The serologic investigation and viral isolation of bluetongue virus in Shangri‐La in Southwest China

Abstract Bluetongue is an arthropod‐borne viral disease of ruminants caused by bluetongue virus (BTV). In China, BTV is relatively common in Yunnan Province with the exception of northern regions around Shangri‐La, where the average altitude is approximately 3,450 metres. Recently, the seroprevalence of BTV has been measured in yaks in Shangri‐La; therefore, this study investigated BTV infections in this area. The serological investigation in five villages in Shangri‐La showed that there were sporadic BTV infections in yaks (20 of 507 positive) during 2014 to 2017, while the seroprevalence of BTV at three goat farms in a nearby river valley was 35%–65% in 2017. Subsequently, 20 sentinel goats were kept on two separate farms in the river valley and monitored for seroconversion between May and September of 2017. Five of the sentinel animals were tested positive for antibodies to BTV by C‐ELISA during the study period, and 13 BTV isolates were isolated from ten sentinel animals. All isolates were identified as the same serotype, and the complete nucleotide sequence of one was determined. The genomic sequences showed that the isolated BTV strain belonged to serotype 21 and had approximately 99.8%–100% homology with three Indonesian BTV‐21 strains (D151, RIVS‐66 and RIVS‐60) between their coding sequences (CDSs) except for Seg4 (99.5%). Besides, our data suggested that this BTV‐21 strain might have also infected some local yaks and sheep.

capsids (Roy, 2007). VP2 and VP5 together construct the outer layer of the capsid, and VP2 is the critical protein for virus attachment to target cells (Roy, 2007). Accordingly, the VP2 protein has the most sequence diversity among BTV proteins and is the basis of serotype classification. To date, 27 serotypes have been identified worldwide (Jenckel et al., 2015;Maan et al., 2011), and 14 serotypes have been confirmed in China since 1979 (Kirkland et al., 2002;Yang et al., 2017).
Yunnan Province is located in a tropical and subtropical region and has diverse climates. It is the main region of BTV prevalence in China, and up to 13 serotypes of BTV have been identified in Yunnan (Kirkland et al., 2002;Yang et al., 2017). However, Shangri-La County, which is located in northern Yunnan Province, has always been excluded from the BTV survey, because it has relatively cold weather and is closed to traffic. To our knowledge, this study is the first to investigate BTV prevalence in this region.
Shangri-La is adjacent to the south edge of the Qinghai-Tibetan Plateau and has an average altitude of 3,450 m. The landform is rugged because there are several rivers that flow across the region. The average annual temperature on the Shangri-La plateau is approximately 5°C, but the climate is relatively warm in summer in the valleys where the lowest altitude is 1,900 m. Yaks, sheep and goats are the main animals raised on the Shangri-La plateau, and yaks are the most prevalent animals raised on the highland.
Recently, anti-BTV IgG seroprevalence in yaks and sheep was measured on the Tibetan Plateau in western China (Ma et al., 2017), and a serological investigation also discovered anti-BTV IgG -positive yaks in Shangri-La (Xiao et al., 2011). Therefore, this study was carried out to investigate the BTV seroprevalence and identify the BTV strains in Shangri-La.

| Sample collection
All the animals in this study were summarized in Table S1, and the operations for sample collection were described as follows. After that, 20 local goats less than 1 year old that were identified as BTV negative by C-ELISA were selected as sentinel animals and assigned to two surveillance sites in Nixi. The animals were kept among other goats between May and September of 2017. Serum samples and EDTA + blood from all sentinel animals were collected every 10 days and used for tests and virus isolation.
In addition, for viral nucleic acid research, EDTA + blood was collected from yaks and sheep in the villages of Xiaozhongdian and Nixi in 2017, and a series of BTV strains isolated in China previously and recently were also used for BTV Seg1 fragments alignment.

| C-ELISA
The C-ELISA method was used to test the anti-VP7 (BTV) antibodies of animal serum samples, as described previously (Chapter 2.1.3, OIE Terrestrial Manual 2009). Briefly, the lysates of BTV-infected BHK-21 cells were used to coat 96-well microplates, and serogroup-reactive anti-VP7 mAb (8A3B6) (Kirkland et al., 2002) was used as a competing antibody. Colour was developed by TMB substrate and stopped by 1 M sulphuric acid.
Optical density (OD) values were read by a microplate reader (Multiskan Ascent, Thermo). Per cent inhibition (PI) was calculated as PI = 1−OD sample /OD negative . Samples with PI ≥ 50% were considered positive.

| Serum neutralization test (SNT)
The SNT is a semiquantitative method to detect anti-BTV specific antibodies in serum. Briefly, starting from a 1/4 dilution, serum samples were diluted in a twofold dilution series. Every diluted sample was added into 4 replicate wells at 50 μl per well.
Approximately 100 TCID 50 (50% tissue culture infective dose) of the newly isolated virus was added at a volume of 50 μl per well and mixed with the diluted serum samples. After that, 1 × 10 4 BHK-21 cells were added to each well in a volume of 100 μl, and after incubation for 4-6 days, the CPE was observed under an inverted microscope. The wells were scored for the degree of CPE observed. Titres were expressed as the final dilution of serum present in the serum/virus mixture when 50% of the duplicate wells were protected according to Reed-Muench methods (Reed & Muench, 1938).

| RT-PCR
For viral nucleic acid detection in animal blood, as well as in the samples of BTV strains isolated previously (see Table S2) and used for Seg1 analysis (Figure 1), high-throughput BTV RNA extraction from the EDTA + blood samples was performed by the MagMAX TM Express-96 (Ambion, USA) with a MagMAX-96 Viral RNA Isolation Kit (Ambion, USA). A pair of universal primers, Bs1p3f (5-TCATCGAATT TCGGGCTAA) and Bs1p3r (5-CCCACATCTT TACAAACCAC), were used to amplify a 429 bp fragment of BTV Seg1. RT-PCR for Seg1 was performed by a GeneAmp 9700 PCR system (Thermo) and a PS One-Step RT-PCR Kit ver.2 (Takara) using primers Bs1p3f and Bs1p3r according to the manufacturer's instructions.
For nucleic acid analysis of the BTV isolates from Shangri-La, whole RNAs were extracted from BHK-21 infected with the isolates and reverse transcribed to cDNA by SuperScript III (Thermo) with random primers. Seg2 fragments of 999 bp were amplified by Pfu (Tiangen) with primers S2-F (5-TAACGCAAAC GGAAATCGCT CA) and S2-R (5-GCGTTATCGG AAGATACCAA ACCA).

| Sanger sequencing and analysis
The PCR products of Seg1 and Seg2 were sent to the Kunming branch of the Tsingke Biological Technology Company for sequencing using primers Bs1p3f/Bs1p3r and S2-F/S2-R, respectively. Sequences in the middle of Seg1 (524-707 bp) and Seg2 (1654-2520 bp) were used for alignment.

| Virus isolation
The virus was isolated by the egg-C6/36-BHK-21 (ECB) virus isolation system, in which the serum was inoculated in an embryonated egg, Aedes Albopictus cell line C6/36 and baby hamster kidney cell line BHK-21, in that order. The ECB virus isolation was performed as previously described (Clavijo, Heckert, Dulac, & Afshar, 2000).
Briefly, the erythrocytes were precipitated from goat blood and lysed by 10 volumes of water. Then, 0.1 ml clarified erythrocytes were inoculated i.v. into embryonated eggs. The eggs were incubated at 33.5°C for 7 days. Embryos that died between 2 and 7 dpi were dissociated in 1 ml of PBS by syringe and clarified by centrifugation at 1,000 g. Then, 0.1 ml of the supernatant was inoculated into a T25 flask with C6/36 cells and 5 ml 2% FBS + (Gibco) MEM (Gibco). The C6/36 cells were cultured at 28°C for 7 days. Then, 1 ml of the supernatant from C6/36 culture was inoculated into a T25 flask of BHK-21 with 5 ml 2% FBS + MEM. If CPE was observed, the flask with BHK-21 was stored at 4°C until further analysis. If no CPE was observed, the culture was blind passaged up to three times before being called negative.

| Viral genomic dsRNA extraction
BTV (YN/2017)-infected BHK-21 cells (48 hpi) were collected by scraping and centrifugation. The cell pellet was suspended in 100 μl medium and lysed by freezing and thawing twice, followed by incubation with RNase A (Takara) at 37°C for 12 hr to delete cellular ssRNA . Then, the viral dsRNA was extracted by TRIzol (Thermo), precipitated with isopropanol and washed with 75% ethanol according to the manufacturer's protocol. The dsRNA was collected by centrifugation (4°C, 5200 g, 5 min) and air-dried.

| Viral genomic cDNA synthesis
Complete viral genomic RNA was amplified by FLAC methods. Both the procedure and the primers used were in accordance with methods by Maan (Maan et al., 2011). Briefly, the 3' end of the viral dsRNA was covalently linked with anchor primer by T4 RNA ligase 1 (NEB) overnight. The anchored dsRNA was purified by an RNA extraction kit (Takara) and used to synthesize cDNA with Superscript III (Thermo).
F I G U R E 1 Phylogenetic analysis of the relationship among the BTV-21 isolates and domestic BTV strains by Seg1 fragment sequence similarity. Sequence alignment was performed by MEGA-X, and the evolutionary tree was built by the neighbour-joining method (bootstrap = 1,000). (1) Shangri-La/2017 (658, 680 and 762) and YN/2017 were the representative isolates from Shangri-La.
(3) Yak, sheep and goat represented the BTV nucleic acid amplified from these animals in Shangri-La S h a n g ri -L a /2 0 1 7 /6 5 8 Ya k 2 Goat 2 Goat 8 Go at 12 G o a t 2 2 S h a n g ri -L a /2 0 1 7 /6 8 0 S h a n g r i-L a /2 0 1 7 /7 6 2

| Amplification and sequencing of the complete virus genome
Viral genomic cDNA was denatured at 95°C and renatured to dsDNA through gradually cooling from 95°C to 25°C. Ten microlitres of dsDNA and 4 μl of 10 μM primer 5-15-1 (Maan et al., 2007) were added to a 100 μl PrimeSTAR-GXL (Takara, Japan) PCR system. Amplification was carried out after denaturation at 95°C for 2 min, followed by 40 cycles of denaturation for 10 s at 95°C, annealing for 10 s at 55°C, extension for 4 min at 68°C and a final extension for 1 min at 68°C.
Amplified complete genomic DNA was sent to the MAGIGEN Company for high-throughput sequencing by a HiSeq 2000 system (Illumina) and SOAPdenovo software.

| Serologic investigation of yaks in Shangri-La
A total of 507 yak serum samples were collected by the local ADCC during 2014 to 2017 in five villages within the Shangri-La district. As shown in Table 1, the C-ELISA tests suggested that a few yaks from three villages had been infected by BTV during 2014 to 2016, and no BTV infection was discovered in the other two villages (Dongwang and Xiaozhongdian). In other words, there were random and sporadic BTV infections among yaks in Shangri-La during 2014 to 2016.

| Serologic investigation of goats in a nearby river valley
Three goat farms in one of the river valleys in Nixi (approximately 28.05 N, 99.51 E) in Shangri-La were selected for serologic investigation. The three farms were located on a hilltop, a hillside and the foot of the hill. Goats were investigated because they are popular ruminants raised in this rocky area. The serologic investigation by C-ELISA suggested that there were notably high BTV infection rates (35.48%-65.63%) among the flocks of goats in the river valley ( Table 2). The clearly different BTV infection rates between the yaks and goats reported here might be attributed to the difference in environmental temperature because a relatively moist and warm climate promotes the propagation of Culicoides, which is a critical factor in BTV transmission among ruminants.

| Serologic surveillance and virus isolation
Two surveillance sites were established in the other two river valleys in Shangri-La. Site 1 and site 2 were located at 27.10 N/100.12 E and 28.17 N/99.70 E, respectively and had altitudes of 2,000 m and 2,800 m, respectively. The linear distance between the two sites was approximately 126 km.
A total of 20 goats identified as BTV negative by C-ELISA were chosen as sentinel animals and kept at the two sites. The sentinel animals were continuously monitored, and the serum and EDTA + blood samples were collected every 10-14 days between May and September of 2017. All the sera collected were tested by C-ELISA, and the results (Table 3) showed that 5 sentinel animals were identified as BTV-infected animals (PI ≥ 50%) and the others were dubious (0 < PI <50%). Since they were clearly positive, the samples from animals #4 and #5 were not collected after the 24 July. Meanwhile, all the EDTA + blood samples collected was used to isolate virus through the ECB virus isolation system. As a result, 13 BTV isolates were isolated from 10 sentinel animals (Table 3).
The isolated virus was used for SNT of sera from the five positive sentinel goats, and the titres were 64, 64, 45, 32 and 128. The high level of titres confirmed that our BTV isolates reliably came from the sentinel goats.

| A strain of BTV-21 was identified
To identify the BTV strains, primary sequencing was performed.

| The isolated BTV was homologous with BTV-21 in Indonesia
Since all the isolates showed accordant properties by SNT and Seg1  (Figure 2, a-c). It has also been suggested that the newly iso-

| D ISCUSS I ON
Thus far, a total of 14 serotypes of BTV (1, 2, 3, 4, 5, 7, 9, 11, 12, 15, 16, 21, 23 and 24) have been isolated in tropical and subtropical areas of China, and all serotypes except type 7 have been found in Yunnan Province (Kirkland et al., 2002;Yang et al., 2017). Recently, BTV traces were found on the Qinghai-Tibetan Plateau in Northwest China, where the average altitude is more than 4,000 metres and the average temperature is approximately 0°C. Additionally, a 13.3% seroprevalence of anti-BTV IgG in yaks in Tibet (Ma et al., 2017) and one case of BTV infection in a yak in Qinghai Province (Hu et al., 2016;Li et al., 2015) have been reported. However, no RNA sequence information for these BTV strains has been reported. Hence, it is meaningful to investigate yaks and BTV in Shangri-La, Yunnan Province.
The Shangri-La plateau, which is located in Yunnan Province in Southwest China, is located in the subtropics and adjacent to the TA B L E 4 Characteristics of the complete genome segments and their deduced proteins from BTV-21 strain YN/2017 Qinghai-Tibetan Plateau and has an average altitude of 3,450 m.
Similar to the Alps, the Shangri-La plateau is rugged and has abundant water resources. Several rivers, including the Jinsha River, which is upstream of the Changjiang River, flow across this plateau.
This study was the first that we are aware of to identify BTV infection in yaks and endemic BTV infection in the relatively warm river valleys in Shangri-La. Serologic investigations show that there was a 0%-22% seroprevalence of anti-BTV IgG in yaks in five villages on the highland during 2014 to 2017 (Table 1), and there was a 50%-65.63% seroprevalence at three goat farms in the nearby valley area (Table 2). However, the direct viral origin of BTV in local yaks is unknown. It is possible that the BTV was transmitted from the viral reservoirs in the valleys to the yaks at higher altitude by Culicoides vectors because the distance between yak farms in the Our preliminary investigation suggested that BTV existed in Shangri-La. Subsequently, 20 sentinel goats were placed at two sites in two valleys in 2017, and several BTV isolates were isolated from the sentinel goats. All isolates were identified as the same strain, and one of the isolates (named YN/2017) was sequenced completely.
Interestingly, the genome of BTV strain YN/2017 was almost identical to those of three Indonesia BTV-21 strains (D151, RIVS-66 and RIVS-60; Table 4) isolated in West Java during 1989 to 1990 (Firth et al., 2017). However, it was clearly different from the domestic strain 5149E, which was isolated in southern China in 2015 and was the first completely sequenced BTV-21 strain in China (Qin et al., 2018).
There are 51 differences between the VP2 amino acid sequences of YN/2017 and 5149E. Similar to the Indian strain KMNO-7 (Shafiq et al., 2013), 5149E has the Seg2 RNA of serotype 21 and the Seg6 RNA of serotype 16, but it is more homogenous with Japanese strains (Qin et al., 2018). Further studies are needed to answer the remaining questions, including the relationship between BTV strains from Shangri-La and the Qinghai-Tibetan Plateau, and the reasons why this strain has maintained a very low incidence of mutation over nearly three decades.

E TH I C S S TATEM ENT
The processes of blood collections from animals (yaks, goats and sheep) were approved by their hosts, and the animal health was protected. The study was approved by the Ethics Committee of Yunnan Animal Science and Veterinary Institute. F I G U R E 3 Sites of collection of the BTV-21 YN/2017 and 5149E strains as well as their homogenous strains. The sites of collection of YN/2017 and its homogenous strains were labelled with red dots, while the sites of collection of 5149E and its homogenous strains were labelled with yellow dots. The map was drawn based on a Google map