Detection of West Nile virus in the tissues of specific pathogen free chickens and serological response to laboratory infection: a comparative study

Abstract

Using an isolate of West Nile virus (WNV) from lineage 1 (Goose/Israel 1998), groups of specific pathogen free chickens were experimentally infected via the subcutaneous or intravenous routes. To evaluate the relative efficiency of detecting the virus in the infected chickens, samples from a range of tissues and organs were examined by virus isolation tests in tissue culture, including Vero, primary chicken embryo liver and fibroblast cells, and polymerase chain reaction (PCR) analyses. Additionally, in order to investigate the serological response of the chickens and produce WNV monospecific antibodies, serum samples were collected from the birds during the trial and analysed for antibodies by virus neutralization (VN) and the plaque-reduction neutralization test (PRNT). No clinical signs or gross pathological changes were seen in any of the inoculated chickens throughout the study. The nested PCR used in the study appeared to be significantly more sensitive at detecting the presence of the virus in both the tissues and the inoculated Vero cell cultures compared with the detection of gross cytopathic changes as observed in infected Vero cell culture. No cytopathic changes were seen in the inoculated avian cell cultures. Following primary inoculation of the chickens there was a weak antibody response 15 days post-inoculation. However, following re-inoculation with inactivated WNV and adjuvant there was a substantial increase in the neutralizing antibody titres when tested 2 weeks later. The results obtained suggested that the PRNT was more sensitive than the conventional VN test. Based on detection of virus and serology there was no evidence of viral transmission to the close contact controls. It can be concluded that the PCR used in this study was more sensitive than virus isolation for the detection of WNV while the PRNT also appeared more sensitive than the conventional VN test.

Détection du virus West Nile dans les tissus de poulets SPF et réponse sérologique à une infection expérimentale: étude comparative

Des groupes de poulets exempts de microorganismes pathogènes spécifiés (SPF) ont été infectés expérimentalement par voies sous cutanée ou intraveineuse avec un isolat de virus West Nile (WNV) de lignage 1 (Goose/Israel 1998). Pour étudier l'efficacité relative de la détection du virus chez les poulets infectés, des échantillons de différents tissus et organes ont fait l'objet d'isolement de virus sur culture cellulaire, incluant les cellules Vero, des cultures primaires de foie d'embryon de poulet et des fibroblastes, ainsi que des réactions de polymérisation en chaîne (PCR). De plus, dans le but d'investiguer la réponse sérologique des poulets et de produire des anticorps monospécifiques anti WNV, des échantillons de sérum ont été prélevés chez les poulets durant l'essai pour détecter les anticorps par les tests de séroneutralisation (VN) et de neutralisation de plaque (PRNT). Aucun symptôme ni lésion pathologique macroscopique n'ont été observés sur aucun des poulets inoculés, au cours de l'expérimentation. La PCR nichée utilisée dans l'étude est apparue être significativement plus sensible pour la détection du virus à la fois dans les tissus et sur les cellules Vero inoculées comparée à la détection des lésions cytopathogènes observées dans les cellules Vero infectées. Aucun effet cytopthogène n'a été observé dans les cultures cellulaires aviaires. Il a été enregistré une faible réponse en anticorps quinze jours après la première inoculation des poulets, cependant, après la ré-inoculation avec le WNV inactivé adjuvé, il y a eu une augmentation substantielle des titres des anticorps neutralisants lors de la prise de sang effectuée deux semaines plus tard. Les résultats obtenus suggèrent que le PRNT a été plus sensible que le test VN conventionnel.

Il n'a pas été mis en évidence de transmission virale aux animaux témoins, contact proche, sur la base de la détection du virus et de la sérologie. Il peut être conclu que la PCR utilisée dans cette étude est plus sensible que l'isolement du virus pour la détection du WNV, et le PRNT est apparu également plus sensible que le test VN conventionnel.

Nachweis des West-Nil-Virus in Geweben von SPF-Hühnern sowie der serologischen Immunantwort nach experimenteller Infektion: Eine Vergleichsstudie

Verschiedene Gruppen von SPF-Hühnern wurden subkutan oder intravenös mit einem Isolat des West-Nil-Virus (WNV) aus der Linie 1 (Gans/Israel 1998) infiziert. Um die Effektivität des Virusnachweises aus infizierten Hühnern zu ermitteln, wurden Proben von einer Reihe von Geweben und Organen mittels Virusisolierungsversuchen in Zellkulturen wie Vero- primäre Hühnerembryoleberzellen und Hühnerembryfibroblasten sowie Polymerasekettenreaktion (PCR)-Analysen untersucht. Außerdem wurden zur Überprüfung der serologischen Immunantwort und zur Gewinnung monospezifischer WNV-Antikörper während des Versuchs von den Hühnern Serumproben gesammelt und auf Antikörper im Virusneutralisations (VN)- und Plaquereduktions (PRNT)-Tests analysiert. In keinem der inokulierten Tiere wurde im Verlauf des Versuchs klinische Symptome oder pathologisch-anatomische Veränderungen festgestellt. Verglichen mit dem Nachweis von zytopathischen Effekten in infizierten Verozellkulturen erschien die verwendete nested PCR bei der Virusermittlung sowohl in den Geweben als auch in den inokulierten Verozellen signifikant sensitiver. In den inokulierten aviären Zellkulturen konnte keine zytopathischen Veränderungen beobachtet werden. Nach der Erstinokulation der Hühner trat 15 Tage später eine geringgradige Immunantwort auf. Zwei Wochen nach der Reinokulation mit inaktiviertem WNV und Adjuvans konnte jedoch ein deutlicher Anstieg der neutralisierenden Antikörpertiter festgestellt werden. Dabei weisen die Ergebnisse darauf hin, dass der PRNT sensitiver als der konventionelle VN-Test ist. Virusisolierungsversuche und Serologie erbrachten keinen Hinweis auf eine Virusübertragung auf benachbarte Kontakttiere. Aus den Ergebnissen kann geschlossen werden, dass die in dieser Studie verwendete PCR für den Nachweis von WNV sensitiver ist als die Virusisolierung mittels Zellkultur und auch der PRNT ein empfindlicherer Test zu sein scheint als die konventionelle Virusneutralisation.

Detección de Virus West Nile en tejidos de pollos SPF y respuesta serológica a la infección en laboratorio: un estudio comparativo

Se infectaron grupos de pollos libres de patógenos específicos (SPF) vía subcutánea o intravenosa con un aislado de virus West Nile (WNV) de la línea 1 (Goose/Israel 1998). Para evaluar la eficiencia relativa de detección del virus en pollos infectados, se examinaron muestras de distintos tejidos y órganos mediante aislamiento vírico en cultivos celulares, incluyendo células Vero, y cultivos primarios de células de hígado y fibroblastos de embrión de pollo, y la técnica de reacción en cadena de la polimerasa (PCR). Además, para investigar la respuesta serológica de los pollos y producir anticuerpos monoespecíficos frente a WNV, se obtuvieron muestras de suero de aves durante la infección y se analizó el nivel de anticuerpos mediante neutralización vírica (VN) y ensayos de neutralización por reducción de placas (PRNT). No se observaron signos clínicos ni lesiones macroscópicas en ninguno de los pollos inoculados durante todo el estudio. La PCR anidada que se utilizó en el estudio fue significativamente más sensible para la detección de la presencia de virus tanto en los tejidos como en los cultivos celulares de células Vero en comparación con la detección de cambios citopáticos en cultivos de células Vero. No se observaron cambios citopáticos en los cultivos de células aviares inoculados. Tras la inoculación primaria de los pollos hubo una respuesta débil de anticuerpos a los 15 días post infección, pero tras la reinoculación con WNV inactivado junto con adyuvante hubo un incremento sustancial en los títulos de anticuerpos cuando se testaron dos semanas después. Los resultados obtenidos sugirieron que la PRNT era más sensible que el ensayo de VN convencional. En base a la detección de virus y serología no hubo evidencias de transmisión vírica a las aves control en contacto estrecho. Se puede concluir que la PCR utilizada en este estudio era más sensible que el aislamiento vírico para la detección de WNV mientras que la PRNT también pareció más sensible que la VN convencional.

Introduction

West Nile virus (WNV) is an arthropod-borne flavivirus primarily maintained in natural transmission cycles between mosquito vectors and birds. The virus may also infect humans, horses and other animals causing a range of symptoms from fever to neurological signs including severe meningoencephalomyelitis and at least 150 bird species and 30 other vertebrates have been reported to be susceptible to infection with WNV (van der Meulen et al., 2005). In mammals, a transient viraemia of low virus titre precedes clinical signs and they are generally considered “dead end” hosts (Komar, 2000; Ostlund et al., 2001). Sporadic outbreaks of disease in humans and horses have been regularly reported throughout the middle east, Asia and eastern and southern Europe (George et al., 1984; Murgue et al., 2001; Weinberger et al., 2001). In 1996 an epidemic in Romania involved hundreds of human cases with 17 fatalities, and recent serological surveys have detected WNV antibodies in wild birds in Poland, the Czech Republic and the United Kingdom (Buckley et al., 2003, 2006). The disease has reached epidemic proportions throughout North America following its introduction to New York in 1999 (Nash et al., 2001) and, although severe disease in humans is only seen in about 1% of infections, 564 deaths were attributed to WNV infection in the USA between 1999 and 2003 (Briese & Bernard, 2005). WNV infections in wild birds are useful risk indicators for human infection, and therefore diagnosis and avian disease surveillance is extremely important (Steele et al., 2000). A range of tests have been developed for the laboratory detection of WNV including direct methods of virus isolation, reverse transcription-polymerase chain reaction (RT-PCR)and real-time RT-PCR as well as indirect serological methods such as virus neutralization (VN), the plaque reduction neutralization test (PRNT), IgM capture enzyme-linked immunosorbent assay (MAC-ELISA), IgG enzyme-linked immunosorbent assay and immunohistochemistry.

The aim of this study was to evaluate the efficiency of detection of virus in the tissues of specific pathogen free chickens infected with a lineage 1 WNV strain (Goose Israel 1998). This was carried out using virus isolation procedures in primary avian cell cultures and a mammalian cell line, and by comparing the results obtained with molecular detection methods. The serological response of the infected chickens was also investigated using VN and PRNT.

Materials and Methods

West Nile virus

The reference strain of WNV used in this study was Goose Israel 1998 (tissue culture passage 5), a lineage 1 strain of the virus and phylogenetically very similar to the strain of virus that appeared in North America in 1999 and recent European strains. Virus propagation was undertaken in Vero C1008 (ATCC) cells that were cultivated at 37°C in Eagle's minimal essential medium containing 10% foetal bovine serum in a 5% CO₂ in air atmosphere according to ATCC recommendations.

Experimental design

Fourteen 3-week-old specific pathogen free White Leghorn chickens were randomly assigned to two groups of five birds and one group of four birds. The birds were housed within a single isolator at ACDP Category 3 within a high biosecure facility for the period of the study; commercial feed and fresh water were provided ad libitum. The groups of chickens were inoculated with WNV as follows; one group of five birds (Group 1) received 10^4.0 median tissue culture infectious doses in 0.2 ml subcutaneously (s.c.) in the base of the neck. One group of five birds (Group 2) received 10^4.0 median tissue culture infectious doses of WNV in 0.2 ml intravenously (i.v.) in the brachial vein. The group of four birds (Group 3) remained uninoculated yet in direct contact with the inoculated birds. The birds were monitored daily for clinical signs during the course of the study. Prior to inoculation and at 3 and 15 days post-infection (p.i.), 1 ml blood was taken from the brachial vein of each bird, the blood sample was allowed to clot at room temperature and serum was separated for serological tests. At 3 days p.i., one bird from each group was killed and tissue samples including the brain, blood, kidney, trachea, heart, liver, spleen, caecum, bursa, skin and lung were collected aseptically from each bird using separate sets of instruments for each individual organ for virological investigations as follows: RT-PCR examination together with virus isolation in Vero C1008 cells and chicken embryo liver and fibroblast cells. At 21 days p.i., the remaining birds in Groups 1 and 2 were inoculated intra-muscularly in the thigh muscle with 0.5 ml inactivated WNV in an oil emulsion in order to prepare hyperimmune WNV antiserum. The oil-emulsion vaccine was prepared using 20% WNV in a mineral oil adjuvant, which prior to inactivation had a titre in Vero cell cultures of 10^5.8 median tissue culture infective doses per millilitre.

At 2 and 3 weeks after re-inoculation blood samples were taken from the brachial vein of each bird, and at 3 weeks the birds were killed and bled out. Serum was removed from the blood clot, centrifuged, heat inactivated at 56°C for 30 min and stored at −70°C prior to testing. The sera, together with reference WNV-positive and WNV-negative control sera (USDA, Ames, Iowa, USA), were individually tested by the VN test, and pools of sera with similar neutralizing titres were subsequently tested by the PRNT.

Cell cultures

Fresh confluent monolayers of Vero C1008 cell cultures were used for attempted virus isolation as well as chicken embryonic liver (CEL) and chicken embryonic fibroblast (CEF) cultures from specific pathogen free chicken embryos prepared and inoculated as described elsewhere (Gough et al., 1988).

Virus isolation on tissues

Twenty per cent (w/v) suspensions of tissue homogenate were made in antibiotic phosphate-buffered saline (pH 7.2) and allowed to stand at room temperature for 1 h. Following centrifugation to clarify the suspensions, supernatants were filtered through a disposable 0.45 µm filter and inoculated onto fresh confluent Vero, CEL and CEF cell monolayers in 25 ml flasks. The cultures were incubated for 1 h at 37°C and then overlaid with 8 to 10 ml fresh maintenance medium warmed to 37°C. Cultures were incubated in a 5% CO₂ in air atmosphere at 37°C, examined daily for cytopathic effects (CPE) using an inverted microscope and frozen at −70°C after 7 days. Cultures in which a CPE was observed were frozen at −70°C and subsequently tested by RT-PCR for the presence of WNV. Cultures in which no CPE was detected were frozen and thawed, and clarified by centrifugation. The supernatants were passaged once more onto similar cell cultures by the same procedure and again monitored daily for CPE for a further 7 days. All the Vero cultures in which no CPE was observed were further tested by RT-PCR for the presence of WNV.

Reverse transcription-polymerase chain reaction

The RT-PCR used in this study employed segment specific primers directed at the C-terminal of the C gene and the N-terminal of the prM gene (nucleotides 233 to 640). Oligonucleotides included forward primer 5′-TTGTGTTGGCTCTCTTGGCGTTCTT-3′ and reverse primer 5′-CAGCCGACAGCACTGGACATTCATA-3′. This first-round one-tube reaction was followed by a second (nested) PCR using forward primer 5′-CAGTGCTGGATCGATGGAGAGG-3′ and reverse primer 5′-CCGCCGATTGATAGCACTGGT-3′ (nucleotides 287 to 390) to produce a WNV-specific 104 base pair product (Shi et al., 2001).

First-round RT-PCR

The master mix was prepared on ice using the QIAGEN OneStep RT-PCR kit, and 5 µl sample RNA was added to 20 µl master mix in 200 µl reaction tubes. The following thermal cycling conditions were used: 50°C for 30 min, 95°C for 15 min, 35 cycles of 94°C for 45 sec, 56°C for 45 sec and 72°C for 1 min, followed by a single step of 72°C for 10 min. Positive control reactions using RNA derived from Goose Israel 1998 WNV together with negative controls using RNase-free water instead of RNA template were included in each run.

Second-round nested PCR

The master mix was prepared on ice using the QIAGEN Taq PCR Core kit, and 0.2 µl first-round product was added to 24.8 µl master mix in 200 µl reaction tubes. The following thermal cycling conditions were used: 95°C for 3 min, 25 cycles of 94°C for 45 sec, 58°C for 45 sec and 72°C for 1 min, followed by a single step of 72°C for 10 min. The PCR product was visualized over ultraviolet light following electrophoresis in 2% agarose gel incorporating ethidium bromide.

Serology

Virus neutralization

Sera were tested by the “diluted sera/constant virus” method in 96-well micro-neutralization plates using two-fold dilutions of sera in modified Eagle's minimal essential medium (MEM) from 1/2 to 1/256 and an equal volume of MEM containing 100 median tissue culture infective doses of WNV. Following a virus-serum reaction time of 1 h at 37°C in a 5% CO₂ in air atmosphere, 25 µl volumes of the virus–serum mixtures were inoculated into wells containing fresh confluent monolayers of Vero C1008 cells. Following gentle rocking of the plates to spread the inoculum over the monolayer, the cultures were incubated at 37°C for 1 h in a 5% CO₂ in air atmosphere to allow adsorption to the monolayer of any virus that had not been neutralized by the WNV antibodies. After adsorption, the cultures were overlaid with 200 µl Vero maintenance medium, sealed and incubated at 37°C in a 5% CO₂ in air atmosphere. The Vero cell monolayers were examined microscopically after 4 to 5 days for CPE and the virus neutralizing titre of the serum calculated as the reciprocal of the highest serum dilution that completely neutralized the virus.

Plaque reduction neutralization test

Test serum was serially diluted in MEM using doubling dilutions from 1/10 to 1/2560 to produce 125 µl volumes in sterile 96-well microplates. WNV diluted in MEM to contain 60 to 70 plaque-forming units/125 µl was added to the serial dilutions of test serum, the plates were sealed and incubated for 1 h at 37°C in a 5% CO₂ in air atmosphere to allow for virus neutralization. Fresh confluent Vero cell cultures in 12-well microdishes were washed once with MEM, and the serum–virus mixture was then added to each of 10 wells of the plate and incubated for 1 h at 37°C in a 5% CO₂ in air atmosphere. A virus-only control well containing 60 or 70 plaque-forming units of virus in 250 µl MEM and a negative control well containing 250 µl MEM only were also included on each plate. Following incubation, 1 ml of 50% carboxymethyl cellulose/50% double strength MEM overlay was added to each well and the sealed plates incubated at 37°C in a 5% CO₂ in air atmosphere. After 4 days incubation, 1 ml of 10% neutral buffered formalin was added to each well and the plates fixed for at least 3 h. Each well of the plates was then washed in cold running tap water to remove the overlay and fixative, the plates were blotted dry and 250 µl of 0.1% crystal violet in phosphate-buffered saline (pH 7.2) added to each well to stain the Vero monolayers and visualize the plaques formed within. The plates were then air dried and the plaques counted in each well. The PRNT titre was calculated as the reciprocal of the dilution of test serum that produced a 90% reduction in the number of plaques formed in the virus control well (PRNT⁹⁰). If plaques were seen in the negative control well, the test was deemed invalid.

Results

Clinical and postmortem findings

All the birds remained clinically normal throughout the period of the study and no gross pathological lesions were seen in the chickens from which samples were taken.

Virus isolation

Four days after inoculation a widespread CPE was seen in the Vero cultures inoculated with samples of kidney and spleen from the s.c.-inoculated chicken and from the kidney of the i.v.-inoculated chicken. The CPE was characterized by the appearance of rounded, refractile cells that progressed to complete destruction of the cell monolayer. On passage, no additional cultures produced a CPE following inoculation with the other tissues examined. The cultures inoculated with samples from the uninoculated contact bird remained normal through two passages. No CPE was detected in any of the CEL or CEF cultures following two passages.

RT-PCR on tissues and cell cultures

PCR amplicons of the predicted size (104 base pairs) were detected in RT-PCR undertaken on RNA extracted from the kidney, spleen, caecum, heart, trachea and lung but not from the brain, liver, skin, bursa and blood of the i.v.-infected chicken (Table 1). Amplicons were also detected in RT-PCR undertaken on RNA extracted from the skin, spleen, kidney, lung and heart from the s.c.-infected chicken but not from the caecum, trachea, brain, liver, bursa and blood. Amplicons of strong intensity were detected in RNA samples derived from heart muscle of both infected birds. No product was amplified from RNA extracted from the tissues of the uninfected in-contact birds. PCR amplicons of the predicted size were also detected in RNA extracts from the infected Vero cultures in which a CPE was seen (Table 1). The CEL and CEF cultures were not available for examination by PCR.

Table 1.  Results of virus isolation (VI) in Vero cells, RT-PCR on RNA extracted from chicken tissues collected 3 days p.i. with WNV and from VI cell culture fluids

	Control birds			s.c.-infected birds			i.v.-infected birds
Tissue	VI	RT-PCR on VI	RT-PCR on tissue	VI	RT-PCR on VI	RT-PCR on tissue	VI	RT-PCR on VI	RT-PCR on tissue
Skin	−	−	−	−	+	+	−	−	−
Bursa	−	−	−	−	−	−	−	−	−
Blood	−	−	−	−	−	−	−	−	−
Caecum	−	−	−	−	−	−	−	+	+
Trachea	−	−	−	−	−	−	−	+	+
Spleen	−	−	−	−	+	+	CPE	+	+
Kidney	−	−	−	CPE	+	+	CPE	+	+
Liver	−	−	−	−	−	−	−	−	−
Lung	−	−	−	−	+	+	−	+	+
Heart	−	−	−	−	+	+	−	+	+
Brain	−	−	−	−	−	−	−	−	−
−, negative;+, positive; ^abcCPE, cytopathic effects seen in Vero C1008 cell monolayers.

Serological results

The results are presented in Table 2. No neutralizing antibody was detected in any of the serum samples 3 days after inoculation. By 15 days, three of the four s.c.-inoculated birds had titres of 2 and one bird a titre of <2; in the i.v.-inoculated group, three birds had titres of 2 and one bird a titre of 8. Two weeks after re-inoculation with inactivated virus (35 days), the birds showed an enhanced antibody response with antibody titres ranging from 16 to 64 in the s.c. group and from 8 to 128 in the i.v. group. When sampled and tested 1 week later (42 days) the antibody titres in the birds showed a slight increase. No WNV antibody was detected in any of the serum samples from the in-contact control birds.

Table 2.  VN titres^a of WNV-infected chickens and in-contact controls

Group and ID number	3 days p.i.	15 days p.i.	35 days p.i.^b	42 days p.i.
Control
434	<2	<2	<2	<2
435	<2	<2	<2	Not tested
436	<2^c	–	–	–
437	<2	<2	<2	<2
s.c.-inoculated
1	<2	<2	32^c	Not tested
2	<2	2	16	32
3	<2	2	16	32
4	<2	2	64	128
5	<2^c	–	–	–
i.v.-inoculated
486	<2	8	128	128
487	<2	2	8	16
488	<2	2	8^c	–
489	<2	2	16	16
490	<2^c	–	–	–
^aTitres expressed as the reciprocal of the highest serum dilution that completely neutralized the virus. A titre <2 was considered negative. ^bTwo weeks post-revaccination. ^cKilled.

Plaque reduction neutralization

The results are presented in Table 3. As can be seen, no PRNT antibodies were detected in pool 1 sera, which was from the contact control chickens. Pools 2 and 3 had titres of 640 and 1280, respectively, and the USDA positive control serum had a titre of 640. No antibodies were detected in the contact control or USDA negative sera.

Table 3.  VN and PRNT⁹⁰ titres of serum pools 1, 2 and 3 together with USDA PRNT reference positive and negative sera

Serum	VN titre^a	PRNT^90 titre^b
Pool 1	<2 (negative)	<10 (negative)
Pool 2	64	640
Pool 3	128	1280
USDA positive	64	640
USDA negative	<2	<10
^aVN titres expressed as the reciprocal of the highest serum dilution that completely neutralized the virus. ^bPRNT^90 calculated as the reciprocal of the dilution of test serum that produced a 90% reduction in the number of plaques formed in the virus control well. For both values, a titre <2 was considered negative.

Discussion

Severe disease and gross tissue lesions described for both experimental and natural WNV infection of other bird species (Steele et al., 2000; Swayne et al., 2001; Malkinson et al., 2002; Komar et al., 2003) were not seen following infection of chickens in the study described here. Senne et al. (2000) obtained similar results in 7-week-old chickens when they used a strain of WNV isolated from a crow in New York, USA. They also carried out histopathological examination on a selection of tissues from the infected birds, and microscopic changes particularly in the heart and kidney, but also in the brain of one chicken, were observed 21 days after inoculation. Although histopathology was not carried out in the study described here, WNV-specific RNA was amplified and detected 3 days p.i. by RT-PCR from the skin, spleen, kidney, lung and heart of the s.c.-infected bird and from the caecum, trachea, spleen, kidney, lung and heart of the bird infected by the i.v. route, but was not detected in tissue from the brain, liver, bursa or blood from birds infected by either route.

Substantial CPE was seen only in Vero cell cultures following the first passage of kidney tissue homogenate from the s.c.-infected bird and both the kidney and spleen tissue homogenate passage from the i.v.-infected bird. WNV-specific RNA was also detected by RT-PCR from the same Vero cell cultures as well as from skin, spleen, lung and heart tissue homogenate passage of the s.c.-infected bird and from the caecum, trachea, lung and heart tissue homogenate passage of the i.v.-infected bird. These results suggest that WNV was concentrated in spleen and kidney tissue in infected birds and are comparable with the findings of Steele et al. (2000), where kidney, spleen and heart samples from both North American and exotic birds, infected during the 1999 outbreak in New York, were consistently found to be positive by RT-PCR, virus isolation and immunohistochemistry, respectively. Virus was neither isolated in cell culture nor detected by RT-PCR in any tissues of the uninfected control birds (Table 1). In the study by Senne et al. (2000), the isolation of WNV in Vero 76 cell cultures from the plasma of all infected birds was reported from 3 to 5 days p.i. and from the spleen, kidney and lung up to 10 days p.i., but not from the brain or liver. In the study reported here no virus was detected by virus isolation or PCR in whole blood samples taken from the s.c. and i.v. birds 3 days after infection. This may have been due to the fact that only one bird from each group was sampled at this time or that the isolate of WNV used in this study was less invasive than the virus used by Senne et al. (2000). It is probable that the latter explanation is the more likely as the dose and s.c. route of inoculation used in this study were similar to those reported by Senne et al. (2000). It is interesting to note that in the present study WNV-specific RNA was detected both directly from skin tissue taken from the s.c.-infected bird and from Vero cell cultures in which skin tissue homogenate from the same bird had been passaged, but not from skin of the i.v.-infected bird. The detection of WNV in skin tissue is perhaps not surprising and concentration of virus in peripheral tissue during the relatively short period that virus is present in birds may aid transmission by arthropod vectors. There was a good correlation between the PCR-positive results obtained from the infected Vero cell cultures and the postmortem tissues, although CPE was only detected in the cultures inoculated with samples of the kidney and spleen. This may have been due to the presence of higher concentrations of virus in these tissues as reported in previous studies. The results of this study suggest that CEL and CEF cultures are not suitable for the detection of WNV as no CPE was detected following two blind passages. It is unfortunate that the inoculated cultures were not available for PCR analysis as replication of the virus may have occurred without detectable CPE, as occurred in the VERO cultures. However, in terms of comparing the sensitivity of the different assay systems, there would be no practical advantage in routinely using primary avian cells for the detection of WNV if the inoculated cultures required further PCR analysis to confirm the presence of the virus.

Virus neutralizing antibodies were first detected in infected birds only at 15 days p.i., and following intra-muscular inoculation of inactivated virus in oil emulsion at 21 days p.i. antibody titres were variable between individual birds. No virus neutralizing antibody was detected in uninfected, in-contact control birds throughout the study (Table 2).

Although the panel of positive and negative serum pools was small in number, comparison of VN and PRNT⁹⁰ titres demonstrated that PRNT⁹⁰ was 10 times more sensitive than the standard neutralization test (Table 3).

The evidence from this study suggests that the nested PCR is more sensitive than virus isolation in Vero cultures for the detection of WNV, although virus isolation may still be useful for the detection of other flaviviruses. Indeed, at the time of writing, the RT-PCR described in this study has been used to test more than 2000 samples of kidney and brain tissue collected from dead wild birds in the United Kingdom, and to date no evidence of WNV has been recorded. The results also confirm that the strategic use of sentinel chickens would be a useful tool in monitoring the presence of WNV in this country.

Acknowledgements

Thanks are due to Dr M. Malkinson, The Kimron Veterinary Institute, Beit Dagan, Israel, for supplying the Goose Israel 1998 strain of WNV.