Hantaan virus antibody prevalence in rodent populations of several provinces of northeastern Thailand
Jean-Paul Gonzalez Center for Vaccine Development-Research, IRD Mahidol University, Nakhon Pathom Province 73170, Thailand. Fax: +66 2 441 0189; E-mail: firstname.lastname@example.org
We conducted a serological survey of 632 rodents from the northeast region of Thailand in order to assess the presence of Hantaan-like viruses that may be a risk to the human population. Rodents were collected from rice fields, houses and domestic gardens in five northeastern provinces and tested for IgG reacting sera to Hantaan antigen using enzyme-linked immunoassays. The overall prevalence of Hantavirus infection in rodents was 2.1% (13/632). Species that tested positive included Bandicota indica (4.3% positive within species), Rattus exulans (2.1%), R. losea (1.6%) and R. rattus (0.9%). Species such as R. exulans and R. losea are candidate hosts of unidentified Hantaan-like viruses in Thailand.
Hantaviruses are members of the Bunyaviridae family and cause zoonotic diseases transmitted to humans through contact with infected rodents. They are single-stranded, negative-sense RNA viruses that reside in Muroid reservoirs and infect humans through infectious aerosols generated from rodent excretions with a transient viremia (Nuzum et al. 1988). Hantaviruses cause haemorrhagic fever with renal syndrome in humans, marked by proteinuria, nephropathy with transient renal dysfunction to acute renal failure, shock and occasional haemorrhage (Chun et al. 1989; Hjelle et al. 1996). More recently described Hantavirus groups also present with respiratory distress syndrome (Elliot et al. 1994; Hjelle et al. 1994). Although viraemia appears as transient, rodents are excellent reservoirs for hantaviruses which disseminate to tissues, including lung, salivary glands and kidney, resulting in long-term, asymptomatic infection despite high-titre antibodies (Lee et al. 1981; Lee 1982; Yanagihara et al. 1985). Vertical transmission probably does not occur in the rodent population, however, exposure to nesting material, grooming, and intraspecies biting lead to horizontal maintenance within the enzootic cycle (Glass et al. 1988; Yanagihara & Gadjusek 1988). In addition, hantaviruses appear closely associated with specific rodent vectors which act as primary host reservoirs for individual viral strains (Puthavathana et al. 1993; Gonzalez 1996).
Hantavirus antibodies have been identified in human and rodent reservoirs worldwide, including South and North America, Europe and Asia (Ahlm et al. 1997; Bharadwaj et al. 1997; Williams et al. 1997; Mills et al. 1998; Calderon et al. 1999). In Thailand Hantavirus antibodies were identified in 3.1% of rodents collected in two provinces from the central plain (Nitatpattana et al. 2000). High antibody prevalence was found in species including Bandicota indica and Rattus norvegicus. Antibodies have also been identified by immunofluorescent antibody technique in 33% of the human population living in slum areas of Bangkok and in close contact with infected rodents, although it has not been established whether they are acting as reservoirs to human disease, or whether the viruses are pathogenic for the human population (Elwell et al. 1985). A more recent investigation suggests that clinically presenting patients with fevers of unknown origin in Thailand are infected with Hantavirus (Suputthamongkol et al. 2001, congress presentation of the Emerging Viral Diseases in Thailand). According to a 1998 study of Hantavirus antibody prevalence by the Pasteur Institute of Cambodia, 2.4% of paediatric patients from Phnom Penh had enzyme-linked immunosorbant assay (ELISA) IgG reacting sera and 5.8% of the Murinae reservoir population, including domestic and field-dwelling rodents, tested positive for Hantavirus antigen (Reynes et al. personal communication).
Given the severe clinical presentations accompanying acute viral illness and widespread geographical distribution, Hantavirus infection is a major global health concern. With the recent suspected risk of pathogenic infection of Hantavirus in Thailand, more information is needed about the risk factors associated with rodent species. Hantavirus strains type Seoul and Thai type have also been isolated from rodents (Xiao et al. 1994) in Thailand, and unidentified Hantavirus-like viruses were detected by PCR (Dr Katrin Leitmeyer and Dr Ho Wang Lee, personal communication). We investigated the antibody seroprevalence of rodents in five provinces of northeast Thailand in order to asses the distribution and geography of potential Hantavirus reservoirs.
Materials and methods
This study involved collection of 1311 rodents from five northeastern provinces (Isaan): Buriram, Surin, Khon Kaen, Kalasin and Nakhon Phanom, during the year 2001, as part of a Thai Department of Communicable Disease study on leptospirosis prevalence (Phulsuksombati et al. 2001) (Fig. 1). Northeastern Thailand is characterized by a tropical climate with seasonal fluctuations in temperature, relative humidity and rainfall. Overall, there are three major seasonal periods including a humid, rainy season (June to November), a cool season (December to February), and a hot season (March to May). The northeast region consists primarily of agricultural land, municipal districts, and animal farms with forested and mountainous areas surrounding the region. The vegetation cover in northeast Thailand consists primarily of rice fields (50%), other field crops (30%) and forested areas (20%).
Rodents were collected from numerous villages located in the Muang district of each province, which have rice fields and gardens in surrounding areas. `Garden' locally refers to a small field of sugar cane, potatoes and/or cassava. Rodents were collected from approximately 80 villages in 16 subdistricts, using locally made live traps (Phulsuksombati et al. 2001). Animals were bled after anaesthesia with CO2, identified and described by recording size, weight, sex and place of entrapment (the collection was funded by the Ministry of Public Health, Nonthaburi, Thailand). If necessary, rats and mice were humanely euthanized according to CDC's (Centers for Disease Control and Prevention) methods for trapping and sampling small mammals for virology testing (Mills et al. 1995).
Enzyme-linked immunosorbant assays for Hantavirus IgG were performed on 632 of the collected rodents that provided adequate specimens for serologic testing. ELISA plates (MaxiSorp; Nunc, Roksile, Denmark) were coated with 1:1000 diluted Hantavirus antigen (Hantaan ROK 84/105 SMRV provided by Prof. Ho Wang Lee, Institute for Viral Diseases, Korea University College of Medicine, Seoul, South Korea) in sterile phosphate-buffered saline (PBS) (pH 7.2) and negative non-infected cell culture 1:100, in sterile PBS, pH 7.2. Cells were dispensed by 100 μl per well, in alternate wells with a positive and a negative antigen and kept overnight at 4 °C. The plates were then washed with 0.5% PBS/Tween and 100 μl of rodent serum (1:100 dilution in PBS/Tween/3% skim milk) was added to the plate wells and incubated for 1 h at 37 °C after washing three times with 0.5% PBS/Tween, 100 μl of horseradish peroxidase (HRP)/antirodent IgG (HRP conjugate goat antirodent IgG, ZYMED) was added and incubated for 1 h at 37 °C. The plates were washed three times with 0.5% PBS/Tween, and 100 μl O-Phenylenediamine substrate was added and incubated for 30 min at room temperature in the dark. The reaction was stopped with 50 μl 4 M H2SO4, and the plates were read in a Metertech Σ490 spectrophotometer at a wavelength of 490 nm. The cutoff value for presence of antibody was defined as three times the negative control absorbance. Fifteen samples of high and low positive sera were checked against the Seoul and Puumala antigens and did not react (Centre National de Reference, Institute de Pasteur, Paris, France; Dr Herve Zeller).
Data were analysed using appropriate non-parametric statistical methods. Fisher exact probability tests were used to examine significant differences between ratios of rodents with positive serology for Hantavirus IgG and the population when expected counts in contingency tables were less than five. Data with higher expected counts were analysed using unpaired chi-square goodness-of-fit tests with a Yates correction for continuity. All statistical tests were performed in relation to a 0.05 confidence interval.
Of the 1311 rodents trapped, 632 were tested for prevalence of Hantavirus IgG reacting antibodies to Hantaan virus antigen. Species were identified as B. indica, B. savilei, Mus caroli, Mus cervicolor, R. argentiventer, R. exulans, R. Losea, R. norvegicus, and R. rattus. Of the 632 specimens tested, 13 had Hantavirus reacting sera (2.1%). The highest seroprevalence was found in the predominantly rice field-dwelling B. indica (4.3%; 9/208) with the remainder of positive rodents belonging to indoor-dwelling rodents including R. exulans (2.1%; 2/95), R. losea (1.6%; 1/64) and R. rattus (0.9%; 1/114). Khon Kaen Province had the highest percentage of seropositive rodents (5.7%), followed by Surin (2.0%), Buriram (1.6%) and Nakhon Phanom (1.5%). In Buriram and Khon Kaen, all positive rodents were B. indica collected from rice fields, whereas seropositive species from Surin included one of each B. indica and R. losea from rice fields and one R. exulans from a village house (Table 1). Positive rodents from Nakhon Phanom were collected from a house (R. exulans) and a domestic garden (R. rattus). None of the rodents from Karasin tested positive for Hantavirus antibody.
Hantavirus (Hantaan antigen) IgG reacting sera from rodent species by collection locale in five northeast provinces of Thailand
Overall, a higher percentage of B. indica (4.3%) were positive than other species with reacting sera (1.5%) in all five provinces. The seroprevalence for this species was significantly higher in Khon Kaen than in all remaining provinces together (Buriram, Surin, Nakhon Phanom and Kalasin) (P=0.006). This spatial difference in distribution of infection remained valid when all combined positive rodent species in Khon Kaen were compared with all combined species of the other provinces (P=0.004).
A similar percentage of rodents with Hantavirus antibody reacting sera were distributed among sex (2.0% female, 2.1% male) and both age classes (1.9% adult, 2.6% juvenile). Statistical analysis of both groups of age class and sex revealed no significant differences (P=0.714 and χ2 < 0.001, P=1.00, respectively) as predicted by the similar percentages. However, the two R. exulans (5.7%, 2/95) showing reactivity to Hantavirus antibodies were males (Table 2).
Hantavirus (Hantaan antigen) IgG reacting sera from female and male rodent species in five northeast provinces, Thailand
Overall, the highest prevalence of Hantavirus reacting antibody found in B. indica (4.3%) corresponds to a previous study of rodent IgG seropositivity in two central Thai provinces (Nakhon Pathom and Nakhon Ratchasima), by Nitatpattana et al. (2000). The authors report that 4.5% of ELISA tested B. indica had sera reacting to Hantaan antigen. In an older study of Hantavirus in central Thailand, 28.6% of bandicoot rats trapped in rural areas of Kanchanaburi province and 20.6% in slum areas of Bangkok tested positive by immunofluorescent assay (IFA) (Elwell et al. 1985; Tantivanich et al. 1988, 1992). Similar research carried out in Chiang Rai province in 1996 (Dr Katrin Leitmeyer, personal communication) resulted in 20.6% prevalence (IFA) of seropositive B. indica.
These studies suggest that B. indica is the predominant reservoir for Hantavirus in Thailand. Other potential reservoirs include R. rattus, R. exulans and R. losea. There is a crossreaction of Hantaan antigen with the Seoul serotype previously identified in Thai and from B. indica and R. norvegicus species (Schmaljohn et al. 1985; Puthavathana et al. 1993; Chu et al. 1994).
Analysis of antibody prevalence in adult and juvenile male and female rodents indicates that there is no statistically significant difference in age class or sex of all species (P > 0.714). Therefore, both juvenile and adult rodents should be considered potential reservoirs and vectors of Hantavirus.
Of the three collection localities, the highest percentage of Hantavirus seropositive rodents were from rice fields, which is consistent with the significantly higher proportion of positive B. indica, a predominantly paddy-field dwelling species. B. indica is the only species with Hantavirus reacting antibody sampled from Buriram and Khon Kaen (Table 2). One B. indica tested positive in Surin along with R. losae and R. exulans. The spatial dynamics of seropositive rodent species are quite different in Nakhon Phanom, however, where only domestic dwelling R. exulans and R. rattus had Hantavirus IgG reacting sera. Similar species have been identified as having Hantavirus reacting sera in central Thailand by Nitatpattana et al. (2000), including R. exulans (3.3% positive) and R. rattus (1.0% positive). Two positive R. norvegicus were collected in the central province of Nakhon Pathom, whereas no rodents of this species tested positive in northeast Thailand.
The fact that similar rodent species have approximately the same percentage of Hantavirus IgG reacting sera to Hantaan antigen in recent studies from northeast and central Thailand suggests that these regions exhibit similar prevalence and spatial distributions of infected rodents. Serologic testing from this analysis and the recent study in the central plain were carried out using ELISA which is a sensitive method for determining sera antibodies. Percentages of positive rodents from previous serologic studies using IFA by Elwell et al. (1985) and Dr Katrin Leitmeyer (personal communication) were proportionally higher than in recent investigations. Hantavirus testing of acute respiratory distress patients in North America also suggests that IFA testing results in a higher fraction of positive results than testing carried out by ELISA (Ksiazek et al. 1995). However, sample sizes (632 from the northeast and 680 in the central region) from the central and northeast provincial studies are sufficient to provide an accurate overall percentage of Hantavirus infected rodents in Thailand.
Hantaviruses have co-evolved with their rodent hosts over the past 30 million years, and spill-over infections from different species seem rare (Xiao et al. 1994; Gonzalez 1996; Nichol 1999). Rodents such as R. rattus and R. norvegicus and B. indica are known as primary hosts of, respectively, Seoul and, Thailand Hantavirus strains (Elwell et al. 1985). On the other hand, species such as R. exulans and R. losea, sampled in central and northeast Thailand, have also Hantaan antigen reacting sera, and/or given strong crossreaction between Hantaan and Seoul serotypes. These species are candidates for either Hantaan, Seoul or other unidentified Hantaan-like viruses. Domestic dwelling species, such as R. exulans, are of particular concern because of their close contact with the human population (Tantivanich et al. 1992). These findings indicate that Hantavirus-infected rodents need to be located and isolated, and that clinicians need to be aware of potential Hantavirus infection in humans.
We would like to thank Dr Darika Kingnate of the Department of Communicable Diseases Control, Ministry of Public Health, Nonthaburi, Thailand, and Miss Yuvaluk Khoprasert of the Agriculture Zoology Research Group, Division of Entomology and Zoology, Department of Agriculture. Special appreciation goes to Prof. Ho Wang Lee, Dr Hervé Zeller and Dr Bernadette Murgue of the Pasteur Institute, France, Mr Noppadol Sanngjun of the Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand, and the staff of the CVD-RCEVD, Mahidol University, Salaya, Thailand.
The study was funded by the Department of Technical and Economic Cooperation, DTEC Thailand; Center for Vaccine Development, CVD Thailand; CDC Thailand; Institut de Recherche pour le Développement (IRD) France.