Several rabies-related lyssaviruses have been associated with bat species in southern Africa, the rarest of these being Duvenhage virus (DUVV), for which only five isolations have been made over five decades. Three of these were from human fatalities, and it is not known which bat species acts as reservoir. In studying a population of Nycteris thebaica in the kingdom of Swaziland, a landlocked country bordering Mozambique and South Africa, we found evidence of the circulation of a lyssavirus. Virus-neutralization assays indicated DUVV-neutralizing antibodies in 30% of the sera collected from this population of N. thebaica, providing the first indication of a Duvenhage virus circulating in this particular species and the first evidence of a bat lyssavirus circulating in Swaziland bats.
The genus Lyssavirus is a group of single-stranded, negative-sense RNA viruses that cause rabies. Rabies virus (RABV) is the prototype lyssavirus, and all subsequently discovered lyssaviral species are collectively known as the rabies-related lyssaviruses. Evidence seems to support that various bat species are the reservoir for 11 of the currently recognized 12 lyssavirus species, with the exception of Mokola virus (Dietzgen et al., 2011). Of these viruses only RABV, Lagos bat virus (LBV), Mokola virus (MOKV), Shimoni bat virus (SHIBV), and Duvenhage virus (DUVV) have been isolated on the African continent, with LBV, MOKV, SHIBV, and DUVV being exclusive to Africa (Nel and Markotter, 2007; Kuzmin et al., 2010). In addition, antibodies that could neutralize West Caucasian bat lyssavirus (WCBV; Kuzmin et al., 2008) and a proposed new lyssavirus species, Ikoma virus (Marston et al., 2012), were identified in a striped leaf-nosed bat in Kenya (Hipposideros vittatus previously reported as Hipposideros comersonni) and an African civet (Civettictis civetta) from Tanzania, respectively. Lagos bat virus has been associated with Old World fruit bats (Family: Pteropodidae), and DUVV, SHIBV, and WCBV with insectivorous bats (Families: Nycteridae, Hipposideridae, and Miniopteridae, respectively). There have been only five reported cases of DUVV infection, and three of these have been fatal human cases (1970, 2006, and 2007), each case followed contact with a small unidentified bat (Van Eeden et al., 2011). In none of the human cases has the bat been found, thus not allowing for identification. The remaining two cases were both from small bats, one from South Africa in 1981 (no bat identification was made) and one from Nycteris thebaica in Zimbabwe in 1986 (Foggin, 1988). Thus N. thebaica is the only bat species linked with DUVV infection.
The kingdom of Swaziland is a small southern African country, landlocked between Mozambique and South Africa. Rabies was first identified here in 1954 and is primarily associated with dogs, with occasional spillover—the first case of bovine rabies was reported in 1975 (Dlamini and Mathunjwa, 1995). Today dog rabies is endemic in Swaziland and all cases are believed to be caused by RABV. Although rabies-related viruses have not been encountered in Swaziland, neighboring South Africa sporadically reports LBV and MOKV and is also the country from where most DUVV cases originated. We studied a population of N. thebaica in northeastern Swaziland for the presence of DUVV in saliva/oral swabs and for the presence of circulating DUVV-neutralizing antibodies in collected sera. This population was previously identified as N. thebaica (Monadjem, 2001, 2005), but DNA barcoding was performed to confirm this identification.
Bats were captured on 15 October 2008 in three culverts passing under the railway track within 10 km of the Mlawula Nature reserve area (26°11′S, 31°59′E, 160-m elevation) directly from the roost (Monadjem, 2001; Monadjem, 2005). After the bats' sex was determined, oral swabs were collected in Dulbecco's Modified Eagle's Medium/Ham's F- 12 (DMEM-F12, Lonza), transported in liquid nitrogen and stored at −80 C. Blood was collected into a serum separator tube using a heparinized capillary tube from a small incision, made with a 25-mm needle in the interfemoral membrane vein. Approximately 150 µL whole blood was collected from each bat and serum was separated by centrifugation and stored at −80 C until use. For DNA barcoding, tissue (wing biopsy) was sampled close to the body from all bats with the use of a 3-mm-diameter biopsy punch. After sampling the bats were released. Swabs were tested for lyssavirus RNA with the use of a reverse transcription–real time polymerase chain reaction assay (Coertse et al., 2010) after total RNA extraction with the use of Trizol (Invitrogen, Carlsbad, California, USA) as described by the manufacturer. Virus-neutralizing antibodies (VNA) in serum samples were determined by a modification of rapid fluorescent focus inhibition test (RFFIT; Kuzmin et al., 2008), with DUVV2006 used as a challenge virus. All samples were tested in duplicate and serially diluted, representing reciprocal titers of 10, 25, 125, and 625. Samples that had a 50% end-point neutralizing titer ≥10 (e.g., <5 of the 10 counted fields contained infected cells at serum dilution 1∶10) were considered positive. For molecular identification of the bat species, the cytochrome c oxidase I region was amplified and analyzed as described by Folmer et al. (1994). Sequences were compared with those available from the Barcode of Life Data system (Ratnasingham and Hebert, 2007). Only limited sequencing information from samples collected in Kenya was available from this database.
We sampled 50 bats from three culverts. No lyssavirus RNA was detected in oral swabs, but virus-neutralizing activity against DUVV was detected in 30% (95% confidence interval [CI], 19–44%) of samples (Fig. 1). No cross neutralization with RABV (challenge virus standard [CVS-11]) was detected indicating the specificity of the reaction. Antibody prevalence varied among roosts, from 15% to 41%. DNA barcoding confirmed the identity of all bats as N. thebaica.
We provide the first evidence of a bat lyssavirus circulating in Swaziland. Although no viral nucleic acid was identified in saliva, the presence of specific antibody suggested exposure to DUVV or antigenically similar virus. However, Duvenhage virus is the only phylogroup I lyssavirus documented to circulate in African bats naturally, and no other closely related lyssavirus has been isolated from African bats. Elsewhere in Africa, for different bat and lyssavirus species, high prevalence of antibody amid very low levels of virus detection was also reported (Kuzmin et al., 2008, Hayman et al., 2012). The genus Nycteris consists of 12 species, of which 9 occur in Africa (Monadjem et al., 2010). These bats have distinctively long ears and a very recognizable nose leaf, which gives the appearance of a slit running the length of the muzzle. Nycteris thebaica is widely distributed throughout southern Africa, where they roost during the day in a variety of shelters including caves, aardvark burrows, road culverts, and tree trunks (Monadjem et al., 2010). The species also uses night roosts, where prey is consumed and conspecifics socialize (Monadjem et al., 2009). These roosts may contain up to several hundred bats. Because lyssaviruses are present in bat populations, persons interacting with these animals should follow appropriate precautions, including vaccination. Available vaccines offer cross protection against some (including Duvenhage virus), but not all African lyssaviruses. Current data clearly underestimate incidence of lyssavirus infection because of sporadic and infrequent surveillance activities. It is important that more active surveillance programs be initiated to increase our understanding of the ecology of DUVV and other obscure lyssaviruses and their associated hosts. Here we have reported additional data to support a hypothesis that N. thebaica should be considered a likely reservoir for DUVV.
We thank Petrus Jansen van Vuren (National Institute for Communicable Diseases—National Health Laboratory Services, South Africa) and postgraduate students of the Virology Research group, University of Pretoria, South Africa for assistance in sample collection. The National Research Foundation, South Africa, provided funding. We thank the Swaziland National Trust Commission for providing a research permit for this work, and assisting with logistics at Mlawula Game Reserve.