Molecular detection, isolation and characterization of Peste‐des‐petits ruminants virus from goat milk from outbreaks in Bangladesh and its implication for eradication strategy

Abstract Peste‐des‐petits ruminants (PPR) is a highly contagious transboundary viral disease of small ruminants, which is endemic in much of Africa, the Middle East and Asia. In South Asia, PPR is of significant concern to the Indian subcontinent including Bangladesh as more than 30% of the world's sheep and goats are farmed in this region, predominantly by small, poor and marginal farmers. PPR virus was detected and isolated from goat milk from field samples from PPR outbreaks (2012–2015) in Bangladesh and its full‐length sequences obtained. Sequence analysis of the partial N gene of Bangladesh isolates showed 99.3%–100% identity whereas 98.2%–99.6% identity was observed when compared with neighbouring Indian viruses. Further analysis of the full‐length genomes indicated that the Bangladesh isolates were 99.3%–99.99% identical among themselves and 98.3%–98.4% identical to neighbouring Indian viruses. These findings further support the transboundary transmission of PPR virus across the Indian and Bangladesh border. In additional, the establishment of a cross‐border strategy between India and Bangladesh will be of paramount importance for the eradication of PPR in this region. Molecular detection and isolation of PPR virus from milk is of significant potential concern for spread of the disease to free areas as the major producers of goat milk globally are PPR endemic countries in particular India and Bangladesh, as well as Sudan. Milk is a noninvasive sample type and bulk goat milk sampling for the detection of PPRV would be of practical significance for regional surveillance of PPRV as progress is made towards the targeted 2030 eradication.


| INTRODUCTION
Peste-des-petits ruminants (PPR) is the most important OIE listed disease of small farmed ruminants in the developing world (Baron, Parida, & Oura, 2011;Parida et al., 2016). The etiological agent, PPR virus (PPRV) is a member of the family Paramyxoviridae and genus Morbillivirus (Banyard et al., 2010). Following the eradication of rinderpest, PPRV has been identified by the Food and Agriculture Organisation (FAO) and World Organisation for Animal Health (OIE) as the next target for eradication by the year 2030. PPRV exists as single serotype, which groups into four distinct lineages (I-IV) based on sequence comparison of the C-terminus of the N gene (Couacy-Hymann et al., 2002) and F gene (Forsyth & Barrett, 1995). PPR was first identified in Cote d'Ivoire (Ivory Coast) in 1942 as an entity distinct from rinderpest (Gargadennec & Lalanne, 1942). With the notable exception of most southern African countries (South Africa, Botswana, Namibia, Zimbabwe, Mozambique and Malawi), it is now recognized to be endemic throughout Africa as well as the Middle East, Central, East and south Asia. Lineage IV is the primary circulating lineage of PPRV in the Middle East and Asia, with recent incursions into China and Tibet Wang et al., 2009) into North (Baazizi et al., 2017;Fakri et al., 2016;Muniraju et al., 2013), Central (Maganga et al., 2013), and East Africa as far south as Tanzania (Lembo et al., 2013;Mahapatra et al., 2015).
Bangladesh is home to the 5th largest population of goats with more than 55 million animals estimated by the FAO in 2014, behind China, India, Nigeria and Pakistan, and has the largest population by land mass (FAO, 2016). The overwhelming majority of these animals are raised in small holdings by poor, marginal and subsistence farmers (Sarker & Islam, 2011). Goat meat makes up greater than 38% of the total meat production in Bangladesh and more than 11% of the goat milk produced globally is produced within Bangladesh (FAO, 2016), and greater than 55% of the milk consumed annually in Bangladesh is from goats, as such the containment of PPRV is of significant concern to the region. This publication describes for the first time the molecular detection as well as isolation and molecular characterization of full-length PPRV from goat milk (noninvasive sample). Further, Bayesian analysis of full-length PPRV genomes and neighbourhood-joining phylogenetic analysis of partial N gene of PPRV from various outbreaks in Bangladesh between 2012 and 2015 were included in the study.

| Sample collection
Samples were collected across a 3-year period from 8 locations (Figure 1) as part of routine diagnostic procedures for PPR in Bangladesh (Table 1). Samples included nasal swabs, tissue samples predominantly from lung, as well as milk and faecal samples.
Selected samples (n = 19) were shipped on dry ice to The Pirbright Institute for confirmation of diagnosis and molecular testing.

| Virus isolation
Attempts were made to isolate virus from tissue samples, nasal swabs, as well as faecal material and from goat milk. The tissue samples and nasal swabs were processed as described previously . For faecal material, where solid pellets were present approximately 1 g of faecal matter was homogenized in 3 ml of M25 buffer using a mortar and pestle; from diarrhoea faecal samples approximately 1 ml of material was diluted into 3 ml of M25 supplemented as above and any solid fragments triturated with mortar and pestle. Milk samples were diluted 1:10 in PBS with antibiotics as above. Homogenates were clarified by centrifugation at 1000 9 g for 15 minutes at +4°C and 500 ll of the supernatant was inoculated onto 70% confluent Vero dog slam cells (VDS) and incubated for 2 hr at 37°C in an atmosphere of 5% CO 2 , before the inoculant was replaced with 5 ml of Dulbecco's Modified Eagle's medium (DMEM) supplemented with 5% foetal calf serum (FCS). The cells were incubated for up to 7 days and blindly passaged or until cytopathic effects (CPE) were observed. The samples were passaged at least five times before declaring negative.
2.4 | RNA extraction, reverse transcription (RT), polymerase chain reaction (PCR), real-time RT-PCR (qRT-PCR) and sequencing Total RNA was extracted from the homogenized tissue samples, faecal matter, nasal swabs and milk samples using Trizol TM (Invitrogen) as per the manufacturer's instructions. In addition, the eluted RNA from faecal samples was further purified using the RNeasy mini RNA Extraction Kit (Qiagen) to remove PCR inhibitors present in faecal material following the manufacturer's protocol after dilution to 100 ll in nuclease-free water. RT-PCR to amplify the C-terminus of the N gene was carried out as previously described (Baazizi et al., 2017). In additional, milk samples were analyzed by qRT-PCR to assess the viral load (Batten et al., 2011) using Superscript III Platinum R one step qRT-PCR system kit (Invitrogen).
For full-length genome sequencing a hemi-nested RT-PCR was performed on tissue samples as described previously  and amplification of the terminal 5 0 and 3 0 ends of the PPRV genome was accomplished via RACE, as previously described (Bao et al., 2012;. The PCR amplicons were purified and sequenced as previously described .

| Sequence analysis
Both partial N and full-length sequences were assembled and analyzed using SeqMan pro (DNAStar Lasergene 13.0). Nucleotide sequences of the viruses were aligned using the CLUSTAL X multiple sequence alignment programme (Thompson, Gibson, & Higgins, 2002) or MUSCLE as appropriate (Edgar, 2004).
For sequence data not generated in this study complete PPRV genome sequences (Supporting information Table S1; n = 37) were obtained from GenBank (as on 15/12/2017). Sequences obtained from live attenuated vaccine virus strains (India/Sungri 96: KJ867542, KF727981 and Nigeria 75/1 (X74443, HQ197753) were removed prior to analysis, these sequences have previously been shown to substantially skew phylogenetic analyses (Muniraju, Munir, Parthiban et al., 2014). To identify the nearest common ancestor and hence likely dates of divergence, the Bangladesh sequence was compared using the coalescent-based Bayesian Markov chain Monte Carlo (MCMC) (Drummond & Rambaut, 2007;Drummond, Suchard, Xie, & Rambaut, 2012) approach to all available full-length PPRV wild-type genomes available (Supporting information Table S1). The general time-reversible nucleotide substitution model with gamma distribution and invariant sites was selected on the basis of Bayes factor results following path sampling (data not shown). Path sampling was performed until the marginal likelihood estimate remained constant (Nr = 16). As has been previously determined the relaxed uncorrelated exponential distribution (UCED) clock model (Drummond, Ho, Phillips, & Rambaut, 2006) was the best fit to PPRV complete genomes (Muniraju, Munir, Parthiban et al., 2014;Parida et al., 2015). As there are very few (n = 4) (KR261605, KT270355, KR140086, and KX033350) full-length sequences available for the surrounding region, further phylogenic analysis was undertaken using the PPRV partial N gene sequence data of the C-terminal region of the N gene.
Partial N sequences to be included in the analysis were selected from GenBank on the basis of accurate annotations including locations and dates of sampling as well as uniqueness. Sequences which  Table S2) to which the 13 sequences generated in this study were added, making it 184 in total. The partial N dataset was aligned using MUSCLE and phylogenetic analyses were performed using MEGA6 (Tamura, Stecher, Peterson, Filipski, & Kumar, 2013). The neighbour-joining tree was generated using the Kimura 2-parameter model and tests for phylogeny performed using the bootstrap method with 10,000 replications and the gaps/missing data removed by pairwise deletion (Kimura, 1980).

| RESULTS AND DISCUSSION
3.1 | Genome detection, virus isolation from PPR infected goat milk and its implication All samples were tested for PPRV using primers targeting the highly variable C-terminus of the N gene followed by hemi-nested PCR using the same primer pair if initial PCR was negative ( Table 2). The faecal samples were found to be weak positive in PCR. Milk sample analysis by real-time RT-PCR revealed high viral load (Table 2).
At present, there are limited data (Wasee Ullah et al., 2016) regarding the successful isolation of infectious virus from faecal material and no attempts are currently documented as to the isolation of virus from milk. Therefore, samples which were positive in PCR were inoculated onto VDS cells and passaged. Virus was successfully isolated from a total of three milk samples ( milk is predominantly produced by small-holders and shipped to regional co-operative processing plants (Hemme, Garcia, & Khan, 2004). This movement of milk and the associated equipment and personal is a possible source of fomites as has been observed for foot-and-mouth disease virus (FMDV) (Donaldson, 1997;Reid et al., 2006). However, these larger milk processing facilities also provide an ideal location for testing for the presence of PPRV genome within the region and the establishment of a robust calibrated test either via conventional or quantitative PCR should be prioritized for bulk milk samples. Although the number of samples tested in this study is relatively low, the load of PPR virus genome detected in goat milk was similar to that of FMDV as reported by Reid et al. (2006). Bulk milk testing has been proposed for surveillance of FMDV (Reid et al., 2006) 1998(95% HPD 1978-1989. As there are so few full-length sequences available from the local region, a further 11 partial N sequences (255 nt) from Bangladesh collected between 2012 and 2015 were sequenced ( Table 2). The partial N gene sequence of Bangladesh/B18/Nihkanchari/2015 was identical to that of Bangladesh/B19/Nihkanchari/2015; therefore was excluded from the analysis. These sequences (n = 10) were added to the equivalent regions extracted from the Bangladesh fulllength sequences (n = 3) (total sequences from this analysis n = 13) and 331 global partial N sequences of which almost half were from the surrounding region (Bangladesh n = 51, India n = 42, Pakistan n = 27, China including Tibet n = 30). Partial N sequences were extracted (as on 15/12/2017) from the GenBank repository F I G U R E 3 Neighbourhood-joining tree using partial N gene sequences. Accession number, country of origin and sampling year of each isolate is shown. All sequences generated in this study are highlighted in red and isolates from the surrounding transboundary region of India are highlighted in green, and closely associated isolates from Pakistan highlighted in blue. All sequences generated in this study have been submitted to NCBI and awaiting accession numbers (Supporting information Table S2) representing the available partial N sequences in GenBank for which accurate annotation details are available, and phylogenetic tree generated and annotated with bootstrap values ( Figure 3). As has been observed previously (Munir et al., 2012;Muthuchelvan et al., 2014), the Bangladesh viruses cluster most closely with viruses from India, China, Tibet, Pakistan, and Iran. In particular, there are extremely strong relationships between viruses from the Indian border region of Tripura and from the Narayanganj and Netrokona outbreaks in Bangladesh, which have been previously sequenced (Muthuchelvan et al., 2014), as well as the isolates sequenced in this study. It is interesting that of the available Pakistani isolates (n = 27) two virus isolates both isolated from camels from the 2012 outbreak also grouped very strongly with virus isolates from Bangladesh ( Figure 3). As Pakistan and Bangladesh do not share a border this further emphasizes the importance of the establishment of an effective regional approach to PPR eradication. This is of particular concern due to the porous nature of the border between India and Bangladesh to prevent the reoccurring transmission of PPR both between these nations but also further afield.
To conclude, this work describes the molecular detection of PPRV genome as well as isolation of virus from noninvasive samples (goat milk) collected from PPR outbreaks in Bangladesh.
While there is currently no evidence for the direct transmission of PPRV through milk, it seems a likely pathway of vertical transmission of PPRV to kids and may be an additional factor in the high prevalence of PPRV mortality among kids (Taylor, 1984). Further investigations are required as to the possible transmission of PPRV between animals from goat milk. In particular, the length which virus remains present in milk and the effect of pasteurization on PPRV viability, as this data will have important implications for the development of effective controls for the export of milk products from PPR endemic regions as well as the development of testing methods for bulk milk storage. In additional, we have sequenced the full-length viral genome of PPRV from milk and tissue samples from three isolates as well as the partial N gene sequence from a further ten isolates and used Bayesian phylogeography to demonstrate the transboundary nature of PPRV infection in the Indian subcontinent and further afield. The close relationships between viruses from Pakistan and Bangladesh serve in particular to emphasize the transboundary nature of PPRV as these countries do not share an immediate border, and highlight the importance of regional approaches to PPR control and eradication.