Full genome sequence analysis
To verify and further analyze selected NDV isolates, full genome sequences were generated for both vaccine type clones (R1954/2011-cl 1, R1973/2011-cl 2) as well as for genotype 2.XXI (R1973/2011-cl 4) and 2.VII.1.1 (AR2178/2014) isolates. The genomes obtained for the lentogenic NDV clones R1954/2011-cl 1 and R1973/2011-cl 2 were 15.135 and 15.150 nt long, whereas the genomes of the virulent strains AR2178/2014 and clone R1973/2011-cl 4 were 15,179 and 15,165 nt long with a 6-nt-insertion ACCCCC or CCTCTC respectively in the untranslated region (UTR) downstream of the NP gene.
The two NDV strains of R1954/2011-cl 1 and R1973/2011-cl 2 showed 99% nt similarity with the NDV strain LaSota C5 (accession: KC844235) (Table 2). On the other hand, the velogenic NDV strain AR2178/2014 showed the highest nt sequence similarity of 99% to the Chinese NDV strain SD04/2011 (accession: JQ015296), whereas strain R1973/2011-cl 4 showed 94% nt similarity with the Japanese NDV strain Osaka/2440/1969 (accession: AB853926).
Phylogenic analysis of the full F gene (Fig. 1) as well as the whole genome sequence (Fig. 2) confirmed the phylogrouping obtained by partial F-gene analysis (Fig. S1). Genome organization of the four fully sequenced NDV strains was found to be identical to other Avian orthoavulavirus -1, containing six genes in the order of 3′-NP-P-M-F-HN-L-5′, encoding at least seven proteins; N, P, V, M, F, HN and L. Analysis of the F protein of the four NDV strains revealed that three strains, namely, R1954/2011-cl 1, R1973/2011-cl 2 and AR2178/2014 (genotype 2.II, and 2.VII.11, respectively) possessed six potential N-glycosylation sites (NGS); five within the ectodomain at amino acid (aa) positions 85, 191, 366, 447, and 471 and one in the cytoplasmic domain at position 542. Strain R1973/2011-cl 4 (genotype 2.XXI) possessed the same five ectodomain NGS and an additional one at position aa 4 (https://prosite.expasy.org/PDOC00001 ).
The HN proteins of the two NDV strains R1973/2011-cl 4 and AR2178/2014 (genotype 2.XXI and 2.VII.1.1, respectively) were 571 aa long, which is 6 aa shorter than those of the NDV strains of genotype 2.II, namely, R1954/2011-cl 1 and R1973/2011-cl 4 (577 aa). The HN protein of all genotypes possessed five potential NGS. Four NGS at positions 119, 341, 433 and 481 were common in the four NDV strains, whereas NDV strains R1973/2011-cl 4 and AR2178/2014 (genotype 2.XXI and 2.VIIb, respectively) possessed an additional NGS at site 508, and strains R1954/2011-cl 1 and R1973-11-cl 2 (genotype 2.II) at position 538.
Comparing the level of homology of the different isolates to the La Sota vaccine strain (Accession No. AF077761) confirms the close relation in between the genotype 2.II viruses. Regarding genotype 2.XXI and 2.VII, both genotypes accumulated mutations compared to La Sota vaccine strain, but these changes were present for both viruses to a comparable level with homology of 81.0 to 85.5%. For the different genes, no striking differences are evident. Comparing the amino acid (aa) composition of genotype 2.XXI and 2.VII.1.1, the homology to 2.II strain La Sota is higher than on the level of nucleotides, with 92 to 87.9 % for NP, M, F, HN and L-protein. It is striking that the homology of the P-protein was considerably lower than the other genes with 80.1 and 81.9% for genotype 2.XXI and 2.VII.1.1 respectively. When focusing on the HN protein, responsible for receptor mediated binding, our data reveal overall aa differences of 11,7% and 11,9 % for 2.XXI and 2.VII.1.1 viruses respectively compared to LaSota vaccine type (Table.3). Mutations accumulated within the protein fragment encompassing the tail, transmembrane part and stalk region (aa 1-125; 22,4% and 25,6%), whereas the globular head (aa 126-570), bearing relevant antigenic sites, was underrepresented and had an aa divergence of only 8,8% and 8,1 % for genotypes 2.XXI and 2.VII.1.1 respectively. Among the aa in the globular head, 21 of 39 and 35 substitutions were identical for both 2.XXI and 2.VII.1.1 viruses. These changes include 6 single point mutations in sites that were described to be part of neutralization sensitive epitopes (Y203H, N263K, N347K, E494D, G495E and V514E) of the HN-protein (Table 3), i.e. site site 23; site 3, site I, site12, site II, site 2 (Iorio et al., 1989; Iorio et al., 1991) (Iorio). In an additional six positions both viruses had changes compared to LaSota (aa, 266, 288, 310, 433, 443, 509) but differed in aa composition. Unique changes were evident on 9 and 10 positions for genotype 2.XXI and 2.VII.1.1 viruses respectively. None of those isolate-specific sites affected known functional sites directly. For the vaccine type isolates only R1973/2011-cl 1 had 5 alterations, that were all located within the globular head of the HN-protein, with mutations in epitope 23 (Y203H) and NA binding site (A502V) that were identical for the genotype 2.XXI and 2.VII.1.1 viruses.