Seoul Virus and Hantavirus Disease, Shenyang, People’s Republic of China

This outbreak was caused by the virus circulating in local wild rats through infection of laboratory rats.

H antaviruses, members of the family Bunyaviridae, genus Hantavirus, cause 2 human zoonoses, hemorrhagic fever with renal syndrome (HFRS) in Asia and Europe and hantavirus pulmonary syndrome in North and South America (1). In their natural hosts, rodents of the families Muridae and Cricetidae, hantaviruses cause chronic infec-tion with no apparent harm (2,3). HFRS has been recognized as a serious public health problem in China since 1955 (4,5). The disease is caused mainly by the Hantaan virus (HTNV), transmitted by the striped fi eld mouse (Apodemus agrarius), and Seoul virus (SEOV), transmitted by the brown Norway rat (Rattus norvegicus) (4,6).
Transmission of hantaviruses among rodents and from rodents to humans generally occurs through inhalation of aerosolized excreta (7). HFRS outbreaks have occurred among farmers and workers during close contact with infected rodents in disease-endemic areas. Hantavirus infections have also occurred among technicians and researchers after handling laboratory rodents. The fi rst report showed that contact with hantavirus-infected laboratory rats caused a HFRS outbreak among 13 doctors and 1 veterinarian at medical research institutions in Japan (8). Since 1975 and1978, laboratory animal-associated HFRS outbreaks have been reported in several countries (9)(10)(11)(12)(13)(14). Dozens of hantavirus infections in laboratory animals also occurred during the 1980s in China (15). Furthermore, 16 HFRS cases associated with laboratory rats occurred in 1983 in the Shanxi province (16). However, only a few reports have attempted to characterize the etiologic agents of the outbreaks and clarify the origin of hantaviruses causing infections in humans and laboratory animals (9,17).
Shenyang City (the capital of Liaoning Province) is located in northeastern China. Shenyang has always been one of the most seriously affected areas in China since the fi rst outbreak of HFRS in 1958 (5,18). A total of 470 HFRS cases were reported in Shenyang in 2005; most of these cases occurred among farmers in the suburbs and the rural areas of Shenyang. Previous studies have shown the presence of 2 hantaviruses carried by rodents: HTNV, carried by striped fi eld mice, and SEOV, carried by Norway rats in Shenyang (18). Serologic and genetic analyses suggest that the HFRS outbreak was caused by transmission of SEOV, which was circulating among local wild rats; the wild rats passed the virus to laboratory rats, which then infected humans. Our study characterizes etiologic agents of these outbreaks among students and clarifi es the origin of hantaviruses causing infections in humans and laboratory animals.

Patients and Serum Samples
HFRS cases were defi ned by a national standard of clinical criteria and confi rmed by detecting antibodies against hantavirus in serum samples obtained in 2006. Serum samples were collected from patients with clinical signs of HFRS and sent to the Shenyang Center for Disease Control and Prevention (Shenyang CDC) for detection of hantavirus-reactive antibodies, and then to the Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention for further serologic and genetic characterization. Shenyang CDC conducted the HFRS epidemiologic studies. Information such as the date of onset of illness, fever, living conditions, history of exposure in dormitory and fi eld, and clinical symptoms and signs was obtained and recorded.

Laboratory Rats and Mice
All laboratory rats (Wistar) and mice (BALB/c) housed in the same animal facility in a pharmaceutical laboratory building were obtained from the Laboratory Animal Center of Shenyang Pharmaceutical University and were sampled. These rodents were generally >6 months of age and had been in the animal facility for >1 month. Serum and lung tissue samples were collected from all laboratory animals, placed in vials, stored immediately at -196 o C, and transported to the laboratory for processing.

Trapping of Rodents
During 2006-2007, wild rodents were captured on the grounds of the animal facility in the pharmaceutical laboratory building, in the vicinity of the laboratory Animal Center of Shenyang Pharmaceutical University during 1 month after the outbreak in 2006, and in a major HFRS-endemic focus in the suburbs of Shenyang in the autumn of 2006 and the spring of 2007 using snap-traps baited with peanuts. Lung tissue samples were taken from dissected animals, placed immediately into vials and stored at -196 o C and, then transported to a laboratory for processing.

Serologic Assays
Human serum samples were tested for immunoglobulin (Ig) G and IgM antibodies against HTNV and SEOV by indirect immunofl uorescent assay (IFA). Serum samples from laboratory rodents were tested for IgG antibodies to SEOV or HTNV. IgG and IgM IFAs were performed with HTNV (strain 76-118)-and SEOV (strain L99)-infected Vero E6 cells. Cells were spread onto slides, air-dried, and fi xed with acetone. Samples were serially diluted in 2-fold steps in phosphate-buffered saline, starting with the initial dilution of 1:2, then added to the cells, and incubated for 90 min at 37°C. Slides were washed in phosphate-buffered saline and incubated with fl uorescein isothiocyanate (FITC)labeled rabbit antihuman IgG and IgM antibodies (Sigma, St. Louis, MO, USA), which are gamma-chain-and muchain specifi c, respectively, at 37°C for 30 min. For rodent samples, FITC-labeled goat antimouse or antirat IgG was used. IgG titers >40 and IgM titers >20 were considered positive.

Detection of Hantavirus Antigen
Viral antigens in the lung tissue (frozen sections) of rats and mice were detected by using indirect IFA as described previously (19), with rabbit anti-SEOV/L99 and HTNV/76-118 hantavirus antibodies and FITC-labeled goat antirabbit IgG (Sigma). Scattered, granular fl uorescence in the cytoplasm was considered a positive reaction ( Figure 1).

Reverse Transcription-PCR (RT-PCR) and Sequencing
Total RNA was extracted from rodent lung tissues by using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer's instructions and subjected to RT-PCR for amplifi cation of partial hantavirus small (S) segment sequences. cDNA was synthesized with avian myeloblastosis virus reverse transcriptase (Promega, Beijing, China) in the presence of primer P14 (20). Partial Ssegment sequences of SEOV (nt 620-999) were amplifi ed from SEOV by using primers HV-SFO and HV-SRO for initial PCR (21), and primers SEO-SF and SEOV-SR for the second round of amplifi cation (22). For amplifi cation of partial S-segment sequence (nt 514-1026) from HTNV, the same primer pair HV-SFO/HV-SRO was used for initial PCR and the primer pair HSF /HSR was used for nested PCR (22).
The PCR products (380 bp and 513 bp, respectively) were gel-purifi ed by using QIAquick Gel Extraction kit (QIAGEN, Beijing, China) according to the manufacturer's instructions and cloned into the pMD18-T vector (TaKaRa, Dalian, China). The ligated products were transformed into JM109-competent cells. DNA sequencing was performed with the ABI-PRISM Dye Termination Sequencing kit and an ABI 373-A genetic analyzer (Applied Biosystems, Carlsbad, CA, USA). At least 2 cDNA clones were used to determine each viral sequence. In case of discrepancy, a third cDNA clone was sequenced.

Patients and Survey Results
From March 8 through April 22, 2006, symptoms of hantavirus infection developed in 8 postgraduate students (5 men and 3 women), who studied at Shenyang Pharmaceutical University located in the center of the Shenyang. All patients met the national clinical criteria of HFRS, required hospitalization, and were treated in Shenyang Infectious Hospital. Fever, proteinuria, and mild hemorrhagic complications were observed in all patients, but without the distinct clinical stages seen in the severe form of the disease caused by HTNV (Table 1). Other clinical symptoms such as weakness, backache, nausea, vomiting, abdominal pain, eyeball pain, and hypotension were not observed.
The 8 students lived in different rooms in the 2 dormitories on the university campus. They had no history of exposure to wild rats in their rooms. Notably, all their roommates had been in good health. Further, the students neither performed fi eld studies nor had a history of exposure to rats or mice in the fi eld during the previous 6 months. All 8 students conducted their research in the same department and had direct contact with a colony of laboratory rats and mice in the animal facility in the pharmaceutical laboratory building. Hantavirus infection did not develop in any person who did not have direct contact with the laboratory rats and mice.

Serologic and Genetic Investigation of Patient Serum Samples
Serum samples from all 8 patients were collected at day 1 of hospitalization (2-4 days post onset of fever). Samples were tested for IgM and IgG antibodies by IFA using SEOV-or HTNV-infected cells ( Table 2). All serum samples showed higher IgM and IgG titers in SEOV-specifi c IFA. In 6 of 8 sera, the IgG titers against SEOV were 4-fold higher; in the remaining 2 serum samples the titers against SEOV were 2-fold higher ( Table 2). These results suggested that the HFRS cases were caused by SEOV.
Total RNA was extracted from all serum samples and analyzed by SEOV S-segment-specifi c or HTNV S segment-specifi c RT-PCR. Hantavirus genome sequences were amplifi ed from 6 serum samples collected soon after the onset of disease by using SEOV S-segment-specifi c primers, not HTNV S segment-specifi c primers. That the HFRS cases were caused by SEOV was confi rmed. Corresponding SEOV strains were designated ShenyangHu3, ShenyangHu4, ShenyangHu5, ShenyangHu6, Shenyan-gHu7, and ShenyangHu8.

Analysis of Laboratory Rats and Mice
Serum samples from all suspected laboratory rats and mice were tested for IgG antibodies against SEOV or  HTNV, and lung tissues were analyzed for the presence of hantavirus antigen by indirect IFA. Hantavirus antibodies were detected in 32 of 139 rats; the hantavirus antigen was detected in 26 of these 32 rats (designated ShenyangW-; Figure 2). Antibodies against HTNV or SEOV, or hantavirus antigen have not been observed in laboratory mice.

Rodent Trapping and Analysis
To investigate whether SEOV strains identifi ed in the patients and laboratory rats originated in the local wild rodent population, 156 Norway rats (R. norvegicus) were trapped in the major HFRS endemic focus during the autumn of 2006 and the spring of 2007 in the vicinity of the Laboratory Animal Center of Shenyang Pharmaceutical University. Four of 156 wild rats were found to be positive for hantavirus antigen by IFA. Hantavirus S-segment sequences were recovered from these animals (corresponding hantavirus strains were designated ShenyangRn-LAC-4, ShenyangRn-LAC-28, ShenyangRn-LAC-41, and ShenyangRn-LAC-137). No rodents had been caught in the pharmaceutical laboratory building, suggesting that the laboratory animal infection occurred in the Laboratory Animal Center.
A total of 299 rodents (56 striped fi eld mice [A. agrarius] and 243 Norway rats [R. norvegicus]) were captured in 2006-2007 during the major HFRS endemic focus in the rural areas of Shenyang, which is ≈15 km from the Laboratory Animal Center. Of these rodents, 11 Norway rats and 1 striped fi eld mouse were found to be positive for hantavirus antigen by IFA. Hantavirus S-segment sequences were recovered from these animals (corresponding hantavirus strains were designated ShenyangRn20, Shenyan-gRn32, ShenyangRn74, ShenyangRn75, ShenyangRn111, ShenyangRn127, ShenyangRn131, ShenyangRn144, She-nyangRn167, ShenyangRn180, ShenyangRn183, and She-nyangAa13).

Genetic Analyses
Partial S-segment sequences were recovered from 6 patient serum samples, 19 laboratory rats (designated She-nyangW-, Figure 2); 15 wild Norway rats, and 1 striped fi eld mouse trapped in the outbreak region. Genetic analysis showed that the partial S-segment sequences recovered from all humans, laboratory rats, and wild rats were very closely related to each other, with 95.6% to 99.8% sequence identity (online Appendix  gov/EID/content/15/2/200-appT.htm). These sequences have a higher level of identity to SEOV (85.5-99.2%) than to HTNV and other hantavirus types. Further comparison showed that the partial S-segment sequences recovered from human and laboratory rats were very closely related to each other, with 98.7% to 99.8% sequence identity. The 5% nucleotide divergence among hantaviruses carried by wild rats suggested that perhaps>1 genetic lineage of SEOV co-circulated in Shenyang. Notably, the sequences of hantaviruses carried by humans and laboratory rats were more closely related to those recovered from the wild rats trapped in the vicinity of the Laboratory Animal Center (ShenyangRn-LAC-4, ShenyangRn-LAC-28, Shenyan-gRn-LAC-41, and Shenyang-LAC-137). Moreover, these sequences also shared a higher homology with those recovered from the lung tissue samples that were collected from the wild Norway rats trapped in the major HFRS endemic focus in the rural areas of Shenyang (ShenyangRn32 and ShenyangRn180). As expected, the partial S-segment sequence recovered from 1 striped fi eld mouse was closely related to those from HTNV. The sequence showed especially high identity (99.0%) to strain Bao14 isolated from A. agrarius in Heilongjiang (23), which is also in northeastern China.

Phylogenetic Analyses
In the present study, phylogenetic analysis of partial Ssegment sequences confi rmed the molecular link between SEOV strains from patients, laboratory Norway rats, and the wild Norway rats trapped in the vicinity of the Laboratory Animal Center and the disease-endemic areas ( Figure  2). As shown in Figure 2, all partial S-segment sequences from humans, laboratory rats, and wild rats fell into the SEOV genetic clade, well separated from other hantaviruses, thus indicating that the HFRS outbreak was caused by SEOV. Notably, the partial S sequences from wild rats were divided into 3 lineages. The partial S sequences recovered from humans and laboratory rats formed 2 groups, and the sequences derived from the wild rats trapped in the vicinity of the Laboratory Animal Center formed another group. Together, these 3 groups formed a lineage that also included the sequences ShenyangRn32 and ShenyangRn183, which were recovered from the wild rats trapped in the major HFRS-endemic focus in the rural areas of Shenyang. This suggests that the HFRS outbreak had been caused by strains belonging to this particular lineage of SEOV.

Discussion
HFRS has been recognized as a serious problem in Shenyang since the fi rst outbreak in 1958 (18)  investigation of a laboratory rats-associated outbreak of hantavirus disease involving 8 postgraduate students in Shenyang. The patients had clinical symptoms and biochemical fi ndings typical of HFRS cases occurring in China. Serologic tests and the analysis of recovered hantavirus genome sequences showed that the outbreak was caused by a transmission of SEOV variants from the local wild Norway rats through the laboratory Norway rats to humans. Serologic tests and phylogenetic analysis indicated that the HFRS cases were caused by the SEOV spread by laboratory rats. HFRS cases associated with laboratoryacquired infections have been reported in several countries (8)(9)(10)(11)(12)(13)16). Notably, hantavirus infections were found to be more common in laboratory Norway rats than in mice and other laboratory animals (8,9,(12)(13)(14)(15). However, only a few investigations gave clear clues as to the origin of hantaviruses circulating in laboratory animals (9,17).
Previous studies have shown the presence of 2 hantaviruses carried by rodents: HTNV carried by the striped fi eld mice and SEOV by the brown Norway rats in Shenyang (18). In the present study, serologic tests showed that all mouse serum samples were antihantavirus antibody-negative, and hantaviral antigens were not identifi ed in the mouse lung tissues and HTNV-specifi c sequences were not detected in human serum samples. These results suggest that human infections were not caused by HTNV, although our data demonstrated that HTNV is circulating in A. agrarius in Shenyang. Both human and laboratory rat serum specimens were anti-SEOV antibody positive, which suggests that the infections were caused by SEOV. Due to the cross-reactivity of sera, exact serotyping for diagnoses of individual patients was not possible. Seroepidemiologic studies may sometimes misidentify the causative hantavirus if typing is based only on ELISA, IFA, or immunoblot analysis (24). Therefore, partial hantavirus S-segment sequences were amplifi ed from the patient sera and laboratory rat lung tissues. All partial S-segment sequences recovered from 6 human and 26 laboratory rats belonged to SEOV; they were closely related to each other, and clustered together on the phylogenetic tree ( Figure 2). These results confi rmed that the HFRS outbreak in Shenyang was caused by SEOV and suggested the likely route of infection was from wild rats to laboratory rats and then to humans.
Analysis of wild rats trapped in the vicinity of human case-patients and the major HFRS epidemic focus allowed comparison of SEOV genome sequences in humans and rats (laboratory and wild). Phylogenetic analysis of the partial S-segment sequences indicated that 3 lineages of SEOV are co-circulating in wild rats in Shenyang ( Figure 2). Notably, the sequences from patients and laboratory rats were clustered within 1 of these 3 lineages. Our results suggest that the viruses carried by the laboratory rats originated from the prevalent SEOV strains circulating in wild Norway rats in this area, and then were transmitted to humans.
In conclusion, our study indicates that the HFRS outbreak was caused by SEOV circulating in local wild Norway rats through laboratory rats. Because hantavirus infection in wild Norway rats is frequent in most regions of China (5), this study reinforces conclusion that vigilance is needed to prevent laboratory-associated cases of hantavirus disease.