An epidemiological study of cryptosporidiosis at the wildlife/livestock/human interface in Mpumalanga Province, South Africa

Abstract

Cryptosporidium spp. is an oocyst-forming apicomplexan protozoan, which infects humans and a large variety of animals. Several species and genotypes are potentially zoonotic and ruminats are considered as an important source of infection. Pre-weaned calves are major hosts for zoonotic C. parvum, and show higher rates of infection than post-weaned or adult animals. Cryptosporidium infection has been demonstrated in a wide variety of wild animals, which may co tribute to environmental contamination. In sub-Saharan Africa, where the HIV infection prevalence is the highest in the world, high incidence of severe and even fatal Cryptosporidium infection have been reported in humans. This study investigated the epidemiology of Cryptosporidium spp. simultaneously in wildlife, indigenous cattle and young children living at the wildlife, livestock and human interface on the western boundary of the Kruger National Park (KNP) in Mpumalanga Province, South Africa. Initially, a pilot study was carried out to assess the zoonotic or anthroponotic importance of Cryptosporidium in diarrhoeic children in South Africa, representing the human group most likely to be infected. This geographically broad study involved hospitals from four provinces in South Africa. Stool samples from hospitalized diarrhoeic children from 0-1 year of age were analysed by microscopy (modified Ziehl-Neelsen (MZN) acid-fast staining) and molecular techniques: polymerase chain reaction (PCR), restriction fragment length polymorphism (RFLP) and sequencing. An overall prevalence of Cryptosporidium infection of 12.2% was revealed, mainly involving species of anthroponotic origin, such as C. hominis (76%) and C. parvum of anthroponotic nature (20%). Only one species of uncertain zoonotic importance (C. meleagridis) was detected in this study. These findings were in accordance with reports from sub-Saharan Africa (including South Africa), where anthroponotic species of Cryptosporidium were responsible for most infections in humans. The study then concentrated on the public health importance of cryptosporidiosis at the wildlife/livestock/human interface of the KNP. Cryptosporidium oocysts were detected in elephant, impala and buffalo samples collected in three different study areas of the KNP; two located close to the boundaries of the KNP and a third one in the centre of the KNP. The MZN staining technique and an immunofluorescent antibody (IFA) test were applied to identify oocysts from faecal samples. The prevalence detected with MZN was higher than that detected with IFA, however both tests found a higher prevalence in elephants (25.8% and 4.2%, respectively) compared to the other species. The prevalence of Cryptosporidium in buffalo was 5.5% and 1.4% with MZN and IFA, respectively, and 4.2% and 1.8% respectively in impala. In the two study areas adjacent to the fence of KNP, the combined prevalence was significantly higher compared to the area in the centre of the KNP. The agreement between the MZN staining technique and the IFA test was assessed for each wildlife species; the estimates of kappa suggested moderate agreement in buffalo and impala and fair to poor agreement in elephant. The above results of were analysed further by the use of molecular techniques in order to reveal the species and genotypes of the parasite in wildlife and in addition faecal samples collected from post-weaned calves. A questionnaire was also conducted among farmers to investigate observed contacts between cattle and wildlife species in grazing areas outside and inside the KNP. Four of the 241 wildlife samples were PCR-positive (2.8% each in impala and buffalo and of 0.0% in elephant) and sequencing revealed the presence of C. ubiquitum in two impala and one buffalo and C. bovis in one buffalo. Cryptosporidium ubiquitum has been commonly found in a large number of animals, including humans. Among calf samples, 8% (4/51) were PCR-positive and were identified as C. andersoni (2/4) and C. bovis (2/4). The probability of contact between cattle and wildlife outside the KNP, observed by farmers, was higher for buffalo (Pr=0.6) and impala (Pr=0.46) than for elephant (Pr=0.04). This suggests that the detection of C. bovis in both cattle and buffalo might be due to direct or indirect contact between these two species. The detection of C. ubiquitum in wildlife, with its zoonotic potential, suggests that Cryptosporidium may be of public health concern for people living at the interface. We further investigated the prevalence of Cryptosporidium infection in cattle and humans, this time targeting younger (pre-weaned) calves and children. Children <5 years were sampled at six rural clinics within the same interface and stool samples were screened by the MZN staining technique. All MZN-positive and suspicious samples of children and samples of 36 calves within the age of 0-4 months were analysed by nested PCR. Eight of the 143 children (5.6%) were positive on PCR, and sequencing identified predominantly C. hominis, while one sample was identified as C. meleagridis. Eleven of the 36 calf samples (30.5%) were PCR-positive and were identified as C. bovis and C. ryanae. Due to limited resources, molecular analysis could not be performed on more samples. Variables such as source of drinking water, age and contact with animals for children, were analysed as potential risk factors for humans and cattle; however, none were statistically significant. In conclusion, the prevalence of Cryptosporidium detected in human and wildlife was low compared to that reported in other studies in Africa. The species and genotypes detected in humans were predominantly of anthroponotic nature; however, the isolation of C. Ubiquitum from buffalo and impala shows that at least one species of zoonotic importance is present at the wildlife/livestock/human interface. The prevalence of HIV/AIDS in our study area is one of the highest worldwide; therefore the potential public health importance of this parasite should be investigated further.