Pyrosequencing analysis for a rapid classification of fowl adenovirus species

Abstract

A rapid fowl adenovirus (FAdV) classification method based on a 30-bp sequence of the hexon loop (L1) was developed using the pyrosequencing technique. FAdV identification is relevant for epidemiological studies and for the adoption of a correct strategy where vaccination is to be used for the control of the disease. FAdV typing is usually performed using polymerase chain reaction coupled with either conventional DNA sequencing or restriction enzyme analysis; however, both methods can be time consuming and/or very expensive to be used as a routine tool. In the present study, polymerase chain reaction and subsequent pyrosequence analysis of the variable hexon L1 region were assessed in order to rapidly differentiate FAdV species. Forty-nine FAdV samples (22 reference strains and 27 field isolates) were tested and the results were compared with those obtained by conventional DNA sequencing. The results clearly demonstrated that pyrosequence analysis provides a new approach for a rapid differentiation and classification of the FAdV species that is faster, more cost-effective and easier to interpret than other techniques commonly used.

Introduction

Fowl adenoviruses (FAdVs), belonging to the Aviadenovirus genus of the family Adenoviridae, are distributed worldwide and known to cause significant economic problems in the poultry industry, mainly due to inclusion body hepatitis (IBH) and hydropericardium syndrome (McFerran & Connor, 1977; Christensen & Saifuddin, 1989; Nakamura et al., 1999; Toro et al., 1999; Hess, 2000; Alvarado et al., 2007), but are also associated with a wide range of clinical presentations including gizzard erosion (Ono et al., 2001, 2003; Muroga et al., 2006) and pancreatitis (Grimes et al., 1977; Nakamura et al., 2002).

FAdVs are mostly transmitted horizontally, being present in all excretions and at high titres in faeces. Vertical transmission is also possible where the virus can be passed through embryonated eggs and can be reactivated in chicks that are a few weeks old, especially if the birds are immunodepressed (McFerran & Smyth, 2000; Grgic et al., 2006).

Various serological tests, such as enzyme-linked immunosorbent assays (Hess, 2000; Balamurugan & Kataria, 2004; Philippe et al., 2007), agar gel immunodiffusion and immunofluorescence (Balamurugan & Kataria, 2004) can be used for antibody detection. However, since serological methods detect only common group-specific antigens, and because of the widespread occurrence of antibodies to avian adenovirus, these techniques are only of minor relevance for their diagnosis. For these reasons, investigations are mainly focused on direct detection and identification of the agent (Hess, 2000) that can be isolated both from sick and healthy birds (McFerran & Smyth, 2000).

FAdVs have been grouped into five species (A to E), based on their molecular structure, and divided into 12 serotypes, based on cross-neutralization assays (Hess, 2000).

Despite the fact that FAdVs are generally considered ubiquitous in poultry farms, certain species and serotypes are known to be associated with primary disease such as IBH, while others cause no clinical disease. Characterization of FAdVs isolated in Canada (Ojkic et al., 2008), New Zealand (Saifuddin & Wilks, 1991) and Australia (Erny et al., 1991) showed that the IBH syndrome is mainly associated with FAdV species E and D, while species A and C have been less frequently found in severe IBH cases (Toro et al., 1999; Ojkic et al., 2008). FAdV classification is therefore very useful for epidemiological tracing and is of critical importance where vaccination is to be used for the control of the disease (Toro et al., 2002; Jadhao et al., 2003; Balamurugan & Kataria, 2004; Steer et al., 2009).

The hexon protein is the major capsid protein of the non-enveloped virion on which type, group, and subgroup-specific determinants are located (Norby, 1969). The hexon gene consists of conserved regions (pedestals), which are located more inside the virion, and variable loops, which protrude from the surface (Roberts et al., 1986; Athappilly et al., 1994), containing the type-specific neutralizing epitopes (Toogood et al., 1992; Adam et al., 1998).Due to the interaction with the immune system, the sequence identity between loop regions of different species is low (Sheppard et al., 1995; Crawford-Miksza & Schnurr, 1996). The hexon loop 1 (L1) represents the most variable region and has proved to be the most useful region in identifying and differentiating FAdV species and serotypes, when used in polymerase chain reaction (PCR) coupled either with conventional DNA sequencing or with restriction fragment length polymorphism (RFLP) analysis (Raue & Hess, 1998; Hess et al., 1999; Meulemans et al., 2001; Steer et al., 2009). However, these methods are time consuming, often require extensive interpretation, and the running cost per analysis can be relatively high to be used as a routine typing tool.

Pyrosequencing has proven to be a rapid, sensitive and high-throughput method for sample screening and genotyping in different fields of disease research (Ronaghi & Elahi, 2002; Elahi et al., 2003; Swan et al., 2006; Deyde & Gubareva, 2009; Deyde et al., 2009; Quince et al., 2009). On the strength of these considerations, the aim of the present study was to develop a rapid FAdV species classification method, based on the pyrosequencing of a 30-bp L1 region of the hexon gene.

Materials and Methods

Viruses

Twenty-two FAdV reference strains (Table 1), representatives of currently recognized distinct serotypes, and 27 field isolates collected in Italy between 2007 and 2008 were used in this study.

Table 1.  FAdV reference strains used in the present study.

Species	Serotype	Strains
A	FAdV-1	CELO
B	FAdV-5	TR22, IBH-2A
C	FAdV-4	KR5, 506, J2-A
	FAdV-10	C-2B
D	FAdV-2	685, P7-A
	FAdV-3	SR49, 75, 75-1A-1
	FAdV-9	A-2A
	FAdV-11	380
E	FAdV-6	CR119
	FAdV-7	YR36, X11-A
	FAdV-8a	TR59, 58, T8-A
	FAdV-8b	764, B-3A
Species and serotypes are indicated according to the International Committee on Taxonomy of Viruses current nomenclature system (http://www.ncbi.nlm.nih.gov/ICTVdb/Ictv/fs_adeno.htm#Genus2).

All the viruses were grown in chicken embryo liver cells obtained from 14-day-old specific pathogen free embryos (Lohmann). Chicken embryo liver cells (1 x 10⁶ cells/ml) in 25-cm² culture flasks were incubated at 37°C with 5% carbon dioxide in Minimal Essential Medium Eagle (Sigma) containing 5% foetal calf serum (Invitrogen), L-glutamine 2 mM (Sigma) and 1% PenStrep (Sigma). Cell monolayers were inoculated with 500 µl viral suspension and, after 1 h absorption, were incubated for 2 to 3 days, until an extensive cytopathic effect could be seen. The infected cells and their supernatants were then harvested and freeze–thawed three times. The suspensions were centrifuged at 3000 x g for 10 min to eliminate cell debris and the supernatants were stored at –20°C.

Hexon L1 conventional PCR amplification

Hexon L1 Genbank sequences CELO AF339914, 685 AF508947, P7-A AF339915, SR49 AF508948, 75 AF508949, 75-1A-1 AF339921, J2-A AF339917, 506 AF508950, KR5 AF508951, TR22 AF508953, IBH-2A AF339916, CR119 AF508954, YR36 AF508955, X11-A AF339924, TR59 AF508956, 58 AF508957, T8-A AF339919, 764 AF508958, B-3A AF339922, A-2A AF339918, C-2B AF339923 and 380 AF339925 were aligned using MEGA software (version 4.1) in order to identify a 30-bp region capable of discriminating different FAdV species. The conserved sequence upstream of this 30-bp region was used to design the PCR forward primer FAdV-Pyro-fw (5'-TGGGTBYTGGACATGGG-3'). Adenoviral dsDNA was extracted using the Roche High Pure PCR Template Preparation Kit, according to the manufacturer's instructions. PCR was performed using the FadV-Pyro-fw primer and the reverse primer Hexon B (5'-TAGTGATGMCGSGACATCAT-3') (Meulemans et al., 2001), in order to generate a PCR product between 740 bp (FAdV-4: J2-A AF339917, 506 AF508950, KR5 AF508951; FAdV-10: C-2B AF339923) and 765 bp (FAdV-5: IBH-2A AF339916). DNA amplifications were carried out in a total volume of 50 µl containing 5 µl viral DNA, 5 pmol each primer, 1.5 mmol MgCl₂, 0.8 mmol dNTPs and 5 units AmpliTaq Gold^® (Roche). The PCR was started with an initial denaturation step of 10 min at 95°C. The temperature profile of the following 40 cycles consisted of 1 min at 95°C for denaturation, 1 min at 55°C for primer annealing and 90 sec at 72°C for elongation. The reaction was terminated by a final elongation step of 7 min at 72°C. The amplification products were separated on a 1.2% agarose gel stained with ethidium bromide and visualized by ultraviolet transillumination. The size of the amplification product was confirmed by comparison with a commercial PCR marker (Roche DNA Molecular Weight Marker VI 0.15—2.1 kbp) in agarose gel and by the following sequence analysis. PCR specificity was appraised against the DNA of haemorrhagic enteritis virus (HEV), egg drop syndrome virus (EDS'76), infectious laryngotracheitis virus (ILTV) and against the cDNA of infectious bronchitis virus (IBV serotype M41) and avian influenza virus (AIV serotype H7).

DNA sequencing and BLAST alignment

Hexon L1 PCR products were purified using High Pure PCR Product Purification (Roche) and sequenced using the same primers used for PCR and the Big Dye^® Terminator v 3.1 Cycle sequencing Kit, according to the manufacturer's protocol (Applied Biosystems Inc.).

The sequences were analysed using an ABI PRISM 3130xl Genetic Analyzer (Applied Biosystems Inc.) and aligned using the BLASTn program (http://blast.ncbi.nlm.nih.gov/Blast.cgi) to confirm the identity of the viral strains. Aiming to verify the relationship between FAdV samples, phylogenetic analysis of the aligned sequences was performed and neighbour-joining distance trees together with the pairwise distance analysis were performed using the MEGA 4.1 program (Tamura et al., 2007).

Pyrosequencing PCR

PCR was performed using the same set of primers as for the previous PCR; however, the Hexon B primer was biotinylated (5'-[Biotag]TAGTGATGMCGSGACATCAT-3') (Figure 1). DNA amplifications were carried out in a total volume of 50 µl using the same template DNA, reagents and cycling conditions as before, with the exception of the primer concentrations, which were increased to 15 pmol.

Figure 1.  Schematic representation of the P1 and L1 regions of the FAdV hexon gene and the primers used in this study for PCR reaction and pyrosequencing analysis.

Pyrosequencing reaction

Amplification products were immobilized onto streptavidin-coated sepharose beads (5 µl/sample) and washed in a series of buffers according to Biotage's recommendations (PyroMark™ Sample Prep Guidelines; Biotage). Single-stranded biotinylated DNA products were hybridized to FAdV-Pyro-Fw sequencing primer in a 96-well plate used at a final concentration of 0.5 µM in 40 µl annealing buffer. Pyrosequencing was performed in a Pyromark ID platform pyrosequencer using Pyro Gold reagents according to the manufacturer's instructions. The related sequences were analysed and compared with a FAdV hexon sequences library using IdentiFire software (Biotage). The library was compiled using the same Genbank sequences used to design the FAdV-Pyro-Fw primer, with the addition of a turkey adenovirus TAdV FJ360746. The consistency of the pyrosequencing technique was appraised, twice repeating the DNA extraction on different days, using different batches of reagents; while the accuracy of the sequences obtained was confirmed by comparison with the results of conventional sequencing.

Nucleotide sequence accession number

FAdV reference strains were previously submitted by the Dr Thierry van den Berg group (Veterinary and Agrochemical Research Centre, Groeselenberg, Brussels, Belgium), which kindly provided us with these viruses (Table 2). GenBank accession numbers were assigned to the nucleotide sequences of the FAdV field isolates as follows: 1279/08, HM592267; 147/08, HM592268; 1474/08, HM592269; 132/08, HM592270; 499/08, HM592271; 829/08, HM592272; 457/08, HM592273; 5670, HM592274; 5705, HM592275; 5363, HM592276; 6169, HM592277; 5341, HM592278; 5630, HM592279; 5441, HM592280; 5997, HM592281; 5765, HM592282; 4587, HM592283; 4158, HM592284; 4220, HM592285; 5231, HM592286; 5339, HM592287; 3890, HM592288; 2496, HM592289; 2797, HM592290; 2161, HM592291; 5083, HM592292; and 3636, HM592293 (Table 3).

Table 2.  Pyrosequencing report of 22 reference strains representing the five FAdV species (A to E) and the 12 different serotypes.

				Score (%)
Sample	BLASTn results	BLASTn sequence homology (%)	Pyrosequence results	First extraction	Second extraction
75-1A-1	AF339921	100	D, serotype 3	100	100
J2-A	AF339917	100	C	100	100
T8-A	AF339919	100	E	100	100
P7-A	AF339915	100	D, serotype 2	100	100
B-3A	AF339922	100	E	100	100
A-2A	AF339918	100	D, serotype 9	100	100
C-2B	AF339923	100	C	100	100
X11-A	AF339924	100	E	100	95.3
YR36	AF508955	100	E	95.3	100
58	AF508957	100	E	95.3	100
75	AF508949	100	D, serotype 3	100	100
506	AF508950	100	C	100	100
685	AF508947	100	D, serotype 2	100	100
764	AF508958	100	E	100	100
CELO	AF339914	100	A, serotype 1	100	100
KR5	AF508951	100	C	100	100
TR22	AF508953	100	B, serotype 5	69.1	100
380	AF339925	100	D, serotype 11	100	100
SR49	AF508948	100	D, serotype 3	100	100
TR59	AF508956	100	E	100	100
IBH-2A	AF339916	100	B, serotype 5	100	100
CR119	AF508954	100	E, serotype 6	100	100
Pyrosequencing analysis, performed after DNA extraction on two different days (first and second), identified all the different species and at least seven serotypes among those tested. No pyrosequencing reaction was observed for the non-template PCR controls tested either with or without beads and sequencing primer (data not shown). The results column shows the FAdV species and serotype from the sequence library to which the tested samples shares homology in the 30-bp pyrosequencing region. The degree of homology is reflected by the score, which is given as a percentage.

Table 3.  Comparison between the BLASTn and the pyrosequence results of the 27 field isolates.

						Species/serotype
Isolation year	Sample	GenBank accession number	BLASTn results^a	BLASTn sequence homology (%)	Strain	BLASTn	Pyrosequence
2008	1279/08	HM592267	AF508956	100	TR59	E/8a	E
	147/08	HM592268	EF685489	98	–	E	E
	1474/08	HM592269	AF508956	99	TR59	E/8a	E
	132/08	HM592270	AF508950	98	506	C/4	C
	499/08	HM592271	EF685649	100	–	E	E
	829/08	HM592272	U46933	99	CELO	A/1	A/1
	457/08	HM592273	FJ360746	91	TAdV-A	TAdV-A	TAdV-A
2007	5670	HM592274	AF508950	99	506	C/4	C
	5705	HM592275	AF339924	98	X11-A	E/7	E
	5363	HM592276	AF508950	99	506	C/4	C
	6169	HM592277	AF508951	99	KR5	C/4	C
	5341	HM592278	AF508956	100	TR59	E/8a	E
	5630	HM592279	AF339924	98	X11-A	E/7	E
	5441	HM592280	AF508948	96	SR49	D/3	D/3
	5997	HM592281	EU931693	99	–	C/4	C
	5765	HM592282	AF508947	99	685	D/2	D/2
	4587	HM592283	AF508950	98	506	C/4	C
	4158	HM592284	AF508950	98	506	C/4	C
	4220	HM592285	EU979367	99	CELO	A/1	A/1
	5231	HM592286	U46933	100	CELO	A/1	A/1
	5339	HM592287	AF508956	100	TR59	E/8a	E
	3890	HM592288	EU931693	99	–	C/4	C
	2496	HM592289	EU979374	99	TR59	E/8a	E
	2797	HM592290	AF508956	99	TR59	E/8a	E
	2161	HM592291	EU861037	98	CELO	A/1	A/1
	5083	HM592292	AF508949	99	75	D/3	D/3
	3636	HM592293	AF508956	100	TR59	E/8a	E
^aBlastn results were obtained by alignment of the field sample consensus sequences (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Results

Hexon L1 PCR amplification and sequence analysis

The primer set FadV-Pyro-fw and Hexon B was used to amplify approximately 765 bp of the FAdV hexon gene. All of the samples tested gave a clear amplification product of the expected size, without aspecific bands in a 1.2% agarose gel (Figure 2). No amplification products were observed for HEV, EDS'76, ILTV, IBV strain M41, AIV subtype H7 and the non-template control (K–).

Figure 2.  Agarose gel electrophoresis of the 740 to 765 bp PCR products from single FAdV strains representing FAdV serotypes 1 to 11 and control samples using HEV, EDS'76, ILTV, IBV strain M41, AIV subtype H7 and non-template control (K–). MW, molecular weight.

Sequence analysis of both reference strains and field samples confirmed the hexon L1 amplification and the FAdV strain identity (Tables 2 and 3).

Based on the sequences of reference strains PCR products, a phylogenetic tree was created in order to confirm the clustering of the different species according to their hexon loop sequences (Figure 3).

Figure 3.  Phylogenetic tree established with the nucleotide sequences of the approximately 765 bp fragment of the reference strains hexon L1 region. The FAdVs are clustered into five species (A to E) and then divided into 12 serotypes according to the ICTV current nomenclature (ICTVdB Management, 2006). Serotypes:

FAdV-1;

FAdV-2;

FAdV-3;

FAdV-4;

FAdV-5;

FAdV-6;

FAdV-7;

FAdV-8a;

FAdV-8b;

FAdV-9;

FAdV-10;

FAdV-11.

Pyrosequencing analysis of FAdV reference strains

Pyrosequencing analysis of the biotinylated PCR products, using FadV-Pyro-fw as sequencing primer, resulted in good quality pyrograms of up to 30 bp for all the samples tested. Figure 4 shows an example of the resultant pyrogram for a reference strain (506), where peak height is proportional to the amount of pyrophosphate (PPi) and thereby to the number of incorporated nucleotides in the sequence.

Figure 4.  Example of pyrogram showing the 30-bp sequence of the hexon L1 of a reference strain (506). The pyrophosphate anion (P₂O₇ ^4-, PPi) is released by the hydrolysis of ATP into AMP when a nucleotide is incorporated into a growing DNA strand by the polymerase. Peak height is proportional to the amount of PPi and thereby to the number of incorporated nucleotides.

The related sequences were analysed and compared with a 30-bp hexon sequence library using IdentiFire software (Biotage). According to the reference strain classification (Table 1), the pyrosequence results showed a perfect match with the respective species (A to E) for all the viruses with scores ranging from 69.1% to 100% (Table 2). In addition it was also possible to distinguish between different serotypes for species D (serotype 11, strain 380; serotype 9, strain A-2A; serotype 3, strains 75, 75-1A-1 and SR49; and serotype 2, strains 685 and P7-A) and species E (serotype 6, strain CR119).

Despite some minor differences in the scores of a few samples, repeating the DNA extraction on different days and using different batches of reagents, the pyrosequencing results were reproducible and viruses were correctly identified, even in the single case where the identity score was as low as 69.1 (Table 2).

Owing to sequence homology it was not possible to differentiate between serotypes 4 and 10 belonging to species C and among serotypes 7, 8a and 8b within species E. Indeed, pairwise distance analysis showed that FAdV-10 strain C-2B had 100% sequence homology with FAdV-4 strains 506, J2-A and KR5 in the 30-bp region analysed by pyrosequencing (data not shown).

Trials were also performed, considering the first 50 bp of FAdV hexon L1, in order to increase the number of identifiable serotypes, but no improvement was obtained (results not shown). Furthermore, this approach was more time consuming and there was an increase in the signal background due to the larger amount of pyrosequence reaction waste products.

The situation was the same for strains belonging to serotypes 7, 8a and 8b, but in this case even considering the whole hexon L1 PCR product it would have been impossible to go further than the species detection, because strains classified in different serotypes, by cross-neutralization assay, were too similar in their nucleotide sequences (Figure 3).

Pyrosequencing analysis of FAdV field samples

Pyrosequence IdentiFire results for the field isolates were consistent with the corresponding BLASTn results (Table 3) and had an identification score with the correct species ranging from 93 to 100% as shown in Figure 5. Of the 27 field samples, isolated both from sick and healthy birds, 11 (41%) belonged to species E, eight (30%) were grouped in species C, while three (11%) and four (15%) belonged to species D and A, respectively. This approach also allowed the identification of two different FAdV serotypes (2 and 3) within one species (D) and a Turkey adenovirus A isolate (sample number 457/08).

Figure 5.  IdentiFire report of 27 field samples collected in the north-east of Italy between 2007 and 2008. This approach also allowed the identification of a Turkey adenovirus A isolate (sample number 457/08) of a distinct genogroup. No pyrosequencing reaction was observed for the non-template PCR controls (K–). The results column shows the FAdV species and serotype from the sequence library to which the tested samples shares homology in the 30-bp pyrosequencing region. The degree of homology is reflected by the score, which is given as a percentage.

Discussion

The present study describes the development of a pyrosequencing-based genotyping method and provides a direct comparison between conventional nucleotide sequencing and pyrosequencing analysis for the identification and differentiation of the five FAdV species (A to E).

Pyrosequence results of all samples tested were consistent with the sequence analysis and BLASTn alignment. Based on a 30-bp fragment of the hexon L1 sequence, it was possible to sort all the FAdVs tested into their correct species (A to E), and to directly recognize several serotypes (2, 3, 6, 9 and 11). In addition, it was also possible to identify a Turkey adenovirus A isolate belonging to a different genus (Siadenovirus).

Previous studies have reported the use of techniques such as microtitre virus neutralization tests and PCR in combination with either conventional sequencing or RFLP analysis to identify FAdV species and/or serotypes. Despite the wide use of these techniques there are limitations, including the high cost per analysis, the lengthy processes involved and the need for extensive interpretation (Grimes & King, 1977; Erny et al., 1995; Steer et al., 2009).

Starting from the PCR products, the pyrosequencing analysis required only 2 h, for 96 samples, to obtain the final result; almost one-half the time compared with RFLP analysis or conventional DNA sequencing. Pyrosequencing analysis does not require any restriction enzyme, terminator enzyme or previous purification step; thus resulting in lower cost per analysis when compared with RFLP and DNA sequencing. In fact, an approximate cost calculation including all the reagents necessary for the reaction resulted in an estimate of €3.87 to 6.03 per sample (for 96 or 25 samples analysed/run, respectively), compared with €9.86 per reaction for the classical sequencing method (data not shown). It is also to be noted that, within the same run (same plate), pyrosequencing of different microorganisms with different primers can be performed, increasing the cost-effectiveness and the time effectiveness of the method.

Moreover the results analysis is automatically performed by the software upon completion of the sequencing and is reported in a clear and accurate format, as shown in Figure 5.

Steer et al. (2009) proposed the use of high-resolution melting curve analysis of the hexon gene region for FAdV classification. However, this analysis was performed only on 12 reference strains, one representative for each serotype, a vaccine strain and just one field sample. Thus its routine application in order to type field samples needs further evaluation.

Based on our sequence analysis of 22 reference strains, it is very difficult to group together certain viruses in the correct serotype because the cross-neutralization classification does not reflect their nucleotide sequences (Figure 3).

Field samples may possess nucleotide mutation in their sequences that could increase the differences with the reference strains (Meulemans et al., 2001). A nucleotide mutation can result in the disappearance of a restriction site in RFLP analysis or a shift in the melting temperature of the high-resolution melting curve analysis, increasing, in both cases, difficulties in interpretation of results. In our study 27 field samples were tested, giving highly accurate nucleotide sequences of 30 bp thanks to the pyrosequencing reaction chemistry (Gharizadeh et al., 2006). Moreover, in case of a doubtful result due to few nucleotide mutations, the location and nature of the mismatches are easily identified and reported; thus the correct identification of the virus is still possible by comparisons with other sequences in the databases (e.g. GenBank), which could also be rapidly incorporated in the reference pyrosequence library.

On the strength of these considerations the present study provides a new approach for a rapid differentiation and classification of FAdV species using the pyrosequence analysis that is faster, less expensive and easier to interpret than RFLP or conventional DNA sequencing.

In addition, the study demonstrates that it is possible to obtain excellent pyrosequence results even with biotinylated PCR products larger than the 300 to 500 bp recommended by the manufacturer (Biotage). These results were probably due to the high specificity of the primers and the increased concentration of the streptavidin-coated beads used (5 µl/sample), which enhanced the intensity of the peaks in the pyrogram.

Finally as demonstrated by characterization of FAdVs isolated in different parts of the world, the majority of IBH-associated viruses were grouped in FAdV species E (Erny et al., 1991; Saifuddin & Wilks, 1991; Ojkic et al., 2008) and species D (McFerran et al., 1976; Ojkic et al., 2008); and these species can be easily and rapidly detected using the pyrosequencing method.

Taken as a whole, the pyrosequencing assay developed here provides a new tool for the detection of FAdV species, including those isolated from field samples, and thus could assist in identifying and rapidly typing different FAdV species for epidemiology tracing and vaccination strategies.

Acknowledgements

The present work was supported by the Italian Ministry of Health (Ricerca Corrente 29/07). The authors wish to thank Dr Thierry van den Berg, Head of Avian Virology & Immunology, Veterinary and Agrochemical Research Centre (VAR), Brussels, for supplying FadV strains 685, 75, 764, 58, 506, YR36, TR22, CR119, KR5, A-2A, X-11A, J2-A, C2B, IBH-2A, B3A, P7-A, 75-1A-1, and T8-A.