It is accepted dogma for avian influenza virus (AIV) that aquatic birds are the natural reservoir for all 16 hemagglutinin (HA) and 9 neuraminidase (NA) AIV subtypes [16]. Infection of land-based poultry with H9N2 viruses has become ubiquitous and endemic in several countries, wherein several bird species are commonly affected and potentially transmit virus to humans and pigs [4]. Usually, poultry infected with H9N2 AIV, unless complicated with other pathogens, show no clinical illness or suffer mild respiratory signs and a drop in egg production [38]. Experimental infection of SPF chickens has shown that H9N2 avian influenza virus is low-pathogenic, but in the last decade, Asian and Middle Eastern countries have faced frequent outbreaks of H9N2 infection with high mortality [10, 21, 38]. On the other hand, co-infection with other pathogens such as Staphylococcus aureus, Haemophilus paragallinarum, E. coli or infectious bronchitis virus can aggravate H9N2 infections, resulting in high mortality rates [10, 23, 31].

Quail have been found to be highly susceptible to H9N2 viruses, with few changes in the HA gene required for efficient replication and transmission in quail [40]. Bobwhite quail express both sialic acid-α2, 3-galactose linked, avian-type receptors, and sialic acid-α2, 6-galactose linked, human-type receptors [21] and thus may be considered another “mixing vessel”, like pigs, for the generation of reassortant viruses from mammalian and avian sources with potentially novel antigenic and genetic features [27]. Furthermore, A/Quail/Hong Kong/G1/97-like H9N2 virus has been assumed to be a donor of the internal gene segments of the lethal H5N1 virus that emerged in Hong Kong in 1997 [19].

Based on phylogenetic analysis of the HA gene of H9N2 viruses, there are, so far, two major genetic lineages: the North American and Eurasian lineages [9]. In the latter, several sublineages have been distinguished: The G1-like sublineage was established in the Middle East and on the Indian subcontinent in the 1990s while other sublineages (Y280 and Ck/bei-like) circulate mainly in countries of the Far East [44]. Extensive evolutionary genetic analysis has indicated that East Asia has been the major source of H9N2 in the Middle East area [17]. Circulation of H9N2 viruses in the Middle East and Northern Africa since the year 2000 has been frequently reported in Israel, Jordan, Lebanon, Saudi Arabia, the United Arab Emirates, Kuwait, Iraq and Libya, where inactivated vaccines have been implemented in some countries to combat H9N2-associated disease of economical importance in poultry [5, 8, 17, 36]. Nevertheless, isolation of H9N2 virus from poultry has not yet been reported from Egypt, whereas the endemic status of H5N1 virus in poultry in Egypt with frequent, and occasionally fatal, transmissions to the human population since 2006 is well known [3].

Here, we describe the first isolation and genetic characterization of H9N2 virus from commercial bobwhite quail in Egypt.

A commercial broiler bobwhite quail (Colinus virginianus) farm with 5000 birds, 35 days old, was routinely surveyed for H5N1 infections during pre-slaughter active surveillance. The flock was vaccinated against H5 with recombinant fowl pox-H5 (Trovac AIV-H5, Merial Select, Inc. Gainesville, USA) and inactivated H5N2 (Volvac AI KV, Boehringer Ingelheim, Mexico) commercial vaccines at 3 and 9 days of age, respectively. At the time of sampling, the examined flock showed neither clinical illness nor unusual mortality rates.

Swab sampling of birds in this study was conducted according to published guidelines of the OIE [39]. Samples were collected by professional veterinarians at the National Laboratory for Quality Control on Poultry Production (NLQP) and General Organization of Veterinary Services (GOVs), Egypt, as a part of the routine nationwide pre-slaughter surveillance programs according to ministerial decree number 221/2006, in which NLQP is responsible for official diagnosis and surveillance of AIV in Egypt. Ten tracheal and cloacal swabs were collected separately on May 28, 2011. The samples were immersed in cooled viral transport medium containing antibiotics and transported immediately to NLQP without interrupting the cold chain [39]. Pools of swab samples were used for virus isolation and/or RNA detection as described below.

Embryonated chicken eggs obtained from a national specific-pathogen-free chicken farm were incubated at 37°C. Fertile 9- to 11-day-old embryonated eggs were inoculated via the allantoic sac route with 0.1 ml of pooled swab materials as recommended [39]. Inoculated eggs were examined daily for embryo mortality. Amnio-allantoic fluid (AAF) was harvested after five days and tested for hemagglutination activity and bacterial contaminants and subsequent passaging.

The infected AAF collected from inoculated eggs was tested for hemagglutination activity according to the recommended protocol [39]. Hemagglutination-positive allantoic fluids were examined using the HI test as described [39], using reference antisera specific for H5N1, H5N2, H6N7, H7N7, H9N2 and H9N7 viruses provided by Istituto Zooprofilattico delle Venezie (IZVSe), Padova, Italy, through FAO Egypt (ECTAD unit). Both tests were carried out using 1% chicken erythrocytes in 96-well V-shape plates (Nunc, Wiesbaden, Germany) [39].

RNA was extracted from a pool of five cloacal and five tracheal swabs by using a QiaAmp Viral RNA Mini Kit (QIAGEN, Hilden, Germany) following the manufacturer’s instructions. RNA extracts, 5 μl per reaction, were reverse transcribed and amplified using a One Step Real-Time RT PCR Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s guidelines for detection of AIV using generic primers and probes targeting the matrix gene [46], H5 gene [2], H7 gene [45] and H9 gene [37]. Each gene was detected in a separate reaction using a Stratagene MX3005P real-time PCR machine (Stratagene, Amsterdam, The Netherlands). For the identification of the N2 gene, a conventional reverse transcription PCR assay described by Fereidouni et al. [14] was used (Biometra Thermocycler machine).

Viral RNA was extracted from the allantoic fluid of H9N2 AIV-infected eggs using a QiaAmp Viral RNA Mini Kit (QIAGEN, Hilden, Germany) and was reverse transcribed using a Super-Script III reverse transcriptase kit (Invitrogen, Carlsbad, CA). Amplification of 225 nucleotides (encoding 75 amino acids) of the HA gene and 334 nucleotides (encoding 110 amino acids) from the NA gene was done using specific primers [14, 25]. Gene sequencing was carried out using a BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems, Foster City, CA) in an ABI PRISM® 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA) as described elsewhere [7].

BioEdit software version 7.0.9.0 [24] was used to analyze and edit the generated sequences of the HA and NA genes. The GenBank database was screened (BLAST) for closely related sequences. A phylogenetic analysis of the newly obtained HA and NA nucleotide sequences using the most homologous H9N2 gene sequences obtained from the BLAST search was carried out using the PhyML maximum-likelihood algorithms with 100 bootstrap replicates using TOPALI v2.5 [35]. A GTR+G model of nucleotide substitutions was chosen according to results obtained with the ModelTest software implemented in the Topali suite [35]. Unrooted HA and NA trees were drawn with Dendroscope [28] and further viewed and edited using Inkscape software version 0.48. Gene sequences of the virus isolated in this study were submitted to the GISAID database, and the accession numbers were WSS72090 and WSS72089 for the HA and NA gene, respectively.

In the present investigation, a hemagglutinating virus was isolated from swab samples obtained from 35-day-old commercial bobwhite quail in embryonated chicken eggs after two passages. The isolated virus was found to be H9N2 using an HI test and RT-qPCR. Tests for other AIV subtypes, namely H5 and H7, gave negative results. The virus isolated in this study had an average nucleotide identity of 99.1% for the HA gene and 98.2% for the NA gene, and 97.3% and 100% identity for the HA and NA proteins, respectively, in comparison to other viruses isolated from the Middle East region. Both the HA and NA genes clustered with those of the G1-like H9N2 virus currently circulating in the region in Gaza, Israel, Lebanon, Iran, the United Arab Emirates and Saudi Arabia (Figure 1). The presence of the amino acid leucine (L) at position 234 (H3 numbering: 226), a part of the receptor-binding domain, in the HA protein of the isolated virus was observed. The NA gene sequence had glutamic acid at position 277 and arginine at position 292, which are known to be associated with sensitivity to neuraminidase inhibitors.

Fig. 1
figure 1

Phylogenetic relation of H9 and N2 genes of A/quail/Egypt/113413v/2011 (H9N2) virus. Phylogenetic trees of the H9 hemagglutinin gene (A) and the neuraminidase (N2) gene (B) of A/quail/Egypt/113413v/2011(H9N2) isolated from 35 days old commercial bobwhite quail in Egypt. Bootstrap replicates (N=100) were calculated using maximum likelihood and a GTR+G model for nucleotide substitution as implemented in software Topali v2.5 [31]. The unrooted trees were visualized with Dendroscope [27] then edited by Inkscape 0.48 software. The virus isolated in this study is highlighted in grey; common H9N2 sublineages were depicted to the right of each tree

Full size image

Infection of different poultry species, including quail, with H9N2 in the Middle East has been reported frequently since the early 2000s [1, 5, 18, 36, 41, 42]. Here, for the first time, the isolation of H9N2 virus in Egypt is described. Despite previous intense surveillance activities conducted by our team and others in the context of H5N1 endemicity in Egypt, H9 viruses had not been detected before [6, 7, 13, 22]. In the present investigation, close sequence and phylogenetic identities with the G1-like H9N2 virus currently circulating in the neighbouring countries could indicate further spread of the virus into Egypt as a part of its extensive circulation in the nearby countries rather than a separate introduction from the Eastern Asian AIV epicentre. Similar observations have been noted with the H5N1 virus where it is endemic in Egypt and occasional sporadic outbreaks of phylogenetically closely related strains have been recorded in neighbouring countries [18, 32]. Possible pathways of introduction, in particular, illegally or legally transported poultry or poultry products, should be investigated. Migratory birds as a source of incursion cannot be ruled out presently, but it seems unlikely. Fereidouni et al. [15] have shown that the H9 lineages circulating in wild birds in Iran are distinguishable from those in poultry.

It is noteworthy that in Eastern Asia, quail were the first land-based birds from which H9N2 viruses were isolated; they still claim endemic status to date [33]. In addition, Guan et al. [20] reported that 17% of quail in live retail poultry markets in Hong Kong were positive for H9N2 AIV. Birds examined in this study were clinically healthy, and mortality was neither reported by the owner nor observed by the sampling team of veterinarians. This is in accordance with results obtained after experimental H9N2 infection of 4-week-old Japanese quail, resulting in efficient virus replication and transmission between quails without inducing any clinical signs [40]. Moreover, quails in the current commercial flock were vaccinated twice with H5 vaccines (as a part of the nationwide vaccination campaign to confront the endemic H5N1 virus infection). Although cross-protection, e.g., through the elicited cellular immune response [43] induced by these H5 vaccines or the anti-NA (N2 subtype) antibodies induced by the H5N2 vaccine [47] against an H9N2 infection, seems unlikely, experimental infections in naïve quail are required to confirm or exclude this cross-protection effect. Khalenkov et al. [30] reported that 90%-100% of chickens previously infected with a recent H9N2 Israeli virus survived an experimental infection with a lethal H5N1 virus. Likewise, results obtained by Imai et al. [29] indicated that chickens vaccinated with H9N2 vaccine were partially protected against challenge with highly pathogenic H5N1 virus. According to these results, we assume that vaccination of birds using H5N2 vaccine can probably modulate infections with the low-pathogenic H9N2 viruses.

Since the first emergence of H9N2 AIV in poultry in 1966 [26], the virus continues to be a surprisingly devastating disease of birds. Due to its zoonotic potential and spread in swine in China, it is also shortlisted as a candidate for a future influenza pandemic [12, 33]. Quail have become another focus of attention as mixing vessels of influenza A viruses from different sources [27, 40, 48].

Since 1998, human infections with H9N2 virus have been reported frequently in Asia [11, 34]. It has been observed that avian H9N2 viruses carry glutamine (Q) at position 234, a part of the receptor-binding domain of the HA protein, which increases the binding affinity of the virus to the avian 2,3-linked sialic acid receptor. Human H9N2 viruses instead carry leucine (L) at this position, similar to human H2 and H3 viruses, which enhances the binding affinity to human 2,6-linked sialic acid receptors and improves direct-contact transmission in ferrets [11, 34, 49]. It is worth pointing out that highly pathogenic AIV H5N1 has established an endemic status in Egypt, even in vaccinated poultry [3]. Human cases of H5N1 infection continue to be reported from the country. Introduction of yet another AIV, which harbours a typical human virus L226 signature, calls for intensified nationwide surveillance, continuous monitoring, prompt control and retrospective studies to elucidate the spread of the virus in all Egyptian poultry sectors. Potential reassortment of the co-circulating H9N2 and HPAIV H5N1 in Egypt poses a major risk not only for poultry but also from a zoonotic point of view. Therefore, the complete genome sequence of the virus isolated in this study should be determined.