Detection and quantitation of gallid herpesvirus 1 in avian samples by 5′ Taq nuclease assay utilizing Minor Groove Binder technology

Abstract

A 5′ Taq nuclease assay utilizing Minor Groove Binder technology and targeting the thymidine kinase gene of gallid herpesvirus 1 (GaHV-1) was designed and optimized for use in diagnosing avian infectious laryngotracheitis. The assay was specific for GaHV-1 in that it did not react with other avian viral or bacterial pathogens. The detection limit was 1.0×10⁻² median tissue culture infectious dose per reaction or 90 target copies per reaction. Fifteen out of 41 diagnostic samples from sick birds reacted in the assay, five of which produced a typical alphaherpesvirus cytopathic effect (CPE) on chicken kidney (CK) cells. Sequencing, using amplicons generated by a polymerase chain reaction with primers flanking the 5′ Taq nuclease amplicon, confirmed the presence of GaHV-1 in six samples (two producing alphaherpesvirus CPE on CK cells, three not producing alphaherpesvirus CPE, and one that was not inoculated onto CK cells). Tracheal swabs taken from 18 healthy broilers did not react in the assay. The ability of the assay to determine viral load in samples was demonstrated. Overall the assay is suitable for the rapid diagnosis of infectious laryngotracheitis.

Introduction

Infectious laryngotracheitis (ILT) is caused by gallid herpesvirus 1 (GaHV-1) (Guy & Garcia, 2008), a member of the Herpesviridae, subfamily Alphaherpesvirinae (Davison et al., 2005). The virus is commonly known as infectious laryngotracheitis virus (Davison et al., 2005).

The virology, pathology, diagnosis and management of ILT were reviewed by Guy & Garcia (2008). ILT presents as a mild to severe contagious respiratory disease characterized, in severe cases, by marked dyspnoea, expectoration of blood-stained mucous and high mortality. Clinical signs may resemble those produced by other acute respiratory diseases including avian influenza, infectious bronchitis and Newcastle disease; hence the capacity to rapidly and reliably differentiate between these infections is required. Outbreaks of ILT in poultry are contained using vaccination followed by disposal of recovered flocks under quarantine. Premises need to be cleaned and disinfected before restocking.

ILT is frequently diagnosed on the basis of histopathology, namely the presence of intranuclear inclusion bodies in respiratory and conjunctival epithelium, and virus isolation in cell culture or embryonated eggs (Tripathy, 1998). Virus isolation can take several weeks, whereas histopathology is generally available within 24 to 48 h using conventional processing. However, diagnosis of ILT based on demonstration of inclusion bodies is considerably less sensitive than virus isolation (Guy & Garcia, 2008).

Several polymerase chain reaction (PCR) assays have been developed for rapid detection of GaHV-1 in poultry (Williams et al., 1992, 1994; Scholz et al., 1994; Abbas & Andreasen, 1996; Abbas et al., 1996; Alexander & Nagy, 1997; Humberd et al., 2002; Pang et al., 2002). All of these assays require the use of ethidium bromide-stained agarose gels and/or DNA hybridization to demonstrate the presence of the expected product after amplification.

Real-time PCR assays using intercalating dyes or dual-labelled probes such as the 5′ Taq nuclease system do not require the use of gel electrophoresis, and offer enhanced sensitivity and reduced susceptibility to cross-contamination compared with conventional PCR (Higuchi et al., 1993; Mackay et al., 2002). However, the sensitivity of intercalating dye assays may be limited by their reduced ability to detect low concentrations of template compared with dual-labelled probe systems (Mackay et al., 2002). Furthermore, as the intercalating dyes bind non-specifically to all double-stranded DNA, melting curve analysis is required to differentiate specific from non-specific products (Mackay et al., 2002).

Real-time PCR assays have recently been described for GaHV-1 detection. Creelan et al. (2006) described a real-time PCR using the intercalating dye Sybr Green followed by melting curve analysis to demonstrate specific amplification of GaHV-1 sequences from tissues taken from experimentally infected birds. Callison et al. (2007) described a 5′ Taq nuclease assay that allowed rapid and sensitive detection of GaHV-1 in samples from experimentally and naturally infected birds. This assay detected virus in significantly more samples than did virus isolation or immunofluorescence, which was attributed to the high analytical sensitivity of the 5′ Taq nuclease assay relative to other methods.

Minor Groove Binder (MGB) dual-labelled probes have several advantages over more conventional dual-labelled probes, such as shorter length, increased specificity, lower background fluorescence and greater dynamic range (Kutyavin et al., 2000). In this paper we report the development and validation of a 5′ Taq nuclease assay for GaHV-1 detection and quantitation utilizing MGB technology.

Materials and Methods

Viruses and bacteria

DNA prepared from the chicken embryo-origin Australian SA2 vaccine strain of GaHV-1 (Fort Dodge Animal Health, Baulkham Hills, Australia) and 15 field strains of GaHV-1, one field strain of fowlpox virus and two field strains of avian adenovirus were used in this study. DNA was extracted from the field strains of GaHV-1 using the method described by Corney et al. (2008). For the other viruses, DNA preparations from previous studies were used as shown in Table 1.

Table 1.  Origins of DNA preparations used in assessing the analytical specificity of the GaHV-1 5′ Taq nuclease assay

Virus or bacterial species	Strain	Source of DNA (culture)
Avian viruses
GaHV-1	SA2	Corney et al. (2007)
GaHV-1	Field strains (n = 15)	Present study (BSL)
Avian adenovirus	Field strains (n = 2)	Corney et al. (2007)
Fowlpox virus	B3666/1	Corney et al. (2007)
Bacteria
Alcaligenes faecalis	ATCC 8750^T	Corney et al. (2007)
Avibacterium avium	ATCC 29546^T	Corney et al. (2007)
Avibacterium paragallinarum	ATCC 29545^T	Corney et al. (2007)
Avibacterium volantium	NCTC 3438^T	Corney et al. (2007)
Avibacterium species A	HIM 790-5^T	Corney et al. (2007)
Bordetella avium	JPDH 591-77^T	Corney et al. (2007)
Bordetella bronchiseptica	ATCC 19395^T	Corney et al. (2007)
Bordetella hinzii	TC 58^T	Corney et al. (2007)
Coenonia anatine	CCUG 46148^T	Corney et al. (2007)
Escherichia coli	92-129302/1	Corney et al. (2007)
Ornithobacterium rhinotracheale	B-3263/91	Corney et al. (2007)
Pasteurella multocida subspecies gallicida	NCTC 10204^T	Corney et al. (2007)
P. multocida subspecies multocida	NCTC 10322^T	Corney et al. (2007)
P. multocida subspecies septica	CIP A215^T	Corney et al. (2007)
BSL, Biosecurity Sciences Laboratory, Yeerongpilly, Australia;^T, type strain.

DNA preparations from 15 bacterial species (Table 1) were also used in this study.

Samples

Forty-one avian samples (comprising 34 fresh tissues and swabs from respiratory tract and lungs, two cloacal swabs, one unidentified tissue, and four livers and spleens) from 17 avian disease cases were analysed by 5′ Taq nuclease assay and by virus isolation on chicken kidney (CK) cells. Most of the samples were from poultry with respiratory signs and came from a mixture of commercial and backyard meat and layer flocks. Only one flock (a commercial layer flock, which was the subject of two cases) was known to have been vaccinated for ILT.

Tracheal swabs were also collected from 18 healthy broilers that were obtained from a commercial poultry farm.

For the 5′ Taq nuclease assay, swabs were vortexed for 15 sec in up to 3 ml virus transport medium (VTM). The VTM consisted of sucrose, 0.218 M; KH2P04, 0.0038 M; K2HP04, 0.0072 M; sodium glutamate, 0.0049 M; bovine serum albumin 1% w/v (SPGA) (Bovarnick et al., 1950) supplemented with 3×10² IU/ml benzylpenicillin sodium salt (CSL, Parkville, Australia), 0.3 mg/ml streptomycin sulphate (Sigma-Aldrich, Castle Hill, Australia) and 0.006 mg/ml amphotericin B (Sigma-Aldrich). DNA was extracted from 200 µl VTM using a QIAamp DNA Mini Kit (QIAGEN, Doncaster, Australia) as per the manufacturer's instructions. Tissue samples were ground in 1 to 2 ml VTM to a concentration of approximately 10% using a mortar and pestle. The suspension was centrifuged for 5 min at 900×g, and DNA was extracted from 200 µl supernatant using the QIAamp DNA Mini Kit. An elution buffer volume of 50 µl (Buffer AE; QIAGEN) was used in all of these extraction procedures.

Alternatively, DNA was extracted from tracheal samples (using approximately 8 mm³ of fresh trachea) on a KingFisher-96 Particle Processor (Thermo Electron Corporation, Vantaa, Finland) using an Ambion MagMAX-96 Viral RNA Isolation Kit (Applied Biosystems, Scoresby, Australia) as per the manufacturer's instructions with the following modifications. The tissue was added directly to 130 µl MagMAX Lysis/Binding solution for extraction, and DNA was eluted from the MagMAX beads using 5 min agitation at 75°C in 50 µl MagMAX Elution Buffer. The beads were air-dried for 2 min before the elution step.

For virus isolation, tissues were ground in VTM to give a 10% suspension that was clarified by centrifugation for 5 min at 900×g. Swabs were vortexed in VTM as described for template preparation. Swab and tissue suspensions were centrifuged for 10 min at 2900×g, and 0.1 ml was inoculated onto CK cells in 24-well Costar cluster plates (Corning Incorporated, Lowell, USA). Cell cultures were incubated for 7 days at 37°C in a 5% carbon dioxide atmosphere and were checked for cytopathic effect (CPE) every 48 h. If no CPE was observed within 7 days, the monolayer was scraped into the medium and 0.1 ml was transferred to a fresh CK culture. If no CPE was observed on this passage within 7 days, the sample was scored negative.

5′ Taq nuclease assay design and optimization

The thymidine kinase (TK) gene was selected as a potential target for a GaHV-1 5′ Taq nuclease assay. A multiple sequence alignment incorporating TK sequences from five GaHV-1 strains was constructed using Pileup (Accelerys GCG, San Diego, California, USA) on the Australian National Genomic Information Service website (http://www.angis.org.au). The GenBank accession numbers (Benson et al., 2000) of the sequences were AF435453, L36139, S83714, Y14300 and D00565. The sequences represented strains from diverse locations including China, the Netherlands (Ziemann et al., 1998) and the USA (Keeler et al., 1991), and were the only GaHV-1 TK sequences available on GenBank at the time. The region of the TK gene represented by Y14300, the shortest of the sequences, was 214 base pairs (bp) and was 100% conserved among the five sequences. A 5′ Taq nuclease assay utilizing MGB technology was designed within this region using Primer Express (Applied Biosystems). The primers and probe were: forward primer TK1 f, 5′-AAAAACTCGCGACGGTATTGA-3′ (positions 1068 to 1088 on S83714); reverse primer TK1 r, 5′-TGAGGCCATGTGCTGGTAAG-3′ (positions 1108 to 1127 on S83714); and probe TK1a MGB, 5′-6FAM-AAACTTGTGYATTTTTT-MGBNFQ-3′ (positions 1090 to 1106 on S83714) which produce a 60 bp amplicon. The degeneracy at position 10 of the probe was to account for a polymorphism identified among Australian GaHV-1 strains (data not shown). A search of GenBank using the Basic Local Alignment Tool (BLAST) (Altschul et al., 1997) on the National Center for Biotechnology Information website (http://www.ncbi.nlm.nih.gov) found no similarities between the amplicon sequence and non-GaHV-1 targets.

Primer and probe concentrations were optimized by checkerboard titrations on a RotorGene 3000 (Corbett Research, Mortlake, Australia) using the same cycling conditions (15 sec at 95°C, 60 sec at 60°C for 50 cycles) as utilized by Corney et al. (2007, 2008), a normalized fluorescence threshold value of 0.1, and DNA from the SA2 vaccine or the Australian field strain 04-128794/1 (Biosecurity Sciences Laboratory, Yeerongpilly, Australia) as the template. The reactions (25 µl final volume) contained 2 µl template and used the RealMasterMix™ Probe mixture (Eppendorf, Hamburg, Germany). A primer concentration of 1.2 µM per primer and a probe concentration of 0.2 µM gave the lowest threshold cycle (C_T) values combined with high normalized fluorescence, and were selected for routine use in the 5′ Taq nuclease assay. All assays included positive control (DNA extracted from SA2) and template-free negative control reactions.

GaHV-1 TKseq PCR design and optimization

To permit sequencing of the 5′ Taq nuclease amplicon and neighbouring DNA, primers flanking the 5′ Taq nuclease amplicon were designed using the sequence alignment generated in the design of the 5′ Taq nuclease assay. The primers were: forward primer TKseq f, 5′-ACCTACCTCCAACGTACATC-3′ (positions 854 to 873 on S83714); and reverse primer TKseq r, 5′-CCCATATCAGCATTCTAGCG-3′ (positions 1248 to 1229 on S83714). These primers produce a 395 bp amplicon. A BLAST search of GenBank found no significant similarities between the primer sequences and non-GaHV-1 targets. Optimum PCR conditions were determined by testing annealing temperatures ranging from 50 to 65°C and by checkerboard titration for Taq polymerase (AmpliTaq Gold; Applied Biosystems) and magnesium chloride concentration in a PCR Express gradient thermal cycler (Thermo Hybaid, Middlesex, UK) using SA2 DNA as the template. Other reaction conditions were as described for AmpliTaq Gold by the manufacturer. The final reaction conditions were: 1 U AmpliTaq Gold in GenAmp PCR Buffer II (Applied Biosystems), 2.5 mM magnesium chloride, 0.5 µM per primer, 200 µM deoxynucleoside triphosphates (PCR Nucleotide Mix; Roche, Castle Hill, Australia) and 2 µl template in a total volume of 25 µl. Cycling conditions were 4 min at 95°C (one cycle), 45 sec at 94°C, 45 sec at 51°C, 30 sec at 72°C (35 cycles) and 10 min at 72°C (one cycle). PCR products were analysed on 1.5% agarose gels and were visualized by staining with ethidium bromide.

Analytical specificity of the 5′ Taq nuclease assay

The DNA preparations listed in Table 1 were tested in the 5′ Taq nuclease assay in duplicate. The bacterial DNA preparations were diluted in 0.1 mM Tris (pH 7) to approximately 10 ng/µl for testing in the 5′ Taq nuclease assay. The SA2 and fowlpox virus DNAs that were tested in the 5′ Taq nuclease assay corresponded to approximately 1.0×10⁴ median tissue culture infectious dose (TCID₅₀) and 1.3×10³ TCID₅₀ (Corney et al., 2007) of virus, respectively. The adenoviruses were not titrated but the presence of virus was confirmed by electron microscopy (Corney et al., 2007). The viral DNA preparations were tested without dilution.

Analytical sensitivity and dynamic range of the 5′ Taq nuclease assay

Serial dilutions of the SA2 vaccine were prepared in phosphate-buffered saline (pH 7.4), and DNA was extracted from each dilution as described for viruses. Each extract was tested in duplicate in the 5′ Taq nuclease assay. The detection limit of the assay in TCID₅₀ was calculated from the highest dilution at which a C_T value was consistently obtained.

To determine the detection limit in terms of target copy number, a 395-bp product incorporating the 5′ Taq nuclease amplicon was produced from the SA2 vaccine using the TKseq PCR, and was purified using a Montage PCR filter unit (Millipore, North Ryde, Australia) as per the manufacturer's instructions. Serial dilutions of the amplicon were prepared in 0.1 mM Tris (pH 7), and were tested in duplicate in the 5′ Taq nuclease assay. The detection limit was determined as described above. The copy number was calculated from the size of the amplicon and the DNA concentration, which was determined using a Qubit Fluorimeter (Invitrogen, Mulgrave, Australia).

The TKseq amplicon dilutions were also used as standards in demonstrating the use of the 5′ Taq nuclease assay for estimating the GaHV-1 load in samples collected for diagnostic purposes.

Amplicon sequencing

DNA extracts from six samples that had C_T values less than 50 in the 5′ Taq nuclease assay were amplified in the TKseq PCR. The samples were selected to cover a range of cases and sample types, including samples that either produced or failed to produce alphaherpesvirus CPE on CK cells. TKseq amplicons were purified using Montage PCR filter units as per the manufacturer's instructions, and sequenced from primer TKseq f and/or TKseq r using a DYEnamic ET Terminator Cycle Sequencing Kit (GE Healthcare Bio-Science, Castle Hill, Australia) as specified by the manufacturer. Sequence analysis was performed on a MegaBACE electrophoresis system (GE Healthcare Bio-Science) at the Genetic Analysis Facility, Advanced Analytical Centre, James Cook University, Townsville, Australia. Alternatively, some purified amplicons were sequenced using a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) as specified by the manufacturer. These sequences were analysed on a 3130×l Genetic Analyser (Applied Biosystems) at the Griffith University DNA Sequencing Facility, School of Biomolecular and Biomedical Science, Nathan, Australia. Sequences were compared with the GenBank database using BLAST on the National Center for Biotechnology Information website.

Estimation of GaHV-1 load in diagnostic samples using the 5′ Taq nuclease assay

A standard curve was constructed using the serial dilutions of the GaHV-1 TKseq amplicon described earlier as standards. Extracts from six positive samples from two cases were tested in parallel with these standards in the 5′ Taq nuclease assay. A standard curve was constructed relating the concentration (target copies/ml tissue homogenate or swab supernatant) to the log C_T values of the DNA standards. The concentration of GaHV-1 extracted from the tissues and swabs was estimated by comparing their log C_T values with the standard curve using software supplied with the RotorGene 3000.

Results

Analytical specificity of the 5′ Taq nuclease assay

The 5′ Taq nuclease assay detected all GaHV-1 field strains and the SA2 vaccine strain. None of the other bacterial or viral extracts listed in Table 1 reacted in the assay.

Analytical sensitivity and dynamic range of the 5′ Taq nuclease assay

The SA2 vaccine was titrated at 1.3×10⁶ TCID₅₀/ml. The highest dilution of vaccine that yielded positive results in the 5′ Taq nuclease assay was 10⁻⁶, corresponding to 1.3 TCID₅₀/ml virus suspension or 1.0×10⁻² TCID₅₀ per reaction (Table 2). The mean C_T at this dilution was 38.09. The reaction efficiency was 0.84. The curve relating copy number to log C_T value was linear over all dilutions down to 10⁻⁵, giving a dynamic range of at least six orders of magnitude.

Table 2.  Mean C_T values for serial 10-fold dilutions of GaHV-1 SA2 vaccine in the 5′ Taq nuclease assay

Dilution	TCID50/ml starting material	TCID50 per reaction	Mean CT value^a
10^−1	1.3×10^5	1.0×10^3	17.86
10^−2	1.3×10^4	1.0×10^2	21.85
10^−3	1.3×10^3	1.0×10^1	25.35
10^−4	1.3×10^2	1.0×10^0	29.59
10^−5	1.3×10^1	1.0×10^−1	32.84
10^−6	1.3×10^0	1.0×10^−2	38.09
10^−7	1.3×10^−1	1.0×10^−3	NR, 36.44^b
10^−8	1.3×10^−2	1.0×10^−4	NR
NR, no reaction. ^aCT values are the mean values obtained from duplicate reactions. ^bCT values are shown for individual reactions; these values were omitted from the calculation of reaction efficiency.

When the TKseq amplicon was used as the template, the highest dilution that yielded positive results in both of the duplicate reactions was 10⁻⁹, corresponding to 90 target copies per reaction (Table 3). The mean C_T at this dilution was 34.21. The reaction efficiency was 0.98. The curve relating copy number to log C_T value was linear down to the detection limit of 90 target copies per reaction, giving a dynamic range of at least nine orders of magnitude.

Table 3.  Mean C_T values for serial 10-fold dilutions of TKseq amplicon in the 5′ Taq nuclease assay

Dilution	Copies per reaction	Mean CT value^a
10^−1	9×10^9	7.79
10^−2	9×10^8	9.82
10^−3	9×10^7	13.26
10^−4	9×10^6	16.73
10^−5	9×10^5	20.11
10^−6	9×10^4	23.65
10^−7	9×10^3	27.16
10^−8	9×10^2	30.41
10^−9	9×10^1	34.21
10^−10	9×10^0	38.9, NR^b
10^−11	9×10^−1	NR
10^−12	9×10^−2	NR
NR, no reaction. ^aCT values are the mean values obtained from duplicate reactions. ^bCT values are shown for individual reactions; these values were omitted from the calculation of reaction efficiency.

Diagnostic samples

Of the 41 diagnostic samples received, 15 samples from eight cases had C_T values ranging from 21.38 to 41.21. Table 4 contains a summary of results for these 15 samples. Five of the 15 samples from four cases produced typical alphaherpesvirus CPE on CK cells. Nine of the remaining 10 samples that reacted in the 5′ Taq nuclease assay did not produce alphaherpesvirus CPE when inoculated onto CK cells. The other sample was not tested on CK cells.

Table 4.  Summary of results obtained for the 15 positive samples

Positive case number	Sample type^a	CT value^b	Copies/ml^c	Herpesvirus CPE	Sequence	Diagnosis^d
1	Trachea	32.48	4.5×10^5	−	GaHV-1	ILT
2	Larynx swab (1)	30.31	2.0×10^6	−	GaHV-1	ILT
	Larynx swab (2)	41.21	1.2×10^3	−	ND
	Larynx swab (3)	32.80	3.7×10^5	−	ND
	Larynx swab (4)	28.80	5.7×10^6	+	GaHV-1
	Larynx swab (5)	31.70	7.9×10^5	−	ND
3	Lung (1)	30.01	ND	−	ND	ILT
	Trachea (1)	23.27	ND	+	ND
4	Cloacal swab	21.38	ND	+	GaHV-1	ILT, mycoplasmosis
5	Trachea	38.19	ND	−	ND	ILT
6	Trachea	24.3	ND	ND	GaHV-1	ILT
7	Trachea (1)	30.63	ND	−	GaHV-1	ILT, fowlpox
	Trachea (2)	24.46	ND	−	ND
8	Trachea (1)^e	27.53	ND	+	ND	ILT, fowlpox
	Trachea (2)^e	30.89	ND	+	ND
ND, not done. ^aValues in parentheses indicate bird number where a disease investigation includes samples from more than one bird. ^bCT values are the mean values obtained from duplicate reactions. ^cCopies/ml swab or tissue suspension. Samples from Cases 1 and 2 only were assayed in parallel with quantification standards. ^dOther investigations contributing to these diagnoses are beyond the scope of this paper. ^eDNA extracted using a KingFisher-96 Particle Processor.

The 26 samples that failed to react in the 5′ Taq nuclease assay did not produce alphaherpesvirus CPE on CK cells.

DNA from six of the samples that reacted in the 5′ Taq nuclease assay (two of which also produced alphaherpesvirus CPE on CK cells) was also amplified using the GaHV-1 TKseq PCR. The samples were from five cases. Sequences derived from all six amplicons matched GaHV-1 (96 to 100% identity) with no other significant matches. The 5′ Taq nuclease amplicon lies within the GaHV-1 TKseq amplicon.

In summary, nine of the 15 samples that reacted in the 5′ Taq nuclease assay either produced alphaherpesvirus CPE on CK cells or were confirmed as containing GaHV-1 by sequencing. These represent seven of the eight positive cases. Samples from the eighth case did not produce alphaherpesvirus CPE and were not available for sequencing.

The positive samples included a cloacal swab that gave a C_T value of 21.38 and that produced alphaherpesvirus CPE on CK cells. The presence of GaHV-1 in the swab was also confirmed by amplicon sequencing. The swab was from a laying hen that had presented with severe conjunctivitis and green diarrhoea. The bird was the fifth on the property to have died with respiratory signs during a 2-week period (Ossedryver, 2007). GaHV-1 was not detected in the trachea from the same bird by 5′ Taq nuclease assay or by virus isolation on CK cells. Furthermore, no tracheal lesions were identified histologically and electron microscopy examination of the tracheal tissue was also negative (data not shown).

Two of the positive cases (Cases 7 and 8) concerned a commercial layer flock that had been vaccinated for ILT.

Healthy broilers

None of the 18 tracheal swabs reacted in the 5′ Taq nuclease assay, and none produced alphaherpesvirus CPE on CK cells.

Estimation of GaHV-1 load in diagnostic samples using the 5′ Taq nuclease assay

Using the TKseq standards, a curve relating the copy number to the log C_T value was produced. This curve was linear down to the detection limit of 90 target copies per reaction, and had a correlation coefficient of 0.99884 as calculated by the RotorGene software, and a reaction efficiency of 0.98. The concentrations of GaHV-1 in the sample homogenates varied from 1.2×10³ to 5.7×10⁶ copies/ml (Table 4). Only the sample yielding 5.7×10⁶ copies/ml homogenate produced alphaherpesvirus CPE on CK cells. This sample and two of the other samples for which virus load was estimated were confirmed as containing GaHV-1 by amplicon sequencing. Sequencing was not performed on the other three samples.

Discussion

This paper describes the design and validation of a 5′ Taq nuclease assay for detecting GaHV-1 in swabs and tissues from birds suspected of having ILT. The usefulness of the assay for estimating viral loads was also demonstrated. The assay uses MGB probe technology and targets the TK gene of GaHV-1. The TK gene is frequently chosen as the target for PCR assays for herpesviruses; for example, previous assays for GaHV-1 (Williams et al., 1992; Abbas et al., 1996), Bovine herpesvirus 1 (Moore et al., 2000) and Equid herpesvirus 1 and Equid herpesvirus 4 (Carvalho et al., 2000). Targeting the TK gene produced an assay that detected all of the strains of GaHV-1 that were tested but did not react with any of the other avian pathogens tested. Although all of the GaHV-1 strains were from Australia, the inclusion of geographically diverse GaHV-1 TK sequences in the assay design process should increase the likelihood that the assay will detect strains from other countries. The assay does not differentiate between the SA2 vaccine and field strains.

The 5′ Taq nuclease assay was also rapid, allowing a turnaround time of several hours compared with the several weeks required for virus isolation.

Most of the bacteria used in testing the analytical specificity of the 5′ Taq nuclease assay were type strains. However, the viruses that were used were mostly field strains. These strains were regarded as representative of the viruses likely to occur in samples being tested for GaHV-1.

The real-time PCR described by Creelan et al. (2006) used the intercalating dye Sybr Green, and required melting curve analysis after amplification had been completed to confirm the specific nature of any amplification. 5′ Taq nuclease assays have the advantage of not requiring the use of melting curve analysis to confirm amplification specificity.

The 5′ Taq nuclease assay compared favourably in terms of analytical sensitivity and dynamic range with the assay described by Callison et al. (2007), which did not incorporate MGB technology. For their assay a limit of quantitation of 100 target copies per reaction, below which the curve relating C_T to target copy number became non-linear, an ultimate detection limit of 25 target copies per reaction and a dynamic range of at least six orders of magnitude were cited. Our 5′ Taq nuclease assay had an analytical sensitivity of 1.0×10⁻² TCID₅₀ per reaction or 90 target copies per reaction. As our assay was still linear at 90 target copies per reaction, this figure represents a quantitation detection limit. The ultimate detection limit for our assay was between 90 and 9 target copies per reaction. Using the TKseq amplicon as the template, our assay had a dynamic range of at least nine orders of magnitude. The dynamic range obtained using serially diluted SA2 vaccine was at least six orders of magnitude. The lower dynamic range obtained using SA2 vaccine as compared with the TKseq amplicon was due to a lower starting concentration of target as indicated by the higher C_T values across the dilution series. The upper and lower limits of dynamic range were not established for either assay. However, an extended dynamic range is one of the expected advantages of MGB probes over other 5′ Taq nuclease probes such as that used by Callison et al. (2007) (Kutyavin et al., 2000).

The 5′ Taq nuclease assay was more sensitive than virus isolation on CK cells, as demonstrated by the fact that only five of the 15 samples that reacted in the 5′ Taq nuclease assay produced alphaherpesvirus CPE on CK cells. Ten of the samples that reacted in the 5′ Taq nuclease assay were either not inoculated onto CK cells (one sample) or failed to produce alphaherpesvirus CPE. However, the GaHV-1 TKseq PCR yielded GaHV-1 sequences for four of these samples. On a case-by-case basis, for seven of the eight positive cases the 5′ Taq nuclease result was supported by virus inoculation onto CK cells and/or by sequencing. This result, together with the failure of the samples from the healthy broilers to react in the 5′ Taq nuclease assay, confirmed the increased sensitivity and specificity of the 5′ Taq nuclease assay. Callison et al. (2007) observed a similar increased detection rate in their 5′ Taq nuclease assay relative to culture, but the specificity of this amplification was not confirmed.

The 5′ Taq nuclease assay had a reaction efficiency of 0.98 when the TKseq amplicon was used as template but a reaction efficiency of 0.84 when the starting material was SA2 virus. The difference in assay performance was most probably due to inefficiencies in template extraction.

The detection limits of the 5′ Taq nuclease assay corresponded to C_T vaues of 34.21 (using GaHV-1 TKseq amplicon as the template) and 38.09 (using GaHV-1 genomic DNA as the template). The highest C_T value recorded for a diagnostic sample in this study was 41.21, which corresponds to 1.2×10³ target copies/ml swab suspension or 9.6 target copies per reaction. The C_T values obtained for this sample are approaching the ultimate detection limit of the assay, as suggested by the trend in C_T values presented in Table 3. These results suggest that samples giving mean C_T values of 42 or less could be considered positive. This criterion for interpreting C_T values is presumptive and could be modified as more samples are tested. It may also differ when instrumentation platforms other than the Corbett RotorGene are used for the 5′ Taq nuclease assay.

GaHV-1 was detected in a cloacal swab by 5′ Taq nuclease assay and its presence was confirmed by virus isolation and amplicon sequencing, confirming that the bird was shedding GaHV-1 in cloacal secretions. A similar finding was reported by Creelan et al. (2006). Sampling for ILT diagnosis is generally restricted to respiratory sites (Guy & Garcia, 2008) and the virus may remain undetected in cases such as this. Furthermore, chicken litter contaminated with infected droppings could serve as a source of GaHV-1 infection. This reinforces the need for adequate clean-up of facilities between successive flocks as a preventative measure against ILT (Guy & Garcia, 2008). This finding also suggests that sites sampled for GaHV-1 testing should include the cloaca.

GaHV-1 was detected in two cases concerning a vaccinated flock with respiratory disease. In both cases, histopathology detected intranuclear inclusions in the respiratory tract that were consistent with ILT (data not shown), and a diagnosis of ILT was made. Live modified vaccines can fail if not handled and administered correctly, and may also reacquire virulence under certain conditions. Factors contributing to vaccine failure were reviewed by Guy & Garcia, (2008).

In conclusion, we have demonstrated that the 5′ Taq nuclease assay is an effective tool for detecting GaHV-1 in a variety of samples including cloacal swabs and for determining viral load if required. It offers the advantages of MGB probe technology over other commonly used real-time PCR systems such as extended dynamic range and not requiring melting curve analysis to determine the nature of amplification products.

Acknowledgements

The present work was funded by the Australian Biosecurity Cooperative Research Centre for Emerging Infectious Disease and Queensland Primary Industries and Fisheries, Queensland, Australia. The authors would like to thank Rod Jenner of Golden Cockerel Pty Ltd for supplying swabs from healthy broilers, and Dr Pat Blackall for supplying bacterial strains used in this study.