Noroviruses (NoV), belonging to the family Caliciviridae, are a major cause of epidemic and sporadic cases of highly contagious gastroenteritis in human. Their transmission follows the fecal-oral way and they are a very important cause of human foodborne outbreaks, in particular due to the consumption of bivalve molluscs. Their genome is made of a positive single stranded RNA and they are classified by genetic analysis into 5 genogroups, each of those containing several genotypes.

NoV can not be multiply in a cell culture system. The RT-PCR became the method of choice for their detection in stool samples, food and environmental samples. It is important to have sensitive techniques that can also detect a broad range of NoV. The quantification of the viral load is possible using real-time RT-PCR and is important, not only to estimate the contamination level of a sample, but also to study and characterize the pathogenesis of a NoV infection.

NoV were detected in several animal species, among which the bovine species. These discoveries raised important questions about a zoonotic transmission and the existence of an animal reservoir. The molecular characterization of the two prototypes of bovine NoV, namely Newbury2 virus and Jena virus, revealed that they are genetically close and associated with human NoV. Among the possible hypothesis, animals could be either passive carriers of NoV, or infected in an active way by these viruses, consequently responsible of a zoonosis. The characterization of NoV circulating in human and animal species is interesting in the aim of studying their transmission routes and the possible inter-species transmission.

An important mechanism for NoV evolution is recombination, of great interest in the study of NoV, leading to modifications of the viral genome by exchange of genomic segments. This creates genetic variation and the emergence of new viruses. Indeed, it is well documented that the recombination often occurs among NoV and contributes to the genetic diversity of these viruses as well as to the appearance of new epidemic strains. The prevalence of recombinant strains of NoV could be underestimated by the fact that the characterization of NoV is usually based only on the partial sequencing of the ARN-dependent ARN-polymerase gene only. Whereas ideally various parts of the genome, mainly the ARN polymerase-ARN dependent and the capsid protein, would be sequenced to identify such viruses. It is necessary to determine precisely the role of recombination in the evolution of NoV in order to understand the evolution of the strains and advantages given to some of these strains. The study of this mechanism will allow to better understand the selective advantage observed with some NoV and will help to elucidate the transmission routes of NoV. The study of these two points (zoonotic transmission and recombinant viruses) is very important. Indeed, the cross of species barrier would affect at the same time the epidemiology and the evolution of these viruses, and would also complicate the capacity to develop a vaccine or a treatment. In other animal species, like horses or birds, no NoV was detected to date but these last years, NoV were described in several animal species (dog, lion, sheep). That lets predict a wider range of target species.

This work enters within the scope of the study of transmission routes of NoV with the objectives, after the development of fast and sensitive methods for detection and quantification, to give an explanation to important matters relating to the evolution of NoV and their host range. To begin, the conventional RT-PCR was used in order to detect NoV in human and bovine. Then, a real-time RT-PCR using SYBR Green technology was developed and uses an internal control made up of RNA added in the same tube. This test is able to detect human and bovine NoV belonging to the genogroups I, II and III and makes it possible to make the distinction between NoV and the internal control by the analysis of melting curves. A 10-fold dilution of the samples appeared the method of choice to get rid of inhibition.

In order to confirm the result directly and to allow the quantification of NoV, a real-time RT-PCR using TaqMan technology was developed. It uses an internal control of RNA and also a RNA standard. In a very interesting way, this method can detect human NoV belonging to the genogroups I, II and bovine NoV of the genogroup III. Inhibitions were effectively raised by a 10-fold dilution of the sample or the addition of bovine serum albumin in the RT-PCR mix.

These two real-time RT-PCR showed a higher sensitivity compared to the conventional RT-PCR.

With the objective of understanding the transmission routes of NoV, the situation in Belgium was studied and human and bovine NoV were detected and analyzed using partial sequencing. NoV belonging to different genogroups were detected: GI and GII in human and GIII in bovine. By analysis of the genetic proximity, bovine NoV are close to genotype GIII.2 and human NoV close to various genotypes, but mainly GII.4.

These analyses allowed the identification of two natural recombinants and a co-infection, the latter showed two different genotypes according to the area of the genome being analyzed (polymerase or capsid). The identification of a hot-spot for recombination at the junction of the ORF (open reading frame) 1 and ORF2 confirms the importance of this part of the genome for this phenomenon.

In order to study the evolution of bovine NoV, a strain detected in Belgium was sequence completely (Bo/B309/2003/BE) and compared with the original strain Newbury2 (detected in the 80s).

On one hand, this study made it possible to set up and validate tools allowing the detection and the study of human and animal NoV regarding their pathogenesis, evolution and transmission routes.

On the other hand, based on the database of samples collected during this study, the phylogenetic analysis of NoV detected and characterized goes in the direction of the studies carried out in other countries tending to show that bovine NoV make a group of distinct viruses, different from human NoV. That suggests that bovine NoV do not represent a risk for the human health. Nevertheless, the hypothesis of a zoonotic infection or an animal reservoir cannot be excluded considering the sequence proximity between human and animal NoV and also the close relationship between the human populations and the animal breedings.

The detection of a co-infection and two natural recombinants shows the evolution opportunities of NoV and the importance of a complete analysis of their genome for their characterization.

This work was the first in the field of NoV in the laboratory and has opened the way for other research subjects on NoV and new PhD thesis, especially on the study of NoV-host interaction (NoV in bovine species) and the study of recombination in vitro (murine NoV).

Scipioni, A., Mauroy, A., Vinje, J., Thiry, E. Animal noroviruses. Vet. J., 2008, 178, 32-45.

Scipioni, A., Bourgot, I., Mauroy, A., Ziant, D., Saegerman, C., Daube, G., Thiry, E. Detection and quantification of human and bovine noroviruses by a TaqMan RT-PCR assay with a control for inhibition. Mol. Cell. Probes, 2008, 22, 215-22.

Scipioni, A., Mauroy, A., Ziant, D., Saegerman, C., Thiry, E. A SYBR Green RT-PCR assay in single tube to detect human and bovine noroviruses and control for inhibition. Virol. J., 2008, 5, 94.

Scipioni, A., Mauroy, A., Mathijs, E., Ziant, D., Daube, G., Thiry, E. Detection of contemporaneous human and bovine noroviruses in Belgium. Submitted for publication.

Introduction

Herpesviruses have mainly co-evolved with their hosts for millions of years. Consequently, different related host species may have been infected by various genetically related herpesviruses. Illustrating this concept, several ruminant alphaherpesviruses have been shown to form a cluster of viruses closely related to bovine herpesvirus 1 (BoHV-1). This latter virus, prototype of this cluster, is a major pathogen in cattle associated with various clinical manifestations including infectious bovine rhinotracheitis (IBR) and infectious pustular vulvovaginitis (IPV). IBR is a disease of major economic concern in many parts of the world and especially in Europe, both in European countries where this viral infection has been eradicated and in those where the control of IBR is currently or will be undertaken. The massive use of vaccination allowed the reduction of the number of clinical IBR cases.

However, the existence of alphaherpesviruses closely related to BoHV-1 has to be taken into account in IBR control. The following viruses can be distinguished: bovine herpesvirus 5 (BoHV-5) responsible for severe meningo-encephalitis in calves, bubaline herpesvirus 1 (BuHV-1) inducing a subclinical genital infection in water buffalo (Bubalus bubalis), caprine herpesvirus 1 (CpHV-1) causing systemic disease in young kids and abortion in adult goats, cervid herpesvirus 1 (CvHV-1) responsible of severe ocular syndrome in red deer (Cervus elaphus), cervid herpesvirus 2 (CvHV-2) and elk herpesvirus 1 (ElkHV-1), which induce a subclinical genital infection respectively in reindeer (Rangifer tarandus) and elk (Cervus canadensis). Ruminant alphaherpesviruses related to BoHV-1 are antigenically and genetically closely related to each other. Their common biological properties allow heterologous infection and therefore the crossing of the species barrier. Thus, experimental cross-infection studies show that BoHV-1 is able to infect buffalo, goats, red deer, reindeer and sheep. Inversely, bovines are susceptible to BoHV-5, CpHV-1, CvHV-1 and CvHV-2.

Consequently, the main objective of the present work is dedicated to afford a better knowledge of the interaction between alphaherpesviruses and their ruminant hosts. To meet the objective, two approaches have been developed: the study of the viral diversity aiming to extend both epidemiological and virological data about ruminant alphaherpesviruses related to BoHV-1 and the study of the heterologous protection aiming to protect a minor ruminant species against infections with BoHV-1 related alphaherpesviruses.

Results

Illustrating the issues associated to the distribution of ruminant alphaherpesviruses related to BoHV-1, an original situation has been described in Belgium. In 2001 and 2002, 28.9% of red deer have been detected seropositive to BoHV-1. In absence of contact between cattle and red deer, it has been suggested that a virus related to BoHV-1 was spreading in the Belgian red deer population. Nasal and genital swabs have been collected in hunted red deer and have been inoculated onto bovine kidney cells in order to isolate the viral agent responsible of the seropositivity of Belgian red deer. A viral isolate was detected from nasal sample of a male fawn hunted in the Anlier forest. The analysis of a 443 bp sequence of glycoprotein B demonstrated that the viral isolate was related to ruminant alphaherpesviruses and especially CvHV-1. This red deer alphaherpesvirus was then characterised by comparison with the other ruminant alphaherpesviruses. The viral isolate was antigenically distinct from the other related viruses although it presumably possesses some common epitopes with CvHV-1 and CvHV-2 because a weak reaction was detected in immunofluorescence. The analysis of BamHI and BstEII restriction profiles showed ruminant alphaherpesviruses and especially the viral isolate can be distinguished by their genomic profile. A phylogenetic analysis of these related viruses has been also undertaken. Partial sequence data from UL27 encoding the glycoprotein B (gB) and US8 encoding the glycoprotein E (gE) have been sequenced. This analysis has confirmed that the viral isolate and related viruses constituted a consistent cluster within the subfamily of alphaherpesviruses.

BoHV-5 and BuHV-1 clustered together and are the most closely related to BoHV-1 1.1, BoHV-1 1.2. CpHV-1 is the most diverging ruminant alphaherpesvirus. CvHV-1 is more related to BoHV-1 than ElkHV-1 and CvHV-2 which is the most closely related to CpHV-1. The Anlier isolate clustered with CvHV-1 and the most closely related virus is ElkHV-1. The viral isolate is indeed a different CvHV-1 strain than the CvHV-1 Banffshire 82 strain. Taken together, these data report the first isolation of a ruminant alphaherpesvirus in wild fauna, which has been named CvHV-1 Anlier.

This isolation demonstrates that a ruminant can be identified as positive to BoHV-1 while it is infected with another alphaherpesvirus. In France, cattle purchased from herds so-called “IBR-free status” are sporadically identified as seropositive to BoHV-1. Among other hypotheses, the likely explanation for this “false” seropositivity would be a cross-infection with a related alphaherpesvirus. The most “plausible” virus leading to such diagnosis misinterpretation is CpHV-1 infecting goat. In this context, an epidemiologic study has been performed in order to determine if a CpHV-1 infection was present in France. The analysis of 2,548 serums in BoHV-1 gB blocking ELISA revealed that a ruminant alphaherpesvirus infection was spreading in Corse-du-Sud while no goat was detected positive in Dordogne and Vendée. Taking into account the results obtained in Dordogne and Vendée, a specificity of 100% has been measured for the BoHV-1 gB blocking ELISA. By testing serums from goats experimentally infected with CpHV-1, the sensibility has been calculated to 93.5%. The analysis in BoHV-1 gE blocking ELISA showed that 22.6% of gB-seropositive serums were also gE-seropositive. The BoHV-1 gE blocking ELISA was therefore not able to differentiate between BoHV-1 and CpHV-1 infections. However, the analysis of serums in cross seroneutralisation has strongly identified antibodies against CpHV-1 in gB-positive serums. The presence of CpHV-1 in Corse-du-Sud associated to high prevalence (61.9%) in all analysed flocks extends the number of countries infected with CpHV-1. Moreover, the difference observed between Corse-du-Sud and Dordogne and Vendée suggests that CpHV-1 is more prevalent in Mediterranean regions or countries than in Central or northern Europe.

CpHV-1 therefore appears as the BoHV-1 related virus which has the highest economic concern in Europe. Goat is a species defined as minor and the veterinary medicine industry has no commercial interest toward vaccine development protecting this species against CpHV-1. As BoHV-1 and CpHV-1 are antigenically and genetically related, a live attenuated BoHV-1 vaccine carrying a deletion in the gene encoding gE could protect goats against CpHV-1 infection. The vaccine safety has been assessed by intranasal inoculation of either BoHV-1 virulent strain or gE deleted BoHV-1 vaccine in two groups of two goats. The length of viral excretion and the peak viral titre have been decreased in the immunised group. In order to assess the vaccine efficacy, a group of goats was immunised twice two weeks a part with the vaccine while a control group was kept uninoculated. Four weeks later, immunised and control goats have been intranasally challenged with CpHV-1. A decrease of the peak viral titre was observed in immunised goats. However, the viral excretion lengthened two days more than in control group. It is concluded that this live attenuated gE-negative BoHV-1 vaccine induces a partial cross-protection against a CpHV-1 nasal infection in goats.

In regards to the results obtained in this preliminary experiment, the vaccine has been assessed against a CpHV-1 genital infection which is the natural route of infection in goats. To reach this goal, a group of goats was immunised twice three weeks a part with the gE-negative BoHV-1 vaccine followed by a subsequent CpHV-1 intravaginal challenge. To analyse the safety and the efficacy of this marker vaccine, two groups of goats served as controls: one was immunised with a virulent CpHV-1 and one was kept uninoculated until the challenge. The vaccine did not induce any undesirable local or systemic reaction and goats did not excrete gE-negative BoHV-1. After challenge, a significant reduction in disease severity was observed in immunised goats. Moreover, goats immunised with either gE-negative BoHV-1 or CpHV-1 exhibited a significant reduction in the length and the peak of viral excretion. Antibodies neutralising both BoHV-1 and CpHV-1 were raised in immunised goats. These data show that the intranasal application of a live attenuated gE-negative BoHV-1 vaccine is able to afford a clinical protection and a reduction of virus excretion in goats challenged by a CpHV-1 genital infection.

Conclusions

This study has been performed in order to afford a better knowledge of the interaction between alphaherpesviruses and their ruminant hosts. The antigenic, genetic and genomic relationships existing between BoHV-1 and ruminant alphaherpesviruses have been analysed. It has been demonstrated that these viruses constitute a strong cluster within the subfamily of alphaherpesviruses. The exact knowledge of the evolutionary relationships among these viruses request complete genome sequencing but this analysis is complicated by the large size of herpesvirus genome. Another possibility would be to compare different regions distributed along the genome of each virus as demonstrated by the study of the BoHV-4 history evolution (Dewals et al., 2006). Such analysis could bring interesting data about recombination and transmission events that occurred in the past between herpesviruses and their ruminant species. Nevertheless, the present work gives a contribution to the knowledge of the natural history of the cluster of ruminant alphaherpesviruses related to BoHV-1. Thus, the first isolation of CvHV-1 in a free-ranging red deer linked to the knowledge of the evolution of red deer populations in Europe allowed an interesting hypothesis about the origin of this virus. In regards of the observed prevalence and the presence of red deer over large area, this virus could be responsible of an endemic infection spreading in Europe. This virus may have been introduced after European red deer importation to constitute the initial stocks of Scottish farms. Consequently, CvHV-1 Anlier strain could be the ancestor of the Scottish CvHV-1 Banffshire 82 strain. However, molecular epidemiologic study of several isolates of the two strains is required to firmly assess this hypothesis.

This isolation has demonstrated that misinterpretation of the serologic status of ruminant can be observed; the Belgian red deer was not infected with BoHV-1 but with its own herpesvirus. These difficulties of interpretation are encountered in France in bovines free of IBR and detected later as positive in serology. As BoHV-5 and ElkHV-1 have not been identified in Europe, that there is a very few number of buffalo farms in France, and that CvHV-1 and CvHV-2 transmission to bovine would be exceptional, the only virus that could be responsible of misdiagnosis in France is CpHV-1. Indeed, goats and cattle can be in close contact in some field situations. The results obtained in the second study did not support this hypothesis.

Indeed, CpHV-1 infection is either absent or at a very low prevalence in continental France. Contrarily, a very high seroprevalence of CpHV-1 was measured in the Corse-du-Sud department. In Corsica, the contact between cattle and goats is very limited because flocks and herds are not mixed. Therefore, the transmission of CpHV-1 from goats to cattle would be only exceptional, such cross-infection being very rare, and a BoHV-1 infection in goats is not likely to occur. As the current sampling is restricted to Dordogne and Vendée, a particular attention to CpHV-1 has to be paid in continental France and especially in the Mediterranean region. Such monitoring can be performed with the BoHV-1 gB blocking ELISA which has been validated in the current study.

The CpHV-1 infection is therefore extended to the entire Mediterranean region and appears to be the most relevant ruminant alphaherpesvirus infection besides BoHV-1. In this context, it is necessary to protect goats to reduce the CpHV-1 circulation in the caprine population and economic concerns caused by this infection. The preliminary assessment of the live attenuated BoHV-1 vaccine carrying a deletion in the gene encoding gE has shown that goats could respond to such vaccination. As goat is completely susceptible to BoHV-1, this species could also be used as an experimental model to study the in vivo BoHV-1 properties. A partial efficacy against CpHV-1 nasal infection has been observed. Moreover, the gE-negative BoHV-1 vaccine has been assessed against CpHV-1 genital infection which is the natural route of infection in goats. The intranasal use of a gE-negative BoHV-1 vaccine enabled a clinical protection and a significant reduction of the viral excretion in goats infected genitally with CpHV-1. These results demonstrate that nasal mucosa can serve as an efficient site for the induction of a specific protective response in the genital tract. In humans, several mechanisms are proposed to explain such immune induction involving antibody secreting cells and leading to specific IgA and IgG responses.

Such mechanisms could be involved in the observed response in vaccinated goats but were not analysed in the current study. The biosafety of such cross protection induced with BoHV-1 vaccine against CpHV-1 has also to be considered. Indeed, viruses can be established in a latent state. In bovine, experiments have shown that gE-negative viruses are less reactivated and reexcreted than wild viruses. Consequently, the risk of reactivation and reexcretion in goats is therefore very low. Otherwise, the emergence of recombinant viruses between CpHV-1 and BoHV-1 after cross infection in goats has to be discussed. Among the cluster of ruminant alphaherpesviruses related to BoHV-1, only two recombinant viruses between BoHV-1 and BoHV-5 were isolated, and no recombinant between BoHV-1 and less closely related CpHV-1 was detected in vitro.

Consequently, the cross vaccination described here is likely to be completely safe. Moreover, the use of already licensed vaccine is interesting knowing that the development of new vaccines to protect minor species is of a weak interest for the pharmaceutical industry.

Thiry J., Keuser V., Muylkens B., Meurens F., Gogev S., Vanderplasschen A., Thiry E. Ruminant alphaherpesviruses related to bovine herpesvirus 1. Vet. Res., 2006, 37, 169-190.

Thiry J., Tempesta M., Camero M., Tarsitano E., Bellacicco A.L., Thiry E., Buonavoglia C. A live attenuated glycoprotein E negative bovine herpesvirus 1 vaccine induces a partial cross-protection against caprine herpesvirus 1 infection in goats. Vet. Microbiol., 2006, 113, 303-308.

Thiry J., Widén F., Grégoire F., Linden A., Belák S., Thiry E. Isolation and characterisation of a ruminant alphaherpesvirus closely related to bovine herpesvirus 1 in a free-ranging red deer. BMC Vet. Res., 2007, 3, 26.

Thiry J., Saegerman C., Chartier C., Mercier P., Keuser V., Thiry E. Serological evidence of caprine herpesvirus 1 infection in Mediterranean France. Vet. Microbiol., 2008, 128, 261-268.

Thiry J., Tempesta M., Camero M., Tarsitano E., Muylkens B., Meurens F., Thiry E., Buonavoglia C. Clinical protection against caprine herpesvirus 1 genital infection by intranasal administration of a live attenuated glycoprotein E negative bovine herpesvirus 1 vaccine. BMC Vet. Res., 2007, 3, 33.

Bovine herpesvirus 1 (BoHV-1), classified as an alphaherpesvirus, is a major pathogen of cattle. Primary infection is accompanied by various clinical manifestations such as infectious bovine rhinotracheitis (IBR), infectious pustular vulvovaginitis, abortion and systemic infection. When animal survives, a latent infection is established in sensory ganglia. A reactivation stimulus can lead to viral re-excretion. In regards to the significant losses incurred by disease and trading restrictions, several European countries have initiated BoHV-1 control programs based on the use of marker vaccines deleted in the gE gene. These marker vaccines, either inactivated or live attenuated, used together with a serological detection of gE-specific antibody (Ab), allow differentiation between infected and vaccinated animals.

The pathogenicity of BoHV-1 considers the existence of coinfection situation in cattle. Previous data supported the frequent rise of BoHV-1 recombinants in cattle after concomitant nasal infections with two BoHV-1 mutants (Schynts et al., 2003b). The situation is even more complex because of the common use of live attenuated BoHV-1 vaccines injected at the intranasal mucosa. Because the nasal mucosa is the natural portal of entry of wild-type BoHV-1, recombination between a field strain and a gE negative vaccine has been envisaged. Therefore, there is a concern about the virulence of gE negative BoHV-1 recombinants rising from coinfection between a virulent BoHV-1 and a gE negative BoHV-1 vaccine.

Intramolecular recombination is a mechanism of the genetic material exchange closely related to the alphaherpesvirus replication cycle (Thiry et al., 2005). It occurs frequently between strains of herpes simplex virus (Brown et al., 1992; Umene, 1985), varicella zoster irus (Dohner et al., 1988), feline herpesvirus 1 (Fujita et al., 1998), PrV (Glazenburg et al., 1994) and BoHV-1 (Meurens et al., 2004a; 2004b; Schynts et al., 2003b). Several studies have shown that mixed infection with two avirulent strains of the same alphaherpesvirus species can result in a synergistic increase in the severity of disease through the generation of recombinant viruses (Brandt et al., 2003; Javier et al., 1986). Presently, no data are available for the assessment of the risk encountered by gE negative recombinants. The gE negative phenotype, acquired from the vaccine strain, is supposed to guaranty the virus attenuation. However, virulence factors inherited from the wild-type BoHV-1 could enhance the virulence of such recombinants.

Three approaches were developed in this thesis. They aimed at (i) generating gE negative BoHV-1 recombinants, (ii) assessing the virulence of such recombinants, (iii) localizing the recombination that lead to their rise.

The first study was devoted to the generation and the biological characterization of gE negative BoHV-1 recombinants. We decided to produce the gE negative recombinants in vitro in order to test the effect of the introduction by recombination of the gE deletion in several BoHV-1 strains and virulence backgrounds. In each coinfection situation, parental strains were one of the seven selected wild-type BoHV-1 and one double deleted BoHV-1, carryong deletions of the gC and gE encoding genes. From plaque purified virions isolated from the coinfection supernatants, 43 gE negative BoHV-1 recombinants were detected by using a double immunofluorescence staining together with PCR characterization. In vitro growth properties were assessed by virus production, one step growth kinetics and plaque size assay. These measurements allowed making comparisons inside the set of recombinants and between the recombinants and their parents. This study demonstrated that some recombinants, in spite of their gE minus phenotype, have biological characteristics close to wild-type BoHV-1. The biological properties were scored according to results obtained in the three tests. The score was also used to identify the recombinants exhibiting biological properties close to the wild-type strains.

In the second study, virulence of gE negative BoHV-1 recombinants was investigated by the inoculation of the natural host. Calves were inoculated with recombinants selected through the biological characterization. After the primary infection and the reactivation, virulence scores as well as virus excretion induced by the gE negative BoHV-1 recombinants were compared to those induced by a highly virulent BoHV-1, by a conventional gE negative vaccine strain, and by the gC-gE negative parental strain. The tested recombinants were more virulent than the gE negative vaccine and gC-gE negative parental strain. The recombinant isolated from a BoHV-1 field strain induced the highest severe clinical score. Latency and reexcretion studies showed that three of the gE negative recombinants were reexcreted. Recombination can therefore restore virulence of gE negative BoHV-1 by introducing the gE deletion in a different virulence background.

The third study aimed at localizing recombination in the BoHV-1 genome. To this end, we compared the use of deletion markers, and a new kind of marker, single nucleotide polymorphism (SNP) discriminated by Taqman assays, to map recombination in bovine herpesvirus 1 (BoHV-1). In vitro coinfection with two BoHV-1 strains allowed a prospective investigation of the recombination dynamics by using SNPs as recombination markers. It can be concluded from the analysis of the progeny obtained in a single growth cycle of co-infected BoHV-1 strains that (i) wild-type BoHV-1 recombined at a high frequency; (ii) multiple distribution of recombination events occurred; (iii) a negative interference was observed; (iv) the long and short segments of BoHV-1 behaved differently in term of observed recombination frequencies; (v) the short segment of BoHV-1 contained most likely hot spots of recombination. These results demonstrated the consequences of recombination on the progeny genome structures obtained after in vitro simultaneous infection. The accurate tool set up in this study will allow addressing the role of recombination in herpesvirus populations in further studies both in vitro and in vivo.

From the data obtained during this PhD thesis, it can be concluded that a virulent gE negative recombinant can be generated by recombination between a live attenuated gE deleted Bo-HV-1 and a virulent BoHV-1 field strain. This result must be considered in the biosafety assessment of live attenuated herpesvirus used as vaccine both in animal and human. Otherwise, the molecular approach of the recombination supported the role of this process in herpesvirus evolution.

Muylkens B., Meurens F., Schynts F., Thiry E. Les facteurs de virulence des alphaherpèsvirus. Virologie, 2003, 7, 401-415.

Muylkens B., Meurens F., Schynts F., de Fays K., Pourchet A., Thiry J., Vanderplasschen A., Antoine N., Thiry E. Biological characterization of bovine herpesvirus 1 recombinants possessing the vaccine glycoprotein E negative phenotype. Vet. Microbiol., 2006, 113, 283-291.

Muylkens B., Meurens F., Schynts F., Farnir F., Pourchet A., Bardiau M., Gogev S., Thiry J., Cuisenaire A., Vanderplasschen A., Thiry E. Intraspecific bovine herpesvirus 1 recombinants carrying glycoprotein E deletion as a vaccine marker are virulent in cattle. J. Gen. Virol., 2006, 87, 2149-2154.

Muylkens B., Thiry J., Kirten P., Schynts F., Thiry E. Bovine herpesvirus 1 infection and infectious bovine rhinotracheitis. Vet. Res., 2007, 38, 181-209.

Recombination which has been described in numerous alphaherpesvirus species, including bovine herpesvirus 1 (BoHV-1), is though to be a major source of genetic variation in herpesviruses. Several studies about recombination assessed some factors that could influence recombination between alphaherpesviruses. An in vivo study about recombination in pseudorabies virus (PrV), specifically investigated the impact of the time interval between inoculation of the first and the second virus on recombination. Their results showed that an interval of 2 h between infections allows recombination in vivo. Among viral factors, a high degree of genome homologies is one of the most important factors influencing homologous recombination. Indeed, recombination was detected between herpesvirus simplex 1 and 2 (HSV-1, HSV-2) but has not been detected between less closely related alphaherpesviruses, such as for example HSV-1 with PrV. In this context, this work was undertaken to study the influence of two parameters on recombination of ruminant alphaherpesviruses: (i) the impact of superinfection delay, and (ii) the influence of genetic relatedness. The first study, about the impact of superinfection delay, assessed BoHV-1 recombination not only at the end of viral cycle but also sooner in the cycle, at the level of concatemeric DNA. Results show that: (i) recombination in BoHV-1 is frequent in vitro with up to 30% of recombinant viruses detected at the end of the cycle, (ii) superinfection delay progressively prevents the rise of recombinant viruses, (iii) mixed concatemers, which reveal recombination between viruses, are not detectable when interval between infection exceed 2 h and (iv) consequently BoHV-1 recombination is only possible in a short window (0 to 6 h) in which superinfection is efficient. The second study was undertaken to assess to what extent in vitro recombination can occur between members of a well-defined group of closely related viruses such as ruminant alphaherpesviruses (BoHV-1 and -5; corvine herpesviruses 1 and 2, CvHV-1 and -2; caprine herpesvirus 1, CpHV-1). Results show that: (i) recombination between same or different strain of BoHV-1 is frequent (26 and 25% respectively), (ii) interspecific recombination BoHV-1/-5 occurs and produces viruses with BoHV-1 or BoHV-5 genetic background, (iii) and interspecific recombinants with less closely related viruses were not detected. The third study was undertaken to further determine the effect of genetic relatedness. Assessment of interspecific recombination at the level of concatemeric DNA reveals that recombination is possible between BoHV-1 and -5 but also between BoHV-1 and CvHV-2 and between CvHV-1 and -2. Results of the three studies emphasize the importance of recombination and highlight its consequences in the context of the extensive use of marker vaccines with gE deleted as a tool in BoHV-1 eradication programs.

Meurens F., Schynts F., Thiry E. La recombinaison chez les alphaherpèsvirus. Ann. Méd. Vét., 2001, 145, 33-43.

Meurens F., Muylkens B., Schynts F., Bourgot I., Billiau A., Thiry E. L’interférence virale chez les Alphaherpesvirinae. Virologie, 2003, 7, 319-328.

Meurens F., Schynts F., Keil G., Muylkens B., Vanderplasschen A., Gallego P., Thiry E. Superinfection prevents recombination of the alphaherpesvirus bovine herpesvirus 1 J. Virol., 2004, 78, 3872-3879.

Meurens F., Keil G.M., Muylkens B., Gogev S., Schynts F., Negro S., Wiggers L., Thiry E. Interspecific recombination between two ruminant alphaherpesviruses, bovine herpesviruses 1 and 5. J. Virol., 2004, 78, 9828-9836.

The bovine herpesvirus 1 (BoHV-1) is a worldwide spread pathogen. The BoHV-1 is responsible for important economic losses which cause its classification in the OIE / B List of pathogens. These losses are linked with reduction of the productions caused by respiratory disorders (infectious bovine rhinotracheitis, IBR) or genital tract affections. Several European countries achieved IBR eradication programmes and became ‘Officially IBR Free’. This status allows them to regulate the import of animal productions from the countries that show a lower sanitary status. The Belgian agricultural sector also should initiate a programme for the control of IBR to keep the benefit of his usual fields for the export of cattle.

Belgium shows a high IBR prevalence, which makes impossible the quick removal of seropositive animals. A control plan should then be based on vaccination programmes.

The gE-negative marker vaccines and accompanying serological tests, that allows for the discrimination between vaccinated animals and animals which were infected by a wild type BoHV-1 strain, lead to the conception of new control plans.

The proposed vaccines decrease the intensity of clinical signs and the frequence and intensity of the episodes of reactivation and rexcretion, which, with latency, are the key events for the persistence of BoHV-1 in a herd.

To test the feasibility of an IBR control plan based on repeated vaccinations, several studies were lead in the field in the Netherlands. They concluded that the most efficient vaccination protocol, applied in dairy herds, consisted in repeated intramuscular administering of attenuated marker vaccines, given every 6 months.

To test IBR vaccination protocols in the context of non-dairy herds compared to dairy herds, the Belgian government took the decision to lead the present study to test the efficacy of repeated administering if IBR vaccines, in the specific Belgian herd management practices.

Two protocols were tested, both in dairy and in dairy/beef mixed herds.

They differed according to the primovaccinations: in the first case, 2 dosis of an attenuated vaccine were given, firstly intranasally, then intramuscularly, with an interval of 4 weeks; in the second case, 2 inactivated vaccines were given, by subcutaneous injection. All booster vaccinations were inactivated vaccines, given subcutaneously.

These protocols were applied in all cattle present in the farm, for 28 months. The seroprevalence in each herd was firstly analysed at the start of the trial by means of individual serological test of all animals in age to enter the trial. The individual serological tests were later followed by means of 5 blood samplings at the start or at the end of the housing periods (total 6 assessments, 5 periods).

The efficacy of the vaccination protocols was studied by the comparison of seroconversion incidences against gE in 3 experimental groups: two groups of hyperimmunized animals and a positive control group, in each type of herd production.

The statistical analysis was lead according two approaches: 1. a surdefencel analysis compared the probability of remaining seronegative against gE between groups, according to time; 2. incidences of seroconversion between groups were compared during the 5 periods.

The second analysis lead to the study of risk factors of seroconversion against the BoHV-1.

The complementary approaches gave similar conclusions about the vaccines efficacy.

In the dairy herds, both protocols reduced the probability of seroconversion against the BoHV-1, according to the time spent in the study, in comparison with the control group.

This information matched with the decreased incidences observed through the intervals in the vaccinated groups. The identified risk factors were: the IBR seroprevalence at the start of each interval, and the increasing age of females. Repeated vaccinations, during at least one year were identified as protection factors. No difference was shown between vaccination protocols. In the mixed dairy/beef herds, the conclusions were identical.

This study gave the opportunity to demonstrate the efficacy of vaccination programmes based on the use of booster administrations with inactivated vaccines as control tools of IBR in Belgium. Risk factors, which decrease the vaccine efficacy, were pointed out. Based on this information, a control programme could be initiated in Belgium.

Dispas M., Schynts F., Lemaire M., Letellier C., Vanopdenbosch E., Thiry E., Kerkhofs P. Isolation of a glycoprotein E-deleted bovine herpesvirus type 1 strain in the field. Vet. Rec., 2003, 153, 209-212.

Dispas M., Lemaire M., Speybroeck N., Berkvens D., Dupont A., Boelaert F., Dramaix M., Vanopdenbosch E., Kerkhofs P., Thiry E. Deux protocoles d’hyperimmunisation au moyen de vaccins marqués réduisent l’incidence de séroconversion envers l’herpèsvirus bovin 1 en cheptels laitiers : résultats d’une étude sur le terrain. Ann. Méd. Vét., 2004, 148, 47-61.

Caprine herpesvirus 1 (CpHV-1) is responsible of neonatal mortality, fertility disorders and abortion. Economic impact of this infection is important in Mediterranean countries which count many goat herds and where the prevalence of CpHV-1 infection is high. Moreover, the CpHV-1 is strongly related to bovine herpesvirus 1 (BoHV-1). Experimental cross-infections have demonstrated that these two viruses are able to cross the species barrier. Their capacity to circulate in the ruminant population is a threat for BoHV-1 control infection. Specific, sensible and efficacious diagnostic methods are needed to identify CpHV-1 infection, but also to distinguish the latter form BoHV-1 infection. To achieve this aim, several direct and indirect diagnostic methods were studied in the first part on this work.

First, ELISA method previously described for the discrimination between BoHV-1 and bovine herpesvirus 5 (BoHV-5) infections was extrapolated for the serological identification of CpHV-1 infection. The reutilisation of this method using two commercial ELISAs, usually utilized for the detection of antibodies directed against BoHV-1 glycoprotein B (gB) and gE, has shown that BoHV-1 gB blocking ELISA can be used for a first diagnostic, but did not permit the discrimination between BoHV-1 andCpHV-1 infections. This discrimination can be performed in some cases with BoHV-1 gE blocking ELISA. A new blocking ELISA using CpHV-1 as antigen and specific CpHV-1 gB monoclonal antibodies was also without having any advantage over BoHV-1 gB blocking ELISA because the discrimination between anti-CpHV-1 and anti-BoHV-1 antibodies was not possible.

Second, viral isolation and restriction enzyme analysis enabled to demonstrate the presence of CpHV-1 in Spain in two goats experimentally reactivated. Two CpHV-1 strains, closely related but different from previously described strains, were isolated from vaginal swabs from these two goats. These methods also allowed CpHV-1 detection in lung from aborted foetus upon experimentally infected goat by CpHV-1.

Third, a PCR allowed demonstrating the presence of CpHV-1 DNA in all taken foetus organs, which confirmed the generalized foetus infection as the cause of abortion.

Fourth, an immunofluorescent diagnostic test on infected cells was developed using monoclonal antibodies with the aim to unambiguously discriminate CpHV-1 from BoHV-1 but also from others related viruses (BoHV-5 and corvine herpesviruses).

The second part of this work was devoted to the CpHV-1 glycoprotein D (gD) characterization insofar as by analogy to other alphaherpesvirus gDs, this glycoprotein is expected to be a potential candidate as antigen for CpHV-1 subunit or recombinant vaccine development. In this final aim, the CpHV-1 gD gene was sequenced. CpHV-1 gD biochemical properties and its kinetic expression were analyzed and compared to those of BoHV-1 gD.

Results showed a relatively high homology between CpHV-1 and BoHV-1 gD amino acid sequences (68.8% of similarity). Moreover, six “cysteine” residues were conserved by other studied alphaherpesviruses, confirming the hypothesis of common ancestor in herpesvirus family. CpHV-1 gD has a molecular mass similar to BoHV-1 gD and contains complex N-linked oligosaccharides, but the presence of O-glycosylation was not demonstrated. Unlike the BoHV-1 gD, the CpHV-1 gD is expressed as a late protein. Results have showed that CpHV-1 shares characteristics with its alphaherpesvirus homologues but also shows divergence which could perhaps influence some CpHV-1 gD functions and probably its antigenicity when compared to other alphaherpesvirus homologues.

Keuser V., Gogev S., Schynts F., Thiry E. Demonstration of generalized infection with caprine herpesvirus 1 diagnosed in an aborted caprine fetus by PCR. Vet. Res. Commun., 2002, 26, 221-226.

Keuser V., Schynts F., Detry B., Collard A., Robert B., Vanderplasschen A., Pastoret P.-P., Thiry E. Improved antigenic methods for differential diagnosis of bovine, caprine and cervine alphaherpesviruses related to bovine herpesvirus 1. J. Clin. Microbiol., 2004, 42, 1228-1235.

Keuser V., Espejo-Serrano J., Schynts F., Georgin J.-P., Thiry E. Isolation of caprine herpesvirus 1 in Spain. Vet. Rec., 2004, 154, 395-399.

Keuser V., Detry B., Thiry J., De Fays K., Schynts F., Pastoret P.-P., Vanderplasschen A., Thiry E. Characterization of caprine herpesvirus 1 glycoprotein D gene and its translation product. Virus Res., 2006, 115, 112-121.

Infectious bovine rhinotracheitis (IBR), a well known cattle disease caused by bovine herpesvirus 1 (BoHV-1) is now eradicated in several European countries. In other European states, especially those experiencing a high seroprevalence, control programmes have been initiated using a marker vaccine deleted in the glycoprotein gE gene. An ideal BoHV-1 vaccine should protect cattle clinically and virologically (prevention of viral shedding, latency establishment and reactivation). To date, all marker BoHV-1 vaccines either attenuated or inactivated, do not efficiently protect cattle virologically, but clinically. Therefore, they need to be not only improved, but also compatible with an eradication program.

In this thesis, the vaccine ability of a recombinant replication-defective human adenovirus type 5 (HAdV-5) expressing BoHV-1 glycoproteins gC or gD, was evaluated in cattle. This vaccine vector should be able to prevent clinical signs and virus excretion against BoHV-1 challenge. In addition, this HAdV-5 based vaccine would provide an immunological marker allowing to differentiate vaccinated from naturally infected animals. Unlike the marker vaccines, adenovirus would not establish a latency following administration.

In the first study, we determined that the apparent prevalence (95% confidence interval) of serum antibodies to HAdV-5 in cattle in the Walloon region of Belgium was 5% ± 2% (p ± 1.96 [p(1-p)/n]1/2. Although the presence of HAdV-5-neutralizing antibody was detected, the results show that the prevalence is very low and that most of the bovine population is not immunized against HAdV-5. Therefore, the vaccination of cattle with recombinant HAdV-5 will not be prevented by pre-existing HAdV-5-neutralizing antibodies. These results confirmed the feasibility of using this virus as a vector in vaccination in cattle.

In the second study the efficacy of recombinant replication-defective HAdV-5 mediated BoHV-1 gD gene transfer in vitro in bovine cell lines and in vivo in cattle upper respiratory tract, was demonstrated. An efficient transfer and long-term expression of gD gene were observed in each bovine cell line, while HAdV-5-mediated gD gene transfer in cattle respiratory tract after intranasal instillation tract showed limited entry and short term expression into well differentiated airway epithelial cells. Therefore, this vector could be inappropriate for gene therapy in cattle where long-term gene expression is required, but instead, is better suited for use in vaccination for which only transient gene expression is required. In fact, in vaccination the expression of immunogenic transgene products on the cell surfaces is aimed to elicit immune response, and consequently, the infected cells are then eliminated.

In the third study, the capacity of replication-defective HAdV-5 expressing the BoHV-1 glycoprotein gC or gD to elicit an immune response and to protect cattle clinically and virologically against virulent BoHV-1 challenge, when administered intranasally in cattle, was also demonstrated. However, the challenge virus excretion was not prevented, but significantly reduced and the duration of viral shedding was shortened. In addition, the highest BoHV-1 neutralizing antibody titres were obtained with HAdV-5 expressing the BoHV-1 glycoprotein gD. In view of the obtained results, recombinant HAdV-5 may be developed as an intranasal vaccine vector in cattle administered either alone or sequentially with non-human adenovirus based vectors.

In the last study, we tested the ability of two soluble formulations, namely chitosan and glycol-chitosan, when used as an intranasal adjuvant, to improve the immunogenicity of replication-defective HAdV-5-gD. The best virological protection was obtained in calves immunized with the HAdV-5-gD adjuvanted with glycol-chitosan which decreased the challenge BoHV-1 virus excretion titres by 0.5-1.5 log when compared to those obtained in calves immunized with the vaccine vector alone. In addition, the shortest duration of viral shedding was also observed in calves immunized with HAdV-5-gD adjuvanted with glycol-chitosan. The obtained data suggest that chitosan based intranasal adjuvants could present a new generation of adjuvants associated with either conventional or recombinant viral vaccines against respiratory diseases in cattle.

Gogev S., Thiry E. Les adénovirus recombinants comme vecteurs vaccinaux. Ann. Méd. Vét., 1999, 143, 323-334.

Gogev S., Lemaire M., Thiry E. Prevalence of antibodies to human adenovirus type 5 in Belgian cattle. Vet. Rec., 2001, 148, 752-754.

Gogev S., Vanderheijden N., Lemaire M., Schynts, d’Offay J., Deprez I., Adam M., Eloit M., Thiry E. Induction of protective immunity to bovine herpesvirus type 1 in cattle by intranasal administration of replication-defective human adenovirus type 5 expressing glycoprotein gC or gD. Vaccine, 2002, 20, 1451-1465.

Gogev S., Versali M.-F., Thiry E. Les chitosanes – nouveaux adjuvants pour la vaccination par voie muqueuse chez les animaux. Ann. Méd. Vét., 2003, 147, 343-350.

Gogev S., de Fays K., Versali M.-F., Gautier S., Thiry E. Glycol chitosan improves the efficacy of intranasally administrated replication defective human adenovirus type 5 expressing glycoprotein D of bovine herpesvirus 1. Vaccine, 2004, 22, 1946-1953.

Gogev S., Georgin J.-P., Schynts F., Vanderplasschen A., Thiry E. Bovine herpesvirus 1 glycoprotein gD expression in bovine upper respiratory tract mediated by a human adenovirus type 5. Vet. Res., 2004, 35, 715-721.

Recombination is thought to be a source of genetic variation in herpesviruses because their nucleotide substitution rate is low. Moreover, the use of a live vaccine deleted in the gene encoding bovine herpesvirus 1 (BoHV-1) glycoprotein gE stress the importance to assess the risk of recombination between the deleted vaccine and wildtype strains which could lead to the appearance of gE deleted recombinants harbouring the virulence of the wildtype strain. In this context, this work was undertaken to study BoHV-1 recombination using two complementary approaches :

• the study of in vivo recombination, in the homologous animal model and following inoculation at the natural site of infection, between two distinguishable strains of BoHV-1
• the in vitro study of the inversion of viral DNA segments, which is a direct consequence of homologous recombination between inverted repeated sequences present within BoHV-1 DNA.

The study of in vivo recombination was made possible by the development of an assay (using PCR and immunofluorescence) allowing a straightforward characterization of progeny viruses generated by the coinfection between Lam gC- and Lam gE- strains, deleted in the gene encoding either gC or gE, respectively. This methodology allowed for the first time to investigate the raise of recombinant viruses, following the coinoculation of cattle by both parental strains, and to follow the evolution of recombinant viruses during both the excretion and reexcretion period. Results demonstrated that recombination is a frequent event in vivo (up to 19,84% of recombinant viruses were detected on day 6 post inoculation) and that the evolution of recombinant viruses depends on the presence or the absence of glycoprotein gE. Moreover, results demonstrate that the risk of spread of a gE- strain, generated from recombination between the gE deleted vaccine and wildtype strains, is weak.

From the structures of BoHV-1 virion and concatemeric DNA, it was demonstrated for the first time that low levels ( 5%) of encapsidated genomes have the L segment in an inverted orientation (IL genomes) while equimolar amounts of IL genomes and P genomes (genomes having the L segment in the prototype orientation) were detected in replicative forms. This result indicates that i) homologous recombination between inverted repeated sequences is frequent during BoHV-replication leading to the formation of equimolar amounts of Il and P genomes, ii) the low levels of IL genomes observed in encapsidated DNA are attributable to a restriction to cleave and package these genomes from concatemeric DNA. The analysis of terminal fragments at the ends of BoHV-1 concatemeric DNA indicated that cleavage to form IL genomes is severely impaired demonstrating that the restriction is more at the level of cleavage that packaging IL genomes.

Schynts F., Lemaire M., Baranowski E., Thiry E. La glycoprotéine gE de l’herpèsvirus bovin de type 1 et les nouveaux vaccins marqués. Ann. Méd. Vét., 1998, 142, 21-32.

Schynts F., Baranowski E., Lemaire M., Thiry E. A specific PCR to differentiate between gE negative vaccine and wildtype bovine herpesvirus type 1 strains. Vet. Microbiol., 1999, 66, 187-195.

Schynts F., Vanderplasschen A., Hanon E., Rijsewijk F.A.M., Van Oirschot J.T., Thiry E. Use of PCR and double immunofluorescence to detect bovine herpesvirus 1 recombinants. J. Virol. Methods, 2001, 92, 99-104.

Schynts F., Meurens F., Muylkens B., Epstein A.L., Mc Voy M., Thiry E. Réplication, clivage-encapsidation et recombinaison de l’ADN des herpèsvirus. Virologie, 2002, 6, 343-352.

Schynts F., McVoy M.A., Meurens F., Detry B., Esptein A.L., Thiry E. The structures of bovine herpesvirus 1 virion and concatemeric DNA: implications for cleavage and packaging of herpesvirus genomes. Virology, 2003, 314, 326-335.

Schynts F., Meurens F., Detry B., Vanderplasschen A., Thiry E. Rise and surdefencel of bovine herpesvirus 1 recombinants after primary infection and reactivation from latency. J. Virol., 2003, 77, 12535-12542.

Bovine herpesvirus type 1 (BHV-1), a member of the Alphaherpesvirinae subfamily, is associated with several clinical manifestations in cattle and particularly with a respiratory form called infectious bovine rhinotracheitis (IBR). Conventional vaccines and especially intranasal live-attenuated vaccines have efficiently contributed to control the disease. Nowadays, BHV-1 eradication programs have been initiated in several European countries or regions. In countries with high prevalence of infection, the control of IBR is associated with the vaccination of cattle with marker vaccines deleted in the gene coding for the glycoprotein E (gE) of BHV-1. One of the major problems encountered to control this infection is the maintenance of the virus in a latent state after infection with both wild-type and conventionally live-attenuated BHV-1 strains. Latently infected animals are usually identified by the detection of BHV-1-specific antibodies in their serum. However, it has been postulated that some infected animals only possess residual antibodies, if any. While they are not detected, such seronegative latent carriers (SNLC) represent a threat for cattle husbandry. The presence of specific maternal antibodies can interfere with the development of an antibody response to vaccination. BHV-1 infection of young calves under passive immunity could lead to viral latency. If this infection is not followed by an antibody response, it could generate SNLC after the disappearance of maternal antibodies.

Several studies were conducted to test this hypothesis. Different strains of BHV-1 were used: the highly virulent strain Iowa and the widely used conventional live-attenuated vaccine strain RLB 106. In the context of BHV-1 control associated with marker vaccines, it was also essential to investigate the effects of the vaccination with the live-attenuated gE-negative vaccine (Difivac) in neonatal calves. It was thereafter determined whether passively immunised gE-negative calves could produce an active antibody response to gE after infection with a field BHV-1 strain (Ciney). All the experiment were conducted in three phases: a first infection phase, a second monitoring phase (for 5 to 18 months) and a third phase where dexamethasone treatment was performed to reactivated putative latent virus. The antibody response was monitored by different serological tests. In addition, the cell-mediated immune response was assessed by an in vitro antigen-specific gamma-interferon (IFN-γ) assay.

The presence of maternally acquired antibodies in calves does neither prevent initial viral replication nor latency of a virulent BHV-1 strain (Iowa), as it was demonstrated in a first experiment. Furthermore, no antibody response could be detected following infection and the results obtained suggested that BHV-1 infection early in life could produce SNLC calves. A second experiment confirmed that the presence of passively acquired antibodies did not prevent virus excretion and establishment of a latent infection. All calves and even those which did not show any antibody response after BHV-1 infection developed a cell-mediated immune response as detected by the specific IFN-γ assay. One out of seven calves became seronegative by virus neutralisation test at 7 months of age like non infected control calves. This calf presented negative IFN-γ results at the same time and was seronegative by ELISA at around 10 months of age. In conclusion, this study demonstrated, for the first time, that BHV-1 SNLC can be experimentally obtained. In addition, the IFN-γ assay was able to discriminate calves possessing only passively acquired antibodies from those latently infected by BHV-1, but it could not detect SNLC.

We thereafter examined whether SNLC could be more easily obtained after infection with a less virulent strain: the widely used live-attenuated temperature-sensitive (ts) BHV-1 vaccine (strain RLB 106). The ts strain established acute and latent infections in all vaccinated calves either with or without passive immunity. In total, four out of 7 calves inoculated under passive immunity became clearly BHV-1 seronegative, like the seven control calves, by several ELISAs and serological reference tests. In contrast to the antibody response, the presence of a passive immunity did not hinder the cell-mediated immune response. The results obtained in this study demonstrate that SNLC can be easily obtained by vaccination with a live-attenuated BHV-1 under passive immunity, whatever the serological test used and despite of a high sensitivity.

Surprisingly, long periods of virus excretion were observed after infection in the presence of specific maternal antibodies. Several studies with human herpes simplex virus and pseudorabies virus indicated that gE plays a major role in humoral immune evasion. We have therefore investigated the consequences of the presence of a specific passive immunity on the virus shedding of the live-attenuated gE-negative BHV-1 vaccine strain (Difivac). The replication of gE-negative strain was considerably and significantly reduces in the maternally immunised calves, in comparison with the non- immune calves. On the other hand, the excretion of a gE-positive conventional vaccine strain (RLB 106) was not reduced and even seemed to be prolonged in the presence of maternal antibodies. These results suggest that BHV-1 gE may also play a role in virus surdefencel in the presence of antibodies.

The effects of the vaccination with the gE-negative BHV-1 vaccine in neonatal seronegative and passively immunised calves on immune response and virus latency were than examined. Like the non-inoculated control calves, all passively immunised inoculated calves became seronegative to BHV-1 and surprisingly remained so post-dexamethasone treatment (PDT). However, all calves which excreted the virus (seven of 10 passively immunised and all six naïve calves) developed a cell-mediated immune response and a booster response was observed PDT. The vaccine virus was recovered PDT from the nasal secretions in 2 calves and BHV-1 DNA was detected by PCR in the trigeminal ganglia from 5 calves. Maternally derived antibodies may interfere with the vaccination. But above all, the obtained results demonstrate that the BHV-1 gE-negative strain can establish latency and can be reactivated and re-excreted in the nasal secretions several months after only one intranasal inoculation. Moreover, BHV-1 DNA sequences were found in the trigeminal ganglia of completely BHV-1 seronegative animals and this feature could also lead to a safety problem in BHV-1 eradication programmes.

Finally, we showed that calves possessing passive immunity from cows vaccinated two or three times with gE-negative inactivated vaccine can develop an active and lasting antibody response to gE after an infection with a field BHV-1 strain. The results demonstrated that under the conditions described, the gE-negative marker also makes it possible to distinguish between passively immunised and latently infected calves.

In conclusion, this work demonstrated for the first time that BHV-1 SNLC can exist and be experimentally obtained by infection under passive immunity, even with a highly virulent strain. A strain effect was evidenced: one SNLC out of 7 calves was obtained with the virulent strain and four with the conventional attenuated BHV-1 vaccine. With the gE-negative vaccine, the seven calves which excreted the vaccine virus became seronegative to BHV-1. The IFN-γ assay appears to be a good complementary test to the serological methods, at least in the acute phase of infection. However this test was not able to detect the SNLC. The failure to easily detect such animals represents a threat for BHV-1 free herds, selection stations, and artificial insemination centres. The vaccination with a live-attenuated BHV-1 conventional vaccine could represent a good model to experimentally produce SNLC in aim to improve the serological diagnostic tools or to develop new approaches in the detection of latency. In addition, this study demonstrated, for the first time, that the BHV-1 gE-negative strain can establish latency not only in seronegative but also in passively immunised calves after only one intranasal inoculation. Moreover, the results obtained indicated that natural BHV-1 infection of calves possessing maternally derived specific antibodies could also represent a good model for the investigation of the immune evasive character of alphaherpesviruses.

Lemaire M., Pastoret P.-P., Thiry E. Le contrôle de l’infection par le virus de la rhinotrachéite infectieuse bovine. Ann. Méd. Vét., 1994, 138, 167-180.

Lemaire M., Meyer G., Ernst E., Vanherrewhege V., Limbourg B., Pastoret P.-P., Thiry E. Latent bovine herpesvirus 1 infection in calves protected by colostral immunity. Vet. Rec., 1995, 137, 70-71.

Lemaire M., Schynts F., Meyer G., Thiry E. Antibody response to glycoprotein E after bovine herpesvirus type 1 infection in passively immunised, glycoprotein E-negative calves. Vet. Rec., 1999, 144, 172-176.

Lemaire M., Weynants V., Godfroid J., Schynts F., Meyer G., Letesson J.-J., Thiry E. Effects of bovine herpesvirus type 1 infection in calves with maternal antibodies on immune response and virus latency. J. Clin. Microbiol., 2000, 38, 1885-1894.

Lemaire M., Meyer G., Baranowski E., Schynts F., Wellemans G., Kerkhofs P., Thiry E. Production of bovine herpesvirus type 1 seronegative latent carriers by administration of a live-attenuated vaccine in passively immunized calves. J. Clin. Microbiol., 2000, 38, 4233-4238.

Lemaire M., Hanon E., Schynts F., Meyer G., Thiry E. Specific passive immunity reduces the excretion of glycoprotein E-negative bovine herpesvirus type 1 vaccine strain in calves. Vaccine, 2001, 19, 1013-1017.

Lemaire M., Schynts F., Meyer G., Georgin J.-P., Baranowski E., Gabriel A., Ros C., Belak S., Thiry E. Latency and reactivation of a glycoprotein E negative bovine herpesvirus type 1 vaccine: influence of virus load and effect of specific maternal antibodies. Vaccine, 2001, 19, 4795-4804.

Bovine herpesvirus type 5 (BHV-5), a member of the Alphaherpesvirinae subfamily, is the causative agent of a fatal meningo-encephalitis in calves. This virus is closely related to bovine herpesvirus type 1 (BHV-1) which primarily causes infectious bovine rhinotracheitis (IBR) and a diverse range of clinical manifestations in cattle. However, these viruses differ markedly in their ability to cause neurological disease in cattle. BHV-5 and BHV-1 produce rhinitis and conjunctivitis when applied to these mucosa but only BHV-5 is able to invade the central nervous system (CNS) and replicate, leading to a fatal meningo-encephalitis. This thesis is included in a research program on the characterisation of genes involved in BHV-5 neuropathogenicity. It is specifically devoted to the comparative neuropathogenicity of BHV-5 and BHV-1 in calves, to the establishment of a rabbit model for BHV-5 neurological acute infection, to the partial characterisation of genes involved in BHV-5 neuropathogenicity and to the characterisation of the expression products of genes encoding glycoprotein gH.

Neuropathogenicity of BHV-1 and BHV-5 was investigated in two independent studies. Firstly, comparative pathogenicity of BHV-1 and BHV-1 was investigated in young susceptible calves for the acute and latent infections. The results showed that, during primo-infection, BHV-5 and BHV-1 diffre markedly in their ability to cause respiratory and neurological diseases although both viruses similarly replicate in the nasal mucosa. Virus isolation and immunohistochemistry demonstrated that BHV-5 N569 is a highly neurovirulent strain which replicates extensively in the CNS but which is also able to invade respiratory and kidney tissues. The route by which the brain of calves is infected is presumably through the trigeminal and olfactory nerves. In addition, excretion profiles after reactivation clearly demonstrated for the first time that BHV-5 is able to establish latency in susceptible calves. The BHV-5 latency sites involved trigeminal ganglia but also different parts of the brain, in contrast to BHV-1 where latency was only detected in the trigeminal ganglia. The second experiment of this study was conducted to evaluate the suitability of the rabbit as a model for BHV-5 and BHV-1 acute and latent infections. The results showed that intranasal inoculation of BHV-5 and BHV-1 in rabbits was followed by the development of a lethal meningo-encephalitis for 66% of BHV-5 inoculated rabbits while all BHV-1 infected rabbits remained healthy throughout this experiment. The presence of BHV-5 in the CNS was confirmed by virus isolation and by immunohistochemical staining of BHV-5 antigen only in BHV-5 infected rabbits showing clinical signs of meningo-encephalitis. All these findings are similar to those observed after experimental infections of calves and confirmed the suitability of a rabbit model for the establishment of BHV-5 neurological acute infection and also as a valuable tool for the comparative study of BHV-5 and BHV-1 neuropathogenicity. BHV-5 is also able to extablish latency in rabbits and to be reactivated by corticoid treatment. In addition, the neuropathogenicity of BHV-5 in rabbits is partly regulated by viral genes which are implicated in brain virus replication.

Preliminaries experiments were also performed to determine BHV-5 genes involved in neuropathogenicity. Firstly, three BHV-1 intertypic recombinant viruses which contain specific BHV-5 genomic regions were constructed and then tested in vivo for meuropathogenicity in the rabbit model. Secondly, a neuropathogenic bovine herpesvirus mutant was isolated from the CNS of rabbits inoculated with viruses resulting from cotransfection of MDBK cells with entire BHV-1 DNA and digested BHV-5 DNA. Preliminary results suggested that one or several genes located in the 38.5kbs of the BHV-5 right end genome could play a role in BHV-5 neuropathogenicity, and more specifically a 2.4kb fragment located between the internal inverted repeat region and the short unit of the BHV-5 virus genome

The last part of this work was devoted to the in vitro characterisation of the BHV-1 and BHV-5 expression products of genes encoding glycoprotein gH. The gH glycoprotein was selected in relation to the putative function of bovine herpesvirus gH in CNS invasiveness. I a first experiment a gH null mutant was constructed in which gH coding sequences were deleted and replaced by the Escherichia coli lacZ cassette. Experiments with the BHV-1 gH null mutant showed that gH of BHV-1 is essential for the infectious virus cycle and is specifically involved in virus penetration and cell-to-cell spread. In the second experiment, the gene encoding BHV-5 glycoprotein gH was localised and sequenced. Synthesis and intracellular processing of BHV-5 gH was analysed in infected MDBK cells using gH cross-reacting monoclonal antibodies (MAbs). The results showed that BHV-5 gH is expressed as a beta-gamma protein detected by radioimmunoprecipitation as early as 3 hours post inoculation and is processed in the Golgi apparatus of MDBK cells by N-glycosylation from a high-mannose precursor to a mature form containing complex-type oligosaccharides. Glycosylation studies indicated that N-linked carbohydrates of BHV-5 gH are essential for the recognition of the protein by the MAbs suggesting that N-linked glycans are involved in protein folding or are targets for the gH cross-reacting MAbs. Plaque-reduction neutralisation assays showed that at least one BHV-1 gH antigenic domain is lacking in BHV-5. The antigenic structure differences between BHV-5 and BHV-1 gH, as well as sequence data comparisons which essentially focused divergences for one aa domain (150-300) in the BHV-5 gH open reading frame, may possible relate to in vivo viral tropism differences.

Meyer G., Lemaire M., Lyaku J., Pastoret P.-P., Thiry E. Establishment of a rabbit model for bovine herpesvirus type 5 neurological acute infection. Vet. Microbiol., 1996, 51, 27-40.

Meyer G., Vlcek C., Paces V., O’Hara M.K., Pastoret P.-P., Thiry E., Schwyzer M. Sequence analysis of the bovine herpesvirus type 1 genes homologous to the DNA polymerase (UL30), the major DNA-binding protein (UL29) and the ICP 18.5 assembly protein (UL28) genes of herpes simplex virus. Arch. Virol., 1997, 142, 89-102.

Meyer G. Hanon E., Georlette D., Pastoret P.-P., Thiry E. Glycoprotein gH of bovine herpesvirus type 1 (BHV-1) is essential for penetration and propagation in cell culture. J. Gen. Virol., 1998, 79, 1983-1987.

Meyer G., Bare O., Thiry E. Identification and characterization of bovine herpesvirus type 5 glycoprotein gH gene and gene products. J. gen. Virol., 1999, 80, 2849-2859.

Meyer G., Lemaire M., Ros C., Belak K., Gabriel A., Cassart D., Coignoul F., Belak S., Thiry E. Comparative pathogenesis of acute and latent infections of calves with bovine herpesvirus types 1 and 5. Arch. Virol., 2001, 146, 633-652.

Lomonte P., Bublot M., Van Santen V., Keil G., Pastoret P.-P., Thiry E. Analysis of bovine herpesvirus 4 genomic regions located outside the conserved gammaherpesvirus gene blocks. J. Gen. Virol., 1995, 76, 1835-1841.

Lomonte P., Bublot M., van Santen V., Keil G., Pastoret P.-P., Thiry E. Bovine herpesvirus 4 : genomic organization and relationship with two other gammaherpesviruses, Epstein-Barr virus and herpesvirus saimiri. Vet. Microbiol., 1996, 53, 79-89.

Lomonte P., Filée P., Lyaku J.R., Bublot M., Pastoret P.-P., Thiry E. Glycoprotein B of bovine herpesvirus 4 is a major component of the virion, unlike that of two other gammaherpesviruses, Epstein-Barr virus and murine gammaherpesvirus 68. J. Virol., 1997, 71, 3332-3335.

Lomonte P., Filée P., Lyaku J.R., Bublot M., Pastoret P.-P., Thiry E. Analysis of the biochemical properties of, and complex formation between glycoproteins H and L of the gamma-2 herpesvirus bovine herpesvirus-4. J. Gen. Virol., 1997, 78, 2015-2023.

Baranowski E., Dubuisson J., Pastoret P.-P., Thiry E. Identification of 108K, 93K and 42K glycoproteins of bovine herpesvirus-1 by monoclonal antibodies. Arch. Virol., 1993, 133, 97-111.

Baranowski E., Dubuisson J., van Drunen Littel-van den Hurk S., Babiuk L.A., Michel A., Pastoret P.-P., Thiry E. Synthesis and processing of bovine herpesvirus-1 glycoprotein H.Virology, 1995, 206, 651-654.

Baranowski E., Keil G., Lyaku J., Rijsewijk F., Van Oirschot J., Pastoret P.-P., Thiry E. Structural and functional analysis of bovine herpesvirus 1 minor glycoproteins. Vet. Microbiol., 1996, 53, 91-101.