PCR-based screening and phylogenetic analysis of rat pegivirus (RPgV) carried by rodents in China.

Rodent-borne pegiviruses were initially identified in serum samples from desert wood-rats in 2013, and subsequently in serum samples from commensal rats in 2014. However, the prevalence and phylogenetic characteristics of rodent pegiviruses in China are poorly understood. In this study, we screened serum samples collected from wild rats in southern China between 2015 and 2016 for the presence of rat pegivirus (RPgV) by PCR. Among the 314 serum samples from murine rodents (Rattus norvegicus, Rattus tanezumi, and Rattus losea) and house shrews (Suncus murinus), 21.66% (68/314) tested positive for RPgV. Out of these, 23.81% (62/219) of samples from R. norvegicus tested positive, which was significantly higher than that for the other species: 7.69% (1/13), 5.88% (2/34), and 6.25% (3/48) for R. tanezumi, R. losea, and S. murinus, respectively (χ2=18.91, P<0.001). Phylogenetic analysis revealed clustering of viral sequences in the main rodent clade. Analysis of the 3 near-full-length genome sequences of RPgV obtained in this study showed that these viruses exhibited mean nucleic acid and amino acid identities of 94.1% and 98.5% with Chinese RPgV strains, and 90.3 and 97.1% with an RPgV strain from the USA, respectively. This study provides novel insights into the geographic distribution of rodent pegiviruses in China, and identifies potential animal hosts for future studies of these pegiviruses.

Since 1995, a number of novel members of the family Flaviviridae have been discovered, and tentatively assigned to a new genus named Pegivirus (formerly GB virus) [10]. The members of this genus are positive-sense, single-stranded RNA viruses of approximately 10 kb in length, which contain only one open reading frame (ORF). Their genomic structure resembles that of hepacivirus [13]. Intriguingly, unlike hepacivirus, pegivirus is primarily lymphotropic, results in asymptomatic infection, and is benign in humans and other animals [5, 6, 9, 14, 17]. A comprehensive understanding of the natural reservoirs for this emerging virus would provide valuable insights into the transmission and geographic distribution of pegivirus. Although pegivirus was first identified in humans, the identification of tamarins infected with the GB hepatitis agent suggested a wider host range for this virus. Homologous pegiviruses were subsequently reported in various mammalian species, including non-human primates (GBV-A or simian pegivirus, SPgV), horses (equine pegivirus, EPgV), pigs (porcine pegivirus, PPgV), bats (GBV-D or bat pegivirus, BPgV), and wild rodents (rodent pegivirus, RoPgV) [1, 2, 8, 10,11,12].

The order Rodentia comprises over 2,200 living species, which represents nearly 40% of the total number of mammals [18]. Numerous rodent species act as wild and synanthropic reservoirs of myriad zoonotic agents. In southern China, Rattus norvegicus and Rattus tanezumi represent 2 of the predominant commensal rodent species. The impact of the frequent interactions between rodents and humans has been actively studied. The identification of new rodent-borne and insectivore-borne viruses has advanced our understanding of both host range and epizootiology. In this study, we screened serum samples from rodents and house shrews in southern China for infection with rodent pegivirus. In addition, partial sequences of the NS3/NS5B gene and 3 near-full-length genome (NFLG) sequences were determined for detailed genetic characterization of this class of viruses.

MATERIALS AND METHODS

Sample collection

Between August 2015 and March 2016, a total of 314 rodents and shrews were captured using cage traps (280 × 120 × 100 mm) from Baiyun (GZ.BY, n=189) and Yuexiu (GZ.YX, n=56) districts in Guangzhou city in Guangdong province and from Xiamen city in Fujian province (FJ.XM, n=69). The collection sites are depicted in Fig. 1. The animals were all captured in or close to human dwellings, and the respective ages were determined based on body weight, as previously described [3]. The animals captured included 43 juveniles (<80 and <23 g for rats and shrews, respectively), 97 sub-adults (80–180, 80–200, and 23–28 g for female rats, male rats, and shrews, respectively), and 174 adults (>180, >200, and >28 g for female rats, male rats, and shrews, respectively). The male-to-female ratio was 1.32. The animals were anesthetized with diethyl ether prior to drawing cardiac blood. All the serum samples were separated by centrifugation, frozen, and stored immediately at −80°C; the stored samples were thawed at 4°C prior to processing.

Fig. 1.

A representative map with information of the collection sites. Triangles indicate the sampling locations.

Identification of rodent species

Rodent species were confirmed by preliminary morphological identification and sequencing of the cytochrome b (cyt-b) gene fragment of the mitochondrial genome, which is a widely recognized technique for classifying biological samples. The primers cytb-F (forward, 5ʹ- CGAAGCTTGATATGAAAAACCATCGTTG-3ʹ) and cytb-R (reverse, 5ʹ- AAACTGCAGCCCCTCAGAATGATATTTGTCCTCA-3ʹ) were used to amplify a fragment containing a partial nucleic acid sequence of the cyt-b gene. The following amplification conditions were used: 95°C for 11 min; followed by 35 cycles of 95°C for 45 sec, 50°C for 45 sec, and 72°C for 90 sec; followed by 72°C for 30 min [7]. The sequences obtained were compared to known DNA sequences from reliable databases and voucher specimens, and analyzed to make high-confidence taxonomic assignments.

RNA extraction and targeted molecular analysis

RNA was extracted from 200 µl of each blood sample using a High Pure Viral RNA kit (Roche, Mannheim, Germany) according to the manufacturer’s instructions, and eluted in a final volume of 50 µl. cDNA was generated from the RNA using random hexamers and a Transcriptor First Strand cDNA Synthesis kit (Roche). Then, pegivirus partial helicase gene sequences were amplified using nested PCR. In addition, RNA-dependent RNA polymerase genes were amplified and compared to reference sequences to confirm the positive pegivirus test result. Specific degenerate primer sets were designed. The rat Pegivirus (RPgV)NS3 gene was amplified using primer pairs as described previously [3]. PCR of the RPgV NS5B gene employed the outside primer pair of RPgV-NS5b-1F (5ʹ-ACCCGACCAACTTTCAAGCCMCC-3ʹ) and RPgV-NS5b-1R (5ʹ-CAGCCGAAGTGAGGTCTGCGTGC-3ʹ), and the inside primer pair of RPgV-NS5b-2F (5ʹ- CCGACCAACTTTCAAGCCMCCYC-3ʹ) and RPgV-NS5b-2R (5ʹ-AAGTGAGGTCTGCGTGCCTAGTA-3ʹ). Two microliters of cDNA were added to a 25 µl reaction mixture using the Premix TaqTM reagents (Takara, Dalian, People’s Republic of China), according to the manufacturer’s protocol. The first and second rounds of RPgV-NS3 gene amplification were carried out under the following conditions: 94°C for 5 min (1 cycle); 94°C for 40 sec, 57°C for 40 sec, and 72°C for 1 min (35 cycles); followed by incubation at 72°C for 10 min; and storage at 4°C. The first and second rounds of RPgV-NS5B gene amplification were carried out under similar conditions with minor modifications: 94°C for 5 min (1 cycle); 94°C for 40 sec, 55°C for 40 sec, and 72°C for 1 min (35 cycles); and 72°C for 10 min (1 cycle). For amplification of the RPgV NFLG sequence, 15 pairs of primers were designed according to the reference sequence (GenBank accession number KJ950934) and used with PrimeSTAR HS (Premix) reagents (Takara, Dalian, People’s Republic of China). The PCR products were separated in a 1% agarose gel, and directly sequenced by the company Sangon Co. (Sangon, Guangzhou, People’s Republic of China). Any anomalies or bases that were ambiguous were clarified by sequencing in both directions.

Sequence and phylogenetic analysis

Sequence editing was performed using BioEdit and DNAMan. The nucleotides were compared to sequences reported in the GenBank non-redundant nucleotide database (NT) using BLASTn and aligned using Clustal W. Nucleic acid similarity were performed online using BLAST (https://blast.ncbi.nlm.nih.gov/ ). All data were exported to the MEGA6 software. A maximum-likelihood phylogenetic tree was constructed using the Tamura-Nei model with 1,000 bootstrap replicates. The similarity between pegivirus sequences was plotted using the SimPlot 3.5.1 software.

Ethics statement

Animals were caught by snap traps with permission of respective authorities. The treatment was carried out in conformity with the guidelines for the Laboratory Animal Use and Care from Southern Medical University and the Rules for the Implementation of Laboratory Animal Medicine from the Ministry of Health, China. The study protocol has been approved by the Animal Ethics and Welfare Committee of the School of Public Health, Southern Medical University.

Accession numbers

The 3 NFLG sequences have been submitted to the GenBank database with the accession numbers MG273686-MG273688.

RESULTS

Prevalence of pegivirus in rodents and shrews

Specimens of rodent and shrew serum were collected, including 219 samples from R. norvegicus, 13 from R. tanezumi, 34 from Ruttus losea, and 48 from Suncus murinus (Table 1). To investigate the frequency of pegivirus infection, each of the 314 samples was individually processed and screened by nested PCR, and 21.66% (68/314) of the serum samples from both Guangzhou and Xiamen cities yielded positive results for pegiviruses. The highest prevalence of pegivirus infection, 51.59% (162/314), was found in Norway rats (R. norvegicus). This was significantly higher than that for the other species: 7.69% (1/13), 5.88% (2/34), and 6.25% (3/48) for R. tanezumi, R. losea, and the insectivore S. murinus (χ2=18.91, P<0.001). The percentages of juvenile (11.63%, 5/43) and sub-adult (12.37%, 12/97) rats positive for infection were similar; however, for adult rats the percentage was higher (29.31%, 51/174).

Table 1.

Distribution of rodent pegivirus in rats and shrews according to geographic locations (%, n)

Species	Collection sites			Total (Positive, %)
	GZ.BY (2015.3-2016.4)	GZ.YX (2015.10)	FJ.XM (2015.10)
Rattus norvegicus	29.08 (41/141)	30.36 (17/56)	18.18 (4/22)	28.31 (62/219)
Rattus tanezumi	33.30 (1/3)		0 (0/10)	7.69 (1/13)
Rattus losea			5.88 (2/34)	5.88 (2/34)
Suncus murinus	6.67 (3/45)		0 (0/3)	6.25 (3/48)
Subtotal	23.81 (45/189)	30.36 (17/56)	8.70 (6/69)	21.66 (68/314)

GZ.BY, Baiyun district, Guangzhou city in Guangdong province; GZ.YX, Yuexiu district, Guangzhou city in Guangdong province; FJ.XM, Xiamen city in Fujian province. There were 2 sample collection regions in Xiamen city: Tong’an district and Huli district.

Phylogenetic analysis of partial helicase and RNA-dependent RNA polymerase gene sequences

Partial (238 bp) NS3 gene sequences and partial (602 bp) NS5B gene sequences were successfully amplified from the serum samples. BLAST was used to align the sequences and determine their similarity. The data obtained are indicative of a close relationship between the rodent pegiviruses identified in the present study and the 2 previously reported (accession numbers KJ950934 and KC815311). Next, 10 out of the 46 typical sample sequences were chosen for further phylogenetic analyses. All the sequence variants in NS3 genome region were congruent with branching orders in the NS5B region, indicating the lack of recombination in these sequences. In general, the pegiviruses formed 3 distinct phylogenetic lineages. The first lineage comprised sequence variants from humans, primates, and bats; while the second lineage was composed of novel porcine pegivirus clustered together with variants from bats and primates. The second human pegivirus (HPgV-2) and sequence variants from rodents and bats were classified into the third lineage. Multiple sequence alignment showed that all the sequences had high levels of nucleotide identity (bootstrap >70) to each other over the NS3 and NS5B regions. These sequences formed a separate clade within the genus Pegivirus (Fig. 2a and 2b) together with 2 other distinct rodent pegiviruses identified from blood samples of wood rats and Norway rats.

Fig. 2.

Phylogenetic tree constructed by the neighbor-joining method based on partial nucleotide sequences of the (a) NS3 (238 bp) gene and (b) NS5B (602 bp) gene of rat pegivirus (RPgV), rodent pegivirus (RoPgV), bat pegivirus (BPgV), human pegivirus (HPgV), human pegivirus type 2 (HPgV-2), porcine pegivirus (PPgV), simian pegivirus (SPgV), and equine pegivirus (EPgV). The RPgV sequences obtained in this study are marked by triangles. Bootstrap values (1,000 replications) >70% are indicated at nodes.

Characterization of the pegivirus genome in rodents

For a detailed phylogenetic analysis of the rodent pegivirus, we included 3 NFLG sequences (missing only the terminal regions) from R. norvegicus, namely YX51 (9,945 nt), SMU217 (9,883 nt), and SMU415 (9,963 nt), representing different collection sites in the Yuexiu and Baiyun districts. The lengths of these genomes ranged from 9,883 to 9,963 nt and contained only one ORF, which encoded a putative multi-functional polyprotein predicted by comparison with polyproteins of rat (GenBank accession number KJ950934) and rodent (GenBank accession number KC815311) pegiviruses. A phylogenetic tree created based on the whole reading frame demonstrated that rat pegivirus belonged to the genus Pegivirus, which is consistent with our analysis using partial NS3 and NS5B sequences (Fig. 3). Overall, the 3 rat pegivirus sequences generated in our study branched very deeply next to the sequences of R. norvegicus and Neotoma lepida, and were appreciably different from the sequences of pegivirus from humans, primates, and other original hosts in this group. All the pegivirus variants found in R. norvegicus were clustered together. The sequences identified in the present study showed a mean nucleic acid identity of 94.1 and 90.3% with the Chinese RPgV strains and the RPgV strain from New York, USA (GenBank accession number KJ950934), respectively. The amino acid identity was similarly high: 98.5% with the Chinese RPgV strains, and 97.1% with both the Chinese strains and the RPgV strains from USA. The differences between the strains identified here and the USA. RPgV sequences were mainly in terms of synonymous mutations at the nucleotide level (Table 2). When compared with pegiviruses from other hosts, the nucleic acid (amino acid) identities of rat-derived pegiviruses isolated in this study were 91.1% (97.3%), 59.3% (56%), 43.8% (32.4%), 45% (34.8%), 42.7% (33.2%), 42.9% (32.9%), 42.3% (31.9%), and 42.2% (32.6%) compared to the complete coding regions of RPgV, RoPgV, BPgV, HPgV-2, human pegivirus type 1 (HPgV-1), PPgV, SPgV, and EPgV, respectively (Table 3). We further analyzed possible recombinations and the differences between RPgV and pegivirus sequences from other hosts using the Simplot software: no recombination event was identified (Fig. 4).

Fig. 3.

Phylogenetic tree constructed by the neighbor-joining method based on amino acid sequences of rat pegivirus (RPgV), rodent pegivirus (RoPgV), bat pegivirus (BPgV), human pegivirus (HPgV), human pegivirus type 2 (HPgV-2), porcine pegivirus (PPgV), simian pegivirus (SPgV), and equine pegivirus (EPgV). The RPgV sequences obtained in this study are marked by triangles. Bootstrap values (1,000 replications) >70% are indicated at nodes.

Table 2.

Identity of the near-full-length genome sequences of rat pegivirus

Aligned region	Polyprotein	Y	E1	E2	X	NS2	NS3	NS4A	NS4B	NS5A	NS5B
Comparison with the Chinese strains
Nucleotide identity (%)	94.1	92.6	94.3	93.9	92.1	91.4	96.6	96.2	94.9	95.4	95.9
Amino acid identity (%)	98.5	94.6	99.3	98.4	99.1	96.6	99.9	100	100	99.5	99.6
Comparison with both the Chinese strains and the American strain (RPgV)
Nucleotide identity (%)	90.3	85.2	92.0	90.8	88.5	88.6	93.3	91.1	91.5	92.5	92.8
Amino acid identity (%)	97.1	83.0	97.1	98.0	97.5	94.1	99.6	99.1	99.6	98.1	99.1

Table 3.

Nucleotide and translated amino acid sequence identity of the near-full-length sequence of rat pegivirus and other pegiviruses

Nucleotide identity (%)										Genbank accession No.
Virus	Present study	RPgV	RoPgV	BPgV	HPgV-2	HPgV-1	PPgV	SPgV	EPgV
Amino		91.1	59.3	43.8	45.0	42.7	42.9	42.3	42.2	Present study
Acids	97.3		59.2	43.8	44.5	42.9	43.2	42.2	42.2	RPgV KJ950934
	56.0	55.4		43.1	45.1	43.1	43.5	42.8	41.6	RoPgV KC815311
	32.4	32.3	32.3		41.7	49.2	47.9	49.2	46.5	BPgV KC796083
	34.8	34.5	34.9	27.8		39.5	41.4	40.1	40.5	HPgV-2 KX528230
	33.2	32.9	32.5	41.6	28.9		47.0	63.6	45.8	HPgV-1 AB003292
	32.9	32.7	32.8	39.2	29.1	40.8		46.9	46.0	PPgV MH345724
	31.9	31.8	31.9	41.7	28.6	65.4	40.0		44.5	SPgV AF070476
	32.6	32.3	32.2	37.1	27.7	38.3	36.7	37.9		EPgV KC410872

RPgV, rat pegivirus; RoPgV, rodent pegivirus; BPgV, bat pegivirus; HPgV-1, human pegivirus type 1; HPgV-2, human pegivirus type 2; PPgV, porcine pegivirus; SPgV, simian pegivirus; EPgV, equine pegivirus.

Fig. 4.

Characterization of rat pegivirus_SMU217 strain. The similarity between RPgV_SMU217 and the reference pegivirus sequences was plotted using Simplot (version 3.5.1) software. The sequences used in the phylogenetic analysis were rat pegivirus (RPgV, KJ950934), rodent pegivirus (RoPgV, KC815311), bat pegivirus (BPgV, KC796081), equine pegivirus (EPgV, KC410872), human pegivirus type 1 (HPgV-1, AB00328), human pegivirus type 2 (HPgV-2, ABT0029A), porcine pegivirus (PPgV, KU351671), and simian pegivirus (SPgV, KF234525).

DISCUSSION

The detection and characterization of animal virus homologs yields valuable insights into host range, epidemiology, and pathogenesis. This is, to the best of our knowledge, the first report of the prevalence of pegivirus-infected rodents in southern China. Rodent pegivirus was first found in 2 mouse species, white-throated wood rats (Neotoma albigula) and deer mice (Peromyscus maniculatus), with the infection rate estimated to be less than 1.5% [5]. Rats are considered to be pests that exist in large population in urban environments and impact human health. One research group assessed the presence and prevalence (15%) of pegivirus in R. norvegicus in New York city [3]. In addition, Nguyen et al. carried out a screening in Vietnam for the presence of rodent pegivirus in 638 serum and 470 liver samples from 8 and 6 rodent species, respectively. In their study, no liver samples tested positive for pegivirus, although 2.3% (2/87) of their R. tanezumi serum samples were pegivirus positive [16]. Likewise, we used PCR to identify rodent pegivirus sequences in serum samples from R. norvegicus and R. tanezumi, which are predominant in southern China. A total of 314 serum samples were collected from 4 regions of 2 cities in China: the Baiyun and Yuexiu districts of Guangzhou city, and the Tong’an and Huli districts of Xiamen city. The prevalence of rodent pegivirus RNA was 21.66%, which was higher than the prevalence reported in USA [3]. The detection of pegivirus infection in R. tanezumi is interesting, since only one sample tested positive in this study. The Asian house rat (R. tanezumi) is closely related to the brown rat (R. norvegicus) and is sympatric with them in southern China. Both are primarily commensal murine rodents, which inhabit human dwellings and livestock sheds. Genetic analysis of the pegivirus sequence from R. tanezumi showed high identity (91–95%) with the Norway rat pegivirus isolates. These observations might provide some initial indication of the route of rodent pegivirus transmission. In addition, the RPgV positivity rates varied from 5.88% in R. losea to 6.25% in house shrews, which mainly inhabit fields. Many different zoonotic viruses are carried by insectivore hosts, for instance, hantavirus, cytomegalovirus, and herpesvirus. However, the only report of an insectivore-borne pegivirus was a partial Tupaia glis NS3 pegivirus sequence, with 64–67% similarity to bat pegivirus (GenBank accession numbers JQ922487-JQ922489). In our study, 3 out of 48 serum samples from the insectivore S. murinus tested positive for pegivirus. As house shrews and murine rodents have overlapping habitats that are in or close to human dwellings, pegivirus may be transmitted between them via fecal-oral contact, exposure of bite wounds to body fluids, close contact, or other transmission routes. Further investigations with an expanded shrew sample set are needed to confirm our findings.

The rat pegivirus has been detected in the bladder, urine, fecal matter, heart, kidney, spleen, liver, and even in brain samples of rats [3], suggesting that RPgV is able to infect rats effectively. However, the route of transmission of rodent pegivirus between individuals has not been investigated in detail. Much more attention must be paid to tissue tropism in this context.

Studies of virus reservoirs are a critical component of any integrated public-health response to emerging zoonotic diseases [4]. The phylogenetic analysis with partial NS3 and NS5B nucleotide sequences from pegiviruses revealed that these viruses were part of the main rodent-derived pegiviruses clade, whilst the pegiviruses found in R. norvegicus and R. tanezumi have expanded the taxonomy of rodent pegiviruses. The pegiviruses tended to be species-specific, and the phylogenetic analysis revealed co-evolution between most of the identified pegivirus and their hosts, except for bat pegivirus (Fig. 2a and 2b). The rat pegivirus occupied the same branch of the phylogenetic tree, indicating that they are adapted to their hosts. However, additional sampling data is required to improve the reliability of this evolutionary history. Also, further studies on the relationship between viral protein and hosts are needed to clarify this issue in a detailed manner. Additionally, the 3 NFLG polyprotein sequences generated in the present study have contributed to our understanding of the phylogeny and biology of pegivirus. Recombination and the difference between RPgV and pegivirus sequences from other hosts were further analyzed in our study using the Simplot software, and no recombination event was identified (Fig. 4), which was consistent with a previous study showing lack of interspecific recombination in the pegiviruses [15]. Rodent pegivirus sequences from distant geographic origins, exemplified by samples collected from New York city in the USA and southern China, were aligned and compared. There was over 90% similarity at the nucleic acid level among the sequences from these regions, which suggests that the dominant strain isolated from R. norvegicus, which has invaded China and the USA, has acquired an indigenous virus due to horizontal transmission.

The genus Pegivirus is a novel genus of the family Flaviviridae. Analysis of conserved gene motifs involved in enzymatic function may elucidate the evolutionary relationship of pegiviruses to each other and to other members of the family Flaviviridae. Comparative phylogenetic analysis of conserved regions in the partial helicase and RNA-dependent RNA polymerase regions were congruent with other rodent pegiviruses, which clustered consistently with pegiviruses of bats, primates, and humans (HPgV-2), suggesting these viruses are ancient members within the same genus. Kapoor et al. has reported that the most conserved regions within viruses of the genus Pegivirus were the NS3 and NS5B genes, with highly divergent sequence in the envelope glycoprotein [5]. In our study, RPgV sequences revealed both similarities and differences from previously identified pegivirus isolates. The prediction of cleavage sites of nonstructural proteins gave comparable sizes to homologs from other pegiviruses. However, 6 N-linked glycosylation sites were present in the pegivirus envelope, compared to 9 sites in the deduced envelope protein of N. lepida pegivirus. Due to constraints in time and facilities, we were unable to determine the whole genome sequences including the 5ʹ- and 3ʹ-flanking regions. The next step could, therefore, be to use the rapid amplification of cDNA ends method for this purpose. Further, the biophysical characterization of rodent pegivirus particles has not been fully elucidated, and their origin and composition remain a mystery. Further studies may help unravel the mechanisms of virus infection and persistence.