Metagenomic Investigation of Ticks From Kenyan Wildlife Reveals Diverse Microbial Pathogens and New Country Pathogen Records

Abstract

Focusing on the utility of ticks as xenosurveillance sentinels to expose circulating pathogens in Kenyan drylands, host-feeding ticks collected from wild ungulates [buffaloes, elephants, giraffes, hartebeest, impala, rhinoceros (black and white), zebras (Grévy's and plains)], carnivores (leopards, lions, spotted hyenas, wild dogs), as well as regular domestic and Boran cattle were screened for pathogens using metagenomics. A total of 75 host-feeding ticks [Rhipicephalus (97.3%) and Amblyomma (2.7%)] collected from 15 vertebrate taxa were sequenced in 46 pools. Fifty-six pathogenic bacterial species were detected in 35 pools analyzed for pathogens and relative abundances of major phyla. The most frequently observed species was Escherichia coli (62.8%), followed by Proteus mirabilis (48.5%) and Coxiella burnetii (45.7%). Francisella tularemia and Jingmen tick virus (JMTV) were detected in 14.2 and 13% of the pools, respectively, in ticks collected from wild animals and cattle. This is one of the first reports of JMTV in Kenya, and phylogenetic reconstruction revealed significant divergence from previously known isolates and related viruses. Eight fungal species with human pathogenicity were detected in 5 pools (10.8%). The vector-borne filarial pathogens (Brugia malayi, Dirofilaria immitis, Loa loa), protozoa (Plasmodium spp., Trypanosoma cruzi), and environmental and water-/food-borne pathogens (Entamoeba histolytica, Encephalitozoon intestinalis, Naegleria fowleri, Schistosoma spp., Toxoplasma gondii, and Trichinella spiralis) were detected. Documented viruses included human mastadenovirus C, Epstein-Barr virus and bovine herpesvirus 5, Trinbago virus, and Guarapuava tymovirus-like virus 1. Our findings confirmed that host-feeding ticks are an efficient sentinel for xenosurveillance and demonstrate clear potential for wildlife-livestock-human pathogen transfer in the Kenyan landscape.

Keywords: metagenomics; pathogen; surveillance; tick; wildlife.

FIGURE 1

Map indicating the counties where ticks were collected and sampled when feeding on domestic and wild animals in Kenya.

FIGURE 2

Heatmap of the detected bacterial pathogens according to the hosts. Pathogen detection rates were calculated as positive tick pools divided by the number of hosts examined.

FIGURE 3

The maximum likelihood analysis of the Jingmen tick virus (JMTV) partial segment 1 sequences (366 nucleotides). The tree is constructed using a gamma-distributed Kimura 2-parameter model for 500 replications. JMTVs included in the analysis are indicated by GenBank accession number, name, isolate/strain identifier, host, and detection region. JMTV sequence characterized in this study is marked. Bootstrap values greater than 75 are displayed.

FIGURE 4

The maximum likelihood analysis of the Trinbago virus partial genome sequences (1,141 nucleotides). The tree is constructed using a gamma-distributed Kimura 2-parameter model for 500 replications. Viruses included in the analysis are indicated by GenBank accession number, name and isolate/strain identifier. Trinbago virus isolate Kenya-P83 characterized in this study is marked (GenBank accession: OM807120). Beihai barnacle virus 1 strain BHGZ is included as the outgroup.

FIGURE 5

The maximum likelihood analysis of the Guarapuava tymovirus-like virus ORF sequences (5,898 nucleotides). The tree is constructed using the general time reversible (GTR) model, gamma distributed with invariant sites (G + I) model for 500 replications. Guarapuava tymovirus-like 1 virus isolate Kenya-P43 characterized in this study is marked (GenBank accession: OM807119). Bootstrap values greater than 50 are displayed.