Elucidating the molecular epidemiology of trypanosomes for evidence-based tsetse-vector control in the Shimba Hills human-wildlife-livestock interface, Kenya

Abstract

Field studies that systematically evaluate the transmission dynamics, reservoir hosts, and genotype diversity of trypanosome parasites are scanty in sylvatic ecologies in sub-Saharan Africa. Such studies are important for providing insights into infection hotspots and strategising targeted control of human sleeping sickness and cattle nagana disease caused by tsetse-borne trypanosomes. This thesis characterised the epidemiology of tsetse-borne trypanosomes in Shimba Hills —a wildlife area in southeast Kenya where nagana is endemic but understudied. Specifically, the thesis assessed the entomological inoculation rates of cattle trypanosome infection and also, unraveled wildlife sources of livestock infection and profiled the genotype diversity of the main trypanosome parasite circulating in populations of tsetse flies and reportedly affecting animal health in Shimba Hills. Using conventional-PCR and amplicon sequencing analyses targeting the Internal Transcribed Spacer-1 (ITS1) gene, 8.62% (60/696) (95% CI: 6.53 — 10.71) of wild-caught tsetse flies collected in biconical traps in Shimba Hills (November 2018 to September 2019) were confirmed positive for DNA of seven species and subspecies of trypanosomes. Trypanosoma vivax was the most prevalent trypanosome in tsetse flies. The other trypanosomes detected in tsetse flies in order of decreasing prevalence were T. congolense Kilifi, T. congolense Savannah, T. simiae Tsavo, and T. godfreyi and similar prevalence for T. brucei sl and T. simiae. Glossina pallidipes were the most widely distributed and abundant among the three species of tsetse flies collected in biconical traps and were more likely to infect cattle. Entomological inoculation rates were homogenous across landscapes but increased significantly towards the wildlife reserve in Shimba Hills. Bloodmeal analyses indicated that tsetse flies had fed on thirteen mammalian species, with DNA confirmation based on PCR-High Resolution Melting analyses of the vertebrate genes 16S and Cytochrome b and amplicon sequencing of the gene Cytochrome Oxidase 1. Warthogs were the preferred hosts of tsetse flies and were associated with increased likelihood of vector infection with cattle trypanosome parasites, including T. vivax. Phylogenetic analyses of amplicon sequences of the ITS-1 DNA and 18S rRNA genes identified endemicity of two T. vivax genotypes in Shimba Hills with >80.00% comprising of the virulent Tvv4 genotype. Tsetse flies confirmed as positive for Tvv4 included those that had blood-fed on warthogs and cattle. Using insecticide-treated fabrics (ITFs) targeted to entomologically defined trypanosome hotspots, a cluster Randomised Controlled Trial (RCT) was designed to rationalise a plan and assess ITF efficacy for the control of tsetse flies incriminated as Tvv4 vectors and having high potential for bloodmeals on cattle and transmission of infections from warthogs to cattle. With or without a blend of waterbuck-mimicking tsetse-repellent odours on cattle, ITFs significantly reduced trypanosome risk in Shimba Hills but with greater impact on T. congolense than T. vivax. Although the intervention was initially effective, an upsurge in infection cases was observed after five months of interventions. Trypanosome infection risks were not significantly different between the intervention arm of the RCT and the control arm. Stomoxys were highly abundant and T. vivax risk was apparently high in the control arm. The present work is the most extensive in Shimba Hills in terms of spatio-temporal coverage of tsetse sample collection and reports the highest diversity of trypanosomes and animal bloodmeal hosts documented in a single epidemiological survey in the area. Spatial entomological risk of cattle trypanosome infections is described for the first time in the Shimba Hills human-wildlife-livestock interface with evidence of high risk of cattle infections from warthogs in locations close to the wildlife reserve. Unequivocal evidence is also presented for the first time of the Tvv4 genotype endemicity in Kenya and plausible contribution to pathologies in cattle in Shimba Hills. As supplementing ITFs with tsetse-repellent did not significantly improve outcome of interventions in Shimba Hills, combined adoption of both technologies for tsetse control may be unnecessary for trypanosome vector management in the area. However, livestock owners in smallholder communities in Shimba Hills may deploy ITFs during tsetse peak seasons while making efforts to maintain these devices for optimal performance. Livestock owners can also apply tsetse-repellent odours on cattle year-round particularly during periods of low tsetse fly abundance when ITF adoption may be unnecessary. This is important given ease-of-adoption and low-cost of the odours and previous findings in Shimba Hills of the odours being effective at reducing tsetse-cattle contacts and infection risk. Stomoxys possible involvement in T. vivax transmission in the control arm of the RCT in Shimba Hills had plausibly compromised intervention-effects on the parasite thus should be assessed in future studies for its role in trypanosome epidemiology in the area.