In vitro and in vivo characterization of equine herpesvirus type 1 (EHV-1) mutants devoid of the viral chemokine-binding glycoprotein G (gG)

Glycoprotein G (gG) of equine herpesvirus type 1 (EHV-1), a structural component of virions and secreted from virus-infected cells, was shown to bind to a variety of different chemokines and as such might be involved in immune modulation. Little is known, however, about its role in the replication cycle and infection of EHV-1 in vivo. Here we report on the function of gG in context of virus infection in vitro and in vivo. A gG deletion mutant of pathogenic EHV-1 strain RacL11 (vL11DeltagG) was constructed and analyzed. Deletion of gG had virtually no effect on the growth properties of vL11DeltagG in cell culture when compared to parental virus or a rescuant virus vL11DeltagGR, respectively, and virus titers and plaque formation were unaffected in the absence of the glycoprotein. Similarly, in the murine model of EHV-1 infection, no significant differences in virulence between the gG deletion mutant and RacL11 or vL11DeltagGR were found at high doses of infection. However, infection of mice at lower doses revealed that the gG deletion mutant was able to replicate to higher titers in lungs of infected mice. Additionally, these mice lost significantly more weight than those infected with RacL11 and a more pronounced inflammatory response in lungs was observed. Therefore we concluded that deletion of gG in EHV-1 seems to lead to an exacerbation of respiratory disease in the mouse.     

Identification of another B-cell epitope in the type-specific region of equine herpesvirus 4 glycoprotein G

Recently, a novel 12-mer B-cell epitope, MKNNPIYSEGSL, in the type-specific region of equine herpesvirus 1 (EHV-1) glycoprotein G (gG) was identified and used as an antigen for enzyme-linked immunosorbent assay (Maeda et al., J. Clin. Microbiol. 42:1095-1098, 2004). Although our prototype strain, TH20p, possesses two repeat sequences containing the B-cell epitope, the EHV-4 NS80567 strain has two repeat sequences that are not identical. One repeat sequence stretch contained the B-cell epitope, while the other contained the 11-mer, MKNNPVYSESL (underlining indicates a different amino acid). In this study, heterogeneity of the type-specific region was compared among Japanese EHV-4 isolates. The 11-mer peptide, MKNNPVYSESL, specifically reacted with sera from horses naturally infected with EHV-4 but not with sera from horses experimentally infected with EHV-4 TH20p. The 11-mer peptide may be another B-cell epitope in the type-specific region.     

Genetic relatedness and pathogenicity of equine herpesvirus 1 isolated from onager,zebra and gazelle

Summary. Equine herpesvirus 1 was isolated from an onager in 1985, a zebra in 1986 and a Thomson’s gazelle in 1996 in USA. The genetic relatedness and pathogenicity of these three viruses were investigated based on the nucleotide sequences of the glycoprotein G (gG) gene, experimental infection in hamsters, and comparison with horse isolates. The gG gene sequences of EHV-1 from onager and zebra were identical. The gG gene sequences of the gazelle isolate showed 99.5% identity to those of onager and zebra isolates. The gG gene sequences of EHV-1 isolated from horses were 99.9–100% identical and 98, 98 and 97.8% similar to gG from onager, zebra and gazelle isolates, respectively. Hamsters inoculated with onager, zebra and gazelle isolates had severe weight loss, compared with hamsters inoculated with horse isolates. The histopathological findings were related to the virulence of each isolate. The results indicated that EHV-1 isolates from onager, zebra and gazelle differ from horse EHV-1 and are much more virulent in hamsters.     

The equine herpesvirus type 1 (EHV-1) homolog of herpes simplex virus type 1 US9 and the nature of a major deletion within the unique short segment of the EHV-1 KyA strain genome.

The DNA sequence of the short (S) genomic component of the equine herpesvirus type 1 (EHV-1)KyA strain has been determined recently in our laboratory. Analysis of a 1353-bp BamHI/PvuII clone mapping at the unique short/terminal inverted repeat (Us/TR) junction revealed 507 bp of Us and 846 bp of TR sequences as well as an open reading frame (ORF) that is contained entirely within the Us. This ORF encodes a potential polypeptide of 219 amino acids that shows significant homology to the US9 proteins of herpes simplex virus type 1 (HSV-1), EHV-4, pseudorabies virus (PRV), and varicella zoster virus (VZV). The US9 polypeptides of the two equine herpesviruses exhibit 50% identity but are twice as large as their counterparts in HSV-1, PRV, and VZV. All five US9 proteins are enriched for serine and threonine residues and share a conserved domain of highly basic residues followed by a region of nonpolar amino acids. DNA sequence and Southern blot hybridization analyses revealed that the Us of EHV-1 KyA differs from the Us of EHV-1 KyD and AB1 in that the ORFs encoding glycoproteins I and E and a unique 10-kDa polypeptide are deleted from the KyA genome. These data demonstrate that the predicted 10-kDa protein unique to EHV-1 is nonessential for replication in vitro and that EHV-1 glycoproteins I and E, like their equivalents in HSV-1 and PRV, are also nonessential. These findings and those reported previously by this laboratory and others reveal that the Us segment of EHV-1 comprises nine ORFs, two of which, US4 and 10-kDa ORF, are unique to EHV-1. The gene order of the Us is US2, protein kinase, gG, US4, gD, gI, gE, 10 kDa, and US9.     

The equine herpesvirus 1 UL45 homolog encodes a glycosylated type II transmembrane protein and is involved in virus egress

Experiments to analyze the product of the equine herpesvirus type 1 (EHV-1) UL45 homolog were conducted. Using an antiserum generated against the carboxylterminal 114 amino acids of the EHV-1 UL45 protein, proteins of M(r) 32,000, 40,000, and 43,000 were detected specifically in EHV-1-infected cells. Neither form of the protein was located in purified virions of EHV-1 wild-type strain RacL22 or the modified live vaccine strain RacH, but UL45 was demonstrated to be expressed as a late (gamma-2) protein. Fractionation of infected cells and deglycosylation experiments demonstrated that the EHV-1 UL45 protein represents a type II membrane glycoprotein. Deletion of the UL45 gene in RacL22 and RacH (LDelta45 and HDelta45) showed that UL45 is nonessential for EHV-1 growth in vitro, but that deletion reduced the viruses' replication efficiency. A marked reduction of virus release was observed although no significant influence was noticed either on plaque size or on the syncytial phenotype of the EHV-1 strain RacH.     

Glycoprotein G deletion mutants of equine herpesvirus 1 (EHV1; equine abortion virus) and EHV4 (equine rhinopneumonitis virus)

Summary. Glycoprotein G (gG) deletion mutants of EHV1 and EHV4, designated EHV1ΔgG and EHV4ΔgG, were constructed. The growth characteristics of the EHV1ΔgG mutants were similar to the parent virus. All of the EHV4ΔgG mutants grew more slowly in cell culture and produced plaques of different morphology including smaller size. The yields of both gG deletion mutant viruses in cell culture were similar to the parent viruses. Sequencing of the genes flanking gG, Southern blot, PCR and western blot analyses of the mutant viruses demonstrated that the deletions were as expected, except for EHV4ΔgG mutants, which in addition to deletion of gG contained unexpected deletions in the adjacent down stream gene ORF 71 (glycoprotein 2). Antisera to EHV1ΔgG and EHV4ΔgG neutralised the respective mutant and the parent viruses to the same titre and these antisera could be distinguished from antisera to the wild type viruses in a gG antibody detection ELISA. The mutant viruses may be useful as vaccine candidates and the deletion of gG may act as a marker to distinguish vaccinated from the naturally infected horses. Present address: Veterinary Medicine R & D, Pfizer Animal Health, 45 Poplar Rd, Parkville 3052, Victoria, Australia.     

Contribution of gene products encoded within the unique short segment of equine herpesvirus 1 to virulence in a murine model

The pathogenesis of three equine herpesvirus 1 (EHV-1) recombinants was assessed in a CBA mouse model. Sequences encoding the majority of glycoproteins I (gI) and E (gE) were deleted from the pathogenic EHV-1 strain RacL11 (L11ΔgIΔgE), and sequences comprising the 3859 bp deletion within the strain KyA U_S segment, which includes genes 73 (gI), 74 (gE), and 75 (putative 10 kDa protein 75), were re-inserted into attenuated KyA (KgI/gE/75). In addition, genes gE and 75 were inserted into KyA to generate the EHV-1 recombinant KgE/75. The insertion of the 3859 bp U_S segment was sufficient to confer virulence to KyA, as indicated by pronounced signs of clinical disease including substantial weight loss. A large plaque morphology was observed in cells infected with KgI/gE/75 compared with KyA, and a small plaque phenotype was observed in cells infected with L11ΔgIΔgE compared with RacL11. These data indicate that gI and/or gI and gE contribute to the ability of EHV-1 to spread directly from cell-to-cell. The deletion of both gI and gE from the pathogenic RacL11 strain did not reduce clinical signs of disease in infected mice, but did decrease mortality compared with RacL11. Furthermore, the insertion of genes 74 (gE) and 75 into the vaccine strain KyA did not alter the attenuated phenotype of this virus. Finally, KgI/gE/75 and RacL11 elicited the production of the proinflammatory chemokines MIP-1, MIP-1β, and MIP-2 in the lungs of infected mice, while KyA did not, suggesting that gI and/or gI and gE contribute to the up-regulation of these mediators of inflammation. These findings show that gI, and/or gI and gE restore a virulent phenotype to the EHV-1 KyA strain, and indicate that virulence factors, in addition to gI and gE, contribute to the pathogenesis of the RacL11 strain.     

A deletion in the gI and gE genes of equine herpesvirus type 4 reduces viral virulence in the natural host and affects virus transmission during cell-to-cell spread

In order to identify the role of the equine herpesvirus type 4 (EHV-4) glycoprotein I (gI) and E (gE) genes in determining viral virulence and their affect on the infection cycle, we constructed an EHV-4 recombinant strain containing a deletion in both gI and gE genes and its revertant. The recombinant was assayed in vitro in order to compare its growth kinetics with the parent and revertant viruses. Our results indicated that a deletion in the genes encoding gI and gE affected cell-to-cell spread of the virus in vitro. In order to assess the pathogenicity and vaccine efficacy of the recombinant in a natural host, colostrum-deprived foals were inoculated intranasally with the recombinant. Clinical signs obtained in foals upon the inoculation with the recombinant were milder than that for the revertant. This suggests that intact gI and/or gE genes are important factors in the expression of virulence in EHV-4 as in seen in the case of other herpesviruses. In addition, full protection against challenge infection was observed in foals, which had undergone a previous inoculation of the recombinant.     

Equine herpesvirus type 1 (EHV-1) induced abortions and paralysis in a Lipizzaner stud: a contribution to the classification of equine herpesviruses

Summary Out of 30 cases of abortion and perinatal deaths in a Lipizzaner stud in Austria 10 mares died after having shown central nervous system disturbances, ataxias and paralysis. The etiological agent of this abortion storm was equine herpesvirus type 1 (EHV-1). The restriction enzyme pattern of the DNA from 5 isolates recovered from fetuses has been analyzed and compared with the known reference strains of EHV-1, -2, -4 and an Austrian vaccine strain. The DNA restriction profiles of the Lipizzaner isolates as well as of the vaccine strain could be identified as being typical of abortigenic strains with minor variations. Such variations on the molecular biological level of the DNA do not justify characterization of the strains as neurovariants. The vaccine strain differed from other isolates investigated with 4 restriction endonucleases (Bam HI, Bgl II, Eco R I, Kpn I) which was due to a deletion in the unique short segment of the genome. The lack of similar DNA bands in two EHV-1 viruses, causing mild respiratory disease, as well as in the vaccine strain Prevaccinol is suggestive of lowered virulence. In contrast to one Lipizzaner isolate tested (strain Austria IV) the Austrian vaccine strain proved to be of strong neurovirulence for suckling mice.With 6 Figures     

Cloning of the genome of equine herpesvirus 4 strain TH20p as an infectious bacterial artificial chromosome

Equine herpesvirus 4 (EHV-4) is a major cause of respiratory tract disease in horses worldwide. The generation of recombinant viruses, which would lead to understanding of viral gene functions, has been hindered by the absence of suitable cell lines and small-animal models of the infection. In the present study, the genome of EHV-4 strain TH20p was cloned as a stable and infectious BAC without any deletions of the viral genes. Mini F plasmid sequences flanked by loxP sites were inserted into the intergenic region between genes 58 and 59. Coinfection of the recombinant virus with a recombinant adenovirus expressing Cre recombinase resulted in the excision of the BAC sequences. Importantly, the resulting recombinant EHV-4 replicated comparably to the wild-type virus in fetal horse kidney cells. The recombinant EHV-4 will facilitate EHV-4 research and provide the opportunity to exploit the power of BAC technology for production of recombinant viral vaccines.     

Modulation of the serological response of specific pathogen-free (EHV-free) foals to EHV-1 by previous infection with EHV-4 or a TK-deletion mutant of EHV-1

Summary EHV-1 was inoculated into specific pathogen-free (SPF) foals in order to study uncomplicated primary responses. Infection resulted in a strong serological response recognizing EHV-1-specific antigens; this contrasts with a previous publication where a weak response was recorded in SPF animals. Antibodies to EHV-1 were readily detected by four techniques (virus neutralization, complement fixation, Western blots and immune precipitation), yet there was comparatively little cross-reaction to EHV-4 target antigen. Re-in-oculation with the same virus strain stimulated antibodies to EHV-1 but no additional antigens were recognized and antibodies cross-reacting with EHV-4 antigens were not enhanced. Having characterized the uncomplicated primary response to EHV-1 in SPF foals, further animals were exposed to either EHV-4 or a thymidine kinase-deficient mutant of EHV-1 prior to challenge with w/t EHV-1 to investigate how these infections might modulate the immune responses to EHV-1 or 4. Primary inoculation with EHV-4 or with a thymidine kinase-deficient mutant of EHV-1 produced productive infections as evidenced by virus shedding and pyrexia. In both these cases, however, in contrast to that with w/t EHV-1, the serological response was very weak. Re-infection of foals primed with either EHV-4 or TK-deficient EHV-1 with w/t EHV-1 resulted in a strong response to EHV-1 antigens detected by all four methods. In addition, in the foals given a primary inoculation with EHV-4, superinfection with EHV-1 resulted in a strong cross-reactive response to EHV-4 target antigens. The relevance of these observations to the interpretation of previously reported serological responses to EHVs in SPF and naturally reared animals is discussed.     

Quantitation of virus-specific classes of antibodies following immunization of mice with attenuated equine herpesvirus 1 and viral glycoprotein D

The antibody responses of CBA/J mice infected intranasally (i.n.) with either the attenuated KyA strain or the pathogenic RacL11 strain of equine herpesvirus 1 (EHV-1) or immunized with recombinant glycoprotein D (rgD) were investigated using the ELISPOT assay to measure EHV-1-specific antibody-secreting cells (ASC) in the regional lymphoid tissue of the respiratory tract. IgG, IgA, and IgM ASC specific for EHV-1 were detected in the mediastinal lymph nodes (MLN) and lungs 2 weeks after i.n. infection with EHV-1 strain KyA or RacL11, or immunization with heat-killed KyA or rgD. EHV-1-specific ASC were present in the MLN and lungs at 4 and 8 weeks, but declined in frequency by fivefold in the lung at 8 weeks. However, i.n. immunized (2 x 10(6) pfu KyA or 50 microgram rgD/mouse) mice infected at 8 weeks with pathogenic EHV-1 RacL11 resisted challenge and showed eight- and tenfold increases in MLN ASC and lung ASC, respectively, by 3 days after challenge. In contrast to the intranasal route of immunization, intraperitoneal immunization yielded ASC frequencies in the MLN and lungs that were only slightly above those of nonimmunized control mice. These data indicate that immunization with infectious or heat-killed EHV-1 KyA, or rgD, induces significant levels of virus-specific ASC both in the MLN and lungs, a specific memory B-cell response, and long-term protective immunity. The finding that the numbers of ASC induced by the pathogenic strain versus the attenuated strain of EHV-1, which were virtually identical, indicated that the ability to generate a B-cell response is independent of and does not contribute to EHV-1 virulence.     

Equine herpesvirus type 1 (EHV-1)-induced rearrangements of actin filaments in productively infected primary murine neurons

Equine herpesvirus type 1 (EHV-1) causes respiratory disease, abortion and neurological disorders in horses. In the present study, we investigated reorganization of the cytoskeleton in neurons infected with two EHV-1 strains: Jan-E (wild-type strain) and Rac-H (attenuated strain). The studies were performed on primary murine neurons, which are an excellent model for studying neurotropism and neurovirulence of EHV-1. We have demonstrated for the first time that EHV-1 infection causes rearrangements in the actin network of neurons that are dependent on the virus strain and its adaptation to cell culture in vitro. Immunofluorescent labeling and confocal microscopy revealed the formation of long, thin projections in neurons infected with the Jan-E strain, which was probably associated with enhanced intracellular spread of the virus. The EHV-1 Rac-H strain caused disruption of the microfilaments system and general depolymerization of actin, but treatment of neurons with cytochalasin D or latrunculin A resulted in limitation of viral replication. It can therefore be assumed that actin filaments are required only at the early stages of infection. Our results allow us to suggest that the actin cytoskeleton participates in EHV-1 infection of primary murine neurons but is not essential, and that other components of the cytoskeleton and/or cellular mechanisms may be also involved during EHV-1 infection.     

Equid herpesvirus type-1 exhibits neurotropism and neurovirulence in a mouse model

Intranasal inoculation of equid herpesvirus type-1 (EHV-1) Brazilian strains A4/72 and A9/92 induced an acute and lethal infection in four different inbred mouse strains. Clinical and neurological signs appeared between the 2nd and 3rd day post inoculation (dpi) and included weight loss, ruffled fur, a hunched posture, crouching in corners, nasal and ocular discharges, dyspnoea, dehydration and increased salivation. These signs were followed by increased reactivity to external stimulation, seizures, recumbency and death. The virus was recovered consistently from the brain and viscera of all mice with neurological signs. Histopathological changes consisted of leptomeningitis, focal haemorrhage, ventriculitis, neuronal degeneration and necrosis, neuronophagia, non-suppurative inflammation, multifocal gliosis and perivascular infiltration of polymorphonuclear and mononuclear cells. Immunohistochemical examination demonstrated that EHV-1 strains A4/72 and A9/92 replicated in neurons of the olfactory bulb, the cortex and the hippocampus. In contrast, mice inoculated with the EHV-1 Brazilian strain A3/97 showed neither weight loss nor apparent clinical or neurological signs; however, the virus was recovered consistently from their lungs at 3 dpi. These three EHV-1 strains showed distinct degrees of virulence and tissue tropism in mice. EHV-1 strains A4/72 and A9/92 exhibited a high degree of central nervous system tropism with neuroinvasion and neurovirulence. EHV-1 strain A3/97 was not neurovirulent despite being detected in the brains of infected BALB/c nude mice. These findings indicate that several inbred mouse strains are susceptible to neuropathogenic EHV-1 strains and should be useful models for studying the pathogenesis and mechanisms contributing to EHV-induced myeloencephalopathy in horses.     

The IR4 auxiliary regulatory protein expands the in vitro host range of equine herpesvirus 1 and is essential for pathogenesis in the murine model

IR4, an early regulatory protein of equine herpesvirus 1 (EHV-1), is not a DNA-binding protein, but interacts with the sole immediate-early protein (IEP) to increase both IEP site-specific DNA-binding and IEP-mediated trans-activation of EHV-1 promoters. To investigate the biological properties of IR4 and ascertain whether this regulatory protein is essential for virus growth, bacterial artificial chromosome methods were employed to generate an IR4-null EHV-1. The IR4 gene was dispensable for EHV-1 growth in non-immortalized equine NBL-6 cells, but virus replication was delayed and was reduced by greater than 10-fold. In addition, replication of the IR4 mutant was abrogated in all other cell types tested, including equine ETCC tumor cells and cells of mouse, rabbit, monkey, and human origin. Further, in contrast to the highly pathogenic parent virus, the IR4 deletion mutant failed to cause disease in the CBA mouse as judged by assessing body weight and clinical signs and was unable to replicate in the murine lung. To define the nature of the block in the replication of the IR4-null virus, molecular analyses were carried out in RK-13 rabbits' cells infected with the IR4-deleted virus and revealed that: 1) the synthesis of the sole IEP was not inhibited; 2) the synthesis of early viral proteins examined was either not affected or was delayed to late times; 3) viral DNA replication was inhibited by more than 99.9%; and 4) synthesis of essential late proteins such as glycoprotein D and glycoprotein K was prevented. These findings indicate that the IR4 protein is required for EHV-1 DNA replication in non-permissive cells, and, like its homologues in other alphaherpesviruses, contributes a function required for virus replication in a variety of cell types.     

Development of a peptide ELISA for discrimination between serological responses to equine herpesvirus type 1 and 4

A peptide-based enzyme-linked immunosorbent assay (ELISA) for discrimination between serological responses to equine herpesvirus type 1 (EHV-1) and 4 (EHV-4) was developed. Three and four peptides for EHV-1 and EHV-4, respectively, were designed and studied initially in the ELISA using sera from foals infected experimentally. The most promising peptide pair, derived from EHV-1 glycoprotein E and EHV-4 glycoprotein G, was evaluated further using acute and convalescent sera from horses infected experimentally and naturally as well as a panel of horse field sera. Ten pre- and post-vaccination serum pairs were similarly tested in the type-specific ELISA. The peptide ELISA was able to identify horses which had been infected with EHV-1 or EHV-4 as derived from the results using acute and convalescent sera collected from natural outbreaks. When applied to a set of field samples, the assay proved robust with respect to determining the EHV-1 and EHV-4 antibody status. Also, the peptide ELISA was able to detect type-specific seroconversion for EHV-1 in vaccinated animals. With further validation, the EHV-1/EHV-4 peptide ELISA described in this study could serve as a reliable and cost-effective alternative to current methods for serological EHV-1 and EHV-4 diagnosis.     

Generation and characterization of an EICP0 null mutant of equine herpesvirus 1

The EICP0 gene (gene 63) of equine herpesvirus 1 (EHV-1) encodes an early regulatory protein that is a promiscuous trans-activator of all classes of viral genes. Bacterial artificial chromosome (BAC) technology and RecE/T cloning were employed to delete the EICP0 gene from EHV-1 strain KyA. Polymerase chain reaction, Southern blot analysis, and DNA sequencing confirmed the deletion of the EICP0 gene and its replacement with a kanamycin resistance gene in mutant KyA. Transfection of rabbit kidney cells with the EICP0 mutant genome produced infectious virus, indicating that the EICP0 gene is not essential for KyA replication in cell culture. Experiments to assess the effect of the EICP0 deletion on EHV-1 gene programming revealed that mRNA expression of the immediate-early gene and representative early and late genes as well as the synthesis of these viral proteins were reduced as compared to the kinetics of viral mRNA and protein synthesis observed for the wild type virus. However, the transition from early to late viral gene expression was not prevented or delayed, suggesting that the absence of the EICP0 gene did not disrupt the temporal aspects of EHV-1 gene regulation. The extracellular virus titer and plaque areas of the EICP0 mutant virus KyADeltaEICP0, in which the gp2-encoding gene 71 gene that is absent in the KyA BAC was restored, were reduced by 10-fold and 19%, respectively, when compared to parental KyA virus; while the titer and plaque areas of mutant KyADeltaEICP0Deltagp2 that lacks both the EICP0 gene and gene 71 were reduced more than 50-fold and 67%, respectively. The above results show that the EICP0 gene is dispensable for EHV-1 replication in cell culture, and that the switch from early to late viral gene expression for the representative genes examined does not require the EICP0 protein, but that the EICP0 protein may be structurally required for virus egress and cell-to-cell spread.