Discussion
In the present study, one hundred and sixteen aMPV-B partial G gene sequences, obtained during routine diagnostics in different European Countries in the last few years, were analysed by sequence and phylogenetic analyses in order to molecularly characterize them.
The G gene, which harbours mutations at variable positions between aMPV-B strains, was conveniently amplified by the routine diagnostic RT-PCR protocol employed in the study, and its sequencing proved useful for epidemiological purposes. Furthermore, its variability can give an indication of whether vaccine or field strains are present, making the differentiation relatively easy and cheap to perform.
Live attenuated aMPV vaccines are widely administered to prevent disease in turkeys and chickens, and the recovery of vaccine-derived strains is not unusual both in vaccinated or in unvaccinated flocks (Banet-Noach et al., 2009; Lupini et al., 2011; Chacon et al., 2011; Cecchinato et al., 2013a; Listorti et al., 2014; Arafa et al., 2015; Bayraktar et al., 2019; Andreopoulou et al., 2019).
The 40% of the aMPV-B strains detected in the present study were classified as vaccine-derived strains, being phylogenetically related and showing high nucleotide identity with live commercial vaccine strains licensed in Europe. As expected, vaccine-derived strains formed separate clusters depending on the vaccine strain of origin and were detected in all tested European countries.
A large part of the vaccine-derived strains analysed in the present study was detected in homologous-vaccinated birds, from approximately two to four weeks after vaccination, and only occasionally in unvaccinated birds. Reversion to virulence of aMPV subtype A or B has been previously demonstrated (Catelli et al., 2006; Brown et al., 2011, Cecchinato et al., 2014). Therefore, the detection of vaccine-derived strains closely related to the applied vaccine, concurrently with respiratory signs, could be reliably linked to the vaccine reversion to virulence.
The detection of the strain aMPV/B/France/GuineaFowl/1060/18 in unvaccinated guinea fowls is noteworthy, since its partial G gene sequence shared 100% nucleotide identity with the vaccine strain 11/94. It could be speculated that the vaccine virus could had been introduced as a contaminant by personnel, fomites, vehicles movement or airborne from neighbouring premises. The field veterinarian reported the presence of a turkey farm at approximately 500 meters from the guinea fowl one. Vaccinal strain 11/94 was applied in the turkey flock and the use of the litter from the turkey farm for fertilization of the surrounding crops was reported. The ability of vaccine-derived aMPVs to spread beyond the administration site is well known and it has been proven by Lupini et al. (2011) following a turkey rhinotracheitis outbreak caused by aMPV subtype A in unvaccinated turkeys.
The remaining 60% aMPV-B viruses analysed in the present study were classified as field strains since they clustered separately and showed a low nucleotide identity with vaccines and vaccine-derived strains.
The phylogenetic tree reconstructed with European sequences showed that the virus has continued to evolve from its first appearance in the ’80s. In fact, more recently detected field strains belonged to clades phylogenetically distant from the older field strains, confirming the previously-reported aMPV tendency to evolve over time (Cecchinato et al., 2010).
Unlike vaccine-derived strains, field strains tended to cluster according to their geographic origin, with few exceptions. Distinct clusters were observed for French, Italian, Romanian and Spanish strains, yet some Italian and Greek field isolates clustered closely together, indicating a potential transmission route between these two countries (Tucciarone et al., 2017; Andreopoulou et al., 2019).
The molecular epidemiology of aMPV within each country was analysed in detail reconstructing country-related phylogenetic trees. Heterogeneous field strain populations seemed to co-exist within single European countries, with the only exception of Romania, where all identified strains were part of just one clade. This last finding could be explained by keeping into account that all the processed samples came from different sites within the same company in which viral circulation could be compartmentalized (Franzo et al., 2020).
French, Italian, and Spanish trees showed a rather heterogeneous field strain population, as the strains fell into more than one cluster, divided in several sub-clusters. The heterogeneity was further deduced from the within-group mean p-distance values. The highest value, indicating the highest heterogeneity of nucleotide sequence, was observed for Italian field strains, presumably because of the larger number of sequences available over a broader time period and coming from different commercial poultry companies.
Furthermore, time-related clustering was visible for Italian and Spanish field strains, as strains detected in the last few years tend to form separate clusters or sub-clusters.
As a common finding, recent field aMPV-B strains, belonging to different and distant phylogenetic clades, were detected in the presence of respiratory signs in vaccinated flocks. Due to the G gene variability observed in recent aMPV-B strains, and knowing that the encoded surface G glycoprotein is a key antigen for vaccine-induced immune protection (Naylor et al., 2007) and can evolve in order to avoid vaccine-induced immunity (Catelli et al., 2010; Cecchinato et al., 2010), re-assessment of protection conferred by commercially available vaccines against currently circulating aMPV field strains might be necessary to improve disease control strategies.
The obtained phylogenetic data revealed that aMPV-B strains clustered together irrespective of the host species where the viruses had been detected. A recent experimental challenge study showed that both chickens and turkeys are susceptible to aMPV-B infection with the same virus isolate (Brown et al., 2019). Moreover Cecchinato et al. (2018) reported an outbreak of respiratory disease in guinea fowls caused by aMPV-B strain identical to the ones circulating in the surrounding turkey flocks. Therefore, no evidences are currently available to support a host-specific adaptation of aMPV variants.
As a final remark, the tree reconstructed on European and extra-European strains further confirmed the geographic clustering of field strains. As a rule, field strains grouped together for country location and time period, suggesting a local evolution tendency of the virus that might have taken place after a single introduction event. The majority of recently detected European field strains were located in a big subtree indicating a certain genetic similarity and supporting the wide circulation of a quite homogeneous aMPV subtype B clade in European countries, with the only exception of two clusters of phylogenetically-distant French and Spanish strains.
In conclusion, the molecular characterization of aMPV subtype B and the differentiation between vaccines and field strains through G gene sequence analysis can be a useful tool towards a correct diagnosis and should be routinely applied in order to better address the control strategies. In this respect, current vaccine research is focused on reducing the issues connected to live attenuated vaccine reversion to virulence or the selection of potentially virulent subpopulations (Franzo et al., 2015). Therefore, considering the associated risks, further efforts should be directed at improving administration and biosecurity measures, in order to reduce their prolonged circulation and spreading.
Several research groups have attempted to develop efficacious, more stable and safer next-generation vaccines (Qingzhong et al., 1994; Tarpey et al., 2001; Kapczynski and Sellers, 2003; Kapczynski, 2004; Chary, Njenga, & Sharma, 2005; Yu et al., 2013; Hu, Roth, Zsak, & Yu, 2017) but, despite this effort, live vaccines still provide the greatest protection after homologous challenge.
Promising reverse genetics systems for aMPV have been developed in recent years for A, B and C subtypes (Naylor et al., 2004; Yu et al., 2010; Laconi et al., 2016) other than being an exceptional tool for the study of the virus properties (Brown et al., 2011), can be used for the development of rationally modified aMPV vaccines (Naylor, Lupini, & Brown, 2010) or recombinant vaccines (Falchieri et al., 2013) expressing foreign genes.
Conflict of Interest Statement
The authors declare no conflict of interest.