key: cord-0999621-nbv8sfoo
authors: Xu, Lin; Wu, Jianmin; Li, Qi; Wei, Yamei; Tan, Zhizhou; Cai, Jianqiu; Guo, Huancheng; Yang, Ling’en; Huang, Xiaohong; Chen, Jing; Zhang, Fuqiang; He, Biao; Tu, Changchun
title: Seroprevalence, cross antigenicity and circulation sphere of bat-borne hantaviruses revealed by serological and antigenic analyses
date: 2019-01-22
journal: PLoS Pathog
DOI: 10.1371/journal.ppat.1007545
sha: 53847cf95615b651c07388b52222211db0b4d303
doc_id: 999621
cord_uid: nbv8sfoo

Bats are newly identified reservoirs of hantaviruses (HVs) among which very divergent HVs have been discovered in recent years. However, their significance for public health remains unclear since their seroprevalence as well as antigenic relationship with human-infecting HVs have not been investigated. In the present study archived tissues of 1,419 bats of 22 species from 6 families collected in 5 south and southwest provinces in China were screened by pan-HV RT-PCR following viral metagenomic analysis. As a result nine HVs have been identified in two bat species in two provinces and phylogenetically classified into two species, Laibin virus (LAIV, ICTV approved species, 1 strain) and Xuan son virus (XSV, proposed species, 8 strains). Additionally, 709 serum samples of these bats were also analyzed by ELISA to investigate the seroprevalence and cross-reactivity between different HVs using expressed recombinant nucleocapsid proteins (rNPs) of LAIV, XSV and Seoul virus (SEOV). The cross-reactivity of some bat sera were further confirmed by western blot (WB) using three rNPs followed by fluorescent antibody virus neutralization test (FAVNT) against live SEOV. Results showed that the total HV seropositive rate of bat sera was 18.5% (131/709) with many cross reacting with two or all three rNPs and several able to neutralize SEOV. WB analysis using the three rNPs and their specific hyperimmune sera demonstrated cross-reactivity between XSV/SEOV and LAIV/XSV, but not LAIV/SEOV, indicating that XSV is antigenically closer to human-infecting HVs. In addition a study of the distribution of the viruses identified an area covering the region between Chinese Guangxi and North Vietnam, in which XSV and LAIV circulate within different bat colonies with a high seroprevalence. A circulation sphere of bat-borne HVs has therefore been proposed.

Introduction Hantaviruses (HVs), members of the genus Orthohantavirus within the family Hantaviridae in the order Bunyavirales, are responsible for two major life-threatening diseases in humans: hemorrhagic fever with renal syndrome (HFRS) in Eurasia and hantavirus cardiopulmonary syndrome (HCPS) in the Americas [1] . Every year around 100,000 HFRS cases and 1,000 HCPS cases are reported worldwide [2] . China suffers severely from epidemic HFRS; in 2017 alone, official statistics reported 11,262 cases with 64 deaths [3] .

HVs are predominantly carried and transmitted by rodents, but insectivores and bats have also been reported as hosts. Several bat-borne HVs are presently known, which show large genetic diversities from currently known rodent-and insectivore-borne HVs. The first reported bat-borne HVs, Magboi virus (MGBV) and Mouyassué virus (MOUV), were identified respectively in Sierra Leone and Côte d'Ivoire of Africa in 2012 [4, 5] . Then two bat-borne HVs, Longquan virus (LQUV) and Huangpi virus (HUPV), were reported in China in 2013 [6] , followed by the detection of Xuan son virus (XSV) at three locations in North Vietnam [7, 8] . We reported the first complete genome of a bat-borne HV, Laibin virus (LAIV), identified from a black-bearded tomb bat in Guangxi Province of China in 2015 [9, 10] . Since then three more complete genomes of bat-borne HVs, Makokou virus (MAKV), Quezon virus (QZNV) and Brno virus (BRNV) have been reported sequentially in 2016, in Central Africa (Gabon), Southeast Asia (Philippines) and Central Europe (Czech Republic), respectively [11] [12] [13] . Most recently, a sister lineage of MOUV was detected in dried blood samples from bats in Eastern Africa (Ethiopia) in 2017 [14] . Of these HVs only three, Laibin, Longquan and Quezon viruses were approved as bat-borne HV species within genus Orthohantavirus in the 10 th report of International Committee on Taxonomy of Viruses (ICTV) released in 2017 [15] . Phylogenetic analysis of bat-borne HVs has indicated that bats might be the natural original hosts of HV: i.e., the viruses first appeared in bats or insectivores, then emerged in rodents [6, [16] [17] [18] [19] . However, due to lack of sufficient bat-borne HV genomic sequences, their evolutionary phylogeny and genetic diversity as well as biological features are poorly understood.

HVs are enveloped and spherical in shape although pleomorphic forms are also found with the diameters ranging from 80-120 nm. Within the capsid is a tripartite negative-stranded RNA genome consisting of small (S), medium (M) and large (L) segments with a total length of about 11.8 kb, respectively encoding nucleocapsid protein (NP), glycoprotein (GP, a precursor for two viral surface glycoproteins, Gn and Gc) and RNA-dependent RNA polymerase (RdRp) [1] . The NP is multifunctional and plays an essential role in viral replication, not only binding viral RNA strands to form a ribonucleoprotein (RNP) to prevent RNA from degradation, but also regulating virus replication and assembly [1, 20, 21] . NP is also the main target for the earliest immune response. Its coding gene is much more conserved than the GP gene, and is therefore commonly used as a diagnostic antigen for HV detection [22] [23] [24] [25] . Different serotypes of HVs can be determined by an at least four-fold difference in two-way cross neutralization tests, and it has been reported that serotype-specific as well as group-common and genus-common epitopes can be found in the NP. Cross-reactivity has been found between different serotypes of HVs in rodents and insectivores [26] [27] [28] . However, the serological and antigenic relationships between bat-and rodent-or insectivore-borne HVs have not yet been studied.

South and southwest China have a high density of bat population consisting of a large number of diverse species. Recently, investigations on bat viruses in this area have revealed many novel viruses, such as coronaviruses [29] [30] [31] [32] [33] [34] , filoviruses [35, 36] and group A rotaviruses (RVA) [37, 38] . Among these, some bat-borne coronaviruses [29, 30, 33] and RVAs [38] have been found to cross species, causing outbreaks of emerging infectious diseases in human and pigs. South and southwest China are also the major epidemic areas of HFRS with all transmission events associated with exposure to rodents [39] . Although increasing number of HVs have been identified in bats, no investigation has been shown their seroprevalence and antigenic characters. The implication of bat-borne HVs to public health is still unclear. In present study, we have conducted systematic etiological and serological investigation on bat-borne HVs in south and southwest China, revealing the antigenic relationships between bat-borne and human-infecting HVs and identifying a geographic region between southwest China and north Vietnam in which divergent bat-borne HVs circulate.

Following viral metagenomic analyses of bat intestines and lungs, 18 contigs with lengths of 110-726 nucleotides (nt) from Laibin (LB), Baise (BS) and Pu'er (PE) cities (see S1 Fig) were annotated to HVs. The highest nt identities of these (72-99%) were shared with the M or L segments of LAIV or XSV (S1 Table) . By pan-HV PCR screening of all intestines and lung tissues, only lung tissue of 9 bats were found to be positive. One of 39 (2.6%) black-bearded tomb bats (Taphozous melanopogon) in BS, Guangxi, 5 of 55 (9.1%) pomona roundleaf bats (Hipposideros pomona) in LB, Guangxi, and 3 of 40 (7.5%) pomona roundleaf bats in PE, Yunnan were also positive ( Table 1) . The viral sequence determined in black-bearded tomb bats in BS showed a 97% nt identity with previously reported LAIV BT20 [9] , and were therefore considered an LAIV variant, BT33. The rest eight from LB and PE showed 93% and 82% nt identities with XSVs identified from Vietnam [7, 8] , indicating they were all XSV variants and therefore respectively named XSV AR18, AR19, AR23, AR28, AR30, PR10, PR15 and PR30.

To isolate infectious viruses, homogenized lung tissues of five XSV ARs and three XSV PRs were separately pooled. The two pooled samples, along with one LAIV BT33 positive lung tissue were thoroughly homogenized by grinding and their filtered supernatants were incubated with African green monkey kidney (Vero) and the E6 clone and baby hamster kidney (BHK-21) cell cultures. During five passages, no CPE was observed, and RT-PCR analyses of all passaged cultures were negative, with no HV isolated.

To gain genetic insight into the HVs, the full genomes of LAIV BT33, XSV AR18, AR23 and PR15 were sequenced and analyzed using previously reported methods [9, 40] . As shown in Table 2 , three gene segments of LAIV BT33 had the same sizes as previously reported LAIV BT20 [9] . Three segments of XSV AR18 and AR23 had exactly the same size (1,753 nt of S, 3,751 nt of M and 6,521 nt of L), while PR15 had similar sized S (1,743 nt) and L (6,522 nt), but its M segment was shorter (3,584 nt) than those of XSV AR18 and AR23, resulting from a 50-aa deletion at the 5' terminal of the coding sequence, corresponding to 6-55 aa of Gn protein. This deletion was confirmed by repeated RT-PCR and sequencing. Currently the function of Gn is largely unknown and 1-17 aa is the signal peptide of Gn responsible for translocation of Gn to Golgi [41, 42] , therefore the deletion may have impact on Gn location. In addition, the highly conserved motif WAASA (polyprotein-recognized pentapeptide) in the M segment of HV was observed in all four strains, but the ORF in the S segment of some HVs (such as Puumala, Tula and Andes viruses) encoding a 7-12 KDa nonstructural protein (NSs) which functioned as an interferon antagonist were not found [43] [44] [45] . Sequence comparison of the four strains with other bat-borne HVs available in GenBank ( Table) .

To construct phylogenetic relationships, 92, 63 and 40 complete coding sequences of hantaviral NP, GP and RdRp respectively were used. Currently there are only 14 NP, 9 GP and 7 RdRp complete sequences of bat-borne HVs available in GenBank. As shown in Fig 1, rodentand insectivore-borne HVs, except for Nova virus (NVAV) [18, 46] and Altai virus (ALTV), showed a similar topology, which classified them within clades I, III and IV in all three trees. Bat-borne HVs showed different topology structures, however, which were all clustered together within clade II in the NP tree, but within three clades (II, V and VI) in the GP tree, or within two clades (II and VI) in the RdRp tree. It is interesting to note that two insectivoreborne HVs, NVAV and ALTV, respectively identified in Talpa moles and Sorex shrews, were genetically closer to bat-borne than to insectivore-borne HVs in the NP and GP trees [18, 46] . 

To characterize the antigenic relationship between bat-and human-infecting HVs the entire NPs of LAIV BT33, XSVAR18 and SEOV GuangzhouRn36 were expressed in E. coli and purified. Polyclonal anti-serum against the rNPs (named anti-L, anti-X and anti-S, respectively, for LAIV, XSV and SEOV) were prepared by immunization of mice, resulting in titers by ELISA of 8,000×, 4,000× and 4,000× respectively. Western blot (WB) analyses showed that SEOVconvalescent human serum (H anti-SEOV) had a significant cross-reactivity with the rNP of XSV but not LAIV, as well as the rNP of SEOV (Fig 2A) . To further characterize this cross antigenicity, three NPs were eukaryotically expressed with an EGFP tag. Further WB analyses with the three anti-rNP sera showed that anti-L reacted strongly with the eukaryotic rNP of XSV but not with that of SEOV, in addition to a very strong reactivity with its own LAIV rNP ( Fig  2B) . In contrast, anti-S showed only a weak cross-reactivity with eukaryotic rNP of XSV and not at all with that of LAIV, although with very strong reactivity with its own SEOV NP. It is interesting to note that anti-X had cross-reactivity with eukaryotic rNPs of both LAIV and SEOV (weak for LAIV and strong for SEOV), in addition to a very strong reactivity with its own XSV rNP. This result unexpectedly showed that bat-borne HVs do share cross reactivity with human-infecting HVs and that significant differences of antigenicity do exist in different bat-borne HVs. In our study XSV was antigenically closer to SEOV than LAIV.

Results of the serological assay of 709 bat sera by ELISA against the three viruses are shown in S2 Fig with 88 

Of 88 bat sera tested by WB, 48 with sufficient volume were further tested for neutralizing antibody (NAb) titers against SEOV by the fluorescent antibody virus neutralization test (FAVNT). Results showed that nine bat sera (18.8%) from four provinces had NAb titers ranging from 32× to 128× (Fig 4) , of which five were both WB and ELISA positive, with the other four negative for both. The positive control (H anti-SEOV) had an NAb titer of 513× (Fig 4) .

Of interest is that one serum from Rousettus leschenaultii bat in Xishuangbanna, BN78, had the highest NAb titer (128×) against SEOV and was WB and ELISA positive only for SEOV, not for LAIV or XSV. Results also showed that some WB and ELISA double-positive bat sera against the three viruses did not neutralize SEOV, including samples BN19, CZ63, CZ26, NP39 and ZJ61.

As bat-borne HVs have only recently been identified, there is insufficient sequence data at present to provide a comprehensive analysis of their genetic diversities. Apart from the four complete genomic sequences reported here (Table 2) , the sequences of representative batborne HVs published to date in GenBank show that many are not of complete genomes or even of a full-length gene segment [5] [6] [7] 11] . Currently Laibin, Longquan and Quezon viruses are the only bat-borne HVs approved by ICTV so far [6, 9, 12] . Among completed genomes, there are the complete coding sequences of the three gene segments of BRNV from the Czech Republic [13] . Here we report the first genomic sequence of XSV and show that XSV and BRNV should be considered new HV species awaiting ICTV consideration. Fortunately, there are partial L gene sequences available for all bat-borne HVs, which allowed construction of a phylogenetic tree (using 314 bp), permitting comparison with rodent-and insectivore-borne HVs (S5 Fig) . This showed that current bat-borne HVs can be classified into nine species as listed in Table 2 , with almost every one having a specific bat genus as host. Of them XSV is the most notable bat-borne HV, which has been found in different locations in present and previous studies [7, 8] , and showing significant nt variation although aa sequences of the genomic segments are conserved. The nt identity between currently identified XSV variants ranges from 76.1%-93.4% (Table 2 ). Furthermore the NP of XSV showed cross antigenicity with both SEOV and LAIV but the NP of LAIV showed no cross antigenicity with SEOV (Fig 2B) , indicating that XSV is antigenically closer than LAIV to SEOV and therefore an ideal focus for gaining insight into the role of bat-borne HVs in public health. Meanwhile the closer Bat-borne hantaviruses in China relationship of insectivore-borne HVs NVAV and ALTV to bat-borne than to insectivoreborne HVs in the NP and GP trees (Fig 1) indicated that HVs from bats and insectivores could share the common ancestry for evolution [18, 46] . In general our sequence comparisons and phylogenetic analyses show that bat-borne HVs had broad genetic diversities and had evolved worldwide within an independent and diverse phylogroup. In this regard, more extensive studies obtaining more complete sequences in extended areas will undoubtedly identify more novel bat-borne HVs in future.

Although nine bat-borne HVs have been identified worldwide, the virus detection rate is low and in limited locations [5, 7, 9, 11] . The present study investigated bats in 22 cities, but the viral RNA was found in only three bat colonies in three cities, and RNA-positive rates were only 1.4% (1/74) for LAIV in BS, 3.0% (5/168) for XSV in LB and 7.5% (3/40) for XSV in PE (see Fig  5 and Table 1 ). In contrast, seropositive rates are higher: 28.1% (9/32) for LAIV and 40% (2/5) for XSV in BS and LB respectively (sera were not collected in PE) ( Table 1) . Low viral RNA detection rates have also been reported in previous publications with 5.6% (1/18) for MGBV in Sierra Leone [5] , 3.1% (1/32) for LAIV BT20 in China [9] , and 0.8% (1/123) for MAKV in Gabon [11] . Unfortunately, the seroprevalence was not reported in these publications. The antibody titers of bat sera against HVs in our study were rather low (the most were 100× and only 18 were 400×) as compared to those of rodent reservoirs which were usually higher and could reach 50,000 [24, 47] , this might be ascribed to the higher diversity of V H (especially FR3) in immunoglobulin genes of bats in comparison to those of mouse, swine and human [48] . The correlation between RT-PCR positivity and antibody positivity about hantavirus infection in rodents or shrews were reported [49, 50] . Song et al. reported that a certain proportion, although not all, of Ussuri white-toothed shrews (Crocidura lasiura) with IgG antibodies against Imjin virus (MJNV, a newly isolated hantavirus) had MJNV RNA detectable by RT-PCR [50] . In our study bats sampled in 2015 and 2016 showed higher seropositive rates, Bat-borne hantaviruses in China but HV RNA was not detected from either seropositive or seronegative bats. All nine RT-PCR positive samples were collected in between 2012-2014, but their antibody titers were not tested since the sera were not collected during that time. Moreover, serology study on bat-borne HVs was not conducted in previous publications, therefore further study is needed to understand the dynamics of HV infection and its antibody response in bats.

Serological epidemiology is important to uncover the real situation of bat-borne HV prevalence, and is critical for eventual estimation of the potential risk of these viruses to public health. Since bat-borne HVs have never been isolated, their NP is a preferential target for serological investigation and antigenic differentiation. It is the main immunogenic protein which contains both serotype-specific and group common epitopes, and is commonly used as diagnostic antigen for HV detection [22] [23] [24] [25] 51] . For these reasons, rNPs of LAIV, XSV and SEOV expressed in E. coli were used to assay all 709 bat serum samples by ELISA, resulting in identification of a large number of seropositive sera (Fig 3) , with many likely to cross react with two or three rNPs (see OD 492 values in S2 Fig). To confirm this, 88 ELISA sera were further tested by WB against all three rNPs with results showing that, except for sera reacting exclusively with one rNP, some could cross recognize two rNPs, mainly the rNPs of LAIV/XSV, or XSV/ SEOV, and seldomly the rNPs of LAIV/SEOV (see S3 Fig) . It is notable that 21 sera could cross recognize three rNPs with 10 showing very strong reactivity against all three rNPs. The role of NP in producing this cross antigenicity was further verified by WB using a combination of eukaryotically expressed rNPs and NP-specific antiserum (see Fig 2B) . To identify NAb against SEOV, 48 bat sera were analyzed by FAVNT, which identified 9 (18.8%) positives. Fig  4 summarizes the results of 48 bat sera assayed by FAVNT, ELISA and WB. Of the nine NAbpositive sera, four (BN5, ZS7, ZS25, ZS27) neutralized but did not react by ELISA or WB, three (BN64, CZ67, NP6) not only neutralized but also reacted with three rNPs by ELISA and WB. The most interesting bat serum was BN78, which neutralized SEOV and reacted with the rNP of SEOV but not with that of LAIV and XSV. BN78 was collected from a Rousettus leschenaultii bat, of this species 51 of 142 individuals showed anti-HV antibody positive (35.9%), the highest among all bat species (Table 1) . Furthermore another Rousettus species (Rousettus amplexicaudatus) was reported to harbor Quezon virus in the Philippines [12] , suggesting that fruit bats in genus Rousettus are likely major reservoirs of HVs. Moreover many sera without neutralizing activity reacted with the three rNPs by ELISA and WB. Altogether, it is interesting to have found multiple patterns of cross-reactivity with three rNPs. Illustration of the complex patterns will be difficult but likely to imply that the bats had been infected with other unknown HVs. The prime example is bat serum BN78. It had the highest neutralization titer against SEOV and exclusively strong reactivity with the rNP of SEOV, indicating that this bat was infected by an unknown HV antigenically very close to SEOV, but not SEOV since both human anti-SEOV convalescent serum (Fig 2A) and SEOV-specific anti-rNP serum (Fig 2B) could also cross react with the rNP of XSV. Altogether, the multiple genetic diversities and different cross-reactivity patterns indicate that more as yet unknown bat-borne HVs circulate in the investigated region, but to uncover them further investigation is needed.

Viruses usually have a defined host range for circulation in nature. It is interesting to note that LAIV BT33 in the present study and LAIV BT20 in a previous study [9] have been identified in different locations but from the same bat species (Ta. melanopogon). All XSV variants in the present and previous studies [7, 8] have come from a single bat species, Hipposideros (Hi. pomona). Apart from MAKV in Gabon, which also came from a Hipposideros sp. (Hi. ruber) , all other six HVs came from bats in six different genera. The host range of currently identified bat-borne HVs is summarized in Table 2 . In general, bat-borne HVs have a huge genetic diversity with different viral species harbored by different genera of bats, showing wide range of hosts. But regarding a given bat-borne HV species its host range may be narrow, restricted mainly to one bat genus. This implies that a given bat-borne HV may have a host tropism.

As shown in Fig 5, 12 of the 13 cities in 4 provinces in which serum collections were made had a positive seroprevalence, with Guangxi having the most positive samples and most seropositive locations (6 of 7 sampled cities were seropositive). Furthermore, two HVs were detected in two of its cities, BS and LB. In 2015, LAIV was identified in LB [9] , in which XSV was found in the present study although from another location within the city, indicating that divergent batborne HVs co-exist in LB. LAIV was also found in BS this time, several hundred kilometers west of LB (see Fig 5) , indicating that LAIV has a broad distribution in Guangxi province. It is notable that XSV has been identified in two north Vietnamese provinces, Tuyên Quang (TQ) and Phú Thȯ (PT), as shown in Fig 5, and in the central Vietnamese province Quảng Nam since 2013 [7, 8] . In present study eight strains of XSVs were identified in LB of Guangxi and PE of Yunnan, indicating that XSV circulates in the vast area between Chinese Guangxi/Yunnan and Vietnam. The accumulated serological and molecular data highly support the proposition that a vast area between China and southeast Asia provides a natural focus for bat-borne HV circulation. In this area natural circulation of genetically divergent bat-borne HVs in their hosts would be maintained, and therefore the concept of a bat-borne HV circulation sphere has been introduced to describe the situation. While there is a lack of sufficient serological data in Yunnan Province, a narrow area between southwest Guangxi and north Vietnam likely forms a main circulation sphere of at least two species of bat-borne HVs (Fig 5) . With more extensive investigations this area may be extended, particularly to surrounding areas in Laos, Myanmar and even Thailand.

In conclusion, the present study has compiled the first profiling of cross antigenicity between bat-borne and human-infecting HVs as well as among bat-borne HVs. It has also revealed the seroprevalence and wide distribution of bat-borne HVs in south and southwest China. A comprehensive analysis based on genetic diversity, seroprevalence, cross antigenicity and host range of the viruses has helped identify an area between China and Vietnam as a main circulation sphere where at least two bat-borne HVs circulate in the bat population. Given the existence of bat-borne HVs genetically and antigenically close to human-infecting HVs, extensive studies should be emphasized in future to assess the potential risk of bat-borne HVs to public health. 

The 1,419 tissues (lungs and intestines) and 709 sera of 1,561 bats used in present study were archived and sub-packed samples stored at -80ºC following collection between 2012 and 2016 in 22 cities of Yunnan, Guangdong, Fujian, Zhejiang and Guangxi provinces, China, and used to investigate viruses than hantaviruses in our previous studies [32, 38, 52, 53] . Bat species were identified morphologically and then molecularly by sequencing the bat mitochondrial cytochrome b gene from muscle tissue [54] . These bats were classified as belonging to 22 species within 9 genera and 6 families: Hipposideridae (n = 598), Rhinolophidae (n = 358), Vespertilionidae (n = 172), Miniopteridae (n = 129), Emballonuridae (n = 74) and Pteropodidae (n = 230). Detailed sample information is shown in Fig 5 and Table 1 . The tissue samples were homogenized and subjected to RNA extraction using the RNeasy Mini Kit (QIAGEN), and the RNA was reversely transcribed into cDNA which was processed for viral metagenomic analysis as described previously [55] . Serum samples were used in serological analyses.

Identified HV-like contigs were subjected to BLASTn and BLASTx search (https://blast.ncbi. nlm.nih.gov/Blast.cgi). The genomic positions of the contigs were decided using Hantaan virus strain 76-118 as the reference. RNA was extracted from bat tissues (intestines with contents and lungs) and screened by RT-PCR using pan-HVs nested primers targeting a 396-bp sequence of the conserved L segment [40] . Positive amplicons were sent for Sanger sequencing (Comate) and the obtained sequences were used to initially determine their phylogenetic locations. Details of the primers can be seen in S2 Table. Virus isolation HV positive lung tissues (~100 mg/bat) were thoroughly ground with DMEM (500 μL/100 mg tissue) in a homogenizer and clarified by centrifugation at 5,000 g for 5 min. Following sterilization by passage through a 0.22 μm Millipore filter supernatants were incubated with BHK-21 cells as well as the HV-sensitive cell lines Vero and Vero E6 [56] (all stored in our laboratory) in 24-well plates. After incubation for 24 h at 37ºC, the cells were washed 2x with PBS, incubated in DMEM with 2% FBS for 14-21 days and observed daily for cytopathic effects (CPE). Cultures, if showing no CPE, were harvested by freeze-thawing 3x and passaged again in the same cell lines. After five passages, the cultures were analyzed for HV by RT-PCR.

To characterize the full genomic sequence and structure of detected HVs, genome-amplifying overlapping primers were synthesized based on the contigs (S1 Table) and representative sequences of previously published HVs including bat-borne HV LAIV BT20 [9] . Since the terminal nucleotide sequences of S, M and L segment are conserved among members of the genus Orthohantavirus [1] , their sequences were used as primers to obtain the end sequences of each segment. The targeting amplicons were amplified using the Phusion High-Fidelity PCR Master Mix with HF Buffer (NEB) with the recommend reaction system, and cloned into a blunt-end pLB vector using a Lethal Based Fast Cloning Kit (Tiangen). Three clones of each amplicon were further identified by PCR and then sent for commercial Sanger sequencing. The complete genomic sequences were obtained by assembling amplicons with overlapped regions using SeqMan in the DNAStar software package. ORFs of each segment were searched by ORFfinder (https://www.ncbi.nlm.nih.gov/orffinder/) in NCBI and the predicted proteins were further confirmed by aligning in BLASTp.

The representative sequences of each classified HV species as well as some unclassified HVs were downloaded from GenBank. Their complete NP, GP and RdRp coding regions (aa) were aligned with those obtained in the present study using the online program MAFFT version 7 (https://mafft.cbrc.jp/alignment/server/). The best-fit substitution model for each tree was selected based on the Akaike Information Criterion (AIC) of Smart Model Selection (SMS) in PhyML (version 3.0) [57] . Phylogenetic trees, including their topology and support for tree nodes, were then inferred using the maximum likelihood method, Subtree Pruning and Regrafting (SPR), approximate Likelihood Ratio Test (aLRT) with the Shimodaira-Hasegawalike (SH-like) tree-selection method in PhyML [58] . Sequence identities were calculated by MegAlign in DNAStar software package.

The complete NP coding sequences (CDS) of bat-borne HV LAIV, XSV strains identified in the present study were amplified with the 5' EcoRI and a 3' XhoI sites at the two ends. The complete CDS of SEOV strain GuangzhouRn36 (1,287 nt, Accession number: GU592948) was optimized using E. coli codons and chemically synthesized (GENEWIZ) with the same restriction enzyme sites (primers shown in S2 Table) . The NP gene fragments so obtained were subcloned into a prokaryotic expression vector pET-28a(+) with a His-Tag at C terminus and used to transfected E. coli strain Rosetta (Tiangen). The rNPs were expressed after induction with 0.5 mM isopropyl-β-d-thiogalactoside (IPTG) and identified by SDS-PAGE and WB using mouse anti-6X His-tag monoclonal antibody and Alexa Fluor 680-conjugated donkey anti-mouse IgG H&L (Abcam) as the respective primary and secondary antibodies. The rNPs of the three HVs were purified and quantified by Ni-NTA His Bind Resin (Novagen) and BCA Protein Assay Kit (CWBio), following which they were identified using an SEOV-convalescent human serum (H anti-SEOV) and a negative human serum (H Neg.) as controls (both stored in our laboratory). The H anti-SEOV serum was collected from a SEOV-infected convalescent patient who was diagnosed at onset by clinical symptoms and RT-PCR.

Specific hyperimmune sera were prepared by injecting four week-old female Kunming mice intramuscularly with purified rNPs of the three HVs. Each injection contained 20 μg protein mixed 1:1 (V/V) with the Quick Antibody-Mouse 5W adjuvant (Biodragon) as recommended by the producers and booster doses with the same formulated rNP were given at 14 days later. At 21 days post boost, blood was collected through heart puncture for serum preparation. Antibody titers were determined by rNP-based ELISA as described below.

Complete CDS of NPs of the same LAIV, XSV and SEOV strains were amplified with primers containing a 5' XhoI site and a 3' EcoRI site. Three NP fragments were fused in-frame to the Cterminal of the enhanced green fluorescent protein (EGFP) tag (239 aa, MW 27 kDa) of eukaryotic expression vector pEGFP-C1. Transient expressions of rNPs were obtained by transfection of BHK-21 cells in 6-well cell plates with the constructed plasmids with the blank vector pEGFP-C1 as control, using Lipofectamine2000 reagent (Invitrogen) according to the manufacturer's protocol. The transfected cells were cultured at 37ºC with 5% CO 2 for 36 h, and then examined microscopically for the expression of fusion proteins with green fluorescence. NP-expressing cell cultures were collected and lysed with Cell Lysis Buffer (CST), and the total protein was quantified using the BCA Protein Assay Kit. Correct expression of rNPs in cell lysates were confirmed by WB using mouse anti-EGFP monoclonal antibody (Abcam) and Alexa Fluor 680-conjugated donkey anti-mouse IgG H&L as the respective primary and secondary antibodies.

Eukaryotically expressed NPs were used as antigens to detect antibodies in mouse hyperimmune sera by WB. Briefly the eukaryotically expressed NPs were separated by SDS-PAGE and transferred onto nitrocellulose blotting membranes (GE Healthcare), blocked with 5% non-fat milk (Promega) at 4ºC overnight, then incubated with the above three mouse anti-NP hyperimmune sera at 1:300 dilution for 2 h. After washing with PBST 3x, the membranes were incubated with Alexa Fluor 680-conjugated donkey anti-mouse IgG H&L 1:1,000 for 50 min. After washing, the membranes were scanned and photographed using an Odyssey imager (LI-COR Biosciences). All WB analyses were repeated at least three times.

ELISAs using rNPs as coating antigen were developed to detect antibodies of bat against LAIV, XSV and SEOV. Briefly 96-well microplates (Corning) were coated with purified prokaryotically-expressed rNPs (200 ng/well in NaHCO 3 -Na 2 CO 3 buffer at pH 9.6) at 4˚C overnight and blocked with 5% non-fat milk (Promega) at 37˚C for 1 h. Then 100-fold PBS-diluted serum samples were added to the wells (duplicate wells per serum) and incubated at 37˚C for 1 h followed by addition of HRP-conjugated goat anti-bat IgG, H&L chain, polyclonal antibody (BETHYL, react specifically with bat IgG, and with light chains common to other bat immunoglobulins) at a 1:20,000 dilution for bat sera, and HRP-conjugated goat anti-human IgG polyclonal antibody (Zsbio) at a 1:200 dilution for human sera (controls). After incubation at 37ºC for 50 min, freshly-prepared O-phenylenediamine (OPD) substrate solution (Sigma) was added to each well for 5 min for color reaction, which was stopped by addition of 2 M sulfuric acid. The OD 492 values were immediately read and blanked by the OD 630 value using a Multimode Microplate Reader (Infinite 200 PRO, Tecan). The cut-off values of bat sera were determined based on the highest CR between results of ELISA and WB.

To validate the ELISA result, WBs using prokaryotically expressed rNPs were performed to detect bat and human sera. Briefly, the rNPs were separated by SDS-PAGE, transferred onto nitrocellulose membranes followed by blocking using the protocol described above. Membranes were then incubated with 1:300 diluted selected bat or human serum for 2 h, followed by 2,000-fold diluted HRP-conjugated goat anti-bat IgG H&L chain polyclonal antibody or 100-fold diluted HRP-conjugated goat anti-human IgG polyclonal antibody for 50 min. All procedures were conducted at ambient temperatures. Bands were pictured by automatic chemiluminescence (Tanon).

To determine the NAb titers of bat and human sera, the FAVNT using 200 TCID 50 SEOV (10 5.31 TCID 50 /0.1mL) in Vero E6 cells was performed by a previously published protocol [59] . Anti-HV monoclonal antibody provided by the Fourth Military Medical University [60] was labeled using FITC [61] . The NAb titer of each serum was calculated using the Spearman-Kärber formula [62] .

To compare the differences of the OD 492 values of bat and human sera to each HV, normal distribution tests were conducted separately, and multiple comparison was performed using t tests (LSD), Student-Newman-Keuls (SNK) test, Tukey's studentized range (HSD) test, Bonferroni (Dunn) t tests and Scheffe's test. Then κ test, an inter-rater agreement statistic, was used to evaluate the consistency between the results of ELISA and WB. All statistics were programed and calculated using the Statistical Analysis System (SAS) version 9.2. Table. (TIF) S1 

Hantavirus infections

China: Disease prevention and control bureau, National health commission of the People's Republic of China

Divergent lineage of a novel hantavirus in the banana pipistrelle (Neoromicia nanus) in Cote d'Ivoire

Hantavirus in Bat

Phylogeny and origins of hantaviruses harbored by bats, insectivores, and rodents

Novel Bat-borne Hantavirus

Molecular phylogeny of hantaviruses harbored by insectivorous bats in Cote d'Ivoire and Vietnam

Novel hantavirus identified in black-bearded tomb bats, China

ICTV taxonomic proposal 2016.023a-cM.A.v2

Phylogenetic analysis of a newfound bat-borne hantavirus supports a laurasiatherian host association for ancestral mammalian hantaviruses

Molecular phylogeny of a genetically divergent hantavirus harbored by the Geoffroy's rousette (Rousettus amplexicaudatus), a frugivorous bat species in the Philippines

Novel hantavirus identified in European bat species Nyctalus noctula

Hantavirus strains in East Africa related to Western African hantaviruses. Vector-Borne and Zoonot

Changes to taxonomy and the International Code of Virus Classification and Nomenclature ratified by the International Committee on Taxonomy of Viruses

Molecular phylogeny of a newfound hantavirus in the Japanese shrew mole (Urotrichus talpoides)

Host switch during evolution of a genetically distinct hantavirus in the American shrew mole (Neurotrichus gibbsii)

Evolutionary insights from a genetically divergent hantavirus harbored by the European common mole (Talpa europaea)

Reconstructing the evolutionary origins and phylogeography of hantaviruses

RNA binding of recombinant nucleocapsid proteins of hantaviruses

Hantaviral proteins: structure, functions, and role in hantavirus infection

Novel serological tools for detection of Thottapalayam virus, a Soricomorpha-borne hantavirus

Rapid, whole blood diagnostic test for detecting anti-hantavirus antibody in rats

Rapid enzyme-linked immunosorbent assay for the detection of hantavirus-specific antibodies in divergent small mammals

Serological diagnosis with recombinant N antigen for hantavirus infection

Monoclonal antibodies to three strains of hantaviruses: Hantaan, R22, and Puumala

Truncated hantavirus nucleocapsid proteins for serotyping Sin Nombre, Andes, and Laguna Negra hantavirus infections in humans and rodents

Serological survey of Seewis virus antibodies in patients suspected for hantavirus infection in Finland; a cross-reaction between Puumala virus antiserum with Seewis virus N protein?

Isolation and characterization of a bat SARS-like coronavirus that uses the ACE2 receptor

Identification of diverse alphacoronaviruses and genomic characterization of a novel severe acute respiratory syndrome-like coronavirus from bats in China

A bat-derived putative cross-family recombinant coronavirus with a Reovirus gene

Detection and characterization of diverse alpha-and betacoronaviruses from bats in China

Fatal swine acute diarrhoea syndrome caused by an HKU2-related coronavirus of bat origin

Discovery of novel bat coronaviruses in south China that use the same receptor as Middle East respiratory syndrome coronavirus

Filovirus RNA in fruit bats

Genetically diverse Filoviruses in Rousettus and Eonycteris spp

The complete genome sequence of a G3P[10] Chinese bat rotavirus suggests multiple bat rotavirus inter-host species transmission events

Group A rotaviruses in Chinese bats: genetic composition, serology and evidence for bat-to-human transmission and reassortment

The national monitoring scheme of Hemorrhagic Fever with Renal Syndrome (trial implementation) Beijing, China: Chinese Center for Disease Control and Prevention

Hantavirus in African wood mouse

Golgi localization of Hantaan virus glycoproteins requires coexpression of G1 and G2

The envelope proteins of the Bunyavirales

Tula and Puumala hantavirus NSs ORFs are functional and the products inhibit activation of the interferon-beta promoter

Tula hantavirus NSs protein accumulates in the perinuclear area in infected and transfected cells

The Andes hantavirus NSs protein is expressed from the viral small mRNA by a leaky scanning mechanism

Complete genome sequence of Nova Virus, a hantavirus circulating in the European mole in Belgium

Long-term patterns of immune investment by wild deer mice infected with Sin Nombre virus

The little brown bat, M. lucifugus, displays a highly diverse V H, D H and J H repertoire but little evidence of somatic hypermutation

Increased detection of Sin Nombre hantavirus RNA in antibody-positive deer mice from Montana, USA: evidence of male bias in RNA viremia

Characterization of Imjin virus, a newly isolated hantavirus from the Ussuri white-toothed shrew (Crocidura lasiura)

Antigenic properties of N protein of hantavirus

Detection and characterization of a novel Norovirus in bats

Genetic diversity of bat orthohepadnaviruses in China and a proposed new nomenclature

Molecular phylogenetic of Hipposiderids (Chiroptera: Hipposideridae) and Rhinolophids (Chiroptera: Rhinolophidae) in China based on mitochondrial cytochrome b sequences

Virome profiling of bats from Myanmar by metagenomic analysis of tissue samples reveals more novel mammalian viruses

Epidemiological surveillance of viral Hemorrhagic Fevers with emphasis on clinical virology

SMS: Smart Model Selection in PhyML

New algorithms and methods to estimate maximum-likelihood phylogenies assessing the performance of PhyML 3.0

Establishment and application of fluorescent antibody virus neutralization test for Hantavirus

Preparation and identification of McAbs against house-rat type of HFRS virus

Hubei: Hubei publishers of science and technology

Animal virology