Background

Tula virus (TULV) is a Puumala-like virus, belongs to the genus hantavirus of family Bunyaviridae, and is primarily carried by Microtus arvalis [1]. TULV was initially discovered in M. arvarlis in Tula, Central Russia in the early 1990s [2]. Since then, M. arvarlis (obscurus [3]) has been found to act as a reservoir host of TULV in many European countries [4, 5], West Siberia [6], and Eastern Kazakhstan [7]. A few reports have shown human cases presenting a fever syndrome with severe headache [8] or hemorrhagic fever with renal syndrome and pulmonary involvement [9, 10]. However, most individuals with serological and/or pathogenic positive test were asymptomatic in European countries [11]. So far, no study has shown the epidemiological situation of TULV in China, even though voles are common rodents in some pasture areas in northern Xinjiang, China.

Methods

Sample collection and preparation

The investigation was performed in September 2015 in Narati mountain pastures in Xinyuan County, Yili region, Xinjiang. Voles were captured by flooding their burrows with a water pump. The voles that were captured alive were euthanized with barbiturate (100 mg/ kg). The voles were initially identified morphologically, then autopsied and approximately 1 g of lung tissue from each specimen was isolated and cut into 0.5-cm-thick slices using a pair of aseptic operation scissors and then placed to a sterile tube containing 5 mL of RNA stabilization reagent (Ambion RNAlater®, Life Technologies, Carlsbad, CA, US). Each tube was stored at 5 °C overnight, then transferred to − 80 °C for long-term storage. The capture of rodents in fields and protocols for using animals were approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University, Urumqi, China (approval IACUC-2015).

RNA extraction, cDNA synthesis, and PCR

Total RNA was extracted from the lung tissues of voles using TRIzol reagent (Thermo Fisher) according to the manufacturer’s instructions. The reverse transcription primer P14 (Table 1) was used for the first strand cDNA synthesis with the GoScript Reverse Transcription System (Promega, Beijing, China) [12, 16]. In brief, A PCR tube containing 2.0 μL of extracted RNA (500 ng/μL) and 1.0 μL of primer P14 (10 pmol/l) was heated in a 70 °C water bath for 5 min, then chilled in ice water for 5 min. After brief centrifugation, 2.0 μL of GoScript 5 × Reverse Transcriptase buffer, 0.5 μL of GoScript Reverse Transcriptase (200 U/μL), 0.5 μL of Recombinant RNasin Ribonuclease inhibitor (40 U/μL), 1 μL of PCR Nucleotide Mix (10 mM), 2.0 μL of MgCl2 (25 mM), and 1.0 μL of RNase-free deionized water were added to the tube. After incubation at 25 °C for 5 min, the tube was placed in a 42 °C water bath for 60 min. Reverse transcriptase was inactivated by heating to 70 °C for 15 min.

Table 1 List of PCR primers used within the study

To identify hantavirus, the L gene segment (438 bp) was amplified by nest-PCR with viral-genus-specific primers, including outer PCR primers HAN-L-F1 and HAN-L-R1 and inner PCR primers HAN-L-F2 and HAN-L-R2 (Table 1) [13, 17]. All of the PCR products were analyzed using 2% agarose gel. To determine the species of the virus, S gene fragments were amplified by nest-PCR with cDNAs from L gene positive samples using reported primers in Table 1 [14].

DNA extraction and molecular identification for species of hosts

Genomic DNA was extracted from approximately 30 mg of lungs of small mammals infected with hantavirus using a Tiangen TIANamp Genomic DNA Kit (Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. Complete cytochrome b (Cytb) gene as a phylogenetic marker was amplified by PCR and the PCR products were subjected to sequencing [15].

Phylogenetic analysis

The target nucleotide sequences were compared to sequences that were available in public databases using BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The other hantavirus sequences for HTNV, DOBV, SEOV, SNV, ANDV, BAYV, PUUV, PHV, ISLAV, and LX307 (which should not be identified as TULV [18]) were set as outgroups. The phylogenetic tree was constructed using MEGA neighbor joining method (MEGA, version 7.0; http://www.megasoftware.net/) [19] and bootstrap values were 70%, calculated from 1000 replicates.

Results

Vole species identification

A total of 198 voles were captured in a Narati pasture (43°23′ N, 83°96′ E, altitude: above 1800 m) in Xinyuan County, Xinjiang, China. These small mammals came from 20 burrow-groups near the streams and each group contained 15–30 occupied burrows.

All of the voles were M. obscurus, initially classified by morphological identification. Two voles from hantavirus positive-groups that showed slight differences in morphology were further identified by PCR amplification and sequencing of Cytb gene, which confirmed that the voles were M. obscurus with identical nucleotide sequences (GenBank accession No: KX058268-KX058269).

Hantavirus detection and evolution analysis

Hantavirus L gene fragments were successfully amplified and sequenced from 31 out of 198 voles, such that 16% of voles were infected with hantavirus. We further amplified and sequenced the S gene of hantavirus from the two L-gene-positive samples to determine the species of the viruses. The sequence analysis showed the nucleotide sequences of two isolates are identical, called Tula xinjiang4 (GenBank accession No: KX270414). Combining with the sequences published in the GenBank, the phylogenetic analysis showed 5 genetic lineages (I–V) likely based on the geographic distribution in Eurasia (Fig. 1). Tula xinjiang4 is very similar to Karata322 (isolated from eastern Kazakhstan, altitude: 1200 m) [7] with which it has been found to share 92.1% nucleotide identity and 98.2% amino acid identity, and Omsk23 (isolated from southwest region of Siberia in Russia), with which it has been found to share 87.5% nucleotide identity and 97.5% amino acid identity. The referenced sequences were isolated from European common voles, M. arvalis (obscurus), M. gregalis, and Pitymys subterraneus, respectively. (Additional file 1: Table S1 and Table S2).

Fig. 1
figure 1

Phylogenetic tree of S gene fragment. The NJ tree was constructed with the S gene fragments (nt 43–1235). Tula virus, strain (GenBank accession number)/carrying host/Location: Tula249 (Z30944), Tula76 (Z30941), Tula175 (Z30943), Tula53 (Z30942), Tula23 (Z30945)/European common voles/Central Russia; Omsk23 (AF442621)/Microtus gregalis/West Siberia, Russia; Karatal322 (AM945877)/ Microtus obscurus/Kazakhstan; GermD5–98 (AF289819), GermD63–98 (AF289821), GermD17–98(AF289820)/Microtus arvalis/Germany; Lodz1 (AF063892), Lodz2 (AF063897)/Microtus arvalis/Poland; Kosice667 (Y13980), Kosice144 (Y13979)/European common vole/East Slovakia; Serbia (AF017659)/Pitymys subterraneus/Serbia-Yugoslaviaia; Germ20 (AF164093)/ Microtus arvalis/Germany; Croatia (AF164094)/Microtus arvalis/Croatia; Koziky47 (AJ223600), Koziky76 (AJ223601)/European common vole/Czech Republic; Moravia02 (Z49915), Moravia86 (Z48573), Moravia93 (Z48574); Moravia94 (Z48741)/M. arvalis/Czech Republic. Hantavirus, strain (GenBank accession number)/carrying host: HTNV, Hantaan virus, 76–118 (M14626)/Apodemus agrarius; DOBV, Dobrava-Belgrade virus, Dobrava (L41916)/Apodemus flavicollis; SEOV, Seoul virus, Z37 (AF187082)/Rattus norvegicus; SNV, Sin Nombre virus, Sin Nombre virus 77,734 (KF537003.1)/ Peromyscus maniculatus; ANDV, Andes virus, C9717869 (NC-003466)/Oligryzomys longicaudatus; BAYV, Bayou virus (L36929)/Oryzonys palustris; PUUV, Puumala virus (KT247597)/Myodes glareolus; PHV: Prospect Hill virus, (Z49098)/ Microtus pennsylvanicus; ISLAV, Isla Vista virus (U19302)/Microtus pennsylvanicus; LX307, unconfirmed hantavirus (HM756286)/Eothenomys milemiletus/Yunnan Province, China. *virus strain reported in this article

Discussion

Few studies have explored the epidemiology of hantaviruses in Xinjiang. Only Seoul virus (SEOV, a hantavirus), found in Rattus norvegicus in Urumqi, Xinjiang, has been assessed in this way [20, 21]. The present study showed for the first time that M. obscurus [3, 22] carries TULV in China. The distribution of hantavirus is likely related to the distribution of rodents [1, 23]. In Xinjiang, there are 69 species of rodents, which belong to 10 families and 34 genera, accounting for 40% of rodent species in China [23, 24]. Among the small mammals, there are 19 species in subfamily Microtinae and 7 species in the genera Microtus [23]. In this way, it is likely that other voles or small mammal species may also serve as hosts for TULV or hantavirus in Xinjiang. This needs further investigation.

Hantavirus infection is a zoonosis and it remains a severe public health problem. However, very few patients infected with TULV in Europe have shown symptoms [8,9,10] with the most serologically positive cases being asymptomatic [11]. It may be relevant whether the populations in these areas where M. obscurus are likely to be infected with TULV.

The two locations, Yili in China and eastern Kazakhstan, where the TULV-infected voles were found, are geographically close and have similar landscapes; both are in the Tianshan Mountains. Viral sequence analysis showed TULV Xinjiang4 to be very similar to Karata322 and the overall lineages appeared closely related to those isolates from central Russia and Siberia, indicating these viruses share a common ancestor (bootstrap support value 98%, on the NJ-tree).

Hantavirus contains three genes, negative-stranded large (L), middle (M), and small (S) genomic RNA segments encode a polymerase, M proteins composed of two surface glycoproteins (Gc and Gn), and the nucleocapsid protein (N protein), respectively. The complete S segment sequence of TULV strain is 1830 nt in size and the sequence includes the following structures: a 5′-untranslated region (UTR, nt 1–42), an open reading frame for the 430 amino acids (AA-long N protein, nt 43–1335) and a 3′-UTR (nt 1336–1830) [7]. In our study, the predicted N-protein sequence revealed that the N protein sequence region (amino acid, AA 244–273) of TULV contains a highly variable region. There are 3 AA (AA 256–258) deletions in Tula Xinjiang4 and other TULV strains, 4 AAs deleted in hantavirus (HTNV) and SEOV (AA 257–260) compared with Prospect Hill virus (PHV) and Puumala virus (PUUV) (Fig. 2). It may be useful to determine whether the deletion of these sequence impacts the function of N protein.

Fig. 2
figure 2

Alignment of partial nucleocapsid protein sequences of Tula virus (TULV). Amino acids that matched those of TULV are highlighted in gray. ORF: Open Reading Frame; UTR: Untranslated Region

Conclusions

TULV was identified in Narati pasture, Yili region, Xinjiang, western China by molecular detection. The high prevalence of the virus in M. obscurus indicates Tula virus is naturally transmitted in this area. It may be useful to determine whether this virus infects other hosts, such as humans.