Phylogenetic analysis of Newcastle disease viruses isolated from waterfowl in the Upper Midwest Region of the United States

Background This study was conducted to characterize Newcastle disease virus (NDV) isolates obtained from waterfowl from the Upper Midwest region of the United States. A total of 43 NDVs were isolated by inoculation of cloacal samples in embryonated chicken eggs. These isolates were obtained from 24 mallards, seven American green-winged teals, six northern pintails, four blue-winged teals, and two wood ducks. Partial sequences of fusion gene were analyzed to determine the pathotypes and genotypes involved. Results Deduced amino acid sequence of the cleavage site of fusion (F) protein revealed that all isolates had avirulent motifs. Of the 43 isolates, 23 exhibited sequence motif of 111GGKQGRL117 at the cleavage site, 19 exhibited 111GEKQGRL117 while one isolate showed 111GERQGRL117. Phylogenetic analysis based on comparison with different classes of NDVs revealed that all 43 isolates clustered with class II NDVs and none with class I NDVs. Within class II, five isolates were phylogenetically close to genotype I NDVs while the remaining 38 were close to genotype II. Conclusion We conclude that more than one genotype of NDV circulates in waterfowl in the Upper Midwest region of the US. Continuous surveillance may help better understand the epidemiology of NDVs maintained in wild bird populations and their relationship to NDVs in domestic poultry, if any.


Background
Avian paramyxoviruses (APMV) belong to genus Avulavirus in the family Paramyxoviridae. The genome of APMV is an approximately 15 kb long, negative-sense, singlestranded RNA molecule. It has six genes that encode for a nucleoprotein (N), a phosphoprotein (P), a matrix protein (M), a fusion protein (F), an attachment protein called hemagglutinin-neuraminidase (HN), and a large polymerase protein (L) [1]. Nine serotypes of avian paramyxoviruses (APMV-1 to APMV-9) have been identified. Of these, APMV-1, also called the Newcastle disease virus (NDV), is the causative agent of Newcastle disease (ND) in poultry. Based on genetic and antigenic analyses of NDV isolates, two major classes (class I and class II) are identified [2,3] and each class has nine genotypes (1-9 genotypes in class I and I-IX in class II) [4,5].
The NDV can cause clinical signs varying from subclinical infections to 100% mortality, depending on the susceptibility of the host and the virulence of the virus. The virus is categorized into velogenic (velogenic neurotropic or velogenic viscerotropic), mesogenic, lentogenic, and asymptomatic enteric strains on the basis of their pathogenesis and virulence. The velogenic strains cause acute fatal infection of chickens of all age groups with clinical findings of nervous signs or extensive hemorrhagic lesions in the gastrointestinal tract. The mesogenic strains are of intermediate virulence and cause moderate respiratory signs with occasional nervous signs while the lentogenic strains cause mild to inapparent infections [1]. The lentogenic strains have been detected in both domestic poultry [6][7][8] and wild bird populations [4,8,9]. Though velogenic strains are considered exotic (exotic Newcastle disease, END) to US poultry, these strains have been isolated occasionally from different avian species in the US [10,11]. During 2002During -2003, California outbreak of END in backyard fowl and commercial poultry resulted in the destruction of about 3.3 million birds and cost $200 million dollars to control the disease [11,12]. Outbreaks of ND have been reported in many countries with considerable economic losses [1]. Such outbreaks warrant continuous surveillance for END in commercial poultry and wild birds.
The surveillance of NDVs in waterfowl is sporadic and often occurs with other monitoring programs such as those for avian influenza viruses (AIV) [13,14]. Wild birds are considered the natural reservoirs of NDVs and mostly harbor lentogenic strains. Studies on genetic diversity among lentogenic strains of NDVs revealed that some of the NDVs from waterfowl and shorebirds were phylogenetically related with NDVs isolated from live-bird markets in the US [4]. It is recommended that epidemiological studies should be continued to determine the prevalence of lentogenic NDVs in wild bird populations [4]. An epidemiological link between isolates recovered from outbreaks in domestic poultry with those obtained from wild bird populations has also been suggested [8,9,15,16]. Therefore, continuous surveillance of wild bird populations may help better understand the NDVs circulating in the environment. This study was conducted to characterize NDV isolates obtained from waterfowl samples. In this study, the cloacal samples from waterfowl from Upper Midwest region of the US were initially screened for AIV by real time reverse transcription-polymerase chain reaction (rRT-PCR); the AIV positive samples by rRT-PCR were inoculated on to the embryonated eggs for virus isolation that yielded NDV in some of them. The NDV isolates were characterized by sequencing to determine the pathotypes and genotypes involved and the changes at the nucleotide and amino acid levels.

Cleavage site analysis
The F gene portion (333 base pairs) corresponding to nucleotide positions 170-502 of GenBank accession number AF217084 was sequenced. Deduced amino acid sequences of the F gene cleavage site were used to determine the pathotypes involved and are shown in Table 1. The fusion gene of virulent NDVs is characterized by the presence of a pair of dibasic amino acids at the cleavage site while in lentogenic strains it is characterized by the presence of monobasic amino acids. None of the isolates had the sequence motif of 111 GR/KRQRK/RF 117 , a characteristic of the virulent strains. All 43 NDVs had an avirulent motif of monobasic amino acids at their F gene cleavage sites. Of the 43 isolates, 23 exhibited sequence motif of 111 GGKQGRL 117 , 19 exhibited the sequence motif of 111 GEKQGRL 117 , and one isolate exhibited the sequence motif of 111 GERQGRL 117 at the cleavage site of F gene.

Phylogenetic analysis
Phylogenetic analysis of partial F gene nucleotide sequences of NDV isolates was done by comparing them with already published F gene sequences of both class I and class II NDVs. None of the isolates clustered with class I NDVs ( Figure 1); all isolates clustered with class II NDVs ( Figure 1). Within class II, all isolates clustered with genotype I or II. Five of the 43 isolates clustered with NDV sequences of genotype I/Ia suggesting them to belong to genotype I (Figure 1) The remaining 38 isolates clustered with genotype II NDVs. These isolates clustered into two groups with 19 isolates in each group. For ease of understanding, we have named these two groups as X and Y ( Figure 1). The isolates in group X were phylogenetically close to genotype IIa NDVs from wild birds from different regions of the US ]. The isolates in group X were also phylogenetically close to a genotype IIa NDV from Argentina [32C/ T.98, AY727881], but the latter was in a different group. None of the already reported NDV sequences of class II     (Figure 1). The sequence homology of genotype II isolates of this study ranged from 95.5% to 100% at the nucleotide level, and the homology as compared to already published sequences of class II genotype II ranged from 90.4% to 100%.

Discussion
This study was conducted to characterize NDVs isolated from waterfowl in the Upper Midwest region of the US.
The initial aim of this study was to isolate and characterize AIV from waterfowl. During the study period, 7458 cloacal samples were collected and of these, 11.9% samples were AIV positive by rRT-PCR. Inoculation of these AIV positive samples in embryonated chicken eggs yielded hemagglutinating viruses and of these, 43 were identified as NDVs by RT-PCR using primer specific for F gene. We were expecting the isolation of AIV rather than NDV on inoculation in embryonated eggs as the samples were initially positive for AIV by rRT-PCR. The possibility of the presence of other hemagglutinating virus(es) in HA positive-AIV negative (by RT-PCR for matrix gene)-NDV negative (by RT-PCR for F gene) allantoic fluid cannot be ruled out and testing of such allantoic fluid is underway in our laboratory. The isolation of NDV from samples that were rRT-PCR positive for AIV indicates that the cloacal sample may have mixed infection with NDV and AIV with concentration of NDV being higher than that of AIV. Hence, the NDV probably overgrew AIV upon inoculation in embryonated chicken eggs. It is to be noted that we tested only AI rRT-PCR positive samples by inoculation in embryonated eggs; testing of more samples might have led to isolation of more NDVs. The isolation of NDV from AIV positive samples indicates the presence of both viruses (AIV and NDV) in waterfowl. The AIV positive allantoic fluid by RT-PCR was not tested for NDV; this testing might provide a better picture of mixed infection of both NDV and AIV. Mixed infection of AIV and NDV in waterfowl has been reported earlier [17,18].
A large amount of sequence data on NDVs isolated throughout the world has been published over the years and is now available for sequence comparison and phylogenetic analysis which can be used to predict the pathotypes and to determine the origin of NDV outbreaks. It has been well established that cleavage of NDV fusion protein is a major determinant for viral virulence. In this study, the F gene sequence of NDVs was used for pathotyping as well as their characterization into different classes and genotypes. None of the isolates was found to be velogenic on the basis of sequence motif of F gene cleavage site. It has been reported that virulent virus has at least one pair of basic amino acids at residues 115 and 116 plus a phenylalanine at residue 117 and a basic amino acid (R) at 113 at the cleavage site whereas lentogenic strains lack dibasic amino acids [19]. All NDV isolates of this study had lentogenic motif at the cleavage site. These results are in agreement with previous studies reporting the detection of lentogenic NDVs in wild birds and domestic ducks [4,9,15,20,21]. None of the isolates had the sequence motif of 111 GERQE/DRL 117 of class I isolates, although the latter have been reported in wild birds and domestic ducks [4,21]. For example, [4] reported seven of the nine genotypes of class I NDVs in waterfowl and shore birds in the US while [21] reported the presence of class I genotype 2 NDVs in domestic ducks in Korea.
Of the 43 isolates, 42 had the sequence motif of 111 GG/ EKQGRL 117 at the cleavage site and were phylogenetically similar to either genotype I or genotype II within class II. This sequence motif has been reported earlier in genotypes I and II of class II NDVs [4]. However, a different sequence motif ( 111 GRRQRRF 117 ) was reported in the lentogenic strains from Australia [22]. One of the isolates had the sequence motif of 111 GERQGRL 117 and this isolate also clustered with class II genotype I strains. This isolate differed from other 42 isolates in the sense that the amino acid lysine was replaced by arginine at position 113.
Overall genotype II viruses were more predominant than genotype I viruses in this study. This finding has the support of [4] who also observed more genotype IIa viruses than genotype I viruses within class II. The NDV isolates in this study were derived only from rRT-PCR AIV positive samples, the possibility of presence of genotypes of both classes (that were not detected in this study) in rRT-PCR AIV negative samples cannot be ruled out. Within class II, the NDV sequences clustered into two different groups. None of the isolates was phylogenetically close to vaccine strains used for comparison. This indicates that in spite of the regular use of live vaccines in poultry throughout the world, their transmission to wild birds may not be a common phenomenon. In an earlier study, [4] also did not detect any vaccine strains in wild birds in the US. Since wild birds have been reported to be a reservoir of NDV [16,23], the mixing of different species at stop-overs during migration and the sharing of common wintering and breeding areas may provide opportunity for virus spread within and between countries and may help perpetuate different genotypes and classes of NDVs in these birds.
The phylogenetic proximity of our isolates with those from the US, China, Korea, and Ireland points to this likelihood.
The presence of class II viruses in wild birds is of concern because this class of viruses has been responsible for several panzootics of Newcastle disease in poultry [24,25].
There are reports suggesting that velogenic NDVs might arise from lentogenic NDVs in nature [23,26]. Further, studies have also suggested that point mutation, and not gene recombination, may be responsible for generating virulent and avirulent strains. For example, the NDV outbreak in Australian poultry during 1998-2000 was caused by a virulent NDV that originated due to mutation in a class II genotype I virus [26]. These authors were of the opinion that lentogenic viruses have the potential to become virulent with the passage of time. Even passaging of NDVs from one host to another has been reported to increase their virulence [16,27]. In addition, the selective forces imposed by a new host environment may also play a role in acquisition of virulence [28]. These findings suggest that the lentogenic strains from wild birds may acquire virulence by waterfowl-to-domestic poultry transmission in nature. In such a scenario we may encounter an NDV outbreak in domestic poultry.
Similar to low pathogenic AIV, the lentogenic NDVs in wild bird populations invariably do not cause obvious disease. Even virulent strains of NDVs that are lethal to chickens, have been isolated from apparently healthy domestic ducks [14,29,30]. Though virulent strains of NDVs were not detected in this study, their presence in the population cannot be ruled out in view of the potential created by the comingling nature and migration patterns of wild birds within and across continents. Thus, continuous surveillance for NDV in wild birds is essential for better understanding of its epidemiology. In conclusion, the present study reveals the circulation of class II (genotypes I and II) lentogenic strains of NDVs in wild birds in the Upper Midwest region of the US. Further studies are needed to determine the true prevalence and implications of various genotypes of NDV within wild bird population.

Conclusion
This study indicates the circulation of class II genotypes I and II NDVs in waterfowl in the Upper Midwest region of the US with an avirulent motif of monobasic amino acids at their F gene cleavage sites. Phylogenetically distant relationship of NDVs of this study with vaccine strains indicates that in spite of the regular use of live vaccines in poultry, their transmission to wild birds may not be a common phenomenon.

Sample collection
Under an NIH funded surveillance program on avian influenza, cloacal and oropharyngeal (OP) swabs were collected from various waterfowl species in Minnesota, South Dakota, and North Dakota from April 2008 to October 2008. The swabs were placed in brain heart infu- sion broth containing antibiotics (penicillin 500 IU/mL, streptomycin 500 μg/mL, neomycin 0.15 mg/mL, fungizone 1.5 μg/mL, and gentamicin 50 μg/mL) and were transported on ice to the laboratory. The initial aim of the project was to test cloacal samples (n = 7458) from waterfowl species for the detection of AIV for which five samples each were pooled and the pools were tested for AIV using rRT-PCR [31]. Individual samples in positive pools were then tested for the detection of AIV by rRT-PCR.

Virus isolation
Individual samples positive for AIV by rRT-PCR (n = 890) were inoculated in 9-day-old specific pathogen free embryonated chicken eggs for virus isolation (VI). Allantoic fluid from inoculated eggs was harvested four days post inoculation and subsequently tested for hemagglutination (HA) using 0.5% turkey erythrocytes. The HA pos-itive allantoic fluids (n = 159) were tested by RT-PCR for the confirmation of AIV as described below.

Total RNA extraction and RT-PCR
Total RNA was extracted from allantoic fluids and a known AIV isolate using QIAamp Viral RNA extraction kit (Qiagen, Valencia, CA). Extracted RNAs were subjected to RT-PCR using primers targeting the matrix gene of AIV [32]. A band of 1027 base pairs was observed in 52 cases indicating them to be AIV. The HA positive allantoic fluids that were negative for AIV (n = 107) were then tested for NDV by RT-PCR. Total RNA extracted from a known APMV-1 was used as a positive control. The RNA was amplified using primers specific to the F gene of NDV [33]. PCR amplification was carried out using Qiagen OneStep RT-PCR kit (Qiagen, Valencia, CA). Amplified PCR products were electrophoresed on 1.2% agarose gel.