Detection of selected pathogens in wildlife from Kruger National Park, South Africa, and Etosha National Park, Namibia

Abstract

Wildlife is naturally exposed and infected with numerous microbials. Wildlife could act as asymptomatic carriers of pathogens critical to animal and/or public health. Literature reported that more than half of emerging infectious diseases (EIDs) in humans are zoonoses, while the majority (70%) of these are linked with wildlife. A series of drivers, mainly anthropogenic (i.e. human encroachment into natural habitats and land-use change), provide conditions that allow a select pathogen to expand and adapt to a new niche thus causing epidemics, emergence and re-emergence of diseases from wildlife. Economic, biodiversity, environmental, sanitary and social damages caused by EIDs can vary from moderate to severe. Although the rate of wildlife-emerging infectious diseases, such as Ebolavirus, Nipah and Hendra viruses, Leptospira spp. and Trypanosoma spp., has been increasing globally in the last decades, there is currently still a serious lack of epidemiological data from wild animals, particularly from species animals presumed to be “non-reservoir” or “non-maintenance” hosts. Large-scale studies are not always feasible in wildlife due to economical and ethical restraints, therefore the use of existing wildlife samples even consisting of small sample size can provide baseline information on pathogens relevant to both animals and human health. On these premises, we designed an observational cross-sectional study to detect epidemiological characteristics (occurrence, prevalence and risk factors) of selected pathogens - namely foot-and-mouth disease virus (FMDV), Brucella spp., Rickettsia spp., Coxiella burnetii and Coxiella-like endosymbionts (CLEs) from wildlife in two national parks namely Kruger National Park (KNP) in South Africa and Etosha National Park (ENP), Namibia. Retrospective samples (blood and serum) were obtained from target host species that included free roaming plains zebra (Equus quagga) (n=70; 40 and 30 from KNP and ENP), greater kudu (Tragelaphus strepsiceros) (n=72; 32 and 40 from KNP and ENP), impala (Aepyceros melampus) (n=21 from KNP) and blue wildebeest (Connachaetes taurinus) (n=30 from ENP). Several serological techniques were used in this study to determine exposure to the pathogens of interest. These included a competitive ELISA (NSPE) for the detection of antibodies against the non-structural proteins of FMDV; the Rose Bengal Test (RBT) and indirect ELISA (iELISA) for the detection of smooth Brucella spp.; and the iELISA for the detection of C. burnetii. Molecular methods consisting of conventional and real-time PCR assays were used to screen for the presence of Brucella spp. (using ITS ribosomal DNA (rDNA) specific for Brucella as well as AMOS-PCR assay that detects B. abortus biovar (bv) 1, 2 and 4, B. melitensis bv 1, 2 and 3, B. ovis and B. suis bv 1), Rickettsia spp. (Rickettsia 16S rRNA PCR confirmed with sequencing, ompB-PCR and ompA-PCR) and Coxiella spp. (real-time 16S rRNA PCR that differentiates between C. burnetii and CLEs DNA in the host species) followed by Sanger sequencing and phylogenetic analysis. The risk factors (animal family, animal species, gender, age, sampling park and sampling area) were assessed for each disease. In KNP, 13/29 (45%; CI: 26-64%) kudu tested serologically positive with NSP competitive/blocking FMD ELISA while all impala were negative. Selected NSP ELISA seropositive kudu samples tested negative for SAT serotypes using solid phase competitive/blocking structural protein (SP) ELISA. Suspect serological brucellosis results were detected by at least one serological test (RBT or iELISA) in 8/29 (28%; CI: 13-47%) kudu, 1/21 (5%; CI: 0-24%) impala, and 3/35 (9%; CI: 2-23%) zebra while brucellosis seropositive samples with both tests include only 3/29 (10%; CI: 2-27%) kudu. Brucella specific DNA were detected in 9/29 (31%; CI: 15-51%) kudu of which none was seropositive with RBT and/or iELISA. Multiplex AMOS-PCR assay followed by B. melitensis specific primers (simplex PCR) of ITS-PCR positive samples did not amplify specific Brucella species. Antibodies against C. burnetii were detected in 6/29 (21%; CI: 8-40%) kudu, 14/21 (67%; CI: 43-85%) impala and 18/39 (46%; CI: 30-63%) zebra, although C. burnetii DNA was absent in all these animals using the real-time PCR C. burnettii probe assay. Moreover, CLEs DNA was detected in 5/28 zebra (18%; CI: 6-37%) and 13/33 kudu (39%; CI: 23-58%), but not in impala. Sequences of three CLEs zebra amplicons confirmed presence of CLE products (~99% homology), and clustered with Coxiella clades C and D using phylogenetic analysis. None of the samples tested positive for Rickettsia spp. In ENP, only 1/29 (3%; CI: 0-18%) wildebeest samples tested serologically positive for FMD. For the serological detection of Brucellosis 2/29 (7%; CI: 1-23%) wildebeest sera reacted faintly to RBT, although molecular test did not detect Brucella DNA in these samples. The iELISA detected C. burnetii in 26/30 (87%; CI: 69-96%) wildebeest, 16/40 (40%; CI: 25-57%) kudu and 26/26 (100%; CI: 87-100%) plains zebra, while the Coxiella real-time PCR assay detected C. burnetii DNA in only 1/28 (4%; CI: 0-18%) wildebeest sample, which also tested serologically negative. Molecular screening using the Coxiella real-time PCR assay detected CLE DNA in 13/40 (33%; CI: 19-49%) kudu and 2/40 (8%; CI: 1-25%) zebra, while all wildebeest tested negative. Sequences obtained from a subset of CLEs positive samples (i.e. two kudu and one zebra) resulted ~99% identical to CLEs of Amblyomma hebraeum (clade D). All wildlife blood from ENP tested negative for Rickettsia spp. Due to the small sample size, our statistical analysis generally lack precision (for the prevalence estimates) and power (for the hypothesis tests and models). Statistical analyses were then conservatively interpreted by viewing confidence intervals and the p value. Risk factor analysis revealed that Q fever seroprevalences were significantly higher in ENP (p-value < 0.001), while FMD and Brucella spp. were significantly higher in KNP (p-value < 0.001). The occurrence of FMD and brucellosis is not surprising in KNP as African buffalo (Syncerus caffer), a reservoir host for FMD and known to be infected with brucellosis, occurs in KNP but is absent in ENP. We also observed widespread exposure to C. burnettii in all animal species targeted and we confirmed presence of CLEs DNA in kudu and zebra from ENP and KNP. Further investigations are needed to clarify whether the C. burnetii iELISA is cross-reacting with CLEs or if these two events are unrelated to each other and must be interpreted separately. Finally, samples from both ENP and KNP tested negative for Rickettsia species. However, blood samples from mammal hosts are not the ideal sample to detect Rickettsia spp. as they primarily target and replicate within endothelial cells.