East Coast Fever and vaccination at the livestock/wildlife interface
Item statusRestricted Access
Embargo end date04/07/2021
Allan, Fiona Katharine
East Coast fever (ECF) in cattle is caused by the tick-borne protozoan Theileria parva, and is transmitted by the three-host tick Rhipicephalus appendiculatus. The African buffalo (Syncerus caffer) is the natural host for T. parva but does not suffer disease, whereas ECF is often fatal in cattle, causing annual losses of more than $300 million. A live vaccine is available, the Infection and Treatment Method (ITM) Muguga Cocktail, but antigenic diversity of the parasite - particularly of buffalo-derived T. parva - results in variable protection. The project aimed to determine the prevalence and genetic and antigenic diversity of the T. parva population in cattle and buffalo, in an area adjacent to the Serengeti National Park (SENAPA), Tanzania, where livestock co-exist with buffalo, as well as ascertaining herd vector control practices in the study area, to help inform future control strategies. Field data were generated by designing and implementing a randomised cross-sectional sampling survey. Cattle were sampled (n=770) from 48 herds and blood samples analysed by diagnostic nested p104 PCR to establish a T. parva prevalence of 5.07% (CI: 3.70-7.00%). This prevalence was lower than in other hyperendemic areas. Half body tick counts were recorded on every cow and although 100% farmers reported seeing ticks on their cattle, tick counts were very low with 78% cattle having zero ticks. A questionnaire survey was created and carried out with 120 farmers, including the 48 sampled herds, to obtain data on vector control. Questionnaire data indicated significant use of acaracide with 79% (CI: 71-85%) of farmers spraying and 41% (32-49%) dipping cattle routinely. Some farmers reported very frequent spraying, as often as every four days. All acaricides used were from the same synthetic pyrethroid drug class, cypermethrin. Local workshops were held to discuss findings and validate results. These data indicate high levels of acaricide use, which may be responsible for the low observed tick burdens and low T. parva prevalence. The vector control is farmer-led and aimed at both ticks and tsetse flies. The levels of acaricide use raise concerns regarding sustainability, as large scale use of a single acaricide compound clearly represents a risk for the development of resistance. A genotyping pipeline was designed to characterise genetic diversity in T. parva field samples. Previous studies have shown that a number of T. parva antigens are recognised by CD8+ T cells in immunised cattle, with several of these antigens demonstrating polymorphism, with greater allelic diversity in buffalo-derived T. parva than in cattle-derived parasite populations. A panel of twelve antigen-encoding genes was investigated, with successful species-specific full length or near-full length amplification of four genes – Tp1, A14, Tp4 and N60. A panel of DNA of known parasite composition stocks, including diverse Theileria species and multiple characterised isolates of both T. parva and the closely related T. sp. (buffalo), was used to validate specificity and sensitivity of primers before applying them to DNA from cattle and buffalo samples from the Ol Pejeta game conservancy, Kenya. The Ol Pejeta site had a known grazing history; buffalo are endemic and no cattle had grazed there for several years and so the infections detected in cattle that were introduced to the area would all be buffalo-derived. This setting allowed for comparison of diversity of T. parva circulating in buffalo with that of parasites acquired by sentinel cattle. Amplicons of antigen genes Tp1, A14 and N60, from two cattle and two buffalo, were sequenced by PacBio long-read technology (RSII platform). Cluster analysis showed that diversity was greater in buffalo parasite populations compared to cattle parasite populations from the Ol Pejeta conservancy in Tp1 and N60, but there were shared parasite populations in cattle and buffalo for all three genes. Diversity was greatest in Tp1 which had 109 clusters (13 cattle, 55 buffalo, 41 shared), A14 showed little diversity, with 23 clusters (10 cattle, 0 buffalo, 13 shared). N60 had 37 clusters (2 cattle, 4 buffalo, 37 shared). After validation of the pipeline using samples from Ol Pejeta, DNA from the Serengeti cross-sectional cattle samples (n = 770), as well as cattle samples from several other timepoints and locations in the SENAPA study area (n = 832), and buffalo samples (n = 22) from the SENAPA study area was analysed for genetic and antigenic diversity of T. parva. Samples positive for T. parva, by p104 nPCR (149 cattle, 22 buffalo), were used to amplify antigen genes Tp1 and N60. PacBio long-read sequencing (Sequel platform) was applied to multiplexed amplicons. These data showed overall that allelic diversity was significantly greater in buffalo-derived parasites compared to cattle-derived, with some alleles shared between buffalo-derived and cattle-derived parasites. Both Tp1 and N60 showed a high degree of polymorphism at the nucleotide level. For Tp1, there were 97 variable loci over the 1618 bp length, resulting in 86 gene allele variants. For N60, there were 14 variable loci over the 983 bp length resulting in 216 gene allele variants. Most alleles were unique to buffalo (55/86 in Tp1; 191/216 in N60), with a smaller proportion unique to cattle (19/86 in Tp1; 14/216 in N60) and relatively few shared (12/86 in Tp1; 11/216 in N60). At the amino acid level, Tp1 showed a large number of variants resulting in four epitope variants observed. In contrast, all nucleotide polymorphisms in N60 were synonymous and this gene was completely conserved at the amino acid level. Sequences identical to the genome reference strain and vaccine component, T. parva Muguga (TpM), were identified in several individuals; 9 cattle and 1 buffalo for Tp1, and 5 cattle and 1 buffalo for N60. There were indications of substructuring, by both genetic distance and phylogenetics, with most alleles found in cattle being closely related to the TpM reference allele, and the most distantly related cluster mainly being found only in buffalo. This, however, requires further sampling and increased numbers to confirm. These results have implications for ITM vaccine efficacy in the study area as well as demonstrating N60 as a potential candidate gene for alternative vaccines where cattle and buffalo interact.