Metagenomic exploration of endosymbionts and pathogens in the tropical lineage of Rhipicephalus sanguineus sensu lato (s.l.) ticks in Colombia

Shotgun metagenomic sequencing of 38 adult tropical-lineage Rhipicephalus sanguineus sensu lato ticks collected from dogs in two Colombian departments identified Francisella tularensis, Anaplasma phagocytophilum and Theileria equi as the dominant pathogens, while the endosymbiont Candidatus Coxiella mudrowiae dominated the bacterial community. The study also recovered 27 medium-quality metagenome-assembled genomes of C. mudrowiae, providing genomic detail about a prevalent microbial partner of the brown dog tick. The findings document microbial diversity and associations in a limited sample, but they do not establish that every detected organism can replicate in the tick or be transmitted to a new host.

The brown dog tick, commonly called R. sanguineus s.l., has a global distribution and is primarily associated with domestic dogs, although it has also been found on wildlife and horses and, rarely, on people. The species complex can carry bacteria, protozoa, fungi, nematodes and viruses associated with disease in animals and humans. Mitochondrial markers have distinguished five lineages within the complex, including a tropical lineage distributed mainly in tropical regions and a temperate lineage found more often in cooler zones, although both can circulate in some countries. Before this investigation, only the tropical lineage had been documented in Colombia. Its close association with dogs creates opportunities for encounters with people and other domestic animals, making characterization of the tick microbiome relevant to veterinary and public-health surveillance.

For the analysis, researchers collected 20 ticks in Yopal, Casanare, including nine females and 11 males, and 18 in Puente Nacional, Santander, including eight females and 10 males. Ticks were removed manually from clearly tick-caused lesions and preserved in a DNA/RNA shielding solution. Researchers used taxonomic keys to identify them morphologically as members of the tropical R. sanguineus s.l. lineage. One longitudinal half of every tick was washed in phosphate-buffered saline, and DNA was extracted with a commercial tissue kit, with the disruption step extended to 12 hours at 56 °C. DNA concentration was measured with a fluorometer and quality was evaluated by agarose-gel electrophoresis. Each sample was sequenced on an Illumina NovaSeq 6000 with paired-end 150-base reads, producing approximately 6 gigabytes of raw data per tick.

After quality control and adapter removal, the investigators filtered reads with minimum length of 150 bases, average quality of 20 and trailing-quality threshold of 20. Host-associated reads were removed by alignment to the available R. sanguineus s.l. genome, and remaining reads were assigned taxonomically with Centrifuge and Kraken reporting tools. Pathogen and endosymbiont screening used a custom database containing 2,153 relevant genomes, from which 524 nonrelevant entries were removed. The retained database included genomes assigned to Anaplasma, Rickettsia, Francisella, Ehrlichia, Borrelia, Coxiella, Babesia, Theileria and selected endosymbionts. Wilcoxon tests compared read abundance by sex and department, while Spearman rank correlations with Benjamini-Hochberg correction assessed relationships among the ten most abundant pathogens and endosymbionts. Separate analyses searched for virulence-associated sequences and antibiotic-resistance markers.

Host sequences represented a mean of 88.5% of reads, with a standard deviation of 6%. Among the remaining reads, an average of 19% was classified, with a standard deviation of 5%, leaving about 81% unclassified. Bacteria accounted for most classified sequences, ranging from 3% to 24%, while archaea represented 0.07% to 0.09% and viruses 0.02% to 0.04%. Coxiellaceae was the predominant bacterial family in every sample, and a Coxiella endosymbiont contributed between 48% and 95% of classified bacterial reads. Other detected genera included Bacillus, Clostridium, Pseudomonas, Rickettsia, Streptomyces, Propionibacterium, Burkholderia, Fusobacterium, Lactobacillus and Campylobacter. The broad microbial composition was similar across departments and sexes, although individual genera differed significantly. Clostridium, Fusobacterium and Rickettsia were more abundant in Santander, while Corynebacterium and Propionibacterium were more abundant in Casanare.

The three leading pathogen signals were F. tularensis, A. phagocytophilum and T. equi, with mean relative abundances of 23%, 7% and 7%, respectively. The standard deviations were 8, 1 and 3 percentage points. The three were generally more abundant in Santander than in Casanare. Within Casanare, male ticks carried significantly higher relative abundances of F. tularensis and T. equi than females, while the same pattern in Santander was not statistically significant. A. phagocytophilum showed the reverse pattern in Santander, where females had significantly higher abundance. Other pathogens in the genera Babesia, Ehrlichia, Borrelia and Coxiella occurred at lower levels, ranging from 2% to 0.02%, and additional species had relative abundances below 0.02% and very few reads. F. tularensis causes tularemia, A. phagocytophilum causes human granulocytic anaplasmosis and anaplasmosis in animals, and T. equi is associated with equine piroplasmosis. Detection of their DNA does not, however, demonstrate vector competence, and metagenomic screening may have difficulty distinguishing F. tularensis from Francisella endosymbionts.

C. mudrowiae accounted for a mean of 47% of reads assigned against the endosymbiont database, with a standard deviation of 18 percentage points. Other endosymbionts included Wolbachia, Coxiella-like and Rickettsia endosymbionts, but their relative abundances ranged from approximately 0.02% to 3%. C. mudrowiae was more abundant in Casanare than in Santander, whereas most other endosymbionts were more abundant in Santander. It was also negatively associated with 11 organisms: three other endosymbionts and eight pathogens. Most other pairs showed strong positive correlations within or between the two ecological groups. These correlations describe patterns in read abundance and do not establish the mechanisms responsible. The inverse relationship is consistent with several possible interactions, including competition for nutrients or genomic resources, but the study did not test whether C. mudrowiae restricts pathogens or whether both respond to shared environmental and host conditions.

The viral component was small but included several families. Reads were assigned to Baculoviridae, Myoviridae and Siphoviridae, with Baculoviridae dominating the viral signal and most of its reads assigned to the genera Alphabaculovirus and Betabaculovirus. These genera contain nucleopolyhedroviruses associated with lepidopteran insects. Other viral families and genera were represented by very few reads, making precise taxonomic assignment difficult. The results therefore show that viral sequences were recoverable from the tick samples, but they do not demonstrate active viral replication, persistence or transmission in R. sanguineus s.l. Archaea likewise accounted for only 0.07% to 0.09% of classified sequences. The detected families included Methanobacteriaceae, Methanocaldococcaceae, Methanosarcinaceae, Thermoplasmataceae and Sulfolobaceae. The authors described this as the first metagenomic identification of archaea in this tick species, while emphasizing that their biological role in ticks remains unclear.

Sample-by-sample assembly and genome binning produced 27 medium-quality C. mudrowiae MAGs, with completeness ranging from 87.21% to 89.44% and contamination from 0.58% to 1.2%. Taxonomic assignments were checked with pubMLST. Four assemblies were unusually long and contained many genes absent from the other MAGs and the two available C. mudrowiae genomes; one contained 2,907 unique genes, largely annotated as hypothetical proteins, and no plasmids were identified. Because these outliers distorted the shared and variable genome estimates, the researchers excluded them from a second pangenome analysis. That analysis identified a core genome of 1,089 genes, or 41% of the pangenome, and an accessory genome of 1,574 genes, or 59%. The recovered MAGs clustered separately from the two previously available genomes in core-genome and single-nucleotide-polymorphism analyses. However, average nucleotide identity exceeded 95%, and a phylogeny based on 40 conserved proteins grouped the new assemblies with C. mudrowiae rather than with other Coxiella endosymbionts or C. burnetii.

Functional annotation showed that the MAGs contained complete pathways for synthesizing biotin, or vitamin B7, and riboflavin, or vitamin B2, as well as the cofactors coenzyme A and flavin adenine dinucleotide. These capabilities are consistent with a proposed role in supplying nutrients that may be scarce in a tick’s blood meals. The genes panE and panD, associated with folic-acid and vitamin B6 biosynthesis, and phoA were not identified, although related pathways have been reported in other Coxiella endosymbionts and C. burnetii. Comparisons showed low protein similarity between the Colombian MAGs and C. burnetii, with phylogenetic separation of the species. The study could not determine how these differences affect protein activity or the endosymbiont’s contribution to tick nutrition, development, reproduction or pathogen acquisition. It also identified resistance markers at much higher levels than virulence-associated reads, with five markers accounting for approximately 80% to 90% of the resistance-marker signal, but the analysis did not establish that the markers were expressed or conferred resistance in the ticks.

Together, the results provide a metagenomic baseline for the tropical R. sanguineus s.l. complex in two Colombian departments while exposing important interpretive limits. Host DNA dominated each library, approximately 81% of non-host reads remained unclassified, and reference genomes were unavailable or inadequate for some organisms. The sample was limited to 38 adult ticks removed from dogs in two municipalities, and the detected pathogens and endosymbionts were not experimentally tested for replication, transmission or pathogenic effects. The negative correlations involving C. mudrowiae therefore require biological validation rather than being treated as evidence of pathogen suppression. The study’s clearest contribution is the simultaneous documentation of pathogen signals, a diverse microbiome and multiple C. mudrowiae genomes in Colombian tropical-lineage ticks. It also identifies priorities for further work: targeted confirmation of F. tularensis and related Francisella endosymbionts, long-term surveillance of F. tularensis, A. phagocytophilum and T. equi, functional testing of the endosymbiont’s metabolic and immune interactions, and broader sampling of the tick complex within Colombia.

Subject of Research: Shotgun metagenomic profiling of pathogens, viruses and bacterial endosymbionts in Colombian tropical-lineage Rhipicephalus sanguineus s.l. ticks.

Article Title: Metagenomic exploration of endosymbionts and pathogens in the tropical lineage of Rhipicephalus sanguineus sensu lato (s.l.) ticks in Colombia

Article References: Paez-Triana, L., Herrera, G., Vega, L., García-Corredor, D., Pulido Medellín, M. O., Paniz-Mondolfi, A., Muñoz, M., & Ramírez, J. D. (2023). Metagenomic exploration of endosymbionts and pathogens in the tropical lineage of Rhipicephalus sanguineus sensu lato (s.l.) ticks in Colombia. Metabarcoding and Metagenomics, 7, Article e109085. https://doi.org/10.3897/mbmg.7.109085

Image Credits: AI Generated

DOI: 10.3897/mbmg.7.109085

Keywords: shotgun metagenomics, brown dog tick, Rhipicephalus sanguineus, tick microbiome, Coxiella mudrowiae, tick-borne pathogens, Francisella tularensis, Anaplasma phagocytophilum, Theileria equi, bacterial viruses, Colombia, Metagenomic