression (fold indcution relative to without inhibitor). Strand: 3p 5p 3p 5p 3p 5p 3p 5p 3p 5p 3p 5p 3p 5p 3p 5p 3p 5p 3p 5p.5 1. A IU/mL IFN-α

128 Chapter 5 signaling in these cells. As a more physiological alternative to reporter induction, we assessed the impact of mir- BART16 expression on the induction of two interferon-stimulated genes (IFIT1 and ISG15) that were highly upregulated following IFN-α treatment (Fig. 3d). Protein levels of both IFIT1 and ISG15 levels were reduced in IFN-α-treated cells expressing either BART cluster 1 entirely or mir-bart16 alone as compared to control cells. These combined results demonstrate that mir-bart16 attenuates endogenous ISRE-promoter activity and impairs production of effector ISGs. A mcherry geometric mean normalized to stimulated control C mcherry+ cells normalized to mirna- cells control control mirna- mirna+, inhibitor- mirna+, inhibitor+ - IFN-α + IFN-α Cluster 1 BART16 mirna(s) mirnamirna+, mirna+, inhibitorinhibitor+ Cluster 1 BART16 mirna(s) B D mirna: BART cluster 1 mir-bart control inhibitor control inhibitor BART16-5p inhibitor BART16-5p inhibitor IU/mL IFN-α IFIT1 TFR mcherry+ cells (%) IFN-α (IU/mL) Control Cluster 1 BART16 Control Cluster 1 1% 49% 45% 1% 47% 51% ISG15 1% 44% 41% 1% 57% 64% BART16 Control Cluster 1 BART16 Fig 3. mir-bart16-5p attenuates type I interferon signaling. (a) 293T ISRE-mCherry cells stably expressing BART cluster 1, mir-bart16, or an empty control vector were treated for 3 hours with 1. IU/mL IFN-α2c before flow cytometric analysis of mcherry reporter expression. Geometric mean fluorescence intensities were normalized to stimulated control cells. Data are representative of two independent experiments. (b) 293T ISRE-mCherry cells without (-) or with (+) BART cluster 1 or mir-bart16 were transduced with a mir-bart16-5p or control inhibitor. The cells were treated for 29 hours with 15, 15, 15, or 15. IU/mL IFN-α2c and the percentage of mcherry-positive cells was determined by flow cytometry. Data are representative of two independent experiments. (c) Bar diagrams of cells treated as in (b) with 15 IU/mL IFN-α2c. Percentage of mcherry + cells was normalized to mirna- control cells (white bars). Replicates from two independent experiments. (d) 293T cells expressing a control vector, BART cluster 1, or mir-bart16 alone were treated with, 15, or 15 IU/mL IFN-α for 29 hours. Expression levels of ISGs IFIT1 and ISG15 were determined in post-nuclear lysates by immunoblotting using specific antibodies. Transferrin receptor (TfR) was included as loading control. 128

129 EBV mirna BART16 suppresses type I IFN signaling Microarray analysis identifies potential mir-bart16 targets To identify potential targets of mir-bart16-5p mediating the interference with type I interferon signaling, we performed gene expression profiling on 293T-ISRE-mCherry cells expressing mir-bart16. Since we previously observed low levels of mirnas when expressed individually, we co-expressed mir-bart16 in conjunction with a helper cluster composed of mir-bart19, mir-bart2, mir-bart13 and mir-bart14. This coexpression was found to boost expression of individual BART mirnas (unpublished observations). Expression and functionality of mir-bart16-5p and mir-bart16-3p in these cells was confirmed using mirna reporter cells carrying a perfect mirna binding site downstream of an mcherry reporter gene (S2 Fig.). We profiled mrna expression by whole human genome microarray analysis on RNA extracts from these cells and control cells only expressing the helper cluster (Fig. 4a). Downregulation of the bona fide mir-bart16-5p target TOMM22 in the BART16 expressing cells (~two-fold, p =.1) supports the validity of our approach. We ranked the genes downregulated in the presence of mir-bart16 according to p-value (S2 Table), and cross-referenced the top-5 genes to a list of genes implicated in IFN signaling extracted from the KEGG (JAK-STAT signaling) and Reactome (interferon signaling) pathway databases (299 genes in total). The histone acetyltransferase (HAT) CBP (CREB binding protein or CREBBP) was the only gene involved in interferon signaling that was downregulated in the presence of mir-bart16 (1.4 fold downregulated, p =.65) (Fig. 4b). CBP acts as co-activator for various transcription factors and has been implicated in interferon induced gene transcription. Besides CBP, its paralog p3 and the HAT GCN5 regulate this process [44,45]. To assess whether downregulation of CBP alone impacts IFN signaling in 293T-ISRE-mCherry cells, we functionally knocked out the CBP gene in these cells using the CRISPR-Cas9 system and monitored responsiveness to IFN-α exposure. Indeed, CBP knockout cells demonstrated reduced reporter expression in response to IFN-α treatment, as did knockout cells lacking functional p3 or IFNAR2 expression (Fig. 4c displaying polyclonal knockout cells and S3a and b Figs. displaying clonal knockout cells). Immunoblotting confirmed reduced CBP expression levels in the knockout cells (Fig. 4d). These results identify CBP as a potential target for mir-bart16-5p, and show that CBP downregulation alone is sufficient to impair IFN-α-induced gene transcription. 5 Inhibition of CREB-binding protein expression by mir-bart16-5p To investigate targeting of CBP by BART16-5p in more detail, we searched for potential mirna target sites in the 3 UTR of the CBP mrna sequence. Indeed, an 8-mer mir-bart16-5p seed match was present in the 3 UTR of CBP (Fig. 5a). To determine whether the this 3 UTR is a direct target for mir-bart16-5p, we cloned this noncoding region downstream of an mcherry reporter gene (Fig. 5b) and monitored its regulation by mir-bart16 (Fig 5c, left and middle columns). Compared to control cells (upper row), expression of either BART 129

130 Chapter 5 A.1 B TOMM22 p-value.1.1 CBP Microarray BART16 Top JAK-STAT and IFN signaling 299 CBP -1 1 C Targeted gene CBP Relative cell number Fold change in expression (log2) 293T-ISRE-mCherry cells #1 #2 Relative cell number grna - IFN-α control + IFN-α grna + IFN-α D grna Con CBP #1 #2 CBP TfR p3 Relative cell number #1 #2 Relative cell number controls Relative cell number non-targeting Relative cell number IFNAR2 mcherry Fig 4. Microarray expression profiling identifies CBP as target for mir-bart16. (a) 293T cells were transduced to stably express mir-bart16 together with the helper cluster or the helper cluster alone (control cells). mrna expression levels of mir-bart16-expressing cells were compared to control cells by whole genome microarray analysis. The Volcano plot depicts fold changes (x-axis, log2) versus p-values (y, axis, -log1). Each dot represents one microarray probe, dots representing CREBB-binding protein (CBP) and established mir-bart16 target protein TOMM22 are indicated. (b) The mir-bart16 targets derived from the microarray analysis (based on p-value) were crossreferenced with genes implicated in JAK/STAT (KEGG database) and IFN signaling (Reactome) databases. (c) CBP or its paralog p3 were functionally deleted from 293T-ISRE-mCherry cells using the CRISPR/Cas9 genome editing system. Cells transduced with a non-targeting grna (grey lines) or a grna targeting the indicated genes were untreated (dashed lines) or stimulated for 29 hours with 15 IU/mL IFN-α2c (black lines), followed by analysis of mcherry expression levels by flow cytometry. Bottom row shows results with a non-targeting grna vector (negative control) or grna targeting IFNAR2 (positive control). (d) Immunoblot showing expression levels of CBP in the cells used in (c). Transferrin receptor (TfR) was used as loading control. Remaining CBP protein is derived from a minority of the polyclonal cells still expressing detectable CBP. 13

131 EBV mirna BART16 suppresses type I IFN signaling cluster 1 (middle row) or mir-bart16 alone (lower row) downregulated reporter levels, indicating that mir-bart16 indeed targets the 3 UTR of CBP. Mutating four nucleotides in the seed-region (Fig. 5a) completely abolished reporter downregulation by both mir-bart16 as well as BART cluster 1, confirming the predicted mir-bart16-5p binding site (Fig. 5c, right column). In line with these results, CBP levels were indeed reduced upon expressing BART cluster 1 or mir-bart16 alone compared to mirna-negative control cells, as shown by immunoblot analysis (Fig 5d). We extended these results to endogenously expressed mir-bart16 during EBV infection by evaluating CBP levels in the EBV-transformed Burkitt s lymphoma cell line Jijoye. Functionally inhibiting mir-bart16-5p in these cells caused an upregulation of CBP protein levels, demonstrating that CBP is actively downregulated by endogenously expressed EBV mir-bart16-5p during latent EBV infection (Fig. 5e and S4 Fig.). Taken together, we show that EBV mir-bart16-5p directly targets the 3 UTR of CBP transcripts and attenuates IFN-α-induced activation of ISRE promoters as well as production of effector ISGs. Discussion Eluding host immune responses is thought to be crucial for the establishment and maintenance of persistent infections by human herpesviruses in the face of a functional human immune system. Here, we revealed a novel function for EBV-encoded mirnas and report that BART cluster 1 mirnas interfere with type I IFN-induced signaling pathways. Using a TuD mirna-inhibitor screen, mir-bart16-5p was identified as a dominant contributor to this effect. Indeed, mir-bart16 expression interferes with ISRE promoter activity and ISG production following IFN-α treatment. mir-bart16 downregulates levels of cellular histone acetylase CBP by directly targeting the 3 UTR of transcripts encoding this protein. Notably, inhibiting endogenously expressed mir-bart16-5p in EBV-transformed Burkitt s lymphoma cells (Jijoye) rescues CBP expression in these EBV-infected cells. 5 EBV comes into contact with the host immune system during various stages of the viral life cycle. B cells, the primary target cells of EBV, express several PRRs that sense EBV components and are able to induce type I IFN production. The adverse effects of IFN treatment during the early phases of infection on establishment of infection suggest that interference with host IFNs is important during latent infection [46]. Inhibiting IFN-mediated antiviral immunity by mir- BART16 may thus favor establishment of latency and viral replication in these cells [1,47]. In addition, pdcs are a major source of type I IFNs and have proven essential in the control of EBV infection in a humanized mouse model [7,14,48]. Although not typically infected by EBV, IFN-mediated effects may be inhibited in these cells through secretion of mirna-containing 131

132 Chapter 5 exosomes from EBV-infected cells [49,5]. Acute EBV infection has recently been reported to induce a compromised systemic IFN response lacking expression of a subset of interferonstimulated genes [6]. It is tempting to speculate that IFN-evasive viral mechanisms, including mir-bart16-5p, contribute to this impaired IFN response. CBP is a key component in the IFN-induced signaling pathway leading to ISG production [51 53]. Downregulation of CBP through the actions of mir-bart16-5p may thus provide an elegant way for EBV to interfere with IFN-induced antiviral responses in infected cells. In addition, a RISC-IP experiment using EBV + primary effusion lymphoma cells identified target sites for both mir-bart16-5p and mir-bart18-5p in the CBP 3 UTR [54]. In line with this, EBV mir-bart18 has recently been implicated in interference with CBP function [55]. Furthermore, a bioinformatics approach suggests that CBP may in addition function as a potential target for the EBV BHRF1 mirnas [56]. Our data support that additional EBV mirnas interfere with IFN-induced effect, as we observed that combined expression of all BART cluster 1 mirnas (including mir-bart18) induced a more potent abrogation of IFN signaling as compared to expression of mir-bart16 alone (Fig. 3). On the other hand, we did not observe a significant rescue of IFN signaling upon mir-bart18 inhibition (Fig. 2a) and altering the mir-bart16-5p target site in the 3 UTR of the mcherry reporter (while leaving the putative mir-bart18 site intact) completely restored reporter expression in the presence of BART cluster 1 (Fig. 5c). Nevertheless, it is striking that EBV encodes at least two and potentially more different mirnas to target CBP and interfere with IFN signaling. It will be interesting to determine whether these mirnas act synergistically to robustly reduce CBP levels during EBV infection, or whether they perform their functions at different stages of the viral life cycle. EBV is not the only herpesvirus interfering with CBP function: the interference of several other (herpes-)viruses with the function of CBP and/or its paralog p3 support the emergence of CBP as an important target for viral immune evasion strategies. For example, the human herpesviruses HSV-1 and KSHV inhibit recruitment of CBP to IRF-3 [57,58] and HSV-1, HCMV, and KSHV induce expression of mir-132, a cellular mirna that downregulates p3 [59]. Besides mir-bart16-mediated CBP downregulation, EBV has adopted additional strategies to interfere with IFN-induced antiviral responses. Components of the IFN-induced signaling pathway are targeted by latency-associated EBV gene products interfering LMP2, EBNA2, EBERs [23,6,61]. Lytic proteins BZLF1, BRLF1, LF2, BARF1, and BGLF5 act at an earlier step and interfere with the production of IFNs directly [18 22,24]. The dedication of several EBV gene products to attenuation of IFN-induced effect suggests that this is an important feature for successful completion of the viral life cycle. How the different IFN-evasion strategies cooperate during natural infection remains to be established. In this respect it is noteworthy 132

133 EBV mirna BART16 suppresses type I IFN signaling A B mir-bart16-5p 3 UTR CBP wild type 3 UTR CBP mutant 3' UCUC U 5' GUGUGUGGGUG AGAUAGAU UACACACCCAC UCUAUCUA 5' CACG ACACA U 3' 5' CACGUACACACCCACACACAUGAAAGUAU 3' mcherry 3 UTR CBP C control 3 UTR CBP wt 3 UTR CBP mut mirna - Relative cell number Relative cell number Relative cell number control mirna+ Relative cell number Relative cell number Relative cell number mir-bart16 Relative cell number Relative cell number mcherry Relative cell number BART cluster 1 5 D 293T cells E EBV + Jijoye cells mirna: ctrl cluster B16 inhibitor: ctrl B16-5p CBP CBP TFR TFR Fig 5. mir-bart16-5p targets CREB-binding protein (CBP). (a) Alignment of mir-bart16-5p with its predicted target site in the 3 UTR of CBP. The bottom line shows the mutated 3 UTR sequence with altered nucleotides underlined. (b+c) 293T-ISRE-mCherry cells were transduced with (from top to bottom) an empty control vector, mir-bart16, or BART cluster 1, together with (from left to right) an empty mcherry control vector, an mcherry reporter containing the wild type 3 UTR of CBP, or the mutated 3 UTR variant depicted in (a). mcherry expression was analyzed by flow cytometry and graphs are shown for mirna- (GFP - ) or mirna + (GFP + ) cells. (d+e) Immunoblots showing CBP levels in (d) 293T-ISRE-mCherry cells transduced with a control vector, mir-bart16, or BART cluster 1; or in (e) EBV + BL cell line Jijoye transduced with empty vector or an inhibitor targeting mir-bart16-5p. Transferrin receptor (TfR) served as loading control. 133

134 Chapter 5 that the EBV-miRNAs including mir-bart16 are expressed during all stages of EBV infection, while protein expression is severely limited during the latent phases infection. This suggests that mir-bart16-mediated immune evasion may be most prominent during latency and in particular latency I, when protein expression is restricted to EBNA1. Besides the role of CBP and its paralog p3 in IFN signaling, these proteins interact with hundreds of transcription factors involved in many signaling pathways, including NF-κB and the IRFs [62 64]. By targeting CBP, EBV may thus simultaneously inhibit IFN production and the effector response. In line with this, our preliminary data suggest that BART cluster 1 expression also interferes with IRF and NF-κB transcriptional activity (data not shown). Rhesus lymphocryptovirus (rlcv) belongs to the same genus as EBV and is frequently used in rhesus macaques as a model system for in vivo lymphocryptovirus infection [65]. Not all mirnas are conserved between EBV and rlcv, yet a mir-bart16 homolog is encoded in the rlcv genome (mir-rl1-19; S5a Fig.) [66,67]. In addition, a potential target site of this mirna is present downstream of the rhesus macaque CBP gene (S5b, c Fig.), indicating that a CBP-targeting mirna may already have existed in a common ancestor. Furthermore, HCMV and KSHV mirnas were recently found to target CBP and interfere with IFN-mediated host responses [55]. The functional convergent evolution of viral mirnas interfering with CBP function among divergent herpesvirus members underscores that evasion of IFN-induced antiviral effects is important during herpesvirus infection and provides further evidence for the importance of CBP as a viral immune evasion target. In conclusion, we implicated EBV-encoded mirnas in thwarting the host antiviral IFNresponses. Identification of the elusive functions of viral mirnas provides novel insight into the elaborate systems EBV and herpesviruses in general have adopted to successfully establish infection and replicate in the presence of a fully competent immune system. Irregularities in the mutual immune interactions between the virus and the host are suggested to be causally involved in IM and EBV-associated malignancies. mir-bart16-mediated interference with immune responses may thus contribute to the maintenance of EBV-associated tumors, in which they are ubiquitously expressed. Acknowledgments This work was supported by Veni grant from The Netherlands Organisation for Scientific Research (NWO) and Marie Curie Career Integration Grant PCIG-GA to R.J.L. M.E.R. was financially supported by Vidi grant from the Netherlands Scientific 134

135 EBV mirna BART16 suppresses type I IFN signaling Organization. E.J.H.J.W., M.E.R. and R.J.L were supported by grant UU from the Dutch Cancer Society (KWF). Materials and methods Plasmids EBV mirna-encoding sequences were amplified by PCR from genomic DNA isolated from EBV + Akata cells using the oligonucleotides indicated in S1 Table. A stretch of ~25 nucleotides encompassing mir-bart16 or two separately amplified stretches encoding EBV mirnas mir-bart3-6 (BART3, 4, 1, 15, 5, 16, 17 and 6) and mir-bart21-18 (together designated as BART cluster 1) were joined in an intron introduced into GFP in the lentiviral expression vector pcclsin. PPT.pA.CTE.4xscrT.eGFP.mCMV.hPGK.NGFR. To boost mirna expression in the microarray gene expression analysis, the BART16 pre-mirna sequence was joined to a helper cluster containing mir-bart19, 2, 13 and 14 (not present in BART cluster 1). Tough decoy (TuD) inhibitors were ordered as oligonucleotides (S1 Table) and cloned downstream a mouse U6 promoter in the psicor-ef1a-zeor-t2a-mametrine vector [68], which was derived from the psicor- EF1A-PuroR-T2A-mCherry vector in which the PuroR-T2A-mCherry cassette was replaced by a ZeoR-T2A-mAmetrine cassette. To construct the lentiviral ISRE-mCherry reporter plasmid the promoter region in front of mcherry of psicor- EF1A-puro-T2a-mCherry was replaced by the ISRE-containing promoter from the pgl3-isre-ta-luciferase reporter. The mcherry-mirna reporters were generated by introducing a short spacer and a single perfect mirna binding site downstream of the mcherry reporter gene in the psicor-ef1a-puror-t2a-mcherry vector. For the mcherry reporter containing the 3 UTR of CBP, a 9 nt part of the CBP 3 UTR (primers listed in S1 Table) was amplified from human HK-1 cell-derived cdna and cloned into the psicor-ef1a-puror-t2a-mcherry vector downstream of mcherry. Mutations in the 3 UTR were introduced by site-directed mutagenesis using the primers listed in S1 Table. For CRISPR-Cas9 genome editing, Cas9 was expressed from an EF1a-puroR-T2A-NLS-hSpCas9 expression cassette in the psicor vector (Jacks Lab, MIT). GuideRNAs (grnas) were cloned downstream of the U6 promoter in the same vector. The pgl3-isre-ta-luciferase reporter, containing five ISRE repeats (GAAACTGAAACT) and an HSV-TK TATA box was a gift from S. Persengiev (Division of Biomedical Genetics, UMC Utrecht, The Netherlands). To construct the lentiviral ISRE-mCherry reporter plasmid the promoter region in front of mcherry of psicor-ef1a-puro-t2a-mcherry was replaced by the ISRE-containing promoter from the pgl3-isre-ta-luciferase reporter. All constructs were verified by Sanger sequencing. Cells EBV + BL Jijoye and HEK293T cells were obtained from American Type Culture Collection (ATCC). Jijoye cells were maintained in Roswell Park Memorial Institute medium (RPMI 164; Life Technologies) and 293T cells in Dulbecco s modified Eagle medium (DMEM; Life Technologies), both supplemented with 1% FCS (Sigma), 2 mm L-glutamine, 1 U/mL penicillin, and 1 mg/ml streptomycin. 293T-ISRE-mCherry reporter cells were generated by lentivirus transduction of 293T cells with psicor-isre-mcherry, single clones were prepared by limiting dilution. 5 Antibodies and reagents Primary antibodies used for immunoblotting were: rabbit anti-cbp (clone A-22, sc-369, Santa Cruz Biotechnology), mouse anti-transferrin receptor (H68.4, 13 68, Life Technologies), rabbit-anti-isg15 (clone 22D2, #2758, Cell Signaling Technology), and rabbit anti-ifit1 (N2C3, GeneTex). Secondary antibodies used were goat anti-mouse IgG (H+L)-HRP (# , Bio-rad) and goat anti-rabbit IgG-HRP (43-5, Southern Biotech). IFN-α2c was obtained from ebioscience (BMS25). Lentivirus transductions Lentiviruses were produced via standard lentiviral production protocols using self-inactivating third-generation packaging vectors. Lentiviral transductions of target cells were performed by spin infections at 1,g for 1.5 hrs at 33 C in the presence of 4 μg/ml polybrene. For the experiments shown in Fig. 1, 2b, 3d and 5a high GFP expressing cells were obtained by fluorescence-activated cell sorting on a BD FACS AriaII. 135

136 Chapter 5 Generation of CRISPR-Cas9 knock-out cells 293T-ISRE-mCherry cells were transfected using Lipofectamine 2 (Life Technologies) with the CRISPR-Cas9 vector expressing both Cas9 and a grna targeting a gene of interest. The cells were cultured in medium containing 2 µg/ml puromycin from one until three days post transfection. Three weeks post transfection, cells were stimulated with IFNα2c at the indicated amounts and durations followed by analysis of mcherry expression by flow cytometry. To generate clonal 293T-ISRE-mCherry knockout cells, 293T-ISRE-mCherry cells were transduced with grna lentiviruses, selected by puromycin and subsequently cloned by limiting dilution. Luciferase reporter assay Cells were transfected in suspension in 96-well flat-bottom plates using Lipofectamine 2 (Invitrogen) according to the manufacturer s instructions. 293T cells (5 x 1 4 per well) were transfected with a total of 2 ng DNA containing 75 ng pisre-firefly luciferase reporter plasmid together with 4 ng of HSV thymidine kinase-driven Renilla luciferase for normalization (pgl3-tk vector), supplemented with empty vector. Starting 16 hours after transfection, cells were stimulated with a final concentration of 5 IU/mL IFN-α2c for 7-9 hours. Following lysis in Passive Lysis Buffer, firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System according to manufacturer s protocol (Promega). Data were normalized for Renilla luciferase activity and presented as percentage firefly luciferase activity relative to the stimulated control sample (=1%), mean ± S.D. 293-ISRE-mCherry reporter assay 293-ISRE-mCherry cells were seeded in 96 well plates. After 24 hours, cells were treated with IFN-α2c at the indicated concentrations for 29 hours, followed by flow cytometric analysis of mcherry expression. Cells were gated for GFP (mirna + ) and mametrine (TuD inhibitor + ) where applicable. Flow cytometry For flow cytometric analysis of fluorescent proteins, cells were fixed with PBS/.5% BSA/.2% sodium azide/1% formaldehyde. Data were acquired on a FACSCanto II equipped with 45, 488, 561 nm laser sources (BD Bioscience) and analyzed using FlowJo software (Treestar). Immunoblotting Cells were lysed in 1% Triton lysis buffer (1% Triton X-1, 5 mm Tris HCl [ph 7.5], 15 mm NaCl, 1 µm Leupeptin (Roche), and 1 mm 4-(2-aminoethyl)benzenesulfonyl fluoride) (Roche) or Radioimmunoprecipitation assay buffer (1% Triton X-1,.5% sodium deoxycholate,.5% SDS, 5 mm Tris HCl [ph 8.], 15 mm NaCl, µm Leupeptin, and 1 mm 4-(2-aminoethyl)benzenesulfonyl fluoride). Lysates generated to detect CBP were sonicated. Subsequently, proteins were denaturated with reducing sample buffer, seperated by SDS-PAGE ( %), and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore). Proteins of interest were detected by incubating the membranes with specific primary antibodies and HRP-conjugated secondary antibodies, followed by visualization using ECL detection kit (Thermo Scientific Pierce) and Amersham Hyperfilm ECL films (GE Healthcare) or ImageQuant LAS 4 imager (GE Healthcare Life Sciences). Densitometric quantification of staining intensity was performed using Quantity One software (Bio-Rad). Microarray analysis 293T ISRE-mCherry cells were lentivirally transduced to coexpress mir-bart16 and the mir-bart19-14 helper cluster or the helper cluster alone as control. Transduced cells were subsequently selected by blasticidin for two weeks. RNA was isolated using Trizol (Life Technologies) and purified using the RNeasy Mini kit according to manufacturer s instructions (Qiagen). Following treatment with DNaseI (Qiagen), RNA quality was analyzed using Agilent 21 Bioanalyzer. Microarrays used were human whole genome gene expression microarrays V2 (Agilent, Belgium) representing H. sapiens 6-mer probes in a 4x44K layout (probe sequences from this array were re-annotated by BLAST-searching against genome build version 71_37 at ENSEMBL). Sample versus common reference hybridizations were performed in balanced dye-swap on biological replicates. cdna synthesis, crna amplification, labeling, quantification, quality control and fragmentation were performed with an automated system (Caliper Life Sciences NV/SA, Belgium), starting with 3 μg total RNA from each sample, as previously described in detail [69]. Microarray hybridization and washing was with a HS48PRO system with QuadChambers (Tecan, Benelux) using 1 ng, 1.3-3% Cy5/Cy3 labeled crna 136

137 EBV mirna BART16 suppresses type I IFN signaling per channel as described [69]. Slides were scanned on an Agilent G2565BA scanner at 1% laser power, 3% PMT. After automated data extraction using Imagene 8. (BioDiscovery), Loess normalization was performed on mean spot-intensities [7]. Gene-specific dye bias was corrected by a within-set estimate [71]. Data were further analyzed by MAANOVA [72], modeling sample, array and dye effects in a fixed effect analysis. P-values were determined by a permutation F2-test, in which residuals were shuffled 1 times globally. After false discovery rate determination (FDR by Benjamini-Hochberg) gene probes were ranked according to p-value (S2 Table) to cross-reference the top-5 genes to a list of genes implicated in IFN signaling (extracted from the KEGG (JAK-STAT signaling; kegg) and Reactome (interferon signaling; pathway databases (299 genes in total)). Supporting information grna: IFNAR2 IRF9 non-targeting #1 #1 Relative cell number Relative cell number Relative cell number Relative cell number #2 #2 Relative cell number grna empty vector grna x- IFN-α + IFN-α + IFN-α mcherry S1 Fig. Knockout of IFNAR2 and IRF9 confirms that ISRE-mCherry-reporter depends on canonical type I interferon signaling pathway. 293T-ISRE-mCherry cells were lentivirally transduced with a control vector (dashed line) or CRISPR grnas targeting IFNAR2 or IRF9 (two grnas per gene). Transduced cells were selected using puromycin for two weeks, followed by stimulation with 15 IU/mL IFN-α2c for 29 hours (black bars) and analyzed by flow cytometry. Grey lines, untreated control cells. 5 Relative mirna reporter expression (% compared to mirna - cells) 1 5 mirna: mirna reporter: control BART16 BART16-5p control BART16 BART16-3p S2 Fig. mir-bart16-5p is functionally expressed in transduced 293T-ISRE-mCherry cells. 293T ISRE-mCherry cells were lentivirally transduced with mir-bart16 together with a helper cluster consisting of mir-bart19-14 or helper cluster alone (control) and selected using blasticidin. Subsequently, a reporter containing a perfect mirna binding site for mir-bart16-5p or mir-bart16-3p downstream of mcherry was introduced and mcherry levels were determined by flow cytometry five days post transduction. Bars depict the percentage mirna reporter expression (geometric mean) in mirna + (GFP + ) cells versus mirna - (GFP - ), average values of two independent experiments ± SD. 137

138 Chapter 5 A CBP * * TfR con CRISPR grna CBP#1 clone 1 clone 2 clone 3 clone 4 clone 5 B Relative cell number 1 CRISPR grna empty vector unstimulated CRISPR grna empty vector +IFN-α CRISPR grna +IFN-α mcherry S3 Fig. Type I interferon response is suppressed in CBP knockout cell lines. (a) 293T-ISRE-mCherry cells were transduced with CRISPR grna for CBP or contol vector, selected using puromycin, and subcloned by limiting dilution. CBP levels in lysates from five different clones were analyzed by immunoblotting and displayed loss of CBP from these cells. Two background bands are marked with an asterisk. (b) Cells described in (a) were left untreated (dashed lines) or stimulated for 29 hours with 15 IU/mL IFN-α2c (black lines) and mcherry levels were determined by flow cytometry. Four out of five CBP knockout clones displayed impaired ISRE reporter induction. Relative cell number 16-5p reporter empty vector mir-bart16-5p inhibitor Relative cell number control reporter mcherry S4 Fig. mir-bart-16-5p inhibitor inhibits mir-bart16-5p in Jijoye cells. EBV + Jijoye cells were lentivirally transduced to express a mir-bart16-5p inhibitor (TuD) or control vector. Following selection with zeocin, these cells were transduced with a mirna reporter containing a perfect mirna binding site for mir-bart16-5p or control sequence downstream of an mcherry reporter gene. mcherry expression levels were determined by flow cytometry. 138

139 EBV mirna BART16 suppresses type I IFN signaling A ebv-mir-bart16-5p rlcv-mir-rl1-19-5p UUAGAUAGAGUGGGUGUGUGCUCU UAUAGAUAGCGUGGGUGUGUGA B mfe: kcal/mol rhesus macaque 3 UTR CBP rlcv-mir-rl1-19-5p 5' A ACACAU C 3' CACACACCCAC CUAUCUAUA GUGUGUGGGUG GAUAGAUAU 3' A C 5' C human rhesus macaque CGUA CACACCCACACACAUCUAUCUAUAC CGUACACACACCCACACACAUCUAUCUAUAC S5 Fig. Comparison of EBV/human versus rlcv/rhesus macaque sequences. (a) Alignment of EBV mir-bart16-5p and rhesus lymphocryptovirus mirna rlcv-mir-rl1-19-5p. (b) Genomic region downstream of the rhesus macaque CBP homolog stop codon was in silico hybridized to rlcv-mir-rl1-19-5p using RNAhybrid. The minimum free energy (mfe) of the energetically most favorable region was calculated. For hybridization of mir-bart16-5p to the human CBP 3 UTR, see Fig. 5c. (c) Alignment of a fraction of the human CBP 3 UTR and the homologous rhesus macaque genome sequence. Nucleotides with predicted interaction with mir- BART16-5p or rlcv-mir-rl1-19-5p are underlined. 5 Continues on next page 139

140 Chapter 5 S1 Table. Oligonucleotides overview. Sequences of oligonucleotides used for cloning. Name Sequence grna sequences for CRISPR/Cas-mediated gene disruption (target sequence underlined) grna_cbp#1 GTGGAAAGGACGAAACACCGAACTGTCGGAGCTTCTACGGTTTTAGAGCTAGAAATAGC grna_cbp#2 GTGGAAAGGACGAAACACCGAATCACATGACGCATTGTCGTTTTAGAGCTAGAAATAGC grna_ifnar2#1 GTGGAAAGGACGAAACACCGTGTATATCAGCCTCGTGTTGTTTTAGAGCTAGAAATAGC grna_ifnar2#2 GTGGAAAGGACGAAACACCGTTCCGGTCCATCTTATCATGTTTTAGAGCTAGAAATAGC grna_ep3#1 GTGGAAAGGACGAAACACCGATCGCTGGCGGACGCCGAGAGTTTTAGAGCTAGAAATAGC grna_ep3#2 GTGGAAAGGACGAAACACCGATGAGAGTTTAGGCCGCTGTTTTAGAGCTAGAAATAGC grna_irf9#1 GTGGAAAGGACGAAACACCGCACCCGAAAACTCCGGAACGTTTTAGAGCTAGAAATAG grna_irf9#2 GTGGAAAGGACGAAACACCGACCATGTTCCGGATTCCCGTTTTAGAGCTAGAAATAG grna_non-targeting GTGGAAAGGACGAAACACCGCCAAAATCTGTGCCATCGCGTTTTAGAGCTAGAAATAGC PCR primers for cloning CREBBP 3 UTR and mutagenesis UTR_CBP_FW CGCGGATCCCATTGTGAGAGCATCACCTTTTC UTR_CBP_RV TGCTCTAGATATGATATAAGAAATGAGTTGG UTR_CBP_mut_FW CCACACACATGAAAGTATACATAATG UTR_CBP_mut_RV CATTATGTATACTTTCATGTGTGTGG PCR primers for cloning BART mirnas BARTcluster1_firstpartFW1 CGGCTAGCCGCTTTAGTGGGAACTAGTGGGAG BARTcluster1_firstpartRV1 CCTTAATTAAGGCTATAGTGAGCCAGGCTTTC BARTcluster1_secondpartFW2 CCTTAATTAAGGCAAGAGCTCCAGGATGTTAGC BARTcluster1_secondpartRV2 CCTTAATTAAGGGCACTACATTGTCTGGCTTAAGG BART19_FW CTAGCTAGCTAGGTGCCAATAACCGGCCCG BART14_RV CCTTAATTAAGGGAAAGGCCTGCTGTCATCGG Bart16_FW CTAGCTAGCTAGTATCGCTGGAAACCGGTGG Bart16_RV CCTTAATTAAGGACTGGGTAATAACCTTGTATG PCR primers for cloning ISRE promoter in mcherry reporter vector ISREclone_FW TACGCGTGCTAGCGAAACTG ISREclone_RV ACATGCATGCCGGAATGCCAAGCTTCCATTA Oligonucleotides for cloning TuDs TuDBart1-3p_sense ATCAACGACATAGTGGATAAACAGCGGTGCTACAAGTATTCTGGTCACAGAATACAACGACATAGTGGATAAACAGCGGTGCTA TuDBart1-3p_antisense CTTGTAGCACCGCTGTTTATCCACTATGTCGTTGTATTCTGTGACCAGAATACTTGTAGCACCGCTGTTTATCCACTATGTCGT TuDBart1-5p_sense ATCAACCACAGCACGTCACTACCATCCACTAAGACAAGTATTCTGGTCACAGAATACAACCACAGCACGTCACTACCATCCACTAAGA TuDBart1-5p_antisense CTTGTCTTAGTGGATGGTAGTGACGTGCTGTGGTTGTATTCTGTGACCAGAATACTTGTCTTAGTGGATGGTAGTGACGTGCTGTGGT TuDBart3-3p_sense ATCAACACACCTGGTGACGGAATAGTGGTGCGCAAGTATTCTGGTCACAGAATACAACACACCTGGTGACGGAATAGTGGTGCG TuDBart3-3p_antisense CTTGCGCACCACTATTCCGTCACCAGGTGTGTTGTATTCTGTGACCAGAATACTTGCGCACCACTATTCCGTCACCAGGTGTGT TuDBart3-5p_sense ATCAACAGCACAACACTACGCAACACTAGGTCAAGTATTCTGGTCACAGAATACAACAGCACAACACTGCACAACACTAGGT TuDBart3-5p_antisense CTTGACCTAGTGTTGTGTAGTGTTGTGCTGTTGTATTCTGTGACCAGAATACTTGACCTAGTGTTGCGTAGTGTTGTGCTGT TuDBart4-3p_sense ATCAACACACCTGGTGCCTAACCACGTGATGTGCAAGTATTCTGGTCACAGAATACAACACACCTGGTGCCTAACCACGTGATGTG TuDBart4-3p_antisense CTTGCACATCACGTGGTTAGGCACCAGGTGTGTTGTATTCTGTGACCAGAATACTTGCACATCACGTGGTTAGGCACCAGGTGTGT TuDBart4-5p_sense ATCAACAGCACACCAGCATTCCGCATCAGGTCCAAGTATTCTGGTCACAGAATACAACAGCACACCAGCATTCCGCATCAGGTC TuDBart4-5p_antisense CTTGGACCTGATGCGGAATGCTGGTGTGCTGTTGTATTCTGTGACCAGAATACTTGGACCTGATGCGGAATGCTGGTGTGCTGT 14

141 EBV mirna BART16 suppresses type I IFN signaling Name Sequence Oligonucleotides for cloning TuDs (continued) TuDBart5-3p_sense ATCAACTTTAGGTGAACTTTTAGCGGCCCACCAAGTATTCTGGTCACAGAATACAACTTTAGGTGAACTTTTAGCGGCCCAC TuDBart5-3p_antisense CTTGGTGGGCCGCTAAAAGTTCACCTAAAGTTGTATTCTGTGACCAGAATACTTGGTGGGCCGCTAAAAGTTCACCTAAAGT TuDBart5-5p_sense ATCAACCGATGGGCAGCTATGAACATTCACCTTGCAAGTATTCTGGTCACAGAATACAACCGATGGGCAGCTATGAACATTCACCTTG TuDBart5-5p_antisense CTTGCAAGGTGAATGTTCATAGCTGCCCATCGGTTGTATTCTGTGACCAGAATACTTGCAAGGTGAATGTTCATAGCTGCCCATCGGT TuDBart6-3p_sense ATCAACTCTAAGGCTAGTAAGGCCGATCCCCGCAAGTATTCTGGTCACAGAATACAACTCTAAGGCTAGTAAGGCCGATCCCCG TuDBart6-3p_antisense CTTGCGGGGATCGGCCTTACTAGCCTTAGAGTTGTATTCTGTGACCAGAATACTTGCGGGGATCGGCCTTACTAGCCTTAGAGT TuDBart6-5p_sense ATCAACCCTATGGATTGGCTCTACCAACCTTACAAGTATTCTGGTCACAGAATACAACCCTATGGATTGGCTCTACCAACCTTA TuDBart6-5p_antisense CTTGTAAGGTTGGTAGAGCCAATCCATAGGGTTGTATTCTGTGACCAGAATACTTGTAAGGTTGGTAGAGCCAATCCATAGGGT TuDBart15-3p_sense ATCAACTCAAGGAAACAATCCTAACCACTGACCAAGTATTCTGGTCACAGAATACAACTCAAGGAAACAATCCTAACCACTGAC TuDBart15-3p_antisense CTTGGTCAGTGGTTAGGATTGTTTCCTTGAGTTGTATTCTGTGACCAGAATACTTGGTCAGTGGTTAGGATTGTTTCCTTGAGT TuDBart15-5p_sense ATCAACGACTTCAGGTGGTCTTTTATGTTTCCCTCAAGTATTCTGGTCACAGAATACAACGACTTCAGGTGGTCTTTTATGTTTCCCT TuDBart15-5p_antisense CTTGAGGGAAACATAAAAGACCACCTGAAGTCGTTGTATTCTGTGACCAGAATACTTGAGGGAAACATAAAAGACCACCTGAAGTCGT TuDBart16-3p_sense ATCAACATATGGATAGAAAAAGGGTGGTGATCAAGTATTCTGGTCACAGAATACAACATATGGATAGAAAAAGGGTGGTGAT TuDBart16-3p_antisense CTTGATCACCACCCTTTTTCTATCCATATGTTGTATTCTGTGACCAGAATACTTGATCACCACCCTTTTTCTATCCATATGT TuDBart16-5p_sense ATCAACAGAGCACACACCCACCGACTCTATCTAACAAGTATTCTGGTCACAGAATACAACAGAGCACACACCCACTTGCTCTATCTAA TuDBart16-5p_antisense CTTGTTAGATAGAGCAAGTGGGTGTGTGCTCTGTTGTATTCTGTGACCAGAATACTTGTTAGATAGAGTCGGTGGGTGTGTGCTCTGT TuDBart17-3p_sense ATCAACACTAAGGGGACACGAAGCAGGCATACACAAGTATTCTGGTCACAGAATACAACACTAAGGGGACACGAAGCAGGCATACA TuDBart17-3p_antisense CTTGTGTATGCCTGCTTCGTGTCCCCTTAGTGTTGTATTCTGTGACCAGAATACTTGTGTATGCCTGCTTCGTGTCCCCTTAGTGT TuDBart17-5p_sense ATCAACCTTGTATGCCTGAAAACGTCCTCTTACAAGTATTCTGGTCACAGAATACAACCTTGTATGCCTGAAAACGTCCTCTTA TuDBart17-5p_antisense CTTGTAAGAGGACGTTTTCAGGCATACAAGGTTGTATTCTGTGACCAGAATACTTGTAAGAGGACGTTTTCAGGCATACAAGGT TuDBart18-3p_sense ATCAACGACGAAGCCCAAAATTACTTCCGATACAAGTATTCTGGTCACAGAATACAACGACGAAGCCCAAAATTACTTCCGATA TuDBart18-3p_antisense CTTGTATCGGAAGTAATTTTGGGCTTCGTCGTTGTATTCTGTGACCAGAATACTTGTATCGGAAGTAATTTTGGGCTTCGTCGT TuDBart18-5p_sense ATCAACTGTATAGGAAGTAAAAGCGAACTTGACAAGTATTCTGGTCACAGAATACAACTGTATAGGAAGTAAAAGCGAACTTGA TuDBart18-5p_antisense CTTGTCAAGTTCGCTTTTACTTCCTATACAGTTGTATTCTGTGACCAGAATACTTGTCAAGTTCGCTTTTACTTCCTATACAGT TuDBart21-3p_sense ATCAACAAACACCAGTGGAAAAGCACAACTAGCAAGTATTCTGGTCACAGAATACAACAAACACCAGTGGAAAAGCACAACTAG TuDBart21-3p_antisense CTTGCTAGTTGTGCTTTTCCACTGGTGTTTGTTGTATTCTGTGACCAGAATACTTGCTAGTTGTGCTTTTCCACTGGTGTTTGT TuDBart21-5p_sense ATCAACGTTAGTTGCCTACCATCACTAGTGACAAGTATTCTGGTCACAGAATACAACGTTAGTTGCCTACCATCACTAGTGA TuDBart21-5p_antisense CTTGTCACTAGTGATCCTAGGCAACTAACGTTGTATTCTGTGACCAGAATACTTGTCACTAGTGATCCTAGGCAACTAACGT Oligonucleotides for cloning mirna sensors sensor_control_sense TAATGTATAGGAAGTGCGAACTTGAG sensor_control_antisense AATTCTCAAGTTCGCACTTCCTATACATTAAT sensor_bart16-5p_sense TAAAGAGCACACACCCACTCTATCTAAG sensor_bart16-5p_antisense AATTCTTAGATAGAGTGGGTGTGTGCTCTTTAAT 5 Continues on next page 141

142 Chapter 5 S2 Table. Top 5 of genes downregulated in BART16 expressing cells. p-values after error correction by false-discovery rate (Benjamini-Hochberg method). Gene symbol Fold change (2log) p-value TOMM CLGN TOMM22P DMRTA RP5-857K SELENBP RAB MAGED CTD-2319I CREBBP/CBP ART RP11-15O PROC AMOT QPRT NREP NBEAP MT-ND HOXA-AS RTN ABAT NBEA IL17A MT-ATP ARL6IP Gene symbol Fold change (2log) p-value CDKN1C PCDH GSTM RP11-324J HMGA RPL RPLP MT-CYB SEPT NAP1L SLX1A-SULT1A SLC47A RASSF CYP26A ROBO TRABD2B ZNF RP11-48N FBLN IGFBP EEF1A RPLP LIN28B TTC39A TSC22D pISRE-mCherry cells were lentivirally transduced (in duplo) to express either mir-bart16 in conjunction with the helper cluster (mir-bart19-14) or the helper cluster alone. Genes that were lower expressed in mir- BART16 expressing cells were ranked according to p-value. The top 5 of this list is displayed. 142

143 References 1. Kutok JL, Wang F (26) Spectrum of Epstein-Barr Virus-Associated Diseases. Annu Rev Pathol Mech Dis 1: Longnecker RM, Kieff E, Cohen JI (213) Epstein- Barr Virus. In: Knipe DM, Howley PM, editors. Field s Virology. Philadelphia: Lippincott Williams & Wilkins. pp Takeuchi O, Akira S (21) Pattern Recognition Receptors and Inflammation. Cell 14: Ivashkiv LB, Donlin LT (214) Regulation of type I interferon responses. Nat Rev Immunol 14: Schoggins JW, Rice CM (211) Interferon-stimulated genes and their antiviral effector functions. Current Opinion in Virology 1: Dunmire SK, Odumade OA, Porter JL, Reyes-Genere J, Schmeling DO, et al. (214) Primary EBV Infection Induces an Expression Profile Distinct from Other Viruses but Similar to Hemophagocytic Syndromes. PLoS ONE 9: e Fiola S, Gosselin D, Takada K, Gosselin J (21) TLR9 Contributes to the Recognition of EBV by Primary Monocytes and Plasmacytoid Dendritic Cells. The Journal of Immunology 185: Severa M, Giacomini E, Gafa V, Anastasiadou E, Rizzo F, et al. (213) EBV stimulates TLR- and autophagydependent pathways and impairs maturation in plasmacytoid dendritic cells: Implications for viral immune escape. Eur J Immunol 43: Quan TE, Roman RM, Rudenga BJ, Holers VM, Craft JE (21) Epstein-Barr virus promotes interferon-alpha production by plasmacytoid dendritic cells. Arthritis & Rheumatism 62: Samanta M, Iwakiri D, Kanda T, Imaizumi T, Takada K (26) EB virus-encoded RNAs are recognized by RIG-I and activate signaling to induce type I IFN. The EMBO journal 25: Iwakiri D, Zhou L, Samanta M, Matsumoto M, Ebihara T, et al. (29) Epstein-Barr virus (EBV)-encoded small RNA is released from EBV-infected cells and activates signaling from toll-like receptor 3. The Journal of Experimental Medicine 26: Sharp NA, Arrand JR, Clemens MJ (1989) Epstein-Barr Virus Replication in Interferon-treated Cells. Journal of General Virology 7: Shigeo K, Kohtaro T, Jiro W, Osamu Y (1986) Inhibition of Epstein-Barr virus infection in vitro by recombinant human interferons alpha and gamma. Virus Research 5: Lim WH, Kireta S, Russ GR, Coates PTH (26) Human plasmacytoid dendritic cells regulate immune responses to Epstein-Barr virus (EBV) infection and delay EBV-related mortality in humanized NOD-SCID mice. EBV mirna BART16 suppresses type I IFN signaling Blood 19: Bowie AG, Unterholzner L (28) Viral evasion and subversion of pattern-recognition receptor signalling. Nat Rev Immunol 8: Randall RE, Goodbourn S (28) Interferons and viruses: an interplay between induction, signalling, antiviral responses and virus countermeasures. Journal of General Virology 89: Feng P, Moses A, Fruh K (213) Evasion of adaptive and innate immune response mechanisms by gammaherpesviruses. Current Opinion in Virology 3: Cohen JI, Lekstrom K (1999) Epstein-Barr Virus BARF1 Protein Is Dispensable for B-Cell Transformation and Inhibits Alpha Interferon Secretion from Mononuclear Cells. Journal of Virology 73: Bentz GL, Liu R, Hahn AM, Shackelford J, Pagano JS (21) Epstein-Barr virus BRLF1 inhibits transcription of IRF3 and IRF7 and suppresses induction of interferon-ß. Virology 42: Hahn AM, Huye LE, Ning S, Webster-Cyriaque J, Pagano JS (25) Interferon Regulatory Factor 7 Is Negatively Regulated by the Epstein-Barr Virus Immediate-Early Gene, BZLF-1. Journal of Virology 79: Wang JT, Doong SL, Teng SC, Lee CP, Tsai CH, et al. (29) Epstein-Barr Virus BGLF4 Kinase Suppresses the Interferon Regulatory Factor 3 Signaling Pathway. Journal of Virology 83: Wu L, Fossum E, Joo CH, Inn KS, Shin YC, et al. (29) Epstein-Barr Virus LF2: an Antagonist to Type I Interferon. Journal of Virology 83: Kanda K, Decker T, Aman P, Wahlstrom M, von Gabain A, et al. (1992) The EBNA2-related resistance towards alpha interferon (IFN-alpha) in Burkitt s lymphoma cells effects induction of IFN-induced genes but not the activation of transcription factor ISGF-3. Mol Cell Biol 12: Shah KM, Stewart SE, Wei W, Woodman CBJ, O Neil JD, et al. (29) The EBV-encoded latent membrane proteins, LMP2A and LMP2B, limit the actions of interferon by targeting interferon receptors for degradation. Oncogene 28: Bartel DP (29) MicroRNAs: Target Recognition and Regulatory Functions. Cell 136: Grundhoff A, Sullivan CS (211) Virus-encoded micrornas. Virology 411: Pfeffer S, Zavolan M, Grasser FA, Chien M, Russo JJ, et al. (24) Identification of Virus-Encoded MicroRNAs. Science 34: Cullen BR (213) MicroRNAs as mediators of viral evasion of the immune system. Nat Immunol 14: Boss IW, Renne R (21) Viral mirnas: tools for 143 5

144 Chapter 5 immune evasion. Current Opinion in Microbiology 13: Xia T, O Hara A, Araujo I, Barreto J, Carvalho E, et al. (28) EBV MicroRNAs in Primary Lymphomas and Targeting of CXCL-11 by ebv-mir-bhrf1-3. Cancer Research 68: Kim S, Lee S, Shin J, Kim Y, Evnouchidou I, et al. (211) Human cytomegalovirus microrna mir-us4-1 inhibits CD8(+) T cell responses by targeting the aminopeptidase ERAP1. Nature immunology 12: Haneklaus M, Gerlic M, Kurowska-Stolarska M, Rainey AA, Pich D, et al. (212) Cutting edge: mir-223 and EBV mir-bart15 regulate the NLRP3 inflammasome and IL-1beta production. Journal of Immunology 189: Bauman Y, Nachmani D, Vitenshtein A, Tsukerman P, Drayman N, et al. (211) An identical mirna of the human JC and BK polyoma viruses targets the stressinduced ligand ULBP3 to escape immune elimination. Cell Host & Microbe 9: Nachmani D, Stern-Ginossar N, Sarid R, Mandelboim O (29) Diverse herpesvirus micrornas target the stressinduced immune ligand MICB to escape recognition by natural killer cells. Cell Host & Microbe 5: Stern-Ginossar N, Gur C, Biton M, Horwitz E, Elboim M, et al. (28) Human micrornas regulate stressinduced immune responses mediated by the receptor NKG2D. Nature immunology 9: Abend JR, Ramalingam D, Kieffer-Kwon P, Uldrick TS, Yarchoan R, et al. (212) Kaposi s sarcoma-associated herpesvirus micrornas target IRAK1 and MYD88, two components of the toll-like receptor/interleukin- 1R signaling cascade, to reduce inflammatory-cytokine expression. Journal of Virology 86: Lei X, Bai Z, Ye F, Xie J, Kim CG, et al. (21) Regulation of NF-kappaB inhibitor IkappaBalpha and viral replication by a KSHV microrna. Nature cell biology 12: Liang D, Gao Y, Lin X, He Z, Zhao Q, et al. (211) A human herpesvirus mirna attenuates interferon signaling and contributes to maintenance of viral latency by targeting IKK[epsilon]. Cell Res 21: Amoroso R, Fitzsimmons L, Thomas WA, Kelly GL, Rowe M, et al. (211) Quantitative studies of Epstein-Barr virus-encoded micrornas provide novel insights into their regulation. J Virol 85: Motsch N, Alles J, Imig J, Zhu J, Barth S, et al. (212) MicroRNA profiling of Epstein-Barr virus-associated NK/T-cell lymphomas by deep sequencing. PLoS One 7: e Qiu J, Cosmopoulos K, Pegtel M, Hopmans E, Murray P, et al. (211) A novel persistence associated EBV mirna expression profile is disrupted in neoplasia. PLoS Pathog 7: e Barth S, Meister G, Grasser FA (211) EBV-encoded mirnas. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms 189: Haraguchi T, Ozaki Y, Iba H (29) Vectors expressing efficient RNA decoys achieve the long-term suppression of specific microrna activity in mammalian cells. Nucleic Acids Res 37: e Paulson M, Press C, Smith E, Tanese N, Levy DE (22) IFN-Stimulated transcription through a TBP-free acetyltransferase complex escapes viral shutoff. Nature cell biology 22/1/22: Kramer OH, Heinzel T (21) Phosphorylationacetylation switch in the regulation of STAT1 signaling. Molecular and cellular endocrinology 29/11/3: Lotz M, Tsoukas CD, Fong S, Carson DA, Vaughan JH (1985) Regulation of Epstein-Barr virus infection by recombinant interferons. Selected sensitivity to interferongamma. Eur J Immunol 15: Dorner M, Brandt S, Tinguely M, Zucol F, Bourquin JP, et al. (29) Plasma cell toll-like receptor (TLR) expression differs from that of B cells, and plasma cell TLR triggering enhances immunoglobulin production. Immunology 128: Iwasaki A, Medzhitov R (24) Toll-like receptor control of the adaptive immune responses. Nat Immunol 5: Meckes DG, Shair KHY, Marquitz AR, Kung CP, Edwards RH, et al. (21) Human tumor virus utilizes exosomes for intercellular communication. Proceedings of the National Academy of Sciences 17: Pegtel DM, Cosmopoulos K, Thorley-Lawson DA, van Eijndhoven MAJ, Hopmans ES, et al. (21) Functional delivery of viral mirnas via exosomes. Proceedings of the National Academy of Sciences 17: Zhang JJ, Vinkemeier U, Gu W, Chakravarti D, Horvath CM, et al. (1996) Two contact regions between Stat1 and CBP/p3 in interferon γ signaling. Proceedings of the National Academy of Sciences 93: Bhattacharya S, Eckner R, Grossman S, Oldread E, Arany Z, et al. (1996) Cooperation of Stat2 and p3/ CBP in signalling induced by interferon-alpha. Nature 1996/9/26: Tang X, Gao JS, Guan YJ, McLane KE, Yuan ZL, et al. (27) Acetylation-dependent signal transduction for type I interferon receptor. Cell 131: Gottwein E, Corcoran DL, Mukherjee N, Skalsky RL, Hafner M, et al. (211) Viral microrna targetome of KSHV-infected primary effusion lymphoma cell lines. Cell Host & Microbe 1: Cox JE, McClure LV, Goga A, Sullivan CS (215) Panviral-microRNA screening identifies interferon inhibition as a common function of diverse viruses. Proceedings of the National Academy of Sciences 112: Callegari S, Gastaldello S, Faridani OR, Masucci MG 144

145 EBV mirna BART16 suppresses type I IFN signaling (214) Epstein-Barr virus encoded micrornas target SUMO-regulated cellular functions. The FEBS journal 281: Xing J, Ni L, Wang S, Wang K, Lin R, et al. (213) Herpes simplex virus 1-encoded tegument protein VP16 abrogates the production of beta interferon (IFN) by inhibiting NF-kappaB activation and blocking IFN regulatory factor 3 to recruit its coactivator CBP. Journal of Virology 87: Lin R, Genin P, Mamane Y, Sgarbanti M, Battistini A, et al. (21) HHV-8 encoded virf-1 represses the interferon antiviral response by blocking IRF-3 recruitment of the CBP/p3 coactivators. Oncogene 21/4/21: Lagos D, Pollara G, Henderson S, Gratrix F, Fabani M, et al. (21) mir-132 regulates antiviral innate immunity through suppression of the p3 transcriptional coactivator. Nat Cell Biol 12: and Regulatory Links Shape Genetic Interactions between Signaling Pathways. Cell 143: Yang YH, Dudoit S, Luu P, Lin DM, Peng V, et al. (22) Normalization for cdna microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Research 3: e Margaritis T, Lijnzaad P, van Leenen D, Bouwmeester D, Kemmeren P, et al. (29) Adaptable gene-specific dye bias correction for two-channel DNA microarrays. Mol Syst Biol 5: Wu H, Kerr MK, Cui X, Churchill GA (23) MAANOVA: a software package for the analysis of spotted cdna microarray experiments. In: The analysis of gene expression data. Springer. pp Nanbo A, Inoue K, Adachi-Takasawa K, Takada K (22) Epstein-Barr virus RNA confers resistance to interferon-a-induced apoptosis in Burkitts lymphoma. EMBO J 21: Aman P, von Gabain A (199) An Epstein-Barr virus immortalization associated gene segment interferes specifically with the IFN-induced anti-proliferative response in human B-lymphoid cell lines. The EMBO journal 9: Vo N, Goodman RH (21) CREB-binding Protein and p3 in Transcriptional Regulation. Journal of Biological Chemistry 276: Yoneyama M, Suhara W, Fukuhara Y, Fukuda M, Nishida E, et al. (1998) Direct triggering of the type I interferon system by virus infection: activation of a transcription factor complex containing IRF-3 and CBP/ p3. The EMBO journal 1998/3/28: Merika M, Williams AJ, Chen G, Collins T, Thanos D (1998) Recruitment of CBP/p3 by the IFN beta enhanceosome is required for synergistic activation of transcription. Molecular cell 1998/7/11: Wang F (213) Nonhuman primate models for Epstein- Barr virus infection. Current Opinion in Virology 3: Riley KJ, Rabinowitz GS, Steitz JA (21) Comprehensive analysis of Rhesus lymphocryptovirus microrna expression. Journal of Virology 84: Walz N, Christalla T, Tessmer U, Grundhoff A (21) A global analysis of evolutionary conservation among known and predicted gammaherpesvirus micrornas. Journal of Virology 84: Lebbink RJ, Lowe M, Chan T, Khine H, Wang X, et al. (211) Polymerase II Promoter Strength Determines Efficacy of microrna Adapted shrnas. PLoS ONE 6: e van Wageningen S, Kemmeren P, Lijnzaad P, Margaritis T, Benschop JJ, et al. (21) Functional Overlap 145

146

147 CHAPTER 6 Summarizing discussion

148 Chapter 6 148

149 Summarizing discussion This thesis reports several novel innate evasion strategies of EBV that interfere with patternrecognition receptor-mediated responses at various levels: with expression of the receptors, with activity of the downstream signaling pathways, and with the induction of effector mechanisms. The ability of EBV to downregulate members of the Toll-like receptor family, an important group of PRRs involved in sensing (viral) infection, is described in chapter 2. To analyze TLR expression levels in productively EBV-infected B cells, we used a system to obtain and study almost pure populations of these cells that naturally express TLR1, TLR6, TLR7, TLR9, and TLR1. Latent EBV infection reduces the mrna levels of TLR1 and TLR1, whereas during lytic infection mrna expression of several TLRs and in particular TLR9 is downregulated. EBV alkaline exonuclease BGLF5 contributes to TLR9 mrna reductions. Even more striking than the TLR9 mrna reduction is the remarkable decline in TLR9 protein levels relatively early after induction of productive infection. This rapid reduction has not been observed for other molecules downregulated during productive EBV infection and cannot be explained by TLR9 turnover alone. These results suggest that EBV employs additional mechanisms to decrease TLR9 levels post-transcriptionally. Notably, we and others have implicated TLR9 in the recognition of EBV, demonstrating the functional relevance of downregulating this receptor during natural infection [1,2]. Also TLR2, another TLR family member involved in sensing of EBV [3 5], was found to be downregulated by the actions of BGLF5, as described in chapter 3. We followed up on the effects of BGLF5-mediated shutoff on immunologically relevant surface proteins and add both TLR2 and CD1d to the BGLF5 target list. The non-classical MHC molecule CD1d presents lipid antigens to invariant natural killer T (inkt) cells and thereby operates on the border between innate and adaptive immunity [6]. inkt cells have recently been acknowledged as important components in the control of EBV infection [7,8]. To obtain a more general impression of the influence of BGLF5-mediated shutoff on surface protein expression, we reduced BGLF5 levels in productively EBV-infected B cells by up to 75% using short hairpin (sh)rnas. The surface proteins tested were downregulated to different extents 24 hrs into productive infection, yet BGLF5 downregulation only marginally rescued their expression. Seemingly in contrast with observations for BGLF5 expressed in isolation, this suggests a limited contribution of BGLF5 to downregulation of surface receptors on naturally infected B cells as a means of immune evasion. This is supported by a recent report describing a limited influence of BGLF5 on HLA I downregulation and subsequent inhibition of T-cell activation [9], opposing earlier studies showing that BGLF5 expressed in isolation exerted a significant inhibitory effect on T-cell activation [1]. It should be noted, however, that both in our study and the report by Quinn et al. [9] BGLF5 was downregulated by 6 149

150 Chapter 6 approximately 75%. Given the involvement of BGLF5 s enzymatic activity, expression levels not necessarily correlate with total activity and the remaining amount of BGLF5 may prevent observation of a more severe phenotype. Furthermore, given the dual role of BGLF5 during viral infection, acting as DNase and RNase, it can be envisioned that BGFL5 is not expressed at limiting levels. Studying a complete knockout phenotype is however complicated by the crucial role of BGLF5 s DNase function during the viral life cycle and has exclusively been performed in 293T cells with a focus on the effects on viral replication, not immune evasion [11,12]. Nonetheless, a limited role for BGLF5 in immune evasion is further supported by observations for the α-herpesvirus HSV-1 and the murine γ-herpesvirus MHV-68 that suggest that immune evasive functions of shutoff mainly extend to newly synthesized induced by type I IFNs, instead of existing surface receptors [13 15]. These observations oppose to a pivotal role for BGLF5 in general immune evasion. Both the downregulation of TLR9 (chapter 2) and CD1d (chapter 3) are less pronounced during expression of BGLF5 in isolation, without the context of viral infection, compared to the downregulation observed in productively-infected B cells. Differences in BGLF5 expression level cannot explain this discrepancy, as the expression levels are nearly equal in both instances. Although other experimental differences may contribute, we favor the idea that for both receptors additional, not yet identified, viral mechanisms synergize with BGLF5 to achieve the optimal timing and extent of downregulation during natural infection. This cooperation between general and more specific EBV evasion tactics is also observed for the adaptive immune system, where the contribution of BGLF5 to HLA I downregulation is limited compared to the effects induced by TAP-inhibitor BNLF2a early during productive infection and viral GPCR BILF1 at a later stage [9]. The mild rescue of a large group of surface proteins following BGLF5 knockdown described in chapter 3 and the likely involvement of additional factors in reducing TLR9 and CD1d surface levels leave interesting opportunities for finding additional, more specific, immune evasion strategies. In chapter 4, we shifted our focus to interference with signaling pathways downstream of TLR ligation that lead to NF-κB activation and production of proinflammatory cytokines. Using active site-directed probes, we establish that the EBV-encoded homolog of the conserved herpesvirus deubiquitinase (DUB), the large tegument protein BPLF1, is functionally expressed during the late phase of productive infection in B cells. BPLF1 interferes with TLRmediated NF-κB activation by removing K63-linked ubiquitin chains from TLR signaling intermediates TRAF6 and IKKγ/NEMO and K48-linked ubiquitin chains from IκBα. Consequently, BPLF1-expressing cells produce less proinflammatory cytokines (e.g. IL-8) following TLR2 stimulation by EBV particles. Around the same time, BPLF1 was reported to improve virus production from lytically infected 293T cells by interfering with LMP1- mediated NF-κB activation [16]. In line with findings by others [17], we observe that BPLF1 15

151 Summarizing discussion is part of the virion-associated tegument. As such, BPLF1 is potentially released in newly infected cells to immediately perform its immunomodulatory function(s). Interestingly, the BPLF1-homolog of the related human gammaherpesvirus KSHV (ORF64) is involved in immune evasion by interfering with RIG-I mediated signaling pathways [18], and ORF64 of murine gammaherpesvirus MHV-68 interferes with STING-mediated DNA sensor responses [19]. It would be interesting to determine whether EBV BPLF1 similarly affects these other PRR responses. The DUB activity of BPLF1 is contained within the N-terminal amino acids of the ~315 amino acid-containing full-length protein. In line with reports on BPLF1 homologs of other herpesviruses [2,21], we observe processing of full length BPLF1 in cells to yield N-terminal active DUB fragments of 25 kda and 32 kda (chapter 4). The 32 kda fragment is present in both cytoplasmic and nuclear fractions, while the 25 kda fragment seems restricted to the nucleus. This corresponds with the reported cleavage of BPLF1 by caspase 1 into an approximately 25 kda fragment that acts as a deneddylase during viral infection [22]. The major group of proteins serving as neddylation substrates is formed by the cell-cycle regulating cullin-ring ligases (CRLs) that localize to the nucleus. An interesting hypothesis is that the 25 kda BPLF1 fragment acts as a deneddylase targeting CRLs in the nucleus, while the 32 kda fragment is responsible for the immune modulation we observe in the cytoplasm. How this processing is spatially and temporally regulated, and at what stage of infection BPLF1 performs various (immunomodulatory) functions remains to be resolved. In this respect it is important to determine the function of the C-terminal part of the BPLF1 protein not directly involved in DUB activity. We speculate that this large part of the protein plays a role in spatially and/or temporally regulating N-terminal DUB activity. Prior to our findings, all studies on BPLF1 were performed using relatively small N-terminal fragments of up to 325 amino acids. The full-length BPLF1 expression construct we generated for our studies thus provides a valuable tool for future studies aimed at identification of the function and regulation of BPLF1 s DUB activity during viral infection. Besides effects mediated by NF-κB activation, other important effector mechanisms of the innate immune system are facilitated by the production of type I IFNs (chapter 1). Binding of these IFNs to receptors on target cells induces production of a plethora of interferonstimulated genes (ISGs) that interfere with the viral replication cycle in numerous ways. Since the identification of viral mirnas in EBV-infected cells approximately ten years ago [23], these mirnas have been implicated in regulation of a variety of cellular and viral processes, including immune evasion [24,25]. Over 9% of the viral mirnas identified to date are encoded by members of the herpesvirus family. Their small size and lack of antigenicity mark mirnas as interesting immune evasion tools, especially during the latent stages of infection when viral protein expression is severely limited. In chapter 5, we describe the 6 151

152 Chapter 6 interference of EBV-encoded mir-bart16 with IFN-α-induced induction of ISRE promoter activity to interrupt concomitant ISG production. mir-bart16 directly targets the RNA messenger encoding cellular histone acetylase CREB-binding protein (CBP). CBP forms a central hub connecting various signaling pathways, including those induced by type I IFN, to the general transcription machinery. As such, CBP is an appealing target for viral evasion strategies, and indeed many viruses target this protein to interfere with IFN responses. Emphasizing its importance during herpesvirus infection, both HSV and KSHV encode proteins to interfere with CBP function [26] and a recent study reports CBP inhibition by mirnas of three different herpesviruses including EBV [27]. Interestingly, similar to our observations for mir-bart16, this study reported EBV-encoded mir-bart18 to interfere with type I IFN signaling by targeting CBP. Besides the EBV-encoded proteins interfering with the production or effector mechanisms induced by type I IFNs (e.g. BZLF1, BRLF1, LF2, and BGLF4), EBV thus encodes (at least) two different mirnas targeting this innate immune pathway, highlighting the importance of attenuating the effects induced by type I IFNs for the virus. EBV mirnas thus likely contribute to the relative resistance of gammaherpesviruses to the IFN response [28]. Conclusions and future perspectives The newly identified EBV innate evasion strategies described in this thesis and recent advances in this field demonstrate that EBV employs an impressive repertoire of mechanisms to interfere with human innate immune responses signaled through PRRs. EBV targets PRR functionality at many levels using both proteins (chapter 2, 3, and 4) as well as viral mirnas (chapter 5). However, it has also become clear that the picture is far from complete and future research is required to provide a more comprehensive overview of EBV evasion strategies and their contribution to natural infection and viral pathogenesis. An interesting aspect remaining largely unresolved is why EBV encodes many different molecules to interfere with single innate immune pathways. Observations for the bettercharacterized adaptive immune evasion strategies may shed light on this issue. For example BNLF2a, BILF1, and BGLF5 have each been implicated in HLA I downregulation and interference with CD8 + T-cell activation [29]. A recent study compared the relative contributions of these molecules to functional evasion of CD8 + T cells during different stages of productive EBV infection by shrna-mediated knockdown of these proteins [9]. BNLF2a mainly acts early after induction of productive infection, while BILF1 is primarily active at later stages; BGLF5 knockdown affects T-cell evasion only to a very limited extent during all stages of productive infection. By analogy, it seems likely that the many EBV innate evasion strategies not act simultaneously but vary in timing of expression or activity during the various stages of latent and productive infection and in the different cell types infected by EBV 152

153 Summarizing discussion (mainly B cells and epithelial cells). For example, evasion molecules LF2, BGLF4, and BPLF1 are present within virion-associated tegument, potentially allowing them to act immediately after infection of a new cell (chapter 4, [3,31]), while (other) non-structural proteins may only act during the productive phase. Proteins may even perform alternating functions during the course of infection, as seems the case for the HSV-1 homolog of BPLF1, UL36. Late during productive infection, UL36 functions as a structural tegument component in virion assembly, while during primary infection UL36 plays a role in the trafficking of the viral capsid to the nucleus and the release of viral DNA following proteolytic cleavage [32 34]. Interestingly, its association with the virion may function to ensure localization of UL36 s DUB activity close to the virion, to prevent premature uncoating of the virion through ubiquitination that releases PRR ligands into the cytoplasm [35]. Whether BPLF1 performs similar dual functions remains to be established, but the observation of processed BPLF1 fragments (chapter 4) provides hints in this direction. Furthermore, recent reports indicate immediate expression of the adaptive immune evasion factors vil-1 and BNLF2a by transfer of virion-associated viral mrnas into newly infected cells, bypassing the need for transcription [36,37]. Even nonstructural proteins may thus be synthesized and active rapidly after EBV has entered a new cell. Finally, exosomal secretion of EBV proteins and mirnas potentially enables immune evasion molecules to act in cells not generally infected by EBV, for example antigen-presenting DCs [38 4]. Taken together, spatial and temporal differences in action of the various innate evasion molecules may explain the need for many different evasion molecules. As described below, improved in vivo (animal) infection models and systems to generate knockout viruses will likely allow us to further resolve the role of innate immune evasion strategies during the various stages of EBV infection. Several examples of EBV and other herpesviruses illustrate that immune activation may actually benefit viral infection or replication. For example, stimulation of TLR9 during primary EBV infection supports B-cell transformation [41], EBV LMP1-mediated NF-κB activation increases B cell-survival [42], and HCMV and MCMV rely on IFI16 activation for successful establishment of infection [43,44]. Evolution has apparently granted these viruses the ability to benefit from (some aspects of) immune activation. This is often accompanied by specific inhibition of certain disadvantageous parts of the immune response. An example of this is the (indirect) induction of TLR7 expression during EBV infection, with concomitant virusmediated upregulation of inhibitory IRF5 splice variants that attenuate TLR7-mediated IRF activation and type I IFN production [45]. Likewise, TLR9 activation seems to be differentially modulated. During primary infection TLR9 ligation favors B cell transformation, activation, and proliferation [41,46]. Following viral gene expression, however, TLR9 expression at the cell surface is reduced [47,48]. Instead of completely inhibiting immune pathways, EBV thus appears to apply its many evasion molecules to specifically steer immune responses in a 6 153

154 Chapter 6 direction required to favor a specific stage of the infection, while attenuating concomitant negative effects. The identification of novel immune evasion strategies is intertwined with the elucidation of the mechanisms the host employs to detect EBV infection. So far, studies aimed at revealing EBV innate evasion strategies have mainly focused on the TLRs and RLRs involved in detection of herpesviruses. Consequently, the role of the nuclear and cytoplasmic DNA sensors discovered more recently has received little attention yet. It is however becoming increasingly clear that these PRRs play an important role in detection of herpesvirus infections, and the DNA genomes carried by herpesviruses are now even suggested to be the main trigger of innate immune responses [35,49]. Reports on the role of these newly-identified PRRs in detecting herpesvirus infection were quickly followed by the revelation of the first herpesvirus encoded evasions mechanisms [5,51]. For EBV, induction of signaling pathways leading to inflammasome formation by IFI16 has recently been reported [52]. Although EBV therefore likely interferes with inflammasome-inducing PRRs, these strategies have not been elucidated yet. Interestingly, a very recent report implicates the deubiquitinase activity of the EBV BPLF1 homolog of murine herpesvirus MHV-68, ORF64, in the inhibition of STINGmediated responses during primary infection [19]. This favors the establishment of latent infection. By analogy, BPLF1, in this thesis described to interfere with TLR-mediated NF-κB activation (chapter 4), may perform a similar role during EBV infection. Also inkt cells and NK cells have recently been implicated in the control of EBV infection [7,53 56]. Patients with selective defects in NK cells frequently suffer from recurrent EBVassociated disease, and a reduced number of inkt cells is associated with fatal EBV infection. inkt cells are activated by the non-classical MHC molecule CD1d that presents lipid antigens. Interestingly, CD1d expression is downregulated in EBV-infected B cells [57]. We also observe reduced CD1d surface levels during productive EBV infection, in part mediated by BGLF5 (chapter 3), and HSV-1 and KSHV also downregulate CD1d expression at the cell surface during lytic infection [58,59]. Activation of NK cells is determined by the balance between inhibitory and activating signals. NKG2D is an activating receptor that has a role in controlling chronic EBV infection [6]. Induction of the EBV lytic cycle increases NKG2Ddependent NK-cell reactivity [61]. These findings, however, do not preclude that EBV also encodes NK evasion molecules to protect lytic EBV-infected cells from an NK-cell attack. For example, BZLF1 contains an HLA-E binding peptide that increases surface expression of HLA-E [62], which is a ligand for the inhibitory receptor NKG2A. Therefore, increased HLA-E surface expression may restrict NK-cell reactivity. The viral GPCR BILF1, interfering with T-cell responses by downregulating HLA I from the cell surface, does not affect HLA-C levels [63], potentially preserving the inhibitory NK cell signal of HLA-C. In addition, preliminary observations in our lab suggest downregulation of NK-cell regulatory molecules like ULBP2 154

155 Summarizing discussion in productively EBV-infected cells. In spite of the identification of novel EBV innate immune evasion strategies, chapter 1 shows that our knowledge on this subject is still fragmented. Missing evasion strategies remain to be identified, and, more importantly, a comprehensive picture of the extent and timing of current evasion strategies is required. An obvious but difficult to address issue is the in vivo relevance and potential synergy of sometimes seemingly redundant evasion mechanisms during the various phases of viral replication. Two main issues have hindered tackling these questions so far: (i) a lack of procedures to easily establish EBV mutants, and (ii) the absence of suitable animal infection models. Classic homologous recombination and bacterial-artificial chromosome (BAC)-based strategies have successfully been applied to prepare mutant viruses and determine herpesvirus gene functions in infected cells [64]. Besides the fact that these methods are time-consuming and laborious, it has been difficult to apply these to study EBV mutants in productively-infected B cells. For example, preparing EBV mutants in our AKBM cells using these methods is difficult due to the many viral genome copies present, and infection and selection of new cells with mutant virus results in low induction efficiencies that hinder biochemical characterization. In chapter 3, we describe the use of shrnas to reduce BGLF5 expression and study BGLF5 s immune-evasive functions during productive infection as an alternative to constructing a knockout mutant virus lacking this essential gene. A major drawback of this method is that complete knockdown is hard to achieve, which may hamper the observation of clear phenotypes. This approach may however be useful for genes essential for viral infection and/or replication; the remaining expression may facilitate production of mutant viruses. A promising new approach for complete knockdown of specific genes is provided by CRISPR/Cas9 mediated genome editing systems. This method has recently been applied to completely remove EBV genomes from latently infected cells [65] and for targeted disruption of EBV mirna expression by deleting the BART promoter region [66]. Despite the many EBV genome copies in a single infected cell, this genome editing method in combination with proper selection strategies delivers a relatively quick and easy way of preparing specific EBV mutants for future studies. Preferably, these EBV mutants are then used to determine the contribution of specific genes to immune evasion by studying phenotypes in vivo. As mentioned, these in vivo are mainly held back by the lack of proper animal infection models due to the strict human specificity of EBV and the lack of homologous viruses in rodents [67]. The closest relative of EBV in mice, MHV- 68, differs in cell tropism and lacks the acute transforming ability that would be necessary to study EBV-induced tumorigenesis [67,68]. Although the results obtained with MHV-68 cannot always be directly translated to EBV infections, the studies with MHV-68 have advanced our understanding of immune responses and immune evasion by gammaherpesviruses in many ways. Importantly, major progress is made in the application of humanized mice carrying 6 155

156 Chapter 6 partly human immune systems (human immune system components or HIS mice) that recapitulate many aspects of EBV infection in humans [67,69 72]. The major limitations of these HIS mice are the lack of infection of epithelial cells, and the limited differentiation of infected B cells to plasma cells, thus restricting studies to latency II [67]. Alternatively, nonhuman primate models may be used to study EBV immune evasion using closely related lymphocryptoviruses naturally infecting monkeys. A well-characterized model involves the use of rhesus macaques that are naturally infected with the rhesus-lymphocryptovirus [73]. Both HIS mice and rhesus macaques are being used successfully in infection studies to gain insight into EBV pathogenesis and immune responses against this human pathogen. Future infection studies with mutant EBV strains will thus provide important information on the relative contribution of immune evasion strategies to viral infection and improve our insight into virus-host interactions and viral pathogenesis. Taken together, the immune evasion mechanisms of EBV identified so far show that this virus employs a wide range of strategies to compromise innate immunity during the latent and replicative phases of its life cycle. These evasion mechanisms also affect adaptive immune activation and are likely to contribute to the development of EBV-associated malignancies. Nevertheless, the picture is far from complete and many aspects require further investigation, especially using model systems mimicking human infections to validate the current results that have primarily been acquired using in vitro systems. A profound understanding of the immune evasive manoeuvres of EBV will aid in the development of novel strategies to combat EBV infections and associated diseases, especially the malignancies caused by this oncogenic human herpesvirus. In addition, the antiviral functions of various host (innate) immune components were only appreciated after they were found to be targeted by viral evasion mechanisms. Revealing novel viral evasion strategies may thus not only increase our understanding of viral pathogenesis, but may also lead to the identification of novel host detection and effector mechanisms involved in the antiviral response. 156

157 Summarizing discussion References 1. Fiola S, Gosselin D, Takada K, Gosselin J (21) TLR9 Contributes to the Recognition of EBV by Primary Monocytes and Plasmacytoid Dendritic Cells. The Journal of Immunology 185: Lim WH, Kireta S, Russ GR, Coates PTH (27) Human plasmacytoid dendritic cells regulate immune responses to Epstein-Barr virus (EBV) infection and delay EBVrelated mortality in humanized NOD-SCID mice. Blood 19: Gaudreault E, Fiola S, Olivier M, Gosselin J (27) Epstein-Barr Virus Induces MCP-1 Secretion by Human Monocytes via TLR2. J Virol 81: Ariza ME, Glaser R, Kaumaya PTP, Jones C, Williams MV (29) The EBV-Encoded dutpase Activates NFkB through the TLR2 and MyD88-Dependent Signaling Pathway. The Journal of Immunology 182: Ariza ME, Rivailler P, Glaser R, Chen M, Williams MV (213) Epstein-Barr Virus Encoded dutpase Containing Exosomes Modulate Innate and Adaptive Immune Responses in Human Dendritic Cells and Peripheral Blood Mononuclear Cells. PLoS ONE 8: e Kinjo Y, Kitano N, Kronenberg M (213) The role of invariant natural killer T cells in microbial immunity. J Infect Chemother 19: Pasquier B, Yin L, Fondaneche MC, Relouzat F, Bloch- Queyrat C, et al. (25) Defective NKT cell development in mice and humans lacking the adapter SAP, the X-linked lymphoproliferative syndrome gene product. The Journal of Experimental Medicine 21: Rigaud S, Fondaneche MC, Lambert N, Pasquier B, Mateo V, et al. (26) XIAP deficiency in humans causes an X-linked lymphoproliferative syndrome. Nature 444: Quinn LL, Zuo J, Abbott RJM, Shannon-Lowe C, Tierney RJ, et al. (214) Cooperation between Epstein-Barr Virus Immune Evasion Proteins Spreads Protection from CD8+ T Cell Recognition across All Three Phases of the Lytic Cycle. PLoS Pathog 1: e Rowe M, Glaunsinger B, van LD, Zuo J, Sweetman D, et al. (27) Host shutoff during productive Epstein-Barr virus infection is mediated by BGLF5 and may contribute to immune evasion. Proc Natl Acad Sci U S A 14: Feederle R, Bannert H, Lips H, Muller-Lantzsch N, Delecluse HJ (29) The Epstein-Barr Virus Alkaline Exonuclease BGLF5 Serves Pleiotropic Functions in Virus Replication. Journal of Virology 83: Feederle R, Mehl-Lautscham AM, Bannert H, Delecluse HJ (29) The Epstein-Barr virus protein kinase BGLF4 and the exonuclease BGLF5 have opposite effects on the regulation of viral protein production. Journal of Virology 83: Murphy JA, Duerst RJ, Smith TJ, Morrison LA (23) Herpes Simplex Virus Type 2 Virion Host Shutoff Protein Regulates Alpha/Beta Interferon but Not Adaptive Immune Responses during Primary Infection In Vivo. Journal of Virology 77: Pasieka TJ, Lu B, Crosby SD, Wylie KM, Morrison LA, et al. (28) Herpes Simplex Virus Virion Host Shutoff Attenuates Establishment of the Antiviral State. Journal of Virology 82: Sheridan V, Polychronopoulos L, Dutia BM, Ebrahimi B (214) A Shutoff and Exonuclease Mutant of Murine Gammaherpesvirus-68 Yields Infectious Virus and Causes RNA loss in Type I Interferon Receptor Knock-Out cells. Journal of General Virology 95: Saito S, Murata T, kanda T, Isomura H, Narita Y, et al. (213) Epstein-Barr Virus Deubiquitinase Downregulates TRAF6-Mediated NF-kB Signaling during Productive Replication. Journal of Virology 87: Johannsen E, Luftig M, Chase MR, Weicksel S, Cahir- McFarland E, et al. (24) Proteins of purified Epstein-Barr virus. Proceedings of the National Academy of Sciences of the United States of America 11: Inn KS, Lee SH, Rathbun JY, Wong LY, Toth Z, et al. (211) Inhibition of RIG-I-Mediated Signaling by Kaposi s Sarcoma-Associated Herpesvirus-Encoded Deubiquitinase ORF64. Journal of Virology 85: Sun C, Schattgen SA, Pisitkun P, Jorgensen JP, Hilterbrand AT, et al. (215) Evasion of Innate Cytosolic DNA Sensing by a Gammaherpesvirus Facilitates Establishment of Latent Infection. The Journal of Immunology Published ahead of print: 1.449/ jimmunol Gredmark S, Schlieker C, Quesada V, Spooner E, Ploegh HL (27) A Functional Ubiquitin-Specific Protease Embedded in the Large Tegument Protein (ORF64) of Murine Gammaherpesvirus 68 Is Active during the Course of Infection. Journal of Virology 81: Wang J, Loveland AN, Kattenhorn LM, Ploegh HL, Gibson W (26) High-Molecular-Weight Protein (pul48) of Human Cytomegalovirus Is a Competent Deubiquitinating Protease: Mutant Viruses Altered in Its Active-Site Cysteine or Histidine Are Viable. Journal of Virology 8: Gastaldello S, Chen X, Callegari S, Masucci MG (213) Caspase-1 Promotes Epstein-Barr Virus Replication by Targeting the Large Tegument Protein Deneddylase to the Nucleus of Productively Infected Cells. PLoS Pathog 9: e Pfeffer S, Zavolan M, Grasser FA, Chien M, Russo JJ, et al. (24) Identification of Virus-Encoded MicroRNAs. Science 34: Boss IW, Renne R (21) Viral mirnas: tools for immune evasion. Current Opinion in Microbiology 13: Grundhoff A, Sullivan CS (211) Virus-encoded 157 6

158 Chapter 6 micrornas. Virology 411: Bowie AG, Unterholzner L (28) Viral evasion and subversion of pattern-recognition receptor signalling. Nat Rev Immunol 8: Cox JE, McClure LV, Goga A, Sullivan CS (215) Panviral-microRNA screening identifies interferon inhibition as a common function of diverse viruses. Proceedings of the National Academy of Sciences 112: Mossman KL (22) Activation and Inhibition of Virus and Interferon: The Herpesvirus Story. Viral Immunology 15: Ressing ME, Horst Dl, Griffin BD, Tellam J, Zuo J, et al. (28) Epstein-Barr virus evasion of CD8+ and CD4+ T cell immunity via concerted actions of multiple gene products. Seminars in Cancer Biology 18: Wang JT, Doong SL, Teng SC, Lee CP, Tsai CH, et al. (29) Epstein-Barr Virus BGLF4 Kinase Suppresses the Interferon Regulatory Factor 3 Signaling Pathway. Journal of Virology 83: Wu L, Fossum E, Joo CH, Inn KS, Shin YC, et al. (29) Epstein-Barr Virus LF2: an Antagonist to Type I Interferon. Journal of Virology 83: Abaitua F, Hollinshead M, Bolstad M, Crump CM, O Hare P (212) A Nuclear Localization Signal in Herpesvirus Protein VP1-2 Is Essential for Infection via Capsid Routing to the Nuclear Pore. Journal of Virology 86: Jovasevic V, Liang L, Roizman B (28) Proteolytic Cleavage of VP1-2 Is Required for Release of Herpes Simplex Virus 1 DNA into the Nucleus. Journal of Virology 82: Schipke J, Pohlmann A, Diestel R, Binz A, Rudolph K, et al. (212) The C Terminus of the Large Tegument Protein pul36 Contains Multiple Capsid Binding Sites That Function Differently during Assembly and Cell Entry of Herpes Simplex Virus. Journal of Virology 86: Horan KA, Hansen K, Jakobsen MR, Holm CK, Soby S, et al. (213) Proteasomal Degradation of Herpes Simplex Virus Capsids in Macrophages Releases DNA to the Cytosol for Recognition by DNA Sensors. The Journal of Immunology 19: Jochum S, Moosmann A, Lang S, Hammerschmidt W, Zeidler R (212) The EBV Immunoevasins vil-1 and BNLF2a Protect Newly Infected B Cells from Immune Recognition and Elimination. PLoS Pathog 8: e Jochum S, Ruiss R, Moosmann A, Hammerschmidt W, Zeidler R (212) RNAs in Epstein-Barr virions control early steps of infection. Proceedings of the National Academy of Sciences 19: E1396-E Verweij FJ, van Eijndhoven MAJ, Hopmans ES, Vendrig T, Wurdinger T, et al. (211) LMP1 association with CD63 in endosomes and secretion via exosomes limits constitutive NF-kB activation. EMBO J 3: Pegtel DM, Cosmopoulos K, Thorley-Lawson DA, van Eijndhoven MAJ, Hopmans ES, et al. (21) Functional delivery of viral mirnas via exosomes. Proceedings of the National Academy of Sciences 17: Meckes DG, Shair KHY, Marquitz AR, Kung CP, Edwards RH, et al. (21) Human tumor virus utilizes exosomes for intercellular communication. Proceedings of the National Academy of Sciences 17: Iskra S, Kalla M, Delecluse HJ, Hammerschmidt W, Moosmann A (21) Toll-Like Receptor Agonists Synergistically Increase Proliferation and Activation of B Cells by Epstein-Barr Virus. Journal of Virology 84: Middeldorp JM, Pegtel DM (28) Multiple roles of LMP1 in Epstein-Barr virus induced immune escape. Seminars in Cancer Biology 18: Cristea IM, Moorman NJ, Terhune SS, Cuevas CD, O Keefe ES, et al. (21) Human Cytomegalovirus pul83 Stimulates Activity of the Viral Immediate-Early Promoter through Its Interaction with the Cellular IFI16 Protein. Journal of Virology 84: Hertel L, De Andrea M, Azzimonti B, Rolle A, Gariglio M, et al. (1999) The Interferon-Inducible 24 Gene, a Member of the Ifi 2 Family, Is Not Involved in the Antiviral State Induction by IFN-a, but Is Required by the Mouse Cytomegalovirus for Its Replication. Virology 262: Martin HJ, Lee JM, Walls D, Hayward SD (27) Manipulation of the Toll-Like Receptor 7 Signaling Pathway by Epstein-Barr Virus. Journal of Virology 81: Younesi V, Shirazi F, Memarian A, Amanzadeh A, Jeddi-Tehrani M, et al. (214) Assessment of the effect of TLR7/8, TLR9 agonists and CD4 ligand on the transformation efficiency of Epstein-Barr virus in human B lymphocytes by limiting dilution assay. Cytotechnology 66: Fathallah I, Parroche P, Gruffat H, Zannetti C, Johansson H, et al. (21) EBV Latent Membrane Protein 1 Is a Negative Regulator of TLR9. The Journal of Immunology 185: van Gent M, Griffin BD, Berkhoff EG, van Leeuwen D, Boer IGJ, et al. (211) EBV Lytic-Phase Protein BGLF5 Contributes to TLR9 Downregulation during Productive Infection. The Journal of Immunology 186: Paludan SR, Bowie AG, Horan KA, Fitzgerald KA (211) Recognition of herpesviruses by the innate immune system. Nat Rev Immunol 11: Li T, Chen J, Cristea IM (213) Human cytomegalovirus tegument protein pul83 inhibits IFI16-mediated DNA sensing for immune evasion. Cell Host & Microbe 14: Johnson KE, Chikoti L, Chandran B (213) Herpes 158

159 Summarizing discussion Simplex Virus 1 Infection Induces Activation and Subsequent Inhibition of the IFI16 and NLRP3 Inflammasomes. Journal of Virology 87: Ansari MA, Singh VV, Dutta S, Veettil MV, Dutta D, et al. (213) Constitutive Interferon-Inducible Protein 16-Inflammasome Activation during Epstein-Barr Virus Latency I, II, and III in B and Epithelial Cells. Journal of Virology 87: Rigaud S, Fondaneche MC, Lambert N, Pasquier B, Mateo V, et al. (26) XIAP deficiency in humans causes an X-linked lymphoproliferative syndrome. Nature 444: Azzi T, Lunemann A, Murer A, Ueda S, Béziat V, et al. (214) Role for early-differentiated natural killer cells in infectious mononucleosis. Blood 124: Chijioke O, Muller A, Feederle R, Barros M, Krieg C, et al. (213) Human Natural Killer Cells Prevent Infectious Mononucleosis Features by Targeting Lytic Epstein-Barr Virus Infection. Cell Reports 5: Pappworth IY, Wang EC, Rowe M (27) The Switch from Latent to Productive Infection in Epstein-Barr Virus- Infected B Cells Is Associated with Sensitization to NK Cell Killing. Journal of Virology 81: Chung BKT (213) Innate immune control of EBVinfected B cells by invariant natural killer T cells. Blood 122: Sanchez DJ, Gumperz JE, Ganem D (25) Regulation of CD1d expression and function by a herpesvirus infection. J Clin Invest 115: Rao P, Pham HT, Kulkarni A, Yang Y, Liu X, et al. (211) Herpes Simplex Virus 1 Glycoprotein B and US3 Collaborate To Inhibit CD1d Antigen Presentation and NKT Cell Function. Journal of Virology 85: Chaigne-Delalande B, Li FY, O Connor GM, Lukacs MJ, Jiang P, et al. (213) Mg2+ Regulates Cytotoxic Functions of NK and CD8 T Cells in Chronic EBV Infection Through NKG2D. Science 341: Seminars in Cancer Biology 18: Wang J, Quake SR (214) RNA-guided endonuclease provides a therapeutic strategy to cure latent herpesviridae infection. Proc Natl Acad Sci U S A 111: Yuen KS, Chan CP, Wong NHM, Ho CH, Ho TH, et al. (214) CRISPR/Cas9-mediated genome editing of Epstein- Barr virus in human cells. Journal of General Virology (Published ahead of print): 1.199/jgv Chatterjee B, Leung CS, Munz C (214) Animal models of Epstein Barr virus infection. Journal of Immunological Methods 41: Stevenson PG, Simas JP, Efstathiou S (29) Immune control of mammalian gamma-herpesviruses: lessons from murid herpesvirus-4. Journal of General Virology 9: Berges BK, Tanner A (214) Modelling of human herpesvirus infections in humanized mice. Journal of General Virology 95: Munz C (214) Viral infections in mice with reconstituted human immune system components. Immunology Letters 161: Strowig T, Gurer C, Ploss A, Liu YF, Arrey F, et al. (29) Priming of protective T cell responses against virusinduced tumors in mice with human immune system components. The Journal of Experimental Medicine 26: Traggiai E, Chicha L, Mazzucchelli L, Bronz L, Piffaretti JC, et al. (24) Development of a Human Adaptive Immune System in Cord Blood Cell-Transplanted Mice. Science 34: Wang F (213) Nonhuman primate models for Epstein- Barr virus infection. Current Opinion in Virology 3: Pappworth IY, Wang EC, Rowe M (27) The Switch from Latent to Productive Infection in Epstein-Barr Virus- Infected B Cells Is Associated with Sensitization to NK Cell Killing. Journal of Virology 81: Ulbrecht M, Modrow S, Srivastava R, Peterson PA, Weiss EH (1998) Interaction of HLA-E with Peptides and the Peptide Transporter In Vitro: Implications for its Function in Antigen Presentation. The Journal of Immunology 16: Griffin BD, Gram AM, Mulder A, van Leeuwen D, Claas FHJ, et al. (213) EBV BILF1 Evolved To Downregulate Cell Surface Display of a Wide Range of HLA Class I Molecules through Their Cytoplasmic Tail. The Journal of Immunology 19: Delecluse HJ, Feederle R, Behrends U, Mautner J (28) Contribution of viral recombinants to the study of the immune response against the Epstein-Barr virus. 159

160

161 Addendum * List of abbreviations Nederlandse samenvatting Dankwoord Curriculum vitae

162 Addendum List of abbreviations AIDS Acquired immunodeficiency syndrome ALR AIM2-like receptor APC Antigen-presenting cell ATP Adenosine triphosphate BART BamHI fragment A rightward transcript BL Burkitt s lymphoma CAEBV Chronic acute EBV CARD Caspase recruitment domain CBP CREB-binding protein CD Cluster of differentiation cdc Conventional dendritic cell CIITA Class II, major histocompatibility complex, transactivator CLR C-type lectin receptors CRISPR Cluster of regularly interspaced short palindromic repeats CRL Cullin RING ligase CSF-1 Colony-stimulating factor-1 DAI DNA-dependent activator of IFN-regulatory factors DAMP Danger-associated molecular pattern dsrna double-stranded RNA DUB Deubiquitinase E Early EBER EBV-encoded non-coding RNA EBNA EBV nuclear antigen EBV Epstein-Barr virus ER Endoplasmic reticulum GPCR G-protein coupled receptor HCMV Human cytomegalovirus HHV Human herpesvirus HIS Humanized immune system HIV Human immunodeficiency virus HL Hodgkin s lymphoma HLA Human leukocyte antigen HSV Herpes simplex virus IE Immediate-early IFI16 Interferon inducible protein 16 IFNAR Type I IFN receptor IKK Inhibitor of NF-κB kinase IL Interleukin IM Infectious mononucleosis inkt Invariant natural killer T cells IPS-1 IFN-β-promoter stimulator 1 IRAK IL-1-receptor-associated kinase IRF Interferon-regulatory factor ISG Interferon-stimulated genes ISGF3 Interferon-stimulated gene factor 3 ISRE Interferon-stimulated response element IκBα Inhibitor of NF-κB α JAK Janus kinase KSHV Kaposi s sarcoma-associated herpesvirus 162

163 List of abbreviations L Late LCL Lymphoblastoid cell line LMP Latent membrane protein MAPK Mitogen-activated protein kinase MAVS Mitochondrial antiviral signaling MCMV Murine cytomegalovirus MDA5 Melanoma differentiation-associated gene 5 MHC Major histocompatibility complex MHV68 Murine herpesvirus 68 MICB MHC class I polypeptide-related sequence B mirna microrna NEMO NF-κB essential modulator NF-κB Nuclear factor κb NK-cell Natural killer cells NLR NOD-like receptors NOD Nucleotide-binding, oligomerization domain NPC Nasopharyngeal carcinoma OAS 2,5 -oligoadenylate synthetase ORF Open reading frame PAMP Pathogen-associated molecular pattern PBMC Peripheral blood mononuclear cells pdc Plasmacytoid dendritic cell PI3K Phosphatidylinositide 3-kinase PKR RNA-dependent protein kinase PML-bodies Promyelocytic leukemia bodies PRR Pattern-recognition receptor PTLD Post-transplant lymphoproliferative disease qpcr Quantitative PCR RIG-I Retinoic-acid inducible gene I RISC RNA-induced silencing complex RLR RIG-I-like receptor shrna short hairpin RNA SOCS Suppressor of cytokine signaling ssrna single-stranded RNA STAT Signal transducer and activator of transcription STING stimulator of interferon genes TAP Transporter associated with antigen processing TBK1 TANK-binding kinase TCR T-cell receptor TGF Transforming growth factor TLR Toll-like receptor TNF Tumor necrosis factor TRAF TNF-receptor associated factor TYK2 Tyrosine kinase 2 UL Unique long vhs Virion host shutoff VZV Varicella Zoster virus XLP X-linked lymphoproliferative disease * 163

164 Addendum Nederlandse samenvatting Ons afweersysteem beschermt ons tegen de vele potentiele ziekteverwerkkers die we dagelijks in onze omgeving tegenkomen. Het deel van het afweersysteem dat we het aangeboren afweersysteem noemen, herkent binnengdringende virussen en bacteriën binnen enkele minuten en is belangrijk voor bescherming tijdens de eerste fase van de infectie. Voor deze snelle herkenning wordt gebruikt gemaakt van receptoren op het oppervlak van cellen die de aanwezigheid van virussen kunnen waarnemen. Een belangrijke groep van deze receptoren wordt gevormd door de familie van de Toll-like receptoren (TLRs). Als deze receptoren virusonderdelen herkennen activeren zij signaleringsroutes die leiden tot het aanzetten van genen met antivirale activiteit, bijvoorbeeld de interferonen. Op deze manier wordt infectie door virussen geremd of voorkomen. De meeste virussen hebben echter methodes ontwikkeld om herkenning en/of eliminatie door het afweersysteem te omzeilen en zo alsnog infecties te veroorzaken. Het Epstein-Barr virus (EBV) veroorzaakt levenslange infecties in meer dan 9% van alle volwassenen over de hele wereld. Vaak zorgt infectie voor weinig problemen, maar in sommige gevallen treedt de ziekte van Pfeiffer (mononucleosis infectiosa) op, of kan het virus bepaalde vormen van kanker veroorzaken. Dit is met name het geval wanneer het afweersysteem wordt onderdrukt, bijvoorbeeld door medicatie of infectie met HIV. EBV infectie kent twee fases. Tijdens de latente fase is het virus in een slapende toestand, ondergedoken voor het afweersysteem. Gedurende de productieve fase worden actief nieuwe virusdeeltjes aangemaakt om overgedragen te worden naar andere personen. EBV wordt herkend door een aantal van de receptoren van het aangeboren immuunsysteem. De wijde verspreiding van EBV geeft echter aan dat dit virus zeer goed in staat is het afweersysteem te ontwijken. Hoe het virus dit voor elkaar krijgt is voor een groot deel onbekend. Om nieuwe en betere behandelmethoden te ontwikkelen tegen aandoeningen veroorzaakt door EBV is het belangrijk een beter inzicht te verwerven in het samenspel tussen dit virus en het afweersysteem. In dit proefschrift is daarom onderzocht hoe EBV interfereert met de eerste herkenning van het virus door het afweersysteem, met name door de Toll-like receptoren. Hoofdstuk 1 geeft een overzicht van de levenscyclus van EBV en de onderdelen van het afweersysteem die een rol spelen bij het tegengaan van infecties met dit virus. Daarnaast worden de reeds bekende ontwijkingsstrategieën beschreven die EBV gebruikt om het afweersysteem te omzeilen. In hoofdstuk 2 beschrijven we dat de expressie van een aantal Toll-like receptoren op het oppervlak van de cel is verminderd tijdens zowel de latente als de productieve fase van EBV infectie. Met name de expressie van één specifieke receptor, TLR9, wordt drastisch verlaagd. 164

165 Nederlandse samenvatting Het virale eiwit BGLF5 is hier mede verantwoordelijk voor. Daarnaast vonden we dat TLR9 een rol speelt bij het herkennen van EBV deeltjes. De verminderde aanwezigheid van deze receptor tijdens EBV infectie lijkt dus een manier te zijn van het virus om herkenning door het afweersysteem tegen te gaan en zo een antivirale reactie te ontwijken. In hoofdstuk 3 gaan we dieper in op het effect van het virale eiwit BGLF5 op componenten van het afweersysteem op het celoppervlak. De niveaus van zowel TLR2 als CD1d, beide betrokken bij de afweerreactie tegen EBV, waren verlaagd in aanwezigheid van BGLF5. Het reduceren van de expressie van BGLF5 door middel van short-hairpin RNAs liet zien dat BGLF5 slechts tot op zekere hoogte betrokken is bij het reduceren van andere eiwitten op het oppervlak van de cel. Dit wijst op het bestaan van nog onbekende, meer specifieke virale ontwijkingsmechansimen. In hoofdstuk 4 worden de effecten van een ander EBV eiwit, BPLF1, beschreven. BPLF1 interfereert met de signaleringsroute die wordt geactiveerd na herkenning van EBV door de TLRs aan het oppervlak van de cel. Door het verwijderen van ubiquitinegroepen van enkele eiwitten in deze route wordt activatie geremd, wat leidt tot een verminderde productie van eiwitten met antivirale activiteit. Ook BPLF1 draagt dus bij aan het ontwijken van de antivirale afweerreactie. In hoofdstuk 5 laten we zien dat EBV naast eiwitten ook virale micrornas gebruikt om de hoeveelheid van een eiwit (CBP) met een functie in het aangeboren afweersysteem te verminderen. Dit leidt tot een verlaagde gevoeligheid voor interferonen, cellulaire boodschappers die antivirale effecten kunnen induceren. De geïnfecteerde cel is daardoor minder goed in staat antivirale reacties te starten en omliggende cellen te waarschuwen voor, en te beschermen tegen, de aanwezigheid van virussen. Tot slot worden deze bevindingen samengevat en bediscussieerd in hoofdstuk 6, waarbij ook de toekomstperspectieven van het onderzoek naar ontwijking van het afweersysteem door EBV worden geformuleerd. Samengevat wordt duidelijk dat EBV verschillende manieren heeft ontwikkeld om herkenning van virusdeeltjes door het aangeboren afweersysteem te voorkomen en de antivirale mechanismen, die daar normaal gesproken op volgen, te remmen. Dit stelt EBV in staat levenslange infecties te veroorzaken. De in dit onderzoek opgedane kennis vergroot ons inzicht in de ingenieuze manieren waarop virussen het menselijk afweersysteem omzeilen en maakt het mogelijk strategieën te ontwikkelen om de complicaties die het gevolg zijn van infecties met het Epstein-Barr virus tegen te gaan. * 165

166 Addendum Dankwoord Nu rest nog het allerlaatste stukje van dit proefschrift, het dankwoord. Als ik terugdenk aan de afgelopen OIO-jaren is het een erg drukke maar bovenal leuke en gezellige tijd geweest. Ondanks de vele wisselingen aan mensen heb ik het altijd erg naar mijn zin gehad binnen de afdeling en stond iedereen altijd open voor samenwerkingen, overleg, en gezelligheid. Vele personen van binnen en buiten de afdeling hebben daardoor direct of indirect bijgedragen aan de totstandkoming van dit proefschrift. Hiervoor wil ik alle collega s, vrienden en familie enorm bedanken, en enkele personen in het bijzonder. Maaike, toen ik begon verhuisde het lab net van Leiden naar Utrecht, en inmiddels ben jij alweer terug in Leiden, waardoor ik daar ook nog wel eens kwam de laatste tijd. Ik wil je bedanken voor alle begeleiding gedurende de afgelopen jaren en alles wat je me hebt bijgebracht over het doen van onderzoek en het schrijven van papers. Alhoewel het soms leek alsof het nooit af zou komen, ben ik zeer tevreden met dit eindresultaat! De vele pagina s handgeschreven correcties hebben me een expert gemaakt in het ontcijferen van jouw handscrift. Emmanuel, jouw positieve en relativerende kijk op alle zaken werkte altijd erg motiverend. Het was prettig om altijd even langs te kunnen lopen om meer (of minder) belangrijke dingen te bespreken. Ingrid, dat jouw naam opduikt boven elk hoofdstuk van dit proefschrift maakt meteen duidelijk hoe belangrijk jouw harde werk en tomeloze inzet zijn geweest voor het slagen van deze projecten, super bedankt daarvoor! Ik vond het altijd erg prettig om met je samen te werken en het was cool-deluxe (of hoe schrijf je dat?) om naast elkaar in het lab bezig te zijn! Jos, bedankt voor de mogelijkheid om bij jou aan de slag te gaan en de ruimte die ik kreeg om deze promotie af te ronden. Marjolein, fijn dat het mirna/ifn project een mooie afronding lijkt te krijgen! De overige (ex-)leden van de afdeling en in het bijzonder van de Wiertz groep (de weirdo s ): Bryan, Danielle, Robert Jan, Rutger, Sytse, Ellen, Ferdy, Mike, Anouk, Jasper, Hanneke, Anna, Elisabeth, Wouter, Femke, en studenten Janneke, Amanda, en Kirsten: bedankt voor de fijne samenwerkingen, ondersteuning en andere bijdragen! De sportieve ondernemingen waren van groot belang om (promotie-)stress kwijt te raken. Naast de wekelijkse wielrentochtjes in wisselende samenstelling met Rutger, Bart, Bas, Alex, Axel, Jelle, Steven, Evelien O., Daphne, Maaike, Claudia, Ron, Manouk, Lindert en Nienke waren er natuurlijk de tochten naar de Koepel en NVMM en de halve-marathon hardlooptripjes naar Parijs en Berlijn! Verder wil ik alle (ex-)oio s bedanken voor de gezelligheid tijdens borrels, feestjes, retraites, de Koepels en (Ardennen-)weekendjes. My ex-roomies: Ana, Fernanda, Cassandra, Evelien O., Rutger, and Jery. The time I have spent in the office definitely changed my life (I m still in doubt whether that is in a good or a bad way though ;-)). I definitely had a lot of fun sharing in all the women-talk, although I was glad when Rutger and later Jery joined to balance the female majority (a little bit at 166

167 Dankwoord least). The occasional coffee break/annoy-visit during the writing of my thesis really helped to acquire new inspiration and energy. I m glad I was still allowed to join the room outings and dinners after I moved out! Ook buiten het lab zorgden menig (sportief) uitstapje voor de nodige ontspanning. De jaarlijkse combinatie van fietsen en bier/whisky-proeverijen met Rutger, Indra, Straat, Bert, Timon, Ron, Arend. Gelukkig was er ook wel eens een dag zónder regen! Gijsbert, Tim, Eric, Bernard, Rene en Arie bedankt voor alle onvergetelijke whisky/wijn/pokeravonden en andere activiteiten! Ook de Dingleberries-activiteiten heb ik altijd erg van genoten. Begonnen als pubquizteam, maar uitgegroeid tot een hechte vriendengroep met (half-) jaarlijke weekendtripjes. Marc, Carlijn, Tieme, Robert, Yannick, Ralph, Lianne, Martijn, Miriam, Gijs, Suzette, Sander en Marieke, bedankt voor alle mooie momenten en (promotie) discussies de afgelopen jaren. Oud-I&Iers, uiteindelijk is toch iedereen gaan promoveren en het grootste deel heeft de promotie zelfs al afgerond. Alhoewel het vinden van een geschikte datum om af te spreken nog weleens lastig was, hebben we vele gezellige avonden beleefd! Maarten, Janita, Koos, Eline, Erik, Eveline, Leon, Margot, ondanks dat we elkaar de laatste tijd minder zien door afstand en drukte zijn onze ontmoetingen altijd al snel weer als vanouds! Paranimfen Marc en Rutger, dat ik jullie beiden (onafhankelijk) enkele weken onderdak heb mogen bieden was uiteraard niet de enige reden om jullie als paranimfen te kiezen. Marc, de laatste volgens mij alweer vijftien jaar hebben we veel samen meegemaakt en zijn we zelfs enkele jaren huisgenoten geweest. De squashpartijtjes en EU-avondjes waren belangrijke momenten om alles weer even door te nemen en op een rijtje te zetten. Rutger, de laatste tien jaar zijn onze paden nooit lang uiteen geweken waardoor er maar weinig dagen zijn in de afgelopen jaren dat we elkaar niet gezien hebben. Fijn dat we daardoor zowel binnen als buiten het lab alles (of in ieder geval veel) konden delen! Ik ben benieuwd wat onze volgende stap gaat worden. Naast collega s en vrienden zorgden ook familie voor vele gezellige momenten. De (paas) weekendjes, verjaardagen, barbecues en andere activiteiten met de familie Hofmans en famillie van Gent zal ik niet snel vergeten! Wouter, Anke, Thijs, Lichelle: bedankt voor de gezellige feestjes, avonden, en stedentripjes. Gelukkig waren er ook altijd wel wat klusjes te doen als ik in de buurt was. Pap en mam, ik waardeer het heel erg dat jullie altijd klaar stonden met hulp en ondersteuning tijdens dit promotietraject, maar ook voor alles daarbuiten! Het is fijn om te weten dat ik altijd op jullie kan terugvallen. Caroline, ik ben ontzettend blij dat ik jou aan mijn zijde had tijdens de afronding van dit proefschrift. Zonder jou was dat een heel stuk zwaarder en zeker saaier geweest! * Michiel 167

168 Addendum About the author Michiel van Gent was born on November 29, 1984 in Nijmegen, The Netherlands. After finishing his secondary school education at Merlet College Land van Cuijk (gymnasium, cum laude) in 23, he commenced his Biomedical Sciences bachelor s studies at Utrecht University. Upon graduating in 26, he continued with the Infection and Immunity master s program at Utrecht University. As part of this program, one internship has been performed at the Virology group of the Veterinary Faculty of Utrecht University under supervision of Dr. Xander de Haan, and a second internship in the lab of Prof. dr. Hidde Ploegh at the Whitehead Institute for Biomedical Research in Boston, MA, USA. During this period he also participated in the X-track honors program. After graduating in 29 (cum laude), he started his PhD under supervision of Dr. Maaike Ressing and Prof. dr. Emmanuel Wiertz at the Virology research group of the department of Medical Microbiology at the University Medical Center Utrecht, The Netherlands. The results of this research are described in this thesis and have been published in various international scientific journals. Currently, he is employed as a scientific researcher in the research group of Prof. dr. Jos van Strijp in the department of Medical Microbiology at the University Medical Center Utrecht, The Netherlands. 168