Molecular studies of the acute infection, latency and reactivation of equine herpesvirus-1 (EHV-1) in the mouse model

ELSEVIER

Virus Research 40 (1996) 33-45

Virus P sea I

I

[

Molecular studies of the acute infection, latency and reactivation of equine herpesvirus-1 (EHV-1) in the mouse model M . K . BaxP'*, K. Borchers b, T. Bartels c, A. Schellenbach b, S. BaxP, H.J.

Field ~

aCentre for Veterinary Science, Cambridge University Veterinary School, Madingley road, Cambridge CB3 0ES, UK bFreie Universitat Berlin, lnstitut fur Vtroiogie, Konigin-Luise, Str. 49, 14185 Berlin, Germany Clnstitut fur Veterinar-Pathologie, Freie Universitat Berlin, Philippstr. 13, 10117 Berlin, Germany

Received 31 July 1995; revised 3 October 1995; accepted 4 October 1995

Abstract

The murine intranasal (i.n.) infection model was used to study the molecular distribution of equine herpesvirus-1 (EHV-1) during acute infection, latency and following a reactivation stimulus. After inoculation, infectious virus was detected in lungs, nasal turbinates, brains and olfactory bulbs during the acute phase. A nested PCR (nPCR) readily detected virus in these tissues and, in addition, virus was detected in spleens and (in the second round of nPCR) in peripheral blood mononuclear cells (PBMC). A digoxigenin-labelled in situ hybridization probe detected EHV-1 DNA in bronchiolar and vascular endothelium in the lungs and in and around germinal centres in the spleens. One month later, although infectious virus was absent from all tissues, the trigeminai ganglia, olfactory bulb and PBMC remained positive for virus DNA although this was detected only on the second round of nPCR. Furthermore, in situ hybridization, using either DNA or RNA probes, suggested that little or no transcription of virus occurred in neural tissues during the 'latent phase'. Following a reactivation stimulus, infectious virus was not isolated from any tissues, however, EHV-1 DNA was detected on the first round of nPCR in olfactory bulb, trigeminal ganglia and PBMC. This suggested a quantitative increase in EHV-1 DNA occurred following reactivation stimulus. The significance of these results is discussed in relation to the molecular state of EHV-1 in different tissues at various stages of infection and the validity of the murine model for studying latency and reactivation of EHV-1 in the horses. Keywords: Equine herpesvirus-l; Acute infection; Latency; Reactivation

1. Introduction Equine herpesvirus-I (EHV-1) is a major equine pathogen, causing respiratory disease, abortion and, occasionally, neurological disease (Allen and * Corresponding author.

Bryans, 1986; Chowdhury et al., 1986). Similar to other members of the Alphaherpesvirinae, EHV-1 infection is characterised by the establishment of latency and reactivation with shedding of virus from time to time (Burrows and Goodridge, 1984; Edington et al., 1985; Browning et al., 1988; Gibson et al., 1992a). However, EHV-1 had been

0168-1702/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved SSDI 0168-1702(95)01255-9

34

M.K. Baxi et al. / Virus Research 40 (1996) 33 45

thought to be different from other members of the sub-family (e.g. herpes simplex, varicella zoster or pseudorabies viruses) in that the infection of neurons has not been demonstrated. When neurological disease does occur in the horse, this is caused by lesions of vascular origin resulting in damage to neurons only as a secondary event (Patel et al., 1982; Ludwig et al., 1987). Furthermore, it appears that lymphoid tissues harbour virus during latency (Welch et al., 1992). It has been observed that virus DNA can be detected by means of polymerase chain reaction in both central nervous system (olfactory bulb) and trigeminal ganglia of ponies during latency (Welch et al., 1992; Slater et al., 1994; Edington et al., 1994); furthermore, infectious virus was reactivated from explanted trigeminal ganglia (Slater et al., 1994). Using in situ hybridization to study ganglion tissue sections from the same ponies we observed that a very small proportion of neurons expressed EHV-l-specific RNA (Baxi et al., 1995). The difficulty of obtaining suitable tissues of known infection status from the natural host hampers research on the relative importance of EHV-1 neurotropism and this justifies the use of laboratory animal model. A murine model for EHV-1 has been developed and is now well-characterised (Awan et al., 1990; Azmi and Field, 1993a,b; Inazu et al., 1993; Csellner et al., 1995). Infection is established by intranasal inoculation and the resulting disease has many features that resemble the natural infection including respiratory clinical signs, infection of the respiratory mucosa (Field et al., 1992a), viraemia, and the production of abortion in pregnant mice (Awan et al., 1990, 1995). Furthermore, following recovery from the acute infection, appropriate stimuli caused reactivation of an apparently latent infection in a proportion of mice (Field et al., 1992b). It has been observed by means of a deletion/insertion mutant of EHV-1 containing the LacZ reporter gene that neurons of the trigeminal ganglion and olfactory bulb expressed fl-galactosidase during the acute phase of the infection (K. Marshall personal communication). However, the anatomical site and cellular state of latent virus was not

determined and, to our knowledge, there has been no published report on the neurological tropism of EHV-1 following intranasal inoculation of adult mice. In the present paper we have used molecular techniques to study the acute and latent infection in mice and the events that occur following a reactivation stimulus.

2. Materials and methods

2.1. Virus and cells EHV-1 strain Ab4 was originally isolated from a clinical case with neurological complications (paresis). The pathogenicity of this strain has been studied in horses (Gibson et al., 1992a, b; Slater et al., 1993) and mice (Awan et al., 1990) and its nucleotide sequence has been determined (Telford et al., 1992). Virus was grown according to published methods (Gibson et al., 1992a). 2.2. Experimental design Three hundred 3 - 4 weeks old Balb/c mice were infected with 5 x 106 pfu/mouse of EHV1 Ab4 intranasally under light anaesthesia. The experiment was divided into 3 phases - - acute (days 1, 3, 5, 8 p.i.), latent (days 30, 35, 40 p.i.) and reactivation. Reactivation of latent virus was attempted at 50 days p.i. using one of the two stimuli. One group was given dexamethasone (DXM), s.c. injection once daily at a dose of 8 mg/kg/day for 7 days and the other group was given cyclophosphamide (CPS), s.c. at the dose of 200 mg/kg/day for 5 days. The third group was kept as an unstimulated control. Tissues were sampled daily for 10 days post-reactivation. Mice were killed by an overdose of pentobarbitone. Lung, nasal turbinate, brain, olfactory bulb, trigeminal ganglia, liver, spleen and peripheral blood lymphocytes were collected at various time points p.i. The tissues were aseptically removed and were fixed in periodate-lysine-paraformaldehyde for 2 h, transferred to 50% ethanol and paraffin embedded.

M.K. Baxi et al. / Virus Research 40 (1996) 33-45

2.3. Clinical assessment Groups of 10 mice were examined and weighed individually for 14 days post-infection. Clinical signs were monitored including respiratory distress, ruffled coats, dragging movements and death. 2.4. Virus isolation Virus isolation from tissue homogenates was carried out using RK-13 cell monolayer according to published methods (Awan et al., 1990). 2.5. Explants and co-cultivation Tissues were aseptically cut into 1-mm 2 fragments and placed on confluent RK-13 monolayers and overlaid with 1% CMC containing 2% fetal calf serum at 37°C in a humidified atmosphere of 5% CO2. Cultures were maintained for 7 days, then sub-cultured twice at weekly intervals if no cpe was observed. 2.6. Alkaline phosphotase anti-alkaline phosphatase (APAAP) technique Paraffin sections were cut on to saline-coated slides and dried overnight at 56°C. After dewaxing in xylene and hydrating in graded ethanol and rinsing in Tris-buffered saline (TBS), the slides were transferred into citrate buffer. The slides were incubated for 3 x 10 min at 95°C in a microwave oven and then subsequently cooled at room temperature. The sections were incubated with polyclonal rabbit anti-EHV serum diluted 1:160 in TBS for 60 min at 37°C. This was followed by washing in TBS and incubation with the first link antibody mouse anti-rabbit immunogiobulin (Dianova, Germany) at a dilution 1:50 in TBS and the second link antibody rabbit anti-mouse (Dianova, Germany) at a dilution 1:20 both for 30 min at room temperature, terminating by rinsing with TBS. Subsequently the sections were incubated with the monoclonal mouseAPAAP complex (Dianova, Germany) also for 30 min at room temperature and afterwards rinsed in TBS. The alkaline phosphatase activity was de-

35

tected by the Naphthol-As Biophosphate (substrate)/New fuschin (chromogen). New fuchsin (5%, 0.2 ml) (Serva, Germany) in 2 N HC1 solution was mixed with 0.5 ml of sodium nitrate and shaken for 1 min. Subsequently 100 ml 0.005 M Tris-buffer (pH 8.7) and t00/zl 1 M Levamisole (Sigma, USA) were added (solution A). Naphthol-As Biophosphate (50 mg) (Chemapol, Czechoslovakia) was diluted in 0.6 ml dimethylformamide (Solution B). Solutions A and B were mixed and filtered. The sections were incubated with the chromogen substrate for 10 min; staining was stopped by rinsing in distilled water. Finally the sections were counterstained with 2% methyl green (10 min) and dehydrated in an ethanol series. 2. 7. Indirect immunoperoxidase Following blocking, with the endogenous peroxidase (0.5% H202) in menthol, incubation was with the same primary antibody as used in the APAAP technique (1:64, 60 min at 37°C) followed by a TBS wash. The sections were incubated for 30 min at 37°C with peroxidase-conjugated goat anti-rabbit IgG (Dianova, Germany) diluted 1:50. Subsequently, peroxidase activity was detected by a solution containing 5 mg diaminobenzidinehydrochloride (Sigma, USA), 10 ml Tris-HCl buffer (pH 7.6) and 5 ~1 H202. After 10 min the sections were washed in running water to stop the reaction. The slides were counter-stained with methyl green. 2.8. DNA extraction High molecular weight (Mr) DNA was extracted from murine tissue by cutting them into small pieces and suspending in digestion buffer comprising 24 mM EDTA, 75 mM NaCI, 1% SDS, 0.2 mg/ml proteinase K incubated at 56°C overnight. The digest was then extracted with equal volumes of phenol, phenol-chloroformisoamylalcohol and finally with chloroformisoamylalcohol. DNA was precipitated with 2 volumes of cold absolute alcohol and 1/10th volume 3 M sodium acetate pH 5.4 at - 2 0 ° C overnight and pelleted at 15000 rev./min and

36

M.K. Baxi et al. / Virus Research 40 (1996) 33-45

washed with 70% alcohol. The DNA was suspended in 100 #1 distilled water. The DNA extracted was checked by electrophoresis and spectrophotometry before use in PCR reactions. For DNA extraction, PBMCs were prepared from heparinized blood by eentrifugation through a Ficoll gradient. Cells were suspended in lysis buffer (75 mM NaC1, 0.4% SDS, 25 mM EDTA, 0.2 mg/ml proteinase K) and incubated overnight at 370C. DNA was extracted 3 times with phenol:chloroform:isoamyl alcohol (25:24:1) and then precipitated with ethanol by incubating at - 700C f o r l h. 2.9. Polymerase chain reaction (PCR) The PCR used in these experiments was a typespecific nested PCR which amplifies a unique region of the EHV-1 gB gene as described by Borchers and Slater, 1993. 2. I0. Probes 2.10.1. DNA probes A 1.993 kb BamHI fragment derived from EHV-1 gB gene was cloned into pUC18 vector. The pUC18/BamHI clone was digested with BamHI. The insert DNA was gel purified and then randomly labelled with digoxigenin (Boehringer Mannheim) according to the manufacturer's instructions. The efficiency of labelling of each restriction fragment was assessed by dot blot assay. 2.10.2. RNA probes In situ hybridization using digoxigenin-labelled R N A probes was carried out in order to study the transcriptional activity of EHV-1. The same R N A probes have been used to detect R N A in the trigeminal ganglia of specific pathogen-free ponies during latency (Baxi et al., 1995). A NarI restriction fragment was isolated from the EHV-1 BamHI E fragment (gift from Dr. G. Lawrence) and cloned in both orientations into the vector pBhiescript + (laBS+ ) (Stratagene). In vitro 'run-off transcriptions of the linearized NarI subclones, using the pBS + T7 RNA polymerase promoter, gave rise to riboprobes which encode

sequences complementary (Ri) and sense (Rii) to ORF63 mRNA (HSV-1 homologue ICP0). In vitro transcription was performed according to manufacturer's recommendations (Boehringer Mannheim). From each transcription reaction ,-~ 0.5-2.0 /zg of RNA was recovered, as determined by incorporation of radioactive tracer ([32p]UTP). 2. I I. In situ hybridization The protocols of Hukkanen et al. (1990) and Arthur et al. (1993) were followed with slight modifications. Thin paraffin wax sections (5 /~m) were deparaffinized in xylene and rehydrated through graded ethanol solutions, fixed in 0.1% glutaraldehyde/phosphate buffer saline (PBS) for 30 min at 4°C and washed in PBS for 5 min. Digestion of the tissue with 100 gg/ml proteinase K in PBS for 15 min at 37°C was carried out to increase the accessibility of the target nucleic acids. Slides to be probed for viral R N A were treated with 100/~ g/ml DNase A (RNase free) for 30 min at 37°C. After washing in glycine/PBS (2 mg/ml) to stop the enzyme reaction, the slides were refixed in 0.1% glutaraldehyde/PBS for 15 min at 40C. Slides were rinsed in 0.1 × SSC before acetylation with 0.25% acetic acid in 0.1 M triethanolamine, pH 8.0 for 10 min. After acetylation the tissues were washed in 0.1 x SSC for 10 min and dehydrated through graded ethanol solutions. Probes were prepared as described below. The hybridization mixture consisted of 50% of deionized formamide/dextran sulphate; 2 x Denhardt's solution; 2 x SSC; 400 ng/mi of herring sperm DNA (10 mg/ml) and 100 ng of EHV-1 gB probe or riboprobes (Ri or Rii) labelled with digoxigenin. The hybridization solution was heated to 100°C for 10 min for the DNA probe and 3 min for the R N A probe, then quenched on ice. Digoxigenin-labeUed probe (20/zl) was added to each section. Glass coverslips coated with Prosil were sealed in place with rubber cement and sections incubated at 42°C for 24 h. When hybridizing with riboprobes the sections were incubated at 720C for 24 h. After hybridization, unbound probe was removed by washing sequentially in 2 x SSC 5 times for 10 rain followed by

M.K. Baxi et al. / Virus Research 40 0996) 33-45

a wash at 55°C for 30 min, and a final 30 min wash in 0.1 x SSC at 42°C, The RNA sections had an additional stringent washing s t ~ with 0.1 x SSC, 30% dcioaiz~ formamide and 10 mM Tris-HC1 pH 7.5 at 75*(2 for 30 rain. Antibody and colour d e v ~ t was carried out with anti-digoxigenin aa~xxlics eouplctl with alkaline phosphatase and d~'doped according to manufacturer's instructions ~ r Mannhoim). Slides were wa.~a~l thoroughly in t a p w a t e r to stop colour reaction prior to fixing in aquamount (BDH).

37

ruffled fur and respiratory distress. Overall, 36% mortality was observed during the acute phase of infection with the maximum number of mice dying on day 3 (19%). No clinical signs were observed in control mice that were inoculated with uninfected RK cell lysate. Three mice were killed on each of days 1, 3, 5 and 8 p.i. and various thumu tested for the presence of infectious virus. Infectious virus was recovered from lung and nasal turbinates up to 8 days p.i. (Fig. 1) with maximum titres on day 3 p.i. Virus was also isolated from brain and olfactory bulb at days 3 and 5 p.i. only. However, infectious virus was not recovered from trigeminal ganglia, spleen, liver or PBMCs. EHV-1 DNA was detected by first round nPCR on days 3 and 5 p.i. in lung, nasal turbinates, brain, olfactory bulb, trigcminal ganglia and spleen (Fig. 2b, Table 1). Lung and nasal turbinates were positive in the first round of nPCR on day 8 p.i. and on the second round nPCR

3. Results

3.1. Acute phase of infection Balb/c mice were infected i.n. with 5 × l0 s pfu/mouse. All mice showed visible clinical signs between days 2-6 p.i., especially weight loss, 7-

e'

4i

:

k

"

J

1

3

5

8

12

35

Reactivation

Days p.i. Fig. 1. Virus replication in respiratory and CNS tissues following EHV-I intranasal inoculation of Balb/c mice. Bars represent geometric mean titre + S.D.; n = 3.

M.K. Baxi et al. / Virus Research 40 (1996) 33-45

38

b

a M

i

2

4

3

M

~

2

3

4

5

6

7

8

3,o.| 2.O63

!

1.636 1,018

d

C M

1

2

3

4

5

6

7

M

8

1

2

3

4

5

6

7

8

3,054 2,063 1.636

3.054 2,063 1,636

....

1.018

.~,

1.018,--

J Fig. 2. The analysis of murine tissues using EHV-l-specific nested PCR to detect EHV-1 DNA. (a) Lanes 1 and 3, negative reagent control; lanes 2 and 4, EHV-1 D N A virus positive control. First round PCR in lanes 1 and 2; 2nd round PCR in lanes 3 and 4. PCR performed on (b) acute, (c) latent and (d) murine tissues following attempted reactivation.. Lanes 1, lung; 2, nasal turbinates; 3, brain; 4, olfactory bulb; 5, trigeminal ganglia; 6, spleen; 7, liver; 8, PBMCs. The arrow head (1~) indicates the 1st round 1,88 kb nPCR product using the EHV 1-specific outer set primers. First round products (1/~1) were amplified with the inner set of 2nd round primers resulting in the 1.28 kb product indicated by an arrow (---,). Results of 2nd round of nPCR are not shown for the tissues found to be positive on 1st round of nPCR. M, 1 kb D N A ladder marker.

M.K. Baxi et al. / Virus Research 40 (1996) 33-45

39

Table 1 R e s u l t s o f nested P C R a n a l y s i s o f tissues t a k e n f r o m mice a d u r i n g acute, l a t e n t a n d r e a c t i v a t i o n p h a s e o f E H V - I infection Tissue

Lung Nasal turbinates Brain Spleen Olfactory bulb Trigeminal ganglia Liver PBMCs

Acute phase Day 3

Day 5

1

2

1

2

1

+ + + + + + . -

+ + + + + +

+ + + + + +

+ + + + + +

+ + . . -

.

. +

. -

Latent phase D a y 30

D a y 35

2

1

1

+ +

. .

Day 8

. +

. .

. -

. . + + . +

2 .

.

. . .

.

. -

.

. +

2

2

.

.

+ . .

. -

1

.

. . . -

. . + +

.

Reactivation

. . + + . +

+ +

+ +

+

+

a N u m b e r o f m i c e tested a t each t i m e p o i n t ; n = 3 d u r i n g a c u t e a n d l a t e n t p h a s e s a n d n = 5 f o l l o w i n g a t t e m p t e d r e a c t i v a t i o n for each d a y ( 1 - 1 0 ) . 1 , 1st r o u n d o f nested P C R . 2 , 2 n d r o u n d o f nested P C R .

olfactory and trigCminal ganglia were also positive. Viral DNA could be detected in PBMCs only on the second round nPCR on days 3, 5 and 8 p.i. All other tissues were negative by nPCR on day 8 p.i. on first or second nPCR. Immunolaistochemistry was carried out on the paraffin-embedded tissues to study the cellular distribution of the viral antigen in target organs especially lung and nasal tissues. In lung tissue, the epithelium lining the bronchioles showed a strong positive reaction with the APAAP technique (Fig. 3a). Cells staining positive for EHV-1 were frequently observed in the lumen of bronchioles. EHV-1 was also detected in endothelial cells lining the blood vessels (Fig. 3b). In some cases, cells in the connective tissue surrounding the vessels also appeared to be positive for viral antigen. No positive stainhig was observed in any of the uninfected control tissues. In nasal tissues the positive reaction for EHV-1 was weak compared to the lung, but EHV-1 positive cells were seen both in the epithelium (Fig. 3c) and in the connective tissue of the mucosal membrane (propria mucosae) (Fig. 3d). The stratified respiratory epithelium showed a focal staining pattern, which included cells of all layers. Tissues obtained post-mortem, were analysed by in situ hybridization to determine the distribution of EHV-1 DNA and RNA. An EHV-1 spe-

cific digoxigenin-labelled probe was used for detection of viral DNA. Uninfected tissues gave negative results in all the cases (Fig. 4a and Fig. 4e), however, the probe detected EHV-l-containing cells in the bronchiol~ epithelium (Fig. 4b) 3 days p.i. and in the desquamated cells in the lumen of the bronchioles (Fig. 4e) 5 days p.i. The maximum number of EH~-I DNA-containing cells was observed in the bronchiolar epithelium on days 3 and 5 p.i. In addition to the positive bronchiolar epithelium, cells containing EHV-1 DNA were also observed in the endothelium lining the small blood vessels in the lung (Fig. 4d) although no evidence of thrombosis or occlusion of these vessels was observed. All other tissues examined were negative using this technique except for the spleen and, in this case, signal was detected in a :very few cells in and around the germinal centres (Fig. 40 in mice examined at day 3 p.i. only. Viral RNA was detected in the lungs using a riboprobe speofic for gene 63. The pattern of positive staining cells obtained with riboprobes was similar to that desedl~l for DNA probes, however, the number of positive staining cells was smaller, and the maximum number of positive cells was observed at days 3 and 5 p.i. No signal was detected using the probe derived from the complementary strand of gene 63.

Fig. 3. Detection of EHV-I antigen in tissues from EHV-1 infected Balb/c mice. (a) Localization of EHV-1 in the lungs of infected mice 3 days p.i. Bronchiolar epithelium is strongly labelled red ( --. ) by the APAAP technique (B). Desquamation of the bronehiolar epithelium is visible in adjacent bronchus (Br). Co) Labelling of the EHV-1 positive endothelial cells ( ~ ) lining the vein (V) above the bronchus (B) by APAAP technique. Immunoperoxidase stained cells positive for EI-IV-I in the nasal turbinate. Specific focal brownish reaction seen both in (c) mucosa and submucosa in the multilayered epithelium and (d) nasal connective tissue.

Fig. 4. Detection of EHV-I DNA by in aitu hybridization using EHV-1 gB specific digoxigenin-labelled probes. (a) Uninfected control mice; section of lung 3 days p.i. Co) Viral DNA-containina cells in the bronchiolar epithelium (--,) in lung 3 days p.i. (c) Desquamation of the bronchiolar epithelium in mouse lung 5 days p.i. EHV-1 DNA- containing cells are visible in the lumen and the epithelium of the infected bronchioles. (d) EHV-1 DNA-containing positive endothelial cells lining the blood vessel (--,). (e) Uninfected control mice; section of spleen 3 days p.i. (f) A few cells containing the viral DNA seen in and around the germinal centres in the spleen from infected mouse 3 days p.i.

I

M.K. Baxi et al. / Virus Research 40 (1996) 3 3 - 4 5

42

Table 2 Detection of EHV-1 DNA by nested PCR followingattempted virus reactivitation Virus

Reactivation

Days post-reactivation 1

EHV-1

No drug Cyclophosphamide Dexamethasone

0/5 0/5 0/5

2

0/5 0/5 0/5

3

0/5 1/5 2/5

4

5

0/5 4/5 5/5

6

0/5 2/5 5/5

0/5 0/5 3/5

7

0/5 0/5 1/5

8

0/5 0/5 0/5

9

0/5 0/5 0/5

10

0/5 0/5 0/5

Period of reactivation:cyclophosphamide,3-5 days; dexamethasone,3-7 days.

3.2. Latent phase of infection Samples were collected at days 30, 35 and 40 p.i., by which time all tissue samples were negative for infectious virus. Tissues collected from groups of 3 mice were co-cultivated using methods previously described for the detection of latent herpes simplex virus, but none of the explant cultures yielded infectious virus. Neither was viral antigen or viral DNA detected in any of the tissues collected using APAAP or in situ hybridization techniques. HOWever, the more sensi- tive nested PCR detected viral DNA (but only on the second round) in olfactory bulb, trigeminal ganglia and PBMCs on each of the above days (Fig. 2c, Table 1). However, viral DNA could not be detected by nPCR in lung, nasal turbinates, brain, spleen or liver. Furthermore, use of digoxigenin-labelled DNA probes spanning the region ofgene 63 (which detected RNA containing cells in equine trigeminal ganglia) did not detect any viral DNA or RNA in any of the tissues positive with nPCR.

3.3. Attempted reactivation of infection Balb/c mice that survived the acute infection were given a variety of stimuli at 50 days p.i. in an attempt to reactivate latent virus. Cyclophosphamide and dexamethasone were administered for 5 and 7 days, respectively. Tissue samples were cultured for virus isolation, but no infectious virus was recovered. In situ hybridization failed to detect any EHV- 1 DNA or RNA positive cells in any of the tissues post-reactivation. Samples were collected from groups of 5 mice for 10 days post-reactivation for nPCR analysis. EHV- 1 DNA was detected on first round of amplification in olfactory bulb, trigeminal ganglia and PBMCs

between day 3-5 in the case of mice treated with CPS and between days 3-7 following DXM administration (Fig. 2d, Table 1). In total, 32% (16/50) mice were positive on first round ofnPCR for EHV- 1 DNA following administration of DXM and 14% (7/50) following CPS. A total of 16/20 of the mice were found to be positive at days 4 and 5 post-reactivation (Table 2). Nasal turbinates (12/100) (in which no EHV-1 DNA was detected at day 45, prior to reactivation stimulus) were found to be positive on second round of nPCR. 4. D i m m s s i o n

The most important findings from this study are: (1) following clearance of infectious virus at the end of the acute phase of infection, EHV-1 DNA remains detectable in neural sites and peripheral blood cells for at least several months. (2) EHV-1 DNA could not be detected in several extra-neural sites including the nasal turbinates and lungs (which are the most important target organs during the acute infection) or spleen. It was of interest, however, that viral replication was detected in the endothelial cells lining small blood vessels while EHV-1 DNA was detected in peripheral blood for at least several weeks. (3) Following attempted reactivation stimuli, there was evidence for an increase in the quantity of EHV-1 DNA present in trigeminal ganglia, olfactory bulb, peripheral blood cells and nasal turbinates although infectioiJs virus was not isolated in these experiments. (4) Despite an extensive search using several sensitive techniques, there was no evidence of EHV-1 RNA expression in ganglionic or olfactory bulb neurons during latency. Confirming previous reports (Awan et al., 1990; Inazu et al., 1993; Csellner et al., 1995), the lung

M.K. Baxi et al. / Virus Research 40 (1996) 3 3 - 4 5

and nasal tissues were found to be the important site for virus replication and infectious virus was isolated up to 8 days p.i. Immunohistochemical studies readily detected viral antigen in the bronchiolar epithelium. Similar studies on ponies experimentally infected with EHV-1 showed rapid dissemination of EHV-1 throughout the respiratory tract, with early replication in the lungs (Kydd et al., 1994). In mice, in situ hybridization for viral DNA and R N A also detected virus in bronchiolar epithelium. However, the number of cells positive for viral transcription was low compared to the number of antigen- and DNA-positive cells. The predilection of EHV-1 for respiratory organs in the mouse emphasizes the resemblance of this model to the natural infection in the horse. The detection of antigen and virus DNA in the endothelial cells lining small blood vessels suggests a further characteristic of the mouse model which closely resembles the infection in the natural host. Infection of vascular tissue is thought to play a crucial role in EHV-1 pathogenesis. Endotheliotropism in the nasopharynx and a role for viraemia in dissemination of EHV-1 to sites of secondary infection has been observed in horses (Kydd et al., 1994). The observation of antigencontaining endothelial cells in the endometrium of aborted mares (Smith et al., 1992) and the small blood vessels in the placental parenehyma in the murine abortion model which contain EHV-1 D N A (Awan et al., i995) suggests that damage to the vascular endothelium is a primary factor in the cause of abortion in mares and mice. The isolation of low levels of infectious virus from the olfactory bulbs and brains during the acute infection are previously unreported findings and form an interesting parallel with the more recent reports of tissue distribution in the natural host (Welch et al., 1992; Slater et al., 1994; Edington et al., 1994). Recently, using a deletion insertion mutant of EHV-1 which contains the lacZ reporter gene, ganglionic and olfactory bulb neurons have been shown to express p-galactosidase (K. Marshall, personal communication) however, EHV-1 antigen was not detected in these tissues during the acute infection by immunohistochemical methods. These observations, together with our previous results that virus can be reactivated from

43

EHV-l-infected mice many months after the primary infection, suggested that neural tissue may be a site for virus latency and reactivation in the murine model. This hypothesis was investigated using the techniques of nested PCR and in situ hybridisation. Various methods for the use of PCR to detect and differentiate EHV-1 and -4 have been reported (Ballagi-Paordany et al., 1990; Sharma et al., 1992; Borchers and Slater, 1993). In our model, the use of nPCR identified a continued presence of EHV-1 genome in certain tissues. During the latent phase, viral DNA was detected in olfactory bulb, trigeminal ganglia and PBMCs on second round nPCR. This suggests that a very low number of genome copies (in the region of 10-104 genomes/ organ) are present. This low level of EHV-1 D N A may explain our failure to reactivate infectious virus from these tissues. Recent studies have shown the presence of latency-associated transcripts (LATs) in the trigeminal ganglia of EHV-I-infected specific pathogen-free ponies, even though the number of positive neurons (0~03%) was very low (Baxi ct al., 1995). However, using identical methods in the present study no transcripts could be detected in any of the putative sites of latency in the mouse. This suggests that the number of transcripts in these tissues is either very low or there is none. It has been shown in HSV latency that the number of cells containing HSV-1 D N A by in situ PCR was greater than the LAT-positive cells detected by in situ hybridization (Ramakrishnan et al., 1994; Mehta et al., 1995). For HSV it has also been suggested that there are neurons which are transcriptionally inactive but harbour latent HSV (Ecob-Prince et al., 1995). One interpretation of our results is that all EHV-l-containing cells are transcriptionally inactive in the mouse or at least remain below the level of detection. Dexamethasone and cyclophosphamide were used as stimuli to reactivate latent virus. Dexamethasone administration has been shown to be a reproducible method for reactivation of EHV-1 in horses (Edington et al., 1985) and in SPF ponies (Gibson et al., 1992a,b,c). Attempted reactiyation using these two drugs did not result in shedding of infectious virus from any of the murine tissues; contrary to results reported earlier (Field et al., 1992b). However, following the attempted reacti-

44

M.K. Baxi et al. / Virus Research 40 (1996) 33-45

vation viral D N A was detected by nPCR in the first round in olfactory bulb, trigeminal ganglia and PBMCs. All these tissues were positive only on the second round prior to stimulation thereby suggesting that there may be a low level of reacti-vation in these tissues. The effect obtained with dexamethasone was seen to be greater and lasted longer compared to cyclophosphamide. On days 4 and 5 after commencing reactivation the above tissues in 80% of mice yielded positive results using this method. This study extends our knowledge of the distribution of virus in the murine intranasal model; particularly in relation to the establishment of neural latency. These results have implications for the natural infection in the horse where, traditionally, lymphoid tissues were regarded as the most important sites for latency and reactivation. However, evidence of neural latency in horse has been recently reported; our results in the mouse model give further evidence of a role for neural tissue and provides a useful system for further study in comparison with the natural host. This will be valuable in our efforts to determine the mechanism of reactivation and the relative importance of the different anatomical sites that may harbour latent virus with the potential for producing recurrent infection.

Acknowledgements M.K.B. is a holder of a Nehru Cambridge Scholarship, Cambridge Commonwealth award and a Jowett Trust award. H.J.F. is in receipt of a British Council Anglo-German academic exchange award and acknowledges the generous support from the Equine Virology Research foundation. K.B. is supported by. a D F G grant BO 1005/3-1. The authors are grateful to Drs. H. Ludwig, S. Efsthathiou, F. Steinbach, V. Bergmann, A.R. Awan and Mr. K. Marshall for helpful discussion throughout this work. We thank T. Leiskau, G. Neuendorff, G. Hahn and U. Wolfinger for technical assistance.

Refcru Allen, G.P. and Bryans, J.T. (1986) Molecular epizootiology, pathogonosis and prophylaxis of equine herpesvirus-1 infections. Prog. Vet. Microbiol. Immunol. 2, 78-144. Arthur, J., Efstathiou, S. and Simmons, A. (1993) Intranuclear foci containing low abundance herpes simplex virus latency-associated transcripts visualized by non-isotopic in situ hydridization. J. Gen. Virol. 74, 1363-1370. Awan, A.R., Baxi, M.K. and Field, H.J. (1995) EHV-1 induced abortion in mice and its relationship to stage of gestation. Res. Vet. Sci. 59, 139-145. Awan, A.R., Chong, Y.-C. and Field, H.J. (1990) The pathogenesis of equine herpesvirus type 1 in the mouse: a new model for studying host responses to infection. J. Gen. Virol. 71, 1131-1140. Azmi, M. and Field, H.J. (1993a) Cell-mediated antiviral response to equine herpesvirus type 1 demonstrated in a murine model by means of adoptive transfer of immune cells. J. Gen. Virol. 74, 275-280. Azmi, M. and Field, H.J. (1993b) Interactions between equine herlg~virus type 1 and equine herpesvirus type 4: T cell responses in a murine model. J. Gen. Virol. 74, 2339-2345. Ballagi-Paordany, A., Klingeborn, B., Flensburg, J. and Belak, S. (1990) Equine herpesvirus type !: detection of viral DNA sequences in aborted fetuses with the polymerase chain reaction. Vet. Microbiol. 22, 373-381. Baxi, M.K., Efsthatiou, S., Lawrence, G.L., WhaUey, J.M., Slater, J.D. and Field, H.J. (1995) The detection of latency associated transcripts of equine herpesvirus 1 (EHV-1) in ganglionic neurons. J. Gen. Virol. 76, 3113-3118. Borchers, K. and Slater, J. (1993) A nested PCR for the detection and differentiation of EHV-1 and EHV-4. J. Virol. Methods 45, 331-336. Browning, G.F~, Bulach, D.M., Ficorilli. N., Roy, E.A., Thorp, B.H. and Studdert, M.J. (1988) Latency of equine herpesvirus 4 (equine rhinopneumonitis virus). Vet. Rec. 123, 518-519. Burrows, R. and Goodridge, D. (1984) Studies of persistent and latent equid herpesvirus-I and herpesvirus-3 infections in the Pirbright pony herd. In: G. Wittman, R.M. Gaskell and H.-J. Rzhia (Eds), Latent Herpesvirus Infections in Veterinary Medicine, pp. 307-319. Martinus Nijhoff, The Hague. Chowdhury, S.I., Kubin, G. and Ludwig, H. (1986) Equine berpesvirus Type 1 (EHV-I) induced abortions and paralysis in a Lipizzaner stud: a contribution to the classification of equine herpesviruses. Arch. Virol. 90, 273-288. CseUner, H., WhaUey, J.M. and Love, D.N. (1995) Equine herpesvirus 1 HVS25A isolated from an aborted foetus produces disease in Balb/c mice. Aust. Vet. L 62, 68-69. Ecob-Prince, M.S., Hassan, K., Denhccn, M.T. and Preston, C.M. (1995) Expression of fl-galactosidase in neurons of dorsal root ganglia which were latently infected with herpes simplex virus type 1. J. Gen. Virol. 76, 1527-1532. Edington, N., Bridges, C.G. and Huckle, A. (1985) Experimental reactivation of equid herpesvirus 1 (EHV-1) follow

M.K. 8 a x i et al. / Virus Research 40 (1996) 3 3 - 4 5

ing administration of corticosteroids. Equine Vet. J. 17, 369-372. Edington, N., Welch, H.M. and Grifliths, L. (1994) The prevalence of latent equid herpesvirus in the tissues of 40 abattoir horses. Equine Vet. J. 26, 140-142. Field, H.J., Awan, A.R., De La Fuente, R. and Gibson, J.S. (1992a) Use of a novel murine model for the study of EHV-1 host responses, immunoprophylaxsis and chemotherapy. In: W. Plowright, P.D. Roosdale and J.F. Wade (F_As),Equine Infectious Diseases, pp. 263-268. R. and W. Publications, Newmarket. Field, H.J., Awan, A.R. and De La Fuente, R. (1992b) Reinfection and reactivation of equine herpesvirus 1 in the mouse. Arch. Virol. 123, 409-419. Gibson, J.S., Slater, J.D., Awan, A.R. and Field, H.J. (1992a) Pathogenesis of equine herpesvirus-1 in specific pathogenfree foals: primary and secondary infection and reactivation. Arch. Virol. 123, 351-366. Gibson, J.S., Slater, J.D., Awan, A.R. and Field, H.J. (1992b) The pathogenicity of Ab4p, the sequenced strain of equine herpesvirus-1, in specific pathogen-free foals. Virology 189, 317-319. Gibson, J.S., O'Neil, T., Thackray, A., Hannant, D. and Field, H.J. (1992c) Serological responses of specific pathogenfree foals to equine herpesvirus-l: primary and secondary infection and reactivation. Vet. Microbiol. 32, 199232. Hukkanen, V., Heino, P., Sears, A.E. and Roizman, B. (1990) Detection of herpes simplex virus latency-associated RNA in mouse trigeminal ganglia by in situ hybridization using nonradioactive digoxigenin-labelied DNA and RNA probes. Methods Mol. Cell. Biol. 2, 70-81. Inazu, M., Tsuha, O., Kirishawa, R., Kawakami, Y. and Iwai, H. (1993) Equid herpesvirus 1 infections in mice. J. Vet. Med. Sei. 51, 119-121. Kydd, J.H., Smith, K.C., Hannant, D., Livesay, G.J. and Mumford, J.D. (1994) Distribution of equine herpesvirus-1 (EHV-I) in the respiratory tract of ponies: implications for vaccination strategies. Equine Vet. J. 26, 466-469. Ludwig, H., Rudolph, R., Chowdhury, S.I., Van Den Bossche, G., Wintzer, H.J. and Krauser, K. 0987) Equine her-

45

pesvirus type 1 (EHV-1) Infektion beim Pferh: neurologische Symptomatik bei einer Warmblutstute mit akutem todlichen Verlauf. Molekulare Charakterisierung des Gehirnisolates und pathologische korrelate. Berl. Munch. Tierarztl Wschr. 100, 147-152. Mehta, A., Maggioncalda, T., Bagasra, O., Thikkavarapu, S., Saikumari, P., Valyi-Nagi, T., Fraser, N.W. and Block, T.M. (1995) In situ DNA PCR and RNA hybridization detection of herpes virus simplex sequences in trigeminal ganglia of latently infected mice. Virology 206, 633-640. Patel, J.R., Edington, N. and Mumford, J.A. (1982) Variation in cellular tropism between isolates of equine herpesvirus-1 in foals. Arch. Virol. 74, 41-51. Ramakrishnan, R., Levine, M. and Fink, D.J. (1994) PCRbased analysis of herpes simplex virus type 1 latency in the rat trigeminal ganglion established with a ribonucleotide reductase-deficient mutant. J. Virol. 68, 7083-7091. Sharma, P.C., Cullinane, A.A., Onions, D.E. and Nicolson, L. 0992) Diagnosis of equid herpesvirus-I and -4 by polymerase chain reaction. Equine Vet. J. 24, 20-25. Slater, J.D., Borchers, K., Thackary, A.M. and Field, H.J. (1994) The trigeminal ganglion is a location for equine herpesvirus 1 latency and reactivation in the horse. J. Gen. Virol. 75, 2007-2016. Slater, J.D., Gibson, J.S. and Field, H.J. (1993) Pathogenicity of a thymidine kinase-deficient mutant of equine herpesvirus type 1 in mice and specific pathogen-free foals. J. Gen. Virol. 74, 819-828. Smith, K.C., Whitweli, K.E., Binns, M.M., Dolby, C.A., Hannant, D. and Mumford, J.A. (1992) Abortion of virologically negative foetuses following experimental challenge of pregnant pony mares with equid herpesvirus 1. Equine Vet. J. 24, 256-259. Telford, E.A.R., Watson, M.S., McBride, K. and Davison, A.J. (1992) The DNA sequence of equine herpesvirus-1. Virology 189, 304-316. Welch, H.M., Bridges, C.G., Lyon, A.M., Gritfiths, L. and Edington, N. (1992) Latent equid herpesvirus 1 and 4: detection and distribution using the polymerase chain reaction and co-cultivation from lymphoid tissues. J. Gen. Virol. 73, 261-268.