The cyprinid herpesvirus 3 (CyHV-3, genus Cyprinivirus, family Alloherpesviridae, order Herpesvirales) is the aetiological agent of a highly contagious and lethal disease affecting koi and common carp (Cyprinus carpio). Common carp is currently the third most important fish species of aquaculture with a world production between 3 and 5 million tons per year. A safe and efficacious attenuated vaccine compatible with mass vaccination is needed to control CyHV-3. In this thesis, we produced an ORF56-57 double deleted (Δ56-57) attenuated recombinant vaccine candidate using prokaryotic mutagenesis. The safety of this vaccine candidate was demonstrated using in vivo bioluminescent imaging system (IVIS), qPCR, and histopathological examination. In comparison to the parental wild type strain, the vaccine candidate replicated at lower levels and spread less efficiently to secondary sites of infection. Transmission experiments allowing contamination through water with or without additional physical contact between fish demonstrated that the vaccine candidate has a reduced ability to spread from vaccinated fish to naïve sentinel cohabitants. Finally, challenge experiments demonstrated that the vaccine candidate induces a protective mucosal immune response at the portal of entry. The first part of this thesis led to the rational development of a recombinant attenuated vaccine against CyHV-3 compatible with mass vaccination of carp.

In the second part of this thesis, the relative contributions of ORF56 and ORF57 to the safety and efficacy profile of the Δ56-57 vaccine candidate have been assessed by producing and phenotyping viruses individually deleted for ORF56 or ORF57. Inoculation of these viruses to carp demonstrated that the deletion of ORF56 did not affect virulence, whereas the absence of ORF57 led to an attenuation comparable to, though slightly less than, that of the Δ56-57 vaccine candidate. To demonstrate further the role of ORF57 as a key virulence factor, a mutant retaining the ORF57 region but unable to express the ORF57 protein was produced by inserting multiple in-frame stop codons into the coding region. Analysis of this virus in vivo revealed a safety and efficacy profile comparable to that of the double deleted virus. These findings show that CyHV-3 ORF57 encodes an essential virulence factor.

CyHV-3 is considered as the prototype of the genus Cyprinivirus which encompasses a growing list of phylogenetically related viruses causing massive economical losses to the aquaculture sector. This thesis led for the first time to the identification of an essential virulence factor of CyHV-3 and to the demonstration that recombinant viruses deleted for this gene are potential vaccine candidates. All cypriniviruses described to date encode orthologues of CyHV-3 ORF57. Consequently, the present work opens new perspectives for the development of recombinant attenuated vaccines against this economically important viral genus.

Boutier M, Ronsmans M, Rakus K, Jazowiecka-Rakus J, Vancsok C, Morvan L, Peñaranda MMD, Stone DM, Way K, van Beurden SJ, Davison AJ, Vanderplasschen A: Chapter Three - Cyprinid Herpesvirus 3: An Archetype of Fish Alloherpesviruses. In Adv Virus Res. Volume 93. Edited by Margaret Kielian KM, Thomas CM. 2015, 93:161-256

Boutier M, Ronsmans M, Ouyang P, Fournier G, Reschner A, Rakus K, Wilkie GS, Farnir F, Bayrou C, Lieffrig F, Li H, Desmecht D, Davison AJ, Vanderplasschen A: Rational development of an attenuated recombinant cyprinid herpesvirus 3 vaccine using prokaryotic mutagenesis and in vivo bioluminescent imaging. PLoS Pathog 2015, 11:e1004690.

Boutier M, Gao Y, Vancsok C, Suarez NM, Davison AJ, Vanderplasschen A: Identification of an essential virulence gene of cyprinid herpesvirus 3. Antiviral Res 2017 145: 60-69.

Cyprinid herpesvirus 3 (CyHV-3) is the etiological agent of a highly contagious and lethal disease affecting koi and common carp worldwide. Beyond its economic importance, this virus turned out to be an interesting subject for fundamental research and is currently considered as the archetype of Alloherpesviridae. The divergence of this family with the Herpesviridae is ancestral, with almost no significant homology between CyHV 3 genes and those of Herpesviridae. Consequently, the extensive knowledge acquired for the latter cannot be used to predict CyHV 3 biological features. Virion transmembrane proteins (VTPs) are well documented in Herpesviridae for their involvement in crucial processes such as entry, immune evasion, morphogenesis and egress of progeny virions from the host cells. In contrast, very little is known about the functions of these proteins in CyHV 3 or in any of the alloherpesviruses. The main objective of this thesis was to provide a first functional characterization of CyHV-3 virion transmembrane proteome. The experimental work performed is summarized in two main chapters.

The first study aimed to update the list of known CyHV-3 VTPs and to determine those that are essential to viral growth in vitro. Using mass spectrometry approaches and mutagenesis experiments, we identified 16 VTPs in the CyHV-3 FL strain, among which 8 turned out to be essential to viral growth in vitro. The non-essential VTPs were further assessed quantitatively for their relative importance in vitro and in vivo. ORF25, ORF64, ORF108, ORF132, ORF136, ORF148, and ORF149 were shown to affect viral growth in vitro; while the lack of ORF148 or ORF25 caused attenuation to a minor or major extent in vivo, respectively. Finally, we showed that a mutant lacking ORF25 was highly attenuated but induced a moderate immune protection under the conditions tested.

Bioinformatic analyses suggested that CyHV-3 ORF27 encodes a VTP. However, no protein corresponding to this ORF was detected in our proteomic analyses. Analyses of genome sequence and protein expression demonstrate that the FL strain, like several other laboratory strains, do not express ORF27 due to presence of various mutations, while it encodes a VTP in field strains. These observation led us to hypothesize that the deletion of ORF27 could confer a selective advantage to viral growth in vitro, while the expression of a functional pORF27 could confer a selective advantage in vivo. These hypotheses were addressed by producing recombinant viruses. In vitro, the CyHV 3 strain lacking ORF27 expression was shown to have a replicative advantage, especially during co-infection with a viral strain expressing this protein. In contrast, in vivo, no difference between both genotypes could be detected in the experimental conditions tested. This observation suggests that the biological functions of ORF27 cannot be revealed in the laboratory conditions used.

In conclusion, this thesis provides a first functional characterization of the virion transmembrane proteome of CyHV-3. It represents a firm basis for further research on alloherpesvirus VTPs.

Vancsok C.,* Peñaranda M.,* Raj V., Leroy B., Jazowiecka-Rakus J., Boutier M., Gao Y., Wilkie G., Suárez N., Wattiez R., Gillet L., Davison A., and Vanderplasschen A. Proteomic and functional analyses of the virion transmembrane proteome of cyprinid herpesvirus 3. Journal of Virology, 2017, 91(21): e01209-17, doi:10.1128/JVI.01209-17.

Myxoma virus belongs to the Poxviridae family and induces a severe and often lethal disease called myxomatosis in the European rabbit. Numerous genes of the virus encode for proteins with immunomodulatory functions. Among them, it has been demonstrated that the virus encodes for an α-2,3-sialyltransferase that catalyzes the transfer of sialic acid to the N and O-glycans of glycoproteins and glycolipids. Glycans, and especially sialic acids, are known to play significant roles in interactions and recognition between cells and between cells and viruses.

The aim of this thesis was to better understand the roles played by the viral α-2,3-sialyltransferase in the pathogenesis of myxomatosis. Using the Lausanne strain as parental strain, we have constructed strains deficient for the M138L gene encoding the α-2,3-sialyltransferase, and a revertant strain. Although no difference could be detected between the strains in vitro, the M138L knockout strains were highly attenuated in vivo. Especially, we observed more neutrophils at the beginning of infection and then an important mononuclear cells infiltrate in the dermis of the primary sites of infection for rabbits infected with the M138L knockout strains.

We then wanted to determine which viral and / or cellular proteins were affected by the viral α-2,3-sialyltransferase, to check if their modification could explain the phenotype that we observed. We showed that the viral M-T7 protein was sialylated by the viral enzyme. M-T7 is a homologue of the interferon-γ receptor which is also able to bind and inhibit several chemokines, to block the chemokine gradient formation and to inhibit the inflammatory cells attraction towards the site of infection. However, we could not detect any effect of sialylation on the secretion or the activity of the M-T7 proteins purified from the four strains, either in vitro or in vivo.

Altogether , these results show that the viral α-2,3-sialyltransferase is a virulence factor of the virus that could act directly or indirectly on several immunomodulatory viral proteins, and so on the immune cells attraction and adaptive immunity development.

BOUTARD B., VANKERCKHOVE S., MARKINE-GORIAYNOFF N., SARLET M., DESMECHT D., McFADDEN G., VANDERPLASSCHEN A., GILLET L. The α2,3-Sialyltransferase encoded by Myxoma virus is a virulence factor that contributes to immunosuppression. PLoS One, 2015, 10(2), 0118806.

Gammaherpesviruses (γHVs) are important human and animal pathogens. Thus, Kaposi’s sarcoma associated herpesvirus (KSHV) and Epstein-Barr virus (EBV) cause several lymphoproliferative disorders and are serious concerns of public health in humans. In animals, problems associated with γHV infections are particularly important in ongulates with viruses such as Ovine herpesvirus 2 or Alcelaphine herpesvirus 1. As most of these viruses replicate poorly in vitro and have no established in vivo infection models, γHVs infecting laboratory mice, such as Murid herpesvirus 4 (MuHV-4), have emerged as efficient and relevant models to study the biology of γHVs both in vitro and in vivo. Despite the fact that γHVs display the classical architecture of herpesviruses, the function of most of their structural proteins is still poorly defined. This is especially true for tegument proteins such as the protein encoded by the open reading frame 63 (ORF63) of KSHV and MuHV-4. The ORF63 gene is conserved among all members of the Herpesviridae family and its homologues have mainly been studies in Alphaherpesvirinae where they play an important role in capsid transport during entry and egress. Interestingly, a potential role in immune evasion has recently been proposed for the tegument protein encoded by KSHV ORF63. To gain insight about the roles of ORF63 in the lifecycle of a γHV, we generated null mutations in the ORF63 gene of MuHV-4. We showed that disruption of ORF63 was associated with a severe growth deficit both in vitro and in vivo. The latter deficit was mainly associated with a defect of replication in the lung but did not affect the establishment of latency in the spleen. From a functional point of view, inhibition of caspase-1 or of inflammasome did not restore the growth of the ORF63-deficient mutant, suggesting that the observed deficit was not associated with the immune evasion mechanism identified previously. Moreover, this growth deficit was also not associated with a defect in virion egress from the infected cells. In contrast, it appeared that MuHV-4 ORF63-deficient mutants failed to address most of their capsids to the nucleus during entry into host cells, suggesting that ORF63 plays a role in capsid movement. Finally, beside the annotated full-length form of pORF63, we showed that a short C-terminal isoform of pORF63 is also expressed during the MuHV-4 cycle and that this isoform might have an important role in the viral lifecycle. Altogether, the results obtained in this work have allowed us to better understand some specificities of the biology of γHVs. In the future, we hope that it could allow the development of new strategies against these infections.

LATIF M.B., MACHIELS B., XIAO X., MAST J., VANDERPLASSCHEN A., GILLET L. Deletion of Murid Herpesvirus 4 ORF63 Affects the Trafficking of Incoming Capsids toward the Nucleus. J Virol, 2015, 90, 2455-2472.

Among gammaherpesviruses, the Macavirus genus is composed of viruses associated with malignant catarrhal fever (MCF) and other phylogenetically related viruses. MCF is described as an acute and fatal lymphoproliferative disease of several species of ruminants, including cattle. Alcelaphine herpesvirus 1 (AlHV-1) and ovine herpesvirus 2 (OvHV-2) are able to persist for life in wildebeest and sheep, respectively, without causing any clinical sign or lesion. Both viruses are the most studied macaviruses. In susceptible species, AlHV-1 and OvHV-2 induce the wildebeest-derived (WD-) and the sheep-associated (SA-) form of MCF, respectively. The economic impact of WD-MCF in sub-Saharan Africa has largely been underestimated, which urges the need to better understand the mechanisms underlying the pathogenesis of MCF. Rabbits are used as experimental model to study WD-MCF. This species develops clinical signs and lesions that are indistinguishable from those observed in ruminants such as bovine. Recent studies have demonstrated that WD-MCF is associated with a severe proliferation and infiltration of latently infected CD8+ T cells in most of organs without detection of any free viral particles (Palmeira et al., 2013). Additionally, latent persistence of viral episomes mediated by the expression of the genome maintenance protein (GMP) encoded by AlHV-1 ORF73 (aLANA) was shown to be essential for WD-MCF induction. Among the rare transcripts expressed during gammaherpesvirus latency, non-coding RNAs such as microRNAs (miRNAs) and GMPs were shown to be expressed. Herpesviral miRNAs play roles in the regulation of latency, lymphomagenesis, cell cycle and immune evasion. GMPs are responsible for episome persistence in latently infected cells during mitosis while evading self-presentation on MHC class I to cytotoxic T cells, a mechanism known as cis-acting immune evasion. This work reflects our interest in the understanding of the mechanisms by which AlHV-1 is able to persist in vivo while remaining hidden from the immune surveillance.

The first study of the present thesis was dedicated to the identification of AlHV-1 encoded miRNAs and their potential roles in MCF induction. We identified 36 potential miRNAs expressed in infected cells derived from MCF-developing calves, of which 32 were expressed on the reverse strand of the genome in two main clusters. After validation of the expression of these 32 viral miRNAs, we showed that deletion of a cluster of 28 viral miRNAs did not affect viral growth in vitro or MCF induction in rabbits. The second study focused on the investigation of the self-inhibition of antigen presentation by aLANA and the potential role of such mechanism in MCF induction in the rabbit model. We demonstrated that, like its orthologues, aLANA was able to inhibit its own presentation to cytotoxic T lymphocytes through a mechanism involving dowregulation of its own protein synthesis. We found that the presence of the GE-rich region in the central acidic repeat domain of aLANA was responsible for this mechanism.

Finally, we generated an AlHV-1 recombinant strain expressing a GE-deleted aLANA protein and observed that the absence of GE did not affect viral growth in vitro and did not impair WD-MCF in the rabbit model, demonstrating that GE is not essential for viral persistence and episome maintenance. However, the induction of MCF was significantly delayed in 2 out of 4 infected rabbits, suggesting that the GE region of aLANA could be necessary to evade anti-viral cytotoxic responses during WD-MCF.

Macaviruses are able to persist in their natural host while cross-species transmission leads to the development of MCF in susceptible species. Recent studies have positioned gammaherpesvirus latency at the center of WD-MCF pathogenesis. Trying to decipher the mechanism by which AlHV-1 is able to persist in vivo and induce WD-MCF while remaining hidden from the immune surveillance is of interest to improve our current knowledge about macaviruses, and more broadly gammaherpesvirus pathogenesis.

Sorel O., Tuddenham L., Myster F., Palmeira L., Kerkhofs P., Pfeffer S. , et al. Small RNA deep sequencing identifies viral microRNAs during malignant catarrhal fever induced by alcelaphine herpesvirus 1. J Gen Virol. 2015;96(11):3360-72.

Sorel O., Benjamin G Dewals. MicroRNAs in large herpesvirus DNA genomes: recent advances. BioMolecular concepts. 2016

Cyprinid herpesvirus 3 (CyHV-3) is a highly pathogenic virus responsible for a lethal disease in both common and koi carp (Cyprinus carpio). The common carp is one of the most important freshwater species cultivated for human consumption. Its colourful subspecies koi is grown for personal pleasure and exhibitions. Both common and koi carp are economically important and since its description in the late 1990s, the CyHV-3 has caused severe financial losses in these two carp industries worldwide. Because of its economic importance and its numerous original biological properties, CyHV 3 became rapidly an attractive subject for both applied and fundamental research.

The objectives of this thesis were to investigate the role of two unrelated innate immune mechanisms of carp in anti-CyHV-3 immunity. The first objective was to determine the role of epidermal mucus as an innate immune barrier against CyHV-3 entry during the early developmental stages of carp. To test this hypothesis, we investigated the sensitivity and the permissivity of carp to CyHV-3 during the early stages of its development. This hypothesis was tested using a recombinant CyHV-3 strain expressing luciferase as a constitutive reporter gene and in vivo bioluminescence imaging system. We demonstrated that carp are sensitive and permissive to CyHV 3 infection since hatching, but that their sensitivity remains relatively low in the two early developmental stages. Similarly to adults, we confirmed that the skin is the main portal of entry for the virus at early stages, and our results stress out the role of epidermal mucus as an innate immune defense of carp against pathogens even and especially at the early stages of development. The results of this study have been published in Veterinary research.

The second objective of this thesis consisted to investigate whether carp express behavioral fever when infected by CyHV-3; and if so, what could be the effect of this innate immune reaction on the development of CyHV 3 disease. When infected by pathogens, both endotherms and ectotherms can express a salutary reaction by increasing their body temperature. While in endotherms this reaction is called fever and depends on intrinsic thermogenesis, ectotherms like teleosts can only upregulate their body temperature by moving to warmer places, hence the term behavioral fever. When studying the pathogenesis of CyHV-3, we observed that carp infected at 24°C (the thermal preference of healthy carp) tended to concentrate around the tank heater when it was running. This observation led us to postulate that infected subjects could express behavioral fever in natural environments where temperature gradients exist. Using multi-chamber tanks encompassing a gradient from 24°C to 32°C, we observed that carp infected by CyHV-3 express a salutary behavioral fever that completely suppresses virus induced mortalities. The relatively late onset of behavioral fever with respect to clinical signs, viral replication and cytokine upregulation led us to postulate that this phenomenon could be delayed by the virus to retain its host at a permissive temperature thereby favoring its replication and spreading. As some herpesviruses have been shown to express soluble decoy cytokine receptors, we hypothesized that CyHV-3 could express such receptor(s) able to neutralize putative pyrogenic cytokines produced by the fish. We found that CyHV-3 ORF12 encodes a soluble decoy receptor for carp Tnfα and that this viral protein makes the virus capable of delaying the migration of infected fish to warmer environments. Remarkably, the study of the molecular mechanism through which the virus alters its host’s behavior led to the discovery of the first pyrogenic cytokine in ectotherms (Tnfα). This study is the first to report the ability of a vertebrate virus to alter the behavior of its host through the expression of a single gene. This second study was submitted for publication when this thesis was printed.

In conclusion, we investigated two innate immune mechanisms expressed by carp against CyHV 3 infection. The results generated in this thesis bring findings related to several scientific fields such as virology, immunology, animal behavior, evolution, ecology and even animal welfare.

1.BOUTIER M., RONSMANS M., OUYANG P., FOURNIER G., RESCHNER A., RAKUS K., WILKIE G. S., FARNIR F., BAYROU C., LIEFFRIG F., Li H., DESMECHT D., DAVISON A. J., VANDERPLASSCHEN A. Rational development of an attenuated recombinant Cyprinid herpesvirus 3 vaccine using prokaryotic mutagenesis and In Vivo bioluminescent imaging. PLoS Pathog, 2015a, 11, e1004690.
2.BOUTIER M., RONSMANS M., RAKUS K., JAZOWIECKA-RAKUS J., VANKSOC C., MORVAN L., PENARANDA M. M. D., STONE D. M., WAY K., VAN BEURDEN S. J., DAVISON A. J., VANDERPLASSCHEN A. Chapter Three - Cyprinid Herpesvirus 3: An Archetype of Fish Alloherpesviruses. In Adv Virus Res, 2015b, pp. 161-256. Edited by K. M. Margaret Kielian & C. M. Thomas: Academic Press.
3.RONSMANS M., BOUTIER M., RAKUS K., FARNIR F., DESMECHT D., ECTORS F., VANDECAN M., LIEFFRIG F., MELARD C., VANDERPLASSCHEN A. Sensitivity and permissivity of Cyprinus carpio to Cyprinid herpesvirus 3 during the early stages of its development: importance of the epidermal mucus as an innate immune barrier. Vet Res, 2014, 45, 100.

Among gammaherpesvirus, the Macavirus genus is composed of viruses associated to malignant catarrhal fever (MCF) and other phylogenetically related viruses. MCF is a frequently fatal lymphoproliferative disease. Three macaviruses inducing MCF have been entirely sequenced: alcelaphine herpesvirus 2 (AlHV-2), ovine herpesvirus 2 (OvHV-2) and alcelaphine herpesvirus 1 (AlHV-1). Sheep carries OvHV-2 asymptomatically while wildebeest is infected with AlHV-1 without developing any clinical signs or lesions. Both viruses represent the most studied macaviruses. In susceptible ruminants, OvHV-2 and AlHV-1 induce the sheep-associated form and the wildebeest-derived form of MCF (WD-MCF), respectively. Economic consequences of WD-MCF are significant in sub-Saharan Africa. WD-MCF is characterized by the proliferation and infiltration of lymphoblastoid T cells surrounding blood vessels and can be considered as a model for peripheral T cell lymphoma caused by a virus. Rabbits are used as experimental model to study MCF. This species develops clinical signs and lesions that they are indistinguishable from those observed in other susceptible species. Until recently, available data on WD-MCF pathogenesis were limited to the simple description clinical signs and lesions. Recently, it was demonstrated that CD8+ T cells proliferate and that this cellular expansion is associated with a severe increase of the viral load in PBMC and lymphoid organs (Dewals et al., 2008). The cloning of the AlHV-1 genome as an infectious and pathogenic bacterial artificial chromosome (BAC) has greatly facilitated the study of individual gene of AlHV-1 (Dewals et al., 2006). Among herpesviruses, viral semaphorins can only be found in members of the Macavirus genus. OvHV 2 encodes Ov3, and AlHV-2 and AlHV-1 encode A3, both genes encoding a semaphorin homolog. Semaphorins are proteins characterized by a conserved amino-terminal domain, the SEMA domain. The roles of the semaphorins on cytoskeleton dynamics have been widely studied. Viral semaphorins could mediate immune evasion mechanisms or viral dissemination and could be involved in specific properties of macaviruses.

The first study of the present thesis was dedicated to the investigation of the pathogenesis of WD-MCF and the role of latency. We investigated the distribution of the AlHV-1 infection in the lesions and demonstrated that the infiltration of CD8+ T cells in different lymphoid and non-lymphoid organs and tissues is directly associated with a non-productive viral infection. The second study focused on the A3 gene of AlHV-1 and its potential functions during WD-MCF. We showed that the A3 gene is expressed during the early phase of the viral infection and encodes a functional semaphorin that was termed AlHV-sema. AlHV-sema was able to induce cell retraction. Despite the observed independent acquisition of pox- and herpesvirus semaphorins, AlHV-sema inhibited phagocytosis by dendritic cells and migration to the draining lymph node through mechanisms similar to poxvirus semaphorin. AlHV-sema could also facilitate viral dissemination or confer immune evasion functions. Next, we investigated whether AlHV-sema could affect WD-MCF induction. We did not observe any effect of the absence of AlHV-sema expression during the development of WD-MCF after rabbit nasal infection.

On the other hand, we improved the knowledge about proteins that compose gammaherpesviruses virions. To this end, MuHV-4 extracellular virions were isolated and structural proteins were identified using liquid chromatography tandem mass spectrometry-based proteomic approaches. These analyses allowed the identification of 31 structural proteins encoded by the MuHV-4 genome which were classified as capsid (8), envelope (9), tegument (13) and unclassified (1) structural proteins. In addition, we estimated the relative abundance of the identified proteins in MuHV-4 virions by using exponentially modified protein abundance index analyses. In parallel, several host proteins were found in purified MuHV-4 virions including Annexin A2. Although Annexin A2 has previously been detected in different virions from various families, its role in the virion remains controversial. Interestingly, despite its relatively high abundance in virions, Annexin A2 was not essential for the growth of MuHV-4 in vitro.

Macaviruses are swine and ruminant gammaherpesviruses responsible for a latent asymptomatic infection in their natural species. The development of MCF in other ruminant susceptible species is due to cross-species transmission. During evolution, the gene selection in susceptible species is certainly reduced due to the fact that these species are dead-end hosts. Thus, it is difficult to address the role of AlHV-1 specific genes in MCF as these genes have evolved in other species. Nevertheless, we brought in this work important insight for our understanding of the pathogenesis of WD-MCF and we identified AlHV-sema as a potential immunoevasion factor.

1.DEWALS B., MYSTER F., PALMEIRA L., GILLET L., ACKERMANN M., VANDERPLASSCHEN A. Ex vivo bioluminescence detection of alcelaphine herpesvirus 1 infection during malignant catarrhal fever. J. Virol., 2011, 85, 6941-6954.
2.MYSTER F., PALMEIRA L., SOREL O., BOUILLENNE F., DEPAUW E., SCHWARTZ-CORNIL I., VANDERPLASSCHEN A., DEWALS B.G. Viral semaphorin inhibits dendritic cell phagocytosis and migration but is not essential for gamma-herpesvirus-induced lymphoproliferation in malignant catarrhal fever. J. Virol., 2015.

Gammaherpesviruses are archetypes of persistent viruses that have been identified in a range of animals from mice to man. They are host-range specific and establish lifelong latency of immunocompetent hosts. Most of the gammaherpesvirinae members are associated with neoplastic diseases. For example, the best studied gammaherpesviruses are Human herpesvirus 4 and 8 that are respectively associated with Burkitt’s lymphoma and Kaposi’s sarcoma. By opposition to its human counterparts, Murid herpesvirus-4 (MuHV-4) is able to replicate to high titers in cultured cells and is therefore an excellent candidate for studying gammaherpesvirus cycle.

On the one hand, we screened the cellular genome for host proteins that are important for the virus biology. RNA interference (RNAi) is a natural process that cells use to turn down, or silence, the activity of specific genes by inhibition of mRNAs. The selective and robust effect of RNAi on gene expression makes it a valuable research tool in cell culture because synthetic dsRNA introduced into cells can induce suppression of specific genes. RNAi may also be used for large-scale screens that systematically shut down each gene in the cell. In this first study, we described the use of high troughput screening by reverse transfection of cells of small interfering RNAs (siRNA), which, combined to fluorimetry explored 9,102 human genes (Ambion silencer druggable genome). The data were analysed for each siRNA (3/gene) and the results were obtained by bioinformatics. 65 genes appeared to be necessary for the completion of the MuHV-4 cycle and likely the initial steps of infection. Moreover, silencing of 110 genes increased MuHV-4 infection suggesting a possible anti-viral function.

On the other hand, we improved the knowledge about proteins that compose gammaherpesviruses virions. To this end, MuHV-4 extracellular virions were isolated and structural proteins were identified using liquid chromatography tandem mass spectrometry-based proteomic approaches. These analyses allowed the identification of 31 structural proteins encoded by the MuHV-4 genome which were classified as capsid (8), envelope (9), tegument (13) and unclassified (1) structural proteins. In addition, we estimated the relative abundance of the identified proteins in MuHV-4 virions by using exponentially modified protein abundance index analyses. In parallel, several host proteins were found in purified MuHV-4 virions including Annexin A2. Although Annexin A2 has previously been detected in different virions from various families, its role in the virion remains controversial. Interestingly, despite its relatively high abundance in virions, Annexin A2 was not essential for the growth of MuHV-4 in vitro.

Altogether, these results could not only help us to better understand the biology of gammaherpesviruses, but should also reveal new therapeutic targets to fight these infections.

1.Francois, S., S. Vidick, M. Sarlet, D. Desmecht, P. Drion, P. G. Stevenson, A. Vanderplasschen, and L. Gillet. 2013. Illumination of murine gammaherpesvirus-68 cycle reveals a sexual transmission route from females to males in laboratory mice. PLoS Pathog 9:e1003292.
2.Francois, S., S. Vidick, M. Sarlet, J. Michaux, P. Koteja, D. Desmecht, P. G. Stevenson, A. Vanderplasschen, and L. Gillet. 2010. Comparative study of murid gammaherpesvirus 4 infection in mice and in a natural host, bank voles. J Gen Virol 91:2553-2563.
3.Vidick, S., B. Leroy, L. Palmeira, B. Machiels, J. Mast, S. Francois, R. Wattiez, A. Vanderplasschen, and L. Gillet. 2013. Proteomic characterization of murid herpesvirus 4 extracellular virions. PLoS One 8:e83842.

The common carp is one of the most important freshwater fish species in aquaculture, and its colourful subspecies koi is grown for personal pleasure and competitive exhibitions. Both two subspecies are economically important. In the late 1990s, a highly contagious and lethal pathogen called koi herpesvirus (KHV) or cyprinid herpesvirus 3 (CyHV-3) began to cause severe financial losses in these two carp industries worldwide. In 2005, CyHV-3 has been classified in the Alloherpesviridae family of the order Herpesvirales. Because of its economic importance and its numerous original biological properties, CyHV 3 became rapidly an attractive subject for applied and fundamental research. However, to date, there is a little information on the roles of individual CyHV-3 genes in the biology of CyHV-3 infection or its pathogenesis. Moreover, there is a lack of safe and efficacious vaccine for the control of CyHV-3 disease. The goal of this thesis was to study the roles of CyHV-3 ORF134 encoding an IL-10 homologue in the biology of the infection. CyHV-3 ORF134 has been predicted to contain an 84 bp intron flanked by 2 exons encoding together a 179 amino acid product. Transcriptomic analyses reveal that ORF134 is expressed as a spliced early-late gene. The identification of the CyHV-3 secretome was achieved using 2D-LC MS/MS proteomic approach. This method led to the identification of 5 viral and 46 cellular proteins in concentrated infected cell culture supernatant. CyHV-3 ORF12 and ORF134 were amongst the most abundant proteins detected. To investigate the roles of ORF134 in the biological of the infection, a strain deleted for ORF134 and a derived revertant strain were produced by using BAC cloning and prokaryotic recombination technologies. Comparison of these strains demonstrated that CyHV-3 ORF134 does not contribute significantly to viral growth in vitro or to virulence in vivo in the present laboratory setting. The present study addressed for the first time the in vivo role of a vIL-10 encoded by a member of the family Alloherpesviridae. This study has been published in Veterinary Research. During the course of the first study, we obtained an unexpected recombination event while we were reconstituting infectious virus from mutated BAC plasmids. To generate a revertant ORF134 Del galK strain, CCB cells were co-transfected with the FL BAC ORF134 Del galK plasmid and the pGEMT-TK vector to remove the BAC cassette inserted in the ORF55 locus (encoding thymidine kinase). One of the clones obtained had an unexpected recombination leading to the deletion of ORF56 and ORF57 in addition to the expected deletion of ORF134. Unexpectedly, this triple deleted strain replicated efficiently in vitro, exhibited an attenuated phenotype in vivo and was proved to confer in a dose dependent manner an immune protection against a lethal challenge. The goal of the second experimental chapter was to investigate the role of the ORF56-57 and ORF134 deletions in the observed safety/efficacy profile of the triple deleted recombinant. To reach this goal, a collection of recombinant strains were produced using BAC cloning technologies, characterized and tested in vivo for their safety/efficacy profile. The results obtained demonstrated that the ORF56-57 deletion is responsible for the phenotype observed and that ORF134 deletion does not contribute to this phenotype significantly. Finally, the immune protection conferred by ORF56-57 deleted recombinant was investigated by challenging immunized fish with a wild type strain expressing luciferase as a reporter gene. In vivo imaging system (IVIS) analyses of immunized and challenged fish demonstrated that the immune response induced by the ORF56-57 deleted strain was able to prevent subclinical infection of the challenge strain. In conclusion, the present thesis addressed both fundamental and applied aspects of CyHV-3. For the first time, it investigated in vivo the roles of a viral IL-10 homologue encoded by a member of the family Alloherpesviridae. Importantly, it identified the ORF56-57 loci as target for production of safe and efficacious attenuated recombinant vaccines .

1.Ouyang P, Rakus K, van Beurden S, Westphal A, Gatherer D, Davison A, Gatherer D, Vanderplasschen A. Interleukin-10s encoded by viruses: a remarkable example of independent acquisitions of a cellular gene by viruses and its subsequent evolution in the viral genome. Journal of General Virology. November 2013.
2.Rakus K, Ouyang P, Boutier M, Ronsmans M, Reschner A, Vancsok C, Jazowiecka-Rakus J, Vanderplasschen A. Cyprinid herpesvirus 3: an interesting virus for applied and fundamental research. Veterinary Research. September 2013, 44:85.
3.Ouyang P, Rakus K, Boutier M, Reschner A, Leroy B, Ronsmans M, Fournier G, Scohy S, Costes B, Wattiez R, Vanderplasschen A. The IL-10 homologue encoded by cyprinid herpesvirus 3 is essential neither for viral replication in vitro nor for virulence in vivo. Veterinary Research. July 2013, 44:53.
4. Raj VS, Fournier G, Rakus K, Ronsmans M, Ouyang P, Michel B, Delforges C, Costes B, Farnir F, Leroy B, Wattiez R, Melard C, Mast J, Lieffrig F, Vanderplasschen A. Skin mucus of Cyprinus carpio inhibits cyprinid herpesvirus 3 binding to epidermal cells. Veterinary Research. August 2011, 42(1):92.

Gammaherpesviruses are the archetype of persistent viruses that have been identified in a series of animals ranging from mice to man. To date the study of transmission of these viruses in natural condition has been limited by the fact that no experimental transmission model exists. Establishment and characterization of a model of transmission are therefore critical points to evaluate strategies of interference with the epidemiological cycle of gammaherpesviruses. We are studying Murid herpesvirus 4 (MuHV-4) which has originally been isolated from naturally infected bank voles (Myodes glareolus). Although serological data indicate that closely related strains are present in wood mice (Apodemus sylvaticus) and domestic mice (Mus musculus), no experimental transmission of MuHV-4 has been demonstrated in laboratory mice, the classically used in vivo model. The objective of this work was therefore to fill this gap. In a first study, we performed a comparative characterization of the infection by MuHV-4 in mice and bank voles. Our results showed that the infectious process, the pathology and the latency establishment are similar in the two species, even if replication is quantitatively lower in bank voles than in mice. It therefore appeared that, Mus musculus represents a suitable host for studying gammaherpesvirus pathogenesis with MuHV-4. These results have been published in Journal of General Virology (J Gen Virol. 2010 Oct;91(Pt 10):2553-63). In a second study, thanks to in vivo imaging, we have been able to observe, for the first time, transmission of MuHV-4 in mice. We firstly showed that MuHV-4 reexcretion occurs in the genital tract of female mice at a period by which latency is considered as established. Ex vivo imaging, histology and PCR allowed us to demonstrate the presence of viral genomes in vaginal tissues and to localize viral replication at the external border of the vagina. We also demonstrated the transient and repetitive presence of infectious viruses in the vaginal cavity. Secondly, we demonstrated the implication of sexual steroid hormones in this re-excretion process. Indeed, we analyzed the infection of untreated mice, ovariectomized mice and ovariectomized mice complemented with estrogens and/or progesterone. These analyses revealed a positive role of estrogens in the observed re-excretion. Finally, based on these results, we tested MuHV-4 transmission in mice by creating different epidemiological conditions. In the conditions tested, vertical transmission did not occur, nor did horizontal transmission between individuals of the same gender. In contrast, we were able to observe sexual transmission to naïve males by serology, in vivo imaging and quantitative PCR. In conclusion, this work has on one hand demonstrated the quality of mice as an in vivo model for MuHV-4 studies and, on the other, it has shown for the first time the existence of re-excretion and sexual transmission of MuHV-4 amongst laboratory mice. The results of this work should therefore have implications for the study of gammaherpesviruses, but also more generally for the study of sexually transmissible infections.

Francois, S., Vidick, S., et al. (2010). "Comparative study of murid gammaherpesvirus 4 infection in mice and in a natural host, bank voles." Journal of General Virology, 91(Pt 10): 2553-2563.

Gammaherpesviruses are responsible of major pathologies either in human or animal populations. However, control of these infections presents a major challenge, as these viruses evolved to coexist with antibody and to resist neutralization. In order to unravel these immune-evasion mechanisms, studying these viruses in vivo in relevant host is an obligatory pre-requisite. Animal gammaherpesvirus, such as Bovine herpesvirus 4 (BoHV-4), allow us to tackle these fundamental questions in a much more accessible form than their human homologues. As neutralizing antibodies act mainly by blocking the entry of virions in target cells, a better knowledge of their entry mechanisms is necessary to improve control of these infections. Herpesviruses use a core fusion machinery composed by glycoproteins gB, gH and gL. gB is a class III viral fusion protein but cannot function on its own and requires the gH/gL heterodimer. Whilst gH acts probably as a co-fusogen, the role of gL is unknown. The PhD work investigated therefore the function of the BoHV-4 glycoprotein L and its protection against neutralizing antibodies. This work is divided in three parts.

1. Bovine herpesvirus-4 glycoprotein L is nonessential for infectivity but triggers virion endocytosis during entry

In order to define more generally what role gL plays in rhadinovirus infections, we disrupted its coding sequence in BoHV-4. BoHV-4 lacking gL showed altered gH glycosylation and incorporated somewhat less gH into virions but remained infectious. However, gL STOP virions showed poor growth associated with an entry deficit. Moreover, a major part of their entry defect appeared to reflect impaired endocytosis, which occurs upstream of membrane fusion itself. It therefore appeared that one of the major functions of gL in rhadinoviruses is to trigger virion endocytosis.

2. Study of the glycoprotein gL as a neutralizing target

In the second part of our work, we used BoHV-4 to determine how gL contributes to virus in vivo cycle and neutralization. While a lack of gL had no impact on the establishment and maintenance of BoHV-4 latency, sera directed against gL deficient virions neutralized significantly less WT virions than sera directed against WT virions. These results suggest therefore that gL-dependent epitopes are the main targets for neutralization of BoHV-4 virions.

3. Study of the glycoprotein L protection against neutralizing antibodies

Finally, we studied how BoHV-4 protects these epitopes from neutralization. Among the mechanisms of resistance to virus neutralization, several studies have highlighted the use of "Glycan shield" by viruses. We therefore studied the role of glycans in neutralizing antibodies evasion by BoHV-4. Firstly, we showed that glycans protect BoHV-4 virions from neutralization. Then, we identified by proteomic characterization of BoHV-4 extracellular virions, four heavily glycosylated protein in the virion that is gB, gL, gH and gp 180 which is potentially the most glycosylated. Gp180 appeared to be highly O-glycosylated and its absence increased the sensitivity of virions to neutralization by immune sera. Antibody had greater access to gB, gH and gL on gp180-deficient virions, including neutralizing epitopes on gL. It appeared therefore that gp180 provides a glycan shield for otherwise vulnerable viral epitopes. This finding is particularly interesting especially because the BoHV-4 possesses a glycosyltransferase encoded by the Bo17 gene. Western blot analyses showed that gp 180 is a target of this glycosyltransferase. As the Bo17 gene undergoes alternative splicing that generates an entire protein enzymatically active or a shorter inactive protein, we produced and characterized recombinant viruses expressing only the long form or the short form of pBo17. With these viruses, we have shown that BoHV-4, via alternative splicing of the gene Bo17, is able to modulate the glycosylation of gp180. In the future, we want to study the consequences of splicing on BoHV-4 neutralization.

In conclusion, we proposed in this thesis that one of the major functions of gL in rhadinoviruses is to trigger virion endocytosis. This work also demonstrated the importance of O-glycosylation to protect neutralizing epitopes including gL. In the future, this work could have both fundamental and applied perspectives.

Lété C, Palmeira L, Leroy B, Mast J, Machiels B, Wattiez R, Vanderplasschen A, Gillet L. Proteomic Characterization of Bovine Herpesvirus 4 Extracellular Virions. Journal of Virology. November 2012, Volume 86 : 11567-11580.

Lété C, Machiels B, Stenvenson PG, Vanderplasschen A, Gillet L. Bovine herpesvirus 4 glycoprotein L is non-essential for infectivity. Journal of Virology. Mars 2012, Volume 86 : 2653-2664.

Machiels B, Lété C, Guillaume A, Mast J, Stevenson PG, Vanderplasschen A, Gillet L, Antibody Evasion by a Gammaherpesvirus O-Glycan Shield. PLoS Pathogens. November 2011, 7(11) : e1002387.

The common carp is one of the most important freshwater species in aquaculture and its colourful subspecies koi is grown for personal pleasure and exhibitions. Both subspecies are economically important. In the 1990s, a highly contagious and lethal pathogen called koi herpesvirus (KHV) or cyprinid herpesvirus 3 (CyHV-3) began to cause severe financial losses in these two carp industries worldwide. Because of its economic importance and its numerous original biological properties, CyHV 3 became rapidly an attractive subject for applied and fundamental research. The goal of this thesis was to identify the portals of entry of CyHV-3 in carp. This information is essential to understand the pathogenesis and the epidemiology of the infection, but also to develop efficacious vaccines. Prolonged CyHV-3 cultivation in vitro leads to the spontaneous attenuation of the virus. To circumvent this problem, the entire viral genome was cloned as a bacterial artificial chromosome (BAC). Then to test the usefulness of the BAC clone, several recombinants strains were generated as described in the first chapter. In the second chapter, we took profit of the CyHV-3 BAC clone to produce a recombinant strain encoding a firefly luciferase (LUC) expression cassette. Infection of carp by immersion in water containing the CyHV-3 LUC strain demonstrated, using bioluminescent in vivo imaging system (IVIS), that the skin, and not the gills, is the major portal of entry for CyHV-3. Fish skin provides mechanical, chemical and immune protection against injury and pathogenic microorganisms. Its mucus layer confers an innate immune protection against pathogen entry. However, there is little in vivo evidence on the role of skin mucus as a first line of innate immune protection against bacterial and viral infections. In the third chapter, we used the CyHV-3 LUC strain and IVIS to investigate the roles of epidermal mucus as an innate immune barrier against CyHV-3 entry. Our results demonstrate that the mucus of the skin inhibits CyHV-3 binding to epidermal cells and contains soluble molecules able to neutralize CyHV-3 infectivity. The skin is the major portal of entry after inoculation by immersion in water containing CyHV-3. While this model of infection mimics some natural conditions in which infection takes place, other epidemiological conditions could favor entry of virus through the digestive tract. Consequently, in the fourth and last chapter, we investigated the role of the carp digestive tract as a viral portal of entry using bioluminescence imaging. We found that feeding carp with infectious materials induces CyHV-3 entry through infection of the pharyngeal periodontal mucosa. In conclusion, this study demonstrated that according to epidemiological conditions, CyHV-3 can enter carp either through infection of the skin (immersion in infectious water) or through infection of the pharyngeal periodontal mucosa (feeding on infectious materials). The existence of these two portal of entry adapted to different epidemiological conditions most probably contributes to the high contagious nature of the virus.

Fournier G, Boutier M, Stalin Raj V, Mast J, Parmentier E, Vanderwalle P, Peeters D, Lieffrig F, Farnir F, Gillet L, Vanderplasschen A. Feeding Cyprinus carpio with infectious materials mediates cyprinid herpesvirus 3 entry through infection of pharyngeal periodontal mucosa. Veterinary Research. January 2012, 43(1) : 6.

Raj VS, Fournier G, Rakus K, Ronsmans M, Ouyang P, Michel B, Delforges C, Costes B, Farnir F, Leroy B, Wattiez R, Melard C, Mast J, Lieffrig F, Vanderplasschen A. Skin mucus of Cyprinus carpio inhibits cyprinid herpesvirus 3 binding to epidermal cells. Veterinary Research. August 2011, 42(1) : 92.

Costes B, Raj VS, Michel B, Fournier G, Thirion M, Gillet L, Mast J, Lieffrig F, Bremont M, Vanderplasschen A. The major portal of entry of koi herpesvirus in Cyprinus carpio is the skin. Journal of Virology. April 2009, 83(7) : 2819-2830.

Costes B, Fournier G, Michel B, Delforge C, Raj VS, Dewals B, Gillet L, Drion P, Body A, Schynts F, Lieffrig F, Vanderplasschen A. Cloning of the koi herpesvirus genome as an infectious bacterial artificial chromosome demonstrates that disruption of the thymidine kinase locus induces partial attenuation in Cyprinus carpio koi. Journal of Virology. May 2008, 82(10) : 4955-4964.

The human gammaherpesviruses, Epstein-Barr virus (EBV) and Kaposi's Sarcoma-associated Herpesvirus (KSHV), are involved in several cancers. These viruses evolved to coexist with the host specific antibodies since they are able to induce reactivation and reexcretion in an immune carrier. Understanding precisely how they enter the target cells is crucial to improve their control and to design original vaccinal strategies. However, entry of gammaherpesviruses is a complex process that requires different cellular receptors and different viral glycoproteins. Among them, the EBV glycoprotein, gp350, binds to CD21 on B cells and is a target for antibodies that neutralize B cell infection. Bovine herpesvirus 4 (BoHV-4) is a gammaherpesvirus that shares many common properties with EBV and KSHV. Thus, the Bo10 gene potentially encodes the EBV gp350 homologous protein. In order to further explore the relationship between gp350 or its homologs, host cells and antibodies, this work focused on the function of the BoHV-4 Bo10 gene in the biology of viral infection. The results obtained are presented in four studies.

1. Bovine Herpesvirus-4 Bo10 gene encodes a non-essential viral envelope protein that regulates viral tropism through both positive and negative effects. Machiels B., Lete C., de Fays K., Mast J., Dewals B., Stevenson P.G., Vanderplasschen A. and Gillet L. Journal of Virology, (2011), 85 (2), 1011-1024.

This study demonstrated that the Bo10 gene encodes a 180 kDa product (gp180), which is incorporated in the virion envelope. This protein is not essential for viral replication but is involved in viral entry. Indeed, Bo10 deleted virus showed a growth deficit associated with a reduced binding to cellular glycosaminoglycans (GAGs). On the opposite, Bo10 deletion improves infection of GAG negative cells. These results suggest that gp180 regulates BoHV-4 tropism both positively and negatively depending upon the presence or absence of its receptor. This mechanism could be shared by the other gammaherpesviruses. We hypothesized that interaction with heparan sulfates of the cell surface displaces gp180 allowing the exposition of viral entry epitopes. A possible rationale for this arrangement would be for gp180 and its homologs to protect an otherwise vulnerable virion protein from antibody neutralization. This hypothesis has been investigated in a second study.

2. Antibody evasion by a gammaherpesvirus O-glycan shield. Machiels B., Lété C., Guillaume A., Mast J., Stevenson P.G., Vanderplasschen A. and Gillet L. Submitted for publication.

This study showed that a lack of gp180 has no impact on the establishment and maintenance of BoHV-4 latency, but markedly sensitizes virions to neutralization by immune sera. The results demonstrated that gp180 hides some different epitopes on gB, gH and gL, including epitopes targeted by neutralizing antibodies. It appeared that gp180 provides part of a glycan shield for otherwise vulnerable viral epitopes and protects them from antibody neutralization. While these proteins are diverse in sequence, the conservation of O-glycosylation sites in all gp350 homologs suggests that this is a general evasion mechanism in the gammaherpesvirinae subfamily.

3. « Natural antibody-complement dependent neutralization of bovine herpesvirus 4 by human serum ». Machiels B., Gillet L., Do Nascimento Brito S., Drion P., Delforge C., Nizet Y., Gianello P., Bona C., Costes B., Markine-Goriaynoff N. and Vanderplasschen A. Microbes and Infection, (2007), 9, 1530-1537.

The importance of human exposure to BoHV-4 and the lack of prophylactic scheme against this virus make human contamination possible. However, to date there is no reported case of human infection by BoHV-4. This observation raises the question of a natural human resistance against this virus. The results of this study highlighted an efficient neutralization of BoHV-4 by human serum, in contrast to sera of various animal species. The mechanism of this neutralization depends of the complement through activation of the classical pathway by natural antibodies raised against the cellular gal epitope. A supplemental experiment revealed that gp180 represents the main backbone for gal epitope exposition. The results of this study allowed us, on one hand, to objectivize the risk of BoHV-4 transmission to humans. On the other hand, this study illustrated a situation where glycosylation of viral proteins did not give a selective advantage to the virus but on the opposite, allowed neutralization. Therefore, glycans that represent a crucial stategy for evasion of antibody response by gammaherpesviruses, could also become the target of original vaccine approaches.

4. Alternative splicing switches tropism of a gammaherpesvirus. Machiels B., Stevenson P.G., Vanderplasschen A. and Gillet L. In preparation.

Herpesviruses have complex lifecycles that involve infection of various cell types. Interestingly, some of them are able to route infection by the use of different cellular receptors and different viral glycoprotein complexes. In this study, we showed that BoHV-4 could use alternative splicing of the Bo10 gene to orientate tropism of progeny virions. Indeed, the spliced transcript of Bo10 produces gp180 while the absence of splicing leads to the reading of a stop codon in the intron. The traduction of this unspliced transcript could generate a truncated protein without transmembrane anchor. Although only gp180 is involved in attachment to cellular GAGs, the function of this truncated protein is not still defined. As a similar gene organization is conserved in the KSHV K8.1 homologous gene, we hypothesized that a similar mechanism could be conserved in different rhadinoviruses.

In conclusion, the study of BoHV-4 Bo10 gene products highlighted the complexity of gammaherpesviruses entry mechanism. This elaborate process could be required to protect vulnerable epitopes from antibody neutralization. This work also demonstrated the importance of O-glycosylation in this mechanism and suggested that this immune evasion strategy is widespread within the gammaherpesvirinae subfamily. The results of this thesis give the first basis for fundamental and applied researches.

Machiels, B., Gillet, L., Nascimento Brito, S. D., Drion, P., Delforge, C., Nizet, Y., Gianello, P., Bona, C., Costes, B., Markine-Goriaynoff, N. & Vanderplasschen, A. (2007). Natural antibody-complement dependent neutralization of bovine herpesvirus 4 by human serum. Microbes Infect 9, 1530-1537.

Machiels, B., Lete, C., de Fays, K., Mast, J., Dewals, B., Stevenson, P. G., Vanderplasschen, A. & Gillet, L. (2011). The bovine herpesvirus 4 Bo10 gene encodes a nonessential viral envelope protein that regulates viral tropism through both positive and negative effects. Journal of virology 85, 1011-1024.

The common carp, cultivated for human consumption, is one of the most important freshwater species in aquaculture with a world production of 3 million metric tons per year. While the common carp is a cheap source of animal proteins, its coloured subspecies koi is grown for personal pleasure and competitive exhibitions. Koi carp can be sold for thousands of Euros per animal. In the 1990s, a highly contagious and dreadful disease started to cause severe economic losses in these two carp industries worldwide. The causative agent of the disease was initially called koi herpesvirus (KHV). It has been recently renamed Cyprinid herpesvirus 3 (CyHV-3) and classified in the Alloherpesviridae family of the order Herpesvirales.

Besides its economical importance, CyHV-3 represents a very interesting subject for fundamental research for the following reasons: (i) it is phylogenetically distant from the vast majority of previously studied herpesviruses; (ii) it can be studied in laboratory by infection of its natural host (homologous host-virus model); (iii) CyHV-3 has by far the largest genome of the order Herpesvirales and most of its genes are non homologous to previously reported viral sequences; (iv) despite the ancestral divergence of the Herpesviridae and the Alloherpesviridae families, and the low to non existing homology between CyHV-3 genes and those of Herpesviridae, CyHV-3 replicates in vitro as all Herpesviridae.

The recent emergence of CyHV-3 makes of this research subject a virgin field to discover. However, the development of fundamental and applied researches requires essential data that have not been produced yet for CyHV-3; for example, the protein composition of the viral particle (structural proteome) is unknown. The identification of these proteins and their coding genes was the first objective of the thesis.

The identification of the CyHV-3 structural proteome was achieved using two complementary mass spectrometry approaches. These methods led to the identification of 40 viral proteins incorporated in mature virions. Based on bioinformatic analyses, these proteins were classified as capsid (3), envelope (13) and tegument (2) proteins. Finally, 22 proteins were unclassified due to the lack of information on their putative location in the viral particle.

The identification of the CyHV-3 structural proteome revealed the presence of 4 paralogous envelope glycoproteins (pORF25, -65, -148 and -149). To our knowledge, this is the first report of a paralogous envelope protein family encoded by a member of the order Herpesvirales. This interesting discovery was at the origin of the second objective of the present thesis: the study of the ORF25 family encoded by CyHV-3. In addition to the 4 paralogous genes listed above, two additional homologous sequences were described in the CyHV-3 genome (ORF26 and -27). These sequences were initially described as being pseudogenes due to the presence of stop codons and mutations in their coding sequence. The objectives of this second study was: (i) to control the sequences available for the paralogues in the gene bank database and to analyze these sequences in silico; (ii) to determine when during the viral replication in vitro these sequences start to be transcribed; and (iii) to investigate whether these sequences, individually or all together, are essential or not for viral replication in vitro.

The identification of the proteins incorporated into CyHV-3 virions and determination of the viral genes encoding these proteins are key milestones for further fundamental and applied research on this virus.

B. Michel, G. Fournier, B. Costes and A. Vanderplasschen. Cyprinid herpesvirus 3. Emerging Infectious Diseases 16, 2010, 1835-43.

B. Michel, B. Leroy, V. Stalin Raj, F. Lieffrig, J. Mast, R. Wattiez, A. Vanderplasschen and B. Costes. The genome of cyprinid herpesvirus 3 encodes 40 proteins incorporated in mature virions. J Gen Virol 91, 2010, 452-462.

All rhadinoviruses encode an ORF73 gene. Its expression product (pORF73) is essential for viral latency but not for lytic infection. pORF73 orthologues exert their functions (i) by maintaining the viral episome through cell division, (ii) by inhibiting the lytic replication and (iii) by modulating cellular events. The structural diversity of pORF73s can be explained by variations in size and amino acid composition. Among ORF73 orthologues, Bovine herpesvirus 4 (BoHV-4) pORF73 is by far the smallest one with only a 253 aa length, corresponding to approximately 20% of largest orthologues. To date, it is unknown if this protein has conserved partially or totally the functions described for other orthologues. This thesis focused on the study of BoHV-4 pORF73.

BoHV-4 has been isolated from different organs from healthy and ill animals. The interest of studying this virus relies on the fact that it provides an "in vivo" model to study rhadinovirus biology such as HHV-8 causing lymphoproliferative diseases but also because it is a potential candidate vector in vaccinology.

The aim of this work was to study the ORF73 gene of BoHV-4 and the functions of its expression product both in vitro and in vivo. Firstly, we analysed the evolution of the ORF73 gene in the BoHV-4 species. Then, we studied its expression product. In the second study, we demonstrated that deletion of ORF73 abrogated both in vitro and in vivo latency. In the third study, we demonstrated that this protein localised in the nucleus and co-localised with mitotic chromosomes. Finally, this work showed the capability of an ORF73 orthologue to induce nucleolus disruption. The results of this work have fundamental and applied as well as vaccinal and virological implications.

Dewals B., Thirion M., Markine-Goriaynoff N., Gillet L., de Fays K., Minner F., Daix V., Sharp P., Vanderplasschen A. Evolution of Bovine herpesvirus 4: recombination and transmission between African buffalo and cattle. J Gen Virol, 87, 2006, 1509-19

Thirion M., Machiels B., Farnir F., Donofrio G., Gillet L., Dewals B. and Vanderplasschen A. Bovine herpesvirus 4 ORF 73 is dispensable for viral growth in vitro but is essential for viral persistence in vivo. J Gen Virol, 91, 2010, 2574-2584

Thirion M., Thiry M., Bureau F., Gillet L., Vanderplasschen A. and Dewals B. Bovine herpesvirus 4 ORF73 expression product co-localizes with chromosomes and induces nucleolus disruption. Submitted for publication

Thirion M., Vanderplasschen A. and Dewals B. Les rôles du cadre de lecture ouvert 73 ORF73) dans la latence des gammaherpèsvirus. Submitted for publication

An increasing number of studies demonstrate that inhibition of host complement activation is crucial for completion of the blood feeding process of hematophagous parasites. Several observations suggest that inhibition of the host complement alternative pathway by tick salivary proteins is crucial for the achievement of blood feeding and efficient transmission of the pathogens transmitted by the parasite. Strongly supporting this conclusion, a salivary protein able to inhibit the alternative pathway was cloned from the American tick Ixodes scapularis (Valenzuela et al., (2000) J. Biol. Chem. 275, 18717-18723). Interestingly, this molecule, termed Isac, has no similarity to any previously reported anti-complement molecules suggesting that it has been acquired through a mechanism of convergent evolution. In addition to this fundamental aspect, Isac is also a promising candidate antigen for the development of an anti-tick vaccine potentially able to induce the reject of the tick and/or to prevent the transmission of the pathogens.

The initial goal of the present work was to clone the orthologue of Isac from the European tick Ixodes ricinus. Interestingly, two different sequences were isolated from the transcriptome of I. ricinus salivary glands. Expression of these sequences revealed that they both encode secreted proteins able to inhibit the complement alternative pathway. These proteins were called I. ricinus anticomplement (IRAC) protein I and II. Further characterization of IRACs using monoclonal antibodies revealed that both proteins are expressed constitutively in I. ricinus salivary glands and are up-regulated during blood feeding. Analysis of a series of individual ticks revealed that all ticks tested express both IRAC I and IRAC II, demonstrating that they are the products of different genes and not of alleles of the same locus. Finally, phylogenetic analyses of the I. ricinus IRAC I and II sequences together with homologues from I. scapularis and I. pacificus demonstrates that ticks belonging to the Ixodes ricinus complex encode a new family of relatively small anti-complement molecules undergoing diversification by positive Darwinian selection.

Phylogenetic analyses of IRACs suggested that these sequences were diversifying by a process of positive Darwinian selection, possibly leading to molecules with different biological properties. In the second study, we tested the hypothesis that each paralogue may have different inhibitory activities against the complement of different natural host species, thereby contributing to broaden the host range of I. ricinus ticks. The data obtained demonstrated that this working hypothesis is correct.

Finally, we addressed the potential of IRAC I and II as candidate antigens for the development of an anti-tick vaccine. Bovine herpesvirus 4 (BoHV-4) recombinants expressing IRAC I or II were produced. Interestingly, we observed that although both recombinants expressed high levels of functional IRAC proteins in vitro, our attempts to immunize rabbits against IRACs via infection with these viruses invariably failed. In order to improve the immunogenicity of IRACs expressed as transgene, a second generation of BoHV-4 recombinants was produced. The latter expressed IRACs as transmembrane fusion proteins on cell surface. Comparison of the vaccine potential of BoHV-4 recombinant viruses expressing either secreted or transmembrane IRAC proteins revealed that while the former did not induce a detectable immune response against IRACs, the latter led to high titers of anti-IRAC antibodies. However, the immune response induced against IRACs did not lead to the reject of the tick but only slightly increased the duration of the blood feeding process.

H. Schroeder, P. Skelly, P. Zipfel, B. Losson, A. Vanderplasschen. Subversion of complement by hematophagous parasites. Developmental and Comparative Immunology 33 (2009) 5-13

H. Schroeder, V. Daix, L. Gillet, J.-C. Renauld, A. Vanderplasschen. The paralogous salivary anti complement proteins IRAC I and IRAC II encoded by Ixodes ricinus ticks have broad and complementary inhibitory activities against the complement of different host species. Microbes and Infection 9 (2) (2007) 247-250

V. Daix, H. Schroeder, N. Praet, J.-P. Georgin, I. Chiappino, L. Gillet, K. de Fays, Y. Decrem, G. Leboulle, E. Godfroid, A. Bollen, P.-P. Pastoret, L. Gern, P. M. Sharp and A.Vanderplasschen. Ixodes ticks belonging to the Ixodes ricinus complex encode a family of anti complement proteins. Insect Molecular Biology 16 (2) (2007) 155-169

L. Gillet, H. Schroeder, J. Mast, M. Thirion, J.C. Renauld, B. Dewals, A. Vanderplasschen. Anchoring tick salivary anticomplement proteins IRAC I and IRAC II to membrane increases their immunogenicity. Veterinary Research, 51, 2009, 1-13

Many gammaherpesviruses encode G-protein-coupled receptors (GPCRs). Several in vivo studies have revealed that gammaherpesvirus GPCRs are important for viral replication and for virus-induced pathogenesis. The gammaherpesvirus Alcelaphine herpesvirus 1 (AlHV-1) is carried asymptomatically by wildebeest but causes malignant catarrhal fever (MCF) following cross-species transmission to a variety of susceptible species (Plowright et al., Nature (1960) 188:1167-1169).

Experimentally, MCF can be induced in rabbits. The A5 ORF of the AlHV-1 genome encodes a putative GPCR. Despite publication of the entire AlHV-1 genome sequence 10 years ago (Ensser et al., J Virol (1997) 71 (9): 6517-6525), little is known about the role of individual AlHV-1 genes in MCF pathogenesis. The study of MCF pathogenesis has been impeded by an inability to produce recombinant viruses, mainly due to the fact that AlHV-1 becomes attenuated during successive passages in culture. In the present study we investigated whether AlHV-1 A5 encodes a functional GPCR and addressed its role in viral replication and in the pathogenesis of MCF.

In order to be able to produce AlHV-1 recombinants, the AlHV-1 genome was cloned as a bacterial artificial chromosome (BAC) stably maintained in bacteria and able to regenerate infectious and pathogenic virions when transfected into permissive cells. Functional analyses of A5 expression product demonstrate that it is a functional vGPCR able to constitutively couple to αi-type G-proteins. Using the AlHV-1 BAC clone, a strain deleted for A5 and a derived revertant strain were produced. In vitro and in vivo studies using these recombinant strains showed that the AlHV-1 A5 vGPCR is not essential for viral replication nor for the induction of MCF in rabbits.

Dewals B., Boudry C., Markine-Goriaynoff N., Desmecht D., Pastoret P. P., Vanderplasschen A. L’herpèsvirus alcélaphin 1, l’agent responsable de la forme africaine du coryza gangreneux. Ann. Méd. Vét., 2003, 147: 373-386.

Boudry C., Costes B., Fournier G., Dewals B.,Vanderplasschen A. Les homologues viraux des récepteurs cellulaires couplés aux protéines G. Virologie, accepted.

Dewals B., Boudry C., Gillet L., Markine-Goriaynoff N., De Leval L., Haig D. M., Vanderplasschen A. Cloning of the genome of Alcelaphine herpesvirus 1 as an infectious and pathogenic bacterial artificial chromosome. J Gen Virol, 2006, 87: 509-517.

Boudry C., Markine-Goriaynoff N., Delforge C., Springael J.-Y., De Leval L., Drion P, Russell G., Haig D. M., Vanderplasschen A., Dewals B. The A5 gene of Alcelaphine herpesvirus 1 encodes a constitutively active G-protein-coupled receptor that is non-essential for the induction of malignant catarrhal fever in rabbits. J Gen Virol, 2007, 88: 3224–3233.

The central role of chemokines in anti-viral defence has been highlighted by the discovery that large DNA viruses (poxviruses and herpesviruses) and at least two RNA viruses have developed strategies to modulate the host chemokine system. Viruses use three known strategies to interact with the chemokine system: virus-encoded chemokines, virus‑encoded chemokine receptors and virus-encoded secreted chemokine-binding proteins (vCKBPs). So far, four families of vCKBPs (vCKBP-1 to 4) have been described. The description of the three first families led to the conclusion that this strategy was unique to poxviruses and gammaherpesviruses. However, recently, the soluble and the membrane-anchored forms of glycoprotein G (gG) from some alphaherpesviruses infecting large herbivores was found to belong to a fourth family of vCKBPs that binds with high affinity to a broad range of CXC, CC and C-chemokines to prevent their interaction with both glycosaminoglycans (GAGs) and cellular receptors (Bryant et al., EMBO J. (2003) 22:833-846).

Glycoprotein G orthologues have been described in several alphaherpesviruses as a minor non essential glycoprotein. Based on the viral species, gG has been reported either as a structural or a non structural protein. However, gG orthologues exist under two different forms, a membrane-anchored form and a secreted form generated by proteolytic cleavage of the former.

Felid herpesvirus 1 (FeHV-1) is an alphaherpesvirus with a worldwide distribution in the cat population. It is the causative agent of feline viral rhinotracheitis, but has also been associated with other clinical disorders.

In the present study, we investigated the function of FeHV-1 gG as a vCKBP. Our results are presented as two original studies. The first study demonstrated that FeHV-1 gG, expressed as a secreted truncated protein or as a membrane-anchored protein exposed at the surface of infected cells, functions as a broad-spectrum chemokine-binding protein. In the second study, we demonstrated that FeHV-1 gG is incorporated into the viral envelope where it acts also as a chemokine-binding protein. Various FeHV-1 gG recombinant strains were produced to reach that conclusion. These recombinants were produced using prokaryotic recombination technologies. Consequently, the present study also described the BAC cloning of FeHV-1 and the development of an original approach based on BAC vectors and restriction endonuclease mediated recombination (REMR) for the production of herpesvirus recombinants.

Costes B., Thirion M., Dewals B., Mast J., Ackermann M., Markine-Goriaynoff N., Gillet L., Vanderplasschen A. Felid herpesvirus 1 glycoprotein G is a structural protein that mediates the binding of chemokines on the viral envelope. Microbes Infect. 2006, 8: 2657-67.

Costes B., Ruiz-Arguello M.B., Bryant N.A., Alcami A., Vanderplasschen A. Both soluble and membrane-anchored forms of Felid herpesvirus 1 glycoprotein G function as a broad-spectrum chemokine-binding protein. J Gen Virol. 2005, 86: 3209-14.

Alcelaphine herpesvirus 1 (AlHV-1), carried by wildebeest asymptomatically, causes malignant catarrhal fever (MCF) when cross-species transmitted to a variety of susceptible species of the Artiodactyla order. There is no available vaccine against AlHV-1. Several studies suggested that the African buffalo (Syncerus caffer) rather than cattle should be considered as the natural host species of bovine herpesvirus 4 (BoHV-4), an apathogenic gammaherpesvirus that is antigenically related to AlHV-1. Rossiter and co-workers (1989) suggested that an evolutionary relationship exists between AlHV-1 and BoHV-4. This hypothesis proposes that the serological antigenic relationship existing between BoHV-4 and AlHV-1 could confer to BoHV-4 infected buffaloes a protective immune response against lethal AlHV-1 infection. In this work, our goal was to generate information and tools that are required to test this hypothesis in the future. Our results are presented as three original studies.

1. The cloning of the AlHV-1 genome as an infectious and pathogenic bacterial artificial chromosome. The study of MCF pathogenesis has been impeded by an inability to produce recombinant viruses, due mainly to the fact that AlHV-1 becomes attenuated during passage in culture. The availability of the AlHV-1 BAC is an important advance for the study of MCF that will allow the identification of viral genes involved in MCF pathogenesis as well as the production of attenuated recombinant candidate vaccines.
2. The antibodies against bovine herpesvirus 4 are highly prevalent in wild African buffaloes throughout eastern and southern Africa. This study revealed that independently of their geographical origin, wild African buffaloes exhibit a seroprevalence of anti-BoHV-4 antibodies higher than 68 %. This result is by far above the seroprevalence generally observed in cattle. Our data are discussed in the light of our recent phylogenetic study demonstrating that the BoHV-4 Bo17 gene has been acquired from a recent ancestor of the African buffalo. Secondly, we investigated the humoral antigenic relationship existing between BoHV-4 and AlHV-1. Our results demonstrate that among the antigens expressed in AlHV-1 infected cells, epitope(s) recognized by anti-BoHV-4 antibodies are exclusively nuclear, suggesting that the putative property of BoHV-4 to confer an immune protection against AlHV-1 could rely on a cellular rather than on a humoral immune response.
3. The phylogenetic analysis of Bovine herpesvirus 4 strains isolated from cattle and from African buffalo. Bovine herpesvirus 4 (BoHV-4) is a gammaherpesvirus which has been isolated from cattle throughout the world. Our phylogenetic analyses led to the following conclusions:
1. BoHV-4 strains from African buffalo and from cattle form clades which have split approximately 700,000 years ago.
2. Since this divergence, inter-clade and intra-clade recombination events occurred at different time in the past.
3. The topology of the tree formed by zebu and taurine BoHV-4 strains is incompatible with a co-speciation process between BoHV-4 and domestic cattle implying that the latter have been contaminated through recent cross-species transmission. A scenario is proposed to explain how BoHV-4 has been transmitted from African buffalo to cattle and how the virus has reached a world wide distribution in the latter host species.

In conclusion, this work generated key information and reagents to test in a near future the hypothesis proposed by Rossiter et al. (1989). The results of this work show the complexity of the interactions existing between viruses and their hosts throughout evolution.

Dewals B., Boudry C., Markine-Goriaynoff N., Desmecht D., Pastoret P.P., Vanderplasschen A. L'herpèsvirus alcélaphin 1, l'agent responsable de la forme africaine du coryza gangreneux. Ann. Méd. Vét., 2003, 147: 373-386.

Dewals B., Gillet L., Gerdes T., Taracha E.L., Thiry E., Vanderplasschen A. Antibodies against bovine herpesvirus 4 are highly prevalent in wild African buffaloes throughout eastern and southern Africa. Vet Microbiol, 2005: 110, 209-20

Dewals B., Boudry C., Gillet L., Markine-Goriaynoff N., De Leval L., Haig D.M., Vanderplasschen A. Cloning of the genome of Alcelaphine herpesvirus 1 as an infectious and pathogenic bacterial artificial chromosome. J Gen Virol, 2006a, 87: 509-517.

Dewals B., Thirion M., Markine-Goriaynoff N., Gillet L., de Fays K., Minner F., Daix V., Sharp P., Vanderplasschen A. Evolution of Bovine herpesvirus 4: recombination and transmission between African buffalo and cattle. J Gen Virol, 2006b, 87: 1509-19.

Several observations suggest that inhibition of the host complement alternative pathway by tick salivary proteins is crucial for the achievement of blood feeding and efficient transmission of the pathogens transmitted by ticks. Indeed, inhibition of complement not only contributes to prevent tick rejection by the host, but it also protects the pathogens contains in the saliva from one of the most important component of the host innate immune response.

Recently, a salivary protein able to inhibit the alternative pathway was cloned from the American tick Ixodes scapularis (Valenzuela et al. (2000) J. Biol. Chem. 275, 18717-18723). Interestingly, this molecule, termed Isac, has no similarity to any previously reported anti-complement molecules suggesting that it has been acquired by ticks through a mechanism of convergent evolution. In addition to this fundamental aspect, Isac is also a promising candidate antigen for the development of an anti-tick vaccine able to induce the reject of the tick and/or to prevent the transmission of the pathogens.

The initial goal of this work was to clone the orthologue of Isac from the European tick Ixodes ricinus. Two different sequences were isolated from the transcriptome of I. ricinus salivary glands. Expression of these sequences revealed that they both encode secreted proteins able to inhibit the complement alternative pathway. These proteins were called I. ricinus anti-complement (IRAC) protein I and II. Further characterization of IRACs using monoclonal antibodies revealed that both proteins are expressed constitutively in I. ricinus salivary glands and are up-regulated during blood feeding. Analysis of a series of individual ticks revealed that all ticks tested express both IRAC I and IRAC II, demonstrating that they are the products of different genes and not of alleles of the same locus. Finally, phylogenetic analyses of the I. ricinus IRAC I and II sequences together with homologues from I. scapularis and I. pacificus demonstrates that ticks belonging to the Ixodes ricinus complex encode a new family of relatively small anti-complement molecules undergoing diversification by positive Darwinian selection.

The identification of the IRAC molecules and the discovery that they belong to a new family of anti-complement proteins lead to fascinating hypotheses. To address these hypotheses, efficient IRAC expression systems needed to be developed. This objective was the second goal of the present work. Two complementary expression systems were developed. They open the possibility to pursue research programs on the IRAC.

Daix V., Schroeder H., Praet N., Georgin J.P., Chiappino I., Gillet L., de Fays K., Decrem Y., Leboulle G., Godfroid E., Bollen A., Pastoret P.P., Gern L., Sharp P.M., Vanderplasschen A. Ixodes ticks belonging to the Ixodes ricinus complex encode a family of anticomplement proteins. Insect Mol Biol, 2007, 16: 155-166.

Gillet L., Daix V., Donofrio G., Wagner M., Koszinowski U. H., China B., AckermanN M., Markine-Goriaynoff N., Vanderplasschen A. Development of bovine herpesvirus 4 as an expression vector using bacterial artificial chromosome cloning. J Gen Virol, 2005, 86: 907-917.

Daix, V., Godfroid E., Pastoret P.P., Bollen A., Vanderplasschen A.. June 2001. New anti-complement proteins PCT 01870143.3.

In this work, we performed in vitro experiments to study the interactions between Bovine Herpesvirus 4 (BoHV-4) and human cells. BoHV-4, a gammaherpesvirus phylogenetically related to several human pathogens, has been isolated frequently throughout the world from cattle and several observations demonstrate the existence of a risk of BoHV-4 transmission to humans. In addition, preliminary in vitro studies have shown that some human cell lines are sensitive and/or permissive to BoHV-4 infection leading to the hypothesis that this virus could represent a risk for human health. In this work, we performed in vitro experiments to study the interactions between BoHV-4 and human cells with the primary goal to assess the risk and the consequences of human infection by BoHV-4. The results obtained are presented as three studies.

In the first study, we performed in vitro experiments to assess the risk and the consequences of human infection by BoHV-4. These experiments revealed that human cell lines from lymphoid and myeloid origins are resistant to infection, whereas epithelial cells, carcinoma and adenocarcinoma cells isolated from various organs are sensitive but poorly permissive to BoHV-4 infection. Secondly, we showed that BoHV-4 non permissive infection of some human cells protects them against apoptosis and that this infection persists through cell divisions. This study stresses the importance to consider insidious effect of non permissive infection when considering the biosafety of animal gammaherpesviruses for humans.

In the second study, we reported the potential of BoHV-4 as a new candidate for viro-onco-apoptotic treatment of some human carcinomas. This conclusion relies on the following observations: (i) BoHV-4 infection of A549 and OVCAR carcinoma cell lines led to apoptosis in a time and dose dependent manner. (ii) Apoptosis was induced by the expression of an immediate early or an early BoHV-4 gene but did not require viral replication. (iii) Cell treatment with caspase inhibitors demonstrated that apoptosis induced by BoHV-4 relied mainly on caspase 10 activation. (iv) Infection of co-cultures of A549 or OVCAR cells mixed with human 293 cells (in which BoHV-4 does not induce apoptosis) demonstrated that BoHV-4 specifically eradicated A549 or OVCAR cancer cells from the co-cultures. (v) Finally, in vivo experiments performed with nude mice showed that BoHV-4 intra-tumoral injections reduced drastically the growth of pre-established A549 xenografts. All together, these results suggest that BoHV-4 may have a potential as a viro-onco-apoptotic agent for the treatment of some human carcinomas. Further identification of BoHV-4 pro-apoptotic gene(s) and the cellular pathways targeted by this or these gene(s) could lead to the design of new cancer therapeutic molecules.

The results of these two first studies illustrate the fact that several herpesviruses encode both anti- and pro-apoptotic genes and that the resulting effect of the expression of these two types of genes is dependent on the cellular proteome. In the future, identification of the genes responsible for these phenomenon's will require the construction of several BoHV-4 recombinants. As this could have been impeded by the difficulty to manipulate the large BoHV-4 genome using classical homologous recombination in eukaryotic cells, we addressed in the third study the feasibility to exploit Bacterial Artificial Chromosome (BAC) cloning and prokaryotic recombination technologies for production of BoHV-4 recombinants. Firstly, we attempted to BAC clone BoHV-4 genome using two potential insertion sites. Secondly, we assessed the possibility to use BoHV-4 as a viral expression vector. BoHV-4 recombinants expressing Ixodes ricinus anti-complement salivary proteins (IRAC I or -II) were produced using a two-step mutagenesis procedure in E. coli. Both recombinants induced the expression of high level of functional IRAC molecules in the supernatant of infected cells.

In conclusion, the results of this work illustrate the complexity of the interactions occurring between BoHV-4 and human cells. This work demonstrates that the outcome of a viral infection relied on the interactions between two variables which are the viral and the cellular proteomes; modification of one or both of these variables leading to potentially opposite effects.

Gillet L., Minner F., Detry B., Farnir F., Willems L., Lambot M., Thiry E., Pastoret P.P., Schynts F., Vanderplasschen A. Investigation of the susceptibility of human cell lines to bovine herpesvirus 4 infection: demonstration that human cells can support a nonpermissive persistent infection which protects them against tumor necrosis factor alpha-induced apoptosis. J Virol, 2004, 78: 2336-47.

Gillet L., Vanderplasschen A. Viral subversion of the immune system. In: Application of gene-based technologies for improving animal production and health in developing countries, 2005, Harinder Makkar and Gerrit J. Viljoen, Dordrecht, Springer, 257-292.

Gillet L., Daix V., Donofrio G., Wagner M., Koszinowski U.H., China B., Ackermann M., Markine-Goriaynoff N., Vanderplasschen A. Development of Bovine Herpesvirus 4 as an expression vector using BAC cloning and prokaryotic recombination technologies. J Gen Virol, 2005, 86: 907-17.

Gillet L., Dewals B., Farnir F., De Leval L., Vanderplasschen A. Bovine herpesvirus 4 induces apoptosis of human carcinoma cell lines in vitro and in vivo. Cancer Res, 2005, 65: 9463-72.

Viral glycosyltransferases have been described in a limited number of viral species. Before the beginning of this PhD thesis, a single virus of vertebrate was know to encode a glycosyltransferase: Myxoma virus (MYXV) that encodes a α-2,3-Sialyltransferase. The goal of the present study was to investigate the Bo17 ORF of Bovine herpesvirus 4 (BoHV-4), an ORF that encodes a potential Core 2 β-1,6–N-acetylglycosaminyltransferase-M (C2GnT-M).

Bovine herpesvirus 4 (BoHV‑4) is a gammaherpesvirus that has been isolated throughout the world from healthy cattle as well as those exhibiting a variety of diseases. The present thesis led to the following discoveries.

The Bo17 gene of BoHV‑4 is the only known viral gene known to date that encodes a functional homologue of the cellular C2GnT‑M, a member of the β-1,6–Nacetylglucosaminyltransferase (β1,6GnT) gene family. β1,6GnTs are involved in the synthesis of (GlcNAcβ1->6)Gal(NAc) linkages and play crucial roles in glycan synthesis. Expression of their products changes during important biological processes such as development, immunodeficiency and oncogenesis.

Phylogenetic analysis revealed that the Bo17 gene was acquired from an ancestor of the African buffalo around 1.5 million years ago. Since the acquisition of the Bo17 gene up to very recently in evolution, recombination has occurred between divergent lineages of BoHV-4.

The pBo17 has all three enzymatic activities exhibited by cellular C2GnT‑M, i.e., core 2, core 4, and I branching activities. Both transcription of Bo17 and activities of pBo17 are conserved among strains of the BoHV-4 species. Bo17 is not essential for viral replication in vitro despite contributing to post-translational modifications of structural proteins.

Because BoHV-4 is the only known virus to date that encodes a homologue of the cellular C2GnT-M, it provides a unique model to study the biological functions of a β1,6GnT in the context of viral infection.

Vanderplasschen A., Markine-Goriaynoff N., Lomonte P., Suzuki M., Hiraoka N., Yeh J.C., Bureau F., Willems L., Thiry E., Fukuda M., Pastoret P.P. A multipotential beta-1,6-N-acetylglucosaminyl-transferase is encoded by bovine herpesvirus type 4. Proc Natl Acad Sci U S A, 2000, 97: 5756-5761.

Markine-Goriaynoff N., Minner F., de Fays K., Gillet L., Thiry E., Pastoret P.P., Vanderplasschen A. L'Herpèsvirus Bovin 4. Ann. Med. Vet., 2003a, 147: 215 – 247.

Markine-Goriaynoff N., Georgin J.P., Goltz M., Zimmermann W., Broll H., Wamwayi H.M., Pastoret P.P., Sharp P.M., Vanderplasschen A. The core 2 beta-1,6-N-acetylglucosaminyltransferase-mucin encoded by bovine herpesvirus 4 was acquired from an ancestor of the African buffalo. J Virol, 2003b, 77: 1784-1792.

Markine-Goriaynoff N., Gillet L., Karlsen O.A., Haarr L., Minner F., Pastoret P.P., Fukuda M., Vanderplasschen A. The core 2 beta-1,6-Nacetylglucosaminyltransferase-M encoded by Bovine herpesvirus 4 is not essential for viral replication despite contributing to post-translational modifications of structural proteins. Soumis pour publication dans J. Gen. Virol., 2004a, 85: 355-67.

Markine-Goriaynoff N., Gillet L., Van Etten J.L., Korres H., Verma N., Vanderplasschen A. Glycosyltransferases encoded by viruses. J Gen Virol, 2004b, 85: 2741-54.