Report of the equine herpesvirus-1 Havermeyer Workshop, San Gimignano, Tuscany, June 2004.

Amongst the infectious diseases that threaten equine health, herpesviral infections remain a world wide cause of serious morbidity and mortality. Equine herpesvirus-1 infection is the most important pathogen, causing an array of disorders including epidemic respiratory disease abortion, neonatal foal death, myeloencephalopathy and chorioretinopathy. Despite intense scientific investigation, extensive use of vaccination, and established codes of practice for control of disease outbreaks, infection and disease remain common. While equine herpesvirus-1 infection remains a daunting challenge for immunoprophylaxis, many critical advances in equine immunology have resulted in studies of this virus, particularly related to MHC-restricted cytotoxicity in the horse. A workshop was convened in San Gimignano, Tuscany, Italy in June 2004, to bring together clinical and basic researchers in the field of equine herpesvirus-1 study to discuss the latest advances and future prospects for improving our understanding of these diseases, and equine immunity to herpesviral infection. This report highlights the new information that was the focus of this workshop, and is intended to summarize this material and identify the critical questions in the field.

Time of transforming growth factor beta 1 inoculation alters the incubation of BSE in mice.

Groups of 20 C57BL/6J mice (10 males and 10 females) were given BSE strain 301C i.p. and subsequently given 2 microg recombinant human TGFbeta1 s.c. at single or multiple times. There was a significant positive correlation between the day of TGFbeta1 administration and incubation time; the later TGFbeta1 was administered after BSE inoculation the longer the incubation time became. The administration of TGFbeta1 at any time point did not significantly alter the distribution or severity of pathology. The effects of TGFbeta1 on BSE pathogenesis appears to be dependent upon its time of administration; early administration shortens the incubation time and late administration lengthens the incubation time.

Impaired motor coordination on static rods in BSE-infected mice.

Scrapie and bovine spongiform encephalopathy (BSE) are both progressive neurodegenerative diseases that are transmissible to mice. The onset of clinical symptoms is more subtle and variable in murine BSE than in murine scrapie. Assessment of behavioural changes that occur throughout disease would aid early diagnosis of disease so that more consistent end points could be made and potential therapies could be investigated. C57BL/6J mice inoculated via the intraperitoneal route with 301C BSE or control inoculum were monitored on a fortnightly basis. The end point was when a mouse showed clinical signs as opposed to behavioural signs of BSE for two consecutive observations. Significant loss of motor function, as assessed by mice balancing on a static rod, was observed consistently from approximately 40 days prior to death. No significant differences in home cage activity (locomotion, rearing) or cognitive function (T-maze alternation) were observed. However, there was an increase in digging by BSE-infected mice from an early stage. This data will aid the standardisation of behavioural tests to characterise and assess the onset of BSE.

Clusterin shortens the incubation and alters the histopathology of bovine spongiform encephalopathy in mice.

Clusterin accumulates in significant quantity in prion protein lesions associated with bovine spongiform encephalopathy (BSE) and we therefore sought to elucidate its ability to alter BSE pathogenesis and incubation time by comparison of wild type C57BL/6J mice and clusterin knock out (ko) mice. The ko mice had a 40 day increase in mean incubation time compared to wild type mice. PrP deposition in the medulla was less aggregated in clusterin knock out mice when compared to wild type BSE infected mice and a more marked astrocytosis, as determined by GFAP staining, was evident. The vacuolation profiles did not differ between the two strains of mice. Taken together these results suggest that clusterin alters the extracellular deposition of PrP(BSE) and accelerates BSE pathogenesis.

Equid herpesvirus 1 infection of endothelial cells requires activation of putative adhesion molecules: an in vitro model.

Antisera to activated equine endothelial cells, which detected surface molecules of 116 kD, 97 kD, 42 kD and 38 kD, were made to investigate the role of endothelial adhesion molecules in equid herpes virus 1 infection. These putative adhesion molecules could be induced by 17-beta oestradiol, chorionic gonadotrophin, or IL-2, as well as by LPS and PWM. In an in vitro flow system, using equine veins or arteries, equid herpesvirus 1 in leucocytes was only transferred to infect endothelial cells if both leucocytes and endothelial cells expressed these surface molecules. Blocking of the membrane molecules with polyclonal antibodies prevented transfer of virus to the endothelial cells, indicating that the adhesion molecules had a key role in effecting transfer of virus. These in vitro observations give particular insight into the reports that in the natural course of infection in horses infection of endothelial cells is restricted to certain tissues, and in a wider context the results illustrate the complexity of factors that may direct tissue tropism.

Equid herpesvirus 1 is neurotropic in mice, but latency from which infectious virus can be reactivated does not occur.

Equid herpesvirus 1 (EHV-1) is the most common cause of virus-induced abortion in horses. After primary infection the virus becomes latent predominantly in the respiratory tract lymph nodes and the genome can also be detected in the peripheral nervous system. The role of mouse as a feasible model for the establishment of latency and reactivation of EHV-1 was investigated. Intracerebral and intranasal infections of 3- and 17-day-old mice were made and virus replication was confirmed by virus isolation and detected by indirect immunofluorescence (IIF) in brain. For reactivation studies, the mice were killed 8 weeks post infection and tissues were collected for cocultivation. In mice from both age groups, infectious virus was not detected by cocultivation. Following attempts to reactivate virus in vivo with corticosteroids, the viral antigen was detected at low levels by IIF and the expression of the gB gene by reverse transcription polymerase chain reaction (RT-PCR) in brain, trigeminal ganglia, olfactory lobe, lung and spleen. Virus was also detected by IIF following incubation of tissue explants in the growth medium containing pokeweed mitogen (PWM). These results show the limitations of the mouse model for investigating EHV-1 latency and highlights the issue of 'ineffective reactivation' of virus.

Infection of endothelial cells with equine herpesvirus-1 (EHV-1) occurs where there is activation of putative adhesion molecules: a mechanism for transfer of virus.

Evidence is presented to show that activation of endothelial and leucoyte adhesion molecules is a key step in transferring virus from infected leucocytes; and determines the restricted tissue tropism. A range of tissues from 2 experimentally infected mares in late pregnancy at 4 and 8 days after infection with EHV-1 were compared with those from normal pregnant and nonpregnant mares. Rabbit antisera to equine activated endothelial cell molecules were used to identify which tissues expressed these molecules in normal nongravid and gravid mares, and to investigate whether the range of tissues was altered in pregnant mares as a consequence of infection. The results indicated that the endothelium of the pregnant reproductive tract did express these molecules. In the 2 pregnant mares infected with EHV-1, the endothelial cells in the nasal mucosa also expressed these activated endothelial cell molecules. Therefore, the sites expressing these molecules closely correlated with those where virus infection of endothelial cells has been described and is consistent with experimental in vitro data, indicating that expression of these molecules is an essential stage in the transference of virus from leucocytes to endothelial cells.

EHV-1 gene63 is not essential for in vivo replication in horses and mice, nor does it affect reactivation in the horse: short communication.

The aim of this study was to investigate the role of immediate early gene (gene63) in the pathogenesis of equine herpesvirus 1 (EHV-1) acute and latent infections in equine and murine models. EHV-1 gene63 mutant virus (g63mut) along with EHV-1 (Ab4) was used for intracerebral and intranasal infection of 3 and 17-day-old mice. Both viruses were recovered at the same frequency from tissues after infection. Two Welsh ponies were infected via the intranasal route with each of the viruses. Acute infection was monitored by virus isolation from nasal swabs and peripheral blood leukocytes. Six weeks post infection, peripheral blood leukocytes were taken from ponies and in vitro reactivation was positive for both viruses. At autopsy, both viruses were isolated by co-cultivation from bronchial and submandibular lymph nodes. These findings indicate that the mutation of EHV-1 gene63 does not play a role in the establishment and reactivation from latency.

Equid herpesvirus 1: platelets and alveolar macrophages are potential sources of activated TGF-B1 in the horse.

Cell mediated responses to Equid herpesvirus 1 (EHV-1) are of short duration in vivo and require considerable expansion to be detected in vitro. Raised serum levels of active transforming growth factor B (TGF-B1) have been shown to depress proliferative T cell responses in experimental infections with EHV-1 in ponies. The present work indicates that latent transforming growth factor B (TGF-B1) is present in circulating platelets, lymph node, bronchial epithelium and alveolar macrophages. Activation of platelets in vitro by thrombin resulted in the release of latent TGF-B1 from platelets, with a pg level of conversion to active TGF-B1, but virus alone did not activate TGF-B1. Exposure of circulating leucocytes to EHV-1 in vivo or in vitro does not result in detection of active TGF-B1 above residual levels that could be attributed to activation of platelets by manipulation. However, alveolar macrophages obtained by lavage at autopsy yield both latent and active TGF-B1 in ng quantities. Bronchial epithelium, and mesenteric lymph node leucocytes had equivalent levels of latent TGF-B1, but horses varied as to whether these tissues were a source of activated TGF-B1 and as to whether EHV-1 activated TGF-B1.

Clusterin in bovine spongiform encephalopathy (BSE).

Clusterin mRNA, detected in increased quantities in the cervical spinal cord of cattle with bovine spongiform encephalopathy (BSE), was localized mainly in the neuroglia (including astrocytes) of the lateral and ventral areas of white matter. Axonal degeneration was also observed in these areas. The dorsal horns of the spinal cord in which BSE prion protein (PrP(BSE)) was deposited did not exhibit strong clusterin "up-regulation" but showed increased clusterin immunolabelling with a punctate distribution in the neuropil. Labelling of adjacent sections of the grey matter in BSE-affected spinal cord and thalamus demonstrated that the clusterin was deposited in association with extracellular PrP(BSE).

Clusterin prevents aggregation of neuropeptide 106-126 in vitro.

The prion/amyloid neuropeptide 106-126 spontaneously aggregates to form fibrillar structures in vitro. The aggregation in vitro could be prevented in a dose-related manner by clusterin, and the specificity of this action was confirmed by reversal with antibody to clusterin. The relevance of these observations is discussed in relation to previous observations that clusterin and PrPBSE colocalise in naturally occurring cases of BSE.

Equine herpesvirus type 1 infects dendritic cells in vitro: stimulation of T lymphocyte proliferation and cytotoxicity by infected dendritic cells.

Equine herpesvirus type 1 (EHV-1) causes respiratory disease, abortion and myeloencephalopathy in horses. As with other herpesviruses, cell-mediated immunity is considered important for both recovery and protection. Although virus-specific T-cell proliferation and cytotoxicity can be detected following in vivo infection, little is known about the role of antigen presenting cells such as dendritic cells (DCs) in these processes. Peripheral blood DCs were shown to express the viral glycoprotein gB perinuclearly following exposure to EHV-1 in vitro, demonstrating EHV-1 replication within them. Co-culture of infected DCs or their supernatants with a susceptible cell line (RK13) demonstrated that EHV-1 infection was productive. In vitro-infected DCs showed cytopathic effects, including loss of viability and syncytial formation. However, they were superior to other antigen presenting cells in stimulating both peripheral blood T-cell proliferation and cytotoxicity. Although ponies which had been intranasally infected with EHV-1 exhibited T-cell proliferation to live virus presented on DCs, the responses began to decline as early as 15 weeks and cease at 22 weeks post-in vivo infection. Cytotoxic responses were not detected 35 weeks after the first intranasal infection but were seen again 7 weeks following a second infection. These findings show that equine DCs, which are infected with EHV-1 in vitro, can stimulate memory T-cell responses but appear unable to circumvent the short-lived memory response found following this infection in vivo.

The equid herpesvirus-1 gene 63 is expressed as a leaky late (gamma 1) transcript and is nonessential for replication in vitro.

The regulation of Equid herpesvirus-1 (EHV-1) gene 63 was investigated using molecular expression studies and its role in viral growth was identified by constructing a gene 63 mutant virus. Metabolic inhibitors were used to show that EHV-1 gene 63 is expressed as a leaky-late (gamma 1) transcript. Transient transfections and subsequent chloramphenicol acetyltransferase (CAT) reporter assays showed that gene 63 was transactivated by EHV-1 gene 64 (immediate early) protein. An EHV-1 gene 63 mutant virus, where the LacZ gene was inserted into the mutated gene 63 open reading frame, showed that gene 63 protein was not essential for efficient EHV-1 growth in tissue culture. These findings indicate that the animal alpha herpesviruses may have evolved different pathways leading to replication.

In vitro reactivation of latent equid herpesvirus-1 from CD5+/CD8+ leukocytes indirectly by IL-2 or chorionic gonadotrophin.

IL-2 and equine chorionic gonadotrophin (eCG) initiated reactivation of equid herpesvirus-1 (EHV-1) from venous lymphocytes at a frequency of 1/10(-5). Indirect immunofluorescence showed that > 80% of virus-positive leukocytes were CD5+/CD8+ with the remaining 20% being CD5+/CD8-/CD4-. Cocultivation demonstrated that the reactivated virus was infectious. In addition, virus was reactivated in vitro from leukocytes of > 70% of horses by the mitogens phytohaemagglutinin (PHA) and pokeweed mitogen (PWM). Transfer of supernatants showed that IL-2 and eCG acted indirectly by causing the release of other mediators from adherent cells; these mediators then reactivated EHV-1 from T cells. Blocking experiments with anti-IL-2 showed that PWM and PHA acted via IL-2 but that eCG did not. This is the first clear definition of the lymphoid cells that harbour latent EHV-1 in vivo and correlates with current RT-PCR and in situ hybridization of latency-associated transcripts in lymphocytes. This method of reactivation in vitro can be used to detect horses carrying latent EHV-1 in vivo and also has the potential to dissect the sequence of events involved in reactivation in vitro.

Detection of latency-associated transcripts of equid herpesvirus 1 in equine leukocytes but not in trigeminal ganglia.

Results from Southern hybridization and PCR amplification experiments using a randomly synthesized reverse transcription-PCR product showed that peripheral blood leukocytes from horses showing no clinical signs of disease expressed a putative latency-associated transcript antisense to and overlapping the 3' end of the equid herpesvirus 1 (EHV-1) immediate-early gene (gene 64). A PCR product derived from this transcript has > or =96% identity with the published EHV-1 sequence. In situ hybridization studies of equine bronchial lymph nodes corroborated these findings and are consistent with reactivation data (D. A. Smith, A. Hamblin, and N. Edington, unpublished data), indicating that EHV-1 latency is established predominantly in CD5+/CD8+ leukocytes.

Transforming growth factor-beta induced by live or ultraviolet-inactivated equid herpes virus type-1 mediates immunosuppression in the horse.

Up to 21 days after exposure to live or ultraviolet-inactivated equid herpesvirus type-1 (EHV-1) autologous serum from ponies caused an immunosuppressive effect if incorporated into T-cell proliferation assays to EHV-1. The suppressive factor in the sera of ponies also inhibited T-cell response to phytohaemagglutinin. Increased levels of circulating activated transforming growth factor-beta 1 (TGF-beta 1) were detected, and the suppressive activity of the serum could be reversed by antibody to TGF-beta 1. In a challenge experiment the ponies which exhibited circulating TGF-beta 1 activity succumbed to infection while the ones with similar magnitudes of T-cell responses, but no TGF-beta 1 activity, were protected. A definition of this immunosuppressive mechanism and its mode of induction must be central to the design of vaccines and to an understanding of the pathogenesis of EHV-1.

Nucleotide sequence of equine MxA cDNA.

A 2.6 kb cDNA species has been isolated from a cDNA library prepared from interferon-alpha stimulated equine peripheral blood leucocytes and the nucleotide sequence determined. The cDNA has a single open reading frame potentially encoding a 660 amino acid polypeptide showing a high degree of homology with known mammalian Mx proteins, including the possession of three consensus GTP-binding motifs. The protein has a calculated pI = 6.1 and in accordance with proposed nomenclature we have designated it equine MxA.

Isolation and characterisation of equine dendritic cells.

Despite their important role in initiating T-cell responses in other species, dendritic cells have not been studied in the horse. A method for isolating blood dendritic cells by adherence and metrizamide gradients was adapted to equine cells. A number of monoclonal antibodies (mAbs), including some which label dendritic cells in other species, were tested for immunochemical reactivity with the isolated blood dendritic cells, and sections of lymph node and spleen. 62 +/- 6% of the isolated blood cells were MHC Class II positive and had typical dendritic cell morphology and only 4 +/- 2% contained non-specific esterase, a marker of mature macrophages. These dendritic cells also expressed MHC Class I, LFA-1, EqWC1 and EqWC2. Amongst the potentially cross-reactive antibodies a mAb against bovine CD1b was the most interesting by staining lymph node, but not blood, dendritic cells. Monoclonal antibodies against equine CD5 (T-cells), surface immunoglobulin (B-cells) and macrophages (CZ2.2) were used to enumerate the contaminating cells in preparations from blood by flow cytometry. 39 +/- 7% of the cells did not express T and B cell markers or CZ2.2 but were large and MHC Class II positive. Comparison of immuno-chemistry and flow data, together with examination of alveolar macrophages and adhered blood cells, all support the view that CZ2.2 detects a myeloid marker not seen on mature macrophages and possibly shared with dendritic cell precursors. The functional capacity of the isolates was assessed in terms of their stimulating ability in the mixed leukocyte reaction (MLR). Dendritic cell enriched isolates were more potent stimulators of MLRs than peripheral blood mononuclear cells or adherent cells. Thus equine dendritic cells isolated from blood express high levels of MHC Class I and II and LFA-1 and stimulate a vigorous MLR. They do not express markers characterising T and B cells but, by virtue of expression of the equine macrophage marker CZ2.2, appear closely related to mononuclear phagocytes.