Molecular characterization of Nipah virus from Pteropus hypomelanus in Southern Thailand

Background Nipah virus (NiV) first emerged in Malaysia in 1998, with two bat species (Pteropus hypomelanus and P. vampyrus) as the putative natural reservoirs. In 2002, NiV IgG antibodies were detected in these species from Thailand, but viral RNA could not be detected for strain characterization. Two strains of NiV (Malaysia and Bangladesh) have been found in P. lylei in central Thailand, although Bangladesh strain, the causative strain for the outbreak in Bangladesh since 2001, was dominant. To understand the diversity of NiV in Thailand, this study identified NiV strain, using molecular characterizations, from P. hypomelanus in southern Thailand. Findings Pooled bat urine specimens were collected from plastic sheet underneath bat roosts in April 2010, and then monthly from December 2010 to May 2011 at an island in southern Thailand. Five in 184 specimens were positive for NiV, using duplex nested RT-PCR assay on partial nucleocapsid fragment (357 bp). Whole sequences of nucleocapsid gene from four bats were characterized. All 5 partial fragments and 4 whole nucleocapsid genes formed a monophyletic with NiV-MY. Conclusions Our study showed that P. hypomelanus in southern Thailand and from Malaysia, a bordering country, harbored similar NiV. This finding indicates that NiV is not limited to central Thailand or P. lylei species, and it may be a source of inter-species transmission. This indicates a higher potential for a widespread NiV outbreak in Thailand. NiV surveillance in Pteropus bats, the major natural reservoirs, should be conducted continuously in countries or regions with high susceptibility to outbreaks.


Background
Nipah virus (NiV) has been genetically characterized as Malaysia (NiV-MY) or Bangladesh (NiV-BD) strains, termed based on the country of their first outbreak, with fruit bats as the main reservoir [1]. Based on currently available complete N gene sequences, NiV-MY was obtained from Malaysia (Pteropus hypomelanus, P. vampyrus, sick pigs and patients) and Cambodia (P. lylei), while NiV-BD was obtained from patients from Bangladesh and India. Outbreak in Malaysia of NiV-MY in 1998 contributed to encephalitis with 39 % fatality rate [2]. Recent encephalitis outbreak (most likely caused by NiV) in Southern Philippines (2014) originated from fruit bats, transmitted to horses and then to humans, which was attributable to either horse slaughter or horse meat consumption [3]. Secondary man-to-man spread was evident, with acute encephalitis, severe influenza-like illness, or meningitis (82 % fatality). NiV-BD has been associated with the Indian (2001) and Bangladeshi (since 2001) outbreaks of encephalitis and respiratory distress syndrome, with a fatality rate of over 70 % [4]. Infectivity and pathogenicity of the two strains may be different, as NiV-BD seems to be associated with higher incidence of respiratory disease. A study found higher level of oral shedding in ferrets infected with NiV-BD than with NiV-MY [5]. However, both strains caused similar respiratory tract lesions in Syrian hamsters [6]. On the other hand, disease does not develop in Pteropid bats, whether infected naturally or experimentally [7][8][9].
Serological studies have demonstrated evidences of NiV infection in multiple bat species, including frugivorous and insectivorous bats. However, viral isolation and molecular characterization was mostly only successful in Pteropus species [10]. NiV-MY was isolated from urine samples of P. hypomelanus [7] and P. vampyrus [11] from Malaysia. NiV-MY has also been found in other Pteropus bats from various countries; P. lylei in Cambodia, nucleocapsid (N) sequence of 1599 bp, shared 98 % identity with the Malaysian patient (AF212302) [12] and P. vampyrus in Indonesia, matrix gene of 251 bp had 100 % identity to NiV-MY in the Malaysian patient (AF212302) and 99.6 % homology with the Malaysian P. vampyrus bat (FN869553) [13]. P. giganteus is believed to be the source of NiV infection in India and Bangladesh via contaminated date palm Fig. 1 Locations in Thailand where bat urines specimens were collected from P. hypomelanus at the south (blue circle) and P. lylei at central (red circle) sap. NiV-BD has been detected from a liver homogenate of P. giganteus captured in India with its partial N gene (205 bp), showing 100 % homology with NiV infected patients in India (FJ513078.1) and Bangladesh (AY988601) [14]. On the other hand NiV was detected in P. lylei from Thailand, with sequences (357 bp) sharing 98-99 % nucleotide homology with Bangladeshi patients (AY988601) [15], but there has been no report of an outbreak in humans or other animals.
The intensive surveillance of NiV in P. lylei in central Thailand has been conducted yearly since 2002. To date, NiV RNA was only detected in P. lylei among two other Pteropus species in Thailand. Fifty two (88.1 %) NiV-BD and 7 (11.9 %) NiV-MY were reported between 2002 and 2008 [15][16][17]. NiV IgG antibody has been found in P. hypomelanus and P. vampyrus from southern Thailand, but NiV RNA could not be successfully detected [16]. The objective of this study was to identify and characterize the strains of NiV found in P. hypomelanus, in southern Thailand.

Methods
Survey location was an island in Southern Thailand, a National Park, where only P. hypomelanus bats roosted in trees along the beach (Fig. 1). Numbers of bat population varied each month due to rain storms (Table 1). Bat urine samples were collected under roost trees with permission from the Department of National Parks, Wildlife and Plant Conservation (No. 0907.1/20713). The samples were collected once in April 2010, and then monthly from December 2010 to May 2011, using a plastic sheet protocol as previously described [18]. Two swabs of pooled bat urine collected from plastic sheet were stored in Lysis buffer (bio-Mérieux), transported on ice to the laboratory within 48 h and stored at −80°C until analysis. The nucleic acid was extracted from bat urine using the NucliSENS easyMAG® extraction kit (bioMérieux). NiV RNA was first screened by duplex nested RT-PCR as previously described [17]. Only PCR positive specimens were further characterized for whole N gene by 3 additional PCR assays. All PCR positive specimens were directly sequenced on hemi-nested PCR products using an automated ABI PRISM 377 model sequencer. The sequence segments were assembled by BioEdit program [19]. Phylogenetic trees were generated by using maximum-likelihood method based on 357 bp (Fig. 2a) and 1599 bp (Fig. 2b) of N gene.
In addition, whole N gene sequences from six urine specimens of P. lylei, collected during 2010-2011 from central Thailand (see the map in Fig. 1), were included in the study. The sample collection and amplification methods were the same as for P. hypomelanus."
Phylogenetic analyses using maximum-likelihood methods from 357 and 1599 bp of N gene are shown in Fig. 1a and b respectively. Sequences were analyzed with available NiV sequences from bats, humans and pigs in GenBank, and six additional whole N gene sequences of NiV-BD obtained in 2010-2011 from urine of P. lylei bats in central Thailand. The NiVs from P. hypomelanus in this study form a monophyletic clade with other NiV-MY ( Fig. 1a and b). All of NiV-MY (partial N gene, Fig. 2a) from P. lylei clustered with P. hypomelanus NiVs from southern Thailand (this study) and Malaysia. Further, the similarity between the whole NiV-BD N gene from Thai P. lylei (KT163251-KT163256) and Bangladeshi patient (AY988601) was 99.1-99.2 % (Fig. 2b). The nucleotide identity between NiV-N gene (205 bp) from P. giganteus in India [14] (JF899339, not included in the phylogenetic tree) and NiVs found in Thailand from P. hypomelanus (no.s 1353, 1753, 8554, 16454) and P. lylei (no. s 1054, 3054, 15053, 15553, 14653, 33654) were 97.1 and 99.5 % respectively.
The visual alignment of the amino acid (aa) sequences was constructed from 533 aa of 1599 nucleotides of N gene open reading frame sequences (Fig. 3). The aa difference of N gene between strains NiV-MY and -BD sequences were found at 5 domains (positions 387, 505, 506, 508 and 521). The N-aa sequences between sick pigs and patients from Malaysia were identical, while one aa sequence difference was observed in NiV isolate from P. hypomelanus in Malaysia (Fig. 3). There were 5-6 aa sequence differences between full-length N in NiV isolates from P. hypomelanus from southern Thailand and P. hypomelanus in Malaysia. However, they had only 1-2 aa sequence differences when compared with NiV isolate from P. vampyrus in Malaysia, and 4-5 aa sequence differences with NiV from P. lylei in Cambodia. Interestingly, N-aa sequences among four NiV-BD isolates from P. lylei in central Thailand (GenBank accession nos. KT163251-KT163254) and from Bangladeshi patient (GenBank accession no. JN808861) were identical (Fig. 3).

Discussion and conclusion
Whether P. hypomelanus or P. vampyrus bats were the spill-over hosts in the Malaysian outbreak has been unclear. Evidence in support of P. hypomelanus as the spill-over host is based on the near identical NiV sequence from this species to the other three NiVs isolated from human patients, and it was isolated shortly after the Malaysian outbreak [7]. In addition, the amino acid sequences obtained from NiV's whole genome differed from NiV from two infected pigs at merely 3 positions [20]. Nonetheless, NiV sequences from P. vampyrus bats have more differing nucleotides to the isolates from the patients during the outbreak than NiV from P. hypomelanus, 98 and 6 nucleotides, respectively.
Our study at the southern island confirmed the similarity between NiV-MY from P. hypomelanus from Thailand and the Malaysian outbreak (in humans and pigs), and was in accord with the finding from one P. hypomelanus from Tioman island, Malaysia, about 960 km from our surveyed site [7]. From our finding, it remains possible that P. hypomelanus is restricted to NiV-MY, as only this strain was found in the 5 pooled urine specimens positive for NiV .   S  30  155  188  191  211  318  380  387  406  429  414  432  436  450  457  481  502  505  506  508  511  518    Unfortunately, there is limited data to support host specificity as only one NiV-MY from P. hypomelanus has been reported from Malaysia [7]. A comprehensive analysis of bats across their migratory routes in Indo-Australian region should be conducted to clearly understand the ecology and transmission of NiV in P. hypomelanus. The co-circulation of NiV-BD and NiV-MY strains within population of P. lylei in central Thailand [15] remains a mystery. High percent similarity of N nucleotide sequences from P. lylei and P. giganteus, and NiV patients in Bangladesh and the identity of whole N-aa among NiV isolates from P. lylei in Thailand and NiV patients in Bangladesh (Fig. 3), extrapolates the possibility of P. lylei bat in Thailand as a source of NiV infection to humans or other animals. The phylogenetic analysis on 357 bp (Fig. 2a) revealed that NiV-MY in P. lylei from Thailand, and P. hypomelanus from Malaysia and Thailand clustered together. This finding introduces the possibility of NiV transmission between P. lylei and P. hypomelanus, in addition to the previously suggested transmission between P. vampyrus and P. lylei [11]. Furthermore, the higher percent similarity of NiV N-aa between P. hypomelanus in Thailand and P. vampyrus in Malaysia than P. hypomelanus in Malaysia, raise the possibility of transmission between P. hypomelanus and P. vampyrus. Molecular characterization of NiV N and other genes, particularly those encoding the envelop glycoprotein from P. vampyrus in Thailand is required to investigate this possibility.
In conclusion, detecting NiV-MY strain in P. hypomelanus in southern Thailand, whose sequences share similarities to that identified in the Malaysian outbreak, and to P. lylei bats in central Thailand, suggests a potential role of P. hypomelanus as a possible source for inter-species transmission. This may result in the emergence of NiV infection to other animals and humans in both southern and central Thailand. Continuing surveillance of NiV infection in bats, humans, and pigs is vital for early detection and can potentially reduce the scale of an outbreak.

Ethics approval and consent to participate
Bat urine samples were collected with permission from the Department of National Parks, Wildlife and Plant Conservation (No. 0907.1/20713). The urine specimens were collected under roost trees, no bat was caught for sample collection.

Consent for publication
Not applicable.

Availability of data and materials
The GenBank accession numbers for Nipah virus sequences reported in this paper are: KT163247-KT163250 for whole N gene and KT163257 for 357 bp partial N gene from Pteropus hypomelanus and KT163251-KT163256 for whole N gene from P. lylei.